But today’s bonus post has an even more immediate trigger—the chaos unfolding in Missouri with respect to which congressional map the state will use in the upcoming midterm elections. That issue is back at the Supreme Court for the second time in three days—for reasons that are complicated, but worth unpacking, as this post attempts to do in detail below the fold. (TL;DR: There are currently conflicting injunctions from the Missouri Supreme Court and a federal district judge in St. Louis, the former barring the state from using the map the legislature passed last year based on its interpretation of Missouri state law; the latter requiring it to do so based on a highly dubious interpretation of federal law.)

And although it’s possible that the U.S. Supreme Court may yet do the right thing (i.e., put back into effect the Missouri Supreme Court’s unanimous ruling from last week by issuing a stay of the federal district court’s TRO from Tuesday night, which would lead to Missouri using its 2022 map in November), there’s a larger point to be made, as well—which is that, no matter what they do later today or tomorrow, the justices in Washington bear at least some of the responsibility for how we got here in the first place.

For those who aren’t paid subscribers, we’ll be back with our continuing coverage of the Supreme Court no later than Monday (and probably sooner). For those who are, please read on.

Read more

Share

If you’re not already a subscriber, please consider becoming one—and upgrading to a paid subscription if and when your circumstances permit:

Subscribe now

I had been waiting until things calmed down (ha!) to devote a post to the scattershot rules governing the release of the justices’ internal papers (especially their case files); how that topic dovetails with broader concerns about the Court’s lack of transparency; and the broader point that, whatever the right answer is, Congress ought to create uniform rules for when and how the justices’ papers become accessible to the public—just like it has for presidential records. Then, on Tuesday, Jodi Kantor published a remarkable story for the New York Times on the same topic, with lots of new reporting on the internal scramble after Justice Scalia’s (unexpected) February 2016 death to figure out when and how his papers would become available for public scrutiny (TL;DR: it’s complicated, but for case files, it’s when the last justice with whom he sat that term dies—so probably at least another 30 years from now for all of them).

After recapping a very busy week of Supreme Court news (especially for the first week of September), today’s post tackles that topic—offering a brief summary of what we know about the papers that are currently unavailable; where folks can find those papers that are accessible (including just how awesome and helpful the folks at the Library of Congress’s Manuscript Division are); and why, whatever the right answer is to the question of when justices’ papers are made public going forward, it ought to be a uniform rule that comes from Congress—and not something that gets left up to the whims of individual justices, or, as in Scalia’s case, their estates. My thanks, especially, to Ava Singh (Georgetown Law ’27), for meticulous assistance in researching this topic.

But first, the (current Court’s) news.

On the Docket

Last week kicked off with Monday afternoon’s 5-4 ruling granting a stay pending appeal in the White House ballroom case, with Chief Justice Roberts writing for himself and the three Democratic appointees in dissent. I had lots to say in Thursday’s bonus post about that deeply misbegotten ruling, and won’t rehash those points here. All I’ll flag for now, because it’s going to be relevant in a minute, is that it was the third straight grant of emergency relief that the Court accompanied with a full (if unsigned) “opinion of the Court.”

Friday morning brought with it the last of the regular summer Order Lists—a series of housekeeping matters, none of which were especially newsworthy. The Court also released on Friday the argument calendar for the “December” 2026 sitting (which starts on November 30). That sitting includes the AR-15 cases on December 2; the transgender parental rights/medical care dispute on December 7; and an important dispute about whether federal law bars states from forcing putative voter registrants to produce documentary proof of citizenship on December 8. But it also includes four open slots across six days—yet more evidence that the upcoming term will once again be back-loaded with big cases (because the justices aren’t filling out the docket in the earlier sessions).

And right before the long weekend, we received the second full Court grant of emergency relief of the week—with the justices blocking a lower court ruling that had frozen an FCC proposal to apply to political parties the requirement that TV stations impose only the “lowest unit charge” on ads by political candidates as Election Day approaches. (In other words, the Supreme Court’s intervention will make ad buys for political parties, especially Republicans, cheaper, at least until November.) It’s not an especially high-stakes case (which may explain why only Justice Jackson publicly dissented), but that ought to have been reason enough for the Court to deny the stay. Instead, after getting the jurisdictional question wrong (per Justice Jackson’s dissent), the Court appears to have made yet more new law on what counts as “irreparable harm,” explaining that having to spend more money to reach voters with election-season ads irreparably harms the applicants by “hampering their efforts to reach the electorate in the critical weeks leading up to the midterms.” The Court has, historically, been adamant that monetary harms are seldom irreparable since having to spend extra money to achieve the same goal can usually be recouped after the fact. Alas.

One last note about Friday’s ruling in National Republican Congressional Committee v. Brown: The Court once again produced an (unsigned) majority opinion—the fourth straight grant of emergency relief to be accompanied by an opinion of the Court. (You have to go back to May 14 to find grants without majority opinions.) I wondered in Thursday’s post about whether the justices have adopted a new norm of always writing an opinion of the Court when they vote to grant emergency relief; four in a row is even stronger evidence that this may be a thing than the three I was relying upon then. And as I suggested on Thursday, that’s both a positive development (the content of the writing notwithstanding), and one that would necessarily reflect a Court once again showing itself to be reactive to public criticism—at least when that criticism is framed in institutional (as opposed to political or ideological) terms.

The emergency docket is also likely to figure prominently during the week to come, thanks to two major, pending emergency applications. The first is in the mail-in ballots case. On Thursday, the Trump administration filed an application seeking a stay of Judge Talwani’s TRO (#37, if you’re keeping count), once again seeking to put back into effect its effort to restrict which mail-in ballots will be delivered by the U.S. Postal Service. But that application was mooted by the preliminary injunction Talwani entered on Friday. So yesterday, the administration withdrew the first application and filed a new one (#38)—this time seeking a stay of the preliminary injunction itself and an immediate administrative stay. As with the previous application, Justice Jackson (as Circuit Justice for the First Circuit) did not issue an administrative stay, but did set a tight deadline for a response—the plaintiffs’ briefs are now due Wednesday at 4 p.m. ET. Thus, it’s at least possible that the Court will rule on this latest application (U.S. Postal Service v. California) by the end of this week—and next week at the latest. The First Circuit is also moving quickly, so it’s also possible (if not likely) that it will rule first, although I think everyone expects it to deny the stay.

And late Friday (although it was docketed on Saturday), Missouri filed an emergency application asking the justices to pause the Missouri Supreme Court’s unanimous ruling that had blocked, on state-law grounds, Missouri’s 2026 gerrymander, which had been designed to create an additional safe Republican seat in the Show-Me State’s eight-member House delegation. (In other words, Missouri is asking the justices to restore the 7R-1D map.) Missouri’s pretty hysterical application (it opens “A federal-election-administration disaster is unfolding in Missouri.”) fails to mention some pretty important facts, including that there are prior examples of states conducting general elections with different districts than the ones that had been in place for the primaries. Its central claim is that the state supreme court’s decision, by requiring this result, violates the federal Constitution—a not-so-subtle effort to invoke the “Independent State Legislature” theory that the Court declined to embrace (or, alas, conclusively reject) in Moore v. Harper. Justice Kavanaugh ordered the challengers to respond by noon today (even though it’s a federal holiday), so this one is likely to move quickly, too.

When the Court rose for its summer recess back in June, I suggested that it was too soon to close the book on OT2025 because we were likely to see a flurry of important emergency applications before the term “officially” ends on the first Monday in October. After a quiet beginning to the summer, things sure are picking up…

The One First “Long Read”:
How the Sausage Gets Made Public

The justices’ papers have long been a subject of intense academic study and controversy. There is a lot to learn about the Court’s internal machinations and deliberations from the justices’ internal records of the proceedings, and forests have been felled about the lessons that can be divined not just with respect to the decisionmaking processes in individual cases (like the initial vote at Conference in Casey to overturn Roe) but the justices’ personal relationships and procedural behavior more generally. If anything, given how much of the Court’s real-time output is shrouded in secrecy (everything from the votes for certiorari to who had, and then lost, an original opinion assignment after oral argument), the internal papers have become even more important to try to piece together how the Court got from A to B.

There’s also a related body of meta-scholarship about the debate over when and how the justices’ papers should become accessible. Professor Susan David deMaine, who’s the Director of the Law Library at the Maurer School of Law at Indiana University, has written extensively on just about all of this—my favorites are her 2018 Law Library Journal article, “Access to the Justices’ Papers: A Better Balance,” and a 2022 co-authored empirical study of how scholars have used the papers of Justices Brennan, Marshall, and Blackmun. (There’s also a great 2013 NYU Law Review article by University of Washington professor Kathryn Watts, “Judges and Their Papers.”) And there have been a series of historical flashpoints, including Justice Marshall’s decision to make his papers available immediately upon his 1993 death—a move that provoked a remarkably sharp and negative response from Chief Justice Rehnquist on behalf of “a majority” of the Court.

Here’s what we know about what we don’t currently know:

I. Nine Former/Retired Justices With Existing Restrictions

Counting Justices Kennedy and Breyer (who, so far as I know, have yet to make any public announcement about what’s going to happen to their papers), there are nine former justices whose case files, in particular, are not yet (fully) accessible to the public. For the seven whose plans have been announced, here are the details (in chronological order based on when they left the Court—which I’ve indicated in parentheses):

Warren Burger (1986): Despite some sources suggesting that Burger’s papers were due to become available later this month (40 years after his retirement), it appears that his papers will in fact become available on December 1, 2033—ten years after the death of Justice O’Connor, the last surviving justice with whom Burger served. Burger’s papers were donated to William & Mary.

William H. Rehnquist (2005): Rehnquist donated his papers to the Hoover Institution at Stanford University, and the papers are being opened on a rolling basis with portions of the papers being opened following the deaths of each of his former colleagues. Currently, papers from OT1987—OT2005 remain closed, presumably because Justice Kennedy was nominated to the Court in 1987 and is still alive. If that approach holds, Rehnquist’s papers will not be fully available until after Justices Kennedy, Thomas, and Breyer have died.

Sandra Day O’Connor (2006): O’Connor gave her papers to the Library of Congress. Papers through OT1990 are available to the public (and have provided a bit of a treasure trove already). But she restricted access to the papers from OT1991—OT2005 until all justices who served with her have retired. Justice Thomas’s retirement would open most of that period up (OT1991—OT2004), leaving only OT2005 for when Chief Justice Roberts retires.

David Souter (2009): Souter’s papers were donated to the New Hampshire Historical Society and are closed until 50 years after his death—so they won’t become available until May 8, 2075. Sigh.

John Paul Stevens (2010): Stevens donated his papers to the Library of Congress starting in 2005, while he was still on the Court. The first installment of papers from OT1975—OT1984 opened in 2020, and the second installment of papers from OT1985—OT2004 opened in 2023. The papers from OT2005—OT2010 will remain closed until October 2030 per Stevens’s instructions. (Importantly, because of the others’ restrictions, Stevens’s papers are the only ones currently open with coverage of OT1994 or any later term, up to and through OT2004.)

Antonin Scalia (2016): Scalia’s papers are held at the Harvard Law Library and there are a number of restrictions on them based upon the type of file. The “correspondence” files will be opened on a rolling basis with the files from 2016 being released in 2047, which is 30 years from the date of creation. The “speaking engagement and event” files will similarly open on a rolling basis, with files becoming available 30 years after the date they were created. Papers from Scalia’s time on the D.C. Circuit also open periodically following the death of judges who participated in individual cases. All D.C. Circuit case files are expected to be open by 2050—but individual items within the case files may be closed longer. The papers from his time on the Supreme Court will open for a particular term upon the death of all other justices who served during that term (the youngest justice to have served with Scalia is Justice Kagan).

Ruth Bader Ginsburg (2020): Ginsburg donated her papers to the Library of Congress. The papers include both personal papers (such as birthday cards and well wishes) and papers relating to her career and time on the Court. Certain papers are available now; another batch will become available on January 26, 2027; and others will become available 25 or 50 years after her death (September 18, 2045 and September 18, 2070), respectively. As for the case files, specifically, those files will remain closed until all judges or justices who participated in that case have died. Because Justices Thomas and Breyer are still alive, that means none of her case files from the Supreme Court are open yet. Certain papers also involve sealed files which must remain restricted until the Court unseals them.

II. Where to Find Everything Else

Professor deMaine’s 2018 article has an incredibly useful appendix that identifies the donations and access restrictions on the most significant collections of papers for every justice to serve on the Court during and since 1900. Although the collections are (literally) all over the place, the most common location is the Library of Congress—where the papers can be accessed through the Manuscript Division.

That gives me an excuse to note that I wrote a lengthy article early in my career that attempted to reconstruct (and then deconstruct) the Supreme Court’s cryptic but important 1948 decision in Hirota v. MacArthur (rejecting a last-ditch effort to get the justices to review the Tokyo war crimes tribunal on the ground that it was really an American operation, and thus subject to American judicial review), a project that relied heavily on access to the justices’ papers from the October 1948 Term. It turns out that the mystery could largely be untangled from the internal papers of Justices Black, Jackson, and Rutledge, specifically—including a draft opinion by Rutledge that my article was the first to make public. There’s just so much more context one can glean from seeing how the justices processed things internally than is often visible from what is ultimately produced for public consumption. (Even Chief Justice Roberts invoked Justice Blackmun’s papers during the oral argument in Dobbs.)

Let me also note here how amazingly helpful the folks at the Manuscript Division were (and continue to be). One need not be an accredited law professor writing an article to access these papers; it’s open to “readers age 16 and above who have a specific need for access to the collections of the Manuscript Division and associated materials.” More details are available here.

III. The Case for Uniformity

I’ve saved the most controversial part for last, but as the haphazardness described above should drive home, it sure would be useful if there were a uniform set of rules to govern how/when the justices’ papers become public. In her 2018 article, Professor deMaine identified the source of the problem as the fact “that the Supreme Court Justices continue to have personal ownership of their working papers, resulting in idiosyncratic retention and access decisions.” (For what it’s worth, the justices have asserted personal ownership of their papers, but there’s no external authority agreeing with them, and it seems odd to think that the government owns every other piece of paper produced by government employees on the job except the ones produced by the justices in their chambers.)

Regardless, Congress, deMaine argued, should assert (or, in my view, re-assert) ownership over the justices’ papers—and create a uniform moving wall for when case files, at the very least, would be released to the public—sometime between ten and twenty years after a justice retired, regardless of who is still serving on the Court. As she suggested,

A window of no fewer than ten years and no more than twenty would allow the Justices, and others represented in the collections, a long stretch of confidentiality and would ensure that no matters discussed in the papers were still in front of the Court. After twenty years, access restrictions would be allowed only on materials the discourse of which would “constitute an unwarranted invasion of privacy.”

As deMaine’s article explains, such a proposal largely tracks a series of recommendations made by the 1977 National Study Commission on Records and Documents of Federal Officials. If that were the rule today, we’d already have access to Rehnquist’s, O’Connor’s, Souter’s, and Stevens’s papers in full (bringing us to the end of OT2009); with Scalia’s (middle of OT2015) and Ginsburg’s (end of OT2019) papers not far behind—versus decades away from public view.

But whatever the “right” answer is, it shouldn’t be controversial to suggest that public awareness of the Court is not well-served by a series of widely varying access rules crafted by the justices (or their estates) to serve the justices’ purposes, rather than the public’s. Judicial deliberations can and should be confidential—for a time. But we already know so much more about the Court because of what we’ve learned from the justices’ internal papers from earlier generations. Having a uniform and predictable set of legislatively imposed rules to govern the release of papers going forward would not only help to increase the Court’s transparency, but it would also reaffirm that the justices’ work, which increasingly affects all of us, also belongs to all of us.

SCOTUS Trivia: Who’s Number Two?

The Library of Congress remains the repository of choice for the justices’ papers. By my count, it holds the “principal” collections of 38 justices’ papers. There’s no official claim to second place, but I’m pretty sure it belongs to the Harvard Law School Library—which houses the principal collections of five justices: Joseph Story, Oliver Wendell Holmes, Jr., Louis D. Brandeis, Felix Frankfurter, and, most recently, Antonin Scalia. Of course, given that Harvard Law graduated twice as many justices as any other law school (18, to Yale’s nine), this particular piece of trivia may not be all that surprising.

I hope that you’ve enjoyed this installment of “One First.” If you have feedback about today’s issue, or thoughts about future topics, please feel free to email me. And if you liked it, please help spread the word!

Share

If you’re not already a paid subscriber and are interested in receiving regular bonus content (or, at the very least, in supporting the work that goes into this newsletter), please consider becoming one:

Subscribe now

This week’s bonus issue will drop on Thursday. As ever, I hope you are all staying safe out there, especially from whatever is in the justices’ internal papers.

Rather than criticize Monday’s ruling in National Park Service v. National Trust for Historic Preservation (Professor Leah Litman has that well in hand), I want to draw out some meaningful lessons that I think we can take away from the Court’s latest misadventure on the emergency docket—one about Chief Justice Roberts (who wrote the dissenting opinion on behalf of himself and the three Democratic appointees); one about Justice Barrett (who continues to vote in ways that I just don’t understand given what she has claimed to be her committments); one about the Court’s internal procedures for handling emergency applications (as reflected in Roberts issuing an administrative stay and then writing the dissent from a stay pending appeal); and one about the Court’s apparent susceptibility to certain types of sustained public criticism (as reflected in the fact that, for the second straight case, we have a brief but deliberate majority opinion formally explaining the Court’s intervention).

I unpack each of these points in more detail below the fold. For those who aren’t paid subscribers, we’ll be back (no later than) Monday with our continuing coverage of the Supreme Court. For those who are, please read on.

Read more

Share

If you’re not already a subscriber, please consider becoming one—and upgrading to a paid subscription if and when your circumstances permit:

Subscribe now

As early as the middle of this week, the Supreme Court may once again be asked to put back into effect President Trump’s effort to interfere with the use of mail-in ballots during the upcoming midterm election cycle. As I suggested after last Monday’s ruling in Trump v. California, Judge Talwani (D. Mass.) moved quickly to issue a temporary restraining order against the U.S. Postal Service’s Final Rule purporting to implement President Trump’s executive order, and the government has already appealed that ruling to the First Circuit (and will surely seek a stay pending that appeal). Once the First Circuit rules, it seems inevitable that the case will go right back to the justices.

One of the questions I raised last week was whether the government would argue (with more than a little chutzpah) that Judge Talwani’s latest intervention should be stayed because it’s “too close” to the election, an invocation of the so-called “Purcell principle”—the idea that that federal courts should stay their hand as election day draws nigh lest they issue rulings that increase the risk of voter confusion.

But the Supreme Court has never previously been asked to apply Purcell where the election-related rule blocked by a lower federal court came from the federal government. Going back to Purcell itself, every single application of Purcell by the Supreme Court over the last 20 years has involved a lower-court injunction against a state (or local) directive. That’s as it should be. Even taking Purcell seriously (which is admittedly challenging) and assuming, for the sake of argument, that it’s a consistent, neutral principle defining the equitable authority of federal courts, there are two independent reasons why Purcell ought never to apply when the federal government is the defendant—and a third reason, for those unpersuaded by the first two, why it ought not to apply in the mail-in ballots cases specifically.

More on that below. But first, the (other) news.

On the Docket

The only formal ruling out of the Court last week was Monday’s per curiam opinion (and dissents) in the first mail-in ballots case, Trump v. California. Although my initial post reacting to that ruling was written in a hurry, I think it still suffices to summarize what the Court did (and, as importantly, what it didn’t do).

I also want to flag one of the more ridiculous recent efforts by states to invoke the Supreme Court’s original jurisdiction—a new motion for leave to file a bill of complaint filed by Iowa and Montana against twelve different states, seeking to have the Supreme Court stop those states from suing (as they currently are) to block Paramount’s merger with Warner Bros. Discovery. In other words, Iowa and Montana are trying to use the Court’s original jurisdiction to speed up a merger of private companies. (Beyond seeking a grant of leave to file, Iowa and Montana have also asked the justices to expedite their consideration of the motion for leave to file.) Whatever one thinks about the Thomas/Alito view of the Court’s original jurisdiction (i.e., that it is mandatory when states are on both sides of the “v.”), the reality is that the Court has repeatedly rejected that view—and has made abundantly clear in recent years that it will not let its original docket be hijacked for political disputes. The only grants of original jurisdiction over the last two decades have all involved cases in which the states are the real parties in interest on both sides—basically disputes over land and water rights. Whatever one thinks of the Paramount/WBD merger, this ain’t that. Instead, this latest suit is just a waste of a lot of time, effort, and taxpayer money.

Speaking of efforts to manipulate the legal system for political purposes, the New York Times published a very well-reported story over the weekend about how the Trump administration has gamed the civil litigation rules to bring various disputes not just in the Northern District of Texas, but before Judge Reed O’Connor, specifically—one of the federal judiciary’s most visible public defenders of the very “judge shopping” in which DOJ is engaging (here’s my response to O’Connor’s speech on the subject). I’ve pointed out before the irony (one might even say hubris) of folks complaining about where suits against the Trump administration have been filed given this kind of behavior; it’s nice to see a discussion of it outside this newsletter.

Turning to this week, we expect the third and final housekeeping summer Orders List this Friday (September 4) at 9:30 ET. It also stands to reason that, at some point, we’ll get the full Court ruling on President Trump’s application to allow him to continue building his new ballroom for the duration of the litigation challenging it; Chief Justice Roberts’s administrative stay (allowing the construction to continue for now) has been in place for 10 days. Obviously, that could also come this week. Or not.

The One First “Long Read”: Purcell and Uncle Sam

The “Purcell principle,” as UCLA professor Rick Hasen has dubbed it, is the idea that, as election day draws nigh, federal courts should refrain from using their equitable authority to rule in ways that might increase voter confusion. It’s named after a remarkably (if typically) cryptic unsigned opinion respecting a grant of an emergency application that the Supreme Court handed down in October 2006. And it has become deeply controversial over the last five years as the justices have wielded it in ways that at least appear to be inconsistent along any axis other than whether the intervention favors Democrats or Republicans. (I make—and defend—this argument in detail in Chapter 6 of The Shadow Docket.) If anything, the Court’s inconsistency has been on even sharper display this term—given its enthusiastic invocations of Purcell to block district court rulings (in the Texas and Alabama redistricting cases) while refusing to apply Purcell to itself (including in the Alabama case, where the district court’s intervention was only necessary because of the Supreme Court’s post-Callais decision to vacate and remand the earlier injunction in that case).

Indeed, the only thread that unites every single application (or not) of Purcell to date is that the relevant change in election rules came from a local or state government—not the United States. Largely because the federal government tends not to try to interfere even in federal elections, there’s just no history of emergency litigation challenging late-breaking changes to federal rules respecting elections.

Beyond a descriptive claim about the twenty-year history of Purcell, that’s also a normative one. Even taking Purcell at face value, there are two independent reasons why it should not apply when the federal government is the defendant, and a third reason, even for those who embrace Purcell in full, why the Trump administration shouldn’t be able invoke it in the mail-in ballots cases specifically.

I. Purcell as a Federalism Doctrine

In a fascinating 2024 student note in the Virginia Law Review, Casey Schmidt argued (persuasively, in my view) that Purcell doesn’t make sense on its own terms (i.e., the “avoiding voter confusion” argument), and that the more principled way to rationalize at least many of the Court’s interventions in this space is to tie the “principle” to federalism—“the idea that decisions about election procedures are primarily the province of the states.”

Schmidt offered that thesis as a better rationalization of at least many of the Court’s previous Purcell-driven interventions, but I think it was also doing work in those cases themselves. Justice Kavanaugh, for instance, has repeatedly described Purcell as establishing the principle that “federal district courts ordinarily should not enjoin state election laws in the period close to an election”—which suggests that it’s about more than just limits on the equitable authority of federal courts. If it were just about the federal equity power (regardless of whose rules the equity power was being invoked to block), there’d be no reason to emphasize that it applies to state rules, specifically. And in the Texas case in December, the majority cast the district court’s injunction against Texas’s new congressional map (which it stayed under Purcell) as “upsetting the delicate federal-state balance in elections.” That balance is not just a creature of Supreme Court jurisprudence; it is reflected in the Constitution itself, Article I, Section 4 of which makes states the first movers when it comes to the rules for even federal elections. Congress can override state rules, of course, but the key is that there’s at least a loose structural hook for Purcell as reflecting that same federalism-driven default rule.

On that reading, the point of Purcell is not to insulate late-breaking changes to election rules from all judicial review; it’s to suggest that any equitable relief against those changes ought to come from state courts—not federal courts. And the Supreme Court’s behavior during the 2020 election cycle largely backs that up. Although the Court intervened repeatedly to stay federal court injunctions that would’ve altered the status quo with regard to election procedures, it declined to intervene in cases from North Carolina and Pennsylvania where the underlying ruling came from a state court. The issue wasn’t a lack of jurisdiction; those cases involved federal claims as well as state ones.

As Chief Justice Roberts put it in a (rare) concurrence in an October 2020 case from Wisconsin (where the Court refused to vacate the Seventh Circuit’s Purcell stay of a federal district court injunction), “Different bodies of law and different precedents govern these two situations and require, in these particular circumstances, that we allow the modifcation of election rules in Pennsylvania but not Wisconsin.” New York’s highest court sounded a similar theme in 2022: “The Purcell doctrine cautions federal courts against interfering with state election laws when an election is imminent, and does not limit state judicial authority where, as here, a state court must intervene to remedy violations of the State Constitution.” States can choose to follow Purcell; the relevant point is that it’s their choice.

The upshot is that federal law just has nothing to say, one way or the other, about whether state courts may issue eleventh-hour injunctive relief in election cases—because that question is one that’s committed to the states as a matter of federalism. By that logic, it ought to be obvious that Purcell would (and should) not likewise constrain federal courts reviewing election procedures grounded in federal statutes and rules—because there is no federalism concern pressuring the federal courts to stay their hand.

II. Purcell and the Vanishing Forum

A related but distinct point is the availability of alternative remedies. As the North Carolina and Pennsylvania cases from 2020 illustrate, in those states that don’t adopt their own version of Purcell, late-breaking changes to local or state election rules remain subject to injunctive relief in state court—so that application of Purcell to lower federal courts does not foreclose all potential channels for judicial review.

In contrast, state courts generally lack the power to issue coercive relief against federal defendants—practically because any such suit would inevitably be removed to federal court, and likely formally, as well (although Richard Arnold famously argued otherwise). That means that, when it comes to efforts to challenge late changes to election procedures imposed by the federal government, federal courts are the only forum in which those challenges can be brought.

That’s more than just a procedural detail, to me. Applying Purcell to state or local governments simply closes the doors of one courthouse. Applying Purcell to the federal government would close the doors of every courthouse, leaving voters without any remedy for even the most nefarious, malicious, or otherwise blatantly unlawful last-minute federal interventions in elections. A federal rule that is concededly unlawful would nonetheless govern a national election, immune from review anywhere, simply because the government waited long enough to issue it. This asymmetry, of course, is a one-way ratchet. The federal government can always cure a timing problem by acting earlier (as the USPS was supposed to have done, here), whereas voters cannot manufacture a forum that does not exist.

III. Heads They Win; Tails We Lose

Even for those who are not (yet) persuaded, there’s one final argument about Purcell in the specific context of the mail-in ballots cases that is worth articulating—which is the government’s unclean hands in even invoking it. After all, this is the same government that argued, as recently as a week ago, that it was too early for litigants to challenge the Trump administration’s efforts to interfere with mail-in balloting, an argument that succeeded in the Supreme Court. But equity can’t—or at least shouldn’t—abide a party arguing one day that it’s too early for a claim to be brought, only to argue the next day that it’s now too late. Whether one frames that in terms of unclean hands or judicial estoppel, the upshot ought to be the same: insofar as Purcell is, at its core, a doctrine about the equity power, equity should not be used to allow a party to manufacture absurd results. (There’s a cryptic 2022 order in a Georgia election case in which the Court said a version of this—vacating an Eleventh Circuit Purcell stay because of the party’s own representations in the lower court.) A “principle” that turns on who is asking, and when it suits them, is no principle at all.

SCOTUS Trivia: The Court’s Longest Vacancy

For … reasons, I was looking at Supreme Court nominations from the Antebellum Era, and came across a tidbit I’d seen before, but haven’t ever used as trivia: The longest vacancy in the Supreme Court’s history. The answer involves the vacancy created upon the April 1844 death of Justice Henry Baldwin, while President John Tyler was still in office. “His Accidency,” the first person to become president other than by election (he succeeded William Henry Harrison upon Harrison’s death), had alienated his fellow Whigs (who expelled him from the party)—and had virtually no support in Congress. He unsuccessfully tried to fill Baldwin’s seat three different times—twice with Edward King, and once with John Meredith Read. (Tyler also struggled to fill the vacant seat of Justice Smith Thompson, although the Senate eventually confirmed his sixth(!) choice—Samuel Nelson—on February 14, 1845.)

After James K. Polk came to office in March 1845, he offered Baldwin’s seat to James Buchanan—who considered the offer before turning it down in favor of becoming Secretary of State. Polk eventually settled on a relatively obscure Western Pennsylvania judge, Robert Cooper Grier, who was confirmed on August 4, 1846, and sworn in on August 10—841 days after the vacancy was created. To put that number in context, it is almost exactly twice as long as the 422-day vacancy between Justice Scalia’s February 2016 death and Justice Gorsuch’s April 2017 swearing-in, which was itself only the second vacancy to exceed a calendar year since the Civil War (the 391-day vacancy created by Justice Fortas’s resignation and filled by Justice Blackmun was the first).

I hope that you’ve enjoyed this installment of “One First.” If you have feedback about today’s issue, or thoughts about future topics, please feel free to email me. And if you liked it, please help spread the word!

Share

If you’re not already a paid subscriber and are interested in receiving regular bonus content (or, at the very least, in supporting the work that goes into this newsletter), please consider becoming one:

Subscribe now

This week’s bonus issue will drop on Thursday. As ever, I hope you are all staying safe out there, especially from the Purcell principle.

My own work is why I want to take seriously the argument that has been made, with increasing volume over the past week, on behalf of Department of Homeland Security General Counsel James Percival: that what he’s been doing is not meaningfully different from what folks like me have been doing, and that the only difference is whose ox is being gored. Florida Attorney General James Uthmeier put a version of it in a letter to the Florida Bar on Monday: “Criticizing any government official, including the Attorney General of Florida, is not professional misconduct. Judges are no exception.” He’s right about that, but utterly wrong that that’s all that Percival has done. My goal in today’s bonus post is to explain, as carefully as I can, exactly why Percival’s conduct is different in kind, and not just degree, from many of the other criticisms of judges floating out there.

The two key factors, in my view, are what the criticism actually says, and who is saying it. The problem isn’t naming judges; judges sign their opinions (except for Supreme Court rulings on emergency applications, anyway), and naming the author of a ruling you think is wrong is how legal commentary has worked in this country since there were rulings to complain about.

Rather, the problem is two-fold: Percival is a government lawyer, and he’s insinuating bad faith and corruption on the part of the judges he’s naming without any proof. And lest you think the latter is my own subjective criterion, it’s the one Rule 4-8.2(a) of the Florida Rules of Professional Conduct already draws. The rule tolerates harsh criticism; what it forbids is a statement about a judge’s “qualifications or integrity” that the lawyer knows to be false or makes with reckless disregard for its truth. You can call a ruling wrong or even lawless. Telling the world that the judge who issued it was acting in bad faith is a different move, and one any lawyer, but especially a government lawyer, had better have some factual basis for making. And it’s because Percival doesn’t have any proof, and because he holds the job he currently holds, that his comments are so profoundly unprofessional—and warrant our repudiation and colloquial sanction, whether or not they also warrant (and will result in) professional discipline.

For those who aren’t paid subscribers, we’ll be back (no later than) Monday with our continuing coverage of the Supreme Court. For those who are, please read on.

Read more

To understand what’s going on, I’m going to try to tee this up as a progression of eight different points:

First, the executive order at issue has two different sets of provisions under challenge. Section 2 of the EO is directed toward DHS, and deals with the sharing of “citizenship lists” between the federal government and states. Section 3 (the really troubling part) is directed toward the U.S. Postal Service, and is the part that, if fully implemented, would allow the USPS to refuse to deliver mail-in ballots from individuals who don’t appear on state-provided lists (which necessarily assumes both that states will provide them and that the lists are accurate). So keep these two different sections in mind.

Second, there are two different cases at the forefront of challenges to the executive order, both pending before Judge Talwani in the District of Massachusetts. One is California v. Trump (a challenge by blue states); the other is League of Women Voters of Massachusetts v. Trump (a challenge by private organizations/voters). Critically, Judge Talwani has issued injunctions against the executive order in both cases. But Monday’s Supreme Court ruling applies only to one of them—the one brought by the states. For unimportant procedural reasons, the League of Women Voters case moved a bit slower, and Judge Talwani’s injunction in that case remains in effect. Even though the Solicitor General had asked the Supreme Court to reach out and pour cold water on that injunction, too, I don’t see anything in the 10-page majority opinion that speaks to the validity of that injunction one way or the other.

Third, late Friday night, the U.S. Postal Service (finally) released the “Final Rule” that was supposed to be published by the end of July that purports to implement Trump’s executive order. It’s that rule, not the executive order itself, that would be the basis for the USPS messing with mail-in ballots come October and November. The rule itself acknowledges the injunctions in both the California and League of Women Voters cases, and provides it won’t go into effect so long as those injunctions remain in force (as one of them still does). And in the League of Women Voters case, Judge Talwani has already indicated that she is inclined to block that rule, too—for the same reasons she enjoined the executive order. That ruling may come as soon as tomorrow—which is when any opposition to such a ruling from the government is due at 8:00 a.m. ET. (The rule is currently set to be “published” in the Federal Register on Wednesday.)

Fourth, and just to be as clear as possible about this, it’s the still-to-come litigation over the USPS Final Rule that’s going to be critical here. In a moment, I’ll get to how the majority opinion from the Supreme Court may and may not influence the trajectory of the League of Women Voters case, but until and unless the injunction of the executive order in that case, or a subsequent TRO or injunction of the USPS Final Rule itself, is vacated by the First Circuit or the Supreme Court, the status quo (that is to say, no mail-in ballot mischief) will persist. The USPS Final Rule itself can’t go into effect until the League of Women Voters injunction is frozen.

Fifth, and turning to the Supreme Court’s actual ruling, the core of the majority opinion is that the blue states brought their challenge too soon. My own view is that both of the dissenting opinions (one by Justice Sotomayor, joined by Justice Kagan; and one by Justice Jackson) already had the better of this argument before the USPS issued its Final Rule last Friday. But to suggest today that these cases are premature is, to use a technical word, bollocks. It’s not going to matter, because the League of Women Voters case will almost certainly be “ripe.” But it’s still a ridiculous rationalization by the Court that I’ll try to write more about later this week.

Sixth, the “ripeness” part of the opinion likely won’t matter much in the League of Women Voters case. Neither will the standing analysis, since voters dealing with a USPS Final Rule are, in numerous material respects, differently situated from blue states challenging the executive order on its face. The part of the majority opinion that may carry over is its discussion of “irreparable harm.” Justice Jackson, as usual, is right about how the majority is once again perverting that concept (as I noted just this morning in discussing the Chief Justice’s “administrative stay” in the ballroom case). But the majority’s wrongheaded irreparable harm analysis will surely apply to a challenge to the USPS Final Rule, too. That should mean that the inevitable emergency application in the League of Women Voters case comes down to the underlying legality of the USPS Final Rule (which would be bad for Trump), except that…

Seventh, the shadow looming over all of this is the “Purcell principle,” the idea that, as election day draws nigh, federal courts should stay their hand rather than issue coercive orders changing the rules for when and how people vote. Purcell is the source of significant controversy and criticism, almost all of it deserved. Among the charges are that (some of) the justices are ruthlessly inconsistent in when they do and don’t apply Purcell—leading to the worry that an injunction against the USPS Final Rule might be deemed “too close to the election,” and stayed under Purcell. That’s a genuine concern, and given the Court’s own repeated transgressions of the Purcell principle earlier this term in Texas, Louisiana, and Alabama, to hide behind it to allow the USPS Final Rule to go into effect would be as indefensible and anti-democratic a ruling as the Court has handed down in a very long time. That brings me to…

Eighth, there’s still a lot to play for in these cases. My own view is that the Court should have denied the stay, especially with respect to section 3, for the reasons the unanimous First Circuit panel articulated in California (Judge Dunlap dissented only as to section 2). But even with tonight’s stay, section 3 remains blocked; the USPS Final Rule remains frozen; and the real question is going to be what happens with the (seemingly inevitable) TRO or injunction against the USPS Final Rule that’s still to come in the League of Women Voters case.

All of this is to say, yikes, but also, stay tuned. And if you’re not already a subscriber, I hope you’ll consider signing up for more posts like this (and also ones written with a bit more … preparation).

Subscribe now

Share

If you’re not already a subscriber, please consider becoming one—and upgrading to a paid subscription if and when your circumstances permit:

Subscribe now

Late Friday afternoon, Chief Justice Roberts issued a one-sentence administrative stay keeping Judge Richard Leon’s injunction against above-ground construction of the new White House ballroom on hold “pending further order” of the Chief Justice or the Court. (In English, Roberts allowed ballroom construction to continue.) There was no reasoning, no vote count, and no dissent—because, as I’ve explained before, these kinds of “administrative” stays are issued by a single justice (the “circuit justice” responsible for the lower courts from which the case came) to buy time for the full Court to act.

The problem with Roberts’s administrative stay isn’t that it was unexplained; it’s that no defensible explanation for it exists. Other than a March 2024 concurrence by Justice Barrett (which was about administrative stays issued by courts of appeals), the justices have never publicly identified any criteria for an “administrative” stay, as opposed to a stay pending appeal. But irreparable harm has to be at least part of the equation. And there’s just no universe in which a party is “irreparably” harmed by a temporary pause in construction; indeed, it wouldn’t be a construction project without unplanned, temporary pauses.

For quirky procedural reasons, it’s possible that Roberts’s Friday afternoon intervention isn’t necessarily a harbinger of how the full Court is going to rule (I’m skeptical, alas). But whether it is or not, the real takeaway from the ballroom case, at least at this juncture, is that it’s long past time for the justices to publicly articulate criteria for when they should (and shouldn’t) issue “administrative” stays—and, even more importantly, to follow them.

More on all of that below. But first, the (other) news.

On the Docket

As expected, the Court released the second of three scheduled summer Order Lists at 9:30 ET last Monday. The only real news on this one was the entirely expected denial of the petition for rehearing in Trump v. Carroll, with the Court showing no interest in reconsidering its refusal to take up President Trump’s appeal of one of the two monetary judgments E. Jean Carroll has obtained against him.

The only other action from the full Court last week was Tuesday’s denial, over no public dissents, of a stay of execution to Florida death-row prisoner William F. Silvia. And the only other ruling of note was Friday’s administrative stay from Chief Justice Roberts in the ballroom case, about which more shortly.

Besides a full Court ruling in the ballroom case, we’re also still waiting for the Court’s disposition of the two emergency applications (one from the Trump administration) in the Trump v. California mail-in ballots case. Those applications have become rather more complicated by two subsequent developments. First, there’s now a separate injunction by Judge Talwani against the USPS provisions of the executive order in the separate League of Women Voters case (which would keep the most important parts of the executive order on hold even if the Court granted stays in Trump v. California). Second, late Friday, the U.S. Postal Service finally submitted its final rule purporting to implement the executive order (which is scheduled to be published this Wednesday), although the rule itself notes that “the Postal Service will not take actions to implement the rule specifically for the 2026 election unless and until the government obtains relief from those injunctions.”

One of the Trump administration’s central arguments in support of a stay in Trump v. California was that any relief against the USPS would be premature—because there wasn’t yet a final rule. Friday’s development pours pretty cold water on that position, and raises the odds that the Court simply ducks its way out of these cases for now (or that any relief it grants is limited to the DHS provisions, with the injunction against the more important USPS provisions remaining intact). It’s all a mess—one, it should be said, that’s entirely of the Trump administration’s making.1

The One First “Long Read”: The Irreparable Harm of Not Building an Unlawful Ballroom

Background

The dispute that reached the Supreme Court last week as National Park Service v. National Trust for Historic Preservation goes back to the announcement in July 2025 that the East Wing would come down to make way for a 90,000-square-foot ballroom, to be paid for with something like $400 million in (supposedly) private donations. Demolition came last fall, construction began in October, and the National Trust for Historic Preservation filed suit in December.

Judge Leon—a George W. Bush appointee, a fact worth mentioning only because of how readily the White House now describes anyone who rules against it as a Democrat—issued a preliminary injunction against construction on March 31. His holding was narrow and statutory. Congress has appropriated money to maintain the White House; it has appropriated nothing for a ballroom. The workaround—private donations routed through the Park Service and then, by way of the Economy Act, to the Office of the Executive Residence—was in Leon’s phrase a “Rube Goldberg contraption,” and it did not supply the missing authority. Two weeks later, Leon modified his injunction to clarify that it applied only to above-ground construction of the ballroom. His April 16 order carved out all underground work, bunkers and shelters included, along with any above-ground work “strictly necessary to ensure the safety and security of the White House and its grounds.” In other words, Leon structured the injunction to allow construction directly related to the bunker and any related national security needs, but to block the ballroom itself.

The government immediately appealed Leon’s April 16 ruling and sought a stay pending appeal. The next day, the D.C. Circuit issued an “administrative” stay, allowing construction to continue while it considered whether to stay the injunction pending an appeal it also expedited. Then, on August 7, instead of ruling on the stay, a divided D.C. Circuit panel affirmed Judge Leon’s preliminary injunction on the merits.

Writing jointly across 100 pages, Judges Millett and Garcia concluded that Congress has “full constitutional control over federal property” and that the government could point to no “express or implied authorization” for what it was building. National security, they added, is “not an automatic get-out-of-law-free card,” especially given that Leon’s injunction specifically exempted below-ground construction and above-ground construction strictly necessary for national security purposes. Having affirmed on the merits, the panel denied the pending stay motion as moot. That last detail is worth pausing on, because the government has since described itself as having been turned down for a stay by two separate courts. What actually happened is that it lost its preliminary injunction appeal on the merits—which ought to be a bigger deal (and carry more weight) in the Supreme Court.

Judge Rao dissented in the D.C. Circuit, in an opinion that formed much of the basis for the Solicitor General’s subsequent application to the Supreme Court. Her principal objection is to the plaintiffs’ standing. The Trust, she wrote, cannot sue in its own right because this litigation is not germane to its statutory purposes, and the germaneness theory the majority relied upon had, in her view, been forfeited by the parties anyway—an argument pitched squarely at a Court that has lately shown a keen interest in party presentation. She also would have rejected the Trust’s reliance on the putative injuries to Alison Hoagland, the architectural historian whose enjoyment of the White House the lower courts found impaired by the construction. As for the injunction itself, Rao called it “a blatant abuse of discretion”: “The district court seized supervision of construction at the White House, and my colleagues affirm this judicial overreach.” Alas, I don’t think the standing portion of Judge Rao’s dissent is crazy (although that’s partly because of how malleable standing doctrine has become). The merits, on the other hand…

What drove the timing from there was how Judges Millett and Garcia structured the mandate: They stayed their ruling for 14 days to allow the Trump administration to seek Supreme Court review. That clock was set to expire at the end of the day last Friday, which is why, when the Trump administration filed for a stay on August 13, it also requested at least an “administrative” stay by last Friday to preserve the “status quo” (one in which construction was continuing).

In its application before the Supreme Court, the government offered two arguments for why allowing Leon’s injunction to go into effect would cause irreparable harm. One was national security: Solicitor General Sauer described the project as “an integrated military complex” that is “vitally required by national security.” The other was momentum: a “250-person crew working 20 hours a day, 7 days a week,” on a project “65% complete in its entirety, and moving quickly toward total completion.”

The Trust’s response, filed last Tuesday, made the obvious rejoinder to the first theory of irreparable harm. Nothing in Leon’s order “prevents them from continuing work on the bunker, securing the East Wing site, or ensuring the safety of the President, his family, and staff.” The security carve-out is transparent on the face of the (modified) injunction. What the injunction stops is just the ballroom. As for the government’s second irreparable harm argument, the Trust said the quiet part out loud: the government has been trying to “outrun judicial review,” building as fast as it can so that it can then tell the justices the project is “beyond the point of no return.” Those “efforts to foil judicial review and arrogate Congress’s exclusive powers,” the brief argued, “should not be rewarded with a stay that allows [the government] to complete a ballroom [it] lack[s] any authority to commence in the first place.”

Friday’s Order and Its Implications

An administrative stay is supposed to be a housekeeping device—to freeze things so that the Court has time to rule on the application in front of it. Roberts’s order says nothing about the merits and doesn’t purport to, and it stands to reason that he’s already referred the application to the full Court, the full Court has already voted on it, and the delay involves the preparation of opinions supporting or dissenting from whatever that disposition is. It is therefore entirely possible that Friday’s order is simply buying time for the Court by “preserving” what had been the status quo since April 17, i.e., that ballroom construction could continue.

There are two problems, though. The first is that there’d be no need to preserve that status quo if the Court had already voted, behind the scenes, to deny the stay. Indeed, especially given the arguments the Trust has made about trying to “outrun judicial review,” even a brief extension of the status quo would be problematic if the end result was foreordained—and adverse to further construction.

Compounding the first problem is the second—the lack of any published standards to guide when a circuit justice should (and shouldn’t) issue an “administrative” stay, a tool that has become much more common in recent years (Roberts’s Friday order is at least the eighth of the October 2025 Term; as recently as the October 2020 Term there were two). You will not find, anywhere in the Supreme Court’s Rules or in any published opinions, a definition of an administrative stay, a standard for granting one, a cap on how long one may last, or a rule about who gets to decide. Did the Chief Justice balance the equities? Did he form a view about the government’s likelihood of success? Did he conclude that the government faced some interim harm? Or did he simply decide that the justices needed more time, and that freezing the injunction was the tidiest way to buy it? The order tells us nothing, and that’s a problem unto itself.

The only real discussion of administrative stays anywhere in the Court’s recent emergency-docket work is Justice Barrett’s concurrence, joined by Justice Kavanaugh, in United States v. Texas—the March 2024 order that let Texas’s SB4 go into effect while the Fifth Circuit sat on the stay motion. Barrett conceded that “there is no jurisprudence of administrative stays, much less a one-size-fits-all test,” thought the gap was tolerable (“Play in the joints seems appropriate for a measure that functions as a flexible, short-term tool”), and I argued at the time that she had reached the right conclusion for the wrong reason. Two more years have not changed my mind.

But Barrett also conceded three things that should cut directly against Friday’s order. First, these stays are “not value neutral”; they reflect “a first-blush judgment about the relative consequences of staying the lower court judgment versus allowing it go to [sic] into effect.” Second, they must be short: “An administrative stay should last no longer than necessary to make an intelligent decision on the motion for a stay pending appeal.” Third, “[t]he time may come, in this case or another, when this Court is forced to conclude that an administrative stay has effectively become a stay pending appeal and review it accordingly.”

On Barrett’s own account, Friday’s order was not value neutral—it embodies a judgment that halting the ballroom by mistake would be worse than letting any further construction go forward by mistake. Given that the injunction exempts anything relating to national security and/or the safety of the President, his family, and White House staff, I just cannot construct the reasoning that gets anyone there on these facts. It’s also worth flagging that the status quo legally is that there’s a preliminary injunction against the construction that the D.C. Circuit affirmed. Barrett acknowledged that “there is no settled way of defining ‘the status quo,’” but on every candidate she floated—before the challenged conduct, before the lawsuit, before the courts got involved—the answer is not a 250-person crew working around the clock on a structure already 65% finished. Staying Leon’s injunction protects the government’s timetable here, not the Court’s ability to deliberate. And the order is not time limited: Roberts set no deadline, so construction proceeds until the Court says otherwise.

In the SB4 case, the justices were reviewing a court of appeals’ administrative stay, and Barrett’s answer thus also included a paean to institutional restraint, i.e., that the Court should let the Fifth Circuit “be the first mover” on the formal stay factors. Here there is no first mover left. The D.C. Circuit affirmed the injunction and mooted the stay motion. The only stay analysis outstanding is the Supreme Court’s own, on an application it has now had before it for 11 days.

If anything, the fact that the government’s strongest argument on the merits is about standing, and not the legality of the construction, only makes this worse. After all, this is the same Court that held in Biden v. Nebraska—in an opinion by the same Chief Justice—that Missouri could challenge student-loan forgiveness on the strength of an attributed injury to MOHELA, a separately incorporated entity that hadn’t sued and could have sued on its own (a holding I described, quite deliberately, as “lawless” for how it defied settled constraints on standing).

Hoagland’s injury here, in contrast, is conventional standing fare, much closer to the ordinary run of environmental and historic-preservation cases than anything in Nebraska. A Court that swallowed that theory in 2023 is poorly positioned to call this one too thin—and if it does, then the D.C. Circuit’s holding that Congress controls construction on the White House grounds becomes unenforceable by anyone.

To bring things full circle, if the government’s best merits argument is that nobody can sue, and the stay lets construction race to completion while the justices deliberate that issue, then the stay isn’t protecting the Court’s ability to decide the merits; it’s effectively dissolving it. That’s why we’re long overdue for the Court to articulate criteria for administrative stays—to make it easier to have at least some faith that these kinds of interventions, which can have dramatic effects on the litigation in which they’re undertaken, are pursuant to neutral principles, and not just case-specific (or president-specific) whims of individual justices.

And whatever those criteria turn out to be, it’s hard to imagine how they could have been satisfied here—no matter what the full Court does next, or when.

SCOTUS Trivia: The Last August Argument

Speaking of emergency litigation before the Supreme Court in August, this Friday marks 68 years since the last time the Supreme Court heard an oral argument in the eighth month of the year. On August 28, 1958, the justices convened in a rare Special Term—the caption in the U.S. Reports reads “August Special Term, 1958”—to take up the Little Rock school board’s request to postpone desegregation for two and a half years. (“Special Terms” are a pre-1990 relic for when the Court needed to convene after it had adjourned the previous term but before the new one had begun. The Court’s 1990 move to a “continuous” term has mooted the need for them.)

The posture was, of course, an emergency one: the board had won a suspension from Judge Lemley, the Eighth Circuit had reversed on August 18 but stayed its mandate, and school was set to open on September 15. The Court heard argument on August 28 on the motion to vacate that stay, with Solicitor General Rankin appearing by invitation as amicus and urging the justices to vacate it forthwith.

Having concluded that the motion could not be resolved without reaching the merits, the Court set an expedited schedule for plenary review, and set argument on the merits for noon on September 11. It affirmed in a per curiam order filed the next day (September 12), with the full opinion—signed individually by all nine justices—following on September 29, at which point Chief Justice Warren adjourned the “Special Term.”

Every other month on the calendar has had at least one Supreme Court argument since 1958. August is the lone holdout—although, the way the emergency docket is going, it may not be for too much longer.

I hope that you’ve enjoyed this installment of “One First.” If you have feedback about today’s issue, or thoughts about future topics, please feel free to email me. And if you liked it, please help spread the word!

Share

If you’re not already a paid subscriber and are interested in receiving regular bonus content (or, at the very least, in supporting the work that goes into this newsletter), please consider becoming one:

Subscribe now

This week’s bonus issue will drop on Thursday. As ever, I hope you are all staying safe out there, especially from standardless administrative stays.

That’s why the lawsuit that was filed by the State of Minnesota in Brownsville, Texas, on Tuesday deserves more attention than it has received. Minnesota Attorney General Keith Ellison has sued Governor Greg Abbott to compel him to sign a rendition warrant for Christian Castro, a former ICE agent charged with firing into an occupied home in Minneapolis who is currently in state custody in Texas. Abbott is refusing to sign the extradition warrant, at least until Minnesota “step[s] up and repay[s] the United States of America and Americans for all the fraud that they’ve committed in their state through federal programs.”

That’s just not how any of this works. As I explain in more detail below the fold, Abbott has no remotely viable legal argument—and he’s destined to lose Minnesota’s new lawsuit. Federal law provides governors with virtually no leeway to resist valid interstate extradition requests—a point the Supreme Court has reiterated every time the question has come up over the last 40 years. And although some might point to New York’s and California’s refusal to extradite doctors indicted by Louisiana for prescribing abortion-related medication via telemedicine, the constitutionally dispositive distinction is that, unlike Castro and Minnesota, those defendants weren’t in Louisiana at the time of their alleged offenses.

Instead, the messier issue is the timing—the distinct possibility that Abbott, perhaps with some help from the Fifth Circuit (and maybe the Supreme Court, too), is not trying to win, but rather is trying to run out the clock, which may expire as soon as next Wednesday.

For those who aren’t paid subscribers, we’ll be back (no later than) Monday with our regular coverage of the Supreme Court. For those who are, please read on.

Read more

Share

If you’re not already a subscriber, please consider becoming one—and upgrading to a paid subscription if and when your circumstances permit:

Subscribe now

On Thursday, the en banc Fifth Circuit finally ruled on the long-pending Alien Enemies Act case that had reached the Supreme Court in April 2025 as A.A.R.P. v. Trump, but that has since been captioned W.M.M. v. Trump. To make a long story a bit shorter, the full court of appeals held that the important questions the case raised about the President’s authority under that 1798 statute had been mooted by the fact that the government had successfully removed the three named plaintiffs under other, more “ordinary” immigration authorities. Some right-wing commentators hailed the ruling as a “win” for the Trump administration, and in the narrowest sense, that’s not completely inaccurate. Given that it was the petitioners who appealed, the fact that their appeal was dismissed and the reason why it was dismissed (because these individuals were removed from the United States) certainly means that they didn’t “win.”

But I want to use today’s post to argue, to the contrary, that the denouement in W.M.M. quietly but emphatically drives home how badly the Trump Administration lost the Alien Enemies Act litigation writ large—as reflected in its clearly deliberate decision to moot these cases rather than defend the validity of President Trump’s March 14 proclamation invoking the act against members of Tren de Aragua (a Venezuelan gang).

Clever lawyering on the government’s part may have allowed it to avoid a precedential holding that the March 14 proclamation is invalid—which, as I’ll explain below, is a problem unto itself. But the reality is that, ever since the famous planes that didn’t turn around on March 15, 2025, every public indication is that no one has been removed from the United States under the AEA. And as problematic as I believe (and have argued) so much of the Supreme Court’s behavior on the emergency docket over the past 18 months has been, I also have to admit, as I explain below, that the justices were key players in the failure of the Trump administration’s misbegotten AEA enterprise. Folks can spin last week’s Fifth Circuit ruling however they want; the real bottom line is that this is how the AEA litigation seems destined to end—with a procedural whimper, not a substantive bang.

More on all of that below. But first, the news.

On the Docket

There was a single ruling by the full Court last week—Thursday’s denial of a stay of execution to Tennessee death row prisoner Anthony Hines, over no public dissents. Justice Kagan also issued a very brief administrative stay in the latest iteration of the long-running litigation between Apple and Epic Games—one that she dissolved herself in denying Apple’s application in chambers on Thursday.

This week will necessarily be busier—if for no other reason than because we expect the second of the Court’s three “regular” summer Order Lists at 9:30 ET this morning. We may also get a ruling on the two pending emergency applications respecting President Trump’s effort to interfere with mail-in ballots via executive order. And as expected, the administration also filed an emergency application last week (#36 if you’re scoring at home—or even if you’re by yourself) in the White House ballroom case. Besides the substantive significance of that case, the application includes 10 brand-new factual declarations by various government witnesses (beyond the declarations submitted in the lower courts)—raising the recurring (and problematic) issue of the factual “record” that’s before the justices on their emergency docket. (Chief Justice Roberts ordered a response by noon ET tomorrow, so this one may move quickly.)

In less significant news, the Court has also released the argument calendar for its November 2026 sitting—a whopping seven cases over five days. I wrote in April about how the Court could decide the same total number of cases each term without an April argument sitting by filling more of the slots earlier in the term; here’s a good example of those slots sitting open.

The One First “Long Read”:
The AEA Litigation … Ends?

I wrote a lot last spring about President Trump’s invocation of the Alien Enemies Act on March 14, 2025 as part of an apparent attempt to engage in mass, summary removals of at least some non-citizens the administration simply deemed to be members of Tren de Aragua—a Venezuelan drug gang that the government claimed had deep connections to the Maduro regime. Without rehashing all of the details, it may be useful to provide a brief procedural recap—to underscore the Supreme Court’s role in at least indirectly thwarting the administration’s (apparent) plan to utilize the AEA as a way of depriving a large number of non-citizens of the ordinary administrative and judicial process before they could be removed.

Round 1: J.G.G.

The litigation started, of course, in the D.C. federal district court in the case captioned J.G.G. v. Trump. On Saturday, March 15 (the same day that the President’s AEA proclamation was actually made public—by which point efforts were already well underway to remove hundreds of non-citizens from the country pursuant to that proclamation), Chief Judge Boasberg issued a pair of TROs—the first one blocking the removal of the named plaintiffs; and the second blocking the removal of any member of a nationwide class of individuals otherwise subject to the proclamation.

It’s worth pausing for a moment to flag the substantive point: the proclamation is based on the view that the United States is subject to a “predatory incursion” by Venezuela because of drug trafficking by gangs like TdA. Leaving aside the dubiousness of the factual predicate (one that the Trump administration’s own intelligence agencies rejected), the whole theory turns the AEA on its head—since that statute was meant to deal with nationals of countries with which the United States is in an actual war, like German, Italian, and Japanese nationals during World War II. Whatever else one might say about drug trafficking, or about the Maduro arrest operation, we were not then, and are not now, at war with Venezuela.

Anyway, the government defied (yes, Judges Katsas, Rao, and Walker—defied) those TROs by refusing to turn around planes that were already in the air, and it also appealed the TROs and sought emergency relief from them first in the D.C. Circuit and, when that failed, in the Supreme Court.

On April 7, 2025, the Supreme Court stayed Boasberg’s TROs by a 5-4 vote, with Justice Barrett joining the three Democratic appointees in dissent. The actual legal basis for the stay was the majority’s (deeply contestable) assertion that challenges to AEA removals had to be brought as habeas petitions, rather than, as in J.G.G., in a single, nationwide action under the Administrative Procedure Act. And because habeas petitions typically have to be brought in the districts in which the petitioners are actually detained, that meant these cases would have to be re-filed in Texas. As I detailed at the time, however doctrinally inconsistent (and wrong) that conclusion may have been, it had the intended and direct effect of channeling AEA claims into district-by-district habeas petitions.

But perhaps because of the stridency of the dissenting opinions in J.G.G., both the unsigned majority opinion and Justice Kavanaugh’s concurrence went out of their way to preemptively reject the Trump administration’s more extravagant claim—not that these cases had to be brought as habeas petitions, but that the petitioners were entitled to vanishingly little process in challenging their potential removals under the AEA.

Instead, the majority opinion in J.G.G. specifically emphasized that “AEA detainees must receive notice after the date of this order that they are subject to removal under the Act. The notice must be afforded within a reasonable time and in such a manner as will allow them to actually seek habeas relief in the proper venue before such removal occurs.” And Justice Kavanaugh buttressed that point in his concurrence, explaining that “all nine Members of the Court agree that judicial review is available.” In other words, even as the Court (wrongly, in my view) channeled AEA claims into a series of distinct, multi-district habeas petitions, it emphasized that such habeas review would be—and would need to be—meaningful. If anything, the way the Court handled the emergency application in the Abrego Garcia case just days later provided further evidence that the justices were contemplating a vigorous role for judicial review even for those who had already been removed from the United States—and not just a rubber stamp.

Round 2: A.A.R.P.

Less than two weeks after J.G.G., the strength of the Court’s commitment to that kind of meaningful review was tested in the A.A.R.P. case. That case was the lead district-wide habeas class action brought on behalf of AEA detainees in the Northern District of Texas—the only district court to hear such a case that did not immediately enjoin or otherwise block AEA removals pending the notice and hearing that the Supreme Court had required. When it appeared that the government was going to remove a group of detainees from the Northern District without abiding by J.G.G., the Supreme Court issued its only grant of emergency relief to date against the second Trump administration—handing down a short but unusually emphatic order shortly before 1:00 a.m. ET on Saturday, April 19 that directly barred the removal of any member of the putative class (that is, any AEA detainee in the Northern District of Texas) while that litigation proceeded. The order came over only two public dissents (from Justices Thomas and Alito), and was backstopped with a more complete opinion on May 16.

As I suggested at the time, the Court’s machinations in A.A.R.P., like its ruling in J.G.G., had two sides. On one hand, here was the Court moving quickly and aggressively to preserve its own mandate in J.G.G., and to communicate, as clearly as it could, that it meant what it had said on April 7 about notice and an opportunity to be heard. On the other hand, rather than pretermitting all of this messy, chaotic, and high-stakes litigation by addressing the substantive merits of President Trump’s March 14 proclamation, or saying anything about the government’s misbehavior in J.G.G., the Court’s moves had been entirely procedural—ensuring a meaningful role for the courts (and the Court) at some future point in answering that question.

Round 3: W.M.M.

With A.A.R.P. quashing any thought of trying to remove folks under the AEA anyway, the Trump administration pivoted to litigating those questions—and to doing so only in the Fifth Circuit (since, under J.G.G., those questions could only be addressed where AEA detainees were being held). This led to the three-judge Fifth Circuit panel’s decision in W.M.M. (what had been the A.A.R.P. case) on September 2, 2025. Over a … lengthy … dissent from Judge Oldham, the court of appeals, in an opinion by Judge Southwick, held that the proclamation was in fact unlawful—principally because Tren de Aragua’s activities did not constitute an “invasion” or “predatory incursion” absent at least some military action.

At that point, rather than take the case to the Supreme Court, the Trump administration decided, instead, to seek en banc review in the Fifth Circuit. That move was another in a series of efforts by the Solicitor General’s office to avoid Supreme Court review on the merits of the President’s AEA proclamation—or, at the very least, to delay it. It’s one thing to delay such review when you’re winning. But recall that the status quo here was a de facto nationwide block on AEA removals. In that posture, not trying to get before the Supreme Court faster was a pretty clear tell. Indeed, it certainly appeared that the government had decided that it wasn’t likely to prevail on the merits if and when the case returned to the Supreme Court—and so the best-case scenario was to (1) vacate the adverse, precedential Fifth Circuit panel ruling; and (2) buy time to pursue other avenues for removing the named plaintiffs in W.M.M. Otherwise, there was no reason to seek further review in the Fifth Circuit rather than immediately petitioning for certiorari in the Supreme Court.

Unsurprisingly, the Fifth Circuit was perfectly happy to oblige—granting rehearing and vacating the panel opinion on September 30, 2025. As last Thursday’s unsigned ruling by the en banc court of appeals explained, “after oral argument before the en banc court, we learned that the government has removed all three named petitioners under the Immigration and Nationality Act—that is, not the AEA.” The result was to moot these petitioners’ appeals. And although the ACLU had asked the Fifth Circuit to substitute other detainees (who are still in U.S. custody) to allow the case to proceed, the court of appeals denied that request in a separate order (albeit without prejudice to the district court allowing the same).

To be sure, the dismissal was accompanied by a pair of thinly-veiled Supreme Court auditions qua concurring opinions from Judges Oldham and Ho trying to salvage the government’s position on the merits (with Oldham reupping his panel dissent and Ho going off on “invasions”). But the fact that those (deeply problematic) analyses came as separate concurrences proves the point: the majority of the en banc court went the procedural route and dismissed the petitioners’ appeals as “moot” rather than giving any credence to the substance of the government’s position.

So now what? I’d be very surprised if the three specific petitioners in W.M.M. try to pursue this litigation any further. Of course, the ACLU may well seek to have other plaintiffs substituted in the district court to continue the challenge, but if the government is bent on mooting those cases, too, it will likely be able to do so. There is also, of course, the ongoing litigation in the D.C. district court over the government’s defiance of Chief Judge Boasberg’s TROs in the J.G.G. case. But on the merits, that case, too, has long-since been mooted. And so far as I know, there are no other live, pending cases in which the government is appealing an adverse AEA ruling by a district court. For now, at least, this appears to be the end of the line for President Trump’s March 14 proclamation—and for whatever plans the Trump administration had for using the AEA as part of its immigration policy. (And for the conspiracy theorists among you who might worry that this was all just a ploy to get out from under the Supreme Court’s A.A.R.P. rulings, specifically, but not to give up on AEA litigation at least for now, let me just say that I don’t think the justices would be especially … impressed by such a maneuver.)

The upshot of all of this, of course, is that there is no adverse, precedential ruling holding that the proclamation was unlawful—and so there’s nothing to stop this administration (or a future one) from trying to rely on this proclamation, or the theory behind it, in future cases. In that respect, the AEA litigation has become the poster child for what Professor Andrew Coan has called “the appellate void”—when the government can effectively frustrate appellate review of rulings blocking a policy simply by not appealing them. Without universal injunctions (mostly foreclosed by the Supreme Court in CASA), nationwide APA relief (expressly foreclosed in the AEA context by J.G.G.), or nationwide class actions (implicitly foreclosed in the AEA context by J.G.G.), the government can litigate in a number of district courts, and then not be bound in any other cases when it loses those. As I suggest in my forthcoming Harvard Law Review Foreword, the Supreme Court bears direct responsibility for exacerbating this phenomenon over the last 18 months, and the AEA litigation is the textbook illustration of how it has done so.

And yet, for as unsatisfying as the AEA litigation has been substantively, it’s worth coming back to the practical bottom line: procedurally, it has been wildly successful from the perspective of the countless non-citizens who might otherwise have been subject to AEA removals. Yes, some of those folks, like the petitioners in W.M.M., have been removed anyway, but only under authorities that were already established. It’s not a “win” for the Trump administration to use its existing powers; if anything, that proves only that the extraordinary claim of new authority may never have been necessary—to say nothing of whether it’s legal.

Instead, the bottom line is that not a single person has been removed from the country under the AEA since March 15 of last year. The first three weeks of that can be traced directly to Chief Judge Boasberg’s TROs. But credit for the remainder has to be given to the Supreme Court, including the three justices in the majority in both J.G.G. and A.A.R.P. (the Chief Justice, Justice Gorsuch, and Justice Kavanaugh)—who went out of their way to (1) insist on notice and a meaningful opportunity to be heard in J.G.G.; and (2) back that up in A.A.R.P. Those were clearly meaningful moves in retrospect, even if their formal effect was to put off (potentially, indefinitely) conclusive judicial review of a patently lawless executive action.

The Court’s more ardent defenders have already held up the AEA litigation as one of the most meaningful examples of this Supreme Court standing up to Trump. I’ve always thought that claim was both slippery (compared to what?) and badly overstated given what the Court didn’t do in either J.G.G. or A.A.R.P. (i.e., reach the merits). But even if that claim is overstated, the trajectory of the AEA litigation since April 2025, potentially culminating in last Thursday’s en banc Fifth Circuit ruling in W.M.M., increasingly demonstrates that it isn’t wrong.

SCOTUS Trivia: A Correction on Last Week’s Trivia

I wrote last week about President Washington’s August 1795 recess appointment of John Rutledge as the second Chief Justice—and how Rutledge doomed his confirmation chances by giving a fiery speech in opposition to the Jay Treaty. But I got something importantly wrong—writing (in a brain fog, apparently) that Rutledge had been the first recess appointment to the Supreme Court.

In fact, that “honor” goes to Justice Thomas Johnson, the first justice to be appointed to the Court after the “original six” (take that, NHL!).1 As my friend (and incomparable Supreme Court historian) Maeva Marcus pointed out, Johnson received a recess appointment in August 1791 (although he didn’t take the oath until September 19), before being formally confirmed by the Senate on October 31. Indeed, Johnson’s recess appointment may be the answer to another puzzle—why the Supreme Court itself has apparently changed its mind about who the shortest-serving justice was, with the Court’s website now (correctly) identifying James F. Byrnes after it had, for a time, identified Johnson.

But this also gives me a chance to share my favorite Thomas Johnson trivia—the letter he sent to President Washington when he resigned from the Court so shortly after joining it. Like many of his contemporaries, Johnson hated circuit riding—and he wasn’t afraid to let Washington know about it, writing, among other things, that “I cannot resolve to spend six Months in the Year of the few I may have left from my Family, on Roads at Taverns chiefly and often in Situations where the most moderate Desires are disappointed: My Time of Life Temper and other Circumstances forbid it.”

Those were the days.

I hope that you’ve enjoyed this installment of “One First.” If you have feedback about today’s issue, or thoughts about future topics, please feel free to email me. And if you liked it, please help spread the word!

Share

If you’re not already a paid subscriber and are interested in receiving regular bonus content (or, at the very least, in supporting the work that goes into this newsletter), please consider becoming one:

Subscribe now

This week’s bonus issue will drop on Thursday. As ever, I hope you are all staying safe out there, especially from unlawful executive proclamations.

Rather than skip a week, though, I thought I’d try something a little different and a lot lighter—a more meta, pre-written post about the history of the newsletter as we are (quickly!) approaching our 500th post (this is #471); and, slightly more slowly, our fourth anniversary. Taking a page from my Letterman-watching youth, I thought I’d do a top-ten countdown of the most widely read posts in this newsletter’s history, drawn from the readership numbers on my end of the Substack dashboard. Mostly, I was just curious to see what the actual answers were (I knew what was #1, but I just don’t check this data often enough to keep tabs on the rest).

Three brief caveats. First, the dashboard data counts cumulative view totals (as of Thursday, August 6)—so they’re not adjusted for how long a post has been up; something from 2023 has had a lot more time to find readers than something from this spring. Second, I’ve counted only written posts, which means the audio episodes and Substack Lives are out, including one that would otherwise have landed at number four. Third, although it should be obvious, views measure attention, not quality (or even whether folks read the whole thing). Some of the posts I’m proudest of are nowhere close to this list (especially some of the earliest posts of the newsletter—when the subscriber base was much … smaller), and at least one entry below is one that I banged out in about ninety minutes (no spoilers).

One more thing worth saying up front, because I suspect it’ll jump off the page: all of the top ten posts were published after January 20, 2025 (which is why I included, as an honorable mention, the pre-2025 post with the most total views). At the end, I reflect a bit upon why that’s happened—and why it’s meaningful.

For fun, I’ll put #s 10-6 above the fold, and #s 5-1 (and the honorable mention) below the fold. Away we go (links to the posts are embedded in the titles if you want to go back and read/re-read them):

#10: 151. The Supreme Court’s (Alien Enemies Act) Patience is Wearing Thin

127,090 views — May 16, 2025

This Friday afternoon post followed up on my earlier state-of-play post about the many cases challenging the effort to use the Alien Enemies Act of 1798 to remove, on a mass and summary basis, individuals the government claimed were Venezuelan members of Tren de Aragua. This post focused on the “per curiam” ruling the Supreme court handed down in the A.A.R.P. case (after the initial order the justices had issued after midnight on Saturday, April 19), especially the majority’s pointed rejection of the dissent’s characterization of what happened in that earlier ruling—a pretty sharp elbow by the Court’s standards—and on the continuing erosion of the presumption of regularity in the Court’s Trump-related rulings.

#9: 145. Justice Alito’s Misbegotten Dissent in A.A.R.P.

127,561 views — April 21, 2025

Speaking of A.A.R.P., this post took a closer look at the dissent Justice Alito filed from the original, April 19 order (somewhat after the fact), joined by Justice Thomas. My view was that the tone was relatively mild for an Alito emergency-docket dissent, but that the substance bespoke a justice grasping for ways to justify not granting the relief the applicants sought rather than genuinely struggling with the relevant procedural questions—and that it portended nothing good for those justices’ votes in future emergency applications.

#8: 200. Five Questions About the Maduro Arrest Operation

134,803 views — January 3, 2026

This is one of the “five questions about” explainer posts I’ve written a bunch more of since last January. I wrote this one while there was still a great deal we didn’t know about the military operation in Venezuela that appeared designed to secure the arrest of Nicolás Maduro and his wife, both apparently under indictment along with four co-defendants. I worked through the administration’s apparent theory that the operation was in support of extraterritorial criminal arrests, and through the objections Maduro would likely raise—head-of-state immunity, foreign official immunity, and the act of state doctrine among them.

#7: 208. The Fifth Circuit Jumps the Immigration Detention Shark

142,356 views — February 7, 2026

Rather than rehash the background and legal arguments I had covered in my December 2025 overview of the broader debate over the Trump administration’s attempt to dramatically expand “mandatory” immigration detention, I used this post to explain two things I wasn’t sure would be well-covered elsewhere: how the administration got this nationwide issue in front of the Fifth Circuit specifically, and what would happen next. Both answers ran through the Supreme Court. Reno v. AADC and Trump v. J.G.G. together closed the door to classwide relief, which is why more than 15,000 individual cases (and counting) have had to be filed challenging the mandatory detention policy.

#6: 156. Federalizing the California National Guard

143,140 views — June 8, 2025

I put out an extra issue that evening in response to the presidential memorandum federalizing 2,000 members of the California National Guard to support DHS activities in and around Los Angeles. My central point was that, even if it was validly invoked (and the Supreme Court would suggest in December that it wasn’t), 10 U.S.C. § 12406 supplies no substantive authority the federal government didn’t already have, so nothing in the memorandum let those troops conduct their own immigration raids or make their own arrests—only to protect federal personnel and property.

***

For those who aren’t paid subscribers, we’ll be back Monday with our continuing coverage of the Supreme Court. For those who are, please read on for the top five and the honorable mention:

Read more

Share

If you’re not already a subscriber, please consider becoming one—and upgrading to a paid subscription if and when your circumstances permit:

Subscribe now

Five weeks after losing at the Supreme Court, President Trump on Thursday signed two new executive orders once again purporting to limit birthright citizenship. Trump has made no bones about his displeasure with the decision in Trump v. Barbara (which he loudly promised to ask the Court to rehear, only to … not follow through). But his and the White House’s efforts to describe these new orders as mere “adjustments” are, perhaps not surprisingly, deeply misleading. The new executive orders reflect a series of efforts to attempt to extend, solely by presidential fiat, the scope of the narrow exceptions to birthright citizenship that the Court reaffirmed on June 30.

As I explain below, although some parts of these new executive orders are likely to survive judicial review either because (1) they’re conditional; or (2) they merely restate what existing law already provides, the important parts are in direct conflict not just with the Fourteenth Amendment itself, but with Barbara’s reaffirmation of the breadth of the Citizenship Clause.

More on all of that in a minute. But first, the news.

On the Docket

The Court, at least as such, made virtually no news last week. There were no decisions by the full Court, and no new major emergency applications were filed. We are, of course, still waiting for the Court to rule on the emergency applications filed by the Trump administration and a group of red states in the mail-in ballots litigation—which I covered in more detail here, and which have been fully briefed since Tuesday. It’s entirely possible that we get that ruling sometime (anytime) this week, although it’s just as possible that we … don’t.

Speaking of emergency applications, we may also get one from the Trump administration this week in the ballroom case, given the 2-1 ruling by the D.C. Circuit on Friday reaffirming “(i) Congress’s full constitutional control over federal property and the funding of its maintenance and development, (ii) the absence of any express or implied authorization for the ballroom construction, and (iii) express statutory limitations that the ballroom construction violates.” In the same ruling, the panel denied as moot the Trump administration’s request for a stay pending appeal, but it also paused its ruling for 14 days to give the government time to ask the Supreme Court for the same relief. After a relatively quiet period with regard to emergency applications from the Justice Department, we may be entering a new, busy one.

Finally, although it’s not really news, Justice Alito gave another “interview” in the Wall Street Journal with James Taranto. This one doesn’t have quite the same stunners as the one from 2023 (in which Alito argued, with a straight face, that “no provision in the Constitution gives [Congress] the power to regulate the Supreme Court”). But it’s still an interesting read, even if it’s also a good reminder of how infrequently the justices are confronted with questions they don’t want to answer. (Now there’s an idea for a future post…)

The One First “Long Read”: Déjà Vu, All Over Again

On June 30, in Trump v. Barbara, a 6-3 majority of the Supreme Court affirmed that children born in the United States to parents who are unlawfully or temporarily present are “subject to the jurisdiction” of the United States and are citizens at birth. Justice Kavanaugh concurred in the judgment and dissented in part; Justices Thomas, Gorsuch, and Alito each dissented.

More important than the vote count is the reasoning. Chief Justice Roberts grounded the Citizenship Clause in the common-law rule the Reconstruction Congress meant to constitutionalize: “the status of citizenship” is “fixed by the place of nativity, irrespective of parentage,” subject to a limited exception for “the children of foreign ministers, and of other persons who may be within our territory with rights of extraterritoriality.” Those exceptions were limited (to three) and narrow: children born in lands the sovereign did not control, children born in areas temporarily outside the sovereign’s control, and children of foreign ministers—the last “by a fiction of extraterritoriality.” The Court also said, specifically, that aliens who come here for “business or pleasure” get no exemption from U.S. jurisdiction, and that their children born here are citizens.

What ties those exceptions together is sovereignty—not on the part of the non-citizen, but on the part of the U.S. government. Each asks whether the United States actually had legal authority over the place and the person at the moment of birth. That was the whole point of the Reconstruction rule. It replaced tests based on parentage or on somebody’s judgment about worthiness with a territorial rule officials would have little room to manipulate. Being “subject to the jurisdiction” of the United States meant being born inside a legal order that could protect you, regulate you, and punish you, no matter who your parents are.

On August 6, the President signed two orders. The first, “Continuing to Protect the Meaning and Value of American Citizenship,” directs federal agencies not to issue or accept documents recognizing citizenship for certain U.S.-born children where neither parent is a citizen. The second, “Ending Birth Tourism,” directs State and DHS to deny visas to suspected birth tourists. Agencies have 30 days to issue implementing guidance.

Everything in the first order hangs on six words from Barbara: “for whom no extraterritorial fiction applie[s].” The White House fact sheet says the new categories are exceptions “the Supreme Court recognized.” Here is the sentence they come from: “Any child who was born ‘under the protection of’ the United States—that is, any child for whom no extraterritorial fiction applied—was made a citizen.” Roberts wrote that to explain why the rule is broad. The order wrongly reads it as authority to make the rule narrower.

Before getting to the unconstitutional parts, let’s start with the parts of the orders that don’t change anything. This is an important baseline, because it allows the administration to say that “some” of the order is “merely” reaffirming settled law:

The accredited-diplomat core. Barbara affirmatively reaffirmed an exception for “the children of foreign ministers,” and existing regulations already implement it: under 8 CFR 101.3, a child born here to “a foreign diplomatic officer accredited to the United States” is not subject to U.S. jurisdiction and is not a citizen. The reason is not complicated. Accredited diplomatic agents are immune from U.S. criminal jurisdiction and largely from civil and administrative jurisdiction, so American law does not reach them in the ordinary way. On this much the order restates settled law, and nobody will sue over it. The problem is how it purports to build out from the foreign ministers themselves. More on that below.

Territories. The order also reaches births in U.S. territories and territorial waters where citizenship isn’t conferred by statute. As of today, that’s exactly one territory: American Samoa. And as I’ve written about before, both the Tenth and D.C. Circuits have held (wrongly, in my view) that the Citizenship Clause doesn’t apply to American Samoa, either (see also this 2017 Harvard Law Review Note)—which means the new executive order merely reflects existing law. The other territories are all covered—Puerto Rico under 8 U.S.C. § 1402, the U.S. Virgin Islands under § 1406, Guam under § 1407, and the Commonwealth of the Northern Mariana Islands under the 1976 Covenant. Congress could repeal those statutes, of course. But the executive order doesn’t (and can’t) do it for them.

Thus, with respect to the territories, the order does … nothing.

Just about everything else is attempting to enlarge, by executive order, one of the existing historical exceptions. The foreign-employee provision widens the diplomat exception. The alien-enemy provision widens the enemy-occupation exception. The surrogacy provision tries to manufacture an extraterritorial fiction out of a private contract. The fraud provision tries to turn misconduct (which could provide the basis for denaturalization) into an absence of jurisdiction. Under Barbara, each should turn solely on where the birth happened and whether American sovereign authority applied at that time. It should not turn on whether officials can pin a disqualifying label on a parent.

“Foreign government employees” beyond accredited diplomats. This provision is easiest to mistake for existing law. It is not. The order does not stop at those diplomats who are outside the jurisdiction of the United States because they enjoy diplomatic immunity. Rather, it purports to create a broader “foreign government employee” category, including an embassy employee who is a national of the country he works for. It also reaches employees of international organizations with international-organization immunity—never mind that international-organization immunity is meaningfully distinct from diplomatic immunity.

This all sweeps well past the current rule. Under settled practice, a child born to a consular officer, a locally hired embassy staffer, or an administrative employee without full diplomatic immunity is a U.S. citizen at birth. So the substitution matters. The existing exception rests on actual immunity from American law; the order rests on foreign employment. But the extraterritorial fiction exists only because the parent is formally beyond the reach of American courts and police. A consular officer who can be prosecuted, sued, and regulated here is subject to U.S. jurisdiction in precisely the sense that Chief Justice Roberts meant in Barbara. As applied to everyone other than the foreign ministers with diplomatic immunity, the order would violate the Fourteenth Amendment.

Surrogacy. The order also purports to deny citizenship to children born in the United States whose parents “engage in a commercial transaction with a surrogate present in the United States, or a territory of the United States, to give birth.” This may be the weakest provision in the package. The problem is not only that surrogacy appears nowhere in the historical exceptions. The trouble runs deeper: it has nothing to do with why those exceptions exist in the first place. A surrogate physically present in the United States is subject to American law in every respect (unless, I guess, the surrogate is a foreign ambassador with diplomatic immunity). Again, the order reasons by analogy: diplomats are treated by fiction as outside territorial jurisdiction, so perhaps a private contract can do similar work. But the constitutional question is not whether an arrangement is commercial or even evasive. It is whether the birth happened under American sovereign authority. To sustain this, the government would have to convince a court that a surrogacy contract does the same jurisdictional work as diplomatic immunity. That was a non-starter even before Barbara.

“Commercial transactions” and fraud. The order also covers parents who “engaged in a commercial transaction to purchase or access birthright citizenship.” That includes paying to get the mother here to give birth. The theory seems to be that fraud somehow generates an extraterritorial fiction. But again, we come back to the same problematic move: A rule about people formally outside American sovereign authority becomes a rule about people who allegedly lied their way in. But Barbara held that visitors here for business or pleasure remain subject to U.S. jurisdiction. A child born in Houston is no less within American authority because of what her mother paid a middleman (like, you know, a travel agent).

A mother’s intent. The order also reaches babies of mothers who lie about their reasons for visiting while pregnant. Consider what that means in practice. A child’s citizenship would depend on a federal official’s after-the-fact judgment about what her mother was thinking at a port of entry sometime before the child was born. Folks have already pointed out that it’s not clear how any of this could be enforced. This isn’t a drafting glitch; it’s what happens when constitutional status rests on an unprovable fact about somebody else. To be sure, there are existing authorities to pursue denaturalization of naturalized citizens who obtained their citizenship through fraud. But even in the most supportive hypothetical, a child born on U.S. soil isn’t obtaining their citizenship from their parents’ fraud; they’re obtaining their citizenship from being born on U.S. soil. They’re not naturalized citizens; they’re birthright citizens.

“Alien enemies.” Finally, this last category deserves more care than it has received, because the exception does have a genuine historical pedigree. Barbara acknowledged an enemy-related exception. It discussed the old rule that if enemies “possess” a “castle or fort,” children born there were not born under the sovereign’s allegiance. So a child born to British parents in an area under British control during the War of 1812, say, would not have been a birthright citizen (even if the Citizenship Clause existed then).

But the order tries to dramatically expand this category to include any parent who belongs to a designated Foreign Terrorist Organization under 8 U.S.C. § 1189, or who is a Specially Designated Global Terrorist under IEEPA and Executive Order 13224. There is a massive gap between the foreign-occupation exception to birthright citizenship and membership in a designated FTO or designation as an SDGT. Indeed, it is entirely possible for groups/individuals to end up on those lists without doing anything in the United States. Even knowingly funding an organization the individual has reason to believe may be engaged in acts of international terrorism is still not within a country mile (or, really, ocean) of being an occupying power on U.S. soil.

Just to put the matter directly, part of why the British child during the War of 1812 couldn’t be a citizen is because he couldn’t be arrested, prosecuted, sanctioned, and deported; he was outside U.S. jurisdiction both formally and practically. Given how regularly the government uses the FTO and SDGT designation processes to impose consequences on U.S. soil, it should be obvious how these processes involve an entirely different category of behavior than what the historical exception was meant to encompass. This goes well past nibbling at the margins; it’s taking the historical exceptions and twisting them into pretzels.

***

It’s pretty obvious that these expansions can’t be reconciled with Chief Justice Roberts’s opinion in Barbara. It’s worth noting, though, that they also can’t be reconciled with Justice Kavanaugh’s opinion concurring in the judgment and dissenting in part. Kavanaugh may believe that Congress can expand the exceptions by statute, but his opinion was unambiguous that the President can’t do it by himself—which is why he agreed that the January 20, 2025 executive order is unlawful (because it’s inconsistent with the current statutes respecting birthright citizenship). The same would be true here.

If the relevant parts of the executive order are almost certainly to be struck down by the Supreme Court, then why go through the motions? Indeed, Stephen Miller called this one of the most important immigration reforms of the presidency.

I have two theories. The first is procedural: Unlike with regard to the January 20, 2025 birthright citizenship order, it’s not as immediately clear who can sue to challenge at least some of these provisions—e.g., the efforts to expand the alien enemy exception. And even with plaintiffs with standing, Trump v. CASA sharply curtailed universal injunctions, so a plaintiff who wins may protect only herself unless she can proceed as a class (which, again, may be only one provision at a time), or find another route to systemwide relief. Barbara itself was a class action, but a challenge to this order may take five or six class actions on behalf of different groups (or an attempt at universal relief by states, who would be stuck trying to figure out how to implement this mess if it ever goes into effect). I still think this will get sorted, but it’s more complicated—and the administration may be planning to battle adverse lower-court rulings on procedural grounds, as much as on substance.

My second theory is that at least part of the goal isn’t to win here, but to continue moving the Overton Window on citizenship—and to more generally try to muddy the waters when it comes to describing categories of non-citizens. Diplomats become indistinguishable from foreign employees. Enemy occupation gets blurred with statutory sanctions listings. Diplomatic extraterritoriality somehow morphs into a template for surrogacy contracts. And sovereign jurisdiction is converted into a subjective after-the-fact judgment about fraud or a mother’s motives. Each move uses a specific historical exception as a justification for executive branch line-drawing. If that works, citizenship stops being a fact fixed at birth and becomes an administrative conclusion reached afterward based upon shifting executive branch priorities. The question would no longer be whether a child was born under American sovereignty. It would be whether officials can pin a disqualifying label on her parent. It’s ugly; it’s wrong; and it’s already succeeding more than I would’ve expected 18 months ago.

That brings me to one last point: President Trump defended all of this by saying the Fourteenth Amendment “was for the babies of slaves” and that people are now building businesses around it, and by claiming the United States is “just about the only country, period,” that does birthright citizenship. The second claim is utterly false; there are at least 32 other countries that recognize a similar rule, and 50–60 more that recognize birthright citizenship in at least some form.

The first claim attempts to be a history lesson, which is what makes it so dangerously misleading. The Citizenship Clause was indeed written to repudiate Dred Scott, the 1857 Supreme Court decision holding that Black people could never be citizens. But the Reconstruction Congress did not swap one caste rule for a narrower one; it extended the promise to “every free-born person in this land.” The idea wasn’t just to retroactively protect “the babies of slaves”; it was to prospectively forestall all future efforts to deny citizenship based on whatever other non-sovereignty-based criteria might be politically popular in any given moment. That is the principle the Supreme Court just reaffirmed in Barbara—a principle that is fundamental to what it means, constitutionally, to be an American, no matter what President Trump, Stephen Miller, or even certain Supreme Court justices might prefer.

SCOTUS Trivia:
The First SCOTUS Recess Appointment

I’ve written before about the phenomenon of “recess” appointments to the Supreme Court—when presidents have temporarily installed justices during formal recesses of the Senate. The first of those happened 231 years ago this Wednesday, when President Washington appointed John Rutledge to be the nation’s second Chief Justice (so that Rutledge could preside over the Court’s August term). In the process, Rutledge became the first justice to ever serve two non-consecutive terms on the Court (Charles Evans Hughes is the only other person to do it).

Unbeknownst to Washington, in late July, Rutledge had delivered a blistering attack on the Jay Treaty with Great Britain at a public gathering in Charleston, South Carolina. In the speech, he said, among other things, that he would rather see the President die than sign it. The Senate had ratified that treaty three weeks earlier, by the barest two-thirds margin. The Federalists who supplied those votes were in no mood to hear it trashed by the man now sitting in Jay’s old chair. Soon, treaty supporters spread word that the speech was proof Rutledge was losing his mind, and pressed Washington to withhold the commission. Washington kept his promise anyway, and Rutledge took his seat.

Once the Senate returned in December, however, it voted 14–10 against him—the first time it ever refused a Supreme Court nominee. After 138 days, Rutledge resigned on December 28 (even though he probably could have served until the end of the Senate session the following year—when his recess appointment expired), and he remains both the shortest-serving Chief Justice and the only recess appointee to the Court whom the Senate declined to confirm.

I hope that you’ve enjoyed this installment of “One First.” If you have feedback about today’s issue, or thoughts about future topics, please feel free to email me. And if you liked it, please help spread the word!

Share

If you’re not already a paid subscriber and are interested in receiving regular bonus content (or, at the very least, in supporting the work that goes into this newsletter), please consider becoming one:

Subscribe now

This week’s bonus issue will drop on Thursday. As ever, I hope you are all staying safe out there, especially from unconstitutional executive orders.

Then, on Monday, Florida Attorney General James Uthmeier asked Congress to open an impeachment inquiry into Eleventh Circuit Judge Nancy Abudu, not for anything she has done as a judge, but for what he suspects she knew during her prior employment at the Southern Poverty Law Center. His theory appears to be that a sitting judge “should not be able to escape impeachment for wrongful conduct occurring prior to her appointment,” even though the evidence of both SPLC’s “wrongful conduct” and Abudu’s responsibility for it both … leave something to be desired.

The complaint about Burroughs is, at bottom, a complaint about procedure: whether a district judge may use an unlabeled interim device to hold a policy in place while she sorts out whether a claim can survive. That’s a fair question, and it’s one I’ve been writing about for years—because, contra what so many right-wing lawyers and commentators have been insisting, district judges did the exact same thing during the Biden administration. If anything, their behavior was, in various respects, even more troubling.

One can say the same thing about the rest of the bill of particulars currently being leveled against the lower courts: From litigants shopping for friendly judges to judges allowing suits against the federal government to go forward based on thin standing arguments to judges issuing creative forms of relief, most of the procedural critiques one hears from the right these days describe behaviors that we saw at least as much—if not more—of between 2021 and 2025. I was critical of much of that behavior then (unlike many of today’s critics) and I remain opposed to much of it today. But the one thing it isn’t is “novel.”

The Abudu letter is a different animal. It isn’t about a ruling; Uthmeier doesn’t cite one. He’s arguing that a sitting federal judge should be removed from office because of where she worked before she was confirmed. And by pretending that the real story is the not-so-novel behaviors we’re seeing in lower courts, right-wing commentators are obscuring what is new—the unprecedented volume of defiance of court orders; the sustained efforts to attack and delegitimize lower federal court judges; and the very real threats that those efforts have engendered. (If you haven’t already read Joanna Slater’s story about Chief Judge McConnell in yesterday’s Washington Post, you should.) In short, we’ve gone from criticizing rulings to seeking to impeach the judges who write them, filing misconduct complaints against them, and listening to the Deputy Attorney General announce that his Department is at “war” with the district courts.

The purpose of today’s post is to document both of these points—how the procedural critiques so often being leveled from the right these days all relate to behaviors that were just as prevalent, if not more problematic, during the Biden administration; and how what’s actually different about this moment is the scale (and source) of these sustained efforts to undermine so many different federal judges, if not the lower federal judiciary itself. I know you’re shocked to hear this, but the same folks who said nothing when district courts engaged in these procedural behaviors during the Biden administration (and who now claim they’re “novel”) have had remarkably little to say about what really is novel—the profoundly dangerous delegitimization campaign in which they are, however unwittingly, necessarily participating.

For those who aren’t paid subscribers, we’ll be back Monday with our regular coverage of the Supreme Court. For those who are, please read on.

Read more

Share

If you’re not already a subscriber, please consider becoming one—and upgrading to a paid subscription if and when your circumstances permit:

Subscribe now

I’ve heard a lot over the past few weeks, especially from conservative defenders of the Supreme Court, about just how “independent” the Court is—how the fact that it ruled against President Trump in (exactly) four significant cases this term should somehow serve as a complete defense to all criticisms of the Court’s behavior, and should stand, instead, as conclusive proof of the Court’s legitimacy. In a New York Times roundtable in which I participated, University of Chicago law professor Will Baude suggested that “this is one of the most independent courts I can imagine at this stage of the second Trump administration.” Cato’s Doug Bandow devoted a column to “The Virtue of Judicial Independence.” A USA Today columnist predicted that “[s]uch independence will surely be forgotten the next time they rule in the president’s favor by those who wish to undermine the court’s legitimacy.” And even Justice Kagan gave a bit of a nod to these arguments in her remarks at the Ninth Circuit Judicial Conference, calling the charge that the Court has been too deferential to this President “a bad rap,” and agreeing that the Court has not been “a puppet” for this administration.

There are two different—but equally significant—problems with this framing, and today’s “Long Read” elaborates upon both of them. First, however much folks want to make out of the Court’s rulings in the tariffs, birthright citizenship, Lisa Cook, and Illinois National Guard cases, there’s still that pesky little denominator—the massive amount of allegedly unlawful behavior by this administration that the Court has green-lit, especially through unsigned and thinly explained orders on the emergency docket. Not only did the Court grant 22 consecutive applications by the federal government last year, but as I argue in my forthcoming Harvard Law Review Foreword, it has gone even further in those rulings and elsewhere to limit the ability of lower federal courts to push back against systemic lawlessness by this administration—and to thereby enable executive branch mischief, even if and when it isn’t affirmatively endorsing it. So even on its face, holding the Court out as a bulwark against Trump is factually contestable, at best.

But even to those who find the “independent Court” framing more descriptively persuasive notwithstanding the contrary evidence, there’s a much deeper second problem: Judicial independence is not, and is not supposed to be, an end unto itself; it is a means to an end. As Justice Anthony Kennedy memorably put it at a 2007 Senate Judiciary Committee hearing, “Judicial independence is not conferred so judges can do as they please. Judicial independence is conferred so judges can do as they must.”

The central problem with this Court, as I’ve argued (ad nauseam) before, is that this Court does what it pleases, when it pleases—because it has become completely unaccountable. An independent Court, like a stopped clock, may still be “right” every once in a while. But a Court that continues to arrogate power from the other branches; that continues to aggrandize the role of lower federal courts (and then doesn’t defend them from external attacks); that continues to overrule prior precedents solely because of its disagreements with them; and that continues to assert that it isn’t (and ought not to be) accountable to anybody is a Court that will only continue to lose public support—and, thus, lose its ability to stand up to other government institutions when it actually needs to—no matter how “independent” it is.

To trumpet the Court’s “independence” as the relevant yardstick is to deeply misunderstand the entire purpose of the enterprise—and the gravity of the danger the Court’s recent behavior has created, and is creating, for itself. Independence may make it possible for the Court to rule against the other branches of government, but it isn’t why the other branches have to comply.

More on all of that below. But first, the news.

On the Docket

The Court issued three orders last week—all of which were denials of stays of execution respecting Florida’s “double execution” on Tuesday of James Duckett and Dominick Occhicone. There were no public dissents from the three orders—the 102nd, 103rd, and 104th full-Court rulings on emergency applications during the October 2025 Term.

Turning to this week, nothing formal is expected from the Court (the next set of summer housekeeping orders isn’t expected until August 17). But it seems distinctly possible, if not likely, that we’ll get some movement on the two emergency applications (including one from the Trump administration) in the mail-in ballot cases—which I covered in detail in last Thursday’s bonus issue.

Finally, for readers who might be prospective applicants for a clerkship in Justice Jackson’s chambers, I’m posting (as I did last year) a letter from her chambers with respect to the process for applying for the October 2027 Term (i.e., 2027-28). I appreciate and commend Justice Jackson’s efforts to provide broader public access to this information—and, whether through this newsletter or more official media (like the Supreme Court website), I hope her colleagues will follow suit.

The One First “Long Read”:
One-and-a-Half Cheers for the “Independent” Court

As I previewed above, my own view is that there are two distinct critiques of the “independent Court” narrative—one descriptive and one normative.

The Descriptive Critique: Is the Court Really “Independent” of Trump?

Those playing up Trump’s losses at the Court this term have a remarkable tendency of playing down his wins. Yes, the Court rejected President Trump’s IEEPA tariffs in February; yes, it struck down his birthright citizenship order in June; and yes, it refused to let him fire Fed Governor Lisa Cook without any process at all. (The fourth ruling in this camp is the denial of a stay in the Illinois National Guard case from December.) But it also gave him Trump v. Slaughter—overruling Humphrey’s Executor outright and subjecting the heads of more than 20 formerly independent, multimember agencies to the President’s whim—and Mullin v. Doe, which cleared the way to strip legal status from roughly 350,000 Haitian and Syrian TPS holders while holding that a district court couldn’t even review most of how the Secretary got there (a ruling that implicates as many as 1.5 million noncitizens who have never committed a crime and who face pretty terrible conditions if/when returned to their home countries). The President’s victories changed the law and the structure of the executive branch permanently, in Trump’s favor, and in ways that no future president is likely to give back. To be sure, Trump’s losses weren’t irrelevant. But other than the Fourteenth Amendment holding in the birthright citizenship case, they were at least legally modest.

Then there’s the emergency docket. This administration has filed 35 emergency applications in roughly 18 months—against 19 from the Biden administration across four years, and just eight from Bush and Obama combined across sixteen years. It has won almost all of them. Bloomberg Law found the justices granting relief in 24 of the first 27 administration-related orders (including, in my own data, 22 in a row). This past term alone, the Court wiped out more than 20 preliminary injunctions and TROs entered against the government. Justice Sotomayor has called the volume “unprecedented in the court’s history.” If you put that number—two dozen-plus, mostly unexplained, mostly unsigned—up against the four losses described above, the claim of “independence” becomes quite a bit murkier (all the more so given the doctrinal moves the justices have made to support those results—which, as I’ve explained, have been Trump-specific).

The interventions have also tended to facilitate much more widespread activity by the executive branch—usually without any suggestion that the activity is legal. With respect to restructuring the executive branch, the Court effectively blessed the government-wide reduction-in-force order in July 2025 (Justice Jackson’s “wrecking ball”); it cleared the dismantling of the Education Department, and it let Trump remove officials at the NLRB, MSPB, and CPSC well before Slaughter explained why. With respect to appropriations, the justices allowed the termination of more than $700 million in NIH grants and then unblocked the first “pocket rescission” in half a century, freeing the President to withhold roughly $4 billion Congress had appropriated—what Justice Kagan called “a presidential usurpation of Congress’s power of the purse.” Andwith respect to immigration, the Court let the administration end TPS for more than 300,000 Venezuelans and stayed an order barring roving patrols in Los Angeles from targeting people by apparent ethnicity, language, and occupation—with only a solo (and deeply problematic) concurrence from Justice Kavanaugh trying to explain why. In each case, the Court’s interventions are what made the policies operative—policies that produced, and continue to produce, massive real-world effects.

Beyond the specific behavior that it has enabled, the Court has also spent much of the last 18 months making it harder for lower courts to halt systemic executive branch lawlessness. Last June’s ruling in Trump v. CASA largely took universal injunctions off the table. Trump v. Boyle told district judges that they are bound not just by what the justices hold but by what they can be inferred to have meant in cryptic orders—a point Justice Kavanaugh restated in Cook, where he wrote that “as a matter of vertical stare decisis, the lower courts have no choice but to follow our lead.” Trump v. J.G.G. foreclosed APA review in a wide swath of immigration cases (especially those arising under the Alien Enemies Act of 1798)—channeling those cases into messy district-by-district habeas litigation. And the Court has also read new constraints into the Tucker Act that has forced countless spending-cutoff suits out of district courts and into the Court of Federal Claims—which generally lacks the power to issue coercive relief (like injunctions).

All of this is coming at the same time as the Court is granting emergency relief even in cases in which the Trump administration had defied the district court; as Justice Gorsuch has chastised lower courts for “defy[ing]” the justices; and as the Court has declined to provide any broader defense of lower-court judges despite withering and deeply personal attacks from the executive branch and its supporters. One might have expected the Supreme Court to be the lower courts’ shield against a program of systemic executive lawlessness, or, at the very least, against unprecedented executive branch attacks on (and defiance of) lower courts. My Foreword argues, to the contrary, that it has been a significant part of the problem.

Against that backdrop, the four big rulings against Trump look considerably smaller than they did on the days they came down. The Guard order was a denial of a stay “at this preliminary stage.” The tariffs ruling closed one statutory door and the administration promptly walked through others. Cook held only that Trump owed Lisa Cook process before firing her, and sent the case back so the government could try to fire her with process. And the birthright citizenship ruling invalidated an order that had never taken effect, because every judge to look at it had already said it was flagrantly unconstitutional. Those decisions aren’t insignificant, but they also aren’t the whole waterfront. A Court isn’t independent just because it rules against the President every once in a while; one also needs to consider what he is able to do anyway—and who let him do it.

The Normative Critique: Should “Independence” Be the Goal?

Regardless of how you slice the data, there’s a deeper problem with the way folks are arguing about this: they’ve turned judicial independence into the benchmark, when it never has been, and isn’t supposed to be, the reason for having an independent judiciary. Independence is instrumental. It matters because of what it produces—a judiciary free to rest its rulings on law rather than on the identity or the preferences of the party in front of it or the threats the judges face if they rule the “wrong” way. Alexander Hamilton’s point in Federalist No. 78 was not that independence is intrinsically valuable but that the judiciary, holding “neither force nor will, but merely judgment,” would need something other than force or will to make its judgments stick.

That authority was to come from diffuse public support—a general willingness, accumulated across decades and across ideological lines, to accept the Court’s rulings as binding even when one strongly disagrees with them. Political scientists have spent forty years documenting that such a base of existing support is what distinguishes such durable judicial authority from specific, transactional approval of particular outcomes. They’ve also argued, that entire time, that this kind of capital is far harder to rebuild than it is to spend. The popularity (and legitimacy) that creates that authority is not something the Court possesses inherently or by structural design; it is something it earns from the public, only over time. That’s why treating independence as a self-justifying good gets the causation backwards, and it lets the Court off the hook for spending down its limited capital—for whether its behavior is continuing to provide the public with reasons to obey it.

When President Eisenhower sent the 101st Airborne into Little Rock in September 1957, he was not acting out of deep agreement with Chief Justice Warren’s opinion for the Court in Brown—he had been conspicuously unenthusiastic about Brown and said so. He acted because defiance of a federal court order by a governor, televised nightly, had become intolerable to a country that took the Court’s word as law. As he put it in a televised address to the nation on September 24, 1957, “[t]he very basis of our individual rights and freedoms rests upon the certainty that the President and the Executive Branch of Government will support and insure the carrying out of the decisions of the Federal Courts, even, when necessary with all the means at the President’s command.” Indeed, Eisenhower continued, “[u]nless the President did so, anarchy would result.”

And when President Nixon released the tapes in response to United States v. Nixon, he did so on the advice that resistance was politically unsurvivable—a unanimous Court, including three of his own appointees, had told him to comply, and there was no constituency for defying it. Neither President obeyed because the Court was “independent.” Both obeyed because the Court was respected enough that disobedience would have cost them more than compliance did. The Court’s power didn’t flow from its independence; it flowed from popular support for its authority.

Last month, in contrast, Gallup found only 33% of the public approves of the Court—the lowest in the 26 years it has asked the question, down from a prior floor of 39% a year earlier—and Republican approval fell from 79% in September 2025 to 58%. The numbers may be sharper among Democrats, but the decline among Republicans should be just as disconcerting to anyone who cares about the Court’s forward-looking authority. Marquette’s pollsters attribute part of it to a “broad perception that the court has sided with the president on more issues than it’s opposed him.” That’s certainly part of it.

But whatever has caused the downturn in public support for the Court, more “independence” isn’t the solution. A Court that wants to regain public trust is a Court that tries to earn that trust back. It’s more than a little difficult to look at the October 2025 Term (or any other recent session of the Court), or the justices’ behavior off the bench, and see any visible evidence of any concerted effort in that direction. Thus, even if the Court rules against Trump more in the years to come (in both absolute and percentage terms), that won’t solve the problem, either. The answer, as ever, is a Court that is responsive, and acts as if it ought to be responsive, to the other institutions of government—with justices, to go back to Justice Kennedy, who do what they must, not what they please.

SCOTUS Trivia: The Rump Docket

Today’s the First Monday in August. From 1790 until 1838, that would have made it the opening day of a Supreme Court Term—one at which, by statute starting in 1802, a grand total of one justice was expected to show up.

I’ve written about this before, but it remains my favorite bit of Supreme Court arcana, and it comes from a pair of terrific 2006 articles by Professor Ross Davies: “The Other Supreme Court” in the Journal of Supreme Court History, and “A Certain Mongrel Court” in the Minnesota Law Review. Davies’s name for the institution is the “rump Court,” and the short version is that the Supreme Court came in two flavors from 1802 to 1838: the familiar en banc February Term, and a summer sitting at which a single justice returned to Washington and constituted the Court all by himself.

The Rump Court was a byproduct of a sustained interbranch row. After the Election of 1800, the lame-duck Federalist Congress passed the Judiciary Act of 1801—the “Midnight Judges Act”—and the incoming Democratic-Republicans repealed it in March 1802. The companion Judiciary Act of April 29, 1802 consolidated the Court’s work into a single annual February Term—conveniently postponing Marbury for fourteen months—while providing that one justice could hold a session each August.

As Davies notes, nobody at the Court ever objected that this was in tension with Article III’s command that the judicial power be vested in “one supreme Court.” The justices were at that very moment writing furious letters to each other about the constitutionality of the repeal and the restoration of circuit-riding. On the rump Court, they were, as Davies put it, perfectly untroubled. Both the en banc Court and the one-justice August Court were simply treated as versions of the same institution.

Part of that may have been because, at most of the August sittings, nothing of consequence happened. A contemporary newspaper account quoted by Davies reported that “[f]or many years past, the business of this court has been entirely pro forma.” But not always. In August 1838, Chief Justice Taney—sitting alone—heard argument and delivered an opinion on an order to show cause in Ex parte Hennen, a case the full Court would go on to decide at the January 1839 Term. Davies makes the persuasive case that this only makes sense if Taney understood himself to be acting as the Court rather than as the Fourth Circuit’s Circuit Justice. He also reprints Taney’s unreported August 1838 opinion in full—which is reason enough to read the article.

Taney’s move drew attention to the whole strange arrangement, and Congress abolished the August Term in February 1839, with only vague gestures toward “efficiency” by way of explanation. Which brings me to Davies’s larger point: for 37 years, Congress subdivided the Supreme Court into two differently constituted bodies, and no one—not the justices, not the litigants, not the bar—thought “one supreme Court” forbade it. Whatever you make of the current proposals to restructure the Court, the early Congresses plainly believed they had more latitude here than we tend to assume. They may have been right.

I hope that you’ve enjoyed this installment of “One First.” If you have feedback about today’s issue, or thoughts about future topics, please feel free to email me. And if you liked it, please help spread the word!

Share

If you’re not already a paid subscriber and are interested in receiving regular bonus content (or, at the very least, in supporting the work that goes into this newsletter), please consider becoming one:

Subscribe now

This week’s bonus issue will drop on Thursday. As ever, I hope you are all staying safe out there, in whatever degree of “independence” you find yourself.

The underlying question in these cases is indeed a massively important one: whether President Trump had the power, as he purported to exercise back in March, to order the U.S. Postal Service to refuse to deliver mail-in ballots to anyone not on a federal citizenship list built and maintained by the Department of Homeland Security and the Social Security Administration (a list that will fail to account for countless citizens who are legally registered to vote in states that allow mail-in voting—including those who live in states that have refused to cooperate with administration demands for their voter rolls).

At least for now, though, the dispute in front of the Supreme Court involves who can challenge Trump’s order and when—not whether the order is legally valid on its face. Thus, even if the Supreme Court grants either or both of the applications filed this week, all it will be doing is putting off resolution of the merits (and, depending upon what the government does, perhaps only for a matter of weeks).

For reasons I explain below the fold, my own view is that such a move would still be problematic in its own right. Either the Trump administration is going to take meaningful steps to implement this executive order before the election (at which point, the standing and ripeness issues necessarily go away) or it isn’t (in which case, it’s impossible to see how the federal government is suffering the kind of irreparable harm that is supposed to be necessary to obtain emergency relief from the Supreme Court). The key point for present purposes is that, whatever headlines it may generate, even a full grant of relief in Trump v. California would almost certainly leave the federal government’s authority to interfere with mail-in voting unresolved—not settled in Trump’s favor.

For those who aren’t paid subscribers, we’ll be back with our regular coverage of the Supreme Court on Monday. For those who are, please read on.

Background

On March 31, President Trump signed Executive Order 14,399, “Ensuring Citizenship Verification and Integrity in Federal Elections.” It is the second of his election orders. The first, from March 2025, sought documentary proof of citizenship for federal voter registration and a ban on counting late-arriving ballots (it has largely been struck down or enjoined).

The new order has two operative parts. Section 2 directs U.S. Citizenship and Immigration Services and the Commissioner of Social Security to compile “state citizenship lists”—state-by-state rosters of people the federal government has identified as citizens who will be 18 or older by Election Day. Section 3 tells the U.S. Postal Service (USPS) to build a system under which it accepts and delivers mail ballots only for voters who appear on state-submitted lists keyed to those federal lists, and only in envelopes bearing new barcodes and logos that USPS would specify.

Together, Sections 2 and 3 purport to hand USPS a veto over which ballots move through the mail. Postmaster General David Steiner has not been coy about this. He told senators that USPS would not deliver mail ballots for states that declined to hand over their absentee voter lists to the federal government—a threat that matters precisely because the states’ refusals all but guarantee that the federal list will be massively incomplete (and not just full of individual errors). Asked whether USPS actually has the authority to do any of this, Steiner said he’d “have to defer back to the courts.”

California v. Trump

So far, the courts to reach the merits have been uniform in their skepticism of the order’s legality. Five suits were filed against the order, and the one that has reached the Supreme Court first came from 23 states and D.C., led by California, Massachusetts, Nevada, and Washington, and filed in Boston (in the U.S. District Court for the District of Massachusetts). In that case (captioned California v. Trump), Judge Indira Talwani granted summary judgment to the states back on June 25.

Two things about Judge Talwani’s ruling have been underplayed. First, Talwani didn’t rest on any single doctrinal hook. She held that Sections 2 and 3 are ultra vires and that they violate the separation of powers—that they are, in her words, “legally void.” As she summarized the ruling in a later order, the “Constitution lodges power over congressional elections in state legislatures primarily and in Congress ultimately. The President plays no role in the regulation of elections.” But the statutory holding is at least as important, and probably harder for the government to dislodge: “No law enacted by Congress delegates authority to control mail-in voting to USPS,” and Congress hasn’t authorized the Postal Service “to sort the mail, let alone to refuse to transmit mail ballots for certain voters.” As in the birthright citizenship litigation, whatever the federal government’s power to take these steps, there’s a far easier argument that the executive branch can’t take them by itself.

Second, Judge Talwani’s June ruling also found that the order’s compressed timeline left “no allowance for Congress’ mandated procedure regarding USPS rulemaking.” An order that can only be carried out by violating the statutes that govern the agency carrying it out ought to be in serious trouble regardless of what one thinks about Article II.

The remedy Judge Talwani imposed was also relatively modest. She enjoined the federal defendants—not the President personally—from implementing Sections 2 and 3 as to the plaintiff jurisdictions, and only for the November 3 election and any elections before it (so this is not a “universal” injunction). Nothing stops the administration from implementing the order in the non-plaintiff states, and nothing about her decision bears in any way on the rules for the 2028 cycle.

Talwani had also, days earlier, refused to dismiss a parallel challenge from the League of Women Voters and other voting-rights groups, finding that “disenfranchisement as to the upcoming election is plausibly likely and imminent.” And on the timing question the government now presses at the Supreme Court, she’d explained her reasoning in an interim order: “[I]n light of the EO’s specific deadlines over the next three months, and the reality that elections will be occurring throughout this period . . . postponing judicial review is impracticable and may inflict significant hardship on Plaintiffs.”

The federal defendants and 12 intervening Republican-led states appealed to the First Circuit and moved for a stay pending appeal. In their stay application, they argued solely that the plaintiffs lacked standing and that their claims weren’t ripe. Tellingly, they did not ask the First Circuit to hold that the order is lawful.

On Saturday, July 25, the First Circuit denied the stay. Most media accounts reported the decision as a divided one, with brand-new Trump-appointed Judge Joshua Dunlap in dissent. But that doesn’t quite capture the nuance. Critically, Judge Dunlap agreed with Judges Gustavo Gelpí and Julie Rikelman with respect to Section 3 of the executive order (the USPS-specific provisions)—i.e., that it should remain paused for the duration of the litigation. As he wrote,

[A]lthough final regulations must normally exist before standing can arise, the specificity of the EO’s stated timelines and the inescapable legal requirements and impending deadlines that States currently face in facilitating mail-in and absentee balloting for immediately upcoming elections together establish sufficient concrete injury from the EO in this unusual case. Unlike Section 2(a), therefore, the Government and Intervenor States have failed to show a substantial likelihood of success as to their argument that Plaintiff States will not suffer actual or imminent harm sufficient to establish standing to challenge Section 3.

And “Because the Government and Intervenor States mount no substantive defense of Section 3, they have no other basis to demonstrate likelihood of success as to the injunction barring implementation of that provision.” Dunlap dissented only as to Section 2, because “Section 2(a) . . . merely orders the provision of information to States; it does not order States, including Plaintiffs, to take any action based on the State Citizenship List.”

To be sure, the majority had a response with regard to Section 2—focusing on a question the government never answered: whether it would prosecute a state official who mailed a ballot to someone missing from the DHS list. “State and local officials are no exception to the commonsense rule that ‘[p]eople do not lightly disregard public officers’ thinly veiled threats to institute criminal proceedings against them if they do not come around.’” And it concluded that letting the order operate for September and November “would sow confusion and threaten disenfranchisement of many eligible voters.”

But whoever has the better argument with respect to Section 2, the key is that the panel was unanimous as to Section 3—and voted to leave it blocked. We’ll come back to the significance of that in a moment.

The D.C. Cases

Complicating matters, both descriptively and analytically, is the eight-page ruling the D.C. Circuit handed down on Tuesday in a different case challenging the executive order—holding, in DSCC v. Trump, that “this case likely is unripe for review in its present posture.” Both Supreme Court applications suggest that there’s a tension between the DSCC and California rulings, but on closer inspection, that isn’t true.

At the heart of the D.C. Circuit’s ruling was the conclusion that the DSCC’s injuries “flow not from Executive Order 14,399 itself, but entirely from predicted future actions contemplated by that Order” (a separate panel had reached the same conclusion on July 17 in the NAACP’s parallel case). The panel thereby affirmed Judge Nichols, who had denied relief earlier this summer on standing grounds. But rather than focusing on standing, the court of appeals rested on the lack of “ripeness”: a party committee suing for its members is suing based upon injuries that depend on someone else’s subsequent choices, so there was nothing yet for a court to review.

In the process, the D.C. Circuit specifically distinguished the facts of the California litigation:

[T]hese Plaintiffs are not directly regulated by the Order, which functions as a directive to federal agencies. Nor are they the States charged with administering elections and adapting their procedures to the requirements contemplated by the Order. Nor do they assert injuries as electoral candidates whose individual campaigns currently are being adversely affected by the ordered instability and uncertainty of the field on which they are competing. Plaintiffs argue instead that their organizations will be harmed by expending resources to counter not the Executive Order, but the ensuing agency actions that Order contemplates. So the record does not establish that the Order “itself” is inflicting “concrete harm” on these Plaintiffs before the agencies act.

The First Circuit, in a case brought by sovereign states rather than an association, found both standing and ripeness satisfied, because the order imposes immediate requirements on California and its co-plaintiffs—including with respect to the lists they must compile, how they must distribute their mail-in ballots, and which officials will be exposed to potential federal investigation the moment the final rule is issued, and not after some future act by a third party. In my view, both of these decisions are correct. But that hasn’t stopped both the federal government and the intervening red states in California from holding them out as in tension with each other—and as thus providing their own basis for relief. As is hopefully clear, that just isn’t accurate; the D.C. Circuit’s DSCC ruling may not have cited the California case, specifically, but the sentence highlighted above expressly distinguished its fact pattern. In that respect, the rulings reinforce each other; they don’t create a split.

The Two Emergency Applications

In its application for a stay in the Supreme Court, filed on Monday, the Trump administration sought both a stay of Judge Talwani’s judgment pending its appeal in the First Circuit and an immediate administrative stay while the justices consider it. The argument is the same one the First Circuit rejected, i.e., that it’s too early. As the application argues, courts “cannot enjoin implementation of Executive Orders that leave open material issues that still need to be resolved by agencies, unless and until those agencies actually take concrete actions that injure the plaintiffs.” The injunction is “especially indefensible because the agencies are still deliberating over how (if at all) to implement the Order, yet the district court preemptively decided that whatever the agencies may choose to do will necessarily be unlawful.” And the equities, as Solicitor General Sauer argues, run only one way: without a stay, “there will not be sufficient time to obtain appellate relief before the November election,” because implementation has to begin “as soon as early to mid-August.”

And yesterday, the 12 states that intervened in California v. Trump to defend the executive order filed their own stay request. (Alabama, Florida, Indiana, Kansas, Louisiana, Missouri, Montana, Nebraska, Oklahoma, South Carolina, South Dakota, and Texas are the applicants, with Missouri Solicitor General Louis Capozzi as counsel of record.) These states want what the federal government wants: a stay of Judge Talwani’s injunction pending the First Circuit appeal, plus an immediate administrative stay. Armed with Tuesday’s D.C. Circuit ruling (which came after the Justice Department’s filing), the states argue quite aggressively that there’s a real tension between the First Circuit’s denial of a stay and the D.C. Circuit’s affirmance of the dismissal in DSCC. Again, there isn’t, but you have to dig fairly deep to see why.

Three Takeaways

All of this leads to three points that should clarify both the stakes of these applications and, as important, what they’re not about.

First, and starting with the latter, neither application asks the Supreme Court to bless Sections 2 and 3 on the merits, and both rest on the same procedural move, telling the Court that no one has a present, concrete injury yet. The government prevailed on this argument against the DSCC and NAACP plaintiffs because they were associations, whose harms run through what some future voter or USPS employee might do; it should fail against the California plaintiffs, who are being told right now, as sovereigns, to take certain concrete and cost-inducing steps. (Again, the First Circuit was unanimous that Section 3 is harming the plaintiff states now.) But even if the Court fails to draw this distinction and grants the applications in full, all that would do is put off, perhaps for only a few weeks, a challenge to the government’s actions. After all, as noted above, one of two things has to be true: Either the government is going to take meaningful steps to implement Section 3 of the executive order or it isn’t. In other words, it’s not clear how the government can even make the procedural case for emergency relief without simultaneously undermining its arguments that the plaintiffs haven’t yet been (or won’t imminently be) injured.

Second, although I’m sympathetic to the First Circuit majority in California, there’s also a lot to commend Judge Dunlap’s approach—which would grant the stay only in part (with respect to Section 2), and deny it in part (with respect to Section 3). Because Section 2 doesn’t directly require the states to do anything, one could reasonably argue that it doesn’t harm the states (or, at least, that it doesn’t harm them irreparably) even if Judge Talwani is right that it’s ultra vires. If that’s the ruling we get from the Supreme Court, my own view is that it would be a significant defeat for the Trump administration—even though the headlines will likely portray it as a partial victory. Here’s a case where the two different pieces are just not of remotely equal importance.

Third, the timing here may also be instructive. In addition to seeking a stay pending appeal, the Trump administration’s application sought an immediate administrative stay—something neither Justice Jackson (who’s the Circuit Justice for the First Circuit, and thus the recipient of the applications) nor the full Court has granted. Justice Jackson has called for a response to both applications (no surprise there), but the plaintiffs’ brief (covering both applications) is not due until 4 p.m. next Monday (August 3). The Court, in other words, is not moving heaven and earth to decide these applications—which suggests that it may not be especially taken with the Justice Department’s own claimed harms, or at least their urgency. Either way, I wouldn’t expect a ruling before the latter part of next week—and perhaps even later if there are any separate writings by the justices. And again, the critical thing to watch for is not whether the Court grants any relief, but whether it grants the applications in full. Whether Section 3 goes into effect is much more important here than whether Section 2 does.

***

I don’t mean to downplay the significance of this litigation, or of the emergency applications in Trump v. California. These cases are a big deal; the executive order is a bigger deal; and Section 3 would be deeply problematic were it to be operative in the 2026 midterm election cycle (or, frankly, ever). But even the best-case scenario for the Trump administration here is a narrow procedural victory in the Supreme Court that puts off Section 3’s reckoning only for a matter of weeks. That’s more relief than the government deserves at this juncture, but it would be far less consequential than it will seem (and than the media will likely report it to be).

Share

If you’re not already a subscriber, please consider becoming one—and upgrading to a paid subscription if and when your circumstances permit:

Subscribe now

One of the quieter casualties of the Court’s just-completed term is the modern descendant of a 1789 statute that, for nearly half a century, was used to allow victims of torture and other grave human rights abuses to walk into a federal courthouse and ask for a remedy. In Cisco Systems, Inc. v. Doe I, a majority comprising all six Republican appointees effectively shut that door—and, as I’ll explain, it did so by overruling a unanimous 2004 precedent without ever fully acknowledging that that’s what it was doing, and through reasoning that says a lot more about this Court’s (lack of) regard for its own past decisions than it does about the relevant clause of section 9 of the Judiciary Act of 1789—known today as the Alien Tort Statute (“ATS”).

To that end, today’s “Long Read” traces how the ATS went from a “legal Lohengrin” to a human rights workhorse and back again; why Justice Barrett’s account of the statute in Cisco doesn’t hold together; and why one case the majority conspicuously ignored—the Abu Ghraib litigation out of the Fourth Circuit—shows just how much we all stand to lose when the Supreme Court closes the courthouse doors to these kinds of lawsuits.

More on that below. But first, the news.

On the Docket

The Supreme Court issued the first of its regular summer order lists last Monday—adding one case to the merits docket for the upcoming term, and denying a number of rehearing petitions (and re-filed applications). Those were the only full Court rulings to come down last week. Speaking of rehearing, although President Trump has publicly committed to seeking rehearing of the Supreme Court’s June 30 decision in the birthright citizenship case, the deadline for filing such a petition is today (technically, it was Saturday, but that bumps it to today). It’s President Trump, so you never know. But I’m not holding my breath.

The One First “Long Read”: The ATS, Then and Now

There are exactly 33 words in the Alien Tort Statute (as amended), and for most of the two centuries after Congress enacted it as part of the Judiciary Act of 1789, they received little attention. The statute gives the district courts “original jurisdiction of any civil action by an alien for a tort only, committed in violation of the law of nations or a treaty of the United States.” That was it. It doesn’t expressly provide a cause of action; it doesn’t list specific, covered wrongs; and there is next to no legislative history to help illuminate why Congress included it in the federal judiciary’s founding charter. Nor did history help; for nearly 200 years, the statute mostly sat on the shelf. Courts invoked it a handful of times, but nobody was quite sure what it was supposed to do.

The modern history of the statute starts with Filártiga v. Peña-Irala, in which a Paraguayan family found their son’s torturer living in Brooklyn and sued him under the ATS for his 1976 killing back home. In 1980, the Second Circuit held that the statute reached the claim—that deliberate torture by a state official violated the law of nations, and that federal courts were open “for adjudication of the rights already recognized by international law.” The court read the ATS not as manufacturing new rights but as providing a domestic forum for a wrong the world had already condemned, and it grounded that reading partly in foreign relations, reasoning that claims “fraught with implications for the nation as a whole” belonged in federal court rather than being left to fifty different state courts.

Filártiga set off a wave of human rights litigation—including significant rulings from at least three different federal courts of appeals (the Second, Ninth, and D.C. Circuits) in the 1980s and 1990s. It also led to an enormous and sustained debate among academics—who fought bitterly over the extent to which U.S. courts could (and should) provide remedies for violations of “customary international law,” i.e., international legal principles and norms that were not part of statutes or ratified treaties, and thus were not (at least obviously) the law of the land for purposes of the Constitution’s Supremacy Clause.

Even as that debate raged on, the Supreme Court didn’t weigh in for almost a quarter of a century. And when it finally did, in Sosa v. Alvarez-Machain, it landed on a compromise that would govern ATS litigation for the next two decades. Justice Souter, writing for the Court, agreed that the ATS is “strictly jurisdictional” and creates no causes of action on its own—but rejected the argument that this made the statute “stillborn.” The First Congress, the Court reasoned, understood that courts would recognize a modest set of common-law claims for violations of international law as well defined and universally accepted as the paradigms the drafters had in mind in 1789 (offenses against ambassadors, violations of safe conduct, and piracy—what the Court now calls the “Blackstone three”).

Going forward, Sosa held, courts could recognize new claims, but only for norms of comparable “definite content and acceptance among civilized nations,” and only with great caution. That was Sosa’s famous second step: narrowing the scope of ATS claims relative to the most extreme lower-court rulings, but leaving federal courthouse doors unmistakably open. Justice Scalia wanted those doors shut altogether, but he didn’t get five votes (or even four). Only Chief Justice Rehnquist and Justice Thomas took the narrowest view of the ATS in 2004.

The Court spent the next two decades narrowing that opening, but it kept declining to overrule Sosa itself. In its 2013 ruling in Kiobel v. Royal Dutch Petroleum, the Court initially granted certiorari to decide whether corporations could be proper defendants under the ATS, but then held, on re-argument, that the ATS wasn’t available in “foreign-cubed” cases—those in which foreign plaintiffs sued foreign defendants for conduct on foreign soil. Only if the claims “touch and concern” U.S. territory “with sufficient force to displace the presumption” against extraterritoriality would Sosa’s framework be available.

Then, in its 2021 decision in Nestlé USA, Inc. v. Doe, the Court held that “general corporate activity”—operational decisions made in U.S. boardrooms—was not enough domestic conduct to let former child laborers sue two American companies over forced labor on Ivorian cocoa farms. That was enough to decide the case, and tellingly, Justice Thomas’s effort to go further and hold that Sosa should be scrapped entirely, couldn’t hold a majority; Chief Justice Roberts and Justice Barrett pointedly declined to join it, with Justice Sotomayor noting separately that such an approach “would overrule Sosa ‘in all but name.’”

Then, on June 23, and without any attempt to reconcile their prior adherence to Sosa, the Chief Justice joined Justice Barrett’s opinion for a six-justice majority in all but relegating Sosa to the scrap heap. In Cisco, plaintiffs alleged that Cisco and two of its executives had helped the Chinese government build the “Golden Shield” surveillance system used to identify, detain, and torture Falun Gong practitioners, and they sued under both the ATS and the Torture Victim Protection Act. Justice Barrett, writing for herself, the Chief Justice, and Justices Thomas, Alito, Gorsuch, and Kavanaugh, held that federal courts “may not create new causes of action” under the ATS at all—including for aiding and abetting—and that the TVPA likewise does not reach those who aid and abet torture. The three Democratic appointees all dissented from the ATS holding, and Justice Sotomayor dissented from the TVPA holding.

Formally, Cisco doesn’t eviscerate the ATS. But it takes a huge bite out of it (and Sosa), by eliminating the power of courts to recognize new international-law claims that were not available in 1789. What’s left is what Scalia wanted in his Sosa concurrence and what Thomas pressed for in Nestlé: the ATS as a jurisdictional grant and nothing more, good for the three offenses Blackstone catalogued in 1789 and, well, that’s it. Justice Sotomayor called this what it is—a “sub silentio overruling of Sosa” that “marks yet another low point in this Court’s esteem for its precedents”—and she’s right. The whole significance of the compromise in Sosa was the Court leaving the door open for ATS suits for offenses other than the “Blackstone three.” By cabining the ATS to those three offenses and nothing more, Cisco renders Sosa a dead letter.

In many respects, reading the opinions in Cisco is like re-reading the opinions in Sosa. It’s the same debate about the relevance of the same sources and arguments to the interpretation of the same statute. And that’s what really bothers me: the majority never really explains why Sosa had to go. Sosa was unanimous on the point that courts retain some power to recognize international-law claims. It was 20 years old. Litigants, human rights advocates, and lower courts had built two decades of doctrine on top of it. Under any of the versions of stare decisis the current justices have professed to follow, that combination usually counts for something.

The Cisco majority engages with almost none of that. Its two stated reasons—that Sosa’s second step was “narrow at the outset,” and that judicially implied causes of action are disfavored today—are arguments about why Sosa was wrong (again, just read Scalia’s Sosa concurrence), not reasons why a wrong-but-entrenched precedent must now be abandoned. Those are different questions, and the Court has spent years insisting they are different questions. “We now think the 2004 Court got it wrong” is not a special justification; it is just disagreement, and disagreement with a prior Court has never been enough, by itself, to justify tossing out its ruling. Beyond its implications for human rights litigation in the United States (about which more in a moment), this is the most alarming part of Cisco: stare decisis really is, as the hosts of Strict Scrutiny so often put it, for suckers.

The other major move in Cisco is Justice Barrett’s suggestion that recognizing ATS claims carries an inherent risk to American foreign relations—that letting these suits proceed invites the courts to meddle in matters the Constitution assigns to the political branches. Especially given how Kiobel and Nestlé had already narrowed the ATS to minimize its impact in cases with foreign implications, that move seems underdefended at best—all the more so in light of a high-profile case that the Court didn’t even discuss: Al Shimari v. CACI Premier Technology.

Al Shimari is the Abu Ghraib case. Four Iraqi men, arrested and detained by the U.S. military in 2003 and later released without charge, alleged that interrogators working for CACI—a Virginia-based government contractor—conspired with soldiers to torture them. After 16 years and five trips to the Fourth Circuit, the case was pared down to two ATS claims, and in 2024 a jury found CACI liable for conspiracy to commit torture and cruel, inhuman, and degrading treatment, awarding the three remaining plaintiffs a total of $42 million. On March 12, 2026—three months before Cisco—the Fourth Circuit affirmed. And it directly confronted the foreign-relations worry that animates Barrett’s opinion. Because Congress had itself endorsed torture claims under the ATS and its companion statutes, the court explained, “allowing this case to proceed lacks any potential to negatively affect foreign relations.” If anything, the court said, the opposite is true: “nothing would be more offensive to the arena of foreign relations than to proclaim that courts of the United States may not provide a remedy to foreign nationals who were tortured by members of the U.S. military.”

The existence of cases like Al-Shimari (and Al-Shimari itself) is, to me, the critical point Justice Barrett’s opinion elides. The foreign-relations objection has at least some force in “foreign-cubed” suits like Kiobel, where U.S. courts are asked to referee a dispute among foreigners over conduct abroad. It has almost no force where the defendant is an American company and the alleged wrong is torture carried out under color of U.S. authority in a manner that, per Kiobel, “touches and concerns” U.S. territory. In those cases, refusing to hear the claim doesn’t protect American diplomacy—it announces to the world that the United States offers no forum when its own contractors and companies are accused of grave human rights abuses. Al Shimari shows how this is supposed to work. There, you had an ATS suit that went to a jury after years of litigation over jurisdictional and procedural predicates. The case not only produced a verdict, but it survived every foreign-relations and separation-of-powers objection that CACI raised. In other words, Al-Shimari illustatres the very accountability that the ATS was revived to provide, delivered without any of the diplomatic ramifications that Justice Barrett’s majority opinion just assumed are inevitable.

Whether Al Shimari itself survives is now an open question—CACI is already arguing that Cisco forecloses the plaintiffs’ aiding-and-abetting theory, and the plaintiffs are countering that torture, like piracy, sits within even the narrowest reading of the 1789 statute. (Perhaps needless to say, I’m deeply sympathetic to the plaintiffs here.) But however that case comes out, its stakes will be deeply significant. For 46 years, the ATS gave torture survivors somewhere to go when no one else would hold their abusers to account—all the more so when the abusers were Americans or American corporations. Cisco has taken most—if not all—of that away. The problem isn’t just what Cisco means for human rights plaintiffs, real as that is. It’s how little work this Court now thinks it has to do before walking away from a precedent it no longer likes—something that the same 6-3 majority would do two more times in the seven days between Cisco and the last hand-downs of the October 2025 Term.

SCOTUS Trivia: The Fifth Seal

For … reasons, I was recently watching a (bad) old West Wing episode (from the sixth season, by which point everything had gone to pot), where Vice President Bob Russell keeps telling his completely unfunny joke about how, if he looks at the Vice President’s seal just so, it kinda looks like it says “President of the United States.”

Somehow, that got me thinking about the Supreme Court’s seal—including its history and evolution. Here’s an image of the current version:

As it turns out, the current seal is the Court’s fifth in its history. On Wednesday, February 3, 1790—during the very first sitting of the Supreme Court—the Court ordered “that the Seal of this Court shall be the Arms of the United States, engraved on a circular piece of Steel of the Size of a Dollar, with these words in the Margin ‘The Seal of the Supreme Court of the United States.’” That original die served the Court for four decades; its last known use came in August 1831, on the papers in Worcester v. Georgia.

Four successive redesigns followed, culminating in the larger, less ornate version the Court ordered in May 1904 and adopted on May 4, 1905—the one still in use today. The symbolism is derived from the Great Seal of the United States, as are most of the design elements. Indeed, besides the Court’s name, the only detail that’s specific to the Court’s seal is also the one with the most meaning: beneath the eagle’s claws sits a single star, representing the “one supreme Court” that Article III creates—a not-so-subtle reminder that the Constitution mandates exactly one such body and leaves control over the existence, size, and scope of the rest of the federal judiciary to Congress.

The seal remains in the custody of the Clerk of the Court, who stamps it on official papers—including the admission certificates issued to new members of the Supreme Court bar and, as I’ve learned the hard way, the official judgment handed down 32 days after a merits case has been decided.

I hope that you’ve enjoyed this installment of “One First.” If you have feedback about today’s issue, or thoughts about future topics, please feel free to email me. And if you liked it, please help spread the word!

Share

If you’re not already a paid subscriber and are interested in receiving regular bonus content (or, at the very least, in supporting the work that goes into this newsletter), please consider becoming one:

Subscribe now

This week’s bonus issue will drop on Thursday. As ever, I hope you are all staying safe out there—especially from human rights abuses, for which there are increasingly few remedies in U.S. courts.

That behavior was on display again late Tuesday, when the en banc Fifth Circuit granted the Trump administration’s unprecedented request for a “precedential stay” of three consolidated district court decisions in the latest dispute over the government’s novel and controversial immigration detention policies. Even though the full court of appeals had already agreed (without even being asked) to rehear a panel decision that had recognized constitutional limits on how long the government can detain certain noncitizens pending their removal before providing them with a bond hearing, it apparently concluded that further intervention was warranted.

As a result, in its unsigned, unexplained order, the Fifth Circuit acted in a way that seems at least outwardly intended to deter district courts in cases other than the three under full court review from providing bond hearings to any of the tens of thousands of other noncitizens currently in immigration detention in Louisiana, Mississippi, and Texas. To get there, the court of appeals had to turn the entire concept of an emergency application on its head—and accept, without any analysis, the Trump administration’s deeply disingenuous arguments about what the district courts in these specific cases had actually done. And the Justice Department is already claiming (wrongly) that this unexplained order somehow resolves the constitutional question at issue—perhaps in an effort to distract from what the Fifth Circuit actually said (i.e., nothing):

All of this may seem hypertechnical, but as I explain below the fold, it comes at an enormous cost. A bond hearing is not—as Brett Shumate asserted—“a constitutional right to roam free”; it allows a noncitizen to demonstrate that they should be released pending their (potentially years-long) removal process—based upon an individualized judicial determination that they pose no threat to themselves or others, and no risk of flight. Without even being eligible for bond hearings, everyone from children to the elderly would be forced to remain in indefinite immigration detention for the duration of their removal proceedings (and pending their removal), even those who have never committed a crime—in conditions that, according to a growing number of reputable accounts (including DHS’s own Inspector General), are increasingly inhumane. That’s problematic enough in the abstract; it’s even more galling when it comes through unexplained and undefended procedural chicanery.

For those who aren’t paid subscribers, we’ll be back Monday with our regular coverage of the Supreme Court. For those who are, please read on.

Read more

Welcome back to “One First,” a newsletter that aims to make the U.S. Supreme Court more accessible to lawyers and non-lawyers alike. I’m grateful to all of you for your continued support, and I hope that you’ll consider sharing some of what we’re doing with your networks.

Share

If you’re not already a subscriber, please consider becoming one—and upgrading to a paid subscription if and when your circumstances permit:

Subscribe now

For as long as I’ve taught National Security Law, the Alien Terrorist Removal Court (“ATRC”), a specialized national security court that Congress created in 1996, has been a recurring curiosity of mine. The idea was to provide a judicial mechanism like the one provided by the Foreign Intelligence Surveillance Act (FISA), in which the government could simultaneously (1) obtain meaningful judicial review in a deportation case while (2) keeping the (classified) evidence secret. Indeed, in a 2014 book chapter, my now-colleague David Cole and I held out the ATRC procedure as a potentially useful illustration of the “cleared counsel” model of national security litigation—in which, instead of providing the subject of an enforcement proceeding with the classified information on which the proceeding depends, the government provides it to a lawyer for the subject (albeit in only a subset of cases before the ATRC). As the chapter reflects, the ATRC had remained an academic curiosity—because it had never actually been used. Until last week.

On Friday, Seamus Hughes (the truly indispensable editor of “Court Watch,” among many other things) broke the news that the Trump administration had filed the first-ever application before the ATRC, leading to the first order in that court’s 30-year history (and its hasty creation of a website). We don’t know much beyond what Seamus has already reported (and what’s now on the ATRC’s website). But as someone who has spent more time thinking about the ATRC than, I suspect, 99.99% of the population, now seems as good a time as any to introduce the court; to flag the constitutional questions its statutory structure raises; and to reflect on how alarming it is that the Trump administration has, after all this time, activated it.

Skipping to the punchline, much like the FISA Court, the ATRC reflects a series of imperfect compromises. And there are reasons to worry about some of the benefits that the ATRC process provides to the government. But—and this is the important part—it also has some protections baked into it that are not currently available (or, at least, not immediately available) to many non-citizens in the “regular” removal process, primarily the robust role of Article III district judges—as opposed to Justice Department immigration judges—in administering it. I have at least some concerns about why the Trump administration is dusting it off now. But there’s no universe in which the ATRC will allow the administration to pursue the kind of mass, summary deportations that it has thus far been unable to achieve through any other immigration authorities. If there’s something nefarious afoot, it’s meaningfully smaller than that.

More on all of this below. But first, the Supreme Court-specific news.

On the Docket

There was only a single full-Court ruling out of the Court last week—Tuesday’s denial, over no public dissents, of a stay of execution to Florida prisoner Dennis Sochor. The only other big news out of the Court last week was Tuesday appearances by Justices Kagan and Barrett at House and Senate budget hearings—the first public testimony by sitting justices since 2019, and their first appearance in the Senate since 2011. I’ve written before about the myriad reasons why this really ought to go back to the annual routine that it used to be; hopefully Tuesday’s hearings were a first step in that direction.

Finally, as I noted in last Thursday’s bonus issue, I’ve been a bit busy with Court-related work myself—and posted to SSRN a draft of my Foreword to the Harvard Law Review’s Supreme Court issue for the October 2025 Term, tentatively titled “The Court Against the Courts.” As I noted in Thursday’s post, I’d welcome your feedback!

The One First “Long Read”: Unpacking the ATRC

In 1996, as part of the Antiterrorism and Effective Death Penalty Act (and as quickly amended in the Illegal Immigration Reform and Immigrant Responsibility Act), Congress created the ATRC—a tribunal largely modeled on the FISA Court, and which was intended to be able to handle individual deportation hearings in cases involving highly sensitive national security information. Although the statutory procedures are quite complex, the Federal Judicial Center has a helpful summary:

The statute authorize[s] the Attorney General to draft an application for removal of a suspected alien terrorist [a term defined by statute], and to submit the application to the removal court under seal. A single judge may grant the application upon a finding that the alien in question has been identified correctly as an alien terrorist present in the United States and that removal under other available legal means would pose a risk to national security. Upon granting a removal application, the court must hold a public removal hearing at which the accused has the right to be represented by counsel and the government bears the burden of proving that the accused is an alien terrorist. To avoid the disclosure of classified information, the government must prepare an unclassified summary of such information which the judge must find to be sufficient for the alien to prepare a defense. If the government meets its burden of proof at the hearing, the court will order the alien removed from the United States.

As for the judges, the ATRC is composed of five already-serving Article III district judges “assigned” by the Chief Justice to also serve on that court, the only criterion for which is that they come from five different circuits. (For more on the Chief Justice’s assignment powers, see the trivia, below). Under the 1996 statute, the Chief Justice may assign to the ATRC judges who are also serving on the FISA Court, but he doesn’t have to (all five of the current ATRC judges are also FISA judges, but that hasn’t always been true). The Attorney General can appeal a judge’s denial of an application under the statute, and either the Attorney General or the non-citizen at issue may appeal the decision of the judge after a removal hearing. Unlike the FISA process (where there’s also a specialized appeals court—the FISA Court of Review), the 1996 statute gives the ordinary D.C. Circuit exclusive appellate jurisdiction over the ATRC.

I started with the judges and the appeals process because this is probably the most important point about the ATRC: The compromise it reflects is one in which the government’s benefits involve the protection of classified information and the burden of proof (about which more shortly), but in which the costs sound almost entirely in who is reviewing these cases. (There are also some more pro-government rules respecting detention of the subject of the hearing both beforehand and afterwards, but the daylight between those rules and the rules in “ordinary” cases has shrunk quite a bit since 1996.) Unlike immigration judges housed within the Justice Department’s Executive Office of Immigration Review (who are less and less independent by the day), the ATRC is staffed not just by Article III district judges, but (especially today) by those with experience handling the only comparable kind of litigation—classified applications and other proceedings brought under FISA.

The ATRC’s Chief Judge, Judge Joan Ericksen (a senior judge on the District of Minnesota), made this abundantly clear in the order she publicly issued on Thursday. Here’s the relevant text:

On July 15, 2026, the United States filed an Application to the United States Alien Terrorist Removal Court pursuant to 8 U.S.C. § 1531 et seq. to remove the respondent in the above-captioned matter from the United States. On July 16, 2026, the Court held a hearing as contemplated by § 1533(c)(1), at which the Court heard attorney argument but no sworn testimony was taken. During the hearing the United States offered to supplement the record with information, and the Court had questions about the nexus that the government alleges between the actions of the respondent and the specific sections and subsections it invokes with respect to those actions. The answers persuaded the Court that the Government could benefit from the opportunity for more thoughtful consideration.

In other words, the ATRC is not going to be a rubber stamp.

That said, it’s also worth highlighting the constitutional concerns militating in the other direction, most of which were flagged (and carefully analyzed) in an excellent 2008 student note in the Duke Law Journal by John Dorsett Niles. The biggest ones involve the evidence. If the cost to the government of the ATRC process is an Article III judge and a D.C. Circuit appeal, the benefit is the ability to rely entirely on evidence that will remain classified—and be presented in most cases only to the judge and only in camera.

If the judge, after considering the evidence on the record as a whole, finds that the Government has met its burden [i.e., that a preponderance of the evidence shows the non-citizen to be an “alien terrorist,” as that term is defined by statute1], the judge shall order the alien removed and detained pending removal from the United States. If the alien was released pending the removal hearing, the judge shall order the Attorney General to take the alien into custody.

In other words, the statute authorizes the government to remove someone from the country based on evidence they never see and don’t meaningfully have an opportunity to contest. As Niles’s note points out, this raises quite severe due process questions—especially for those non-citizens whose due process rights most closely approximate those of citizens, i.e., lawful, permanent residents (LPRs, or “Green Card” holders).

Mindful of those concerns, Congress provided a special procedure for cases in which the subject of the proceeding is an LPR. Under 8 U.S.C. § 1534(e)(3)(F), the court in such a case must appoint a “special attorney to assist the alien,” “by reviewing in camera the classified information on behalf of the alien, and . . . by challenging through an in camera proceeding the veracity of the evidence contained in the classified information.” In other words, an LPR is entitled to have a lawyer challenge the evidence being used against them, even if the lawyer can’t actually consult with the subject of the proceeding. This approach is obviously not perfect, but it reflects at least a genuine attempt on Congress’s part to balance the government’s interest against that of the non-citizen.

At the other end of the spectrum, for better or worse (my own view skews toward the latter), non-citizens with far weaker claims to comparable due process protections (including those who have never had lawful immigration status) can be removed from the country under this provision without ever being told why; other than the government, only the judge assigned to the case (and the D.C. Circuit) would ever know the reasons. And in the middle, perhaps the hardest cases are those in which the subject is a non-citizen with substantial lawful connections to the United States, just not LPR status. Those individuals would have the strongest argument, in my view, that the statute fails to accord them the constitutionally required process—because they don’t get the benefit of the “special attorney” even though they have substantial due process rights.

As Niles notes, the seriousness of these constitutional concerns is almost certainly a big part of why, across 30 years, the government had never previously utilized this procedure. And we have no way to know what it is that prompted the Trump administration to change course now. But it seems worth flagging three things that we do know—which ought to bear on both how big a story this is and how alarmed we should be by it.

First, on its face, the statute contemplates a robust and meaningful role for the district judge—in lieu of the subject and their attorney in most cases. And Thursday’s order by Chief Judge Ericksen is a promising sign that she, at least, takes that role seriously. Courts could do a lot, in this context, to mitigate the constitutional concerns by ensuring not only that the government is held to the statutory requirements in every particular, but that its evidence that the subject really is an “alien terrorist” is robust and substantial.

Second, because of the meaningful (if secret) judicial review the statute contemplates, there is no universe in which the ATRC could become a vehicle for mass, summary deportations along the lines of what the Trump administration has already attempted—including through the Alien Enemies Act of 1798. The ATRC is exactly what the executive branch-dominant process before immigration judges and the Board of Immigration Appeals is not—it’s insulated from political pressure. However the ATRC’s judges come out in individual cases, the substance and timing of the review the statute contemplates would be deeply antithetical to any attempt to use its procedures in a large number of cases at any one time. It’s a scalpel, not a bazooka.

Third, and militating in the other direction, the fact that the government had never previously used this process in the 30 years it’s been on the books speaks to the discomfort that just about everyone who’s ever looked carefully at the statute has come away with. Democratic legal systems abhor secret evidence—not just because of the grave fairness concerns it can create but because of the corrosive effects of having these kinds of decisions made out of the public eye, no matter how faithful to the relevant legal criteria they may be. The FISA process may affect a heck of a lot more people, but no one is being removed from the country based solely on a secret decision made by a FISA judge based upon secret and uncontested evidence.

Even in a world in which the five ATRC judges are doing their level best to put the government through its paces, and to look out for the interests of the non-citizen whose liberty is at issue, it is more than a little difficult to have faith in such a process—no matter what we think of the administration that’s invoking it. All the more reason, in my view, for abolition of the ATRC—or, at the very least, substantial reform—to be added to the laundry list of lower court-directed measures that will be sorely needed the next time we have a Congress that cares. Until then, here’s another envelope-pushing move by the Trump administration to which we’ll have to pay close attention, and another example of Article III district judges serving as the principal and most meaningful check to hold the executive branch accountable.

SCOTUS Trivia: The Chief Justice’s Assignment Power

One of the Chief Justice’s more interesting statutory authorities is the power to “assign” (a verb that has intentionally different legal implications from “appoint”) already serving Article III judges (and justices) to temporary service on other courts. The ATRC is one of four distinct courts (or panels) staffed entirely by Chief Justice assignments; the FISA Court, the FISA Court of Review, and the Judicial Panel on Multidistrict Litigation are the other three. But the Chief Justice also has the (oft-used) authority to assign active and senior district and circuit judges (and retired justices) to temporary service on other courts. (This practice is generally referred to as “sitting by designation.”)2

Chief Justice Roberts’s approach to these assignments has come under some fire in the past, including in a July 2013 New York Times study by Charlie Savage that raised questions about whether Roberts’s assignments to the FISA Court had skewed toward judges appointed by Republican presidents to a meaningfully greater degree than the assignments made by his predecessors. I’ll just say that, for all of the criticisms directed toward the Chief Justice these days, the current compositions of these courts reflect a pretty impressive cross-section of highly regarded judges appointed by presidents of both parties. There are plenty of problems with the ATRC (and the FISA Court, while I’m at it). But in my view, the identity—and quality—of the judges is pretty far down the list.

I hope that you’ve enjoyed this installment of “One First.” If you have feedback about today’s issue, or thoughts about future topics, please feel free to email me. And if you liked it, please help spread the word!

Share

If you’re not already a paid subscriber and are interested in receiving regular bonus content (or, at the very least, in supporting the work that goes into this newsletter), please consider becoming one:

Subscribe now

This week’s bonus issue will drop on Thursday. As ever, I hope you are all staying safe out there.

For better or worse, I have the honor and privilege of having been invited to write the Foreword for this year’s Supreme Court issue, covering the October 2025 Term (to be published in November 2026). And now that I have a completed draft, I thought I’d use today’s bonus post to introduce (and share) the draft, and, below the fold, to reflect a bit on both (1) the evolution of the Foreword in general; and (2) the specific reasons I settled on the focus of this year’s piece.

To spoil the punchline, here’s the abstract:

The Supreme Court sits atop a judicial system on which it is deeply dependent. It lacks the capacity to superintend more than a fraction of the cases filed each year, and it relies on the “inferior” courts beneath it to develop records, sharpen issues, and enforce compliance with coercive orders. The Chief Justice (of the United States, not just the Supreme Court) is steward of that whole system—not merely of the Court at its apex.

Since President Trump returned to office in January 2025, that system, and federal district courts in particular, has been subjected to a sustained, multi-front assault without precedent in American history: a historic flood of litigation, rhetorical and impeachment-based attacks from the political branches, escalating threats against judges and their families, and mounting executive defiance of court orders.

One might have expected the Supreme Court to be the lower courts’ shield. In this Foreword to the Harvard Law Review’s Supreme Court issue for the October 2025 Term, I argue that it has instead been a significant part of the problem. Through an accelerating series of emergency-docket interventions—and through its merits rulings, its rhetoric, and its silences—the Court has stayed, narrowed, or vacated district court orders, often without full briefing, argument, or reasoned explanation, while treating trial judges’ considered factfinding and legal conclusions as provisional placeholders to be revisited at the justices’ discretion. Using seven sets of cases from the past eighteen months, I show that the Court has done significant damage to the formal, practical, and moral authority of federal district courts to hold the federal government to account—all while arrogating much of that same authority to itself, with respect to factfinding, law-finding, the timing and scope of appellate review, and the balancing of the equities.

The cumulative effect has been to recast the district judge from an Article III adjudicator whose orders bind the United States into something closer to an adjunct of the justices. That transformation, I contend, is best understood as one more expression of the judicial aggrandizement—and, ultimately, the “autocratic judging”—that has come to define the Roberts Court, and it poses a serious threat to the rule of law. Re-empowering the district courts, accordingly, should be central to any serious program of court reform. The story of the Supreme Court since early 2025 is, in the main, a story about the Court against the courts.

And for those who would like to read more (warning: it’s … long), the full draft can be downloaded from SSRN at this link. Needless to say, I’d welcome any and all feedback!

For those who are not paid subscribers, we’ll be back Monday with our continuing coverage of the Supreme Court. For those who’d like to know a bit more about the Foreword in general and my contribution, in particular, please read on.

Read more

Share

If you’re not already a subscriber, please consider becoming one—and upgrading to a paid subscription if and when your circumstances permit:

Subscribe now

Given the paucity of news out of the Supreme Court last week, I’m devoting today’s newsletter to an ongoing and increasingly messy litigation battle between the Trump administration and hospitals providing gender-affirming care to transgender minors, and the related litigation between the Federal Trade Commission (FTC) and the World Professional Association for Transgender Health (WPATH)—the professional body whose Standards of Care shape how gender dysphoria is treated worldwide.

This litigation did not arise in a vacuum. Rather, it’s the latest development across nearly eighteen months in which the Trump administration has made ending gender-affirming care an explicit policy goal and directed federal agencies to investigate the doctors and organizations that (legally) provide or support it. But the Trump administration’s animus and shady litigation behavior does not make every criticism WPATH has faced baseless, nor does it resolve the underlying, genuinely contested questions about the evidence behind WPATH’s guidelines. The goal of today’s post is thus to provide enough of a background on what’s going on in these cases for folks to (hopefully) reach their own conclusions. And before diving in, let me express my thanks and appreciation to one of my superstar research assistants, Abby Liman, for truly superlative assistance with this post.

Background

WPATH is a nonprofit professional association that has, for more than fifty years, published the “Standards of Care” (SOC) that much of the world’s clinical, insurance, and legal infrastructure treats as the authoritative guide to treating gender dysphoria. On the other side are the Justice Department the FTC—a federal agency tasked, among other things, with certain consumer protection responsibilities, which has been investigating WPATH and peer organizations, including the American Academy of Pediatrics and the Endocrine Society, over their public statements on gender-affirming care for minors.

Back in 2022, WPATH had published the eighth edition of its Standards of Care and, in doing so, dropped the specific age limits that prior editions had attached to hormones and certain surgeries. That editorial decision—and internal documents about how it was made—has become the centerpiece of the criticism against the organization. But it also became fodder for the Trump administration, which has made hostility to gender-affirming care—and to transgender rights more broadly—an explicit policy priority. On January 20, 2025, the President issued an executive order announcing that the federal government would recognize only two sexes and enforce “all sex-protective laws” accordingly; eight days later, a second order, “Protecting Children from Chemical and Surgical Mutilation,” declared it federal policy not to “fund, sponsor, promote, assist, or support” the “transition” of minors and to “rigorously enforce all laws” limiting such care.

Those orders set the tone, and the enforcement agencies followed. In April 2025, Attorney General Pam Bondi issued a memorandum accusing clinicians of deploying “junk science” to justify “maiming and sterilizing children” and directing DOJ components to investigate providers; a June 2025 Civil Division memo made those investigations a stated priority, and DOJ’s own press release announcing the first wave of subpoenas quoted Bondi vowing that “medical professionals and organizations that mutilated children in the service of a warped ideology will be held accountable.”

That context matters for two reasons. First, it is the evidentiary backbone of the retaliation and improper-purpose findings that courts would repeatedly make about those subpoenas: judges did not infer animus from thin air but from the administration’s own words—including a DOJ lawyer’s concession in one hearing that “the executive branch wants to reduce or eliminate gender-related care to minors,” and FTC Chairman Andrew Ferguson’s memorandum promising to “fight back against the trans agenda” by investigating “the doctors, therapists, hospitals, and others who deceptively pushed” such care. Second, it is precisely what makes the two sides’ conduct so hard to sort: an administration that has openly announced its goal of ending a category of medical care, and then deploys investigative tools against the providers and professional bodies that support it, invites the suspicion that any given enforcement action is pretextual—even in the instances where the underlying legal theory might otherwise stand on its own.

One last note before seguing to the nationwide litigation: I’ve seen the FTC/WPATH imbroglio described elsewhere as “multidistrict litigation,” but it is not an MDL in the technical sense of a single consolidated proceeding under 28 U.S.C. § 1407. Instead, what’s happening is a campaign playing out across multiple districts at once. The geographic sprawl is not an accident, and is itself one of the central features (and, to my mind, problems) of the underlying dispute.

The Trump Administration and the Nationwide Losses It Incurred

The litigation kerfuffle began in July 2025, when the Department of Justice issued more than twenty administrative subpoenas to hospitals providing gender-affirming care, demanding the identities and medical records of minor patients. Those efforts met with near-uniform objections from district courts across the country, most finding they had been issued for an improper purpose.

Judge Chun in the Western District of Washington set aside the subpoena to Seattle Children’s Hospital, citing the government’s “threadbare justification” and “strong evidence” of improper purpose. Days later, Judge Joun in the District of Massachusetts quashed the subpoena to Boston Children’s Hospital, pointing to the administration's own executive orders and memoranda as evidence of its objective. Judge Whitehead, also in the Western District of Washington, quashed a subpoena in the QueerDoc matter, concluding that the government “issued the subpoena first and searched for a justification second.” Judge Kearney in the Eastern District of Pennsylvania quashed a demand for children’s medical and psychological records, finding the link between patient identities and any fraud theory “tenuous at best.” Judge Bissoon in the Western District of Pennsylvania granted a motion to quash in the UPMC matter, writing that the government’s demand carried “more than a whiff of ill intent.” Magistrate Judge Chung in the District of Colorado recommended that a subpoena to Children’s Hospital of Colorado be held unenforceable as a “smokescreen,” expressly “join[ing] the chorus” of courts before him. And Judge Rubin in the District of Maryland quashed the subpoena to Children’s National Hospital as “a pretext to fulfill the Executive’s well-publicized policy objective.”

The District of Columbia litigation belongs in the same story but occupies a distinct branch of it. Where the seven quashals concerned the DOJ’s administrative subpoenas, the D.C. cases challenged a different instrument—the FTC’s civil investigative demands (CID)—and three organizations brought parallel suits there: the Endocrine Society, the American Academy of Pediatrics, and WPATH. The doctrinal frame also differed: rather than resting on the “improper purpose” standard that governs subpoena enforcement, Chief Judge Boasberg analyzed the CIDs as likely First Amendment retaliation, granting a preliminary injunction in the Endocrine Society case on May 7 and, the same day, enjoining the WPATH CID by relying on that companion opinion. Yet the two branches were expressly linked: Boasberg grounded his reasoning for enjoining the CIDs in the rulings by a “slew of courts” that had already characterized the DOJ’s subpoenas as a "smokescreen" for pressuring providers to abandon gender-affirming care, citing the Colorado and Western District of Washington decisions among them. The connection ran the other way too—the Rhode Island Child Advocate, in turn, leaned on the same cross-district consensus in its own motion to quash. What ultimately set the D.C. cases apart was how they ended: not in a merits ruling but in the FTC's withdrawal of the CIDs on the day it filed a new lawsuit in Texas, which mooted the preliminary-injunction motions in D.C.

Shopping for Chief Judge O’Connor

Indeed, it was only after this unbroken run of losses that the government turned to the Northern District of Texas—and the Fort Worth division, specifically, without ever actually explaining why that’s where this nationwide investigation has been centered. (Although Fort Worth is not technically a “single-judge division,” the two judges who hear 90% of the civil cases filed there—Reed O’Connor and Mark Pittman—are both … pretty far to the right end of the ideological spectrum, at least as federal judges go.)

To remind folks, O’Connor, a George W. Bush appointee who became chief judge in 2025, has for years been one of the go-to judges for conservative litigants, a reputation built in part on the outsized power of one-judge divisions where a plaintiff can effectively choose its jurist (including the Wichita Falls division in which O’Connor hears 100% of new civil cases). He has authored some of the highest-profile district-court rulings of the last decade—striking down the Affordable Care Act and portions of the Indian Child Welfare Act among them—and has himself publicly defended the existence of single-judge divisions (and the judge-shopping it invites), blasting the judiciary’s leadership for “caving to political pressure” when it moved to curb the practice. (I published a response to O’Connor’s speech here.)

Importantly, it’s not quite accurate to refer to what the government is doing in Fort Worth as “judge-shopping” (since Fort Worth isn’t a single-judge division). But once one of these cases was assigned to O’Connor, the federal government appears to be have been relying on the claim that all of these other disputes are “related” to produce the same result—having O’Connor handle all of the nationwide litigation relating to the Trump administration’s investigations of gender-affirming care for minors.

The conduct that has drawn the sharpest “exceeded his power” criticism is not the handling of the WPATH litigation (which we’ll come back to in a moment), but a companion episode: the DOJ’s subpoena to Rhode Island Hospital, which kicked off the latest maneuvering in Fort Worth. On April 30, 2026, DOJ filed a petition to enforce its administrative subpoena against Rhode Island Hospital, demanding five years of records on every minor who received gender-affirming care—including patient names, Social Security numbers, diagnoses, and family details, and asserting that its investigation was “being carried out in the Northern District of Texas.” Chief Judge O’Connor granted the motion to compel that same day, without opposition and without notice to the hospital, ordering production within fourteen days. After the District of Rhode Island separately quashed the subpoena, O’Connor on May 18 ordered the hospital to turn over all responsive records to his court for in camera “safekeeping” pending appeal, accusing it of trying to “circumvent the authority of this Court and the Fifth Circuit”; the hospital’s appeal and emergency-stay request to the Fifth Circuit were denied.

O’Connor pressed ahead anyway, and—most controversially—barred the hospital from seeking further relief in any court other than his own, the Fifth Circuit, or the Supreme Court. In effect, O’Connor issued an injunction aimed not at the executive branch or the plaintiffs, but at another federal court, a maneuver for which, as one commentator put it, “[n]either the Constitution nor any statute comes close to awarding O’Connor, or any other judge, this power.”

At the same time, enforcing an administrative subpoena is a bread-and-butter judicial function, and O’Connor grounded his order in a finding that the government was “engaged in a legitimate investigation” acting “pursuant to its statutory authority.” He also ordered the records produced to the court itself, to be held privately, rather than handed straight to the government—a protective measure, not a wholesale disclosure (which is why the First Circuit ultimately refused to intervene in the Rhode Island case).

The Fort Worth venue engineering is not a one-off, which is part of why it has drawn such scrutiny. Beyond the Rhode Island case, DOJ began issuing grand-jury subpoenas out of the Northern District of Texas to reach other hospitals thousands of miles away—Stanford’s Packard Children’s in California and NYU Langone in New York—even though nothing about those investigations was local to Texas. Judges in both places said so bluntly. Judge Pitts in the Northern District of California found that the government had “issued its grand jury subpoena to avoid another loss and force Packard and its patients to pursue any challenge . . . in a forum that DOJ deems friendlier,” and that DOJ “cannot reasonably invoke comity and the separation of powers as cover for its forum shopping”; Judge Failla in the Southern District of New York likewise granted emergency relief to a class of minor patients. As she put it in ruling from the bench, “[t]he subpoena, expressly targeting members of a particular and uniquely vulnerable group, both shocks the conscience and rises to the level of the most egregious official conduct.”

Perhaps because of the wall of resistance it met in trying to enforce these subpoenas, the government finally resorted to another tack on June 17—when the FTC and four states filed a standalone lawsuit against WPATH and, the same day, withdrew the D.C. CIDs. The FTC then moved for a TRO and preliminary injunction asking O’Connor to bar WPATH from seeking relief in any court other than the Northern District of Texas, the Fifth Circuit, or the Supreme Court. That gambit produced the one moment of restraint: on July 3, after WPATH sought emergency relief in D.C., O’Connor issued a docket order stating that, “[a]fter consultation with Chief Judge Boasberg (D.D.C.),” he would defer ruling on the TRO to permit “orderly consideration” of the parallel injunction requests—a de-escalation that, as one observer noted, O’Connor had conspicuously declined to extend to the Rhode Island Hospital dispute.

Meanwhile, just last Friday, Chief Judge Boasberg rejected WPATH’s attempt to use its D.C. injunction to freeze the FTC’s new Texas enforcement suit, explaining that his May order “did not protect WPATH from complying with all information-seeking processes—only the CID that the Court held was likely retaliatory,” and that his role was not “to prejudge every permutation of the FTC’s future actions as unconstitutionally retaliatory.” Alas, it’s hard to disagree with Boasberg’s conclusion; once the FTC pulled the CIDs and refiled its theory as a standalone lawsuit in Texas, the D.C. injunction had little left to bite on, since it had enjoined a specific administrative demand that no longer existed—not all enforcement action by the FTC writ large.

***

Stepping back from the details, the upshot is that, for more than a year, district judges across the country reached a remarkable, largely unprompted consensus that these subpoenas were pretextual—a form of horizontal dialogue that lent the judiciary’s conclusions real institutional weight. But venue engineering of the kind on display in the Rhode Island and grand-jury episodes threatens to override that consensus altogether: if a single, reliably sympathetic forum can absorb every dispute and insulate the government’s preferred outcome from the accumulated judgment of a dozen other courts, that horizontal consensus becomes largely irrelevant to how these cases actually resolve. (Folks might recall that these were some of the loudest arguments against “nationwide injunctions,” arguments the Supreme Court endorsed last June, at least in that specific context, in CASA.) That’s not to say that the end result of all of this procedural maneuvering by the federal government is inconsistent with the rules as they currently stand; my own view is that it isn’t. But the above account seems like a pretty strong argument for why it should be.

The mischief caused by this kind of venue engineering is compounded by the fact that the FTC itself no longer operates with the independence from the White House it once had, following the Supreme Court’s Slaughter decision—raising the prospect that agency enforcement actions like the WPATH suit become a recurring tool rather than an isolated episode, and that the ability to judge- and forum-shop will only increase the politicization and ideological extremism of these disputes.

Three Takeaways

All of this leads me to what I see as three takeaways from what’s going on in these cases—one about the executive branch; one about the underlying merits; and one about the courts.

First, the combination of FTC forum-shopping and direct political control is a troubling one. Indeed, there are serious questions about the government’s litigation conduct in these cases, to say nothing of the FTC’s attempt to assert enforcement authority over noncommercial speech by a nonprofit like WPATH. The specter of the government playing “heads we win, tails you lose” by withdrawing the CIDs only after it filed its deceptive practices lawsuit in a cherry-picked forum is not an especially becoming one, regardless of where one comes down on the merits. Not that long ago, we could’ve taken some solace from the fact that government lawyers could generally be trusted to turn square corners. But for better or worse, the combination of the Trump administration’s behavior and the direct control over the FTC the President now enjoys thanks to Slaughter ratchet up the appearance—if not the reality—of some truly sketchy behavior by the executive branch.

Second, and militating at least a little bit in the other direction, the government has gone to great lengths to suggest that concerns about WPATH’s Standards of Care are more than frivolous. The UK’s Cass Review—commissioned by the National Health Service, not by U.S. culture warriors, albeit still deeply contested—found that WPATH’s SOC8 was a low-quality clinical guideline (and that most international guidelines trace back to WPATH or the Endocrine Society, rather than independently support it). Similarly, several European health systems have restricted these procedures on evidentiary grounds. WPATH’s First Amendment objections to the FTC’s investigation may have been vindicated by Chief Judge Boasberg in the D.C. litigation, but its science remains at least somewhat under challenge.

That said, even reasonable evidentiary disagreement (if that’s what this is) is not the same as fraud, and the picture here strikes me as, at best, contested rather than settled: WPATH has argued with more than a little force that the Cass Review is itself an “outlier,” and that a draft German guideline developed with Austria and Switzerland—under review by 27 professional societies—broadly aligns with SOC8 and does not restrict puberty blockers. The Cass Review has also been subject to its own significant academic criticisms. And even for those who find the Cass Review and arguments based upon it more persuasive than I do, “the evidence is somewhat weaker than you said” is a very different accusation against WPATH than “you knowingly deceived consumers for profit,” which, contra the FTC’s allegations, this ain’t.

More fundamentally, WPATH’s core legal position—that a nonprofit professional association’s clinical guidance and advocacy are protected speech, and that a consumer-protection agency should not be policing a live scientific debate—is not a dodge. It is a serious First Amendment argument that a serious judge (Boasberg) found likely to prevail. WPATH’s stated defense that it “supports individualized patient care, not a ‘one size fits all’ approach,” and that its guidelines rest on “established scientific standards, expert consensus, and patient-centered values,” is a legitimate framing of what guidelines are supposed to do—offer clinical judgment, not consumer advertising. And the associational-privacy stakes of the FTC’s original records demand—compelled disclosure of members’ internal communications—are precisely the kind of intrusion the First Amendment has long guarded against, which is why the chilling effect Chief Judge Boasberg credited should worry even those who distrust WPATH’s science.

Third, and finally, one of the real problems created by forum- and judge-shopping is the concern that these kinds of questions are being steered to the judges most likely to answer them in a particular way, and not necessarily in a way that reflects the median viewpoint of neutral Article III jurists. I’ve long suggested that one of the most important ways to curb the most extreme forms of judge- and forum-shopping is for Congress to provide clearer guidance on where these kinds of lawsuits can be brought—and, indeed, to require some degree of randomness in cases with true nationwide implications.

To date, most of the focus of that conversation has been on suits against the government, but especially after Slaughter, it seems increasingly clear that government-initiated suits can also raise the same suite of concerns. After all, if the Trump administration can run nationwide administrative enforcement proceedings through a single division in the Northern District of Texas, then a future Democratic administration could do the same through the Northern District of California, or the District of Massachusetts, or … you get the point.

I don’t imagine I’m going to persuade anyone about who’s right on the science, medicine, or policy here. Indeed, I’m not entirely sure myself. But if nothing else, perhaps folks can (finally?) be persuaded that this kind of strategic litigation behavior makes it harder to have faith that the best answer is “the courts.”

I hope that you’ve enjoyed this installment of “One First.” If you have feedback about today’s issue, or thoughts about future topics, please feel free to email me. And if you liked it, please help spread the word!

Share

If you’re not already a paid subscriber and are interested in receiving regular bonus content (or, at the very least, in supporting the work that goes into this newsletter), please consider becoming one:

Subscribe now

This week’s bonus issue will drop on Thursday. As ever, I hope you all stay safe out there—especially from subpoenas issued by the Northern District of Texas.