I. Introduction
The “administrative state” is a term that has been in use in academia for some time, but it emerged to occupy a more prominent place in public debates during the presidencies of Barack Obama and, especially, Donald Trump. Occasionally confused with the “deep state”—an amorphous concept most often used to refer to an alleged conspiracy of behind-the-scenes figures who run the government in defiance of its elected officials—the “administrative state” is a relatively straightforward concept. It reflects the reality in modern American government that most policy is not made by Congress, but is instead made by administrative agencies to whom Congress delegates policymaking authority.Footnote 1 As Congress has become increasingly paralyzed due to the ideological polarization in the United States, the agencies of the administrative state have increasingly been relied upon to make policy.
This state of affairs is evident in most of the major policy debates of the day, where the disputes often center around decisions made by the executive, usually through an administrative agency. On the environment, for example, while Congress has not enacted any significant legislation since the amendments to the Clean Air Act in 1990, the Environmental Protection Agency (EPA) has been aggressive in its regulatory activity on greenhouse gas emissions. Major legislation was defeated in Congress in 2009, but the EPA has been busy since that point with a series of major rulemakings pertaining to greenhouse gases. Policy on immigration and citizenship has been made in like manner: while Congress has been largely inactive, the executive—through the Department of Justice (DOJ)—has been the originator of several major immigration policy initiatives. In the Obama administration, the DOJ granted legal status to immigrants under policy initiatives like “DACA,” whereby the children of those who immigrated illegally into the United States were extended status and benefits in the absence of any overt legislative warrant. And the Trump administration was equally active, though in the other direction, as its DOJ launched major initiatives restricting travel from foreign nationals as well as asylum applications. In the arena of healthcare, while Congress itself enacted the Affordable Care Act in 2010, most of the actual policymaking has come from the administrative agencies tasked with the law’s implementation, since the law itself is extremely vague.Footnote 2 It was the Department of Health and Human Services (HHS), for example, that made the policy requiring insurance plans to cover contraceptives, which has become one of the most controversial issues in healthcare policy today. And, as a final example, consider recent attempts to extend federal regulation over the Internet: “net-neutrality” legislation was considered by Congress for years without coming close to enactment, until the Federal Communications Commission (FCC) under President Obama took matters into its own hands and enacted the policy through the regulatory process—a policy that was later rescinded by the FCC under the Trump administration.
In addition to the lead role taken by administrative agencies in the promulgation of all of these policies, they have something else in common: all of them have gone on to be the subject of extensive litigation in the federal courts, which is where the fate of such policies is often settled. The favored strategy of opponents has been to litigate as opposed to legislate. This choice was particularly evident during the Obama administration, where the opposition party controlled at least one house of Congress for the majority of the president’s tenure. But instead of taking the more straightforward approach of opposing the administration’s policymaking ventures through its own legislative activity—especially by using the power of the purse—Congress frequently took the president to court. Opponents to President Trump’s policies did likewise. And as the major policy disputes in the country have thus played out in its courtrooms, increasing attention has naturally been paid to the legal issues pertinent to administrative policymaking, and especially to the constitutional questions that go to the heart of the administrative state. Opponents of administrative action have raised important constitutional objections to administrative action that has been thought, for a long time, to be settled law. And in addition to constitutional law, administrative law is now drawing increased public attention. As courts are more frequently asked to consider the statutory and constitutional limits on administrative policymaking, more scrutiny is being paid to the major doctrines that courts have developed to govern these questions since the origins of the administrative state in the 1930s. Obscure questions previously of concern only to a narrow band of legal scholars —“Chevron deference” or the meaning of the Administrative Procedure Act (APA), for example—have now become part of the public discourse. Scholarship on administrative law—both skepticalFootnote 3 and supportiveFootnote 4—has correspondingly been on the rise.
Has the increased administrative policymaking, and the subsequent challenges to it in the legal arena, led to a reconsideration of the constitutional interpretations that originally accommodated the rise of the administrative state? There is some evidence that such a reconsideration is underway, but in order to understand it we must first consider how the constitutionality of the administrative state came to be established.
II. The Legal Accommodation for the Administrative State
Gary Lawson’s seminal article on “The Rise and Rise of the Administrative State” is still the best explanation of how the federal courts, over the period of time since the 1930s, came to accommodate the role of the administrative state, in spite of its fairly obvious tensions with the basic separation-of-powers structure of the Constitution.Footnote 5 (The best proof of this is that the Progressive-Era fathers of the administrative state were themselves quite open about the fact that it was a legal novelty intended to operate outside the sphere of constitutional government).Footnote 6 The administrative state was built on three main pillars, each of which clashes with core constitutional principles.
The first pillar was the congressional delegation of discretionary and regulatory power to the executive—especially to an enlarged national administrative apparatus which, it was contended, would be much more capable of managing the intricacies of a modern, complex economy because of its expertise and ability to specialize. The second pillar was the combination of powers—legislative, executive, and judicial—into single entities within the administrative apparatus. Proponents argued that it would be much more efficient for a single agency, with its expertise, to be made responsible within its area of competence for setting specific policies, investigating violations of those policies, and adjudicating disputes. The third pillar was the insulation of administrators from political control. For the Progressive fathers of the administrative state, its legitimacy stemmed not from consent but from science. This is why the Constitution’s unitary executive, where the president exercises all of the executive power because he is the lone executive officer accountable to voters, needed to be transformed by the independence of administrative agencies.
By the late 1930s, federal courts had given way on all three of these pillars. The Supreme Court ceased applying the non-delegation principle after 1936,Footnote 7 and allowed to stand a whole body of statutes whereby Congress delegates significant discretionary power to executive agencies. Single federal agencies are also now regularly permitted to exercise all three powers of government—legislative, executive, and judicial. And the courts have permitted the weakening of the political accountability of administrators and the shielding of a large subset of agencies from most political controls. The most common way this shielding is accomplished is by statutes that limit the president’s freedom to remove agency personnel.Footnote 8
As constitutional restraints on the national administrative state were eroded, federal courts came to rely on a growing body of administrative law to govern the scope of national administrative power. This body of law is grounded in the APA of 1946 and the precedents that have been established as courts have applied that law (along with the specific, “organic” statutes that give life to individual agencies) during the growth of the administrative state over the last seventy-five years. Initially intended to rein in national administrative power after the courts had loosened the constitutional restraints on it in the 1930s, the manner in which the APA has been interpreted has led, for the most part, to even greater discretion for national bureaucracies in both procedure and substance.
On procedural questions, the APA was thought to be a check on the discretion of agencies by means of the many trial-like steps it lays out for formal agency rulemaking and adjudication. Affected parties are given, in these steps, significant rights to participate in the decision-making process and to present their own evidence and cross-examine witnesses, among other things. However, in several landmark cases from the 1970s, the Supreme Court greatly narrowed the category of administrative actions to which these formal procedures apply. In the case of United States v. Florida East Coast Railway from 1973, the Court construed the triggering language for formal procedures so narrowly as to virtually eliminate formal rulemaking as a viable category of administrative law.Footnote 9 And in 1978, the Court strictly limited the procedural restraints that could be imposed on agencies engaged in informal rulemaking in the case of Vermont Yankee v. Natural Resources Defense Council, thereby reducing the ability of affected parties to challenge agency decision-making in independent, Article III federal courts.Footnote 10
As these precedents have developed on the procedural side, judicial deference to national administrative power on substantive questions has come to be even greater, though this is a somewhat more recent development. In one respect, such deference seems perfectly consistent with the basic tenets of the administrative state: national bureaucracies were created because the national government was taking on many of the police powers that had previously been handled at the state and local level, and it needed the expertise of administrative agencies to accomplish the task. The federal courts concluded, not unreasonably, that the administrators in the bureaucracy were the experts on the substance of the policies that they had been created to implement, and that judges should not substitute their own, amateur understanding of policy for substantive decisions made by national administrators. Courts thus adopted a sharply deferential posture to the substance of agency decision-making.
The difficulty with this principle, however, comes in the fact that much of the substance of what agencies do involves interpreting the laws they are charged with implementing; and interpretation of law is supposed to be the province of the independent judiciary. The Clean Air Act, for example, places certain requirements for expensive pollution-control equipment on “stationary sources” of pollution; but the Act does not define “stationary source.” Does a single factory—which may contain a number of different emitting devices—constitute a single “stationary source,” or is each discrete emitting device within a single factory its own “stationary source,” thus requiring the factory to make a potentially crippling investment in a multitude of diverse control devices? This seems like an obscure question (as administrative cases often are), but it had major economic consequences, affecting the profitability of an entire industry and the jobs of many workers. And since Congress did not clearly address this question in the legislation, did it intend for an agency to step in and, effectively, make the law on the question? How much latitude do agencies get to fill in the gaps left by legislation, much of which, in our time, has become broad and vague?
In the question posed by the example, which was at issue in the 1984 case of Chevron v. Natural Resources Defense Council, the Supreme Court concluded that gaps in the law are to be filled in by the agency charged with its implementation—that when Congress does not directly address a question in the statute, that lack of clarity is in itself a kind of express intent that the agency should have the power to do so, and that courts reviewing agency action are to grant significant deference to the agency’s interpretation of the law.Footnote 11 That conclusion established what is known as the “Chevron doctrine,” which has become the most important principle in American administrative law. With it, we have gone from the old, constitutional understanding—that for executive agencies to implement policy the legislature must first enact law giving them warrant to do so—to the understanding of national administration today—that when Congress fails to enact a policy, that failure or void can itself be understood as a warrant for national administrators to make policy on the basis of their own expertise. Much of contemporary administrative law—and, thus, much of our understanding of the power of the administrative state today—comes from court decisions applying the Chevron doctrine to a wide variety of administrative action. It is unsurprising that critics of discretionary policymaking by agencies point to Chevron as a principal target for reform, along with its companion doctrine from Auer v. Robbins establishing judicial deference when agencies interpret their own regulations.Footnote 12
III. A Reconsideration of the Administrative State?
The accommodation described above proceeded unabated for four decades following the New Deal, as constitutional challenges to the pillars of the administrative state were rarely raised, much less acted upon. Yet it may be the case that some reconsideration is now underway—a reconsideration that began, in small ways, in the 1970s, but which has lately become more serious. The analysis of this potential reconsideration will proceed by breaking it down into discrete categories of law. The article will look first to the major constitutional challenges: those pertaining to appointment and removal of agency personnel, followed by the potential revisiting of the non-delegation doctrine.
A. Appointment and removal
Presidential control over personnel in the executive branch was considered by the Constitution’s framers to be a critical means of ensuring that national administration remained consistent with the regime’s republican character. Since the president is the only elected official in the executive branch, a “unitary executive” was essential to holding administrators accountable to the people, even if indirectly.Footnote 13 Unitary control over administration comes in the president's ability to appoint and remove the most important agency personnel. First consider the appointment power—a critical question for the viability of the administrative state today—where the Constitution specifies that employees who are “officers of the United States” must be appointed under the methods of the Constitution’s appointments clause. Defenders of the administrative state have sought to remove as many administrators as possible from the umbrella of the appointments clause, as the need to undergo the constitutional process of appointment creates a burden in staffing and empowering administrative agencies. So the question is, who exactly, among agency personnel, are “officers” requiring the constitutional method of appointment?
An early, but slight, blow was struck against the administrative state on this question with the Supreme Court’s 1976 decision in Buckley v. Valeo. Footnote 14 In its attempt to shield an agency—in this case, the Federal Elections Commission (FEC)—from the influence of the president, Congress had devised a rather odd method of appointing commissioners,Footnote 15 which led to the Court’s striking down part of the statute on separation-of-powers grounds—the first time it had done so since the 1930s. Buckley held that FEC commissioners were indeed “officers,” and that they therefore needed to be appointed in accordance with the forms of the Constitution.Footnote 16 The statute in this case failed to do that. Yet Buckley was very much an outlier, as courts in subsequent cases typically took great pains to find that agency personnel did not rise to the level of exercising “significant authority,” which was the bar set by Buckley to trigger the Constitution’s appointment process for “officers.”
In its 2018 term, however, in the case of Lucia v. SEC, the Supreme Court seriously undercut this functionalist reading of Buckley and the appointments clause, finding that Administrative Law Judges (ALJ’s) are indeed “officers,” and thus are subject to the constitutional mode of appointment.Footnote 17 ALJ’s are a staple of the administrative state, exercising considerable power in many agencies, even though—prior to the Lucia decision—their appointment was deemed to be outside the bounds of the Constitution’s prescribed methods for “officers.” The Lucia case is especially significant in that it arose from a controversy under the Dodd-Frank financial-regulation law, which is the model law for advocates of an independent administrative state.
The removal power over agency personnel raises similar issues, as the aim for advocates of the administrative state is to shield agency personnel, as much as possible, from presidential control. The landmark case on this question is Humphrey’s Executor v. United States, from 1935—a case, without which, the administrative state in its present form could not exist.Footnote 18 It allows Congress to create agencies in the executive branch but to insulate them from presidential control. The Supreme Court later expanded on this precedent in the Morrison v. Olson case from 1988, where it permitted even a federal prosecutor—as pure an executive official as one can find—to be shielded from presidential control.Footnote 19
But the Court has since started retreating from its constitutional revisionism in removal power cases. The greatest example of this came when it ruled against a new agency created by the Sarbanes-Oxley law: the Public Company Accounting Oversight Board (PCAOB). The Board answers to the Securities and Exchange Commission (SEC), and is appointed by the SEC, but its members may not be removed at will. This arrangement effectively sets up a double-layer of insulation from presidential control: the SEC cannot remove Board members at will, and the President cannot remove SEC commissioners at will. This is too little control for the president, the Supreme Court concluded in the case of Free Enterprise Fund v. PCAOB, decided in 2010.Footnote 20
In this case, and also in the Lucia case, the Court claimed it was doing nothing radical, and that it was keeping within its post-1930s functionalist precedents. That is technically true, but the momentum has clearly shifted—a shift that has not gone unnoticed by defenders of the administrative state like Justice Stephen Breyer. In his dissent in the PCAOB case, Breyer pointed out that if the Court now objects to shielding agencies from presidential control, the implications for the administrative state are clear going forward.Footnote 21 It will be increasingly difficult for the Court to continue to walk the tightrope of returning to constitutional formalism while also pretending to be faithful to precedent like Humphrey’s or Morrison. The most obvious candidate for further movement would seem to be the Consumer Financial Protection Bureau (CFPB). This agency is the crown jewel of the Dodd-Frank law, with a constitutionally questionable structure, even as administrative agencies go: a single administrator, not removable by the president, whose funds are not subject to annual congressional appropriations. In PHH v. CFPB, a panel of the United States Court of Appeals for the District of Columbia Circuit actually struck down a major piece of the CFPB’s structure on removal power grounds in 2016,Footnote 22 before that panel’s judgment was reversed by the full circuit sitting en banc. Footnote 23 Due to a peculiar set of facts, the case was never appealed up to the Supreme Court. But a similar case is already in the appellate pipeline, where the law has been struck down by a district judge in the Second Circuit on identical grounds.Footnote 24 One indication of the potential fate of the CFPB at the Supreme Court level is that the author of the original District of Columbia Circuit panel opinion was now-Justice Brett Kavanaugh. Kavanaugh’s opinion held the agency’s structure to be unconstitutional on removal-power grounds.
B. Delegation
While courts have entertained constitutional challenges to the administrative state on appointment- and removal-power grounds, until very recently there has been no indication that any reconsideration of the administrative state’s structure might be underway with respect to its first main pillar: delegation. Without the ability of Congress to enact broad directives and to delegate to administrative agencies the power to legislate specific rules under these directives, the administrative state as we know it could not exist. As mentioned above, courts made this accommodation in the 1930s, and have swatted back subsequent attempts to revive the doctrine of non-delegation. A major example of this unwillingness to revisit the delegation question came in 1989 with the Supreme Court’s decision in Mistretta v. United States. Footnote 25 This case featured a constitutional challenge to the U.S. Sentencing Commission on non-delegation grounds, where challengers pointed out that Congress had, in the Sentencing Commission, effectively created a second legislature to exercise purely Article I powers; its only duties were legislative in nature—to come up with criminal penalties for a variety of federal crimes as part of the federal government’s move from indeterminate to determinate sentencing. The Court’s accommodation of the administrative state and abandonment of non-delegation was confirmed in direct language: “Our jurisprudence has been driven by a practical understanding that in our increasingly complex society, replete with ever changing and more technical problems, Congress simply cannot do its job absent an ability to delegate power.”Footnote 26 If the U.S. Sentencing Commission was able to survive a non-delegation challenge, it was difficult to imagine what administrative entity could not.
But the 1989 Supreme Court was populated by at most one or two originalists.Footnote 27 The present makeup of the Court is decidedly different, of which there could be no clearer indication than the bomb that was figuratively dropped in the Gundy v. United States case from the recent term. On the surface the case is unremarkable, since it continued the practice of upholding congressional delegations of legislative authority to administrative entities. Specifically, in the Sex Offender Registration and Notification Act (SORNA), Congress had delegated to the Attorney General the authority to make rules concerning the applicability of registration requirements to pre-Act offenders, and the Court upheld the constitutionality of the delegation. But the opinion of the Court was only signed by a plurality of four justices, who were the only four to affirm the Court’s decades-long accommodation of delegation to administrative agencies.Footnote 28 In dissent, three justices took direct aim at the constitutionality of delegation, contending that the principles of the Constitution require the Court to revisit its long-standing abandonment of the non-delegation principle. Writing for the dissenters, Justice Gorsuch calls the SORNA delegation an “extraconstitutional arrangement” where Congress was attempting to “endow the nation’s chief prosecutor with the power to write his own criminal code.”Footnote 29 And he extends the logic of his criticism to the manner in which Congress undertakes much of its legislative work today: “The Framers understood … that it would frustrate ‘the system of government ordained by the Constitution’ if Congress could merely announce vague aspirations and then assign others the responsibility of adopting legislation to realize its goals.”Footnote 30 This is exactly how Congress prefers to operate today—as Justice Kagan herself recognizes in her opinion for the Court: “if SORNA’s delegation is unconstitutional, then most of Government is unconstitutional.”Footnote 31 While it would be more accurate to say that if SORNA’s delegation is unconstitutional, then much of the administrative state’s rulemaking power is unconstitutional, Justice Kagan’s comment nonetheless captures the threat to the administrative state posed by the reasoning of the dissenters. Their reasoning is relevant in this case because a fourth justice with originalist leanings—Justice Alito—concurred in the Court’s judgment but indicated sympathy with the dissent’s desire to revive the non-delegation principle. The case was heard by only eight of the justices, and Justice Alito noted that there were not, among the eight, enough votes to revisit the Court’s delegation jurisprudence—even if, presumably, he were one of those in favor of revisiting it. His separate opinion concurring only in the Court’s judgment leaves little room for doubt where he might come down should a fifth vote materialize in favor of reviving non-delegation: “If a majority of this Court were willing to reconsider the approach we have taken for the past 84 years, I would support that effort.”Footnote 32 The significance of Justice Kavanaugh’s ascension to the Court, just a few months after Gundy was handed down, was lost on no one, as his originalism in separation-of-powers cases had been well established during his tenure on the Court of Appeals. And Justice Kavanaugh himself underscored the point by going out of his way to make an unusual separate statement on an otherwise unsigned order in the case of Paul v. United States during his first term. This statement praised the Gorsuch dissent in Gundy and noted that it “may warrant further consideration in future cases.”Footnote 33
C. Deference
As the main constitutional pillars of the administrative state—delegation and freedom from presidential removal—have lately drawn scrutiny, there have also been important developments with the major deference doctrines of administrative law. These are doctrines, as explained above, that limit a court’s freedom to undertake de novo review of agency interpretations of statutes and of their own regulations. Without such deference—especially Chevron deference to agency statutory interpretation—the discretionary rulemaking power of the administrative state would be substantially curtailed. Chevron has been controversial since it was first decided in 1984, underscored by the fact that even the Supreme Court at the time could not quite agree on what it had done.Footnote 34 And since then its applicability has been limited in a variety of ways; in more recent cases, those limitations have become more evident and more robust. A brief review of the basics of the so-called “Chevron test” will allow us to see how this is the case.
The basic holding of Chevron is that when the meaning of a statute is unclear, courts are to defer to any reasonable interpretation made by the agency, and are not to engage in their own de novo interpretation of the statute. This formulation gives rise to a two-step test, where courts reviewing agency interpretations first determine if a statute is clear, or if Congress has left a gap for the agency to fill. If the statute is clear, the analysis stops, and the reviewing court goes with the clear meaning of the statute, giving no deference to the agency. If the statute is unclear, the next step is for the court to defer to the agency’s interpretation of it, as long as it is not unreasonable. Considered straightforwardly, this approach grants very considerable power to agencies, since it is not unusual for statutes to be vague. But limits to Chevron are increasing, as are direct assaults on its fundamental reasoning.
First, courts have limited the kinds of agency action to which the Chevron methodology applies. Within a couple of decades after Chevron, the Supreme Court had made clear that its deference doctrine was only to be applied in instances of major agency action where Congress intended such action to have the “force of law,” thus excluding from deference common kinds of agency actions like “policy statements, agency manuals, and enforcement guidelines … .”Footnote 35 The scope of Chevron was further limited in United States v. Mead, where the Court said that the Chevron doctrine should not be applied as a hard and fast rule, but was instead subject to a case-by-case analysis of congressional intent in writing the relevant statute.Footnote 36 These limitations on Chevron were substantial enough to draw vigorous objections from advocates of judicial deference.Footnote 37
An even more serious limitation on Chevron has come in what is being called the “major questions” doctrine—more serious because it has been invoked precisely when the Court is considering agency action on the most prominent public policy issues. Raised initially in litigation over the Food and Drug Administration’s (FDA) attempt to regulate tobacco in the 1990s, this exception to Chevron was employed even more recently when the Court considered agency action under the controversial Affordable Care Act (ACA or “Obamacare”). In cases where, under the Court’s precedents, Chevron would otherwise seem to apply, the Court in these cases has said that major or contentious issues of public policy merit an exception to the assumption in Chevron that ambiguities in the law are indicative of Congressional intent for the agency to make policy. The Court finds it implausible that Congress would intend to make major policy determinations by means of saying nothing about such determinations in a statute. As the Court reasoned about tobacco regulation in 2000, “we must be guided to a degree by common sense as to the manner in which Congress is likely to delegate a policy decision of such economic and political magnitude to an administrative agency.”Footnote 38 And, as it reasoned in 2015 about a major unresolved question in the Obamacare legislation, “had Congress wished to assign that question to an agency, it surely would have done so expressly.”Footnote 39
All of this narrowing of the range of agency actions to which Chevron deference applies has come to be called “Step Zero” of the Chevron test—an initial analysis that must be undertaken that may short-circuit the deference process before it even begins. And once it begins, courts have, since the early days of Chevron, often been willing to leave no stone unturned during Step One, using extensive statutory construction to avoid finding that a law is ambiguous and thus merits Chevron deference to the agency’s interpretation.Footnote 40 There are now, in other words, an increasing number of options available to judges who wish to find reasons to avoid triggering Chevron, though lower courts may be more constrained in straying from the Chevron formula than the Supreme Court.
An even more recent phenomenon in the scaling back of Chevron is developing in Step Two cases—cases where Chevron has been triggered and courts are bound to defer to any agency interpretation that is not unreasonable. Ordinarily, once the analysis reaches this stage, these have been easy wins for the agency. Courts may not substitute their own reading of the law for the agency’s at Step Two; they must simply verify that what the agency has done reflects an interpretation that a reasonable person could find plausible. There are some very recent examples at the Supreme Court level, however, of the Court’s increasing unwillingness to rule automatically in favor of the agency at Step Two. Twice in highly significant litigation over Obama-administration environmental policies, the Court refused to defer to the agency even though it reached the highly deferential second step of the Chevron process. In Utility Air Regulatory Group v. EPA (2014) and again in Michigan v. EPA (2015), the Court paid lip service to the Chevron process, but used Step Two to invalidate the EPA’s interpretations of the Clean Air Act.Footnote 41 Admittedly, this more robust use of Step Two is still in its infancy and thus it would be premature to make too much of it at this point. But there is no question that, when combined with the other ways in which the Court has limited the applicability of Chevron described above, the momentum is going in the direction of a curtailment of judicial deference to agency statutory interpretations—and these interpretations are often the principal foundation for agencies’ discretionary policymaking power. It does not seem out of bounds to ask if Chevron is not, at this point, mere window-dressing, at least at the Supreme Court level: the Court of late has mouthed fidelity to the structure of the Chevron test, but in practice has discovered for itself ample means of avoiding Chevron’s actual consequences.
What lies behind this momentum? It is difficult to miss that underneath all of this tinkering with the scope and mechanics of the Chevron process lies an increasing skepticism of the fundamental reasoning for judicial deference itself. While the most direct attacks on the principle of deference remain—for now—confined to Supreme Court concurrences and dissents, these concurrences and dissents have been joined by a growing number of justices. Moreover, in those recent cases where the Court has upheld the deference regime, it has saved it by allowing its scope to be considerably narrowed. (In this respect, the phenomenon is not unlike what we have seen with limits to the president’s removal power over agency personnel: those limits have been upheld, as described above, but only by means of narrowing their scope; the momentum there seems clear enough).
Examples of direct assaults on the principle of deference itself have become increasingly common. In Michigan v. EPA, Justice Thomas wrote a separate concurrence to contend that the entire regime of judicial deference to agencies is unconstitutional. Among other points, he argues that agency “interpretations” are, for all practical purposes, policymaking actions, and thus violate the separation of powers by doing in the administrative state what the Constitution reserves for the legislature alone.Footnote 42 A broader group of justices raised deep concerns with Chevron in 2013 due to its potential to leave agencies as the final authority on the extent of their own powers under the law. If the power of an administrative agency is defined in and limited by the statute that creates it, and if, under Chevron, courts are to give deference to agencies when they interpret such statutes, then this means that courts must defer to agencies even when agencies interpret statutes to determine the extent of their own powers. This scenario led Chief Justice Roberts to conclude, in a three-justice dissent, that “the Framers could hardly have imagined … the authority that administrative agencies now hold over our economic, social, and political activities,” and to warn that “the danger posed by the growing power of the administrative state cannot be dismissed.”Footnote 43 And even Justice Scalia, who had been the strongest cheerleader for Chevron deference during most of his tenure on the Court, had pretty clearly come to be a skeptic in his final years.Footnote 44
More recently Justice Gorsuch contended that judicial deference “sits uneasily” with the Constitution’s vesting of the judicial power exclusively in the Article III courts. Citing both older and more recent precedents, Gorsuch argued that
a core component of that judicial power is ‘the duty of interpreting [the laws] and applying them in cases properly brought before the courts.’ As Chief Justice Marshall put it, ‘[i]t is emphatically the province and duty of the judicial department to say what the law is.’ And never, this Court has warned, should the ‘judicial power … . be shared with [the] Executive Branch.’Footnote 45
Gorsuch’s reasoning was applied in this case to the question of judicial deference to agency interpretations of their own regulations (“Auer deference”), but there is nothing in it that would not apply equally to Chevron-style deference on interpretations of statutes.Footnote 46 Moreover, this constitutional reasoning was endorsed by three other justices (Thomas, Alito, and Kavanaugh), and a fifth—Chief Justice Roberts—wrote separately to indicate that, while he was not joining Gorsuch’s reasoning with respect to Auer, he considered Chevron an entirely different matter that might warrant reconsideration in a future case.Footnote 47
These most recent arguments were from the Supreme Court’s Kisor v. Wilke decision handed down in 2019—a decision, many will be quick to point out, that upheld the Auer doctrine of deference to agency interpretations of their own prior, vague regulations. But the opinion upholding the Auer regime represented a mere plurality of four justices, and even these justices were only able to save Auer by substantially limiting its applicability. It is not clear how much of Auer deference has actually survived from the limitations placed on it by Justice Kagan in her Opinion for the Court, where she emphasizes that it is “just a ‘general rule’” that ‘does not apply in all cases’.” She attempts to save Auer by explaining that it is “not the answer to every question of interpreting an agency’s rules,” and by going on to enumerate an extensive set of circumstances under which deference would not apply.Footnote 48 Combined with statements, above, from a majority of the Court that this tepid upholding of Auer does not apply to future consideration of Chevron, it seems reasonable to surmise that judicial deference is due for major reconsideration.
The specifics of individual cases make it difficult to predict what the vehicle for such a reconsideration might be, though the Justice Department’s recent ban on “bump-stock” mechanisms for firearms raises the appropriate issues and is currently the subject of a Petition for Writ of Certiorari to the Supreme Court. The agency’s ban on bump-stocks came in a rule that was based on its interpretation of the term “machinegun” used first in the National Firearms Act of 1934, then in the Gun Control Act of 1968 and subsequent federal statutes.Footnote 49 Lower courts, in upholding the rule, employed the Chevron analysis, finding the term “machinegun” to be ambiguous and deferring to the agency’s interpretation of the term to include bump-stocks.Footnote 50 In their appeal to the Supreme Court, the petitioners call for Chevron to be overruled, and make their case on the basis of several constitutional and statutory objections that have been raised in this article and voiced by several of the justices in recent cases.Footnote 51 Whether the Court, for legal or political reasons, chooses not to take up Chevron in this particular vehicle, a reconsideration of Chevron deference seems likely in the not-too-distant future.
D. Restraints on agencies through “arbitrary and capricious” review
The narrative thus far with respect to constitutional and legal restraints on the administrative state has pointed to the long period of permissiveness since the 1930s, followed by a relatively recent trend toward reconsideration of agency discretion. Yet there is one area of law where no reconsideration seems in the offing, because courts have been able to check agency discretion for a long time: by employing Section 706 of the APA, which allows reviewing courts to invalidate agency action deemed “arbitrary and capricious.”Footnote 52 This provision of the APA applies to agency policymaking, as opposed to the agency legal interpretations that are the objects of Chevron and Auer deference. Read plainly, the meaning of the APA is straightforward: when the law directs agencies to make policy, they must show that they have reasons for it—that the policy is the result of some discernible reasoning process. While reviewing courts do not get to substitute their own reasoning for that of the agency, they do get to ensure that the agency does more in justifying its actions than what a parent might say to an inquisitive child: “because I said so.” Courts have come to call this area of the law “hard look review,”Footnote 53 indicating that they must see in the reasons presented by agencies evidence that the agency took a “hard look” at the policy question.
This area of review is worth mentioning in the context of this essay because it has become, of late, a principal vehicle for courts to stymie the policies of executive agencies. By one recent count, federal courts had invalidated actions of the Trump administration at least sixty-three times over the president’s first two years in office.Footnote 54 A substantial portion of these rulings invoked arbitrary and capricious or “hard look” review in concluding that the administration had not provided sufficient reasoning to justify its actions, and applied the standards to major policy areas such as the environment, immigration, and the census. As presidents have come to rely on executive action to make or change policy—a trend as evident in the Obama Administration as in the Trump Administration—litigating on “hard look” grounds has become a favored method of the administration’s opponents.
On the one hand, “hard look” review is sometimes embraced by those who are wary of the administrative state’s power, as it can be a check on administrative discretion in cases where fundamental rights are affected. On the other hand, there is reason to question the desirability of this trend from the perspective of republican principles. Presidents making regulatory policy changes often do so because they have campaigned on the policies and feel as if they have an endorsement from voters to move administrative agencies in a particular direction.Footnote 55 Such was certainly the case with Trump Administration action on the environment and immigration, as it was with Obama-era policies on the environment. Yet the courts have maintained that such democratic reasons will not, as a rule, be acceptable in the “hard look” review process. The major controlling case here is Motor Vehicle Manufacturer’s Association v. State Farm (1983), in which the Supreme Court disallowed Reagan administration changes in regulations on automobile passive restraint systems. Reagan had campaigned in 1980 on the issue of reducing the regulatory burden on domestic automobile manufacturers, contributing to his election victory in states like Michigan. When he came into office his administration justified rescinding certain automobile regulations by arguing that the regulatory changes came out of a change in political leadership in a democratic country—that the voters had the prerogative, through their election of the chief executive, to affect regulatory policy. But the Court was explicit in rejecting that reasoning as insufficient.Footnote 56 Instead, the Court insisted on seeing technical reasons for the policy change, which is a point worth emphasizing: the aim of the Court was not to stand up for individual rights in an instance of democratic excess. The guiding principle was neither republicanism nor liberalism but the supremacy of agency expertise, and that has become the standard in these kinds of cases.
IV. Conclusion
Courts’ continued reliance on arbitrary and capricious review, combined with recent trends portending a reconsideration of the constitutional and administrative-law pillars of the administrative state, suggest that agencies may be entering an era where they will have less discretion to make policy. This comes at a time when, given the divide and intransigence in Congress, the reliance on executive agencies for making policy is very much on the rise. It is not clear what the specific vehicles for this increased judicial scrutiny will be. But this development should be an opportunity for those of us who are inclined to value the founders’ constitutional order to recur to the fundamental principles of that order. As mentioned above with respect to the “hard look” review process, an increase in judicial scrutiny of agency action may well be a welcome development for those of us who see in the founding the pre-eminence of individual rights, and who place great emphasis on the founders’ mechanisms for protecting those rights even when they thwart the wishes of the people. But the founders also sought to establish a republic, and while they surely did not believe that republican principles could justify intrusions on individual natural rights, this is not the same thing as saying that a reliance on the judiciary should be the principal means for resolving the tensions that might arise between the regime’s republican and liberal aims. In the context of recent constitutional and legal challenges to today’s administrative state, skeptics of agency discretionary power ought therefore to ask whether or not increased judicial supervision is the best mode—in a republic—for reining in agency discretion. Would a new era of judicial supremacy in this area simply replace the discretion of one set of unelected rulers with another? And might the founders have promoted other modes for citizens to protect themselves from arbitrary rule—specifically, the people’s ability to control their own fate through the elected branches of government? While it is easy to see why advocates of individual liberty should embrace recent trends in the courts to curb agency power, it can also be argued that other, more republican, remedies would be preferable, including the exercise by Congress of the ample powers provided to it in the Constitution to rein in—and even to eliminate—the administrative state.
