1

The Legality of Administrative Law

THIS CHAPTER begins the sketch of law’s abnegation in broad strokes, to set the main outlines of the picture. The following chapters fill in critical details. I will begin with the great 1932 decision in Crowell v. Benson1 and its subsequent history. As we will see, Crowell attempted to create a synthesis or compromise between the claims of administration and the claims of law. In later decades, however, things fell apart.
Critics of the administrative state, who tend to be either originalist or libertarians or (frequently) both, portray this process as the conquest of law by the state. For concreteness I will examine two wholesale criticisms of the legality of the administrative state offered by Gary Lawson and Philip Hamburger. The significance of these theories, for my purposes here, is that they claim—explicitly or implicitly—that the administrative state and the law are antithetical and that the original Constitution, with its commitments to legality and the separation of powers, was shoved aside by political necessity and the expanding state. I argue to the contrary that from a lawyer’s point of view, the main justifications for ever-increasing abnegation were not political but internal and legal. The very institutions of the original Constitution, functioning as they were originally created to function, decided for excellent reasons (from a lawyer’s point of view) to create the administrative state and to abnegate authority to it. Together, the two pieces of the chapter try to offer a unified overview of law’s abnegation and its internal motivation.

Baselines

We can imagine, if only dimly, a system in which judicial review is even more relaxed than it is today; relative to that baseline, our system is highly legalist and intrusive. But from the internal lawyer’s point of view, the right place to begin is not with abstract and rather silly comparisons of hypothetical systems across possible worlds. The right place to begin is doctrinal and historical. The proper benchmark, I believe, is the great hinge between the world of classical legalism and the modern system of administrative law—a great case that attempted to settle a stable equilibrium compromise between the claims of law and the imperatives of bureaucratic government, and that in many ways provided a foundation for the structure of the Administrative Procedure Act (APA). Crowell v. Benson, decided in 1932, was a sweeping attempt to mediate the conflict between law and the administrative state in general terms. Although the attempt failed, it was a grand failure, one that casts its author, Chief Justice Charles Evans Hughes, in an admirable light.
The failure both provides the baseline or benchmark for my claim that law has abnegated authority to the administrative state, and structures the elements of that claim—accounting both for what Crowell did, and for what it left undone. Every important element of the Crowell compromise has become unstable; and in every case, the instability has resulted in a movement in one and the same direction, toward ever-greater judicial deference to administrative discretion. The arc of administrative law bends toward abnegation.
The critical point, or so I will claim, is that the arc results from the law working itself pure. It is not that the law was overcome by external force. Rather, the very arguments that Hughes used to compromise with the administrative state, taken to their logical end, implied that the law should cede authority altogether. The unfolding logic of deference in administrative law represents, not a triumph of state force over reason, but a flowering of reason.

Hughes, between Two Masters

“No man can serve two masters; for either he will hate the one, and love the other; or else he will hold to the one, and despise the other.”2 Charles Evans Hughes, of all people, ought to have understood this; the son of a minister, some of his contemporaries mockingly dubbed him—but only behind his back—“Charles the Baptist.” As Dan Ernst has shown, Hughes attempted to domesticate the administrative state, rejecting two alternative visions: the vision of the reactionary lawyers who wanted to subjugate the administrative state to common-law baselines, through constitutional law; and the vision of the radical progressives and New Dealers, who wanted to subjugate law to the administrative state.3 Hughes attempted to serve both masters, combining and harmonizing both of these impulses in a unifying framework that would adapt the legalist commitments of the common lawyers to new circumstances, while making room for progressive institutional innovation—especially the development of a professional civil service and bureaucracy that would more efficiently process adjudication on the industrial scale needed in the modern state.
Crowell involved the constitutionality of administrative tribunals adjudicating workmen’s compensation claims occurring on navigable waters. As Hughes sketched the main outlines of the synthesis:
  • 1. Courts would decide all questions of “law” de novo, without deference to administrators.
  • 2. Administrative tribunals could (but need not) be given exclusive power by Congress to decide questions of “fact” in cases involving “public rights”—cases between government officials and citizens.
  • 3. In cases involving “private right,” between citizen and citizen—the heartland of the common law—administrative agencies could be given the power to determine the facts, subject to somewhat deferential judicial review for “substantial evidence.” Here Hughes split some differences. On the one hand, he rejected legalist proposals that courts should redetermine all facts de novo, on a new record generated in the trial court, or should at least review the administrative record without deference, asking whether the tribunal’s determination was supported by the “weight of the evidence.” On the other hand, he attempted to preserve a domain in which courts would not wholly defer to administrative determinations of fact.
  • 4. Following on the last point, Hughes also ruled that courts would make de novo factual determinations in two subcategories of cases—those involving
  • (a) “jurisdictional facts,” whose existence or nonexistence supplied the predicate for the statutory “jurisdiction” of the administrative body;4
  • (b) “constitutional facts,” whose existence or nonexistence supplied the predicate for the constitutional power of Congress to confer jurisdiction on the administrative body in the first place. For Hughes, the critical examples involved Congress’s power over admiralty cases and its power over interstate commerce.

The Lions of Constitutional Law

An obvious question was how the Crowell synthesis, established in 1932, would survive the transition from the old constitutional regime to the new in 1937. In the short run, Hughes’s synthesis was enormously influential. Its major outlines were embodied in the administrative state’s large-scale treaty with law: the Administrative Procedure Act of 1946, which said that reviewing courts would “decide all relevant questions of law,”5 specifically including constitutional questions, and that agency factual determinations in formal adjudication would be subject to review only for “substantial evidence.”6 Yet even at the moment of its triumph, there were ominous signs. To understand them, some constitutional background is necessary.
In constitutional law, Hughes had also attempted to steer a middle course. On the one hand, he assembled coalitions in favor of ample congressional lawmaking power under the Commerce Clause and voted to uphold ample grants of statutory authority to administrative agencies. Yet on the other, Hughes also organized judicial resistance to either plenary national lawmaking or to delegations of legislative power so sweeping as to amount to abdication. In perhaps his most famous opinion, A.L.A. Schechter Poultry Corp. v. United States,7 which invalidated the National Industrial Recovery Act on both Commerce Clause and nondelegation grounds, Hughes tried to draw lines, create a narrow but real version of the nondelegation doctrine, and allow judicial intervention in important cases without inviting it in routine cases.
It was not obvious, however, that Hughes’s balancing act could survive the sweeping away of the old regime in 1937–1938. In Jones & Laughlin Steel8 in 1938, Hughes himself tacked back, upholding expansive national lawmaking powers. On the nondelegation front, when the reconstructed Court decided in 1944 in Yakus v. United States9 that Congress could grant agencies the power to set prices in the “public interest,” it was hardly clear what if anything was left of Hughes’s version of the nondelegation doctrine or of his moderating approach to constitutional law proper. A book by Charles Curtis, writing in 1947, described the Justices as “Lions Under the Throne” and said that Bacon’s metaphor “hit … precisely on the relation of the Court with the New Deal.”10 In a world that could be so described—and with a great deal of justice—the Hughesian synthesis of administrative law seemed vulnerable in the extreme. Despite its triumph in 1946, the Hughesian synthesis was unstable and eventually came undone.
To be sure, one might see the Hughesian synthesis and the APA as a set of “substitute safeguards” that implement constitutional values of legality and rationality under changed circumstances, under a new constitutional order. On this view, the Hughesian synthesis became all the more necessary and desirable precisely because of the collapse of the classical constitutional safeguards of the old regime, such as the separation of powers. But that turns out not to be a view that the law has subscribed to over time. As I will discuss in this chapter (with reference to the general notion of “substitute safeguards”), in Chapter 2 (with reference to the separation of powers), and indeed throughout, the law has progressively relaxed the safeguards that Hughes wanted to put into place at the level of administrative law. Those safeguards, especially de novo and nondeferential judicial decision of legal questions, have seemed to modern judges and lawyers to be indefensible on legal grounds themselves.

Internal Tension and Instability

The reason for this basic instability in the framework is that the elements of Hughes’s synthesis sat uneasily together. Indeed there was a warning sign written right across the face of Hughes’s opinion. To many readers, Crowell did not convey the impression of an integrated synthesis, but instead seemed internally conflicted. The great public law scholar David Currie later characterized Crowell as schizophrenic.11 In its first part, allowing agencies principal control over factual determinations even in cases of private right, Crowell extolled the benefits, in accuracy and efficiency, of entrusting workmen’s compensation claims to administrative tribunals. Indeed Hughes implicitly praised the new system for furthering the cause of justice for all, relative to the expensive baseline of common-law litigation. In its second part, carving out jurisdictional and constitutional facts for de novo review, Crowell warned darkly that the administrative state represented a “government of a bureaucratic character, alien to our system.”12 Logically, there was no necessary contradiction between the specific holdings of the two parts, but the deep premises and attitude of each were inconsistent with the deep premises and attitude of the other. The overall impression was that of a house divided against itself, one that could not stand. The first half, extolling the efficiencies of administrative justice, and the second half, fearing the advent of an alien bureaucracy, seemed as though written by different pens. Hughes was attempting to serve two masters simultaneously, and ended up serving neither wholeheartedly.
Hughes had tried to go so far and no further. However, it was hardly obvious that the logic of his arguments could be confined in such a way. After all, no one has ever drawn clear and crisp distinctions among fact-finding, law-interpretation, and policy-making; the three activities bleed into one another in an integrated course of activity by which agencies set legally enforceable government policies within a certain domain. What would happen if the arguments that persuaded Hughes to commit fact-finding to administrative tribunals—arguments from justice, the inadequacy of the common law, expertise, and efficiency—also applied to law-interpretation, for example? In that case the rationales for one half of the synthesis would sweep too broadly, undermining the other half of the synthesis.
And in fact the Crowell synthesis has proved unstable over time and has come largely undone; the balance Hughes attempted to strike has tilted ever farther in the direction of administrative power. The judges, working through the implications of the doctrines and principles by which law constrains the administrative state, have decided that the logic of the Hughesian synthesis, fairly understood, implied greater deference to administrative agencies than even Hughes himself understood. The judges have thus discarded a number of Crowell’s central tenets.

The Collapse of Crowell

Consider the elements of Hughes’s synthesis and how they have fared.
(1) Legal Interpretation. For many of the same reasons that Crowell offered to justify deference to agency fact-finding, fe...