TABLE OF CONTENTS

Introduction

Courts, litigants, and scholars should not be confused by the ongoing debate about nationwide or so-called “universal” injunctions: the proper scope of remedies under the Administrative Procedure Act (APA) and other statutes providing for judicial review of agency action is “erasure.” This Article aims to save scholars’ recent progress in showing the legality of stays and vacatur under the APA from muddled thinking that conflates these forms of relief with other universal remedies that face growing criticism.

Begin with first principles. When a federal court reviews a legislative enactment that conflicts with a source of higher law (i.e., the Constitution), the court engages in what is essentially a choice-of-law analysis: the court chooses to apply the higher law to the parties in the case at hand and declines to apply the conflicting lower law to those parties. It does not “strike down” the lower law or repeal it, any more than a court choosing to apply Ohio law rather than Michigan law to a tort suit “strikes down” the unchosen Michigan law. To “strike down” the statute in this way would be to exercise legislative, not judicial power—and courts may only exercise the latter. Once the right law has been identified, the remedy is to apply that law to the parties in the case. Nationwide or “universal” injunctions that intend to deliberately affect parties beyond the case exceed the judiciary’s equitable powers, and perhaps the judicial power altogether. But the increasing frequency of such overbroad remedies flows from the fallacy that a court, in finding that the legislative enactment must yield to a higher law in a given controversy, has “erased” the statute. Correct the fallacy, and the proper scope of the remedy comes into focus.

But the “erasure” conception of judicial review is not a fallacy in the context of federal agency action. Federal agency action is subject to review under statutes like the APA that authorize courts to “set aside,” “postpone the effective date of,” “reverse,” or grant other relief directed at agency action itself, rather than at the officials responsible for carrying out agency action. These statutes reflect the principle that Article III courts review agency action analogously to decisions by Article I courts. Federal courts can thus review agency action much like a bankruptcy court’s judgments or a magistrate judge’s report and recommendations. This power to invalidate unlawful agency action exists in other places as well. For example, courts are also allowed to invalidate unlawful agency action taken under the Clean Air Act and the Securities Exchange Act of 1934. This Article’s arguments defending universal remedies under the APA apply equally to agency action reviewed under these provisions.

Professor Mila Sohoni and others have shown that Congress designed judicial review of agency action under the APA to replicate the appellate review model, whereby a superior court judgment takes as its object the inferior court’s judgment and invalidates that initial judgment if it is unlawful. Agency action reviewed under the APA and similar statutes essentially stands as an inferior-court judgment, subject to vacatur if the reviewing court finds it unlawful. This view is consistent with that of lawyer and academic Jonathan Mitchell, whose extensive work criticizing universal injunctions expressly carves out review of agency action. Congress’s decision to subject agency action to these broad remedies is the result of its post–New Deal understanding that agency rulemaking is an exercise of nationwide quasi-judicial, quasi-legislative power that must be checked by judicial review of matching scope. Thus, stay and vacatur of agency action in these contexts are presumptively lawful and appropriate remedies, whereas universal injunctions of presidential action and universal injunctions against enforcement of statutes are not.

Part I of this Article surveys scholarship that shows that the APA authorizes federal courts to issue relief that undoes the agency action under review. That work has established that vacatur is ordinarily the appropriate remedy for an agency rule found to be unlawful. Part II draws on that work to explain that the APA similarly authorizes universal preliminary relief from agency action. Part III shows why the Constitution not only permits, but requires, unlawful agency action to be subject to vacatur. Part IV applies the preceding discussion to contemporary debates about other forms of universal relief. This article aims to keep these debates from overspilling their proper doctrinal banks and disfiguring judicial review of federal agency action, where universal remedies should remain the norm as Congress intended.

I. The APA Instructs Courts to Invalidate Unlawful Agency Action

Scholars have demonstrated that courts truly “strike down” or “erase” unlawful agency action reviewed under the APA. Moreover, in both a recent stay grant and concurrence, Justice Brett Kavanaugh recognized that this distinguishes judicial review of agency action from judicial review of statutes, where universal injunctions are increasingly (and appropriately) suspect. And judicial practice wholeheartedly agrees that whatever may be said of universal injunctions involving statutes, courts should issue universal remedies for unlawful agency action. Every circuit has effectively recognized that the APA authorizes it to vacate a rule,1 and the D.C. Circuit often does so “five times before breakfast.” Because scholarship and practice firmly establish vacatur of federal agency action, this Part only summarizes the primary reasons—textual and historical—that others have advanced in support of it.

A. History Supports Continued Use of Vacatur

Reviewing the APA’s history sets the stage for analyzing its text. Section 706 of the APA directs that “the reviewing court shall . . . hold unlawful and set aside agency action” that is “found to be” unlawful. When Congress was debating and drafting the APA, both Congress and the Executive Branch understood that the phrase “set aside” prescribed judicial invalidation of unlawful regulation. Just four years before the APA’s enactment, the Emergency Price Control Act of 1942 gave an Emergency Court of Appeals exclusive jurisdiction “to stay, restrain, enjoin, or set aside, in whole or in part, any provision of th[e] Act . . . or any provision of any such regulation [authorized by the Act] . . . or to restrain or enjoin the enforcement of any such provision.” This statute shows that Congress understood “set aside” to be an action against an entire provision of a statute or regulation, and distinct from an order “to restrain or enjoin” a provision’s “enforcement” against plaintiffs to a lawsuit. Congress recognized these as different remedies and authorized both in the Emergency Price Control Act. Accordingly, Congress knowingly authorized the greater “set aside” remedy in Section 706 of the APA.

The Executive Branch shared this understanding that to “set aside” agency action meant to invalidate it wholly. The 1941 Attorney General’s Report on Administrative Procedure, in discussing judicial review of agencies’ formal rulemaking, explained that a “judgment adverse to a regulation results in setting it aside.” That sentence shows then-Attorney General Robert H. Jackson’s understanding that the object of the reviewing court’s judgment is the regulation itself, not the regulation’s application in the case at hand.

President Franklin D. Roosevelt used the term “set aside” to denote invalidation too. In an address designed to sell his court-packing scheme to Congress, Roosevelt lamented that “[s]tatutes which the Congress enacts are set aside or suspended for long periods of time” by federal courts. Roosevelt was upset that courts were preventing whole pieces of New Deal legislation from taking effect, not merely exempting individual plaintiffs from compliance. To him, and to the public he addressed, “set aside” meant to invalidate entirely.

B. The APA’s Text Undergirds the Judicial Consensus Favoring Vacatur

That background informs the meaning of APA Section 706, which directs that “the reviewing court shall . . . hold unlawful and set aside agency action” that is “found to be” unlawful. The APA’s definitions section states that “‘agency action’ includes the whole or a part of an agency rule.” As Mila Sohoni explained, “the statute makes agency action the object of the court’s review.” This posture replicates the “appellate review model” in which an appellate court takes an inferior court’s judgment as the object of its review and sets it aside—that is, invalidates it—if the appellate court finds the judgment is unlawful.

Another way to understand this review structure is by analogy to bankruptcy law. When a federal court reviews agency action under the APA, the relationship between the court and agency is like that between an Article III federal district court and the Article I bankruptcy court under its supervision. The inferior actor, be it an agency or bankruptcy court, takes the first shot at determining legal duties and obligations, but that determination has no force if the reviewing court finds it inconsistent with law. Both scenarios ensure that the final arbiter of legal rights and obligations is an actor that the Constitution itself creates and entrusts with Article III judicial power.

Jonathan Mitchell agrees that “[u]nlike judicial review of statutes, in which courts enter judgments and decrees only against litigants, the APA . . . [goes] further by empowering the judiciary to act directly against the challenged agency action.” This statutory design “enables the judiciary to formally revoke an agency’s rules . . . in the same way that an appellate court formally revokes an erroneous trial-court judgment.”

In fact, the majority of the APA’s drafters assumed that most administrative agencies would regulate through quasi-judicial adjudication, not rulemaking. As Professor Reuel Schiller observed, “[b]efore the 1960s agencies acted mainly through case-by-case adjudications,” and “[m]ost traditional administrative actions—ratemaking, for example—were based on judicial models.”2 The New Deal expansion of administrative agencies may be understood as a proliferation of what looked to Congress like a cornucopia of Article I courts. Given that “[a]dministrative proceedings looked like mini-trials, where the rights of individual actors were adjudicated,” it is not surprising that “critics of the Roosevelt administration, who aggressively pushed for the passage of the APA, focused their energies on making agency adjudications more like common law trials.”3 It is also clear that the judicial review provisions of the APA re-constitutionalized agencies by placing them in an appellate-review chain of command under Article III courts. When an Article III court sets aside an unlawful agency action, be it a narrow adjudicatory order or a nationwide rule, that action ceases to have any force.

Contrary readings of Section 706’s “set aside” language are implausible. Best read, it cannot mean, as Professor John Harrison argued, that a court should only decline to apply a rule to the parties who challenged it. The statutory text instructs courts to “set aside” an unlawful “rule,” not enjoin agency personnel from enforcing the rule against parties. In defining “agency action,” the APA equates an agency “rule” with an “order” produced through trial-like agency adjudication. It then instructs courts to “set aside” agency actions that are unlawful. Congress thus drew the easy analogy between judicial review of lower court orders and court-like agency adjudication orders and prescribed the same remedy for both. It then extended this appellate review analogy to review of agency rules as well.

The fact that everyone in 1946 expected agencies to do most of their regulating through court-like adjudication orders rather than quasi-legislative rulemaking does not mean that agency rulemaking stands outside the appellate-review model.4 For one thing, agencies cannot skirt judicial review by regulating more people with less process than Congress expected in the 1940s. Moreover, by the time of the APA, agencies had long been promulgating regulations with nationwide scope through individual actions. They simply called these regulations “orders” instead of “rules,” and courts had granted universal set-aside relief against them in at least three pre-APA cases. In the illustrative example of United States v. Baltimore & Ohio Railroad Co., the Supreme Court affirmed a three-judge district court’s ruling that an Interstate Commerce Commission “order” requiring railroads to install a power reverse gear on their locomotives be “vacated, set aside, and annulled,” and that its enforcement be “perpetually enjoined.” The Congress that passed the APA understood that courts would review and vacate agency action that sought to regulate nationally.  Indeed, the APA commands courts to do so.

Another flawed textual argument against universal APA remedies is that the APA sets forth remedies in Section 703, not Section 706, and so Section 706’s “set aside” language does not address remedies at all. Solicitor General Elizabeth Prelogar advanced this argument at oral argument in United States v. Texas, echoing Professor Harrison. Section 706, however, authorizes courts, in appropriate circumstances, to “compel agency action,” which is most definitely a remedy. And the APA’s structure leads one to expect to find final remedies in Section 706, right after Section 705 introduces (some universal) interim remedies. Section 703, on the other hand, is labeled “[f]orm and venue of proceeding,” and makes no reference to remedies besides once using the word “injunction” to specify that the permissible “form[s] of legal action[ ] includ[e] actions for . . . writs of prohibitory or mandatory injunction or habeas corpus.” Section 703’s authorization of a specialized form of proceeding that happens to have the word “injunction” in its name does not make Section 703 a remedies provision, let alone an exclusive one. The APA’s text and structure suggest that courts have been right all along—they should ordinarily vacate unlawful administrative rules and remand them to the agency for reconsideration.

II. The APA Authorizes Universal Interim Relief

Less has been written on the interim remedies available under the APA, although the importance of those remedies has only increased. Indeed, a growing number of high-profile challenges to agency action have reached the Supreme Court not through petitions for certiorari, but in an emergency posture.5 As Justice Neil Gorsuch has observed, in this setting, interim remedies control the challenged action’s fate for months or years during litigation, and often practically become final remedies when litigation outlives the challenged action. This Part shows that the APA grants courts authority to stay agency rules from taking effect pending appeal to the Supreme Court or denial of certiorari, even if courts lack authority to universally enjoin statutes and direct presidential action.

The textual argument for universal interim remedies under the APA is perhaps even stronger than that for universal final remedies. The APA’s interim remedies provision, Section 705, grants that “the reviewing court . . . may issue all necessary and appropriate process to postpone the effective date of an agency action or to preserve status or rights pending conclusion of the review proceedings.” This grant of judicial discretion to “postpone the effective date of an agency action” combines with the APA’s definition of “agency action,” which includes a “rule” or “order,” to make a challenged agency action itself an object of interim remedies, just as agency action is the object of final remedies.

The operative language makes sense only in terms of a universal interim pause. “[P]ostpone the effective date of an agency action” is most naturally read to mean that the agency action—a rule or order—takes no effect as to anyone anywhere, not that it takes effect as to everyone but the parties to a legal challenge. And reason accords with text: only universal interim remedies match the scope of final relief available, as is necessary to protect parties adequately. Moreover, these universal remedies avoid the practical difficulties of carving individual parties or jurisdictions out of a rule or order applicable elsewhere. Put plainly, since Section 706 creates the universal final remedy of vacatur, it only makes sense that Section 705 would create a universal interim remedy capable of preserving the possibility of a universal final remedy. One would expect this congruence between interim and final remedies.

Moreover, when a court determines that tailored relief is practicable and otherwise appropriate, Section 705 authorizes the court to issue a preliminary injunction tailored as “necessary and appropriate . . . to preserve the status quo or rights pending conclusion of the review proceedings.” The APA presents courts with injunctive relief as an alternative option to postponing a rule’s effective date, which strongly suggests that a judicially postponed agency action is postponed universally, not only so far as necessary to preserve the status quo (whatever that might mean) or the rights of parties.

A final textual hint lies in Section 705’s parallel grant of authority to an agency to “postpone the effective date of action taken by it, pending judicial review.” The SEC recently exercised this power in response to legal challenges to its rule requiring companies to provide climate-related information in their registration statements and annual reports.6 The agency decided to stay the rule on its own initiative pending final judicial review. Just as an agency may postpone its rule or order wholesale, so may a reviewing court. That is the only sensible reading of the APA’s deployment of the same phrase in the same section to describe the interim relief available from agencies and courts.

Again, judicial precedent shows practice agrees with the text. The Supreme Court itself has universally stayed two agency actions pending final merits review in recent years, showing that it understands the APA to authorize interim relief that runs against a rule itself. In 2016, it stayed the EPA’s “Carbon Pollution Emission Guidelines for Existing Stationary Sources: Electric Utility Generating Units.” And in 2022, the Court stayed OSHA’s vaccine mandate. In the latter case, the Court remarked without criticism that the Fifth Circuit had stayed the agency action universally in earlier proceedings. Justice Kavanaugh, joined by Justice Amy Coney Barrett, addressed this practice directly in his concurrence in the Court’s narrowing of a district court’s stay against an Idaho statute. He distinguished universal stays of statutes from stays against “new federal regulations, given the text of the APA.”

The APA’s text and the behavior of courts and agencies confirm that courts may issue universal interim remedies against unlawful agency action.

III. The Separation of Powers Requires the Availability of Universal Judicial Remedies for Unlawful Agency Action

Applying the plain meaning of the APA’s text and history makes sense in the broader separation of powers scheme of the Constitution. The Constitution establishes the federal judiciary as a branch of government coequal with the legislative and executive. Thus, it is unsurprising that no provision of the Constitution authorizes the court to “set aside” the work of Congress when it reviews a statute for conflict with the Constitution. Instead, recall that constitutional review involves what are essentially choice-of-law principles. That is because both the Constitution and any duly enacted statute have the status of “law,” but, as Justice Clarence Thomas has observed, one has a principled and textual right-of-way in any possible collision.

Administrative agencies, by contrast, are not creatures of constitutional creation, but statutory hybrids within the executive branch, exercising delegated powers that are considered mixed quasi-executive, quasi-judicial, and quasi-legislative in nature. And while an “agency action” might look like law because it has legal consequences for regulated parties, it is not “law” in the same sense that the Constitution and federal or state statutes are law. Rather, an agency action is the executive branch’s enforcement of the laws enacted by Congress reduced to a rule or an order. As the Supreme Court has noted, “an agency literally has no power to act . . . unless and until Congress confers power upon it.” Thus no agency’s regulation can “‘operate independently of’ the statute that authorized it.”

Nevertheless, agency rulemaking binds parties with the force of legal consequences without undergoing the compromise-inducing ordeal of passage through both houses of Congress and presentment to the President. This constitutional process restrains exercises of pure legislative power and commands greater judicial respect for legislative commands that survive the process. Maintaining the Constitution’s allocation of powers requires courts to counterbalance this relative lack of front-end checks on agency action with relatively greater judicial review on the back end.

To see why that is so, consider the alternative. If courts could not universally vacate agency action that unlawfully regulates with universal effect, then the Constitution’s allocation of powers would be distorted by executive branch “lawmakers” insufficiently accountable to either congressional or judicial review. And indeed, there is much to be said for the argument that Congress has legislated relatively less as executive agencies have issued more rules.

It is, therefore, logically symmetrical that the APA authorizes courts to issue remedies running against an agency rule itself, even when courts may not issue universal remedies against statutes that are subject to greater front-end checks. Congress simultaneously gave agencies authority to make rules with universal effect and courts commensurate power to prevent the universal injustice of an unlawful rule by issuing universal relief. If the Constitution permits the first move, separation of powers requires, or at least permits, as Justice Byron White suggested, the second move as well. Otherwise, agency action would evade judicial review as to all regulated persons who do not join a successful lawsuit. Such a system would deny full protection of law to those without the resources and wherewithal to challenge unlawful regulation. It would also create a legal patchwork that would undermine the effectiveness of many rules and greatly complicate compliance and enforcement efforts.

Nevertheless, some, including Professor Samuel Bray and Chief Judge Jeffrey Sutton of the Sixth Circuit, have suggested that universal APA remedies exceed the limits of the judicial power that Article III vests in the federal judiciary.7 These critics have argued that federal courts may not hand out remedies “in the abstract” because Article III empowers them only to resolve concrete “Cases” or “Controversies.” A universal remedy might exceed that limit to the extent it goes beyond redressing the injury in the case or controversy presented. Samuel Bray’s work provides historical grounding for this critique by showing that principles of traditional equity did not permit universal injunctions.

Universal remedies under the APA, however, remain within Article III limits because they are legal, not equitable, remedies created by Congress and available only to resolve true cases or controversies. Critics of universal remedies are correct that the separation of powers generally prohibits remedies that reach beyond parties except where equity would have permitted at the Founding. Courts may not invent new equitable remedies that were unknown in England and America before the Founding because that would allow federal judges to unilaterally expand their own power into the legislative realm of general policymaking. Since only the legislature has the power to bind the sovereign people at large, Congress is the proper institutional actor to decide how far judicial remedies can reach before becoming quasi-legislative. These principles make remedies created by Congress (like stay and vacatur of agency action) presumptively constitutional and remedies invented by courts (like universal injunctions) constitutionally suspect.

History explains why remedies created by statute are presumptively lawful “legal” remedies in the fullest sense, whereas court-created remedies are best understood as remedies in “equity” and strictly conscribed. In medieval England, Parliament possessed absolute power to create new causes of action by statute.8 No separation of powers concerns were conceivable because Parliament was both the legislature and the high court. Any expansion of judicial remedies approved by statute was, by definition, lawful. Equity, by contrast, was an extra-legal device by which the King’s appointed Chancellor could, within broad limits, supersede the requirements of law that bound courts when he believed that justice so required. In this sense, the Chancellor could “make law” independent of the courts and Parliament.

When the U.S. Constitution created a separate legislature and judiciary, it also created a potential separation of powers problem by placing the powers of equity in the hands of federal judges. If courts expanded equity to include remedies unknown at the Founding, the federal judiciary would gain a share of legislative power. To avoid potential separation of powers problems, federal courts must respect the traditional limits on equitable remedies that Samuel Bray identified, which did not permit courts to issue sweeping injunctions impacting the ability of a separate branch of government to enforce a statute.9 Statutory remedies, by comparison, present fewer separation of powers concerns because they represent the legislature’s judgment that those remedies do not encroach on legislative power. Thus, broad statutory remedies like those in the APA and Federal Rule of Civil Procedure 23, are not subject to the traditional limits of equity. They simply are not equitable remedies.

Some might argue that universal remedies might nonetheless exceed the judicial power of Article III. If so, even a statute could not authorize universal remedies because Congress may not expand the Constitution’s limits on judicial power. And indeed, according to Supreme Court doctrine, Article III contains some limits on how far Congress can expand judicial remedies. Congress could not, for example, delegate to courts authority to write statutes in the form of judicial opinions addressed to the public at large, or to give government officials legal advice. But judicial review of agency action does not depend on Congress expanding Article III. It comports with Article III’s case or controversy requirement, avoids advisory opinions, and gives courts no quasi-legislative power to create new, generally applicable legal obligations.

The Constitution bars Congress from authorizing courts to declare law independent of a concrete case or controversy because that would commingle legislative and judicial power by allowing courts to “prescribe[ ] the rules by which the duties and rights of every citizen are to be regulated” or to issue advisory opinions. But judicial vacatur of an unlawful agency rule never approaches these boundaries. It is available only if a party “aggrieved by [the] agency action” brings a concrete case, and it never declares new policies, just prevents new policies from taking effect. Nor does it purport to undo any act of Congress. It is inherently judicial power in the sense that it is a strictly negative power, activated only by a concrete dispute, to prevent subordinate actors from transgressing the boundaries of law. In this sense, the judiciary acts as a faithful agent of Congress in reviewing executive branch agencies’ actions. And an opinion vacating a rule is not advisory. The very reason for recent objections to judicial vacatur is that it has too significant an effect. These essential characteristics of stay and vacatur of agency action show that it is precisely the type of power that Article III contemplates.

Some of the anxiety about the universal scope of APA remedies, which gets inaptly expressed as an Article III concern, may be resurfacing doubt about the constitutionality of agency rulemaking. Analyzing APA review through the appellate review model illustrates this point. In that model, a trial court judgment ordinarily affects only parties to the underlying suit. Thus, only those parties are affected when an appellate court invalidates that judgment. There is a correct intuition that the appellate court in this scenario could not act with nationwide effect without exceeding its authority. To reach nationwide results, the appellate court would need to reach beyond the trial-court judgment before it, by purporting to invalidate other similar trial-court judgments or decree binding rules of primary conduct in the abstract.

But the intuition that nationwide effect must exceed rightful judicial authority leads critics astray when a federal court sits in appellate-style review of agency action. Here, the judgment below—the agency rule under review—had a nationwide effect, unlike the party-bound effects of true court judgments. Thus, when a court invalidates an agency rule, that invalidation has a nationwide effect too. That may seem strange, but the strangeness lies in the scope of regulatory authority the “lower court” (the agency) exercised, not in the reviewing court’s routine exercise of judging only the object placed before it. What really triggered the intuition that something illegal is ongoing was the agency action.

This again stands in contrast to statutes that have nationwide effect but came into being through procedures designed to ensure something like national consensus: bicameralism and presentment. Such federal law passes a gauntlet far more daunting than informal (notice-and-comment) rulemaking. So again, to reconstitutionalize Congress’s choice to give an agency subordinate to a unitary executive (the President) the power to make nationwide pronouncements, Congress placed those agencies under the direct appellate-style review of the federal judiciary.  If Congress has the power to adopt statutes that create agencies with nationwide quasi-legislative power, it surely has the power to adopt a statute that cabins such nationwide quasi-legislative power.

Also worth noting is that the injunctive class action mechanism from Federal Rule of Civil Procedure 23—a dramatic expansion of equitable remedies—is likely unconstitutional if Congress may not expand remedies beyond traditional limits. And there is no principled line to be drawn between Rule 23’s expansion of injunctive relief to third parties only before the court in a representative capacity and the APA’s choice to broaden relief by directing remedies at agency action itself, rather than expand the equitable tradition of enjoining the officers charged with carrying out agency action. If anything, Rule 23 is more suspect because it expands courts’ power to bind private actors by injunction, whereas the APA’s universal remedies only create a power to negate agency action that was authorized by statute in the first place.

The APA’s text, historical background, and constitutional principles all demonstrate that courts have correctly concluded that the “ordinary result” for unlawful rules is “that the rules are vacated—not that their application to the individual petitioners is proscribed.”10

IV. Consequences Stemming from the Distinction Between Equitable Remedies and the APA Remedies of Stay and Vacatur

Contemporary debates surrounding universal injunctions and stays of new laws do not apply to APA remedies. It is easy to conflate the APA remedies of stay and vacatur of agency action with the equitable remedies of stay and injunction (preliminary and permanent) against statutes and presidential action. But, the APA’s judicial review provisions create a unique remedies paradigm that is independent of courts’ inherent equitable powers and general grants of statutory authority. This Part differentiates the universal remedies available under the APA from universal injunctions and from stays against state and federal statutes. It shows that concerns about the latter judicial inventions say nothing about the legality or propriety of the APA’s universal remedies.

A. The APA Offers the Universal Remedies of Stay and Vacatur, Which Are Distinct from Preliminary and Permanent Injunctions

The universal remedies that the APA makes available are different in kind from universal injunctions. Courts and litigants have sometimes confused this distinction, especially in the context of interim remedies, by using the terms “preliminary injunction” or “temporary restraining order” interchangeably with “stay.” For example, the Supreme Court recently decided an application for a partial stay of a rule implementing Title IX as though it were a request for a preliminary injunction.  And the Eleventh Circuit has generated still more confusion by purporting to apply the traditional stay factors to applications for injunctions pending appeal, while also heightening the likelihood-of-success prong in apparent recognition of the extraordinary nature of preliminary injunctive relief. The Supreme Court’s jurisprudence helps elucidate the difference between these remedies.

The Supreme Court’s comparison of stays and preliminary injunctions in Nken v. Holder most visibly illustrates the important difference between stays and preliminary injunctions. To receive a stay, the Court explained, a party must show a likelihood of success on the merits, irreparable injury absent a stay, and that the private and public interests favor a stay (though it remains an open question whether the likelihood-of-success prong requires a showing of “certworthiness” for “emergency” stay applications in the Supreme Court). The Court has acknowledged that “[t]here is substantial overlap between these and the factors governing preliminary injunctions,” but has maintained that stays are distinct from preliminary injunctions. “[A] stay operates upon” a “proceeding itself,” the Nken Court noted, “either by halting or postponing some portion of the proceeding, or by temporarily divesting an order of enforceability.” It does not directly bind any person to act or refrain from acting.

The Nken Court wrote in the context of a stay of a judicial proceeding, but the Court confirmed that the Nken standard applies to a request to stay administrative action this term in Ohio v. EPA, as other courts have done in the past. Section 705’s grant of authority for courts to “postpone the effective date of an agency action” makes agency action the “proceeding” against which a stay operates. So, it is no surprise that the Supreme Court has recognized universal stays of agency action as an appropriate interim remedy. By contrast, an injunction, whether preliminary or permanent, “is a judicial process or mandate operating in personam.” It “is directed at someone, [not something,] and governs that party’s conduct.” And because an injunction infringes the enjoined person’s liberty, it must be as narrow as justly possible. As Samuel Bray has observed, historically, injunctions did “not control the defendant’s behavior against nonparties.”11 Thus, a stay must be as broad as the action or proceeding it operates against, whereas the proper scope of an injunction is to prohibit the enjoined party (whether private actor or government official) from taking the unlawful action plaintiffs complained of.

Importantly, that distinction holds as to final APA remedies because vacatur, just like a stay, acts against agency action itself.

B. The APA’s Universal Remedies Do Not Raise Significant Separation of Powers and Federalism Issues

The nature of the action that APA remedies run against is as significant as the nature of the remedies themselves when it comes to understanding why the APA’s universal remedies are lawful. APA remedies run against agency action. And agency action is, at best, a quasi-constitutional chimera of quasi-legislative, quasi-executive, and quasi-judicial power, as the Court recognized in Humphrey’s Executor v. United States, and as Justices Robert Jackson and White have suggested. This hybrid nature carries two important implications for the legality of universal APA remedies.

First, it means that judicial invalidation of agency action does not unbalance the tripartite separation of powers the Constitution establishes. Judicial suspension of or refusal to apply federal legislation always results in a clash of the coordinate departments of government since a statute (unlike an agency action) is the purely legislative act of a coequal branch of government. Universal remedies against enforcement of federal statutes interfere with what could reasonably be viewed as the Constitution’s main focus—Congressional action. They tend toward power imbalance by shifting power from Congress to the courts. On the other hand, the APA’s universal remedies reduce the power of governmental actors that are unknown to the Constitution. They serve to return the allocation of federal power closer to the Constitution’s equilibrium point. They do this by using an enactment of Congress (the APA) to review and sometimes negative the actions of creatures of Congress (executive branch agencies). Universal APA remedies thus benefit separation of powers principles. In contrast, separation of power principles suffer when these universal remedies touch on federal statutes.

Second, the federal origin of agency action meaningfully distinguishes it from state laws that parties challenge in federal court. The APA directs review only of federal government action, so remedies that wholly incapacitate that action do no violence to federalism. When federal courts issue stays against enforcement of state laws, however, they irreparably infringe the States’ retained lawmaking power, harming the Constitution’s vertical balance. So, the federalism reasons to avoid federal court-issued remedies against state laws have no purchase when it comes to the APA’s universal remedies.

C. The APA Does Not Authorize Courts to Enjoin Presidential Action

Equally important is that APA universal remedies face none of the problems unique to universal injunctions directed against presidential action. The APA authorizes universal remedies only against “agency action,” which does not include presidential action. In fact, the APA does not authorize judicial review of presidential action at all. A different framework of constitutional review applies instead. So, the legal arguments against universal injunctions of presidential action do no harm to universal remedies against agency action under the APA.

The primary argument against universal injunctions of presidential action is that no source of law authorizes them. No statute grants district courts general power to issue universal injunctions, nor does the courts’ inherent constitutional authority include any such power. Thus, courts lack general power to enjoin executive action. But courts do enjoy express statutory authorization to issue stays and vacatur of agency action under the APA. And this Article has already explained that these remedies are consistent with the separation of powers and Article III limits that might prevent Congress from authorizing universal injunctions against presidential action by statute. As elsewhere in the universal remedies debate, APA remedies stand above the fray. It should be little wonder, then, that courts have universally accepted without question that the APA provides universal remedies against agency action.

V. Conclusion

The APA is best understood as making universal remedies the default relief for unlawful agency action. The legal profession should not allow separate questions about universal injunctions to unsettle this consensus. APA remedies do not face the most serious legal problems that federal courts create when they enjoin enforcement of statutes beyond their jurisdiction (or, in the case of state laws, created by a separate system of government). If courts, lawyers, and scholars want to debate the legality of both universal remedies against agency action and universal injunctions, they must have two separate debates. This Article has focused on the APA debate to emphasize that that debate is largely settled, and rightly settled too.

* * *

T. Elliot Gaiser is the Solicitor General of Ohio.

Mathura J. Sridharan is the Director of Ohio’s Tenth Amendment Center and serves as a Deputy Solicitor General in the Ohio Attorney General’s Office.

Nicholas A. Cordova is an associate at Boyden Gray PLLC and former Simon Karas Fellow to the Ohio Solicitor General.

  • 1E.g., Harmon v. Thornburgh, 878 F.2d 484, 495 n.21 (D.C. Cir. 1989); Nat’l Org. of Veterans’ Advocs., Inc. v. Sec’y of Veterans Affs., 48 F.4th 1307, 1317 (Fed. Cir. 2022); N.H. Hosp. Ass’n v. Azar, 887 F.3d 62, 77 (1st Cir. 2018); Nat’l Black Media Coal. v. FCC, 791 F.2d 1016, 1020, 1024 (2d Cir. 1986); Prometheus Radio Project v. FCC, 652 F.3d 431, 453–54, 453 n.25 (3d Cir. 2011); N.C. Growers’ Ass’n v. United Farm Workers, 702 F.3d 755, 759 (4th Cir. 2012); Chamber of Com. of the U.S. v. Dep’t of Labor, 885 F.3d 360, 363 (5th Cir. 2018); Mason Gen. Hosp. v. Sec’y of Dep’t of Health & Hum. Servs., 809 F.2d 1220, 1231 (6th Cir. 1987); H & H Tire Co. v. Dep’t of Transp., 471 F.2d 350, 355–56 (7th Cir. 1972); Menorah Med. Ctr. v. Heckler, 768 F.2d 292, 297 (8th Cir. 1985); Nat. Res. Def. Council v. U.S. EPA, 526 F.3d 591, 594 (9th Cir. 2008); Zen Magnets, LLC v. Consumer Prod. Safety Comm’n, 841 F.3d 1141, 1155 (10th Cir. 2016); Alabama v. Ctrs. for Medicare & Medicaid Servs., 674 F.3d 1241, 1244 (11th Cir. 2012).
  • 2Reuel E. Schiller, Rulemaking's Promise: Administrative Law and Legal Culture in the 1960s and 1970s, 53 Admin. L. Rev. 1139, 1145 (2001).
  • 3Id. at 1145–46.
  • 4See Samuel L. Bray, Multiple Chancellors: Reforming the National Injunction, 131 Harv. L. Rev. 417, 425–45, 438 n.121 (2017).
  • 5See, e.g., Ohio v. EPA, 144 S. Ct. 2040, 2058 (2024) (granting stay application for stay of ozone transport regulations); Garland v. Blackhawk Mfg. Grp., 144 S. Ct. 338 (2023) (challenging the regulation of gun parts as “firearms”); Biden v. Nebraska, 143 S. Ct. 2355 (2023) (rejecting student loan nullification); United States v. Texas, 143 S. Ct. 1964, 1980–86 (2023) (Gorsuch, J., concurring in judgment) (criticizing universal remedies in a challenge to immigration guidelines); Nat’l Fed’n of Indep. Bus. v. Dep’t of Lab, 142 S. Ct. 661 (2022) (addressing the COVID-19 vaccine mandate). Still more high-profile litigation reaches the lower courts in an emergency posture. See, e.g., The Enhancement and Standardization of Climate-Related Disclosures for Investors; Delay of Effective Date, 89 Fed. Reg. 25,804 (April 12, 2024) [hereinafter Climate-Related Disclosures; Delay of Effective Date] (staying rule requiring registrants to provide certain climate-related information in registration statements and annual reports in response to Eighth Circuit litigation seeking stay); Texas v. EPA, 2024 WL 3384818, at *1 (D.C. Cir. July 9, 2024); see also W.V. by & through Morrisey v. U.S. Dep’t of Treasury, 59 F.4th 1124 (11th Cir. 2023) (granting a permanent injunction against a Treasury rule); Texas v. EPA, 662 F. Supp. 3d 739 (S.D. Tex. 2023), appeal dismissed, 2023 WL 8295928 (5th Cir. Oct. 6, 2023); Kentucky v. Fed. Highway Admin., 2024 WL 1402443, at *1 (W.D. Ky. Apr. 1, 2024).
  • 6Climate-Related Disclosures; Delay of Effective Date, 89 Fed. Reg. at 25,804.
  • 7See Bray, supra note 4, at 433, 471–72.
  • 8See John Baker, An Introduction to English Legal History 221, 354 (2019).
  • 9See Bray, supra note 4, at 420–21, 425–28.
  • 10See also Data Mktg. P’ship, LP v. Dep’t of Lab., 45 F.4th 846, 859–60 (5th Cir. 2022); Regents of the Univ. of Cal. v. U.S. Dep’t of Homeland Sec., 908 F.3d 476, 511 (9th Cir. 2018), rev’d on other grounds, vacated in part sub nom. Dep’t of Homeland Sec. v. Regents of the Univ. of Cal., 140 S. Ct. 1891 (2020); Black Warrior Riverkeeper, Inc. v. U.S. Army Corps of Eng’rs, 781 F.3d 1271, 1290 (11th Cir. 2015).
  • 11Bray, supra note 4, at 421.