Numerous academics and even a few jurists have applauded these developments, hoping for a decades-overdue correction in federal administrative law. In 1952, Justice Robert Jackson warned that administrative agencies had “become a veritable fourth branch of the Government, which has deranged our three-branch legal theories much as the concept of a fourth dimension unsettles our three-dimensional thinking.” In decades since, commentators have challenged the administrative state’s departures from rule-of-law principles, deprivations of due process, and challenges to the Founders’ conception of limited government. Yet this is far from a consensus view, as others perceive such “anti-administrativism” as a more ominous trend and wish to rescue the administrative state from its critics before it is too late.

Harvard Law School professors Cass Sunstein and Adrian Vermeule are in the latter camp. In Law & Leviathan: Redeeming the Administrative State, they attempt a rescue. In the face of what they repeatedly describe as a “fundamental assault” on the premises of administrative law, Sunstein and Vermeule seek to explain why administrative law, in operation, is fundamentally moral and sound. The heart of some modest critiques may be true, they concede, but the leviathan of the book’s title is sufficiently constrained by law to preserve its moral legitimacy.

Surrogate safeguards / Rather than offer the full-throated defenses of the administrative state each has offered elsewhere, in Law & Leviathan Sunstein and Vermeule suggest administrative law has developed a set of “surrogate safeguards” that enable the administrative state to protect public welfare while preventing the worst abuses of bureaucratic excess. Aligned with a set of principles articulated by the legal philosopher Lon Fuller in his 1964 book The Morality of Law, these safeguards embody an “internal morality” of administrative law that serves to “both empower and constrain the administrative state.” By requiring agencies to follow their own rules, limiting retroactive rulemaking, and ensuring rules are clear, consistent, stable, and non-contradictory, these safeguards serve to “legitimate, rather than curtail, the administrative state” — and for our authors that is all to the good.

These surrogate safeguards do not derive from constitutional text, nor are they to be found in the 1946 Administrative Procedure Act (APA). Rather, Sunstein and Vermeule explain, they have developed within administrative law over the past 70 years as courts have confronted various challenges to agency action. These safeguards respond to “many of the concerns and objections of those who are deeply skeptical of the administrative state,” but ultimately serve to vindicate, rather than undermine, the prerogatives of administrative law. Accordingly, our authors argue, these safeguards should be accepted as a sufficient response to concerns raised by the administrative state’s critics — those anti-administrativists that Sunstein and Vermeule dismissively deem “the New Coke.” This is a reference to the common-law judge Edward Coke, “a (putatively) heroic opponent of Stuart despotism” who has inspired some of the administrative state’s contemporary critics.

At times, the authors seem to suggest that these surrogate safeguards are baked into administrative law itself. As they note, the APA represented something of an accommodation between constitutional formalism and the post–New Deal administrative state. As Justice Jackson famously observed in Wong Yang Sung v. McGrath (1950), the APA “settle[d] long-continued and hard-fought contentions” through “a formula upon which opposing social and political forces have come to rest.” This “formula” did not give administrative agencies carte blanche, but it stretched prevailing understandings of due process and separation of powers just enough to provide sufficient flexibility and force to be effective. Like all compromises, this formula may have been unsatisfying — and it is neither explicated in the text of the APA nor wholly derivable from the Constitution’s text — but it embodied a set of principles that “offer a powerful rejoinder to many, though certainly not all, of the objections to the administrative state.”

Deference to agencies / While situating the origins of such surrogate safeguards in the APA-era, when pointing to examples our authors sometimes focus on more recent doctrinal developments, including some they themselves opposed. As a consequence, parts of their argument seem to be something of a rearguard action, meant to preserve as much of the administrative state — and agency discretion — as can be salvaged in an age in which devotees of the New Coke may seem ascendant. Nowhere is this clearer than in their treatment of Auer deference, under which courts are obligated to defer to an agency’s reasonable interpretation of its own ambiguous regulation.

Auer deference takes its name from Auer v. Robbins, a 1997 decision in which Justice Antonin Scalia, writing for a unanimous Court, held that an agency’s interpretation of its own regulation must be “controlling” unless it is “plainly erroneous or inconsistent with the regulation.” Under Auer it did not matter how the interpretation was issued, so long as the interpretation represented the official and authoritative position of the agency. Nor, under Auer, did it matter whether the agency’s interpretation was at odds with prior understandings of how the relevant regulatory standards might apply. In Auer itself the Court deferred to the interpretation offered in an agency amicus brief filed at the Court’s request that adopted a non-intuitive (and perhaps politically motivated) interpretation of the Labor Department’s rules concerning which supervisory employees (in this case, police officers) were eligible for overtime. Although the idea that courts should give weight to agency understandings of their own regulations was well-established, Auer embraced a more categorical rule of deference to agencies’ interpretations of their own regulations than had been the norm.

Auer was an inviting target for anti-administrativists because the rule it created was so prone to abuse. Residual ambiguity is rather easy to find in federal regulations concerning complex and technical areas of administrative law. Accordingly, under Auer, regulated firms had little choice but to acquiesce to post-hoc agency interpretations of potentially ambiguous regulatory text. Provided they were reasonable — a low bar in federal court — such interpretations would likely prevail in any subsequent legal proceedings. This was so no matter the form in which the interpretation was expressed.

Under the Chevron doctrine as it has evolved, an agency seeking deference for its interpretation of an ambiguous statutory provision would need to conduct a rulemaking or otherwise act with the force of law to obtain this protection. Under Auer, however, obtaining deference for the interpretation of a regulation required nothing of the kind. A simple guidance document, “Dear Colleague” letter, or other casual missive would suffice, so long as the agency could convince a court that the underlying rule contained residual ambiguity and the interpretation represented the agency’s “fair and considered judgment on the matter in question.” Accordingly, Auer allowed agencies to adopt and revise regulatory interpretations with the stroke of a pen.

Many commentators viewed Auer as something of an outlier within administrative law, unmoored and unsupported by other deference doctrines. Even Scalia came to view it as an aberration, violating the fundamental separation-of-powers principle that “he who writes a law must not ajudge its violation.” Sunstein and Vermeule championed a different view, however. In a 2017 University of Chicago Law Review article, they celebrated “The Unbearable Rightness of Auer” and rejected any calls for its reform. They argued Auer’s downsides were more than outweighed by the value of yielding to agency competence, expertise, and accountability. Given the alternative of instructing generalist judges to offer up controlling interpretations of agency rules, they concluded, “The balance cuts hard in the direction of Auer.”

Sunstein and Vermeule saw nothing wrong with Auer deference as it stood, but the Supreme Court seems to have felt otherwise. When Auer came before the Court in 2019’s Kisor v. Wilkie, a majority of the Court voted against overturning Auer outright, but not a single justice offered an unqualified endorsement of Auer deference as it then stood. While Justice Elena Kagan’s opinion for the Court sought to explicate some of Auer’s virtues, this portion of her opinion only garnered the assent of three other justices, one short of a Court majority. Meanwhile, opinions encompassing the entire Court embraced a suite of conditions and constraints to be appended to Auer in the future. If anything was “unbearable,” it was allowing Auer to continue unrestrained.

As Sunstein and Vermeule remark with some understatement, “Justice Kagan took pains to note that she was merely restating and expanding upon limitations already present in the case law” (emphasis added). A less charitable interpretation would be that the Court had to emasculate Auer deference in order to save it. As Chief Justice John Roberts and Justice Brett Kavanaugh each noted in their separate opinions, there are likely to be few cases in which agencies prevail with the help of Kisor-constrained Auer that would not have prevailed had Auer been overruled outright.

Insofar as Kisor v. Wilkie embodies Fuller’s principles that law should be transparent, consistent, and relatively stable, it is not clear it helps the authors’ case. Kisor represents a fairly late-stage correction that the authors themselves had opposed. Whereas in 2017 Sunstein and Vermeule trumpeted the need to maintain Auer deference so as to preserve agency flexibility, in Law & Leviathan they acknowledge that, under Kisor, agency inconsistency “counts against the agency” as much as if Auer unabridged were no longer on the books. As a technical matter, Auer survives, as it was not overruled. As a practical matter, it is “hedge[d] round with Fullerian constraints” it previously lacked. Despite their earlier position, in Law & Leviathan Sunstein and Vermeule consider this a “vindication of the internal morality of law.”

Herein, perhaps, lies the heart of the authors’ project: not a defense of administrative law as it could or should be, but a defense of as much of existing administrative law as can be maintained. In this sense, Law & Leviathan adopts a defensive crouch, seeking to preserve as much territory as possible as the administrative state’s defenders seek higher and more secure ground from which to repel the forces of New Coke. In this light, the “surrogate safeguards” are as much a reserve line of defense for the modern administrative state as they are a principled accommodation of the anti-administrativists’ critique. In the authors’ ideal world, such safeguards might not be necessary, but their project here is explicitly about identifying and defending a compromise that may serve as a “non-ideal second best.”

In Law’s Abnegation: From Law’s Empire to the Administrative State (2016), Vermeule argued that “the long arc of the law has been steadily toward deference.” In this prior telling, courts had become ever more deferential to administrative agencies as the logic of the law worked itself toward greater consistency and coherence. The mere existence of Law & Leviathan suggests Vermeule’s prior pronouncements may have been a bit premature. As Chief Justice Roberts’ refusal to embrace Auer shows, one need not drink the anti-administrativist New Coke to think deference has gone too far. The law’s internal pressures may now be pushing against abnegation.

Conclusion / In the end, Sunstein and Vermeule are likely correct that the administrative state is here to stay. Future court decisions are more likely to trim doctrine around the edges than they are to disembowel the administrative state. Accordingly, the task of modern administrative law will be to accommodate the actual needs of administration with constitutional constraints and liberal values to reach a “sort of equilibrium accommodation.” In this endeavor, Fullerian principles may prove useful, particularly insofar as they “both channel and enable” administrative law in ways that are responsive to anti-administrativist concerns about separation of powers and due process of law. Yet, just as Kisor departed from our author’s preferences, so too this new equilibrium may lie some distance from the location Law & Leviathan describes, let alone that which the authors may prefer.

Law & Leviathan offers an insightful perspective on the 20th century’s accommodation between law’s morality and the administrative state. Time will tell whether the accommodation it describes is more past than prologue.