Portrait of an Equilibrium


Review of TOCQUEVILLE'S NIGHTMARE: The Administrative State Emerges in America, 1900-1940, by Daniel R. Ernst

Oxford University Press, 2014


         Although Dan Ernst ends his account of the emergence of the American administrative state in 1940, the true climax, at least from the lawyer’s point of view, occurs in 1932. In that year the great Chief Justice Charles Evans Hughes undertook his titanic effort to forge a charter of compromise, a treaty of peace, between the administrative state and the rule of law. The case was Crowell v. Benson, involving an agency charged with deciding workman's compensation cases involving injured maritime workers. Hughes’s opinion in many ways laid down lines of demarcation that were written into the Administrative Procedure Act of 1946, the great framework statute or quasi-constitution of the administrative state. It is a tribute to Hughes that his effort created an equilibrium that outlasted the turbulent years of his Chief Justiceship -- despite the intervening constitutional revolution of 1937, after which the courts stopped trying to enforce narrow readings of the national government’s power over interstate commerce, and stopped trying to police statutory grants of authority from Congress to the executive (the so-called “nondelegation doctrine”).

Having paid due tribute, however, it must be said that the equilibrium Hughes brought into being is a thing of the past. The line of demarcation between administration and law, the frontier of the administrative state, has shifted markedly, with law giving way to administration across almost every margin identified in Crowell -- in large part because law has abnegated its authority to administration. Ernst is not wholly clear about whether the equilibrium he identifies persists all the way into the present, doubtless because the story from 1940 to the present is not the story he is trying to tell. But to understand the significance of his book, it is important to understand that what it offers is a portrait of a particular equilibrium, one that has since vanished. The mid-century attempt to domesticate the American administrative state, described so skillfully by Ernst, ultimately came undone, and it is a live question whether anything else has taken its place.

Ernst’s narrative is highly readable and strikes just the right balance among the historian’s love of detail, the lawyer’s need for conceptual organization, and the political theorist’s addiction to high-level principles. Let me begin with the level of political and constitutional theory. The high-level frame of the book is a choice or contest among possible visions of the relationship between law and administration. Traditional lawyers were afflicted by “Tocqueville’s nightmare,” a vision of a centralized administration abusing its powers and trampling on legal rights. (The nightmare persists, of course, as Philip Hamburger’s recent book shows). The main alternatives or competitors may be understood as different conceptions of “the rule of law.” One alternative, championed by Ernst Freund, was the German idea of the Rechtsstaat -- the rule-of-law state founded on clear positive enactments that would fix the metes and bounds “where the sovereign’s will prevailed and where it yielded to the will of the individual” (p. 2). The Rechtsstaat ideal, however, lost out to a different conception of the rule of law, championed by Hughes among others -- a modified and updated version of Albert Venn Dicey’s ideal that subjected all official action to review by ordinary common-law courts.

After the emergence of the administrative state, the original version of the Diceyan ideal was a non-starter. Ernst shows convincingly that even some traditional lawyers came to understand and appreciate the expertise and efficiency of relatively nonpolitical agencies, who were more professional and less liable to be overrun by patronage politics than other potential suppliers of lawmaking, such as legislative committees, and more knowledgeable and less expensive than the common-law judges and the elaborate processes of litigation. Such lawyers reinvented themselves as transactional engineers, shepherding clients through the administrative process -- not “officers of the court” but “officers of the state” (6). Yet lawyers like Hughes also worked to translate or adapt Dicey’s commitments in the new environment, developing an approach that retained a crucial role for judicial review of administrative action. As Richard Fallon has observed in a different but related context, the translated Diceyan approach attempted not so much to get every given case right, but instead to provide an overall scheme of review that would suffice to keep the administrative state within the bounds of law.

            The framework erected in Crowell v. Benson had multiple components. Speaking very roughly, the main elements were that (1) courts would review all questions of law de novo, without deference to agencies; (2) in adjudication between private parties (cases of “private right”), agencies could decide the facts subject to deferential judicial review for “substantial evidence,” on a formal record developed within the agency itself; (3) however, as to “jurisdictional facts” and “constitutional facts,” judicial review would be based on independent fact-finding, without deference. (There is an interpretive question, discussed most clearly by Mark Tushnet, whether the categories of “jurisdictional fact” and “constitutional fact” are different, or ultimately the same; nothing here or in Ernst’s narrative turns on that issue). Later Hughes opinions fleshed out this quasi-judicial framework for agency decisionmaking, perhaps the most famous being the rule of Morgan v. United States in 1936, according to which “the one who decides must hear” -- meaning that the administrative official who decides the case must personally hear and consider the evidence. Crowell, Morgan and ancillary cases all worked towards a general requirement that agencies must, at least presumptively, decide cases by making reviewable findings on a defined record -- a requirement that Ernst, showing an admirable grasp of the practical importance of legal technicalities, calls “the key to understanding the twentieth-century origins of the administrative state in America” (3).

Viewed in the broad, the new equilibrium had two main features. The one Ernst emphasizes is that the equilibrium arrangements implicitly took courts and judicial procedure as the gold standard, and attempted to judicialize agency procedure as the price of administrative power. Rather than make decisions themselves, courts were to review what agencies did, but also wielded doctrines that commanded or encouraged agencies to use court-like procedures. The second important feature was the compromise character of the arrangements, apparent on their face. Hughes attempted to accommodate and trade off two grand imperatives: the desire to allow a “prompt, continuous, expert and inexpensive method” of decisionmaking, and the need to prevent the administrative state from evolving into what Crowell called “a government of a bureaucratic character alien to our system” -- Tocqueville’s nightmare. Respecting both imperatives, the Crowell framework inevitably had a roughly optimizing character. The Administrative Procedure Act later adopted a similar approach. Justice Robert Jackson famously described the Act as a charter of tradeoffs, one that “settles long-continued and hard-fought contentions, and enacts a formula upon which opposing social and political forces have come to rest.” Among the Act’s key tradeoffs and compromises are its elaborate network of rules that partly respect, but partly abandon, the traditional separation-of-powers notion that the adjudicator must not also be a prosecutor or a rulemaker; the Act separates functions, but only below the level of the agency itself, which can as a general matter make rules, prosecute cases, and decide the cases itself.

Where does the Crowell compromise stand today? Both in terms of what it addressed, and in terms of what it failed to address, it no longer fairly represents the prevailing equilibrium between administration and law. The main elements of the framework have come undone, in ways that have shifted power from courts to agencies. In the main, moreover, this shift has come about through the action of judges and courts themselves -- a process of abnegation rather than conquest. The internal logic of legal argumentation has pushed towards ever-greater judicial deference to agencies.

As to questions of law, even before the Administrative Procedure Act was enacted, the Hearst Publications decision in 1944 indicated that courts might decide, in certain cases, that it is best to defer to agencies’ understanding of legal questions. Hearst and successor cases left it unclear whether such deference should apply only to “mixed questions” of law and fact, or also to “pure questions of law,” and courts struggled with the issue for another generation. By the time of the (in)famous Chevron decision in 1984, however, the courts were willing to adopt a general and largely fictional presumption that Congress intends courts to defer to agencies whenever statutes are ambiguous or silent. Lightly cloaked in that fiction, courts abnegated authority to agencies, in large part -- as the Chevron opinion also admitted -- because of the increasing complexity and political sensitivity of “legal” questions in the administrative state. Traditionalists about the separation of powers have never fully reconciled themselves to Chevron, but for the rising generation it is an intellectual fixture that structures their legal intuitions. A world of genuinely de novo judicial decisionmaking on all “legal” questions is now unimaginable.

            As to questions of fact, deference under the substantial evidence test holds sway, while the doctrines of “jurisdictional fact” and “constitutional fact” are largely moribund. The latter occasionally surfaces in free speech litigation and a few other areas; the former is a historical curiosity. Indeed the very category of agency “jurisdiction” was declared logically incoherent by the Court itself, through Justice Scalia, in City of Arlington v. FCC in 2013, over the dissent of the Chief Justice. Agencies have no “jurisdiction,” just statutory authority in a given domain, and the only question is whether they have reasonably interpreted the scope of that authority -- so the reasoning went. To the traditional legal mind, City of Arlington is a shocking repudiation of the basic premises of legality. It effectively allows agencies to receive deference on the limits of their own power, making them “judges in their own cause.” The Court could barely bring itself to devote a paragraph to dismissing this venerable common-lawyers’ trope; instead the Court emphasized the countervailing danger that judges would transfer power to themselves through excessively intrusive review. With friends like these, law needs no enemies.

            But perhaps the major expansion of the administrative state since Crowell has come not in the areas it addressed, but in an area it said almost nothing about: agency rulemaking. Agencies may act like little courts, as in Crowell, or like little legislatures, making general rules with the force and effect of law. Ernst focuses on adjudication because his subjects did, by and large; Crowell is about agency power to adjudicate, and as Ernst explains, Hughes’s doctrinal contributions center on judicializing agency procedure in adjudication on a formal record. Partly that is because the background law of due process imposes far fewer constraints on agencies acting like mini-legislatures, but partly the explanation is just that contemporaneous agencies made fewer rules.

All that changed in the decades after World War II, especially in the 1970s. Agency rulemaking became central, encouraged by academics and judges who believed that adjudication lacked the virtues of generality and clarity that rules could afford. To some extent, rulemaking was cabined and legalized by (other) judges in the 1960s and 1970s who developed doctrines that required agencies to use extra procedures, beyond those required by the Act; required agencies to give elaborate explanations for rules; and forced agencies to run the gauntlet of “hard look” review -- judicial scrutiny of the rationality of agency decisionmaking. Yet this new domestication of rulemaking was itself partly checked by the Supreme Court, which rejected judicial attempts to proceduralize rulemaking in the Vermont Yankee decision in 1978, and which periodically reins in excessively intrusive “hard look” scrutiny by lower courts.

Despite ongoing controversies over how intrusive judicial review of rulemaking really is (most of the recent empirical work denies that it hampers agencies very much), the larger picture is clear. In recent years a polarized Congress is increasingly hamstrung, and apart from blue-moon events like the Affordable Care Act, the main “legislative” events are agency rulemakings that seriously affect the national economy, as when the Environmental Protection Agency prescribes standards governing whole industries. Revealingly, such rulemakings take place under increasingly antiquated statutes enacted with a view to other problems entirely -- statutes that agencies must fold, spindle and mutilate to make them useful in new circumstances, given congressional paralysis. As Jody Freeman and David Spence put it, under “old statutes” agencies take the lead in addressing “new problems.” One wonders what Hughes, Dicey or for that matter Tocqueville would think of a world in which agencies not only (in effect) rewrite their own authorizing statutes, but have indeed taken center stage in the drama of government. The same forces have also encouraged increasing centralization of authority within the executive branch, as the Office of Information and Regulatory Affairs, an arm of the Presidency, oversees major agency rulemaking.

In both what it addressed and what it failed to address, therefore, the Crowell equilibrium no longer obtains; the effort to domesticate the administrative state, along lines that common lawyers like Dicey might recognize, has failed, and it is an open question whether any such effort can succeed. Ernst observes that even if the equilibrium he portrays no longer obtains, his narrative “remains relevant because it shows that the builders of the new administrative state did not succumb to alien ideologies” (144). His book thus rebuts “a complaint that has gained in popularity since the eruption of the Tea Party movement in 2009: the statebuilders of the early twentieth century abandoned an American tradition of individualism….” (7). This is true and important, yet it is also true that the Tea Party complaint could simply be relocated in time, say to 1984, the year that Chevron was decided -- and indeed Chevron is a major target of recent revanchist works, by Hamburger and others, that damn the administrative state as unconstitutional. All this may limit the significance of Ernst’s book for current lawyers, yet it adds a poignant note that makes the book well worth reading as history, even apart from its great intrinsic merit. The legal statesmen like Hughes who confidently grasped the reins of state, trading off principled imperatives and balancing the claims of administration and law, in a grand effort to “preserve, not renounce, fundamental principles of American government” (144) -- these statesmen were, in the end, no more than flies of a long but fading summer.  

ADRIAN VERMEULE is the John H. Watson Professor of Law at Harvard Law School. He is the author or co-author of eight books on public law and legal theory, most recently The Constitution of Risk (2014).