At 5:45 on the evening of May 19, 2025, six men in an immigration detention facility were handed a single document telling them they would be flown to South Sudan — a country none of them were from, and that the State Department treats as unsafe for nearly all U.S. travel. A federal district court had already found the removal plan violated due process; the First Circuit left that framework in place while the government sought emergency relief at the Court. See the Supreme Court docket summarizing the posture.
On June 23, 2025, the Supreme Court responded in a single paragraph: it granted a stay of the district court’s preliminary injunction, without resolving the underlying merits and without explaining why the lower courts were wrong. Justice Sotomayor’s dissent, joined by Justices Kagan and Jackson, walked through what had already happened on the ground — including the May 19 notice window the government later defended in the district court — and called the grant “so gross an abuse of the Court’s equitable discretion.”
After the flight plan shifted, reporting drew on government filings describing detainees held at Camp Lemonnier in Djibouti in extreme heat inside converted containerized housing. The district court kept trying to impose structured process; the administration held a press conference.
This is Department of Homeland Security v. D.V.D. — one data point in a workload that has quietly become the center of gravity for the Roberts Court. In his October 2025 wrap of the October 2024 Term, Stephen Vladeck counted 140 rulings by the full Court on emergency applications, compared with 55 signed merits opinions — a ratio that would have read like a typo a decade ago.
The permission slip Congress wrote
In 1948, Congress passed a statute authorizing emergency stays pending certiorari. The relevant section, 28 U.S.C. § 2101(f), gives the Supreme Court (and individual justices) power to pause lower court orders while a case is still in flight. The theory is sound: some harms cannot be unwound after the fact.
Congress wrote the permission slip. Congress can amend it.
For much of the Court’s history, that authority was used sparingly — the kinds of problems where delay really is the same as denial. Across the combined sixteen years of George W. Bush and Barack Obama, the federal government asked the Supreme Court for emergency relief eight times total — a figure Vladeck used as a benchmark when the modern surge began.
Then, on February 9, 2016, the Court issued an unexplained order blocking the Obama administration’s Clean Power Plan before any court of appeals had ruled on its legality. The vote split 5–4 along appointment lines.
When the line moved
What happened inside the Court stayed mostly private for a decade — until April 2026, when the New York Times published sixteen pages of internal memos circulated among the justices in the weeks before that stay. Secondary analysis — including Vladeck’s close read — describes Chief Justice Roberts as the driving force: pressing an industry-focused irreparable-harm frame while barely engaging the government’s counter-harms, and treating the application as if it were an ordinary stay pending appeal even though the posture was not.
The memos also preserve a sharper line from Justice Kagan: that granting the stay without lower-court input would be, in her view, “unprecedented” absent full briefing.
Roberts’s own contemporaneous language, as relayed in that reporting and summarized by Vladeck, is closer to conceding the posture was “unusual” than to pretending nothing had changed. That is the birth certificate of the modern emergency docket as a policy instrument: not because emergency orders never existed before, but because the Court chose to normalize a form of intervention it simultaneously knew was novel.
What followed was a ratchet. Donald Trump’s first administration filed 41 emergency applications. Joe Biden’s administration filed 19 — fewer, but enough to ratify the new baseline as bipartisan practice. Trump’s second term opened with the same tool at even higher velocity; by late March 2025 alone, the administration had already reached eight applications — matching the Bush-plus-Obama sixteen-year total.
Justice Kagan’s recent dissents from emergency stays in removal-power fights have stressed how much work the Court is doing for the executive without ordinary merits process — for example, warning that the Court risks transferring operational control of multi-member agencies to the President through repeated interim orders (see her dissent in Trump v. Slaughter).
The reform conversation’s blind spot
In July 2024, President Biden stood at the LBJ Presidential Library and announced term limits, a binding ethics code, and a constitutional amendment reversing the presidential immunity decision — ideas his own December 2021 commission had catalogued years earlier.
The critique he was responding to was real. The ethics scandals were real. So was public erosion of confidence.
Here is what that conversation systematically underweights: each of those headline proposals requires conditions that do not currently exist. A constitutional amendment requires supermajorities in Congress and three-quarters of the states. Serious term limits likely require the same survival math. An ethics code Congress passes still has to be enforced against a Court that gets the last word on what binds it.
Those are not footnotes. They are the central political fact.
Meanwhile, the branch that has been doing much of the immediate separation-of-powers work — often in one-paragraph orders — sits outside the frame of the op-ed solutions.
The upstream lever Congress keeps not pulling
The executive branch, across administrations, uses emergency applications to bypass lower courts when the stakes feel existential to the sitting president. The eight applications across Bush and Obama are less a morality play than a snapshot from an era before the tool was stress-tested as a default instrument.
The Court has institutional reasons to preserve the emergency docket in something like its current shape. Writing reasons binds you; binding reasons let lower courts distinguish you; litigants can test consistency. An unexplained order is harder to argue with — which is also why it is harder to legitimize.
Congress is not powerless on procedure. Two statutory levers could matter without constitutional amendment theater — and neither was the centerpiece of Biden’s three-part package.
The first is direct: amend § 2101(f) and tighten the statutory predicates for extraordinary interim relief so the familiar four-factor stay inquiry cannot collapse into vibes on the fastest track.
The second is upstream: much of the pressure that forces “emergency” posture in national policy fights comes from the remedy structure of nationwide injunctive relief. Congress has authority to reshape how district courts calibrate scope, panels, and venue — the kind of structural debate law professors were already having in the Virginia Law Review’s empirical shadow-docket symposium and related 2024–2025 commentary.
Instead, on June 27, 2025, the Court curtailed universal injunctions on the emergency docket in Trump v. CASA — with Justice Barrett writing that universal injunctions “likely exceed the equitable authority that Congress has granted to federal courts,” and Justice Alito warning in concurrence that, unless class certification is taken seriously, the universal injunction could “return from the grave” as nationwide class relief.
Congress could have led that remedial conversation on a slower track. It did not have to outsource the first cut to a one-paragraph shadow docket posture.
The working hypothesis
The shadow docket’s expansion is not a malfunction — it is the predictable output of a system in which every actor with the structural power to constrain the ratchet has a vested interest in keeping it wound.
That includes the Court, which gains leverage from opacity; the executive, which gains speed; and Congress, where no individual member profits from tying the hands of a future president of their own party. In December 2025, Senate Democrats (and House Democratic partners) introduced the Shadow Docket Sunlight Act — transparency-focused, and, as introduced, without Republican co-leadership in the Senate. That legislative fact is less an attack on Democrats than a datum: the cross-partisan coalition required to tie the Court’s procedural hands does not exist at the price point reformers are actually willing to pay.
The people who pay the cost — the detainees described in the D.V.D. record, a withholding recipient identified as O.C.G. whom Judge Brian Murphy described as “the banal horror of a man being wrongfully loaded onto a bus,” the lower court judge overruled without reasoning to distinguish — do not convene blue-ribbon panels. They are the residual claimants of a power transfer nobody in the room is being asked to surrender.
What would change my mind
-
If the Shadow Docket Sunlight Act — or any bill that materially constrains emergency-order procedure at the Supreme Court — attracts Republican co-leadership in a Senate where Republicans hold the majority, that would falsify the “purely partisan incentive trap” read of why these bills stall.
-
If a future administration, of either party, holds federal emergency applications to the pre-2017 baseline for a full presidential term and states that restraint as explicit enforcement policy, that would be strong evidence against the “inevitable ratchet” claim.
-
If courts treat a statutory codification of stay standards for Supreme Court emergency relief as constitutionally barred — not as a policy fight, but as a settled separation-of-powers holding — then the statutory lever I am pointing to is weaker than this piece assumes, even if the incentive map stays the same.
Until one of those happens, the permission slip stays on the table. And the Court keeps issuing orders in the dark — to people who were never given enough process to argue with what they were never shown.
Related: The One-Way Door — a paired look at institutional architecture: not the emergency docket, but the ethics stack that fails in real time when load-bearing rules are removed.
If you found this useful, the best thing you can do is forward it to one person who would push back on it. I’d rather be wrong in public than right in private.