Kisor and the Piecemeal Dismantling of the Modern Administrative State

The administrative state is under attack. We see this nearly every day in a president who regularly and brazenly slams the administrative state and administrative actors, including his own appointees. We see it in an administration that seems to work overtime to undermine and obstruct the expert work of administrative agencies and their longtime professionals. And we see it from a White House that moves at every chance to consolidate executive power and to politicize the agencies, putting politics above the agencies’ law-enforcement priorities.

In addition to these blatant political attacks, the administrative state is under assault in another, less palpable way: it is under constitutional attack. While perhaps less public than the political attacks, the constitutional assault on the administrative state is far more dangerous. That’s because a constitutional assault could produce lasting, even permanent, changes to the bureaucracy—changes that could bring the administrative state directly and solely under the political wing of the White House, thus threatening agency independence, agency expertise, and, ultimately, the agencies’ apolitical enforcement of the laws.

The constitutional assault on the administrative state comes in four parts. First, opponents attack independent agencies. Drawing on Justice Scalia’s sole dissent in Morrison v. Olson,[1] opponents of the administrative state increasingly argue that any measure of independence within the executive branch impermissibly encroaches upon the president’s unitary executive power in violation of the separation of powers. This argument is gaining traction, and the Court may soon reevaluate its holding in Morrison. If so, this could deal a significant blow to independence within the executive branch.

Second, opponents attack broad congressional delegations of power to administrative agencies under the nondelegation doctrine. The Court’s current approach to the nondelegation doctrine allows nearly any congressional delegation, no matter how broad, to stand up to a separation-of-powers challenge. This approach empowers the administrative state, because it allows agencies to regulate under even the most sweeping congressional delegations. In this way, it also recognizes the comparative advantage that expert agencies have in regulating the details in any given policy area.

Just this Term, the Court rebuffed a challenge to a congressional delegation under the nondelegation doctrine.[2] But three justices—Chief Justice Roberts and Justices Thomas and Gorsuch—argued that the act at issue did delegate too much authority to an agency,[3] and a fourth, Justice Alito, indicated that he would be willing to reevaluate the Court’s hands-off approach to the nondelegation doctrine if a majority on the Court were willing.[4] Justice Kavanaugh could well provide a fifth vote for that majority. (Justice Kavanaugh did not participate in the decision because he was not confirmed in time for oral arguments.) If so, we can expect a revitalized nondelegation doctrine, restricting agencies’ powers by limiting their ability to regulate under broad congressional delegations.

Third, opponents attack agencies’ expansive authority to interpret vague laws and the courts’ deference to the agencies’ interpretations when they do. Under “Chevron deference,” courts defer to an agency’s interpretation of a vague statute.[5] Restricting Chevron deference would therefore deal a sharp blow to agencies’ regulatory powers (and at the same time enhance the powers of the courts). Challenges to Chevron deference are gaining steam, with a majority on the Court now apparently open to reconsidering Chevron.

Finally, opponents attack agencies’ expansive authority to interpret their own vague regulations, and the courts’ deference to the agencies’ interpretations when they do. This is called “Auer deference.”[6] And it came before the Court this Term.

To appreciate the importance of Auer deference, it helps to understand that federal administrative agencies issue thousands of regulations on matters that span a truly mind-boggling range of topics. As a result, agency regulations can be highly specialized, obscure, and even arcane. They can also be vague. When courts are called upon to interpret and apply those regulations, they sometimes understandably have a hard time.

Auer deference can help. Under Auer deference, a court defers to an agency’s interpretation of its own ambiguous regulations unless that interpretation is “plainly erroneous or inconsistent with the regulation.”[7] It’s based on the simple idea that an administrative agency is best suited to interpret its own regulations. That’s so for two reasons. First, the agency wrote its own regulations, and, as author, has the best insight into what those regulations mean. Next, the agency has experience and expertise that others (in particular, judges) don’t have, and that experience and expertise can often inform an interpretation of otherwise ambiguous regulations. As Justice Breyer wryly observed during oral argument in Kisor, the Court in a prior case had “deferred to the understanding of the FDA that a particular compound should be treated as a single new active moiety, which consists of a previously approved moiety, joined by a non-ester covalent bond to a lysine group. Do you know how much I know about that?”[8]

Auer deference is not only an important tool to help the courts interpret and apply ambiguous agency regulations. It’s also a powerful tool for the agencies themselves to define those regulations. In other words, Auer deference empowers the expert agencies in enforcing the law.

But that empowerment was threatened when the Court took up Kisor v. Wilkie.[9] In that case, opponents of Auer deference argued that Auer ought to be overturned. The Court rejected that challenge, and Auer survived—but just barely, and probably not for long. In short, the Court held on to a scaled-back Auer deference under stare decisis. But a majority of justices couldn’t agree on much else. Coming out of Kisor, we now have four justices (Justices Ginsburg, Breyer, Sotomayor, and Kagan) who think that Auer deference is consistent with the Administrative Procedure Act (APA) and the separation of powers, and four (Justices Thomas, Alito, Gorsuch, and Kavanaugh) who think that it violates the APA and the separation of powers. The fifth, Chief Justice Roberts, merely said that the difference between the two camps was not that great, suggesting that there’s little left to Auer deference, that it may as well be overruled, and that the Court will apply it sparingly, if at all, in the future. As Justice Gorsuch put it, “today’s decision is more a stay of execution than a pardon.”[10]

This doesn’t bode well for the future of Auer deference. And, along with the other three lines of attack, it doesn’t bode well for the future of the administrative state.

This article will first trace the history of Auer deference. Next, it will describe some of the criticisms of Auer deference and how we arrived at Kisor’s challenge to it. It will then examine the highly fractured Kisor ruling and try to make sense of the various opinions. Finally, this article will offer some predictions about what Kisor could mean for the future of Auer deference, and for the future of the administrative state.

I. The History of Auer Deference

The idea that the courts should defer to an agency’s interpretation of its own regulations traces back to the late nineteenth century, and maybe even earlier.[11] For example, in a case that many point to as the one setting this principle, the Court in United States v. Eaton appeared to defer to an agency’s interpretation of its own regulation, even if only as an alternative basis for affirming the agency’s interpretation, in a dispute over the emergency appointment of a vice consul of Siam by the consul general.[12] A State Department regulation authorized “the diplomatic representative” to appoint an emergency vice consul when the incumbent could not fulfill the duties of the office,[13] as here. But there was a glitch: the consul general himself had obtained a leave of absence from the president because he was sick and unable to discharge his duties.[14] So the question arose: if the consul general was unable to discharge his duties, did the regulations authorize him to appoint a vice consul (which, of course, was an exercise of his duties)? The State Department itself impliedly answered yes, by recognizing the appointment of the vice consul.

The Court in Eaton agreed. It surveyed the purpose and text of the regulation and concluded that the regulation authorized the consul general to make the emergency appointment.[15] The Court then wrote, “The interpretation given to the regulations by the department charged with their execution . . . is entitled to the greatest weight, and we see no reason in this case to doubt its correctness.”[16]

The Court refined its approach to agency deference in 1944, in Skidmore v. Swift & Co.[17] In that case, seven employees of the Swift & Company packing plant sought to recover overtime pay for staying on premises after hours, in the company’s fire hall or within hailing distance, several nights a week in order to answer fire alarms.[18] The Fair Labor Standards Act (FLSA) required employers to pay their employees overtime pay for overtime work, but it did not say whether on-call duties, like those of the plaintiffs, counted as overtime work. In resolving the question, the Court looked to the agency’s interpretation of the FLSA, in an interpretative bulletin and the agency’s amicus brief in the case.[19] In language worth quoting at length—and notably bereft of citation—the Court wrote that agency interpretations:

are not, of course, conclusive, even in the cases with which they directly deal, much less in those to which they apply only by analogy. They do not constitute an interpretation of the Act or a standard for judging factual situations which binds a district court’s processes, as an authoritative pronouncement of a higher court might do. But the Administrator’s policies are made in pursuance of official duty, based upon more specialized experience and broader investigations and information than is likely to come to a judge in a particular case. They do determine the policy which will guide applications for enforcement by injunction on behalf of the Government. Good administration of the Act and good judicial administration alike require that the standards of public enforcement and those for determining private rights shall be at variance only where justified by very good reasons. The fact that the Administrator’s policies and standards are not reached by trial in adversary form does not mean that they are not entitled to respect. This Court has long given considerable and in some cases decisive weight to Treasury Decisions and to interpretative regulations of the Treasury and of other bodies that were not of adversary origin.

We consider that the rulings, interpretations and opinions of the Administrator under this Act, while not controlling upon the courts by reason of their authority, do constitute a body of experience and informed judgment to which courts and litigants may properly resort for guidance. The weight of such a judgment in a particular case will 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.[20]

Today, we call this sliding-scale deference to an agency’s interpretation “Skidmore deference.”

The Court once again refined its approach to deference the following Term, and this time took it in a new direction, in Bowles v. Seminole Rock & Sand Co.[21] In Seminole Rock, the administrator of the Office of Price Administration brought a claim against Seminole Rock & Sand Company to enjoin it from selling crushed stone in violation of Maximum Price Regulation No. 188.[22] That regulation created a price ceiling for certain building materials and consumer goods as part of a larger regulatory effort to set maximum prices across the nation’s economy in order “to combat wartime inflation.”[23] In particular, the regulation provided that “the maximum price for any article which was delivered or offered for delivery in March, 1942, by the manufacturer, shall be the highest price charged by the manufacturer during March, 1942 (as defined in [Section] 1499.163) for the article.”[24] Section 1499.163(a)(2) in turn defined the “[h]ighest price charged during March, 1942” as “[t]he highest price which the seller charged to a purchaser of the same class for delivery of the article or material during March, 1942 . . . .”[25] The administrator argued that the provision applied to Seminole Rock, because Seminole Rock actually delivered covered articles in March 1942. Seminole Rock, on the other hand, argued that the regulation didn’t apply, because it didn’t charge for those articles in March 1942.

The Court first noted the significance of the agency’s interpretation. In a brief passage, without citation or meaningful analysis, the Court explained:

Since this involves an interpretation of an administrative regulation a court must necessarily look to the administrative construction of the regulation if the meaning of the words used is in doubt. The intention of Congress or the principles of the Constitution in some situations may be relevant in the first instance in choosing between various constructions. But the ultimate criterion is the administrative interpretation, which becomes controlling weight unless it is plainly erroneous or inconsistent with the regulation.[26]

Looking at the language, the Court noted that the regulation itself “recognizes the fact that more than one meaning may be attached to the phrase ‘highest price charged during Mach [sic], 1942’” and that it was reasonably susceptible to the two very different interpretations offered by the parties.[27] But using a textual analysis, the Court went on to hold that the administrator’s interpretation was the right one: “the highest price charged for an article delivered during March, 1942, is the seller’s ceiling price regardless of the time when the sale or charge was made.”[28] The Court then wrote that “[a]ny doubts concerning this interpretation . . . are removed by reference to the administrative construction of this method of computing the ceiling price.”[29] The Court referenced the administrator’s interpretation in a bulletin that was available to manufacturers, wholesalers, and retailers; in a quarterly report to Congress; and through the uniform position of the Office of Price Administration “in the countless explanations and interpretations given to inquirers affected by this type of maximum price determination.”[30] The Court concluded that its “reading of the language of [the regulation] and the consistent administrative interpretation of the phrase ‘highest price charged during March, 1942’ thus compel[led]” its decision in favor of the administrator of the Office of Price Administration.[31] Given the Court’s strong language on deference, its equally important independent textual analysis, and its conclusion that seems to merge the two, it’s not entirely clear how much the Court actually relied on deference to the agency in that case.

Over twenty years later, the Court clarified its position on deference—and underscored the primacy of deference—in Auer v. Robbins.[32] In Auer, sergeants and lieutenants of the St. Louis Police Department sued for overtime pay that they claimed the department owed them under Section 7(a)(1) of the FLSA.[33] That provision requires employers to pay their employees at least time-and-a-half overtime for work beyond a forty-hour workweek, with certain exceptions.[34] U.S. Department of Labor regulations define one such exception—to an “employee [who] earn[s] a specified minimum amount on a ‘salary basis.’”[35] Under the regulation, an employee is paid on a salary basis “if under his employment agreement he regularly receives each pay period on a weekly, or less frequent basis, a predetermined amount constituting all or part of his compensation, which amount is not subject to reduction because of variations in the quality or quantity of the work performed.”[36] The plaintiffs claimed that they were not paid “on a salary basis,” because the department’s police manual authorized the department to take certain disciplinary actions that would reduce an officer’s pay, and because one “sergeant was actually subject[] to [such] a disciplinary deduction.”[37] If the plaintiffs were right—and they were “subject to reduction because of variations in the quality or quantity of the work performed”—they would not have been exempt, and they would have been entitled to overtime pay under the FLSA.

The Court asked the secretary of Labor to weigh in on the question at the petition stage (before the Court agreed to hear the case). The secretary filed an amicus brief that, apparently for the first time ever, offered the secretary’s interpretation of the regulatory language:

Although petitioners’ interpretation of the regulation is not implausible, the Secretary does not interpret the “subject to” language in that way. An employee who is otherwise paid on a salaried basis does not lose that status simply because he is theoretically “subject to” an unpaid suspension of less than one week for violating a rule that is not a major-safety rule. Rather, salaried status is lost under the regulations only when employees are “subject to” such a disciplinary suspension as a practical matter. Thus, where no deductions have actually been made, or where there have been only isolated instances in which disciplinary suspensions have occurred, or where suspensions have occurred only under unusual circumstances, employees who are not themselves suspended do not lose their salaried status.[38]

Importantly, the government provided no citation, no history, and no other basis for the secretary’s reading. Again, the government apparently stated this reading for the first time in its amicus brief at the petition stage in the very case in which it asked the Court to adopt its preferred reading.[39]

Still, the Supreme Court deferred to the secretary’s reading. In a shockingly brief paragraph, without significant analysis, and quoting Seminole Rock, Justice Scalia, writing for a unanimous Court, explained why: “Because the salary-basis test is a creature of the Secretary’s own regulations, his interpretation of it is, under our jurisprudence, controlling unless ‘plainly erroneous or inconsistent with the regulation.’ That deferential standard is easily met here.”[40] Underscoring the extreme deference that the Court gave to the secretary’s interpretation, the Court was utterly unconcerned with the fact that the secretary only offered his interpretation in a legal brief for the Court in this very case. Justice Scalia explained, again without significant analysis, in another astonishingly brief and telling passage, that the secretary’s interpretation of the regulation “is in no sense a ‘post hoc rationalizatio[n]’ advanced by an agency seeking to defend past agency action against attack . . . . There is simply no reason to suspect that the interpretation does not reflect the agency’s fair and considered judgment on the matter in question.”[41] In response to the petitioner’s assertion that the FLSA’s exemptions must be “narrowly construed against . . . employers,” Justice Scalia concluded, “A rule requiring the Secretary to construe his own regulations narrowly would make little sense, since he is free to write the regulations as broadly as he wishes, subject only to the limitations imposed by the statute.”[42]

Today, we call this rigid and fixed-level deference to an agency’s interpretation “Seminole Rock/Auer deference,” or merely “Auer deference.” And we distinguish it from the sliding-scale (and lower level) Skidmore deference discussed above.

II. Issues with Auer Deference

In the years following Auer, the Court refined and even scaled back its approach to deference. But its rulings left some confusion as to the precise contours of Auer deference and thus opened Auer deference up to attack. Four lines of criticism emerged.[43] First, critics argued that the scope of Auer deference was uncertain and indeterminate.[44] They claimed that the Court itself, let alone lower courts,[45] could not apply it in a consistent, sensible way. Justice Thomas summarized this critique:

[Auer deference] has even been applied to an agency’s interpretation of another agency’s regulations. And, it has been applied to an agency interpretation that was inconsistent with a previous interpretation of the same regulation. It has been applied to formal and informal interpretations alike, including those taken during litigation. Its reasoning has also been extended outside the context of traditional agency regulations into the realm of criminal sentencing.

The Court has even applied the doctrine to an agency interpretation of a regulation cast in such vague aspirational terms as to have no substantive content.[46]

Next, critics argued that Auer deference encouraged agencies to write vague regulations, so that they could write new rules under the guise of interpretation, and thus evade judicial review. According to critics, this gave agencies expansive and unchecked power to effectively issue new regulations without subjecting them to judicial scrutiny, or any other democratic scrutiny for that matter.[47] Justice Scalia summarized this critique:

But when an agency interprets its own rules—that is something else. Then the power to prescribe is augmented by the power to interpret; and the incentive is to speak vaguely and broadly, so as to retain a “flexibility” that will enable “clarification” with retroactive effect. “It is perfectly understandable” for an agency to “issue vague regulations” if doing so will “maximiz[e] agency power.” . . . Auer deference encourages agencies to be “vague in framing regulations, with the plan of issuing ‘interpretations’ to create the intended new law without observance of notice and comment procedures.”[48]

Third, critics argued that Auer violated the APA.[49] They claimed that judicial deference to an agency’s interpretation of its own regulation violated the judicial-review provision in the APA,[50] and that deference violated the APA’s requirement for notice-and-comment rulemaking.[51] As to judicial review, Justice Scalia argued that the Court had ignored the APA’s “directive that the ‘reviewing court . . . interpret . . . statutory provisions [and] determine the meaning or applicability of the terms of an agency action,’” and had instead held that agencies may resolve ambiguities in statutes and regulations.[52]

As to notice-and-comment rulemaking, Justice Scalia argued,

By supplementing the APA with judge-made doctrines of deference, we have revolutionized the import of interpretive rules’ exemption from notice-and-comment rulemaking. Agencies may now use these rules not just to advise the public, but also to bind them. After all, if an interpretive rule gets deference, the people are bound to obey it on pain of sanction, no less surely than they are bound to obey substantive rules, which are accorded similar deference. Interpretive rules that command deference do have the force of law.[53]

Finally, critics argued that Auer deference violated the separation of powers in (at least) three ways. First, critics argued that Auer deference impermissibly put both the lawmaking and the law-executing power in the same hands. As Justice Scalia explained, “when an agency promulgates an imprecise rule, it leaves to itself the implementation of that rule, and thus the initial determination of the rule’s meaning.” He continued that “[i]t seems contrary to fundamental principles of separation of powers to permit the person who promulgates a law to interpret it as well,” and that to do so threatens liberty, “because apprehensions may arise, lest the same monarch or senate should enact tyrannical laws, to execute them in a tyrannical manner.”[54]

Second, critics argued that Auer deference impermissibly restricted the courts’ power to interpret and apply the law. Justice Thomas critiqued Auer deference as “a transfer of the judge’s exercise of interpretative judgment to the agency,” and noted that an administrative agency “lacks the structural protections for independent judgment adopted by the Framers, including the life tenure and salary protections of Article III.”[55] Justice Thomas concluded that because “the agency is thus not properly constituted to exercise the judicial power under the Constitution, the transfer of interpretive judgment raises serious separation-of-powers concerns.”[56]

Finally, critics argued that Auer deference “undermines the judicial ‘check’ on the political branches.”[57] As Justice Thomas asserted, the Court has “repeatedly declined to exercise independent judgment” in the face of challenges to agency interpretations, “[i]nstead, [deferring] to the executive agency that both promulgated the regulations and enforced them.”[58] Justice Thomas warned, “When courts refuse even to decide what the best interpretation is under the law, they abandon the judicial check. That abandonment permits precisely the accumulation of governmental powers that the Framers warned against.”[59]

All this meant that the Court became more skeptical of Auer deference, and even declined to apply it in certain cases. Christopher v. SmithKline Beecham Corporation, a 2012 case, illustrates this. In Christopher, the Court declined to apply Auer deference to an agency’s interpretation of its own regulations that the agency only announced in litigation—and then offered shifting reasons.[60] The case arose when pharmaceutical sales representatives sued their employer for overtime pay under the FLSA and Department of Labor regulations. The FLSA generally requires employers to pay time-and-a-half to their employees for any work over forty hours a week. But it exempts workers “employed . . . in the capacity of outside salesman.”[61] So the question for the Court was whether the plaintiffs, whose job entailed calling on physicians and obtaining nonbinding commitments to prescribe the corporation’s drugs in appropriate cases, “sold” a product.

The Department of Labor issued regulations and guidance that touched upon the issue, but did not firmly settle it, at least not in this particular context.[62] When other similar cases started percolating in the lower courts, the department, starting in 2009, filed amicus briefs arguing that representatives are not exempt. But its reasons changed. Initially, the department argued that “a ‘sale’ for the purposes of the outside sales exemption requires a consummated transaction directly involving the employee for whom the exemption is sought.”[63] After the Court granted certiorari, however, the department refined the “consummated transaction” requirement and argued that “[a]n employee does not make a ‘sale’ for purposes of the ‘outside salesman’ exemption unless he actually transfers title to the property at issue.”[64] Still, the petitioners and the Department argued that its new interpretation was entitled to controlling deference.

The Court disagreed. The Court first noted that Auer deference is inappropriate in cases when an agency’s interpretation is “plainly erroneous or inconsistent with the regulation,”[65] when it “does not reflect the agency’s fair and considered judgment on the matter in question,”[66] or when it is a “’post hoc rationalization[n]’ advanced by an agency seeking to defend past agency action”[67] or merely a “convenient litigating position.”[68] It then ruled that Auer deference was inappropriate in this case for a related reason: The department’s interpretation failed to give “fair warning of the conduct [a regulation] prohibits or requires.”[69] The Court warned that Auer deference “creates a risk that agencies will promulgate vague and open-ended regulations that they can later interpret as they see fit, thereby ‘frustrat[ing] the notice and predictability purposes of rulemaking,”[70] and suggested that the department’s late-breaking interpretation in this case actualized that risk. It concluded that while Auer deference was inappropriate, “[w]e instead accord the department’s interpretation a measure of deference proportional to 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’”[71]—in other words, Skidmore deference.

But while Auer deference came under increasing fire, and while the Court continued to refine it and scale back in reaction, objections to the doctrine itself never gained the attention of a majority on the Court. Only Justices Scalia and Thomas were openly hostile to it. Chief Justice Roberts and Justice Alito wanted to wait for an appropriate case in which to review it.[72] That case came in Kisor.

III. Kisor’s Challenge to Auer Deference

The case arose when James Kisor first applied for Veterans Affairs (VA) disability benefits in 1982.[73] Kisor claimed that he suffered from post-traumatic stress disorder (PTSD) based on his service as a U.S. Marine from 1962 to 1966, including service in the Vietnam War, where he experienced intense and deeply traumatic combat.[74] In support of his claim, Kisor offered a letter by a counselor at the Portland Veterans Center identifying “concerns that Mr. Kisor had towards depression, suicidal thoughts, and social withdraw[a]l.”[75] The counselor wrote that “[t]his symptomatic pattern has been associated with the diagnosis of Post-Traumatic Stress Disorder.”[76] Nevertheless, the VA Regional Office in Portland denied Kisor’s claim, concluding that “posttraumatic stress neurosis, claimed by vet” was “not shown by evidence of record.”[77] (In evaluating Kisor’s claim, the VA failed to request his service personnel records, as required by VA regulations.[78] These records could have connected any disability to Kisor’s service. But the decision denying benefits did not even address service-connection; instead, the VA simply concluded that Kisor did not have PTSD at all.)

On June 5, 2006, Kisor moved to reopen his claim. In support, he submitted new evidence, including an extensive report by a psychiatrist who concluded that it was “clear that the claimant was evincing symptoms of P.T.S.D. back in the 1980’s.”[79] In addition, the VA requested and obtained Kisor’s service personnel records in order to determine whether any disability was service-connected. Based on the new evidence, the VA reversed course and determined that Kisor suffered from PTSD that his PTSD was service-connected. It assigned a disability rating of fifty percent and awarded him benefits, including retroactive benefits as of June 5, 2006.[80]

Kisor appealed the decision, arguing that he was entitled to a higher disability rating and retroactive benefits as of the date of his original claim. In March 2009, a decision review officer at the VA regional office concluded that Kisor was entitled to a disability rating of seventy percent, but that he was not entitled to retroactive benefits back to the date of his original claim. As to retroactivity, the VA reasoned that “[a]t the time of the [original] decision the veteran did not have a clinical diagnosis of post traumatic stress disorder.”[81]

The Board of Veterans Appeals affirmed and similarly denied retroactive benefits. The Board pointed to the VA regulation on new evidence, which states,

[A]t any time after VA issues a decision on a claim, if VA receives or associates with the claims file relevant official service department records that existed and had not been associated with the claims file when VA first decided the claim, VA will reconsider the claim . . . . Such records include, but are not limited to: (i) Service records that are related to a claimed in-service event, injury, or disease . . . .[82]

The Board read “relevant” to mean only those records that “would suggest or better yet establish” the component of the claim that the VA found missing in the prior judgment.[83] The Board said that the component of the claim found missing in the 1983 denial was Kisor’s PTSD, not its connection to his service. But the Board ruled that Kisor’s new evidence (his service personnel records) went to service-connection, not to his PTSD. As a result, the Board concluded that Kisor’s new evidence would not “suggest or better yet establish” the component of the claim found missing in the earlier judgment (PTSD). The Board explained that the new evidence was thus not “outcome determinative” and “not relevant to the decision in May 1983 because the basis of the denial was that a diagnosis of PTSD was not warranted, not a dispute as to whether or not the Veteran engaged in combat with the enemy during service.” The Board thus rejected Kisor’s claim for retroactive benefits. The Veterans Court denied his claims, too, for substantially similar reasons.[84]

Kisor appealed to the U.S. Court of Appeals for the Federal Circuit. He argued that the Board and the Veterans Court “mistakenly interpreted the term ‘relevant’ . . . as related only to service department records that countered the basis of the prior denial.”[85] Kisor claimed that the term applied more broadly, to any record that has “any tendency to make the existence of any fact that is of consequence to the determination of the action more probable or less probable than it would be without the evidence.”[86] Under this interpretation, Kisor said that his new evidence (the service personnel records) was relevant, because the records “demonstrate that he was subjected to the trauma of combat, thereby establishing his exposure to an in-service stressor.”[87]

The Federal Circuit rejected that argument. The court first noted that it would apply Auer deference to the VA’s interpretation of its own regulation: “As a general rule, we defer to an agency’s interpretation of its own regulation ‘as long as the regulation is ambiguous and the agency’s interpretation is neither plainly erroneous nor inconsistent with the regulation.’”[88] The court ruled that the term “relevant” was, indeed, ambiguous, and that the VA’s interpretation satisfied this test. In particular, the court said that “[t]he Board’s ruling was thus based upon the proposition that . . . ‘relevant’ means noncumulative and pertinent to the matter at issue in the case”—an interpretation that was neither “plainly erroneous [n]or inconsistent with the VA’s regulatory framework.”[89] The court affirmed the Veterans Court and rejected Kisor’s claim for retroactive benefits.

Kisor appealed to the Supreme Court, arguing that the Court should overturn Auer and Seminole Rock and reverse the Federal Circuit’s deference to the VA’s reading of its own regulation.

Kisor argued first that Auer deference is inconsistent with the APA.[90] Kisor claimed that Auer deference violates the APA’s judicial-review provision, because it requires courts to defer to an agency’s interpretation of the agency’s own regulations—and not interpret those regulations for itself.[91] Kisor contended that Auer deference also violates the APA’s requirement for notice-and-comment rulemaking, because it allows an agency effectively to make new rules (by interpreting existing ones) without notifying the public and considering public comments.[92] Kisor said that Auer deference is in tension with the APA’s treatment of agency interpretive rules, which the APA exempts from notice-and-comment requirements but which, as a result, get no special deference in the courts.[93]

Kisor argued next that Auer deference injects unpredictability into agency action, in tension with the principles behind the APA. In particular, he claimed that Auer deference “invites vague regulations, which limit the public’s ability to conform conduct to law.”[94] Quoting Justice Scalia, he explained,

Because the agency (not Congress) drafts the substantive rules that are the object of those interpretations, giving them deference allows the agency to control the extent of its notice-and-comment-free domain. To expand this domain, the agency need only write substantive rules more broadly and vaguely, leaving plenty of gaps to be filled in later, using interpretive rules unchecked by notice and comment. The APA does not remotely contemplate this regime.[95]

Kisor contended that this problem was especially acute when a new administration seeks to enforce its own new priorities,[96] and when an agency is self-interested.[97] He claimed that the policy justifications for Auer deference—that the agency has special insight into its own rules and a special expertise in its own regulatory programs—are not strong enough to overcome these problems.[98] He said that if those policy justifications carry any weight, they support only a lesser Skidmore deference.[99]

For largely the same reasons, Kisor argued that Auer deference violates the separation of powers. Here, he added only that Auer deference impermissibly mixes the law-making authority with the law-executing authority, making it “incompatible with the separation-of-powers principles that animate the Constitution. To conclude otherwise ‘would violate a fundamental principle of separation of powers—that the power to write a law and the power to interpret it cannot rest in the same hands.’”[100]

Finally, Kisor argued that stare decisis cannot save Auer. He claimed that Seminole Rock was poorly reasoned (or relied on no reasoning at all), and that Auer has produced all the problems described above.[101] He contended that Auer, as a judicially (not legislatively) created doctrine, has less staying power under stare decisis, and that, as a “deference regime,” may in fact have no staying power under stare decisis.[102] And he asserted that private parties’ reliance interests on Auer are low, “because it does not authorize any particular result with respect to any particular rule. Indeed, one of Auer’s principal effects is to promote legal instability. At its core, Auer deference gives the force of law [even] to an agency’s interpretation of a regulation adopted after a dispute begins.”[103]

The Court declined to outright overturn Seminole Rock and Auer. In a sharply divided and highly fractured ruling, the Court held by a five-to-four margin that Seminole Rock and Auer survived under principles of stare decisis. Justice Kagan penned the majority opinion, joined by Chief Justice Roberts and Justices Ginsburg, Breyer, and Sotomayor. Justice Gorsuch wrote the principal minority opinion, arguing that the Court should overrule Seminole Rock and Auer. But at the same time, the Court (by the same five-to-four margin) sharply curtailed Auer deference in the interest of giving lower courts more determinate guidance about how to apply it. The Court then divided four-to-four over interpretations of Auer’s history, and whether Auer deference violates the APA and the separation of powers. (Chief Justice Roberts did not join either faction on these questions.) The only thing that all the justices agreed on was this: The Federal Circuit mis-applied Auer deference to Kisor’s case and should take another crack at it.

Justice Kagan’s majority opinion first explained some of the problems with the Court’s prior treatment of Auer deference, and how to clean them up.[104] Specifically, she wrote that the Court has sent “mixed messages” with regard to Auer deference, sometimes applying it “without significant analysis” or “careful attention to the nature and context of the interpretation.”[105] In order to address these problems and to provide guidance to lower courts for future applications of Auer deference, her opinion sought to “reinforce[] some of the limits inherent in the Auer doctrine.”[106]

As an initial matter, Justice Kagan wrote that “a court should not afford Auer deference unless the regulation is genuinely ambiguous.”[107] In making this determination, the Court held that courts must use all their tools of construction, including “the text, structure, history, and purpose of a regulation, in all the ways it would if it had no agency to fall back on.”[108] “Doing so will resolve many seeming ambiguities out of the box, without resort to Auer deference.”[109]

Next, if a genuine ambiguity remains, the Court held that the courts must determine if the agency interpretation of its regulation is “reasonable.”[110] This means that an interpretation “must come within the zone of ambiguity the court has identified after employing all its interpretive tools.”[111] The Court noted that at this stage some courts have granted greater deference to agency interpretations to their own rules than to agency interpretations of statutes.[112] But the Court said that this was wrong: The standard for both kinds of deference is whether an agency’s interpretation falls “within the bounds of reasonable interpretation.”[113] And the Court was clear that this “is a requirement an agency can fail.”[114]

Finally, the Court held that, even if an agency interpretation clears these first two hurdles, Auer deference is only appropriate in certain circumstances. Rather than specifying a test, the Court “laid out some especially important markers”[115] that derive from and support the purposes of Auer deference in the first place.

For one, the agency interpretation “must be the agency’s ‘authoritative’ or ‘official position,’ rather than any more ad hoc statement not reflecting the agency’s views.”[116] That does not mean that an agency must have subjected its interpretation to formal notice-and-comment rulemaking. But it does mean that “[t]he interpretation must at the least emanate from those actors, using those vehicles, understood to make authoritative policy in the relevant context.”[117]

For another, “the agency’s interpretation must in some way implicate its substantive expertise.”[118] The Court held that this could be quite broad, including all those policy areas that Congress has delegated to an agency. But at the same time, it has limits. Thus, “deference ebbs when ‘[t]he subject matter of the [dispute is] distan[t] from the agency’s ordinary’ duties or ‘fall[s] within the scope of another agency’s authority.’”[119] Moreover, the Court noted that “[s]ome interpretive issues may fall more naturally into a judge’s bailiwick.”[120] In such a case, a court should look to comparative institutional competence: “When the agency has no comparative expertise in resolving a regulatory ambiguity, Congress presumably would not grant it that authority.”[121]

For a third, “an agency’s reading of a rule must reflect ‘fair and considered judgment’ to receive Auer deference.”[122] The Court held that this rules out deference for an agency’s “convenient litigating position” and “post hoc rationalizatio[n].”[123] It also rules out deference for a “new interpretation . . . that creates an ‘unfair surprise’ to regulated parties.”[124]

The Court summarized its holding: “When it applies, Auer deference gives an agency significant leeway to say what its own rules mean. In so doing, the doctrine enables the agency to fill out the regulatory scheme Congress has placed under its supervision. But that phrase ‘when it applies’ is important—because it often doesn’t.”[125] The decision “maintained a strong judicial role in interpreting rules. What emerges is a deference doctrine not quite so tame as some might hope, but not nearly so menacing as they might fear.”[126]

Having so clarified Auer deference, the Court went on to hold that stare decisis “cuts strongly against” overruling Auer.[127] The Court held that a long line of Court cases reaffirms Auer; that lower courts have applied Auer or Seminole Rock thousands of times; and that overruling Auer would therefore “cast doubt on many settled constructions of rules,” and would even “allow relitigation of any decision based on Auer, forcing courts to ‘wrestle [with] whether or not Auer’ had actually made a difference.”[128] “It is the rare overruling that introduces so much instability into so many areas of law, all in one blow.”[129] Moreover, the Court held that whatever the Court has said about Auer deference, because Auer is not a “constitutional case” Congress could change it.[130] The Court noted that Congress “could amend the APA or any specific statute to require the sort of de novo review of regulatory interpretations that Kisor favors.”[131] Congress, however, has allowed the Court’s “deference regime [to] work side-by-side with both the APA and the many statutes delegating rulemaking power to agencies. It has done so even after we made clear that our deference decisions reflect a presumption about congressional intent.”[132] The Court held that given this history, “we would need a particularly ‘special justification’ to now reverse Auer”—a justification that Kisor simply could not provide.[133]

Finally, in applying Auer to Kisor’s case, the Court vacated the Federal Circuit’s judgment and remanded the case. (The Court was unanimous in its final judgment, although Justices Thomas, Alito, Gorsuch, and Kavanaugh did not specifically join this portion of the Court’s ruling.) The Court held that “the Federal Circuit jumped the gun in declaring the regulation ambiguous,” because it failed to grapple with the “text, structure, history, and purpose” of the rule.[134] And it held that the lower court too quickly assumed that Auer deference should apply even if the regulation were ambiguous. In particular, the Court held that the Federal Circuit failed to consider that a single Board member’s reading of the regulation “reflects the considered judgment of the agency as a whole.”[135] The Court vacated the judgment of the Federal Circuit and remanded the case for reconsideration of whether and how Auer deference applies.[136]

From there, the Court split even more sharply—four-to-four, without a majority holding, and with Chief Justice Roberts declining to join either faction—over the history and purposes of Auer deference, and over whether Auer deference violates the APA and the separation of powers.

The disagreement about Auer’s history and purposes is significant, because the doctrine’s evolution helps explain whether it violates the APA and the separation of powers, and even its durability under principles of stare decisis. Justice Kagan started her assessment of history and purpose by noting that judicial deference to an agency’s interpretation of its own regulations is a practical and sensible solution to the problem that “[f]or various reasons, regulations may be genuinely ambiguous.”[137] She argued that deference to an agency’s interpretation of its own rules traces back to 1898, when the Court in United States v. Eaton wrote that “[t]he interpretation given to the regulations by the department charged with their execution . . . is entitled to the greatest weight.”[138]

Justice Kagan then argued that “Auer deference (as we now call it) as rooted in a presumption about congressional intent—a presumption that Congress would generally want the agency to play the primary role in resolving regulatory ambiguities.”[139] In other words, Auer deference is a result of congressional intent (and not merely judicial say so): “we have thought that when granting rulemaking power to agencies, Congress usually intends to give them, too, considerable latitude to interpret the ambiguous rules they issue.”[140] In a passage worth quoting at length, she went on to explain the familiar reasons for Auer deference, and the related reasons why the Court has presumed that Congress intends it:

In part, that is because the agency that promulgated a rule is in the “better position [to] reconstruct” its original meaning. Consider that if you don’t know what some text (say, a memo or an e-mail) means, you would probably want to ask the person who wrote it. And for the same reasons, we have thought, Congress would too (though the person is here a collective actor). The agency that “wrote the regulation” will often have direct insight into what that rule was intended to mean. . . .

In still greater measure, the presumption that Congress intended Auer deference stems from the awareness that resolving genuine regulatory ambiguities often “entail[s] the exercise of judgment grounded in policy concerns.” . . .

And Congress, we have thought, knows just that: It is attuned to the comparative advantages of agencies over courts in making such policy judgments. Agencies (unlike courts) have “unique expertise,” . . . can conduct factual investigations, can consult with affected parties, can consider how their experts have handled similar issues over the long course of administering a regulatory program. And agencies (again unlike courts) have political accountability, because they are subject to the supervision of the President, who in turn answers to the public. . . .

Finally, the presumption we use reflects the well-known benefits of uniformity in interpreting genuinely ambiguous rules.[141]

In sum, according to Justice Kagan, there are good and long-standing reasons for the courts to defer to an agency’s interpretation of its own regulations, that is, to apply Auer deference. These include the comparative institutional competence between agencies and the courts, and the need for consistency in regulatory interpretation. Congress has recognized this, and has given its consent to Auer deference.

Justice Gorsuch expressed a very different view, in an opinion joined by Justices Thomas, Alito, and Kavanaugh (but not Chief Justice Roberts). He argued that before 1945 the Court treated an agency’s interpretation of its own regulations merely as evidence of the law (and not the law itself), using a sliding scale of deference.[142] He wrote that this approach is reflected in Skidmore, where the Court held that “an agency’s interpretation of the law is ‘not controlling upon the courts’ and is entitled only to a weight proportional to ‘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.”[143] He argued that deference to an agency’s interpretation only arose in 1945, in Seminole Rock, where “the Court declared—for the first time and without citing any authority—that ‘if the meaning of [the regulation were] in doubt,’ the agency’s interpretation would merit ‘controlling weight unless it is plainly erroneous or inconsistent with the regulation.”[144] Justice Gorsuch argued that Auer deference, as we know it today, only arose when, “[f]rom the 1960s on, this Court and lower courts began to cite the Seminole Rock dictum with increasing frequency and in a wider variety of circumstances, but still without much explanation. They also increasingly divorced Seminole Rock from Skidmore.”[145] In short, according to Justice Gorsuch, Auer deference came about only because the courts misread earlier precedent and let their misreading snowball.

Justice Gorsuch argued that because of this slipshod and careless evolution, modern courts have “‘mechanically applied and reflexively treated’ Seminole Rock’s dictum ‘as a constraint upon the careful inquiry that one might ordinarily expect of courts engaged in textual analysis.’”[146] Moreover, he argued that the doctrine is hopelessly indeterminate, and that lower courts don’t know how to apply it.[147]

As to Auer’s consistency with the APA, Justice Kagan argued that the now-clarified Auer deference is perfectly consistent, and she flatly rejected Kisor’s arguments to the contrary. She argued first that Auer deference (again, as clarified) allows for plenty of room for independent judicial review, consistent with the APA’s judicial review provision. She noted that in determining whether to apply Auer deference, courts must initially “apply all traditional methods of interpretation to any rule, and must enforce the plain meaning those methods uncover.”[148] And even then, courts must still “determine whether the nature or context of the agency’s construction reverses the usual presumption of deference . . . and whether the interpretation is authoritative, expertise-based, considered, and fair to regulated parties.”[149] Justice Kagan argued that all of this constitutes “meaningful judicial review.”[150] She also noted that the APA’s judicial-review provision does not specify a particular standard of review, and she argued therefore that the Court’s presumption that Congress intended to delegate to agencies “considerable latitude to construe its ambiguous rules” does not run afoul of it.[151]

Justice Kagan also argued that Auer deference does not circumvent the APA’s notice-and-comment rulemaking requirements. Justice Kagan wrote that an agency’s “interpretive rules, even when given Auer deference, do not [impermissibly] have the force of law,”[152] and that for all the reasons described above, Auer deference still gives the courts (and not the agency) the final authority to determine the legality of an agency’s reading of its own rules. Moreover, Justice Kagan noted that under its clarifications to the Auer doctrine, “an agency has a strong incentive to circulate its interpretations early and widely.”[153] According to Justice Kagan, that means that “the doctrine of Auer deference reinforces, rather than undermines, the idea of fairness and informed decisionmaking at the core of the APA.”[154] Finally, she flatly rejected Kisor’s argument, echoing Justice Scalia, that Auer deference encourages agencies to issue vague regulations.[155]

Again, Justice Gorsuch took a very different view. He argued that Auer deference flies in the face of the APA’s judicial-review provision. He noted that this provision requires reviewing courts to “decide all relevant questions of law” and to “set aside agency action . . . found to be . . . not in accordance with law,”[156] and that this review should be de novo.[157] Yet according to Justice Gorsuch, Auer deference means that the courts do not independently “decide all questions of law”; instead, the courts cede that responsibility to the agencies:

A court that, in deference to an agency, adopts something other than the best reading of a regulation isn’t ‘decid[ing]’ the relevant ‘questio[n] of law’ or ‘determin[ing] the meaning’ of the regulation. Instead, it’s allowing the agency to dictate the answer to that question. In doing so, the court is abdicating the duty Congress assigned to it in the APA.[158]

Moreover, Justice Gorsuch argued that Auer deference violates the APA’s notice-and-comment requirement. He claimed that Auer deference erases the distinction between a formal agency regulation (which requires notice-and-comment rulemaking and carries the force of law) and an agency interpretation of an existing regulation (which does not require notice and comment and does not carry the force of law). That’s because “[u]nder Auer, courts must treat as ‘controlling’ not only an agency’s duly promulgated rules but also its mere interpretations—even ones that appear only in a legal brief, press release, or guidance document issued without affording the public advance notice or a chance to comment.”[159] Thus, Auer deference circumvents the APA’s notice-and-comment requirement.

Justice Gorsuch also disputed Justice Kagan’s argument that Auer deference conforms to a presumption about congressional intent. He claimed that Congress never expressed this intent, and the APA suggests the opposite.[160] Moreover, he contended that the APA, which came just one year after Seminole Rock, was designed to expand judicial review of agency action, not restrict it (as Auer deference does).[161] And as described above, Auer deference was not a part of the firmly established background common law when Seminole Rock came down—indeed, judicial deference to an agency’s interpretation of its own regulations “was in a confused state”—so the APA cannot represent a codification of that (not-yet-existing) common law.[162]

Finally, Justice Kagan argued that Auer deference does not violate the separation of powers. In a brief analysis, drawing in part on her arguments why Auer deference doesn’t violate the APA, Justice Kagan wrote that courts still “retain a firm grip on the interpretive function” by deciding whether and how Auer deference applies, so that agencies do not usurp the role of the judiciary.[163] In other words, Auer deference is simply a rule of interpretation, not a requirement that courts mechanically adopt an agency’s interpretation wholesale. Moreover, she noted that the Court has long rejected claims that modest comingling of legislative and executive functions—including the kind of comingling that may arise under Auer deference, where executive agencies play a modest role in lawmaking by interpreting their own regulations—do not violate the separation of powers.[164]

Again, Justice Gorsuch took a dramatically different view. He argued that Auer deference “sits uneasily with the Constitution,” because it outsources to executive agencies the judicial function of interpreting the law in violation of the separation of powers.[165] In short, “Auer tells the judge that he must interpret these binding laws to mean not what he thinks they mean, but what an executive agency says they mean.”[166] Worse, Justice Gorsuch added that agencies (unlike judges) are often self-interested, and therefore likely to interpret their regulations according to their interests (and not according to the best reading of the law).[167] He argued that under longstanding principles, Congress cannot direct the courts to interpret the law in a particular way, and, by analogy, neither can the executive branch.[168] In sum,

When we defer to an agency interpretation that differs from what we believe to be the best interpretation of the law, we compromise our judicial independence and deny the people who come before us the impartial judgment that the Constitution guarantees them. And we mislead those whom we serve by placing a judicial imprimatur on what is, in fact, no more than an exercise of raw political executive power.[169]

Justice Gorsuch went on to contest Justice Kagan’s policy arguments in support of Auer deference.[170] Justice Gorsuch argued that Justice Kagan was wrong to say that regulatory interpretation is a matter of what the agency intended. Instead, he claimed that courts should interpret what the regulatory text means.[171] And he contended that Justice Kagan was also mistaken when she argued that agencies’ expertise supports Auer deference, because “even agency experts ‘can be wrong . . .’”[172] Justice Gorsuch argued that agencies are therefore only entitled to the lesser Skidmore deference.[173] Moreover, Justice Gorsuch argued that Justice Kagan was wrong to say that consistency in regulatory interpretation justifies Auer deference; instead, he claimed that the courts, without applying Auer deference, could promote a durable and consistent interpretation of an agency’s regulations just as well as, or better than, the agency.[174]

Finally, Justice Gorsuch argued (with only Justices Thomas and Kavanaugh joining) that stare decisis did not justify upholding Auer. As an initial matter, he claimed that stare decisis might not apply at all, because Auer does not merely settle a particular case; it “prescribe[s] an interpretive methodology governing every future dispute over the meaning of every regulation.”[175] He then claimed that Auer has “no persuasive rationale,” fails as a “workable standard,” and is “out of step with how courts normally interpret written laws.”[176] Furthermore, “the explosive growth of the administrative state over the last half-century has exacerbated [its] potential for mischief.”[177] Finally, he asserted that Auer “has generated no serious reliance interests.”[178]

Justice Gorsuch concluded by claiming that under the majority’s limits, “courts may find that [Auer deference] does not constrain their independent judgment any more than Skidmore.” And then in an ominous warning about Auer’s future he wrote:

But whatever happens, this case hardly promises to be this Court’s last word on Auer. If today’s opinion ends up reducing Auer to the role of a tin god—officious, but ultimately powerless—then a future Court should candidly admit as much and stop requiring litigants and lower courts to pay token homage to it. Alternatively, if Auer proves more resilient, this Court should reassert its responsibility to say what the law is and afford the people the neutral forum for their disputes that they expect and deserve.[179]

In separate concurrences, Chief Justice Roberts and Justice Kavanaugh (joined by Justice Alito) echoed Justice Gorsuch’s premonition. All three agreed that “the distance between the majority and Justice Gorsuch is not as great as it may initially appear,”[180] suggesting that the majority’s restrictions on Auer deference may all but neuter it, and even that the Court may soon outright overrule Auer. And all three agreed that the majority’s holding says nothing about the “distinct” question of judicial deference to an agency’s interpretation of statutes enacted by Congress,[181] suggesting that Chevron deference may be next on the chopping block.

IV. Auer’s Future and the Future of the Administrative State

Although the Court in Kisor declined to overturn Auer, the future of Auer deference is nevertheless bleak, for two reasons. First, the Court so sharply curtailed Auer deference that it is not obvious how and when it will apply, if ever. As to the Supreme Court, Kisor’s highly fractured ruling gives credence to Justice Gorsuch’s prediction that “today’s decision is more a stay of execution than a pardon.”[182] Remember that three justices—Chief Justice Roberts and Justices Kavanaugh and Alito—agreed that “the distance between the majority and Justice Gorsuch is not as great as it may initially appear.”[183] Remember, too, that Justices Alito and Kavanaugh joined most of Justice Gorsuch’s opinion. This all suggests that a five-justice majority (the conservatives) will continue to narrow Auer deference, until the doctrine as a practical matter disappears. In other words, future disputes over Auer’s application will simply be proxy disputes over its validity, with a five-justice majority in effect ruling that it is invalid. Whether the Court reevaluates Auer yet again in future Terms, or not, the doctrine appears to be on the way out.

In the meantime, as to the lower courts, Kisor’s limits on Auer’s application serve not only as guideposts, but also as warnings: Apply Auer deference rarely, and only in a narrow band of cases. Even in that band, lower courts may shy away from Auer deference out of an abundance of caution, reading the handwriting on the wall. In short, Auer deference will likely all but disappear in the lower courts.

Second, given the line-up and the various opinions, it’s possible that the Court will reevaluate Auer in coming Terms, and outright overrule it. Recall that the Court split four-four on the substantive questions (whether Auer deference violates the APA or the separation of powers), with only Chief Justice Roberts declining to opine on those questions. (He only agreed to uphold Auer under principles of stare decisis.[184]) If, as predicted, Kisor’s limits on Auer deference mean that the doctrine in effect goes away, Chief Justice Roberts may decide in a future case that stare decisis alone can no longer support the case. If so, then five justices would hold the view that Auer deference violates the APA and the separation of powers, and the Court would overrule Auer. In this way, Kisor is less a familiar “landmine” for a future Court to overrule Auer, and more a spot of cancer that will continue to grow until it destroys Auer itself.

If the future of Auer deference looks bleak, so too does the future of the administrative state, or at least one other critical aspect of it—Chevron deference. That’s because the opinions in Kisor suggest that Chevron deference could be next on the Court’s chopping block. Many of the reasons why four justices argued that Auer deference violates the APA and the separation of powers also apply to Chevron deference. And Chief Justice Roberts had only this to say about it: “Issues surrounding judicial deference to agency interpretations of their own regulations are distinct from those raised in connection with judicial deference to agency interpretations of statutes enacted by Congress.” A five-justice majority now seems primed to reconsider Chevron.

***

Critics of the modern administrative state have moved to dismantle key aspects of it by fashioning constitutional arguments to challenge four of its key features: executive agency independence; executive agency authority to regulate under broad congressional delegations; judicial deference to executive agencies’ interpretations of ambiguous statutes; and judicial deference to executive agencies’ interpretations of their own regulations. Kisor goes to the fourth feature. And while the case didn’t accept the constitutional challenge, it didn’t fully reject it, either. Instead, the Court sharply limited the doctrine, and probably marked the beginning of its end. Taken together with other recent Court rulings, Kisor is an important step in the judicial piecemeal dismantling of key features of the modern administrative state.

Schwinn is the Professor of Law, UIC John Marshall Law School, University of Illinois Chicago. Many thanks to Christopher Wright Durocher, ACS Senior Director of Policy and Program, and Violet Rush, ACS Law Fellow, for their outstanding editorial work on this piece. All errors, of course, are my own.

[1] Morrison v. Olson, 487 U.S. 654, 697 (1988) (Scalia, J., dissenting).

[2] Gundy v. United States, 139 S. Ct. 2116 (2019). For analysis of Gundy, see William D. Araiza, Toward a Non-Delegation Doctrine That (Even) Progressives Could Like, Am. Const. Soc’y S. Ct. Rev.                        (2019).

[3] Gundy, 139 S. Ct. at 2131 (Gorsuch, J., dissenting).

[4] Id. (Alito, J., concurring).

[5] Chevron, U.S.A., Inc. v. Nat. Res. Def. Council, Inc., 467 U.S. 837 (1984).

[6] Auer v. Robbins, 519 U.S. 452 (1997).

[7] Bowles v. Seminole Rock & Sand Co., 325 U.S. 410, 414 (1945).

[8] Transcript of Oral Argument at 10, Kisor v. Wilkie, 139 S. Ct. 2400 (2019) (No. 18–15).

[9] Kisor, 139 S. Ct. 2400.

[10] Id. at 2425 (Gorsuch, J., concurring).

[11] See Brief of the American Federation of Labor and Congress of Industrial Organizations as Amicus Curiae in Support of Neither Party at 8–9, Kisor, 139 S. Ct. 2400 (No. 18–15) (tracing judicial deference to an agency’s interpretation of its own regulations even earlier, to the early- and mid-nineteenth century). Justice Sotomayor at oral argument pointed to “cases in the early 1800s” as the starting point. Transcript of Oral Argument at 23–25, Kisor, 139 S. Ct. 2400 (No. 18–15). At least four justices on the Court dispute this. In Kisor, Justice Gorsuch, joined by Justices Thomas, Alito, and Kavanaugh, wrote that “Eaton . . . simply followed the well-worn path of acknowledging that an agency’s interpretation of a regulation can supply evidence of its meaning.” Kisor, 139 S. Ct. at 2427 (Gorsuch, J., concurring).

[12] United States v. Eaton, 169 U.S. 331 (1898).

[13] Id. at 338.

[14] Id. at 340.

[15] Id. at 337–42.

[16] Id. at 342–43.

[17] Skidmore v. Swift & Co., 323 U.S. 134 (1944).

[18] Id. at 135.

[19] Id. at 138.

[20] Id. at 139–40. Skidmore isn’t the only case in this post-Eaton period in which the Court applied deference to an agency’s interpretation of its own regulations. See Brief of Administrative Law Scholars in Support of Affirmance at 5 n.3, Kisor v. Wilkie, 139 S. Ct. 2400 (2019) (No. 18–15).

[21] Bowles v. Seminole Rock & Sand Co., 325 U.S. 410 (1945). Importantly, Seminole Rock did not cite Skidmore, but was nevertheless supported by a substantial body of precedent. See Brief of Administrative Law Scholars in Support of Affirmance at 5 n.3, Kisor, 139 S. Ct. 2400 (No. 18–15) (noting that the brief for the United States in Seminole Rock invoked these cases).

[22] Seminole Rock, 325 U.S. at 412.

[23] Id. at 413.

[24] Id. at 414.

[25] Id.

[26] Id. at 413–414 (emphasis added).

[27] Id. at 415 (“The phrase might be construed to mean only the actual charges or sales made during March, regardless of the delivery dates. Or it might refer only to the charges made for actual delivery in March.”).

[28] Id. at 416.

[29] Id. at 417.

[30] Id. at 418.

[31] Id. at 418.

[32] Auer v. Robbins, 519 U.S. 452 (1997).

[33] Id. at 455 (citing 29 U.S.C. § 207(a)(1)).

[34] 29 U.S.C. § 207(a)(1).

[35] Auer, 519 U.S. at 455 (citing 29 C.F.R. §§ 541.1(f), 541.2(e), and 541.3(e) (1996)).

[36] 29 C.F.R. § 541.118(a).

[37] Auer, 519 U.S. at 460.

[38] Brief for the United States as Amicus Curiae at 9–10, Auer, 519 U.S. 452 (No. 95–897). The government went on to provide “factors relevant to determining whether the exemption is lost for employees who are covered by the terms of a manual that on its face permits disciplinary suspensions . . . .” Id. at 10.

[39] The government reiterated its interpretation in its amicus brief on the merits. Brief for the United States as Amicus Curiae at 21–22, Auer, 519 U.S. 452 (No. 95–897).

[40] Auer, 519 U.S. at 461 (quoting Robertson v. Methow Valley Citizens Council, 490 U.S. 332, 359 (1989) (quoting Bowles v. Seminole Rock & Sand Co., 325 U.S. 410, 414 (1945))).

[41] Id. at 462. (citations omitted).

[42] Id. at 462–63 (citations omitted).

[43] These are the four principal lines of critique that seem to have gained traction with some justices and several scholars. But there’s another line of critique—that Auer is wrong because it was based on a faulty reading of Seminole Rock. See generally Jeffrey A. Pojanowski, Revisiting Seminole Rock, 16 Geo. J. L. & Pub. Pol’y 87 (2018) (arguing that Seminole Rock was simply based on a routine application of (the weaker) Skidmore deference and thus does not support the (mightier) Auer deference). At the same time, Auer deference certainly had its defenders. See, e.g., Cass R. Sunstein & Adrian Vermeule, The Unbearable Rightness of Auer, 84 U. Chi. L. Rev. 297 (2017) (defending Auer deference against many of the attacks described below); Scott H. Angstreich, Shoring Up Chevron: A Defense of Seminole Rock Deference to Agency Regulatory Interpretations, 34 U.C. Davis L. Rev. 49 (2000) (defending Auer deference).

[44] See, e.g., Kristin E. Hickman & Mark R. Thomson, The Chevronization of Auer, 103 Minn. L. Rev. Headnotes 103, 105 (2019) (arguing that Auer deference is of “uncertain scope and application”); cf. Sanne H. Knudsen & Amy J. Wildermuth, Unearthing the Lost History of Seminole Rock, 65 Emory L.J. 47 (arguing that the courts have transformed Seminole Rock from its origins and that Auer deference is an unwarranted expansion of Seminole Rock); Matthew C. Stephenson & Miri Pogoriler, Seminole Rock’s Domain, 79 Geo. Wash. L. Rev. 1449 (2011) (arguing for a more limited domain for Seminole Rock deference).

[45] See, e.g., Hickman & Thomson, supra note 44 at 111 (noting “the glut of recent cases in which members of the same court are openly divided on the proper application of Auer” and citing Kisor v. Shulkin, 880 F.3d 1378 (Fed. Cir. 2018); United States v. Havis, 907 F.3d 439 (6th Cir. 2018); Marsh v. J. Alexander’s LLC, 905 F.3d 610 (9th Cir. 2018) (en banc); Turtle Island Restoration Network v. U.S. Dep’t of Commerce, 878 F.3d 725 (9th Cir. 2017); Real Alternatives, Inc. v. Sec’y Dep’t of Health & Human Servs., 867 F.3d 338 (3d Cir. 2017); and G.G. ex rel. Grimm v. Gloucester Cty. Sch. Bd., 822 F.3d 709 (4th Cir. 2016), vacated,137 S. Ct. 1239 (2017); Kevin O. Leske, Splits in the Rock: The Conflicting Interpretations of the Seminole Rock Deference Doctrine by the U.S. Courts of Appeals, 66 Admin. L. Rev. 787 (2014) (describing the inconsistencies in application and confusion in understanding Auer deference in the federal appellate courts).

[46] Perez v. Mortg. Bankers Ass’n, 135 S. Ct. 1199, 1214 (2015) (Thomas, J., concurring in the judgment) (citations omitted).

[47] See generally Matthew C. Stephenson & Miri Pogoriler, Seminole Rock’s Domain, 79 Geo. Wash. L. Rev. 1449, 1459–66 (2011) (arguing for limits to Auer deference based on separation-of-powers concerns). But see Daniel E. Walters, The Self-Delegation False Alarm: Analyzing Auer Deference’s Effects on Agency Rules, 119 Colum. L. Rev. 85 (2019) (concluding, based upon an empirical study of federal rules from 1982 to 2016, that agencies did not measurably increase the vagueness of their regulations).

[48] Decker v. Nw. Envtl. Def. Ctr., 568 U.S. 597, 620 (2013) (Scalia, J., concurring in part and dissenting in part) (citations omitted); see also John F. Manning, Constitutional Structure and Judicial Deference to Agency Interpretations of Agency Rules, 95 Colum. L. Rev. 612, 617 (1996) (arguing that Auer deference “also contradicts a major premise of our constitutional scheme and of contemporary separation of powers case law—that a fusion of lawmaking and law-exposition is especially dangerous to our liberties”). Justice Scalia, of course, wrote Auer. Yet in the years following Auer, he became one of its most vocal opponents. He explained that he changed his position because he had “become increasingly doubtful of its validity,” for separation-of-powers and other reasons. Talk Am., Inc. v. Mich. Bell Tel. Co., 564 U.S. 50, 68 (2011) (Scalia, J., concurring).

[49] See generally Jonathan H. Adler, Auer Evasions, 16 Geo. J.L. & Pub. Pol’y 1 (2018) (“Auer deference undermines—and facilitates the evasion of—basic administrative law principles”); Robert A. Anthony, The Supreme Court and the APA: Sometimes They Just Don’t Get It, 10 Admin. L.J. Am. U. 1, 8 (1996) (arguing that Auer deference violates the judicial-review provision, the notice-and-comment requirement, and “the spirit of the APA”).

[50] 5 U.S.C. § 706 (providing for judicial review of certain agency actions).

[51] 5 U.S.C. § 553 (requiring agencies to notify the public of a proposed new rule and to accept and consider public comments on a proposed new rule).

[52] Perez v. Mortg. Bankers Ass’n, 135 S. Ct. 1199, 1211 (2015) (Scalia, J., concurring in the judgment) (citations omitted).

[53] Id. at 1211–12 (Scalia, J., concurring in the judgment).

[54] Talk Am., Inc. v. Mich. Bell Tel. Co., 564 U.S. 50, 68 (2011) (Scalia, J., concurring) (citation omitted); see also Decker v. Nw. Envtl. Def. Ctr., 568 U.S. 597, 621 (2013) (Scalia, J., concurring in part and dissenting in part) (“In any case, however great may be the efficiency gains derived from Auer deference, beneficial effect cannot justify a rule that not only has no principled basis but contravenes one of the great rules of separation of powers: He who writes the law must not adjudge its violation.”); Paul J. Larkin, Jr. & Elizabeth H. Slattery, The World After Seminole Rock and Auer, 45 Harv. J.L. & Pub. Pol’y 625, 627–29 (arguing that Auer deference “effectively empowers one party to a lawsuit—a federal agency—to decide a legal issue in any case where the federal government is a party” in violation of the maxim that “[a] judge must hear both sides of a case before deciding it” and the related “adversarial system of adjudication” developed by the English courts).

[55] Mortg. Bankers Ass’n, 135 S. Ct. 1199, 1219–20 (2015) (Thomas, J., concurring in the judgment).

[56] Id. at 1219–20. Taking a somewhat different tack, Professor Molot argues that:

[A] tension emerges between judicial practice under Chevron and Seminole Rock and the judiciary’s original role in our constitutional framework. If we gain something in dispute resolution by shifting authority to resolve legal ambiguity from judges to agencies, we also lose an influence over lawmaking that was an important component of the Founders’ constitutional design.

Jonathan T. Molot, The Judicial Perspective in the Administrative State: Reconciling Modern Doctrines of Deference with the Judiciary’s Structural Role, 53 Stan. L. Rev. 1, 8 (2000).

[57] Mortg. Bankers Ass’n, 135 S. Ct. at 1220 (Thomas, J., concurring in the judgment).

[58] Id. at 1221.

[59] Id.

[60] Christopher v. SmithKline Beecham Corp., 567 U.S. 142, 153–59 (2012).

[61] 29 U.S.C. § 213(a)(1).

[62] See Christopher, 567 U.S. at 148–49 (describing the Department’s regulations and guidance).

[63] Id. at 153–54 (quoting Brief for Secretary of Labor as Amicus Curiae at 11, In re Novartis Wage & Hour Litigation, 611 F.3d 141 (2d Cir. 2010) (No. 09–0437)).

[64] Brief for the United States as Amicus Curiae at 12–13, Christopher, 567 U.S. 142 (No. 11–204).

[65] Christopher, 456 U.S. at 155 (quoting Robertson v. Methow Valley Citizens Council, 490 U.S. 332, 359 (1989)).

[66] Id. (quoting Auer, 519 U.S. at 462).

[67] Id. (quoting Auer, 519 U.S. at 462 (quoting Bowen v. Georgetown Univ. Hosp., 488 U.S. 204, 212 (1988))).

[68] Id. (quoting Bowen, 488 U.S. at 213).

[69] Id. at 156 (quoting Gates & Fox Co. v. Occupational Safety & Health Review Comm’n, 790 F.2d 154, 156 (D.C. Cir. 1986) (Scalia, J.)). The Court went on: “Indeed, it would result in precisely the kind of ‘unfair surprise’ against which our cases have long warned.” Id. (citing Long Island Care at Home, Ltd. v. Coke, 551 U.S. 158 (2007); Martin v. Occupational Safety & Health Review Comm’n, 499 U.S. 144 (1991); and NLRB v. Bell Aerospace Co., 416 U.S. 267 (1974)).

[70] Id. at 158 (quoting Talk Am., Inc. v. Mich. Bell Tel. Co., 564 U.S. 50, 69 (2011) (Scalia, J., concurring)).

[71] Id. at 159 (quoting United States v. Mead Corp., 533 U.S. 213, 223 (2001) (quoting Skidmore v. Swift & Co., 323 U.S. 134, 140 (1944)).

[72] See, e.g., Decker v. Nw. Envtl. Def. Ctr., 568 U.S. 597, 615 (2013) (Roberts, C.J., concurring) (“It may be appropriate to reconsider [Auer deference] in an appropriate case. But this is not that case.”).

[73] As a general matter, a veteran qualifies for VA disability benefits if the veteran (1) suffered from a disability (2) in the line of duty. 38 U.S.C. § 1110 (“For disability resulting from personal injury suffered or disease contracted in the line of duty, or for aggravation of a preexisting injury suffered or disease contracted in the line of duty . . . .”). The VA refers to this as a “service-connected” disability. See 38 C.F.R. § 3.303 (stating the factors the VA uses to determine whether a disability is “service-connected”). The VA pays compensation “from the period of service in which said injury or disease was incurred, or preexisting injury or disease was aggravated . . . .” 38 U.S.C. § 1110.

[74] Brief of Petitioner at 13–14, Kisor v. Wilkie, 139 S. Ct. 2400 (2019) (No. 18–15).

[75] Id. at 15.

[76] Id.

[77] Id. at 16–17.

[78] Id. at 16.

[79] Id. at 17 n.4.

[80] Id. at 17.

[81] Id. at 18.

[82] 38 C.F.R. § 3.156(c)(1)(i) (emphasis added).

[83] The Board’s reasoning is set out in Kisor v. Shulkin, 869 F.3d 1360, 1364 (Fed. Cir. 2017), vacated sub nom. Kisor v. Wilkie, 139 S. Ct. 2400 (2019).

[84] Shulkin, 869 F.3d at 1364–65.

[85] Id. at 1366 (quoting Brief of Appellant at 5, Kisor, 139 S. Ct. 2400 (No. 18–15)).

[86] Id. (quoting Brief of Appellant at 9–10, Kisor, 139 S. Ct. 2400 (No. 18–15)).

[87] Id.

[88] Id. at 1367 (quoting Gose v. U.S. Postal Serv., 451 F.3d 831, 836 (Fed. Cir. 2006)).

[89] Id. at 1368.

[90] Kisor noted that Congress enacted the APA in 1946, the year after the Court decided Seminole Rock, but thirty-one years before the Court ruled in Auer. He argued that “[w]hatever could have been said about Seminole Rock prior to 1946, the deference doctrine should not have survived the APA’s enactment.” Brief of Petitioner at 26, Kisor, 139 S. Ct. 2400 (No. 18–15).

[91] Brief of Petitioner at 27, Kisor, 139 S. Ct. 2400 (No. 18–15). Kisor noted that the statute authorizing the Federal Circuit to review decisions of the Court of Veterans Claims contains language similar to the APA’s judicial-review provision. Id.

[92] Brief of Petitioner at 28–31, 33–36, 45–47, Kisor, 139 S. Ct. 2400 (No. 18–15).

[93] Id. at 31–33.

[94] Id. at 37.

[95] Id. at 37–38 (quoting Perez v. Mortg. Bankers Ass’n, 135 S. Ct. 1199, 1212 (2015) (Scalia, J.)).

[96] Id. at 39.

[97] Id. at 40.

[98] Id. at 41–43.

[99] Id. at 43.

[100] Id. at 45 (quoting Decker v. Nw. Envtl. Def. Ctr., 568 U.S. at 619 (Scalia, J.)).

[101] Id. at 47–48.

[102] Id. at 48–51.

[103] Id. at 51 (citations omitted).

[104] Kisor, 139 S. Ct. at 2414–18. Chief Justice Roberts did join this portion of the opinion, making it the Court’s opinion.

[105] Id. at 2414–15.

[106] Id. at 2415.

[107] Id.

[108] Id.

[109] Id.

[110] Id.

[111] Id. at 2416.

[112] Id.

[113] Id. (quoting Arlington v. FCC, 569 U.S. 290, 296 (2013)).

[114] Id.

[115] Id.

[116] Id.

[117] Id.

[118] Id. at 2417.

[119] Id.

[120] Id. Justice Kagan provided two examples: “one requiring the elucidation of a simple common-law property term . . . or one concerning the award of an attorney’s fee.” Id. (citations omitted).

[121] Id.

[122] Id.

[123] Id.

[124] Id. at 2418.

[125] Id.

[126] Id.

[127] Id. at 2422.

[128] Id. (quoting Transcript of Oral Argument at 30, Kisor, 139 S. Ct. 2400 (No. 18–15)).

[129] Id.

[130] Id. at 2422–23 (citations omitted).

[131] Id.

[132] Id.

[133] Id. at 2423.

[134] Id. at 2423–24.

[135] Id. at 2424.

[136] Id.

[137] Id. at 2410. Justice Kagan set out five examples of such regulations and the interpretation problems they raise. Here’s one that perhaps best illustrates the problem:

An FDA regulation gives pharmaceutical companies exclusive rights to drug products if they contain “no active moiety that has been approved by FDA in any other” new drug application. Has a company created a new “active moiety” by joining a previously approved moiety to lysine through a non-ester covalent bond?

Id. (citations omitted). Justice Kagan asked, “To apply the rule to some unanticipated or unresolved situation, the court must make a judgment call. How should it do so?” Id. at 2411.

[138] Id. at 2412 (quoting United States v. Eaton, 169 U.S. 331, 343 (1898)).

[139] Id. at 2412.

[140] Id.

[141] Id. at 2412–13 (citations omitted).

[142] Id. at 2426–29 (Gorsuch, J., concurring). Contrary to Justice Kagan, Justice Gorsuch argued that the Eaton Court applied deference to the agency’s interpretation only after it decided for itself the meaning of the regulation based upon a textual analysis. Id. at 2427.

[143] Id. at 2427 (quoting Skidmore v. Swift & Co., 323 U.S. 134, 140 (1944)).

[144] Id. at 2428 (quoting Bowles v. Seminole Rock & Sand Co., 325 U.S. 410, 414 (1945)). Even then, he argued, it wasn’t entirely clear what this meant. Id. 2428–29.

[145] Id. at 2429.

[146] Id. (quoting Knudsen & Wildermuth, Unearthing the Lost History of Seminole Rock, 65 Emory L.J. 47, 53 (2015)).

[147] Id. at 2430.

[148] Id. at 2419 (plurality opinion).

[149] Id.

[150] Id.

[151] Id.

[152] Id. at 2420 (citing Perez v. Mortg. Bankers Ass’n, 135 S. Ct. 1199, 1208 & n.4 (2015)).

[153] Id. at 2421.

[154] Id.

[155] Id. (“But the claim has notable weaknesses, empirical and theoretical alike. . . . No real evidence—indeed, scarcely an anecdote—backs up the assertion. . . . [And] strong (almost surely stronger) incentives and pressures cut in the opposite direction.”).

[156] Id. at 2432 (Gorsuch, J., concurring) (quoting 5 U.S.C. § 706).

[157] Id. at 2433.

[158] Id. at 2432.

[159] Id. at 2434. He argued, contrary to certain amici in the case, that it didn’t matter that the VA’s interpretation in this case came about in an adjudicative proceeding: whether an agency’s interpretation comes from an adjudicative proceeding, a press release, or some other source, it is not binding upon the courts. Id. at 2435. He also argued, contrary to Justice Kagan, that under Auer deference agency interpretations do have the force of law: “While an agency interpretation, just like a substantive rule, ‘must meet certain conditions before it gets deference,’ ‘once it does so [Auer makes it] every bit as binding as a substantive rule.” Id. (quoting Perez v. Mortg. Bankers Ass’n, 135 S. Ct. 1199, 1211–12 (2015) (Scalia, J., concurring)).

[160] Id. at 2435.

[161] Id. at 2435–36.

[162] Id. at 2436–37.

[163] Id. at 2421 (“If [Kisor’s separation-of-powers] objection is to agencies’ usurping the interpretive role of the courts, this opinion has already met it head-on. Properly understood and applied, Auer does no such thing. In all the ways we have described, courts retain a firm grip on the interpretive function.”).

[164] Id. at 2421–22 (plurality opinion).

[165] Id. at 2437 (Gorsuch, J., concurring).

[166] Id. at 2439; see also id. at 2440 (“Under Auer, a judge is required to lay aside his independent judgment and declare affirmatively that a regulation means what the agency says it means—and, thus, that the law is what the agency says it is.”).

[167] Id. at 2439.

[168] Id. In a related vein, Justice Gorsuch also argued that if Auer were a congressionally enacted stated, it would impermissibly encroach upon the judicial function in violation of the separation of powers. Id. at 2440.

[169] Id.

[170] Only Justices Thomas and Kavanaugh joined this portion of Justice Gorsuch’s opinion; Justice Alito did not.

[171] Kisor, 139 S. Ct. at 2441–42.

[172] Id. at 2443 (quoting Larkin & Slattery, The World After Seminole Rock and Auer, 42 Harv. J. L. & Pub. Pol’y 625,647 (2019)).

[173] Id.

[174] Id.

[175] Id. at 2444.

[176] Id. at 2446.

[177] Id.

[178] Id. at 2447.

[179] Id. at 2448.

[180] Id. at 2424 (Roberts, C.J., concurring in part); id. at 2448 (Kavanaugh, J., concurring).

[181] Id. at 2425 (Roberts, C.J., concurring in part); id. at 2449 (Kavanaugh, J., concurring).

[182] Id. at 2425(Gorsuch, J., concurring).

[183] Id. at 2424 (Roberts, C.J., concurring); id. at 2448 (Kavanaugh, J., concurring).

[184] Id. at 2424–25 (Roberts, C.J., concurring).