Law & Leviathan is an accessible, nicely executed defense of a curiously ill-defined beast—the “administrative state.” Cass Sunstein and Adrian Vermeule, both prolific writers and renowned teachers at Harvard Law School, refrain from cheap polemics, for the most part; indeed, they seek common ground with conservative and libertarian critics of administrative government. Although much can be said for their analysis, their core argument remains unconvincing.

Many of the book’s chapters are previously published law review essays, and the title suggests greater ambition than the content would warrant. “Law” turns out to mean chunks of administrative law. “Leviathan” appears only sporadically, as a scarecrow conjured up by overwrought conservative and libertarian critics of administrative government. The authors’ cringe-inducing term for the “cluster of impulses stemming from a belief in the illegitimacy of the modern administrative state” is, alas, “the New Coke.” Rock-ribbed critics of the administrative state, they write, “frequently refer to the specter of tyranny or absolutism…and they valorize a (putatively) heroic opponent of Stuart despotism: the common-law judge, symbolized by Edward Coke.”

* * *

Both authors have supplied firm (though somewhat varying) defenses of an administrative state with broad discretion to promote the common good. To their minds, assaults on that edifice on supposedly originalist grounds project modern-day fears and values on constitutional materials that simply will not support totemic “New Coke” positions—say, a robust, judicially enforced doctrine that would bar Congress from delegating legislative powers to the executive; or full-scale judicial review of agency decisions. In this book, however, the authors bracket those contentions and instead propose to provide “a structure that can transcend the current debates and provide a unifying framework for accommodating a variety of first-order views, with an eye to promoting the common good and helping to identify a path forward amid intense disagreements on fundamental issues.”

The authors’ lodestar is Justice Robert Jackson’s opinion in Wong Yang Sung v. McGrath (1950), which famously characterized the 1946 Administrative Procedure Act (APA) as a result of “long-continued and hard-fought contentions” and as a compromise on which “opposing social and political forces have come to rest.” Sunstein and Vermeule readily acknowledge that the compromise formula of the APA has long been superseded by fundamental changes in agency organization and procedures and by a substantial body of administrative common law. Still, the authors “aim to recover and renew the force of the principles emphasized in Wong Yang Sung.” Modern administrative law, they argue, has an inherent morality—precepts or principles that protect the rule-of-law values championed by, well, “the New Coke.”

In elaborating those principles, the authors rely on Lon L. Fuller’s famous account of The Morality of Law (1964)—in relevant part, an explication of the minimum conditions a legal system must satisfy to be called “legal” in a moral or rule-of-law sense. Such a system, Fuller argued, must rest on rules, instead of ad hoc decisions. The rules must be publicized, prospective (not retroactive), understandable, consistent, observable, and reasonably stable over time. And there must be a rough congruence between the rules and actual administration.

* * *

Sunstein and Vermeule argue that modern administrative law satisfies those demands most of the time, largely as a result of judge-made doctrines. For example, courts generally require agencies to follow their own rules. As the authors convincingly explain, that doctrine is not easily traced to the APA, let alone the due process clause. It is best understood as reflecting the judges’ Fullerian intuitions. For another example, there “appears to be broad consensus on the [Supreme] Court for the proposition that [judicial] arbitrariness review should impose a heightened burden of justification on agencies when ‘serious reliance interests’ are at stake.”

That, mind you, was written before the Supreme Court held in Department of Homeland Security v. Regents of the University of California earlier this year that the Trump Administration could not unilaterally revoke the Obama Administration’s equally unilateral Deferred Action for Childhood Arrivals (DACA) program without taking serious consideration of the reliance that “dreamers” and other constituencies had placed in the program. That decision, alongside others, lends support to the authors’ contention that “administrative law has developed surrogate safeguards for the values underlying the rule of law.” And that, in a nutshell, is the “redemption” of administrative law. Just as New Deal opponents accepted the APA settlement, Sunstein and Vermeule posit, so the modern Court’s surrogate safeguards “might be acceptable or at least tolerable as a non-ideal second best” for contemporary critics of the administrative state.

Sunstein and Vermeule have several important things right. First, regardless of one’s views of the constitutional order, we will not abolish the Environmental Protection Agency or overrule the New Deal. In that sense, constitutionalists must commit to some set of second-best rules; the only question is what they will look like. Second, some (not all) of the federal courts’ common law-ish doctrines, developed in the shadow of, but at some remove from, both the Constitution and the APA, have in fact made administrative law more regular and law-like. And third, the Supreme Court’s and the D.C. Circuit’s recent decisions lend plausibility to the authors’ analysis. Even so, count me among the skeptics.

For one thing, anyone familiar with the authors’ imposing oeuvre will wonder whether they are actually wedded to the supposed consensus espoused in this book. Sunstein has propounded a “behavioral economics” that would have impartial bureaucrats—themselves free, or so they would have us believe, from bias—“nudge” deplorable citizens into welfare-enhancing choices. Vermeule has powerfully argued that administrative law is really no law at all but a collection of black and gray holes and, as such, wholly incapable of constraining an “unbound executive”—and a good thing, too. “Surrogate safeguards” endorsed by authors with those views may not be worth very much.

* * *

On a more substantive note, Law & Leviathan is usefully read side-by-side with Richard A. Epstein’s recent exploration of The Dubious Morality of Modern Administrative Law. A law professor at New York University and senior fellow at the Hoover Institution, Epstein argues that formal rule of law constraints work best in the context of a classical-liberal regime that rests on property rights, freedom of contract, and protection against uncompensated takings. Once those substantive commitments go by the boards, procedural rule-of-law requirements are bound to prove ineffectual.

Law & Leviathan illustrates the force of that criticism. “A central goal of the rule of law,” the authors write,

is to allow people to have room to maneuver—to create a sphere of action in which citizens do not have to worry about what their government will do…. Many people have been concerned that the administrative state can turn into a form of absolutism, in which citizens must constantly be fearful of what public officials might do. The internal morality of law offers a response.

Does it, really? The passage conflates personal autonomy with government predictability and, correspondingly, reliance. But those are different things, and the difference matters. Legally deportable “dreamers,” the Supreme Court opined in the aforementioned DACA decision, may rely on a government “non-enforcement” policy that was probably unlawful from the get-go; and they may do so despite the government’s explicit declaration that it might change its position any day of the week, for any reason. All the while, none of us may rely on a legally protected, private “sphere of action” beyond government interference: ain’t no such thing under the APA, under administrative common law, or in the Sunstein-Vermeule framework. “A homeless person,” the authors write in a dreadful passage,

is deprived of access to shelter by virtue of the law of property, which is emphatically coercive…. [Modern agencies] did not impose law or coercion where unregulated freedom previously flourished. They substituted one regulatory system for another.

With all respect, there ought to be a safeguard against academic childishness.

* * *

Equally trenchant and apropos is Epstein’s indictment of the administrative law profession’s “single systematic error—the high level of abstraction.” Once one examines administrative law in action, he argues, it turns out to violate rule-of-law constraints at every corner. There is no way to decide between that assessment and the rosier picture of Law & Leviathan without going into the weeds. Professor Epstein’s former colleagues at the University of Chicago Law School, however, are indeed enamored with abstractions, to the point of evasion. “We might favor quite significant reforms” of the administrative state, Sunstein and Vermeule write—with no explanation of what those might look like. “The Constitution and the administrative state attempt to channel and constrain, rather than eliminate or minimize, executive discretion.” Very fine, then: how well has that worked for us, in the age of Obama and Trump?

Sunstein and Vermeule are curiously mum about that pressing question. “Some agency practices,” they acknowledge,

do raise serious constitutional questions…. Sometimes agencies violate the law. Sometimes they act arbitrarily. They can be unfair. They can be influenced by powerful private interests; they might even do their bidding. They might not use their expertise. They can threaten liberty. They can reduce welfare and act in ways contrary to the common good.

Indeed. And because these kinds of things tend to happen a lot, federal courts—and much of the country—are anxiously searching for means of disciplining a feckless but unbound executive.

Professors Sunstein and Vermeule, it is fair to say, do not share that sense of urgency. They fail to provide a single unequivocal example of agency overreach; and, irresponsibly to my mind, they dismiss concerns on that score as a “cluster of impulses.” I commend their search for common ground in a heated debate. But before I sign on to their redemptive enterprise, I’d like to get some sense of where the boundaries are.