A Review of: Law’s Abnegation: From Law’s Empire to the Administrative State, by Adrian Vermeule

Note from the Editor: This book review discusses and critiques Adrian Vermeule’s new book about the administrative state. The Federalist Society takes no positions on particular legal and public policy matters. Any expressions of opinion are those of the author. Whenever we publish an article that advocates for a position, as here, we offer links to other perspectives on the issue, including ones opposed to the position taken in the article; those can be found on this page under “related content.” To join the debate, please email us at info@fedsoc.org.

Criticism of the modern administrative state is an enduring theme in American legal and political discourse.^[1] This is unsurprising since the administrative state is a complex bureaucracy that operates in ways that are often incomprehensible to members of the general public, that exercises regulatory authority over everything from the air we breathe^[2] to our use and enjoyment of our property^[3] to our speech^[4] to our access to potentially life-saving medicines,^[5]and that seems to consolidate powers that the Constitution is designed to keep separate. In Federalist 51, James Madison described the “accumulation of all powers, legislative, executive, and judiciary, in the same hands” as “the very definition of tyranny.”^[6] Today, federal agencies routinely make rules that govern our conduct, investigate whether those rules have been violated, adjudicate alleged violations, and impose heavy fines or imprisonment upon violators.^[7] If this is not “tyranny,” it does raise uncomfortable questions—questions that have been pressed for more than a century.

Recent years have seen the publication of striking academic and judicial attacks on the legal and political-philosophical premises on which the administrative state rests, as well as the jurisprudence that has facilitated its expansion—jurisprudence characterized by judicial deference to all sorts of assertions of administrative power.^[8] Few scholars have defended the administrative state and administrative jurisprudence against such critiques more zealously than Adrian Vermeule.^[9] Vermeule’s new book, Law’s Abnegation, is a cogent presentation of a bright, optimistic vision of our administrative jurisprudence—one comforting to what Vermeule refers to as the “traditional legal mind.”^[10] On Vermeule’s account, judicial deference to administrative power is the result of abnegation,^[11] that is, the voluntary renunciation of power by those whose “province and duty” it is to “say what the law is.”^[12] To borrow Francis Bacon’s iconic image, our judicial “lions” have leashed themselves “under the throne”—which is to say, under the bureaucracy.^[13] Adopting an interpretive model developed by legal philosopher Ronald Dworkin, Vermeule contends that judicial deference to administrative power both fits with and justifies our administrative jurisprudence—that it is not only consistent with our institutional history, but is normatively desirable. In Dworkin’s terms, deference gives our administrative jurisprudence integrity. And, Vermeule avers, deference is here to stay.

Like all of Vermeule’s work, Law’s Abnegation is taut, insightful, and provocative—a must-read for anyone interested in administrative power and the duty of judges who confront it. Yet Vermeule’s case for judicial deference is ultimately unpersuasive. In Part I of this article, I summarize Vermeule’s arguments for deference. In Part II, I critique those arguments. In Part III, I sketch an alternative approach that better equips judges to discharge their constitutional duties in cases involving administrative power.

I. The Case for Deference

A. The Sound of Dworkin’s Silence

The title and introduction to Vermeule’s book take their inspiration from Ronald Dworkin’s influential 1986 volume, Law’s Empire.^[14] In that volume, Dworkin offered a systematic articulation of his theory of the law, which he called “law as integrity.”^[15] Law as integrity has two components: “fit” and “justification.”^[16] Dworkin’s ideal judge—“Hercules”—adjudicated cases in a way that was both consistent with “language, precedent, and practice” and which served as the best normative justification of the law as a whole.^[17]

As evinced by his description of courts as “capitals of law’s empire,” Dworkin’s legal universe was profoundly jurocentric.^[18] Vermeule thus finds it curious—and significant—that Dworkin utterly failed to discuss the administrative state, which, as Vermeule notes, had “come to structure citizens’ experience of government” long before Dworkin was born.^[19] Vermeule attributes this omission to “willful self-blinding”; he contends that Dworkin knew that there was no answer that he could give concerning the courts’ long-since-established practice of deferring to administrative power that was consistent with his jurocentric vision.^[20] Vermeule adopts an insight by David Dyzenhaus, who has written that “[t]here is no room in [Dworkin’s] account for administrative agencies that have an authority to make or interpret the law in the sense that such administrative decisions are ones to which courts have reason to defer.”^[21] And yet our judiciary has ratified precisely such authority, in what Vermeule describes as “a considered, deliberate, voluntary, and unilateral surrender” on the part of “the law” (by which Vermeule means primarily, although not exclusively, Article III judges) to administrative power.^[22]

It is the project of Vermeule’s book to demonstrate that judicial deference to administrative power both fits with and justifies “the settled fabric of [administrative] law as it has developed across the Anglo-American world.”^[23] Law’s empire, in Vermeule’s telling, has been undone from within, not by coercion or treachery, but on the basis of “valid lawyerly reasons” that “good Dworkinians” ought to accept.^[24]

B. Why Deference Fits

Tracing the trajectory of our administrative jurisprudence entails identifying a baseline from which we can measure the withdrawal or advance of the administrative state upon law’s “heartland,” that is, “courts, and judicial review.”^[25] Vermeule chooses Crowell v. Benson, a 1932 case in which the Supreme Court upheld against an Article III challenge a statute that empowered administrative tribunals to adjudicate workmen’s compensation claims arising from activities on navigable waters.^[26] According to Vermeule’s summary of Crowell, Chief Justice Charles Evans Hughes, writing for the majority, sought to “achieve a stable accommodation of the claims of law and the imperative of bureaucratic government” by distinguishing between questions of “law” and questions of “fact,” as well as between “ordinary” facts, “jurisdictional” facts and “constitutional” facts.^[27] Questions concerning the interpretation of the law, jurisdictional facts, and constitutional facts were to be determined by the courts de novo—without deference to administrative power—but Congress could give agencies exclusive power to decide any ordinary questions of fact in cases involving “public rights” (cases between government officials and citizens).^[28] In cases involving “private rights” (between citizen and citizen), Congress could give agencies power to determine ordinary facts, subject only to deferential judicial review to ascertain whether those determinations were supported by “substantial evidence.”^[29] These distinctions and categories would subsequently be incorporated into the Administrative Procedure Act of 1946 (APA).^[30]

Hughes’s attempt to synthesize an accommodation between law and administrative power failed. “[E]very important element of the Crowell framework has come unglued,” and the result has been the steady advance of judicial deference to administrative power.^[31] Federal courts now defer broadly to administrative agencies’ interpretations of statutes, as well as to agencies’ interpretations of their own regulations.^[32] Hughes failed to anticipate the rise of informal, off-the-record rulemaking, which is now the principal method of administrative decision-making.^[33]Even “hard-look review,” a standard for judicial scrutiny that is used to implement the APA’s instruction that courts set aside actions that are “arbitrary, capricious, or not in accordance with law,”^[34] is in practice highly deferential to agencies.^[35]

Vermeule finds the failure of the Hughesian synthesis instructive. It is not “a tale of the conquest of law’s empire from without,” a top-down coup initiated by progressive political scientists and politicians and ratified by judges who bowed to mere political will or expediency.^[36] It is, rather, the victory of a thoroughly orthodox understanding of the law over a species of idolatry, the latter of which treated “the classical separation of powers as an inviolable command, whatever the sacrifice required to respect it, even if those sacrifices worked to the overall detriment of law itself.”^[37]

As for latter-day idolaters—among them Philip Hamburger,^[38] Gary Lawson,^[39] Jeremy Waldron,^[40] and the late Justice Antonin Scalia^[41]—they come in for rough treatment. Vermeule takes seriously arguments that the administrative state (or at least core features of it) is unconstitutional and that certain judicial doctrines which command judicial deference to administrative power are illegitimate. Yet he ultimately finds these arguments to suffer from crippling flaws.

Vermeule’s criticism of both Hamburger and Lawson focuses on the nature of executive power and the way in which the supposed departures from the Constitution that these scholars identify came about. He charges Hamburger in particular with a profound misunderstanding of the theory which animates current legal doctrine concerning “delegation”—whereby Congress statutorily authorizes agencies to issue general rules that bind members of the public. Everyone agrees, argues Vermeule, that legislators cannot subdelegate legislative power that is constitutionally delegated exclusively to Congress by “We the People” through Article I.^[42] But, he goes on, the statutory authorization of agency rulemaking involves legislators legislating and executive officials executing the law, as they should, and Hamburger is wrong to claim that this arrangement is a subdelegation of legislative power to the executive branch.^[43] The Supreme Court has so held in its better moments,^[44] even if it is not above indulging a fiction that an “intelligible principle” must guide the exercise of delegated discretion to ensure that it is an exercise of executive rather than legislative power.^[45]

Further, Vermeule contends that it is incoherent and pointless for Hamburger and Lawson to complain about the status quo.^[46] For in Vermeule’s telling, the rise of the administrative  entities that Hamburger and Lawson decry because they (allegedly) consolidate legislative, executive, and judicial powers was facilitated “by judges exercising the intrinsically and quintessentially judicial power for statutory interpretation and judicial review” that Hamburger and Lawson believe it proper for them to exercise.^[47] To call this a “dereliction of duty”^[48] on the part of the judiciary is to indict oneself of “hubris”^[49] and indeed to be “unfaithful . . . to the original public understanding of the Constitution”^[50] by departing from the founding generation’s allowance for the “liquidation of ambiguous written legal rules by practice and precedent.”^[51] The administrative state, Vermeule points out, has been so entrenched by “consistent recognition by Congress, President, and Court that capacious delegation of statutory authority is fundamentally legitimate.”^[52] Vermeule summarizes his argument elegantly: “When critics of the administrative state call for a return to the classical Constitution, they do not seem to realize they are asking for the butterfly to return to its own chrysalis.”^[53]

For his part, Vermeule admires both butterfly and chrysalis. As he sees it, the “classical lawmaking institutions”^[54] have properly recognized that we “inhabit a different world of policy-making than did the theorists of the eighteenth century.”^[55] In this world, those institutions must “trade off the quality of policy,”^[56] the “impartiality”^[57] of decision-making, and indeed the “very goal of minimizing abuses of power”^[58]—once of primary importance—in order to capture the benefits of “timeliness”^[59] and “expertise.”^[60] Owing to the rate of change in the policy environment in an increasingly complex economy, “[l]egislative institutions are structurally incapable of supplying policy change at the necessary rates.”^[61]

Vermeule concludes his book by summarizing various doctrines that exemplify and facilitate deference. Certain examples are obvious. Take the doctrine of deference associated with Chevron USA, Inc. v. Natural Resources Defense Council, Inc.^[62] and United States v. Mead Corp.,^[63] which requires judges to defer to “reasonable” agency interpretations of “ambiguous” statutory language where Congress has demonstrated an intention to delegate law-interpreting power to the agency. “Chevron deference” disempowers lawyers and judges by allowing agencies to choose between a range of permissible policies rather than insisting that agencies arrive at one legally correct answer.^[64] But other examples are counterintuitive. In SEC v. Chenery Corp.,^[65] the Court ruled that agency actions can be upheld only on the rationale that the agency itself articulated when taking action, and thus that the agency may not employ post hoc rationalizations during litigation.^[66]Allowing post hoc rationalization would make it easier for those actions to survive judicial review, but it would also empower the agency “lawyers who formulate ex post reasons that are presented to a court.”^[67] Thus, Chenery constrains lawyers and empowers “scientists, engineers, and other technical experts, political appointees within agencies, and civil servants,” all of which help formulate policy before the fact.^[68] Chenery, too, is a kind of legal retreat from administrative power, even if it does not initially appear to be.

C. Why Deference is Justified

Even if deference to administrative power fits with our law, it falls to Vermeule to offer a normatively appealing justification for it—as Dworkin put it, to derive from existing legal materials “an overall story worth telling now.”^[69] According to Vermeule, the synthesis attempted by Hughes in Crowell has come undone because the very reasons which drove the synthesis in first place counseled in favor of broader deference to administrative power—and those reasons are good reasons.

What are these good reasons? Vermeule writes that “the implicit question [in Crowell] is whether judicial review, at the margin, adds net value to the process of institutional decision-making that begins with agency decision-making.”^[70] Hughes concluded that “judicial review promises little additional benefit and threatens to impose incremental delay and litigation costs that will make the overall system worse, not better.”^[71] Hughes failed, however, to properly apply this marginalist analysis to certain kinds of factual questions and to questions of law.^[72] It fell to subsequent courts to do so, and they concluded that agencies had a comparative advantage in answering questions of law as well as questions of fact. Whereas Hughes “assumed that courts were naturally superior to agencies on questions of law,”^[73] it later became clear that it was “impossible to disentangle legal questions from policymaking decisions, at least as to the complex regulatory statutes that predominate in the modern state,”^[74] and, thus, that “agencies, at least as compared to courts, were better positioned both to make ultimate value choices relevant to regulatory questions . . . and also to determine facts, causation, and the likely consequences of alternative interpretations.”^[75]

Vermeuele candidly acknowledges that judicial deference to administrative power carries with it “risks of error and abuse.”^[76] Yet he urges that “accept[ing] increased risks of official abuse and distorted decision-making” is required “in order to give government officials more power to suppress ‘private’ abuses, in order to increase the activity level of the government as a whole, and in order to give administrators sufficient information to combat the evils that arise in complex sectors of the economy.”^[77] Thus, it is all to the good that courts have gradually developed a jurisprudence that does precisely that by following “a predictably and sensibly deferential review of agency policy judgments.”^[78]

D. Toward a Deferential Future

Vermeule does believe that our administrative jurisprudence can be improved, and he has several suggestions for nudging it in what he believes to be the right direction—that is, toward more deference. Perhaps the most intriguing of these suggestions concerns judicial review of agency actions under Section 706(2)(A) of the APA, which provides that agency actions can be overturned if found to be “arbitrary, capricious, or not in accordance with law.”

Vermeule contends that judges should expressly recognize that agencies facing conditions of uncertainty “may have excellent reason to make some decision or other,” yet not a particular decision—and that arbitrary-and-capricious review must be sensitive to this reality.^[79] Surveying the case law, Vermeule finds that judges evaluating the rationality of agency actions “for the most part”^[80] do allow agencies to make what he terms “rationally arbitrary decisions,” but that reviewing courts have yet to entirely jettison “a cramped and erroneous conception of rationality” that “requires agencies to do the impossible by giving reasons as to matters where reason has exhausted its powers.”^[81] While agencies must act on the basis of reasons, Vermeule argues that judges must “recognize[] that limits of time, information, and resources may give agencies good second-order reasons to act inaccurately, nonrationally, or arbitrarily in a first-order sense.”^[82]

What does a rationally arbitrary decision look like? Vermeule offers the example of a 2007 decision by the Fish and Wildlife Service to delist the Yellowstone grizzly bear as a “threatened” species, despite the potential that a decline in the prevalence of whitebark pine might limit a source of sustenance for grizzlies.^[83] The agency rested its decision on the grounds that “grizzlies are notoriously flexible and adaptable about their sources of food . . . bears have proven they can go without [whitebark pine] . . . and other populations of grizzlies have flourished despite the loss of whitebark pine.”^[84] In 2011, this decision was held “arbitrary” by a panel of the Ninth Circuit on the ground there was no evidence in the record “demonstrating grizzly population stability in the face of whitebark pine declines.”^[85] Yet, as Vermeule points out, “there was no information in the record either way . . . and no cost-justified procedure for obtaining such information.”^[86] The agency had to make some decision, and neither pessimism nor optimism concerning grizzly population stability was warranted by the evidence.

Vermeule calls for judges to explicitly and consistently apply a kind of “thin rationality review” that is generally consistent with what they are doing already.^[87] Concretely, this means that courts should not impose even a presumptive requirement of quantified cost-benefit analysis on agencies as a measure of rationality,^[88] require agencies to conduct comparative policy evaluations,^[89] compel agencies to demonstrate the superiority of a chosen policy to past choices, make agencies opt for any particular assumptions (whether pessimistic or optimistic) in the face of uncertainty, demand a “rational connection between the facts found and the choices made,”^[90]or require agencies to explain or convey their reasons “to the satisfaction of a panel of generalist judges.”^[91]

II. Critique

Vermeule’s book offers a profound challenge to critics of judicial deference to administrative power. It presents judicial deference as consistent with the Constitution, consistent with the APA, consistent with precedent, normatively desirable, and in some sense inevitable. In what follows, I will challenge Vermeule’s case for fit and justification and address his argument that deference is in some sense inevitable.

A. Unfitting

In making his case for fit, Vermeule attaches tremendous significance to what he takes to be the fact that law was not violently “overcome” by administrative power or compromised through “treachery.”^[92] Rather, judges, acting freely and in “good faith,” decided to defer to administrative power for legal reasons that they found convincing.^[93] Voluntariness and good faith constitute the foundation of his argument that deference cannot be characterized as “abdication” and that public understanding originalists should not find the result legally objectionable.^[94]

One need not reject Vermeule’s premises to recognize the weakness of his further argument. To take a key example upon which Hamburger’s critique has focused much attention: If in fact Article III’s authorization of “[t]he judicial power” imposes upon judges a constitutional duty to exercise independent judgment concerning the meaning of a statute or a regulation, without regard to the beliefs or desires of government officials concerning what the statute or regulation means, voluntarily chosen deference to the latter would amount to a partial relinquishment of the judicial power.^[95] Such relinquishment would be the very definition of abdication.^[96] Thus, when Vermeule claims that myriad doctrines which Hamburger and Lawson criticize “were developed by judges exercising the intrinsically and quintessentially judicial power for statutory interpretation and judicial review,” he begs the question: Were they really exercising that power?^[97] Or were they declining to exercise power that is delegated to them and ratifying power that is not delegated to the executive branch—neither of which they have the legal power to do?

Vermeule does not provide a convincing answer to this question. He asserts that Lawson is “surely unfaithful to the original public understanding of the Constitution” in claiming that the Constitution bars the kind of delegation that has produced and which perpetuates the administrative state.^[98] Yet although Vermeule alludes at various points to his theories of the proper scope of legislative, executive, and judicial power,^[99] he does not demonstrate that they are grounded in the Constitution’s text, as enriched by the publicly available context at the time of its enactment.^[100] His originalist case, such as it is, ultimately rests upon the “liquidating force of the consistent recognition by Congress, President, and Court that capacious delegation of statutory authority is fundamentally legitimate.”^[101] Vermeule claims thatoriginalists must embrace such delegation because “the founding generation allowed for ‘liquidation’ of ambiguous written legal rules by practice and precedent.”^[102] Here, Vermeule begs another key question. Even if the founding generation allowed for the liquidation of ambiguous legal rules through practice and precedent, and even if that allowance was somehow incorporated into the Constitution’s original meaning, is it really the case that the written rules in question are ambiguous?^[103]

We know from intense Founding-era debates over the wording of constitutional provisions that there was widespread agreement that the Constitution’s language ought to be precisely drafted. “Anti-Federalist” opponents of the proposed, unamended 1787 Constitution warned that imprecise grants of federal power would give rise to usurpations and abuses; the Constitution’s “Federalist” supporters responded that the Constitution’s terms had been drafted as precisely as possible.^[104] In such a context, it cannot be assumed that constitutional language is ambiguous—that is, that it can bear two or more distinct meanings, one of which could potentially authorize the power grants that Vermeule believes to be constitutionally legitimate.^[105] Lawson, Hamburger, and others have adduced considerable evidence that the subdelegation of legislative power is unambiguously forbidden by the Constitution and that many modern congressional grants of power to agencies are unambiguously instances of subdelegation.^[106] Vermeule cannot establish an ambiguity against the weight of the evidence—at the very least, he must show that the evidence is in equipoise.

Vermeule does not adduce sufficient evidence to even muddy the waters. He cites Jerry Mashaw’s scholarship (which purports to trace administrative power back to the Founding Era),^[107] cites a line of cases beginning in the late nineteenth century,^[108] and observes that, “with the arguable exception of Justice [Clarence] Thomas, no modern Justice has fundamentally contested the legitimacy of delegation.”^[109] The Founding-era governmental practices documented by Mashaw are relevant to original public understanding,^[110] but, like all such practices, their existence is not dispositive of their own constitutionality, nor is it proof that the relevant text is ambiguous; those responsible for the practices may well have been mistaken about whether their actions were authorized by the Constitution, and Vermeule makes little effort to show that their beliefs were justified. Both Hamburger and Michael Greve have highlighted salient differences between the practices identified by Mashaw (think of the Steamboat Inspection Service, which regulated a single piece of equipment on a single type of vessel) and the kind of delegation that pervades the modern administrative state (think of the Occupational Safety and Health Act, which authorizes the Secretary of Labor to issue standards that are “reasonably necessary or appropriate to provide safe or healthful employment or places of employment”^[111] and impose those standards on industries across the nation).^[112] Finally, the late nineteenth century cases are obviously not evidence of original public understanding, nor are the opinions of modern Justices.

Originalists who have any interest in staying grounded in reality certainly must acknowledge that Congress, President, and Court have recognized capacious delegation of statutory authority for quite some time. But Vermeule fails to demonstrate that such delegation—however longstanding and widespread—fits with the Constitution’s original meaning.^[113]

B. Unjustified

Vermeule evidently believes that both the Constitution and the APA are shot through with vagueness^[114] and ambiguity.^[115] Interpreting the text can thus only get us so far—there will be contexts in which it does not yield a single determinate answer. Chevron, Auer, and thin rationality review can all be understood as rules of construction that are necessary to give legal effect to vague or ambiguous constitutional and statutory guarantees.^[116] Vermeule’s justification for all of these constructions is simply stated: agencies have a “comparative advantage” in resolving the questions at stake, and judges add little marginal value to agency decision-making.^[117]

Insofar as a normatively desirable but unfitting prescription would lack integrity, Vermeule’s justification depends upon the premise that it is fitting for judges to engage in a particular kind of marginalist analysis. There is a general sense in which judges, like all fallible human beings who operate under the constraints of time and space, necessarily must decide how to focus their intellectual efforts, and it behooves them to consciously do so with reference to value that they seek to capture. Yet Vermeule has something more specific in mind. Judges, he argues, should understand themselves to be part of a “process of institutional decision-making that begins with agency decision-making,” and they ought to judge with an eye to adding something to that process.^[118] Is it fitting for judges to engage in this kind of marginalist analysis?

Robert Natelson, Guy Seidman, and Gary Lawson have convincingly argued that the Constitution’s structure and content reflect its character as a fiduciary document—a document that entrusts government officials (the fiduciaries) with discretionary power to act on behalf of members of the public (the beneficiaries) for limited purposes, through specified means.^[119] Owing to the vulnerability of beneficiaries, the law imposes a set of stringent duties on private fiduciaries, including the duty to follow the beneficiary’s instructions, the duty of utmost good faith (that is, honesty), and the duty to take reasonable care and competently pursue the beneficiaries’ interests.^[120] With discretionary power in the hands of private fiduciaries, those who entrust them with that discretionary power are correspondingly vulnerable; with discretionary power in the hands of public fiduciaries, the entire citizenry is correspondingly vulnerable.^[121] Founding-era writings are replete with references to government officials generally and judges in particular as fiduciaries, whether “agents,” “trustees,” or “representatives.”^[122]

It is doubtful that the marginalist analysis commended by Vermeule is compatible with judges’ fiduciary duties. Judges’ fiduciary duties are centrally concerned with ensuring that officials in the other branches of government adhere to their fiduciary duties by making an independent determination of what the law is in cases, regardless of whether that helps or hinders the other branches’ goals.^[123] Vermeule’s focus on contributions to decision-making that begins with agencies risks transforming the judicial role in cases involving administrative power from an independent one into a collaborative one.

Even if it is appropriate for judges to engage in marginalist analysis along the lines that Vermeule suggests, any such marginalist analysis must incorporate the very real risks of abuse of administrative power. Vermeule rightly warns of the dangers of relying upon unsupported generalizations about agencies’ motivations and judgements.^[124] Yet there is no denying that the discretionary power wielded by agencies is susceptible of being abused. There is, for instance, a rich literature on the phenomenon of “regulatory capture,” wherein the comparative overrepresentation of regulated private interests in the process of agency decision-making results in agency bias in favor of these interests rather than the public interest.^[125] The development of hard-look review of the kind deployed in Motor Vehicle Mfrs. Assn. of the United States, Inc. v. State Farm Mut. Automobile Ins. Co.^[126] can be understood as being in part the product of judicial recognition that agencies do not always seek legally legitimate goals.^[127]

Now consider the thin rationality review that Vermeule believes to be the dominant form of arbitrary-and-capricious review at present and which he urges judges to explicitly embrace: “[A]gencies must act based on reasons, where the set of admissible reasons includes second-order reasons to act inaccurately, nonrationally, or arbitrarily.”^[128] We can safely predict that agencies will seldom fail to offer such reasons, if they understand that this is all that they have to do before judges will uphold their actions. But how, if at all, can thin rationality review safeguard citizens against the risk that agencies will offer insincere, pretextual reasons to conceal their illegitimate ends?

Vermeule points to several methods of “‘flushing out’ an agency’s real motives.”^[129] These include “mandating that the agency make decisions on a formal record; mandating that the agency respond specifically to comments even if there is no formal record; allowing cross-questioning of agency experts; and checking the fit between the agency’s findings and its conclusions.”^[130] It is striking that Vermeule includes the last method, considering that he elsewhere states that judges should not “require agencies to be able to explain or convey their reasons, to the satisfaction of a panel of generalist judges.”^[131] Judges “checking the fit” between findings and conclusions with any rigor would seem to be requiring agencies to do precisely that. If, on the other hand, by “checking the fit” Vermeule simply means requiring agencies to point to factual findings and to point to a legitimate reason for action without inquiring into the connection between findings and action, the problem of pretext remains.

Officials can also fall short of their legal duties without acting in bad faith, that is, without deliberately seeking legally illegitimate ends. Officials may err as a consequence of taking insufficient care in the pursuit of legally legitimate ends.^[132] Some well-documented psychological biases that can distort judgment are particularly pronounced amongst experts—these include egocentrism and overconfidence.^[133] The impact of such biases can be diminished by the “expectation that one may be called upon to justify one’s beliefs, feelings, and actions to others” and will suffer negative consequences if one fails to do so.^[134] Judicial review can offer a means of diminishing the impact of psychological biases and thus encouraging care by ensuring accountability, but it cannot do so if it solely requires officials to offer valid reasons for their actions. Officials who pursue legitimate ends but fail to take sufficient care will easily be able to offer valid reasons for their actions and thus escape negative consequences for them.

Finally, Vermeule marshals little evidence to justify his apparent confidence in agency high-mindedness and care. The closest he gets is his citation to other scholars’ findings that agencies often go above and beyond the courts’ interpretations of their procedural obligations under the APA.^[135] This is interesting and encouraging, but—as Vermeule admits—insufficient to support any generalizations.^[136] Vermeule also never considers whether agencies might act differently if judges were to explicitly and consistently embrace his deferential counsel. If we accept Vermeule’s marginalist terms, it is possible that an increased risk of abuse would be outweighed by the benefits that would be captured, but Vermeule does not sufficiently account for those potential costs in his calculations. Nor, for that matter, does he show that the costs he does acknowledge are or would be outweighed by the benefits of deference along the lines that he urges. Instead, readers are treated to summary assertions that the trade-offs are worth making. Again, if we accept his terms, perhaps they are—but he who asserts must prove, and Vermeule does not prove his marginalist case.^[137]

C. Correctable

Some of the most evocative language in Vermeule’s book is deployed in the service of his argument that legal resistance to administrative power will prove futile. Were the administrative state “abolished,” he predicts that it would “be created again, in a kind of eternal recurrence,” with the judiciary’s aid.^[138] Vermeule reaches all the way back to the seventeenth century in support of this claim, invoking Sir Edward Coke’s “maxim” that, “in a doubtful thing, interpretation goes always for the king.”^[139] Vermeule suggests that Coke was identifying a “baseline tendency” in the law that has “gathered strength over time” as “judges and lawyers come to doubt their own epistemic competence.”^[140]

The language is memorable, but Vermeule’s argument lacks substance. Coke was decrying a “tendency” that was less a product of judicial reflection about “epistemic competence” than of premises concerning royal prerogative power. Royal officials at the time pressured judges to defer to prerogative power that was said to be superior to law.^[141] Some judges did give way to such pressure, but Coke did not, and he urged others to follow his example. What Vermeule refers to as Coke’s “maxim” was in fact a rueful observation that judges often failed to discharge what Coke believed to be their duty of independent judgment and instead followed the path of deference to royal power.^[142] The architects of the administrative state certainly did not believe that judicial deference to administrative power was inevitable—they knew well that they were arguing for a fundamentally different conception of the nature and limits of government than that which was reflected in previous American legal materials.^[143] It was because they knew that the judiciary could not be expected to simply embrace their arguments for consolidating government powers that had traditionally been vested in specialized branches that those arguments went hand-in-hand with calls for judicial deference.^[144]

There is nothing incoherent about arguing that voluntarily chosen judicial deference, even deference predicated upon good legal arguments, should be voluntarily abandoned in the face of better legal arguments against deference. American legal history is littered with precedents which rested upon premises that have since become discredited and which are no longer “good law.”^[145]What critics of judicial deference to administrative power must do is highlight the weaknesses of the legal arguments upon which deference rests and chart an alternative course with sufficient detail to guide judges in resisting assertions of administrative power.

III. Restoring Law’s Supremacy

Vermeule’s most forceful criticism of opponents of administrative power and judicial deference is that the latter do not have a plan. If administrative power and judicial deference are not deeply rooted in our nation’s history and traditions, they have nonetheless been embraced by all three branches of our government for more than a century. Even if Hamburger, Lawson, and others are correct that the administrative state is unconstitutional and the judiciary has abdicated its duty in ratifying it, how should we even begin to repair the damage that has been done?

I share Vermeule’s doubts that many judges are likely at present or in the foreseeable future to lead a charge against administrative power in the name of the Constitution or the rule of law. Yet judges are legally and indeed morally bound to maintain the rule of law by giving effect to the “Supreme Law of the Land.”^[146] It is imperative that they understand their constitutional duties and evaluate assertions of administrative power in a manner that equips them to discharge those duties; so long as they hold judicial office, they must not evade those duties. Further, by focusing specifically on judicial duty, we can avoid presenting judges with a seemingly impossible task.

A. The Letter and the Spirit of the Law

We have seen that Vermeule believes that adjudication in cases involving administrative power ought to be a particular kind of marginalist enterprise. In his view, judicial scrutiny ought to reflect judges’ potential contributions to a process that begins with agency decision-making—and it should be deferential because judges have little to contribute.

There is a stark contrast between this conception of the judicial role and that which informed the drafting of Article III. During the Founding Era, judges were not viewed as part of a decision-making process that commenced in the other branches. Rather, judges were understood to have a duty to exercise independent judgment in accordance with the law of the land in cases properly before them, without deference to the beliefs or desires of government officials or members of the general public and without imposing their own extralegal beliefs or desires.^[147] It was thought that judges contributed to the proper functioning of the system of government of which they were a part, not by thinking institutionally, but by focusing on the merits of particular cases.^[148] In economic terms, this case-specific focus enabled judges to capitalize upon their comparative advantage, which did not lie in wholesale system design, but in retail evaluation of whether particular government actions were lawful.

What could persuade judges to focus their attention in this way? As Hamburger has shown, more than life tenure or undiminished salaries, a particular conception of judicial duty—and a commitment to fulfilling it—was understood to be essential.^[149] That duty was symbolized by and assumed through an oath. The oath initially had religious significance; American judges who often found themselves isolated from their communities^[150] in evaluating the lawfulness of legislative enactments steeled themselves to fear not men but only God.^[151] As they faced down hostile legislative majorities, judges could take comfort in the fact that they were emulating the divine lawgiver in seeking to arrive at an accurate understanding of the law and to impartially give effect to it.

Judges did not have the luxury of infinite time to spend on getting the right answer in any given case, and they inevitably found themselves interpreting and applying written instruments the text of which was insufficient to produce determinate answers to particular questions. Thus, they were forced to rely upon default rules of construction. The distinction between the linguistic meaning of a provision of a written instrument and that instrument’s fundamental purpose or function—whether a contract or a constitution—was expressed through a Christian trope: the distinction between the “letter” and the “spirit.”^[152] Where interpretation of the letter—the linguistic meaning of the text—did not yield a determinate answer, judges had recourse to the spirit—the original function or purpose of that text.^[153]

All of this might seem rather remote from the concerns of contemporary judges. But even in a more secular age, the concept of judicial duty and its association with the oath holds the potential to shape how judges approach cases, and the distinction between the letter and the spirit can be of use in resolving them. In a compelling recent paper, Richard Re has detailed how the oath required of all government officials to “support this Constitution”^[154] can “give[] rise to personal moral obligations” even today.^[155] Re explains that “[n]o hand—either dead or alive—forces individuals to run for office, take the oath, or lead others to think that they will take ‘the Constitution’ seriously.”^[156] Once officials do make such a promise, however, they are legally entrusted with power that they would not otherwise possess—power over their fellow citizens, power to ensure that their fellow citizens are not subjected to unlawful exercises of power.^[157] The oath thus “functions as a bridge between the document and the duty to obey it”—more specifically, it creates a morally binding promise “to adopt an interpretive theory tethered to the Constitution’s text and history.”^[158]

The distinction between letter and spirit also captures an enduring truth. Written instruments are calculated to serve particular functions, and they would be without value if they did not do so. Having recourse to the function, or spirit, of the law where the letter fails—as it may—can equip judges to give effect to the law as best they can. Discerning the spirit of the law entails investigation into the context in which the law was enacted, with an eye to identifying the function or functions that particular provisions would have reasonably be understood to serve. Judges may not, however, disregard the letter in search of the spirit—to do so would violate the duty of good faith that is imposed upon them qua fiduciaries.^[159]

1. Following the Letter: Independent Judgment

What then is a judge who is conscious of and faithful to his or her duty to give effect to the law—both letter and spirit—to do in cases involving administrative power? At least two of the doctrines of deference that Vermeule celebrates are prohibited by the letter of the law—that is, by its text.

Chevron deference and Auer deference require judges to defer to agencies’ “reasonable” interpretations of statutes or regulations, respectively, upon finding that the relevant language is “ambiguous.” To the extent that judges accord such interpretive deference, they cannot be said to exercise independent judgment, which entails an independent effort to ascertain the meaning of the law and give effect to it.^[160] Because the duty of independent judgment is imposed upon them by Article III’s authorization of “[t]he judicial power,” judges violate Article III in declining to discharge it.^[161]

Any argument that judges are merely deferring to the law when they defer to agency interpretations of statutes and regulations is vulnerable to two fatal objections.^[162] First and most fundamentally, Congress has no power to dictate how judges exercise their constitutionally delegated power—and independent judgment lies at the core of “[t]he judicial power.”^[163] Second, the notion that Congress generally intends for courts to defer to agencies is—as Vermeule has pointed out elsewhere—“rankly fictional.”^[164] As Aditya Bamzai has shown in an important constitutional and statutory critique of Chevron deference, the relevant text of the APA, enriched by the context in which it was adopted, is best understood as instructing judges to engage in independent review—consistent with the Hughesian synthesis.^[165] The most one can say on behalf of the view that judges are deferring to the law when they accord Chevron and Auer deference is that is that Chevron and Auer are the law—but that just sends us back to the initial question about whether Chevron and Auer were correctly decided.

In addition, due process of law entails—among other things—impartial adjudication, free from bias towards either party.^[166] Both Chevron and Auer deference require judges in cases involving assertions of administrative power to favor the legal position held by the most powerful of parties—the government. That the bias is systematic and the product of adherence to a perceived legal principle rather than dependent upon the proclivities of individual judges only makes it more troubling because it makes it more certain to influence judges’ deliberations.^[167] According Chevronand Auer deference thus entails violating the Fifth Amendment.

2. Following the Spirit: Judicial Engagement

Although the Supreme Court has maintained that arbitrary-and-capricious review of agency actions is more rigorous than the modern “rational-basis test,” which serves as the default standard of review in constitutional cases, the Court has done little to ground that understanding in the letter or the spirit of the APA.^[168] The APA does not sketch the contours of hard-look review or even suggest such a framework—nor, for that matter, does it sketch or suggest a different framework. Any approach to arbitrary-and-capricious review is necessarily a matter of construction rather than interpretation. Accordingly, judges must seek out the spirit of Section 706(2)(A); this in turn requires study of the publicly available context in which the APA was enacted into law.

The story of the APA’s enactment is one of hard-fought compromise.^[169] That compromise was forged between New Deal Democrats with undiluted faith in technocratic administration on the one hand, and Republicans and conservative Democrats who had become increasingly concerned with what Dean Roscoe Pound described as “administrative absolutism”^[170] on the other.^[171] The former sought the ratification of the New Deal vision of government-by-experts; the latter called for extensive constraints on executive power.^[172]

Neither side got everything that it wanted. The APA provides for some separation of rulemaking, prosecution, and adjudication, some means through which regulated industries can challenge administrative decisions, and some judicial review. But it accepts what Vermeule’s frequent co-author Cass Sunstein has described as the “enduring legacy of the [New Deal] period”: “[the] insulated administrator, immersed in a particular area of expertise, equipped with broad discretion, and expected to carry out a set of traditionally separated functions.”^[173]

If we understand the function of the APA generally and Section 706(2)(A) in particular as a means of ensuring that the administrative state can capture the goods that it was created to provide whilesafeguarding citizens against the abuse of agency discretion, the hard-look review showcased in State Farm is well-tailored to accomplish it. As defined and deployed in State Farm, hard-look review is a form of judicial engagement that can equip judges to discharge their constitutional duties.^[174] State Farm involved a 1982 decision by President Ronald Reagan’s National Highway Traffic Safety Administration (NHTSA) to revoke regulations issued by his predecessor’s administration.^[175] Those regulations would have required vehicles produced after a certain date to include either airbags or automatic seat belts. The NHTSA determined that manufacturers would voluntarily include seat belts rather than airbags and that the regulation would not sufficiently increase seatbelt usage to justify its costs, given that “so many individuals will detach the mechanism.”^[176] Deploying a framework with both procedural and substantive dimensions,^[177] the Court determined that the agency had erred in failing to consider viable alternatives and in making a policy choice that was unreasonable in light of the evidence in the record. The Court pointed out that the NHTSA’s claim that “detachable automatic seat belts cannot be predicted to yield a substantial increase in usage” flew in the face of “empirical evidence on the record, consisting of surveys of drivers of automobiles equipped with passive belts, [which] reveal[ed] more than a doubling of the usage rate experienced with manual belts.”^[178]It also criticized the agency for failing to consider requiring the installation of airbags, even though the agency had “acknowledged the lifesaving potential of the airbag” and despite the fact that airbags cannot be detached.^[179]

The hard-look review showcased in State Farm is comparable to the rationality review that served as the default standard of constitutional review prior to the Supreme Court’s decision in Williamson v. Lee Optical,^[180] and which we find today in cases in which the Court applies “rational basis with bite”; while deferential, it is not toothless.^[181] It requires an actual, rather than a hypothetical, fit between evidence and action.^[182] It requires judges to review the record to determine whether the agency considered the evidence before it in light of contextually relevant factors prior to making a decision.^[183] And while litigants ultimately bear the burden of rebutting a presumption that the agency is acting lawfully, that presumption is rebuttable.^[184]

Implementing the State Farm model of arbitrary-and-capricious review more consistently would, of course, be costly. Time, information, and other resources are scarce, both for judges and for agency officials, and Vermeule is right that “[d]ollars and lives may be lost” if agencies cannot act quickly in certain contexts.^[185] And yet there is ample reason to believe that the benefits of hard-look review outweigh the costs. As Sunstein observed several decades ago, “[t]he requirement of detailed explanation has been a powerful impediment to arbitrary or improperly motivated agency decisions,” and it addresses lingering concerns about the “uneasy constitutional position of the administrative agency” by ensuring that agencies will be held accountable for their decisions.^[186] It provides a framework for ensuring that agency officials comply with the same fiduciary duties that the Constitution imposes on all government actors. It thereby promotes the actual and perceived legitimacy of administrative power, as the APA was designed to do.

Vermeule’s concern that demanding a “rational connection between the facts found and the choice made” demands too much of agencies that must act under conditions of uncertainty is valid. Yet it may be possible to address that concern without doing away with hard-look review entirely. The “rationally arbitrary” decisions that Vermeule regards as critical to the functioning of the administrative state are not arbitrary in the sense of being the product of mere will. They are reality-based, context-sensitive decisions, grounded in the (limited) information available to the decision-makers.^[187] Nothing prevents agencies from explaining in detail why they decided as they did, as well as why any other decision would have been more, less, or equally rational. Judges should be aware that they could be misled concerning uncertainty, but if they are convinced that there is uncertainty and that the agency has outlined legally legitimate second-order reasons for its decision, judges could allow the agency to proceed.

B. Legislative and Executive Duty: The Need for Constitutional Engagement

On Vermeule’s account, judges arrived at deference because there were and are good legal reasons for them to defer. Were judges to become convinced that there are better legal reasons to engage, it stands to reason that the arc of administrative jurisprudence could bend away from judicial deference and toward judicial engagement.

Yet such a change requires that officials in the other branches discharge their own constitutional duties. If Congress continues to enact statutes granting vast and unspecified powers to agencies and agencies continue to argue that they are entitled to deference when their actions are challenged, the judiciary will continue to face enormous pressure to defer. The pressure upon judges to defer will be diminished considerably if the other branches act consistently with their own constitutional duties, neither enacting statutes that purport to subdelegate legislative power nor asking for deference when their actions are challenged in court.

Because the judiciary has acquiesced in broad delegation and itself forged the abovementioned doctrines, relieving this pressure will require legislators and executive officials to articulate and act upon alternative visions of constitutional and statutory meaning. Independent constitutional deliberation by members of branches that are associated with “will” and “force” rather than “merely judgment” may sound quaint and unrealistic, but examples of such deliberation can be found throughout American history.^[188] While the judiciary’s status as a separate branch of government that neither formulates nor executes policy and its relative insulation from extralegal pressures gives it certain institutional advantages in evaluating the legality (if not the wisdom) of particular actions,^[189] nonjudicial actors can and do deliberate independently about the meaning of our law and the principles that undergird it.^[190]

Indeed, legislators and executive branch officials are obliged by their oaths to independently interpret the Constitution and construct rules for implementing it in the statutes they enact and execute. Like judges, legislators and executive officials are elevated to public office only through processes authorized by the Constitution and only after taking an oath of fidelity to “this Constitution.”^[191] Congress is empowered to enact measures that are “necessary and proper”^[192] for carrying delegated powers into execution, and to “lay and collect Taxes, Duties, Imposts, and Excises” in order to “provide for the . . . general Welfare”^[193]; the President is required to “take Care that the Laws be faithfully executed.”^[194] All of this language, writes Natelson, sounds in fiduciary law and discloses a “purpose . . . to erect a government in which public officials would be bound by fiduciary duties.”^[195] Thus, like judges, legislators and executive branch officials are public fiduciaries with corresponding duties, including the duty to follow the letter and the spirit of their constitutional instructions.^[196]

IV. Conclusion

Law’s Abnegation is the work of a legal scholar of the first rank at the height of his considerable powers. If Vermeule’s central thesis is ultimately unconvincing, the problem may lie less with the advocate than with his cause. Broad judicial deference to administrative power may well be the product of serious investigation into extant legal materials and careful reflection upon them in a world very different from that which the Framers knew. But even if Vermeule has advanced the best possible argument for deference, reclaiming territory long since abandoned by the courts to administrative power is neither absurd nor unwise. To those who would take up that task, Law’s Abnegation is not only a challenge, but a potential source of inspiration. Decades’ worth of abnegation will not easily be corrected, but our judicial lions have this consolation—that the letter and the spirit of “the Supreme Law of the Land” is on their side.

Suggested citation: Evan Bernick, Lions Under the Bureaucracy: Defending Judicial Deference to the Administrative State, 18 Fed. Soc’y Rev. 78 (2017).