So-Called “Administrative Stays” in Trump 2.0

Essay - by Online Edition - Volume 104

Introduction

The first few scenes of the Trump presidency sequel have been action-packed. The White House’s news-getting activity has triggered similarly newsworthy happenings in the federal courts. Lower courts have put temporary halts on executive actions relating to DEI programs,[1] birthright citizenship,[2] federal funding,[3] and the firing of federal officers,[4] just to name a few.

In stopping a number of the new administration’s initiatives, these courts have relied on a potent new tool: the so-called “administrative stay.” Although administrative stays of lower court orders are familiar, before 2022, it seems there was not a single clear example of an “administrative stay” of executive action.[5] Recently, such “administrative stays” have proliferated. Courts have issued them to stop executive action from the get-go, even before issuing more traditional stopgap relief like temporary restraining orders (TROs), preliminary injunctions, or (for agency action) stays pending appeal. And they have done so in cases in which no party has requested an administrative stay or any stay at all.

For example, the United States District Court for the District of Columbia issued an “administrative stay” temporarily blocking President Trump’s removal of Hampton Dellinger from his position as Special Counsel of the Office of Special Counsel even though the parties had requested only a TRO and preliminary injunction.[6] That same lower court also administratively stayed an internal Office of Management and Budget memorandum ordering a temporary stop on the disbursement of federal funds.[7] And the United States District Court for the Eastern District of Virginia issued an administrative stay that temporarily prevented the Director of National Intelligence from firing certain intelligence officers.[8]

An administrative stay of executive action is a novel and extraordinary exercise of judicial power. It halts executive action by “setting aside” the underlying legal basis for it.[9] And critically, unlike a TRO or a preliminary injunction, both of which require the plaintiff to clear a high bar before persuading the court to halt executive action,[10] administrative stays require no analysis of traditional considerations, like irreparable injury or the merits.[11] What administrative stays do require is hard to say. The most influential judicial opinion to ever address the issue explained that “there is no jurisprudence of administrative stays.”[12] Instead, courts may issue administrative stays without applying any test at all.[13]

So when issuing these administrative stays of executive action, the district courts have not engaged in any of the searching analysis that normally accompanies TROs and preliminary injunctions. Indeed, these lower courts have not applied any discernible legal test at all. And they have issued this relief on their own initiative, simply as a means to “freeze[]” the “legal proceedings” and buy time to “rule on” the parties’ actual “request[s] for expedited relief,” like plaintiffs’ requests for TROs and preliminary injunctions.[14]

What could possibly be the legal basis for issuing such extraordinary and novel relief to freeze executive action without the request of any party and without the application of any test? So far, the courts have relied on their inherent authority and the All Writs Act.[15]

But these so-called “administrative stays” of executive action are conceptually confused and rest on a doubtful legal basis.

To be fair, it is not surprising that courts might commit errors in issuing emergency relief. Emergency relief has long been an understudied area of federal court practice.[16] Perhaps the “most important and influential casebook ever written,”[17] Hart & Wechsler’s The Federal Courts and the Federal System, only recently added any extended discussion of the emergency docket.[18] Naturally, then, the jurisprudence of emergency relief remains arcane and confused.

Still, as Justice Kavanaugh recently urged, “[i]t is critical to appreciate the significance of the decision that” the courts are “being asked to make in emergency cases” involving government action.[19] By issuing emergency relief, the court temporarily blocks consequential and contentious action taken by a democratically accountable branch of government. Although the relief issued is only temporary, even the specific form of emergency relief analyzed here, an administrative stay of executive action, may have the practical effect of putting government action on ice for months at a time.[20] Thus, the decision whether to issue emergency relief is often the whole ballgame.[21]

In critiquing district courts’ recent “administrative stays” of executive action, then, this Essay offers critical guidance for navigating the murky waters of the federal courts’ emergency dockets.

Part I provides background on administrative stays, stays, TROs, and preliminary injunctions.

Part II argues that several recent orders labeled by the district courts as administrative stays are not actually stays at all, administrative or otherwise. They are injunctions. Since these injunctions have involved none of the traditional TRO or preliminary-injunction considerations, they cannot be TROs or preliminary injunctions. Instead, they are better thought of as “administrative injunctions.”

To be sure, I do not analyze each “administrative stay” of executive action any court has issued to determine whether it is an injunction in disguise. I use Dellinger v. Bessent as an example.[22] My analysis of Dellinger, though, should provide a roadmap for distinguishing so-called “administrative stays” from true administrative stays. So it should illuminate much of the litigation that either has already occurred or is yet to come.

Part III then considers each possible legal basis for issuing these injunctions mislabeled as administrative stays: the All Writs Act, inherent authority, 28 U.S.C. § 1331, and § 705 of the Administrative Procedure Act (APA). None appears to authorize this newfangled form of relief.

Part IV turns to the more complex issue that could arise soon: the lawfulness of potential orders that might rightly be thought of as administrative stays. Here, there are three possible sources of authority: the All Writs Act, inherent authority, and § 705 of the APA. Although I do not draw firm conclusions in this Part, I provide guidance for courts and commentators in thinking through the propriety of true administrative stays of executive action. And I suggest that under current doctrine, an administrative stay of action taken by an agency acting as a minicourt (an agency issuing an order, not a rule) might be available under courts’ inherent powers or the All Writs Act. But in any other context, an administrative stay of executive action is probably not available.

 I. Emergency Relief: The Fundamental Forms and Concepts.

As relevant here, there are four standard forms of emergency relief: stays, preliminary injunctions, administrative stays, and TROs. All are temporary. But stays and preliminary injunctions are longer term forms of relief than administrative stays and TROs.

In this Part, I begin by laying out the basic distinction between stays and preliminary injunctions. Then, I explain the basic distinction between administrative stays and TROs, including their relationships to stays and preliminary injunctions, respectively.

 A. Stays and Preliminary Injunctions: The Basic Distinction.

Stays and preliminary injunctions have “some functional overlap,” but they are formally distinct.[23]

Functionally speaking, “[b]oth can have the practical effect of preventing some action before the legality of that action has been conclusively determined.”[24] In other words, both “halt” some action.[25] Both do so only pending the final decision on the merits. And both require application of demanding four-part tests,[26] tests which have “substantial overlap.”[27]

But stays and preliminary injunctions take different forms. An injunction “is a means by which a court tells someone what to do or not to do.”[28] Stays, by contrast, operate on the underlying legal authority claimed as a basis for the action at issue. As the Supreme Court put it in Nken v. Holder, “a stay” halts action by “temporarily suspending the source of authority to act” rather than “by directing an actor’s conduct.”[29] When an appellate court stays a lower court’s order, for example, the stay “operates upon the judicial proceeding itself . . . by temporarily divesting [the] order of enforceability.”[30] When a reviewing court stays an agency rule, the stay operates upon the rule itself, by temporarily divesting it of enforceability.[31] In other words, unlike a preliminary injunction “which merely thwarts the enforcement of” a legal rule, a stay “suspend[s]” the rule “or delay[s] its effective start date.”[32]

To illustrate the distinction, consider the permanent versions of these preliminary forms of relief. A preliminary injunction is a temporary version of a permanent injunction, temporary in the sense that “it does not persist after the judgment.”[33] A stay is also temporary in that same sense. But it is a temporary version of vacatur.[34] Like a stay, “[v]acatur operates on the legal status of a” given legal authority, such as a rule or an order, “causing” that legal authority “to lose binding force.”[35] Or as Nken put it—employing the language of the APA supposedly authorizing vacatur[36]—a stay “set[s] aside” a given legal “authority.”[37] It just does so “temporar[ily].”[38]

With those formal distinctions in mind, one can see how the two forms of relief accomplish similar ends, even though they do so in different ways. The injunction blocks executive action by ordering executive officials not to take the action. The stay blocks executive action by “temporar[ily] setting aside” the underlying “source of the Government’s authority to” act.[39]

 B. Administrative Stays and TROs.

While administrative stays are stopgap versions of stays, TROs are stopgap versions of preliminary injunctions.

More specifically, administrative stays “freeze legal proceedings until the court can rule on a party’s request for expedited relief.”[40] As Justice Barrett has explained, the normal four-part Nken test for stays, involving considerations like likelihood of success on the merits and irreparable injury, is “not always easy” to undertake “in haste, and an administrative stay buys the court time to deliberate.”[41] So if anything, issuing an administrative stay just “reflects a first-blush judgment about the relative consequences” of the case.[42] Of course, “[b]ecause an administrative stay precedes a ruling on a stay pending appeal, the Nken factors” may well be “on the court’s radar.”[43] But at its core, the administrative stay is “a flexible, short-term tool” that “courts apply” without applying any test at all.[44]

Just like preliminary injunctions, TROs temporarily order a defendant to do or not do something. TROs operate only “until there is an opportunity to hold a hearing on the application for a preliminary injunction.”[45] To obtain a TRO, unlike an administrative stay, applicants need to clear the same hurdles as when obtaining a preliminary injunction.[46] And once a TRO’s brief lifespan expires, a plaintiff must obtain a preliminary injunction to continue putting a stop to the defendant’s action.[47]

So in sum, just as TROs are to preliminary injunctions, administrative stays are to stays. Both are even shorter term versions of their counterparts. But there is a key difference between TROs and administrative stays. The party moving for a TRO must clear a demanding test before a court will grant a preliminary injunction. But courts may issue administrative stays without applying any test at all, including the test for granting an ordinary stay.[48]

 II. Properly Categorizing So-Called “Administrative Stays”: The Dellinger Example.

With these formal distinctions in mind, one can see why some of the Trump cases do not really involve administrative “stays” at all. Of course, the courts may continue calling them “stays” all they want. But that label does not control.[49]

Consider the recent “administrative stay” in Dellinger v. Bessent.[50] There, President Trump ordered the removal of Hampton Dellinger, the single head of the Office of Special Counsel. Dellinger sued and sought a TRO. Before granting that TRO, the district court issued a “brief administrative stay” to “preserve the status quo.”[51]

Issuing a “stay”—administrative or otherwise—makes no sense in a case like Dellinger. Recall that “a stay . . . temporarily suspend[s] the source of authority to act.”[52] It does not “direct[]” any “actor’s conduct.”[53] So what was the “source of authority” for the action taken in Dellinger—that is, President Trump’s removal of Dellinger—upon which the district court’s administrative stay might act? Presumably, the Vesting and Take Care Clauses of Article II.[54] So by issuing a stay, apparently the District Court for the District of Columbia “temporarily suspend[ed]” Article II itself. But that is obviously absurd and impossible.

Yet this Article II suspension theory seems to be the most plausible way to conceive of how an administrative stay would operate in this context. Consider an analogy to the removal of an illegal alien. Those aliens often seek stays of removal. As the Court explained in Nken, “[a]n alien seeking a stay of removal” seeks “the temporary setting aside of the source of the Government’s authority to remove.”[55] In that context, the authority to remove is generally the immigration judge’s order. In this context, though, the authority to remove is Article II of the Constitution. Since a true “stay” operates on the underlying “authority to remove,” it would operate here on Article II of the Constitution itself.

To be sure, one might instead conceive of the President’s firing decision like an informal adjudication. Thus, the President issues an “order” that is sufficiently like a lower court judgment such that the court directly reviews and acts upon the order itself, just like it acts upon the order of removal by an immigration judge.

That, though, makes little conceptual sense. There is no such direct review of presidential action.[56] So no one in Dellinger even thought to ask the district court to “vacate” the President’s firing decision as the remedy at the end of the litigation. But as explained above, a stay is just like short-term vacatur. So review of President Trump’s decision to fire Dellinger is unlike a court’s review of an immigration judge’s order. There, the court may directly review and act upon the order of the immigration judge itself, so the remedy sought results in the actual wiping away of the order altogether.[57]

Anyway, it is hard to see what a true administrative stay would even accomplish. If the district court merely stayed President Trump’s singular decision to fire Dellinger, that might have the effect of re-installing Dellinger. It would be as if Dellinger were never fired. But that would presumably not stop President Trump from just firing Dellinger again since the stay would not have “direct[ed]” President Trump’s “conduct . . . with the backing of [the court’s] full coercive powers.”[58] Indeed, President Trump would appear to be under no order whatsoever not to fire Mr. Dellinger a second time. And since administrative stays entail no analysis of the merits, nor did the district court engage in any, even committed defenders of judicial supremacy should find no conflict with the court’s nonexistent merits analysis.

So what the district court is really aiming to do in a case like Dellinger is issue an injunction. The court is seeking to “direct[]” President Trump’s “conduct,”[59] barring him from removing Dellinger, while it proceeds to consider the merits of the case (and the merits of pending emergency motions).

That much is obvious once one looks at the next stages of the litigation. The administrative stay in Dellinger operated only as a precursor to a TRO and a preliminary injunction. But administrative stays are supposed to be precursors to stays, not these other distinct forms of relief.[60] As Justice Barrett has explained, “an administrative stay precedes a ruling on a stay pending appeal.”[61] Consider the core example of a stay: a stay of a district court’s order. It would be odd to ask for a stay and a preliminary injunction. Often, the very thing to stay is the preliminary injunction entered by the district court. The fact that a district court will proceed to consider a TRO, a preliminary injunction, and ultimately a permanent injunction after first issuing an “administrative stay” of executive action helps reveal that the relevant relief is an order by the court “tell[ing] someone what to do or not to do.”[62] That is an injunction.

Since these injunctions, though, have involved none of the traditional TRO or preliminary injunction factors, they cannot be TROs or preliminary injunctions. Instead, they are better thought of as “administrative injunctions.”

 III. The Lack of Authority to Issue “Administrative Injunctions.”

The concept of an “administrative injunction,” though, might seem strange. A super-short-term preliminary injunction already exists: namely, the TRO. That alone casts serious doubt on the legality of administrative injunctions. Ordinarily, the availability of one form of relief forecloses others.[63] And it would be big news if there were a way to get a de facto TRO (in the form of an administrative injunction) without having to clear the normal, demanding test.

Still, the administrative injunction has begun to proliferate in the lower courts.[64] And applicants have begun to actively request “administrative injunction[s]” in high-profile emergency-relief cases.[65]

There is no discernible legal basis, though, for issuing administrative injunctions, which involve none of the traditional preliminary-injunction considerations. To be sure, I do not analyze every potential legal basis for preliminary relief in any hypothetical case involving executive action. I analyze the four potential bases that would apply across a host of cases. None, I conclude, is promising.[66] Moreover, my analysis does suggest that barring specific language authorizing courts to issue injunctions without considering the four traditional factors enumerated in Winter v. Natural Resources Defense Council, Inc.,[67] no statute should be read to authorize such relief.

I begin by assessing the two bases that the district courts have routinely invoked in issuing so-called “administrative stays” of executive action: the All Writs Act and the federal courts’ inherent authority to manage their dockets.[68] Neither source of law can bear the extraordinary weight that the district courts are putting on it as of late. Next, I consider whether 28 U.S.C. § 1331 offers a basis for administering an administrative injunction. It is unlikely. Erie and its progeny cast serious doubt on whether § 1331 authorizes remedies. To the extent that § 1331 does offer a basis, it authorizes courts only to administer equitable remedies which can trace their lineage to the Founding. The test-less administrative injunction would not seem to qualify. Finally, I analyze § 705 of the APA. Section 705 appears to be the least likely basis of all. It will often not apply in these cases. Even when it does, its text, especially interpreted in light of recent Supreme Court precedent, would seem to authorize only ordinary preliminary injunctions requiring analysis of the four Winter factors.

 A. The All Writs Act.

The district courts have pointed to the All Writs Act, 28 U.S.C. § 1651, when issuing administrative injunctions to halt executive action.[69] On the books since the Judiciary Act of 1789, the All Writs Act provides that federal courts “may issue all writs necessary or appropriate in aid of their respective jurisdictions and agreeable to the usages and principles of law.”[70] The All Writs Act serves as a critical source of legal authority for stays of lower court orders[71] and preliminary injunctions.[72]

The All Writs Act is an unlikely basis for issuing injunctive relief without consideration of the traditional factors. The All Writs Act does not authorize entirely novel forms of equitable relief. The Act only ensures that federal courts can leverage “familiar procedures”[73] and “historic aids”[74] in exercising their powers. But these injunctions appear novel. The most that the district courts have mustered by way of precedent when issuing these injunctions are 2022 and 2024 district court orders from the District Court for the District of Columbia and the Eastern District of Texas.[75]

Instead, if history provides guidance in determining whether courts may issue injunctive relief, the Supreme Court’s read of that history counsels in favor of the four-factor Winter test: Per the Supreme Court, the Winter test “reflect[s] a ‘practice with a background of several hundred years of history.’”[76] Thus, it is that test which is generally “applicable to cases in which injunctions are sought in the federal courts.”[77]

 B. Inherent Authority.

The district courts have also invoked their “inherent authority to manage [their] docket[s]” when issuing administrative stays of executive action.[78] When these administrative “stays” are understood for what they really are—administrative injunctions—it is unlikely that a court’s inherent authority to manage its docket authorizes them.

Although the federal courts possess inherent powers,[79] those powers are historically limited.[80] So this potential basis for issuing administrative injunctions fares no better than the All Writs Act. If anything, it fares worse. As explained above, there appears to be no historical basis whatsoever for issuing preliminary injunctions without analysis of any of the typical preliminary-injunction factors. So there seems to be little reason to think that power is somehow implicit either in the grant of “judicial Power” to Article III courts or in the structure of the Constitution.[81] Furthermore, the Supreme Court has warned that lower courts must exercise their inherent powers with “great caution,”[82] “restraint and discretion,”[83] and “care.”[84]

To be sure, “[p]rior cases have outlined the scope of the inherent power of the federal courts,”[85] and those cases have held that federal courts have inherent authority “to manage their own affairs so as to achieve the orderly and expeditious disposition of cases.”[86] But that language cannot be read to authorize administrative injunctions. The authority of courts to manage their own affairs has been deemed to permit measures like imposing courtroom decorum rules and sanctioning conduct that “abuses the judicial process.”[87] It has never been held, as far as I’m aware, to permit enjoining executive action outside the courtroom on a temporary basis without consideration of the normal equitable factors such as likelihood of success on the merits and irreparable injury.

 C. 28 U.S.C. § 1331.

Another possible basis for administrative injunctions is 28 U.S.C. § 1331, the grant of federal question jurisdiction to the federal courts. To be sure, the federal courts have hesitated recently to find authority under § 1331 to issue equitable relief.[88] And they have repeatedly rejected the provision as a source of other remedies,[89] or basically any other law.[90] After all, as the late federal courts scholar Daniel J. Meltzer put it, “the lesson of Erie” is that “grants of jurisdiction alone are not themselves grants of lawmaking authority.”[91] Still, the Court has treated equitable remedies as unique for unclear reasons.[92] And Grupo Mexicano de Desarrollo, S.A. v. Alliance Bond Fund, Inc.[93] treated the general jurisdictional statutes as a basis for administering equitable remedies, and it did so in determining whether a certain type of preliminary injunction was available.[94]

But general jurisdictional statutes are an unlikely basis for issuing an injunction without considering the traditional factors.

Grupo itself explained that the inquiry under general jurisdictional statutes is “the very inquiry” under the All Writs Act.[95] So if the relief is not available under the All Writs Act, it is not available under a general jurisdictional statute like § 1331.

Anyway, under Grupo, federal courts’ power to grant equitable relief under a general jurisdictional statute is limited to those remedies “traditionally accorded by courts of equity.”[96] Specifically, equitable relief is available only if it can trace its lineage to the High Court of Chancery in 1789.[97] As the Supreme Court put it, “the equitable powers conferred by the Judiciary Act of 1789 did not include the power to create remedies previously unknown to equity jurisprudence.”[98] So to the extent the general jurisdictional statutes are read to authorize courts to administer equitable remedies, the Supreme Court has seemed to limit the courts’ authority to administering those remedies available at the Founding, probably in part because of the obvious tension with Erie. Indeed, when the Court has recognized authority to administer remedies under jurisdictional statutes, it has generally cabined the available remedies to those available at the Founding[99] or those otherwise enshrined in existent Supreme Court precedent,[100] at least in part because of Erie.[101]

Under that test, the administrative injunction potentially fares even worse under general jurisdictional statutes than it does under the All Writs Act. Again, there is no apparent Founding-era basis for recognizing the administrative injunction. It appears no older than a few years old.[102]

To be fair, Grupo may leave open the possibility that courts may issue remedies that the High Court of Chancery itself could not issue so long as issuing that remedy does not amount to a “substantial expansion of past practice.”[103] But given the Supreme Court’s consistent warnings that even preliminary injunctions issued to applicants who have cleared the four demanding hurdles imposed by Winter are “‘extraordinary’ equitable remed[ies] . . . ‘never awarded as of right,’”[104] it is hard to see how a preliminary injunction issued without any analysis whatsoever would not be such an extraordinary remedy so as to amount to a “substantial expansion of past practice.”

 D. The Administrative Procedure Act.

Another possible source of authority for issuing administrative injunctions, though not cited by any of the district courts to date, is § 705 of the Administrative Procedure Act. But although possible, § 705 is not a promising source of authority for issuing administrative injunctions.

First, in many cases the APA will not even be on the table. For instance, in a case like Dellinger, the action ultimately to be enjoined would seem to be the President’s termination of Dellinger. But the APA does not apply to presidential action.[105] And the APA might not be available in other suits, too, for the simple reason that it does not apply to much executive branch action.[106]

Even when the APA is on the table, it is unlikely to authorize an administrative injunction for at least three reasons.

First, § 705’s own text permits issuance of a preliminary injunction only “as may be required and to the extent necessary to prevent irreparable injury.”[107] But an administrative injunction of the type courts have been issuing does not require irreparable injury.

Second, § 705 does not authorize novel remedies, and the administrative injunction appears novel. As Professor Louis Jaffe, in many ways the godfather of modern administrative law,[108] explained decades ago, § 705 is tied to “the power granted under the All Writs statute.”[109] Indeed, § 705 did not seek to “change existing law” concerning judicial review of agency action.[110] So § 705 does not authorize administrative injunctions for the same reasons the All Writs Act doesn’t: The administrative injunction appears to be a wholly unheard-of remedy, both at the time of the adoption of the All Writs Act and the APA.[111]

Third, Supreme Court precedent makes clear that § 705 should not be read to authorize administrative injunctions. Just this past term in Starbucks Corp. v. McKinney,[112] the Supreme Court held that “[w]hen interpreting a statute that authorizes federal courts to grant preliminary injunctions,” a court should “not lightly assume that Congress has intended to depart from established principles.”[113] The Supreme Court “has consistently employed this presumption when interpreting a wide variety of statutes that authorize preliminary and permanent injunctions.”[114] Those established principles require courts to “adhere to the traditional four-factor test” “absent a clear command from Congress” to the contrary.[115]

Nothing in the APA amounts to a “clear command” that courts may issue injunctions without application of the traditional four-factor test. On the contrary, as just noted, § 705 itself refers to the “irreparable injury” consideration.[116] And it may refer to the other considerations by authorizing only “necessary and appropriate process” issued upon “such conditions as may be required.”[117] Moreover, § 702 reaffirms that the APA does not “affect[]” the “duty of the court to . . . deny relief on any other appropriate . . . equitable ground.”[118] Thus, rather than authorize equitable relief absent consideration of the traditional factors, the APA seems to require such consideration.

To see this, contrast the language in the APA with the language discussed by the Supreme Court in Starbucks Corp. In Starbucks Corp., the Supreme Court was tasked with interpreting a statute that authorized injunctive relief if the court “deem[ed]” it “just and proper.”[119] Not even that language, the Supreme Court held, amounted to the requisite “clear command” displacing the traditional four-factor test.[120] Indeed, the Court emphasized, it had previously held in the 1944 case of Hecht Co. v. Bowles[121] that “the Emergency Price Control Act of 1942’s instruction that an injunction ‘shall be granted’” upon a showing that a defendant acted unlawfully was insufficient to jettison the traditional test.[122] If the language in those statutes “did not displace the presumption that traditional equitable principles apply,” then the APA’s text, passed just two years after the Supreme Court’s decision in Hecht, “does not either.”[123]

 IV. Analyzing True Administrative Stays of Executive Action.

In this final Part, I look ahead to potential litigation that might arise in which courts will issue true administrative stays of executive action. Although space and time do not permit me to analyze in full the propriety of an administrative stay as to each possible type of action the executive branch might take,[124] I do offer guidance. Specifically, I show how the analysis from previous Parts can guide courts in trying to determine the legal basis for issuing true administrative stays of executive action. Moreover, I suggest that true administrative stays of executive action, if permitted at all, are probably permitted only in the context of agency adjudication.

The analysis in the previous Parts suggests that administrative stays of executive action might be grounded in three general sources of authority: inherent powers, the All Writs Act, and the APA.[125] I take the first two together before turning to the APA.

 A. Inherent Powers and the All Writs Act.

Start with inherent powers and the All Writs Act. As noted above, these authorities demand a historical basis. But as far as I’m aware, there is no historical basis for federal courts “staying” executive action,[126] especially without any analysis of the traditional equitable factors. As noted, the earliest apparent example of an administrative stay of executive action comes from 2022.

That said, Nken v. Holder and Scripps-Howard Radio, Inc. v. FCC[127] open the door to the possibility of stays of agency orders. In Nken, the Court assessed whether the Illegal Immigration Reform and Immigrant Responsibility Act of 1996’s restrictive standards for injunctive relief applied to requests to “stay” immigration judge removal orders.[128] The Court concluded that it did not; instead, the traditional stay test governed.[129]

To reach that conclusion, Nken conceived of the relationship between the immigration judge’s order and a reviewing court as akin to the relationship between a lower court’s order and an appellate court.[130] Indeed, Nken grounded the power to stay the immigration judge’s order in the traditional authority of “appellate courts” to stay lower court orders[131] both under the All Writs Act and the courts’ inherent powers.[132]

In likening agency orders to lower court orders, Nken echoed the Supreme Court’s decision in Scripps-Howard Radio, Inc. v. FCC. There, the Court reasoned that federal courts’ “traditional equipment for the administration of justice,” like “stay[ing] the enforcement of a judgment pending the outcome of an appeal,” carried over into the context of agency adjudication.[133] Just as an appellate court can stay a lower court order, a reviewing court can stay an agency order.

Read for all their worth, Nken and Scripps-Howard could cut in favor of concluding that federal courts can administratively stay agency orders. Just as reviewing courts’ power to stay lower court orders carries over to agency orders, perhaps reviewing courts’ power to administratively stay lower court orders carries over to agency orders.

But it’s not clear cut. After all, in linking review of agency orders with review of lower court orders, the Scripps-Howard Court was quite explicit that it had nothing to say about the proper standard for staying agency action: “We merely recognize the existence of the power to grant a stay. We are not concerned here with the criteria which should govern the Court in exercising that power.”[134] But the standard is the critical aspect of administrative stays: Their key feature is that they don’t have a test. And later cases have read Scripps-Howard as carrying only “the traditional stay” into the agency context.[135] Nken also only endorsed the “traditional” stay test for stays of agency orders. Nken wasn’t confronted with, nor did it say anything about, standardless administrative stays. But just because a source of law authorizes one kind of stay doesn’t mean it authorizes another.[136] And since administrative stays of executive action do not themselves appear “traditional,” neither Nken nor Scripps-Howard would seem to embrace them.

Regardless, even if these cases are read to provide support for the availability of administrative stays of agency orders, they don’t provide support for administrative stays of other types of executive action, such as agency rules. Recall that both Scripps-Howard and Nken proceeded upon the assumption that the relationship of reviewing courts to agencies was like the relationship of appellate courts to lower courts. But those cases were operating in a context of agencies issuing discrete orders to particular individuals that were in turn reviewed by courts. And it was only in that unique context—the context in which agencies act as “minicourts,” to borrow Adrian Vermeule’s nomenclature[137]—that the Supreme Court applied reviewing courts’ traditional powers to stay lower court orders to the agency actions under review. In other contexts, however, the analogy doesn’t hold up. Outside of the context of agency adjudication, there would not appear to be anything akin to a lower court order being reviewed on appeal.[138]

 B. The APA.

The APA itself provides no more promising basis.

The argument in favor of the APA providing a basis for issuing administrative stays of agency action might go like this. As Professor Jaffe argued, § 705 “relates the power granted under the All Writs statute to the review of administrative orders.”[139] In other words, § 705 makes the remedies available under the All Writs Act available when courts review “administrative orders.” Since the All Writs Act contemplates administrative stays of lower court orders, § 705 contemplates administrative stays of agency orders.

The invocation of agency “orders” might suggest that if courts may issue administrative stays, it is only in the context of reviewing agency orders.[140] Indeed, as noted above, Nken and Scripps-Howard might be read to support the availability of such relief under the All Writs Act itself although it is unclear.[141] That would be dispositive at least for Jaffe himself since he meant by his comment only that § 705 authorized the precise remedies already available under the All Writs Act.[142]

Regardless, it is doubtful administrative stays may issue under the APA at all—whether in the context of agency adjudication or rulemaking. The critical tool for determining what remedies the APA authorizes is the APA’s text. And the APA’s text seems to take the extraordinary remedy of administrative stays off the table.

Section 705 permits reviewing courts to issue stays only “to the extent necessary to prevent irreparable injury.”[143] This makes “irreparable injury” an “indispensable condition” to issuing stays under § 705, as the influential[144] Attorney General’s Manual issued shortly after the passage of the APA explained.[145] But administrative stays do not require consideration of irreparable injury.

In addition, § 705 limits courts to issuing stays only “[o]n such conditions as may be required.”[146] And it authorizes only “necessary and appropriate process.”[147] That could be read to require the other traditional considerations. That reading is reaffirmed by the APA’s insistence that it does nothing to “affect[]” the “duty of the court” to “deny relief on any other appropriate legal or equitable ground.”[148] But administrative stays do not require the other traditional considerations either.

Until recently, that’s exactly how courts have read § 705. Courts have issued stays of agency action only after considering the Nken factors. They have not issued administrative stays of agency action void of analysis.

So it is much more likely that only ordinary stays of agency action are available under § 705, not administrative stays.[149]

Conclusion

Although early in Trump 2.0 courts have issued orders labeled as administrative stays to halt executive action, that label is inapt. These so-called “administrative stays” are better thought of as injunctions.

But this newfangled form of injunctive relief does not seem to have any legal basis. The only injunctions generally available during the pendency of litigation require courts to apply the four-factor Winter test. Yet courts have issued these de facto injunctions without applying that test or any other test.

More litigation, though, is surely coming. Courts will soon presumably issue “administrative stays” of executive action that really are stays. But courts will need to point to legal authority to issue this relief. It is unlikely such authority exists, at least outside the context of judicial review of agency orders—i.e., action taken by an agency acting as a minicourt.

  1. . Nat’l Ass’n of Diversity Officers in Higher Educ. v. Trump, No. 25-cv-00333, 2025 WL 573764, at *1, *30 (D. Md. Feb. 21, 2025). ↑
  2. . Washington v. Trump, 764 F. Supp. 3d 1050, 1052–54 (W.D. Wash. 2025). ↑
  3. . Order at 1–2, 4–5, Nat’l Council of Nonprofits v. Off. of Mgmt. & Budget, No. 25-cv-00239 (D.D.C. Jan. 28, 2025). ↑
  4. . Minute Order, Dellinger v. Bessent, No. 25-cv-00385 (D.D.C. Feb. 10, 2025) (administrative stay); Dellinger v. Bessent, No. 25-cv-00385, 2025 WL 471022, at *1, *14 (D.D.C. Feb. 12, 2025) (TRO). ↑
  5. . See infra note 75 and accompanying text. ↑
  6. . Minute Order, Dellinger v. Bessent, No. 25-cv-00385 (D.D.C. Feb. 10, 2025) (administrative stay); Dellinger, 2025 WL 471022, at *14 (TRO). ↑
  7. . Order at 1–2, 4–5, Nat’l Council of Nonprofits v. Off. of Mgmt. & Budget, No. 25-cv-00239 (D.D.C. Jan. 28, 2025). ↑
  8. . Order at 1–2, Doe 1 v. U.S. Off. of the Dir. of Nat’l Intel., No. 25-cv-00300 (E.D. Va. Feb. 18, 2025). ↑
  9. . See Nken v. Holder, 556 U.S. 418, 429 (2009). ↑
  10. . Winter v. Nat. Res. Def. Council, Inc., 555 U.S. 7, 20 (2008) (holding that to obtain a preliminary injunction, the plaintiff must show that he “is likely to succeed on the merits, that he is likely to suffer irreparable harm in the absence of preliminary relief, that the balance of the equities tips in his favor, and that an injunction is in the public interest”); see also Morgan Stanley DW Inc. v. Rothe, 150 F. Supp. 2d 67, 72 (D.D.C. 2001) (“The court considers the same [Winter] factors in ruling on a motion for a temporary restraining order . . . .”). ↑
  11. . See, e.g., United States v. Texas, 144 S. Ct. 797, 798 (2024) (Barrett, J., concurring in denial of applications to vacate stay). ↑
  12. . Id. at 799. ↑
  13. . See id. ↑
  14. . Order at 4, Nat’l Council of Nonprofits v. Off. of Mgmt. & Budget, No. 25-cv-00239 (D.D.C. Jan. 28, 2025) (internal quotation omitted). ↑
  15. . See, e.g., id. ↑
  16. . See William Baude, Foreword: The Supreme Court’s Shadow Docket, 9 N.Y.U. J.L. Liberty 1, 4 (2015). ↑
  17. . Akhil Reed Amar, Law Story, 102 Harv. L. Rev. 688, 688 (1989) (reviewing Paul M. Bator et al., Hart & Wechsler’s The Federal Courts and the Federal System (Foundation Press, 3d ed. 1988)); see also Massachusetts v. EPA, 549 U.S. 497, 520 n.17 (2007). ↑
  18. . William Baude, Jack Goldsmith, John F. Manning, James E. Pfander & Amanda L. Tyler, Hart & Wechsler’s The Federal Courts and the Federal System 371–99 (8th ed. 2025) [hereinafter Hart & Wechsler]. ↑
  19. . Labrador v. Poe ex rel. Poe, 144 S. Ct. 921, 928 (2024) (Kavanaugh, J., concurring in the grant of stay). ↑
  20. . Cf. United States v. Texas, 144 S. Ct. 797, 799–800 (2024) (Barrett, J., concurring in denial of applications to vacate stay) (noting concerns with administrative stays lasting too long, thus pushing off the normal analysis of the traditional stay factors). ↑
  21. . See Hart & Wechsler, supra note 18, at 371; id. at 371–72 (discussing Trump v. Sierra Club, 140 S. Ct. 1 (2019)); id. at 386 n.4 (providing two more examples of emergency relief being the only relief); cf. Lackey v. Stinnie, 145 S. Ct. 659, 665–67 (2025) (explaining that plaintiffs were not “prevailing part[ies]” under 42 U.S.C. § 1988(b) despite having received a preliminary injunction against enforcement of a statute that was later repealed). ↑
  22. . Minute Order, Dellinger v. Bessent, No. 25-cv-00385 (D.D.C. Feb. 10, 2025) (administrative stay). ↑
  23. . Nken v. Holder, 556 U.S. 418, 428 (2009). ↑
  24. . Id. ↑
  25. . See generally Note, Halting Administrative Action in the Supreme Court, 137 Harv. L. Rev. 2016, 2016 (2024). ↑
  26. . Compare Winter v. Nat. Res. Def. Council, Inc., 555 U.S. 7, 20 (2008) (four-part preliminary-injunction test), with Nken, 556 U.S. at 434 (four-part stay test). It is beyond the scope of this Essay to discuss the potential differences between these tests and the tests the Supreme Court uses. See generally Hart & Wechsler, supra note 18, at 375–77, 377 n.2, 380–381; Halting Administrative Action, supra note 25, at 2019–25. ↑
  27. . Nken, 556 U.S. at 434. It is also beyond the scope of this Essay to discuss the potential differences between the Winter and Nken tests. ↑
  28. . Id. at 428. Of course, there are some legal remedies, such as mandamus, which also tell people what to do. But I set those legal remedies aside for purposes of this Essay. Those legal remedies are not really temporary forms of relief. Cf. Samuel L. Bray & James E. Pfander, Remedies in the First Hundred Days of Trump II: A Gently Adversarial Collaboration, 78 Stan. L. Rev. Online (forthcoming 2025) (manuscript at 9), https://ssrn.com/abstract=5263422 [https://perma.cc/5E7H-TC9C] (“The temporary restraining order helps answer the quest for a speedy remedy, but it remains provisional. A mandamus from King’s Bench settled the matter once and for all.”). And none of the orders I discuss here seem to “require actions that are narrow and discrete,” like mandamus, but actions that are “open-ended and indeterminate.” Samuel L. Bray, The System of Equitable Remedies, 63 UCLA L. Rev. 530, 559 (2016). ↑
  29. . Nken, 556 U.S. at 428–29. Throughout I leave to the side the potentially different context Nken refers to where the term “stay” is used when “postponing some portion of [a judicial] proceeding.” Id. at 428. ↑
  30. . Id. ↑
  31. . Id. ↑
  32. . Jonathan F. Mitchell, The Writ-of-Erasure Fallacy, 104 Va. L. Rev. 933, 950–51 (2018). ↑
  33. . Samuel L. Bray, The Purpose of the Preliminary Injunction, 78 Vand. L. Rev. (forthcoming 2025) (manuscript at 9–10), https://papers.ssrn.com/sol3/papers.cfm?abstract_id=4922379 [https://perma.cc/VVA6-BA4F]. ↑
  34. . For further elaboration of this point as it relates to the debates about vacatur under the APA, see Brief of State of Florida as Amicus Curiae in Support of Respondents at 16, 18–20, United States v. Texas, 143 S. Ct. 1964 (2023) (No. 22-58), 2022 WL 16239734, at *16, *18–20. Of course, I do not weigh in on those debates. One might, for instance, think courts cannot stay rules under the APA either. Cf. infra note 149 (briefly noting this possibility but expressing no view on the matter). ↑
  35. . John Harrison, Vacatur of Rules Under the Administrative Procedure Act, 40 Yale J. on Reg. Bull. 119, 119 (2023). ↑
  36. . See 5 U.S.C. § 706(2) (“The reviewing court shall . . . hold unlawful and set aside agency action . . . .” (emphasis added)); see also Corner Post, Inc. v. Bd. of Governors of the Fed. Rsrv. Sys., 144 S. Ct. 2440, 2460–61 (2024) (Kavanaugh, J., concurring) (explaining that the “set aside” language in 5 U.S.C. § 706(2) is “understood” by many “to authorize vacatur of unlawful agency rules” but that more recently some have argued “that the APA’s authorization to ‘set aside’ agency action does not allow vacatur”). Yet again, though, I set aside the debates about vacatur for the time being. ↑
  37. . Nken v. Holder, 556 U.S. 418, 429 (2009). ↑
  38. . Id. ↑
  39. . Id. ↑
  40. . Rachel Bayefsky, Administrative Stays: Power and Procedure, 97 Notre Dame L. Rev. 1941, 1942 (2022); see also Cobell v. Norton, No. 03-5262, 2004 WL 603456, at *1 (D.C. Cir. Mar. 24, 2004) (“The purpose of this administrative stay is to give the court sufficient opportunity to consider the merits of the motion for a stay pending appeal and should not be construed in any way as a ruling on the merits of that motion.”). ↑
  41. . United States v. Texas, 144 S. Ct. 797, 798 (2024) (Barrett, J., concurring in denial of applications to vacate stay). ↑
  42. . Id. ↑
  43. . Id. at 799. ↑
  44. . Id. ↑
  45. . 11A Charles Alan Wright & Arthur R. Miller, Federal Practice and Procedure § 2951 (3d ed. Supp. Apr. 2024). ↑
  46. . See Gordon v. Holder, 632 F.3d 722, 723–24 (D.C. Cir. 2011). ↑
  47. . See Dellinger v. Bessent, No. 25-5028, 2025 WL 559669, at *3 (D.C. Cir. Feb. 15, 2025) (“Beyond the timeframe of a TRO, the district court may grant a preliminary injunction to provide relief that extends until the lawsuit is resolved.”). ↑
  48. . One might think this distinction reinforces the traditional notion that the status quo is at least presumptively the moment before judicial intervention. See Respect Maine PAC v. McKee, 131 S. Ct. 445, 445 (2010) (“[U]nlike a stay, an injunction ‘does not simply suspend judicial alteration of the status quo but grants judicial intervention that has been withheld by lower courts.’” (quoting Ohio Citizens for Responsible Energy, Inc. v. Nuclear Regul. Comm’n, 479 U.S. 1312, 1313 (1986) (Scalia, J., in chambers))). But cf. Hart & Wechsler, supra note 18, at 376 n.1 (“[T]he Court’s stay jurisprudence provides ‘no settled way of defining “the status quo.”’” (quoting Texas, 144 S. Ct. at 798 n.2)). After all, the fact that courts may issue administrative stays of lower court orders without analysis suggests that administrative stays are not extraordinary or disruptive forms of relief. By contrast, the fact that courts may issue TROs only after applicants satisfy a demanding test suggests that they are drastic and disruptive forms of relief. See Sampson v. Murray, 415 U.S. 61, 86 n.58 (1974) (“It is because the remedy is so drastic . . . that the authority to issue temporary restraining orders is carefully hedged . . . .” (internal quotation omitted)). ↑
  49. . See, e.g., McCarthy v. Briscoe, 429 U.S. 1317, 1317 n.1 (1976) (Powell, J., in chambers) (treating an application for a “stay” as an application for an “injunction” because “the applicants” sought “affirmative relief”); Sampson, 415 U.S. at 74–76 (explaining that although the lower courts had called “the mandatory retention of respondent” a “stay,” it was really a temporary “injunction”). ↑
  50. . For background on the case’s procedural history, see generally Dellinger, 2025 WL 559669. ↑
  51. . Minute Order, Dellinger v. Bessent, No. 25-cv-00385 (D.D.C. Feb. 10, 2025) (administrative stay). ↑
  52. . Nken v. Holder, 556 U.S. 418, 428–29 (2009). ↑
  53. . Id. at 429. ↑
  54. . See U.S. Const. art. II, § 1, cl. 1; id. § 3; see also Seila L. LLC v. CFPB, 140 S. Ct. 2183, 2197 (2020) (discussing these sources of authority for removal); Application to Vacate the Order Issued by the United States District Court for the District of Columbia and Request for an Immediate Administrative Stay at 12, Bessent v. Dellinger, 145 S. Ct. 515 (2025) (No. 24A790), 2025 WL 543195, at *12 (relying on the same legal bases for removal). ↑
  55. . 556 U.S. at 429. ↑
  56. . Cf. Franklin v. Massachusetts, 505 U.S. 788, 801 (1992) (holding that the APA does not “allow review of the President’s actions”); id. at 828 (Scalia, J., concurring in part and concurring in the judgment) (explaining that “[r]eview of the legality of [p]residential action can ordinarily be obtained” only by “a suit seeking to enjoin the officers who attempt to enforce the President’s directive”); Thomas W. Merrill, Article III, Agency Adjudication, and the Origins of the Appellate Review Model of Administrative Law, 111 Colum. L. Rev. 939, 942 (2011) (explaining that historically courts would only review executive branch action under the prerogative writs such as mandamus or habeas corpus). ↑
  57. . See, e.g., Macapagal v. INS, 68 F. App’x 109, 110 (9th Cir. 2003) (vacating final order of removal); see also Nken, 556 U.S. at 429–30 n.*. ↑
  58. . Nken, 556 U.S. at 428; cf. Michael T. Morley, Public Law at the Cathedral: Enjoining the Government, 35 Cardozo L. Rev. 2453, 2463 (2014) (“An injunction is the only mechanism that gives a person the immediate ability to attempt to prevent violations of a constitutional or statutory right ex ante.”). ↑
  59. . Nken, 556 U.S. at 428. ↑
  60. . Hart & Wechsler, supra note 18, at 393 (“An administrative stay temporarily stays a lower court order while the Court deliberates whether to stay definitively that order pending appeal.”). ↑
  61. . United States v. Texas, 144 S. Ct. 797, 799 (2024) (Barrett, J., concurring in denial of application to vacate stay) (emphasis added). ↑
  62. . Nken, 556 U.S. at 428. ↑
  63. . See Alexander v. Sandoval, 532 U.S. 275, 290 (2001). But cf. Chambers v. NASCO, Inc., 501 U.S. 32, 47 (1991) (counseling hesitation before finding that the inherent powers of the court have been displaced). ↑
  64. . See, e.g., Alabama v. U.S. Sec’y of Educ., No. 24-12444, 2024 WL 3981994, at *2 (11th Cir. Aug. 22, 2024) (noting that the Eleventh Circuit had previously granted an “administrative stay” of executive action that “enjoined” the Department of Education “from enforcing” a new rule); Career Colls. and Schs. of Tex. v. U.S. Dep’t of Educ., 98 F.4th 220, 233 (5th Cir. 2024) (noting that the court had “granted a temporary administrative injunction” at an earlier stage of the litigation) cert. granted in part sub nom. Dep’t of Educ. v. Career Colls. & Schs. of Tex., 145 S. Ct. 1039 (2025); Trump v. Thompson, No. 21-5254, 2021 WL 5239098, at *1 (D.C. Cir. Nov. 11, 2021) (granting “an administrative injunction”). ↑
  65. . See, e.g., Emergency Application for [a]n Emergency Injunction or Writ of Mandamus, Stay of Removal, and Request for an Immediate Administrative Injunction at 3, AARP v. Trump, 145 S. Ct. 1034 (2025) (No. 24A1007), 2025 WL 1171734, at *3; Reply in Support of Emergency Application for an Emergency Injunction or Writ of Mandamus, Stay of Removal, and Request for Immediate Administrative Injunction at 15, AARP, 145 S. Ct. 1034 (No. 24A1007), 2025 WL 1171737, at *15. ↑
  66. . Along with the reasons laid out in the discussion that follows, an additional reason to think such injunctions aren’t available in cases like Dellinger is that the injunction appears to be an order to the President. See Dellinger v. Bessent, No. 25-5028, 2025 WL 559669, at *13–14, *14 n.2 (D.C. Cir. Feb. 15, 2025) (Katsas, J., dissenting). That might run into the additional problem that the Supreme Court has long held that it has “no jurisdiction of a bill to enjoin the President in the performance of his official duties.” Franklin v. Massachusetts, 505 U.S. 788, 802–03 (1992) (quoting Mississippi v. Johnson, 71 U.S. (4 Wall.) 475, 501 (1866)). Still, the analysis of the legal bases for administrative injunctions shows that they are almost certainly not available against any executive official. ↑
  67. . 555 U.S. 7, 20 (2008); see also supra note 10 (listing the four factors). Again, I do not address any potential distinctions between the tests applied by the lower courts and Supreme Court. See supra note 26. ↑
  68. . See, e.g., Minute Order at 1, Dellinger v. Bessent, No. 25-cv-00385 (D.D.C. Feb. 10, 2025) (administrative stay) (citing the All Writs Act “and a court’s inherent authority to manage its docket” (internal quotation omitted)); Order at 2, Texas v. U.S. Dep’t of Homeland Sec., 756 F. Supp. 3d 310 (E.D. Tex. Aug. 26, 2024) (No. 24-cv-00306) (same); Order at 4, Nat’l Council of Nonprofits v. Off. of Mgmt. & Budget, No. 25-cv-00239 (D.D.C. Jan. 28, 2025) (same). ↑
  69. . See supra note 68 and accompanying text. ↑
  70. . 28 U.S.C. § 1651(a). ↑
  71. . Nken v. Holder, 556 U.S. 418, 426 (2009) (explaining that authorization to issue stays is “preserved in the grant of authority to federal courts” under the All Writs Act). ↑
  72. . See Ohio Citizens for Responsible Energy, Inc. v. Nuclear Regul. Comm’n, 479 U.S. 1312, 1312 (1986) (Scalia, J., in chambers) (explaining that “writ[s] of injunction” may be sought under “the All-Writs Act”). That said, it is unclear whether the All Writs Act as originally understood supplied a basis for courts to issue preliminary injunctions. One might think the basis for such equitable relief was the early Process Acts. Cf. Tyler B. Lindley, Anachronistic Readings of Section 1983, 75 Ala. L. Rev. 897, 916 (2023) (noting that the Process Acts of 1789 and 1792 authorized courts to issue “traditional forms and modes of proceeding in equity jurisdiction”). But such emergency relief originalism is beyond the scope of this Essay. ↑
  73. . Harris v. Nelson, 394 U.S. 286, 300 (1969). ↑
  74. . Adams v. U.S. ex rel. McCann, 317 U.S. 269, 272–73 (1942). ↑
  75. . See Order at 3, Chef Time 1520 LLC v. Small Bus. Admin., 646 F. Supp. 3d 101 (D.D.C. Dec. 1, 2022) (No. 22-cv-03587); Order at 1, Texas v. U.S. Dep’t of Homeland Sec., 756 F. Supp. 3d 310 (E.D. Tex. Aug. 26, 2024) (No. 24-cv-00306). ↑
  76. . Starbucks Corp. v. McKinney, 144 S. Ct. 1570, 1576 (2024) (quoting Weinberger v. Romero-Barcelo, 456 U.S. 305, 313 (1982)). But cf. Samuel L. Bray, The Supreme Court and the New Equity, 68 Vand. L. Rev. 997, 1014 (2015) (suggesting the Supreme Court’s treatment of the “history of equity” is somewhat constructed). ↑
  77. . Starbucks Corp., 144 S. Ct. at 1576 (quoting Weinberger, 456 U.S. at 313). ↑
  78. . See supra note 68 and accompanying text. ↑
  79. . See United States v. Hudson & Goodwin, 11 U.S. (7 Cranch) 32, 34 (1812). ↑
  80. . See Chambers v. NASCO, Inc., 501 U.S. 32, 58 (1991) (Scalia, J., dissenting) (“I agree with the Court that Article III courts . . . [have] the authority to do what courts have traditionally done in order to accomplish their assigned tasks.”); see also Missouri v. Jenkins, 515 U.S. 70, 124 (1995) (Thomas, J., concurring) (noting that the inherent “remedial authority of the federal courts” should be exercised “in a manner consistent with our history and traditions”); Union Pac. Ry. Co. v. Botsford, 141 U.S. 250, 252–57 (1891) (holding that federal courts do not possess inherent authority to order medical examinations of plaintiffs because common-law courts had never claimed such authority); Amy Coney Barrett, Procedural Common Law, 94 Va. L. Rev. 813, 852 & n.119 (2008) (explaining that history determines whether federal courts possess a given inherent power). ↑
  81. . Barrett, supra note 80, at 847–52 (explaining that the inherent powers of federal courts might either come from Article III’s vesting clause or constitutional structure more generally, but that either way “the argument” for existence of the claimed inherent power “requires a historical inquiry”). ↑
  82. . Ex parte Burr, 22 U.S. (9 Wheat.) 529, 531 (1824). ↑
  83. . Chambers, 501 U.S. at 44. ↑
  84. . Degen v. United States, 517 U.S. 820, 823 (1996). ↑
  85. . Chambers, 501 U.S. at 43. ↑
  86. . Link v. Wabash R.R. Co., 370 U.S. 626, 630–31 (1962) (emphasis added). ↑
  87. . In re Petition for Ord. Directing Release of Recs., 27 F.4th 84, 89 (1st Cir. 2022) (internal quotation omitted). ↑
  88. . See Armstrong v. Exceptional Child Ctr., Inc., 575 U.S. 320, 327–31 (2015). ↑
  89. . See Hernandez v. Mesa, 140 S. Ct. 735, 742 (2020) (explaining that “the grant of federal question jurisdiction” does not provide a sufficient “manifestation of congressional intent” to authorize creating “a damages remedy”); see also Egbert v. Boule, 142 S. Ct. 1793, 1809 (2022) (“[I]f we were called to decide Bivens today, we would decline to discover any implied causes of action . . . .”); Sosa v. Alvarez-Machain, 542 U.S. 692, 731 n.19 (2004). ↑
  90. . Tex. Indus., Inc. v. Radcliff Materials, Inc., 451 U.S. 630, 640–41 (1981); see also Thomas Koenig & Christopher D. Moore, Of State Remedies and Federal Rights, 75 Cath. L. Rev. (forthcoming) (manuscript at 59–60), https://papers.ssrn.com/sol3/papers.cfm?abstract_id=4462807 [https://perma.cc/L77G-6ZU8]. ↑
  91. . Daniel J. Meltzer, Customary International Law, Foreign Affairs, and Federal Common Law, 42 Va. J. Int’l L. 513, 541 (2002); see also Sosa, 542 U.S. at 744 (2004) (Scalia, J., concurring in part and concurring in the judgment). ↑
  92. . See Richard H. Fallon, Jr., Constitutional Remedies: In One Era and Out the Other, 136 Harv. L. Rev. 1300, 1359–60 (2023). ↑
  93. . 527 U.S. 308 (1999). ↑
  94. . See id. at 318. ↑
  95. . Id. at 326 n.8. ↑
  96. . Id. at 318–19; see also id. at 322 (“We do not question the proposition that equity is flexible; but in the federal system, at least, that flexibility is confined within the broad boundaries of traditional equitable relief.”). ↑
  97. . Id. at 318. ↑
  98. . Id. at 332. ↑
  99. . See Nestle USA, Inc. v. Doe, 141 S. Ct. 1931, 1940 (2021) (plurality opinion in part). ↑
  100. . See Egbert v. Boule, 142 S. Ct. 1793, 1803–04 (2022) (holding that “[i]f there is even a single ‘reason to pause before applying Bivens in a new context,’ a court may not recognize a Bivens remedy” and that there is such a reason “in most every case” (quoting Hernandez v. Mesa, 140 S. Ct. 735, 743 (2020))). ↑
  101. . See Nestle, 141 S. Ct. at 1938–39; Hernandez, 140 S. Ct. at 741–42. ↑
  102. . See supra notes 64, 75. ↑
  103. . Grupo, 527 U.S. at 329; see also Bray, supra note 76, at 1010 n.61 (arguing that Grupo in fact leaves open this possibility because (1) “many of the sources [Grupo] cited to determine the remedies that were ‘traditionally accorded by courts of equity,’ were actually from the nineteenth and twentieth centuries,” and (2) much language “in the opinion suggest[s] that the historical inquiry is broader than 1789 and that incremental change is not ruled out” (quoting Grupo, 527 U.S. at 319)). ↑
  104. . Starbucks Corp. v. McKinney, 144 S. Ct. 1570, 1576 (2024) (quoting Winter v. Nat. Res. Def. Council, Inc., 555 U.S. 7, 24 (2008)). ↑
  105. . Franklin v. Massachusetts, 505 U.S. 788, 801 (1992). ↑
  106. . See 5 U.S.C. § 701. ↑
  107. . Id. § 705. ↑
  108. . See Daniel B. Rodriguez, Jaffe’s Law: An Essay on the Intellectual Underpinnings of Modern Administrative Law Theory, 72 Chi.-Kent L. Rev. 1159, 1159 (1997); see also Nathaniel L. Nathanson, The Contributions of Louis L. Jaffe to Administrative Law, 89 Harv. L. Rev. 1667, 1667 (1976). ↑
  109. . Louis L. Jaffe, Judicial Control of Administrative Action 661–62 (1965). ↑
  110. . Id. at 661 (quoting S. Rep. No. 79-752, at 44 (1945) (comments of Attorney General Clark)). That said, there is one potential exception not relevant here. See id. ↑
  111. . See supra subpart III(B). ↑
  112. . 144 S. Ct. 1570 (2024). ↑
  113. . Id. at 1576 (quoting Weinberger v. Romero-Barcelo, 456 U.S. 305, 313 (1982)). Although I do not in this Essay analyze more specific statutes authorizing injunctive relief, the reasoning in Starbucks equally casts doubt on whether any other statute might authorize preliminary injunctions issued without analysis of the four Winter factors—i.e., what I have called “administrative injunctions.” ↑
  114. . Id. ↑
  115. . Id. ↑
  116. . See supra note 107 and accompanying text. ↑
  117. . 5 U.S.C. § 705. ↑
  118. . Id. § 702. ↑
  119. . 144 S. Ct. at 1574 (quoting 29 U.S.C. § 160(j)). ↑
  120. . See id. at 1576. ↑
  121. . 321 U.S. 321 (1944). ↑
  122. . Starbucks Corp., 144 S. Ct. at 1577 (quoting Hecht, 321 U.S. at 322). ↑
  123. . See id. ↑
  124. . Cf. City of Arlington v. FCC, 569 U.S. 290, 304 n.4 (2013) (explaining that agency action takes many forms even though all those forms amount to the exercise of executive power). ↑
  125. . 28 U.S.C. § 1331 has never been held to be a basis for issuing stays. Given that as well as the problems Erie poses to recognizing § 1331 as a basis to create remedial law, see supra notes 88–91, 99–101 and accompanying text, § 1331 does not provide a promising basis for issuing stays. Anyway, the analysis under § 1331 would effectively mirror that under the All Writs Act. See Grupo Mexicano de Desarrollo S.A. v. Alliance Bond Fund, Inc., 527 U.S. 308, 326 n.8 (1999). ↑
  126. . Cf. Merrill, supra note 56 (explaining that historically courts would only review administrators’ actions under the prerogative writs such as mandamus or habeas corpus). ↑
  127. . 316 U.S. 4 (1942). ↑
  128. . Nken v. Holder, 556 U.S. 418, 423–26 (2009). ↑
  129. . Id. at 426. ↑
  130. . See id. at 426–27, 433. But cf. Garland v. Dai, 141 S. Ct. 1669, 1678 (2021) (challenging this characterization of the relationship between agency orders and reviewing Article III courts). ↑
  131. . See Nken, 556 U.S. at 426–27 (explaining that the power to stay the immigration judge’s order was “consonant with the historic procedures of federal appellate courts” (quoting Scripps-Howard, 316 U.S. at 13)); id. at 429 n.*. ↑
  132. . Id. at 426 (noting that “[a]n appellate court’s power to hold an order in abeyance . . . has been described as ‘inherent,’ preserved in the grant of authority to federal courts” under the “All Writs Act” (quoting In re McKenzie, 180 U.S. 536, 551 (1901))). ↑
  133. . Scripps-Howard, 316 U.S. at 9–11; see also Sampson v. Murray, 415 U.S. 61, 72–73 (1974) (discussing Scripps-Howard). ↑
  134. . Scripps-Howard, 316 U.S. at 17. ↑
  135. . Sampson, 415 U.S. at 73–74. ↑
  136. . An analogy to preliminary injunctions should help illustrate the point. Although certain forms of preliminary injunctions are undoubtedly available, the Court in Grupo still found that another form of preliminary injunction was not. See Grupo Mexicano de Desarrollo, S.A. v. Alliance Bond Fund, Inc., 527 U.S. 308, 333 (1999). To be sure, that deals with remedies of different scope, rather than remedies with a different test. Still, the point remains that authorization to issue one kind of remedy does not entail authorization to issue a different kind of remedy—even if the two are closely related. And administrative stays are not ordinary stays. See United States v. Texas, 144 S. Ct. 797–99 (2024) (Barrett, J., concurring in denial of applications to vacate stay) (contrasting administrative stays and ordinary stays). ↑
  137. . Adrian Vermeule, No, 93 Texas L. Rev. 1547, 1560 n.76 (2015) (reviewing Philip Hamburger, Is Administrative Law Unlawful? (2014)). ↑
  138. . To be sure, one might argue that the entire modern administrative edifice is built on what Thomas Merrill has called an “appellate review model,” in which courts relate to all agency actions as appellate courts do to trial courts. Merrill, supra note 56, at 940. But Professor Merrill himself does not suggest that the appellate review model has anything to do with remedies, see id.; cf. FCC v. Pottsville Broad. Co., 309 U.S. 134, 141 (1940) (explaining that the relationship of reviewing court to agency is not always like the relationship of appellate court to lower court), although others have suggested that, see Mila Sohoni, The Power to Vacate a Rule, 88 Geo. Wash. L. Rev. 1121, 1133–34 (2020) (arguing that vacatur follows “from the appellate review model . . . incorporated into the APA”). Anyway, Professor Merrill argues that the appellate review model’s adoption in the administrative law context traces back only to “the early decades of the twentieth century.” Merrill, supra note 56, at 942. So it should not affect the remedies available under the courts’ inherent powers or the All Writs Act. ↑
  139. . Jaffe, supra note 109. ↑
  140. . Jaffe may also have referred to agency orders, though, simply because that was the primary form by which agencies were originally expected to make policy. See Antonin Scalia, Vermont Yankee: The APA, the D.C. Circuit, and the Supreme Court, 1978 Sup. Ct. Rev. 345, 376. ↑
  141. . One might think the “appellate review model” would not counsel in favor of drawing such a distinction between agency orders and other agency action for purposes of the APA. See Sohoni, supra note 138 (arguing that the appellate review model applies equally to agency rules and orders). But as discussed above, Professor Merrill himself never argued the appellate review model had anything to do with remedies. See supra note 138. Anyway, the text of the APA, as discussed below, seems to authorize only ordinary stays, not administrative stays. ↑
  142. . See Jaffe, supra note 109 (offering only one potential exception not relevant here). ↑
  143. . 5 U.S.C. § 705. ↑
  144. . See Vermont Yankee Nuclear Power Corp. v. Nat. Res. Def. Council, Inc., 435 U.S. 519, 546 (1978) (observing the manual’s influence on the Court’s interpretation of the APA). ↑
  145. . Tom C. Clark, U.S. Dep’t of Justice, Attorney General’s Manual on the Administrative Procedure Act 106 (1947). ↑
  146. . 5 U.S.C. § 705. ↑
  147. . Id. ↑
  148. . Id. § 702. Although somewhat equivocal, the Attorney General’s Manual seems to support this reading. See Attorney General’s Manual, supra note 145, at 106 (noting that § 705 “does not require the issuance of stay orders automatically upon a showing of irreparable damage” but that instead, “reviewing courts may balance the equities,” and “should take into account” the public interest (internal quotations omitted)). ↑
  149. . For the same reasons some have argued that the APA as originally understood did not permit vacatur of rules (as opposed to orders), see, e.g., Harrison, supra note 35, at 121 n.6, perhaps the APA as originally understood did not permit stays of rules but only stays of orders. That issue, though, is beyond the scope of this Essay. And both scholars and judges have pushed back on those who suggest vacatur is not available in reviewing agency rulemaking under the APA. See, e.g., Sohoni, supra note 138; Corner Post, Inc. v. Bd. of Governors of Fed. Rsrv. Sys., 144 S. Ct. 2440, 2460–70 (2024) (Kavanaugh, J., concurring). Moreover, under current Supreme Court doctrine, agency rules are frequently stayed. See, e.g., West Virginia v. EPA, 577 U.S. 1126, 1126 (2016). So any argument that § 705 does not permit stays of agency rules would presumably meet just as concerted opposition as the arguments against vacatur have. ↑

Download PDF