The Emergency Docket Explained: How the Supreme Court Decides Cases Before It Decides Them

On June 29, 2026, the Supreme Court overruled Humphrey’s Executor in a signed majority opinion by the Chief Justice. Rebecca Slaughter lost. The decision was a full merits ruling — briefed, argued, and written. But the case had arrived at the Court as something very different: an emergency application asking the justices to pause a district court’s order reinstating Slaughter at the FTC. In September 2025, the Court granted that stay. For nine months, the FTC operated as though Humphrey’s Executor were already overruled. No briefing had been filed on that question. No argument had been held. And yet, as anyone watching the case understood, the stay was the ending. The written opinion in June only ratified what the September order had already decided. In Slaughter’s case, the emergency docket wasn’t a procedural way-station. It was the whole case.

This is the detour between the threshold doctrines of judicial review (standing, finality, ripeness/exhaustion/mootness, and scope limits) and the merits machinery that comes next. Every episode in the threshold series was about doctrines that answer when or whether a case ever reaches the merits. The emergency docket operates on a different premise — it exists for the cases where waiting for the merits would decide the outcome. And over the last several years, it has become one of the most consequential mechanisms in American public law. Thirty applications for emergency relief were filed by the federal government between January 2025 and early November — against nineteen across the previous administration’s full four years, and eight across the sixteen years of the two administrations before that combined. The volume alone is a story. The format in which those cases are decided is the deeper one.

What the emergency docket actually is: a federal court’s authority to decide applications fast — in days or weeks — through stays, injunctions pending appeal, and applications to vacate lower-court stays. These applications arrive because something is happening in the meantime. Someone’s about to be executed or deported. An election is days away. A policy is about to take effect, or is about to be blocked. Courts have always had to act in these windows, because the alternative is often that the question gets decided by default while briefing takes its ordinary months. What’s new isn’t the format itself. What’s new is the volume, the stakes, and the growing frequency with which these fast-posture rulings are shaping doctrine in ways that look much more like full merits decisions than their formal label suggests. The Court itself has given varying answers about how much weight these orders carry. In Trump v. Boyle (July 2025), the Court granted a stay in two sentences, said the application was “squarely controlled” by an earlier two-page emergency order (Trump v. Wilcox), and added: “Although our interim orders are not conclusive as to the merits, they inform how a court should exercise its equitable discretion in like cases.” Justice Kagan dissented, pointing out that only another under-reasoned order was holding up Wilcox, and that next time there would be two unexplained orders to cite instead of one. The reasoning, she wrote, would be turtles all the way down.

The four-factor framework for granting a stay was restated in Nken v. Holder (2009): likelihood of success on the merits, irreparable harm to the applicant, substantial injury to the other parties, and the public interest. Likelihood and irreparable harm matter most, but the Court has never said how to weigh them. A court applying this test in a signed opinion shows its work. A court applying it in a one-paragraph unsigned order doesn’t need to. The second factor — irreparable harm — is where recent cases have moved the most. Chief Justice Rehnquist, writing alone on an emergency application in 1977, said that “any time a state is stopped by a court from carrying out statutes its legislature has passed, it suffers a form of irreparable injury.” In 2012, Chief Justice Roberts, also writing alone on an emergency application, quoted that line. Justice O’Connor, in another single-justice emergency application in between, had written that an injunction against the government “intrudes on a coordinate branch.” Three statements, three single-justice opinions, none of them an opinion for the full Court. Last year, in Trump v. CASA, the Court put them together: a universal injunction improperly intrudes on a coordinate branch and prevents the government from enforcing its policies against non-parties. That was enough to justify interim relief, without the Court having to decide whether the underlying policy was lawful. Justice Sotomayor’s dissent gave the counter-case: “The majority is apparently willing to say” that the government is irreparably harmed and entitled to emergency relief by universal injunctions “even if what [it] wants to do is break the law.”

How much the format matters is on sharpest display in Trump v. Cook, the Lisa Cook case decided June 29 with Slaughter. The administration had moved to remove Cook from the Federal Reserve Board in August 2025. District court injunctions kept her in office starting in September. The government appealed and sought emergency relief at the Supreme Court. The Court didn’t rule. For nine months. Then, in June, it set the case for argument in January, held full oral argument, and issued a signed majority opinion — formally captioned “application for stay,” formally resolved by one line (“application denied”). The majority wrote: “Although this extraordinary case arises on the Court’s interim docket, the Court has had the benefit of not only amici and oral argument, but months of internal consultation and deliberation.” Translation: when the Court wants an emergency-docket decision to carry weight everywhere, it can give it the full treatment. The Cook opinion ran five separate writings. The entire operation took the Court months. The disposition was still, formally, interim relief. The merits had not been “decided.” And yet the Court had, in substance, decided them. In Trump v. Slaughter, decided the same morning, this is the move the Court used to end Humphrey’s Executor. The emergency posture and the full merits posture had converged.

That convergence is partly a response to criticism of the format. Written opinions accompanied almost none of these applications in the mid-2010s. The share grew to twenty-three percent in 2023-24, thirty-one percent in 2024-25, and thirty-nine percent in the term that ended October 2025. On October 14, 2025, the Court granted an FCC application with a five-page unsigned opinion — unremarkable anywhere else, four and a half pages more than these orders usually get. The institution is arguing with itself in public. And it is making rules out loud as it goes. In May 2025, in Trump v. Wilcox, the Court let the president remove members of the labor board and merit systems board in a two-page unsigned opinion, and added a sentence saying the Federal Reserve is “a uniquely structured, quasi-private entity that follows in the distinct historical tradition of the First and Second Banks of the United States.” The Federal Reserve wasn’t a party to that case. The claim was undefended by any adversary. The authority cited in a footnote didn’t actually support the proposition. Justice Kagan said in dissent that the sentence appeared out of the blue. Eight months later, when the Court heard Cook, the Solicitor General expressly acknowledged what Wilcox had said and conceded that the administration wasn’t challenging the statute protecting Fed governors from removal without cause. A sentence in an unsigned emergency order became the premise of the government’s argument in another case — decided the same morning as Slaughter. The Fed’s carve-out held. Independent agency structure for everything else was ended.

The emergency docket’s effects go well beyond the removal cases. In 2022, OSHA issued an emergency standard requiring large employers to mandate vaccination or testing. Challenges went up on emergency applications. The Court heard arguments within weeks. It blocked the standard in an unsigned opinion holding that OSHA had likely exceeded what Congress authorized. The same day, on a parallel application, it allowed a narrower rule for federally funded healthcare facilities to take effect. OSHA withdrew the enforceable standard weeks later. There was never a merits judgment. The ruling about likelihood of success ended the whole thing. In August 2025, the Court fractured over terminated NIH research grants in a way neither side had asked for — Justice Barrett joined one group of four to let grant terminations take effect and a different group of four to leave the underlying directives blocked. Every piece had five votes but only one justice supported the whole combination. Justice Jackson called it “a cautionary tale about lawmaking on the emergency docket,” leaving grantees “on a likely futile, multi-venue quest for complete relief.” And the single most-cited election-law principle of the last two decades — the Purcell principle, that federal courts shouldn’t change the rules of an election when the election is close at hand — comes from a three-page unsigned emergency order from 2006 that never defined “close at hand.” Lower courts apply it constantly. The justices have openly disagreed about what it actually requires.

The emergency docket’s growth is tied to the growth of broad district-court relief. A single district judge issuing an injunction that reached beyond the parties produced an immediate government application to pause the order — first to the court of appeals, then to the Supreme Court. In Trump v. CASA, the Court restricted universal injunctions. Government applications fell from twenty-one in the previous term to nine in the term that just ended. Substantive emergency applications overall dropped from forty-nine to thirty-six. The grant rate dropped from nineteen of twenty-one to five of nine. Correlation, not proof — the government’s litigating posture may have changed, the subject matter definitely changed. But the timing fits, and the change in volume matters as much as the change in doctrine. The emergency docket’s reach into American public law isn’t primarily a story about judicial design. It’s substantially a story about how often one litigant chose to knock on the door. That variable is the one no one votes on. And the fast track isn’t evenly available. Getting to the Supreme Court on an emergency application in three days takes a legal team that can produce a serious filing in that window. The federal government can always do that. Large institutional litigants can. Most people can’t. Justice Barrett has described the underlying tension about as well as anyone: interim applications “routinely require the Court to balance the lock-in risk of saying too much against the transparency cost of saying too little.” The criticism was never that interim relief exists. It’s that consequential law is being made in unsigned orders with no reasoning. The response, in the two biggest separation-of-powers cases of the term, was to slow down and write. With the caveat that slowing down is discretionary.

Listen Now

Apple Podcasts | Listen on our site

What We Cover

Full Rough Transcript

Opening

Gwen: Welcome to Administrative Remedies, because you can't fix what you don't understand. Brought to you in part by the University of Tulsa College of Law's Mabee Legal Information Center. I'm Gwendolyn Savitz, an associate professor here at TU.

Marc: And I'm Marc Roark, a professor at the College of Law. We'll be breaking down complex doctrines with real-life analogies and examples, to demystify the world of administrative law for everyone trying to understand how government actually works.

Gwen: Agencies are the main way the federal government gets things done. It's not through Congress, for reasons we'll be addressing over the course of the series.

Marc, we've talked about this before. On June 29th of this year, the Supreme Court overruled precedent that had stood since 1935. This was Humphrey's Executor, the case that let Congress protect the heads of independent agencies from being fired at will. They voted six to three, the Chief Justice wrote, and —

Marc: They ruled that Rebecca Slaughter's removal from the Federal Trade Commission was fine, and there was no for-cause protection for her. Yeah, and the Court decided this as a merits decision. And that means that there was full briefing, that there was argument, and a signed opinion by a justice.

Gwen: Yes, but it arrived at the Court as a request to pause the lower court's order. Soon after Trump got into office, he removed Rebecca Slaughter. A district court held that it was bound by Humphrey's Executor and ordered her reinstated. As it was, that was the law of the land. The government asked the Supreme Court for a stay — an order pausing the reinstatement while the appeal ran. That was an emergency application. There had been no merits briefing. There was no argument in the Supreme Court.

Marc: And instead of just granting the motion or denying it.

Gwen: Right. So the Court decided to treat it as a petition for certiorari. It took the case before the D.C. Circuit had been able to enter a final judgment, and it wrote the question presented, naming Humphrey's Executor and asking whether it should be overruled.

Marc: It hadn't even been asked to take on the case in full. This seems like a very unusual step, to assign it for a full merits decision.

Gwen: It was indeed unusual.

Marc: Okay. And the circuit never got to finish its consideration.

Gwen: Yeah. That matters not just for this decision, but — normally we want to let these questions move through the circuit so that you get a bunch of different judges working on it. They can disagree over things. They can sharpen the issues that are actually at stake.

In this one, the Supreme Court agreed to stay the district court's opinion. It decided that in September of 2025, and it didn't issue the order in Slaughter until the following June. However, as we'll see, based on that stay, it was obvious to everybody watching how this case was going to come out. For the nine months before they'd officially said anything, the Commission operated as though the question were settled — which, again, that's how everybody viewed it, basically. So the emergency posture, what the Court did right then when it decided to stay the district court's opinion, that was immediately consequential.

Segment 1 — What the thing actually is

Marc: OK, let's start with the name, because I've heard three different ways of referring to this.

Gwen: All right. What have you heard?

Marc: OK, well, you've got the shadow docket, which is what we as academics typically talk about when we talk about these things. And we also talk about the interim docket, which is probably the most accurate, but also the least memorable. And the emergency docket, which is just the straightforward way of describing this.

Gwen: We're using emergency docket today because it is probably the clearest term for somebody who's not steeped in what's going on here. A lot of these cases come to the Court as emergency decisions. Someone's about to be executed or deported, or there's an election that's days away. If a court has to say what happens in the meantime, it's going to need to come through this emergency procedure. These can be stays, injunctions pending appeal, or applications to vacate a lower court stay. Federal courts have always had to do this, because the alternative is that we don't act at all and we potentially put in place really damaging current situations.

Marc: Okay. And so what makes that a story?

Gwen: All right. So there's a few things we care about. First, these applications are decided fast — days or weeks. Again, a lot of the time that matters, because what's at stake is going to be decided itself in days or weeks. There's generally no oral argument. The briefing is going to be thin. And the Court generally rules on these in a short, unsigned order with little or no explanation, although there might be a concurrence or a dissent attached that goes into more detail.

Marc: OK. And so no signature, no reasoning. And this is binding. It raises the question, what is a lower court supposed to do with one of these decisions coming out? If the Supreme Court stays an injunction in a two-paragraph order, is that a holding the district judge in another circuit has to follow?

Gwen: Yeah. So obviously it binds the parties. And beyond that, things get a little murkier. An interim order, by definition, doesn't finally resolve the merits. It's interim. It is in the meantime. It's a prediction about who's likely to win. It's not a ruling that somebody did win. But the Court has also said that these orders do carry weight. In July 2025, in Trump v. Boyle, it granted a stay in two sentences and told the lower courts why they should have seen it coming. Here, the Court said:

Marc: The application is squarely controlled by Trump v. Wilcox. Although our interim orders are not conclusive as to the merits, they inform how a court should exercise its equitable discretion in like cases.

That's a generous reading of "inform," though. A district judge is supposed to extract a rule from an unexplained order and then figure out whether her case is a like case.

Gwen: Yes, and that was the point Kagan made in a dissent — that the Wilcox order had offered three sentences, so only another under-reasoned order was holding it up. And next time, she said, the majority would now have two orders to cite, and the reasoning would be turtles all the way down.

Marc: So the weight of the format is contested while the format is deciding things.

Gwen: Yes. And this is being worked out order by order. We've also had a big change in volume, though. This is why people started to take more notice of it. So one count found thirty applications for emergency relief filed by the federal government between January of 2025 and early November. That was against nineteen over the entire four years of the previous administration, and eight in the sixteen years of the two administrations before that combined.

Marc: OK, so eight in sixteen years, and then thirty in ten months. Those are two very different sizable numbers.

Gwen: They really are. So of those thirty, twenty-four were granted. But then the numbers started to turn around, or at least the granting did. In the term that ended October 4th, the Court ruled on nine applications from the administration and granted five. The term before that, it had granted nineteen of twenty-one. So we're looking at different numbers, different windows, and different grant rates. This isn't a clean story about the Court consistently siding with the administration.

Marc: Okay. And so from ninety percent down to about half, and a fraction of the filings. Something has clearly changed.

Gwen: It has. Before we start talking about how this works, though, it's important to note that the government is the one choosing which of these petitions to bring. So a high success rate on its own wouldn't automatically mean that the Court was heavily favoring the government. It could be the government was only bringing the very strongest cases. Most people would probably say that was not what happened in Trump's first year in his second term. But it is worth noting.

Marc: So we could call it selection bias, right? Either way, this becomes a real caution and not really a full answer, because it explains the win rate without really explaining the volume.

Gwen: Yes. Adam Feldman also published a study on exactly that this week. He's looking at executive action applications back to 2017. So this administration won about eighty-two percent of the ones it filed in the second term. It won seventy-eight percent in the first, against sixty-seven percent under Biden. So that's a gap, but it's not a crazy high gap. And if we look at all executive action disputes, the success rate comes much closer together. One reason is that there are different kinds of cases in the pile. Across this administration, through the October 12th cutoff, immigration accounted for fourteen government applications against four under Biden, plus ten more about removing or controlling executive officers.

Marc: OK, so part of what looks like favoritism is the government bringing a different mix of cases to the Court.

Gwen: Right. It's potentially because the government's bringing a different mix. But even if that's part of it, it certainly isn't necessarily all of it.

Marc: And none of it explains the volume. Thirty applications in under ten months against eight across the Bush and Obama years. And then you have the drop this year.

Gwen: Yes. That does connect us in with the rest of the season. A lot of these applications existed because a single district judge entered relief that reached beyond the parties. As soon as one judge was halting a national policy, the government asked the court of appeals to stay it. And if that failed, it went up to the Supreme Court.

Marc: So the emergency practice grew partly out of the same thing the Court was narrowing in Trump v. CASA, when it restricted universal injunctions.

Gwen: There's kind of an awkward relationship. But this broad district court relief is driving a large volume of emergency applications. And then the Court restricted broad relief. So the following year, once district courts understood they couldn't do that anymore, there would have been less for the government to challenge. Could be why the government's applications fell from twenty-one to nine, and substantive applications overall from forty-nine to thirty-six. That's correlation. It isn't proof. And the government's litigating posture may simply have changed. And the subject matter definitely changed. But the timing fits.

Segment 2 — Why an interim order is often the whole case

Marc: And in property litigation, the fight over who holds the asset while the case runs is frequently the only fight that matters. A boundary dispute, a contested title, a foreclosure — whoever's in possession is collecting the rents, making improvements, establishing facts on the ground. And by the time you get to the judgment, three years have gone by and the equities have moved all around. And so everybody who does this work knows that preliminary relief is not preliminary in any practical sense.

Gwen: Yes, generally the same thing here. The stay of an injunction lets the policy operate. An injunction pending appeal stops it. Either way, the rules people actually live under for the next two or three years are set by this order that formally is deciding nothing.

Marc: And for a policy with a four-year horizon, two or three years is most of its life. We just spent an episode on one of these, right? The White House ballroom case.

Gwen: Yes, it was an interim docket order. It was five to four, with the Chief Justice dissenting along with the liberals. This was an order to let construction continue while the litigation ran. There was no merits judgment on whether the project needed Congress's authorization. And by the government's own estimates, the structure would be substantially finished before the Court could hear the case.

Marc: Which is the whole thesis in one order. Right. Nobody decided anything, and the building is going to go up.

Gwen: Yes. Slaughter is additional proof. We could see the ending there. She won in the district court in July of 2025, and she was out from the stay in September until the decision in June, at which point she lost. The interim order was the final answer. But if it had gone the other way, she would have spent fifteen months off the Commission, counting from the time of removal. There would have been all the votes taken without her, and no remedy for months.

Marc: And the merits in the case is real, and it may also be an epilogue.

Gwen: Yeah. And notice the stuff we've been talking about for the last few episodes. We were looking at doctrines that decide when a court decides to engage. Finality, ripeness, exhaustion. These are all built around the idea that a court should wait for the right moment. Emergency practice runs on the opposite instinct. It exists precisely for the cases where waiting would decide the outcome.

Marc: Okay. The two bodies of doctrine answer the same question with opposite defaults.

Gwen: They answer different questions, but they interact. One asks whether review is available at all, and the other asks what happens while that review runs. In the NIH case, Justice Barrett was asking whether the challenged guidance was even final agency action — on an emergency application. So the threshold questions don't disappear just because the posture is fast.

Marc: And with different rules and different levels of explanation.

Gwen: And the fast track isn't evenly available. Getting to the Supreme Court on an emergency application takes counsel who can produce a serious filing in days. The federal government can always do that. And repeat-player industry groups can do that. But for most litigants, it's going to be out of reach.

Marc: And the same sorting we saw in the standing episode — that's formally open to everyone, is really practically only available to whoever is already staffed for it. And this isn't only a separation of powers story. Has an agency rule ever lived or died this way?

Gwen: One good example is from January of 2022. OSHA issued an emergency standard requiring large employers to mandate vaccination or testing. Challenges went up on applications for emergency relief. The Court heard arguments within weeks, and it blocked the standard in an unsigned opinion. It was holding that the agency had likely exceeded what Congress authorized. But the same day, on a parallel application, it allowed a narrower rule for federally funded health care facilities to take effect.

Marc: So two agency rules with opposite results. Both resolved on a prediction about who would likely win rather than a judgment that anyone had.

Gwen: Yes. And then OSHA withdrew the enforceable standard weeks later. There was never a merits judgment. There was just a ruling about the likelihood of success. And that ended the whole thing.

Marc: So the property point again — the thing labeled temporary was the thing that decided it.

Segment 3 — Whose hand is on it

Gwen: Yeah. So this is where we can start looking at what this is based on. There's a couple of places where Congress has set up this type of authority. In section 2101(f) of the Judicial Code, it lets a justice stay a final order that's headed there on certiorari. And the All Writs Act says that federal courts may issue all writs necessary or appropriate in aid of their jurisdictions, agreeable to the usages and principles of law.

Marc: Okay, there are words of limitation in there — in aid of jurisdiction, agreeable to usages and principles — but nothing that tells a court when to grant a stay and when to refuse one.

Gwen: Yeah, it's about as thin as the grants we've dealt with so far the entire season. The criteria are judicial. The four-factor framework was restated in Nken v. Holder, a 2009 immigration case, even though it's much older than that. At the Supreme Court, there's also an additional element: the applicant has to show that the Court would likely grant review of a contrary decision. But here are the basic four elements.

Marc: One, whether the stay applicant has made a strong showing that he is likely to succeed on the merits. Two, whether the applicant will be irreparably injured absent a stay. Three, whether issuance of the stay will substantially injure the other parties interested in the proceeding. And four, where the public interest lies.

Four factors. And I notice the Court doesn't say how to weigh those?

Gwen: It says that the first two matter most — the likelihood of success and irreparable harm — but it doesn't have any sort of formula, or, and this is the part that matters for how the practice has developed, any obligation to say which factor carried the day. A court applying this in a signed opinion shows its work. A court applying it in a one-paragraph unsigned order doesn't need to.

Marc: Okay. The second factor. What's the government's irreparable harm when a court pauses a program? What makes that more than the ordinary cost of waiting for an appeal?

Gwen: All right. Justice Rehnquist wrote about that in 1977. So these emergency applications first go to a single justice. Each one is covering a particular circuit or circuits. And the justice can decide alone instead of referring it to the full Court. So in 1977 Rehnquist did that, and said any time a state is stopped by a court from carrying out statutes its legislature has passed, it suffers a form of irreparable injury. Then in 2012, Roberts quoted that line when he was also deciding an emergency application by himself. And we'd had another one in between. Justice O'Connor, in yet another one of these applications, had written that an injunction by the court against the government intrudes on a coordinate branch.

Marc: And so three opinions, each by one justice, and none of them an opinion for the full Court.

Gwen: Yes. In CASA, though, last year, the Court put O'Connor's coordinate branch language and Roberts's statute language together, and said that a universal injunction improperly intrudes on a coordinate branch and stops the government from enforcing its policies against people who aren't parties. Then that is enough to justify interim relief.

Marc: OK, enough to establish the harm without deciding whether the policy was lawful.

Gwen: Yeah. So, two things the Court already settled. That the government was likely to win the question actually in front of it — not whether the order was constitutional, but whether a single district judge could block it for the whole country. And second, that the people who sued would still be protected, because narrow injunctions covering them would stay in place.

Marc: So the Court never said the order was lawful.

Gwen: Right. It said that question wasn't before it, and that the government asking for emergency relief doesn't have to show it will win the whole case. It only has to show it's being harmed right now by a lower court's mistake.

Marc: So being blocked more broadly than the law.

Gwen: Yes, that's the harm. And Sotomayor's answer was that this entire position was self-refuting. Here's what she said:

Suppose an executive order barred women from receiving unemployment benefits, or Black citizens from voting. Is the government irreparably harmed, and entitled to emergency relief, by a district court order universally enjoining such policies? The majority, apparently, would say yes.

Marc: So the majority's answer being that even an unlawful policy doesn't give a court unlimited power to stop it.

Gwen: Yes, that's the heart of it. The majority is saying that an injunction can exceed the court's authority whether or not the policy is lawful. Her position is different in kind. She's saying that by enjoining the government from violating settled law, the orders don't cause it any harm at all. And then she takes the Maryland line and turns it around. If the injury being stopped is carrying out laws enacted by the people's representatives, look at what this executive order would override. The Fourteenth Amendment citizenship guarantee, and the statute that Congress passed to codify it.

Marc: So she's saying the Constitution and the statute protected birthright citizenship, and it's the executive order that would override them, not the injunction.

Gwen: Yes. And she said it strained credulity to treat the executive branch as irreparably harmed by injunctions that direct it to continue following settled law. And then there's a good line that everyone's quoting — that the Court endorsed the radical proposition that the Executive is harmed irreparably whenever he cannot do something he wants to do, even if what he wants to do is break the law.

Marc: And is this an argument that justices are only having in the opinions?

Gwen: She's talked about it in public too. At the University of Kansas in April, she said that there was a belief among the majority that whenever they stop the executive branch from doing something it wants to do, that that is irreparable harm to the government, but that others like her believe irreparable harm can happen to the people being affected too.

Marc: And that isn't a criticism of any particular order. It's an observation about what the format permits.

Gwen: Yes, the way the format permits it as the Court's now interpreting it. But there are strong defenses of the practice. These applications arrive fast. There can be a number of them. Full opinions at this kind of speed would be impossible. So unsigned orders let a court act without committing to reasoning that it also hasn't had time to develop. But there have been explanations getting longer lately. Last term, when the Court granted relief, at least one justice publicly noted disagreement in sixteen of eighteen cases. And when it denied relief, that happened in six of eighteen.

Marc: So these are dissent clusters where the Court acts. Which makes sense. Denying an application leaves the world as it is.

Gwen: Yes. And the writing has been increasing for the last decade. Written opinions accompanied almost none of these applications in the mid-2010s, but twenty-three percent in 2023 to '24, thirty-one percent in 2024 to '25, and thirty-nine percent in the term that just ended.

Marc: Give me an example.

Gwen: On October 14th, the Court granted an application from the FCC in an unsigned opinion that ran for five pages. Five pages would be unremarkable anywhere else, but on this docket, it's four and a half pages more than these orders usually get.

Marc: So the institution is arguing with itself in public.

Gwen: Yes, in public and in the orders. We can also see the rules actively being made here. In May of 2025, in Trump v. Wilcox, the Court let the president remove members of the labor board and merit systems board while their cases proceeded. Two pages that were unsigned. And then it added this: the Federal Reserve is a uniquely structured, quasi-private entity that follows in the distinct historical tradition of the First and Second Banks of the United States.

Marc: The Federal Reserve wasn't a party, for one. And number two, the First and Second Banks of the United States weren't necessarily very popular. There are so many things wrong with that, but —

Gwen: Yeah. Part of it was that the Federal Reserve wasn't even in the case. This was not an issue that had been raised directly, but it had been raised by implication. If those officers could be removed, why not a Fed governor? But there was no record about the Fed's structure, no adversarial testing of the historical claim. Justice Kagan said in dissent that the sentence appeared out of the blue, and that the authority cited for it — a footnote in an earlier case — didn't even support it.

Marc: And the administration presumably treated that sentence as law the next morning.

Gwen: Yes. And by the time the Court heard the Lisa Cook case eight months later, the Solicitor General expressly acknowledged what Wilcox had said about the Federal Reserve, and told the justices that the administration wasn't challenging the statute that protected its governors from removal without cause. It was instead arguing that the president had cause to remove Cook, and that his determination wasn't subject to judicial review.

Marc: So a distinction the Court drew in an unsigned emergency order became the premise of the government's argument in another case.

Gwen: Right. A case decided the same day. So this was the June 29th we started with. In Slaughter, the Court ended Humphrey's Executor and held that the president could remove FTC commissioners at will. In Trump v. Cook, five to four, it refused to let him remove Lisa Cook from the Federal Reserve Board. The Chief Justice wrote both opinions.

Marc: So the carve-out held.

Gwen: Yes, it did, which surprised nobody. But notice the timing here. Cook had been protected by district court injunctions since September. The government had appealed and sought emergency relief from the Supreme Court, and the justices left that application unresolved until June. So for nine months, Cook stayed in office under the district court's orders. The Supreme Court hadn't decided whether she should be removed.

Marc: And the district court had already given the justices a workable arrangement. Cook stayed in office and the Federal Reserve continued operating, while the Supreme Court didn't have to disturb any of it while it considered the case.

Gwen: Yes. So leaving the application unresolved meant that the district court's order was still in place. It doesn't necessarily say why the justices waited, but it meant that they could take their time without changing the situation on the ground. It's not clear the same thing would happen if the district court hadn't put the injunction in place.

Marc: And when it finally acted, it still didn't decide whether the president actually had cause.

Gwen: Right. It was decided on narrower grounds — that the statute's for-cause protection required notice and an opportunity to respond, and that Cook hadn't received either. Here's what they actually said:

Marc: Whether a Governor should be removed, it is true, is a decision only the President can make, short of impeachment. But that does not mean that he may make that decision for any reason, or no reason.

Gwen: And the Court didn't decide whether there was cause. It decided the application on a narrow ground. It read the Act's for-cause protection to carry a right to notice and an opportunity to respond, and she hadn't gotten either. So some explanation of the evidence, some avenue for response, a deadline for responding. The government argued a social media post had supplied that, and the Court said it wouldn't do.

Marc: So a statutory notice-and-response holding, in a case about central banking, in a line of argument the government had conceded because of a sentence that first appeared in an unsigned order about the labor board.

Gwen: Yeah, that's the chain. It, again, wouldn't have to automatically be sinister. Courts do signal. Preserving an obvious distinction in advance is a way of limiting the disruption your own holding is going to have. And a listener might reasonably think that this was a responsible action for the Court to take. It was important to people that they understand the Federal Reserve was safe.

Marc: Though the responsible version and the improper version produced the same artifact. A distinction about an institution that wasn't a party, drawn without briefing or argument focused on it, in a document nobody signed.

Gwen: Yes. And sometimes assembled out of two different majorities. In August 2025, the Court fractured over terminated research grants at NIH. Justice Barrett joined four justices to let the grant terminations take effect, and four different justices to leave the underlying directives blocked. Every piece had five votes, but only one justice supported the whole combination. It was an outcome neither side had asked for. Justice Jackson called it a cautionary tale about lawmaking on the emergency docket, and said grantees were left on a likely futile, multi-venue quest for complete relief.

Marc: OK, so challenge the policy in one court and then chase your money in another.

Gwen: Yes. And Wilcox isn't the only rule like that. We see this in elections too. There's a principle that federal courts shouldn't change the rules of an election when the election is close at hand, because late judicial changes could confuse voters and election officials. And lower courts apply this constantly.

Marc: Where does that come from?

Gwen: Purcell v. Gonzalez. A three-page unsigned opinion from 2006, decided on an emergency posture, that didn't announce or define "close at hand." And it's been shaping election litigation for twenty years. The justices have even disagreed openly about what it actually requires.

Marc: A twenty-year-old rule with no definition, built from a document written in days. So this isn't automatically a new phenomenon.

Segment 4 — The format is changing

Gwen: It isn't. And there's one more thing in these, and it's about the docket itself. So we should look at how these were handled. Slaughter arrived as a stay application, became a full merits case. It was argued and decided with an opinion. Cook arrived as a stay application in September '25, and instead of ruling in two paragraphs, the Court set it for argument in January and issued a signed majority opinion in June.

Marc: So Slaughter became a merits case. Cook stayed an application and got the full treatment anyway.

Gwen: Yes. Cook is captioned "on application for stay," and the disposition is one line: application denied. The Court was predicting who would likely win, not deciding it. Justice Kavanaugh wrote separately to say so expressly. And it took argument, a majority opinion, and five separate writings, and months to get there.

Marc: Did the Court say anything about doing it that way?

Gwen: It did. Here's where we talked about it in the opinion itself.

Marc: Although this extraordinary case arises on the Court's interim docket, the Court has had the benefit of not only amici and oral argument, but months of internal consultation and deliberation.

Which answers my question from earlier, at least partly, right? If the Court wants one of these to carry weight everywhere, it can give it the full treatment.

Gwen: Yes. Justice Barrett has talked about this underlying tension about as well as anyone. She said this in Mirabelli v. Bonta: interim applications routinely require the Court to balance the lock-in risk of saying too much against the transparency cost of saying too little.

That kind of sums it all up. The criticism was never that interim relief exists. It's that consequential law is being made in these unsigned orders with no reasoning. The response to that criticism, in the two biggest separation of powers cases of the term, was to slow down and write.

Marc: With the caveat that the slowing down is discretionary.

Gwen: Absolutely.

Marc: Everything we've described rewards the side that can file in three days. What does that mean for the lawyer who can't?

Gwen: Build the harm record before you need it.

Marc: So the interim order decided part of the case whichever way the merits land.

Gwen: Yes. That's the last big point. The volume that turned an ordinary judicial power into one of the most consequential mechanisms in American public law wasn't set by Congress or by the Court. It was set in large part by how often one litigant chose to knock on the door.

Marc: And that's the variable nobody votes on.

Closing

Gwen: Right. So next week, we're back to regular episodes. We're talking about the administrative record — what the judge is actually allowed to look at once the challenger gets in the door.

So that does it for today's episode of Administrative Remedies. Thank you for joining us today. Please, if you enjoy this podcast and enjoy this episode, give us a like on Spotify, iTunes, or whatever platform you're listening on. And be sure to tune in next time, where we'll continue to dive into the contours of administrative law. Because remember, you can't fix what you don't understand.Related Guides

Administrative Remedies is produced by Professor Gwendolyn Savitz and Professor Marc Roark at the University of Tulsa College of Law.