Taking Stock of the Birthright Citizenship Cases, Part I: Unpacking Trump v. CASA, Inc.

It has been just over a month since the Supreme Court issued Trump v. CASA, Inc. on June 27. In the immediate wake of the decision there was a good deal of confusion and disagreement about exactly what the Court held; whether it was right to do so; how important the Court’s decision was for the future of district court injunctions against Executive branch wrongdoing; and the impact of the decision, in particular, on President Donald Trump’s attempted redefinition of birthright citizenship. That confusion and disagreement has not dissipated in the interim.

Moreover, there have been several important developments in the past couple of weeks, including these:

• Last Wednesday, July 23, the U.S. Court of Appeals for the Ninth Circuit became the first appellate court to issue a decision on the merits: It held that enforcement of Trump’s Executive Order No. 14160, “Protecting the Meaning and Value of American Citizenship” (the “Citizenship Order”), would violate both the Fourteenth Amendment and 8 U.S.C. § 1401. The court of appeals also held that the four plaintiff States in a challenge in the Western District of Washington had standing to sue (over a dissent from Judge Patrick Bumatay), and that the district court’s “universal” injunction is appropriate because it is necessary to guarantee that the plaintiff States obtain complete relief from injuries they would suffer if the defendants implemented the Citizenship Order.

• Two days later, District Court Judge Leo Sorokin in the District of Massachusetts similarly held that it would be inappropriate to trim back his earlier “universal” preliminary injunction because it, too, is necessary to guarantee that the plaintiff States (and two localities) in the case before him won’t suffer injuries if and when the Government implements the Citizenship Order.

• In a case in the federal district court in New Hampshire (Barbara v. Trump), Judge Joseph Laplante certified a plaintiff class of all children born in the United States after Feb. 19, 2025 whose parents are described in the Citizenship Order, and he issued a preliminary injunction against implementation of the Order as to such children. Judge Laplante stayed that order for seven days to give the Government an opportunity to appeal, but for some reason the Government did not appeal, and therefore the injunction took effect on July 17.

• In the CASA case itself in Maryland, the plaintiffs asked Judge Deborah Boardman to certify a similar class of children born in the United States after Feb. 19, 2025 and to issue a similar classwide preliminary injunction. On July 16, Judge Boardman issued an “indicative ruling” that if and when the court of appeals remanded the case back to her for the limited purpose of adjudicating the plaintiffs’ class-related motions, she will grant the requested classwide preliminary injunction. On Tuesday, the court of appeals dismissed the pending appeal in CASA and remanded the case to Judge Boardman so that she can (i) ensure that her previous injunction complies with the Supreme Court’s decision in CASA; and (ii) rule on the plaintiffs’ motion for class-wide relief, “so that appellate review of the merits of that relief can come sooner rather than later.”

• In the past few days, at least half a dozen executive departments and department components have issued one form or another of “guidance” about how they’ll implement the Trump Citizenship Order:
  • The Department of State (for purposes of passport issuance)
  • The Department of Health & Human Services (for purposes of providing “Federal public benefits” defined by Title IV of the Personal Responsibility and Work Opportunity Reconciliation Act of 1996 (PRWORA))
  • HHS’ Center for Medicare and Medicaid Services (CMS) (also for “federal public benefits” purposes, including Medicaid and the Children’s Health Insurance Program (CHIP))
  • The Department of Agriculture (also for purposes of assessing eligibility for “federal public benefits”)
  • The Social Security Administration (for purposes of issuing Social Security numbers)
  • The Department of Homeland Security’s U.S. Citizenship and Immigration Services (USCIS) (to establish the meaning of “unlawfully present” for purposes of the Trump Citizenship Order; to establish the meaning of “presence” that is “lawful but temporary” in the Order and which immigration statuses or other types of lawful presence “may fall into that category”; and to establish whether certain children born in the United States to a person whose presence is “lawful but temporary” would have lawful immigration status for purposes of the Order)

In light of these developments and the forthcoming oral arguments on the merits of the Citizenship Order in the U.S. Court of Appeals for the First Circuit this Friday, I thought it would be useful to pull together an article that comprehensively addresses an array of matters related to CASA and the pending litigation about the Trump executive order on birthright citizenship.

Part I describes the various actions the Solicitor General (SG) asked the Court to take on the “shadow docket” in CASA.

Part II identifies the aspects of the SG’s application that the Court either rejected or chose not to resolve. Among other things, Part II addresses two questions currently being litigated in the lower courts: (i) whether some of the plaintiffs in the cases lack standing and (ii) whether the courts can and should issue relief to nonparties (including even “universal” injunctions) in order to secure “complete relief” to the plaintiffs themselves.

In Part III, I offer several reasons why the Court’s (tentative) holding in CASA about the absence of broad statutory authorization for federal courts to issue universal relief might not have nearly the practical significance that many scholars, judges, Justices, and Executive branch officials have suggested. In so doing, I also discuss the status and prospects of current class actions that plaintiffs have filed in some of the birthright citizenship cases themselves.

Part IV analyzes the merits of Justice Amy Coney Barrett’s opinion on the question of federal courts’ statutorily authorized equitable power to issue injunctions protecting nonparties.

Finally, in Part V, I discuss three things about the Court’s opinion that are much more troubling than its account of that statutory question.

In subsequent pieces, I’ll turn to the merits of the suits and examine several of the most prominent arguments the Department of Justice is offering in defense of the legality of Trump’s Citizenship Order.

I. What the Solicitor General Asked the Court to Do in CASA

Section 1 of the Citizenship Order declares that the privilege of U.S. citizenship under the Citizenship Clause of Fourteenth Amendment and 8 U.S.C. § 1401 does not extend to two categories of persons born in the United States to a father who wasn’t a citizen or lawful permanent resident at the time of the person’s birth: (1) those whose mothers were “unlawfully present” in the United States at the time of birth; and (2) those whose mothers were lawfully in the United States on a temporary basis, such as on a student, work, or tourist visa, on the date of the person’s birth.

For present purposes, it’s important to stress that Trump’s reading of the Citizenship Clause and § 1401 is inconsistent with the unbroken, consensus understanding of birthright citizenship that all three branches of the federal Government have had for the past 127 years. (In an earlier piece, I explained that DOJ’s initial defense of the Citizenship Order in district courts was flatly inconsistent with the Supreme Court’s holding in United States v. Wong Kim Ark (1898) and with well-established DOJ views, and that the litigators had failed even to try to explain why they were disregarding binding Court and DOJ precedent. Subsequently, DOJ has slightly changed the nature of its legal arguments; insists (however unconvincingly) that Wong Kim Ark does not preclude those new arguments; and has acknowledged that its current view about the scope of birthright citizenship deviates from the Executive branch’s longstanding position, which it now insists was mistaken all along.)

Section 3 of Trump’s Citizenship Order requires the Secretary of State, the Attorney General, the Secretary of Homeland Security, and the Commissioner of Social Security to “take all appropriate measures to ensure that the regulations and policies of their respective departments and agencies are consistent” with Section 1’s gloss on birthplace citizenship, and to ensure “that no officers, employees, or agents of their respective departments and agencies act, or forbear from acting, in any manner inconsistent with th[e] order,” i.e., with the Trump definition of birthright citizenship. Section 3(b) of the Order directs executive agencies to “issue public guidance within 30 days … regarding this order’s implementation with respect to their operations and activities.” Citizenship Order § 3(b); see App., infra, 17a; id. at 58a-59a; id. at 107a.

Section 2 of the Order includes a more specific directive: It forbids any federal department or agency from issuing documents recognizing U.S. citizenship, or accepting documents issued by State, local, or other governments or authorities “purporting to recognize U.S. citizenship,” to persons described in the two categories of Section 1 of the Citizenship Order. That particular prohibition on issuing or accepting documents—unlike the general directives in Section 3—by its terms applies only to persons born after Feb. 19, 2025.

Many different plaintiffs—individuals, organizations and States—have sued to enjoin federal officials from implementing the Citizenship Order. Three U.S. district courts, in the Districts of Massachusetts, Maryland and the Western District of Washington, not only concluded that the Citizenship Order is likely unlawful, but also entered “universal” preliminary injunctions barring the defendant executive officials from applying the Order to any covered persons, whether or not they are plaintiffs in the three cases.

In each of those three cases, the Government asked the court of appeals to stay the preliminary injunctions as applied to nonparties covered by the Citizenship Order. In the Massachusetts and Washington cases, it also asked the courts of appeals to hold that the State plaintiffs lacked standing to sue—which in the case of the Massachusetts case would mean staying the injunction entirely, since the plaintiffs there are all States (as well as the District of Columbia and San Francisco). And in the Maryland case brought by five individuals and two membership organizations, DOJ asked the court of appeals to limit the injunction to the individual plaintiffs and the eleven members of the organizational respondents who are personally identified in the complaint, primarily because (DOJ argued) an organization lacks standing to sue on behalf of members who are not themselves injured by the Government’s actions.

Each of the three courts of appeals denied the Government’s requests.  The Solicitor General then filed three distinct but virtually identical applications in the Supreme Court, each asking the Court to “partially” stay the scope—not the substance—of the district courts’ so-called “universal” preliminary injunctions.  (The applications were made by then-Acting SG Sarah Harris, before the Senate confirmed John Sauer to be Solicitor General. Sauer was in office by the time reply briefs were filed, however, and I’ll refer to him throughout this article.)

The Solicitor General’s applications were very unusual. He did not ask the Court to review the district courts’ holdings on the legality of the Trump Citizenship Order. Instead, he asked the Court “only” to do four other, more discrete things to limit the scope of the preliminary injunctions: (i) to “stay” the preliminary injunctions to the extent they offer relief to nonparties; (ii) to hold that the State plaintiffs lack standing to sue under the Court’s prudential “third-party standing” doctrine—a holding that would thereby limit the parties protected by the Washington injunction and require dismissal of one of the Massachusetts cases entirely; (iii) to limit the injunction in the Maryland case to cover only the five individual respondents and the eleven members of the organizational respondents identified in the complaint—but not any other, unidentified members of the plaintiff organizations; and (iv) to stay the provisions of at least two of the district court injunctions that might have been read to preclude executive agencies from developing and issuing public guidance about the their plans to implement the Citizenship Order.

II. What Did the Court Not Resolve in CASA?

The Court did not do several things the Solicitor General had requested.

1.

In footnote 2 of its opinion, the Court simply declined to address the SG’s arguments that the State plaintiffs lack “third-party” standing and that the organizational plaintiffs lack Article III standing to sue on behalf of unidentified members. It’s not altogether certain what that unexplained refusal means for purposes of the Government’s stay application. What’s clear, however, is that the Court did not stay or amend the preliminary injunctions on the basis of any alleged standing deficiencies by the State and associational plaintiffs. It’s therefore fair to say that the Court effectively denied those aspects of the Solicitor General’s application.

The questions of State and associational standing remain open for consideration by the lower courts, however. Therefore, in the next two sections I’ll describe very briefly what the Government’s standing arguments are and why the Supreme Court probably won’t deny standing to any of the plaintiffs in the birthright citizenship cases themselves.

2.

DOJ is making two different arguments about why the plaintiff States in the Massachusetts and Washington cases don’t have standing—one constitutional, the other “prudential.” (The SG raised only the latter, “prudential standing” argument in the Supreme Court, but DOJ is relying on both arguments in the lower courts.)

DOJ argues first that the States don’t satisfy the constitutional conditions for standing because any harms they will suffer, if and when the Executive branch begins to implement the Citizenship Order, are too speculative and contingent upon the ways in which the agencies decide to implement the Order, and because those harms are allegedly “self-inflicted” due to the fact that the States have voluntarily chosen to provide medical benefits to undocumented foreign nationals (which is the principal context in which the States allege they will suffer their injuries). This question of Article III standing, which wasn’t before the Supreme Court in CASA, is the issue that divided the judges on the Ninth Circuit panel last week: The majority (Judges Ronald Gould and Michael Daly Hawkins) held that the plaintiff States are virtually certain to sustain (mostly economic) injuries if the Order is implemented (see pp. 18-19), and that the harms would not be self-inflicted. Judge Bumatay disagreed (see pp. 63-76). If you’re interested in these issues, check out those opinions. I won’t elaborate further on them here because I doubt the Supreme Court will have occasion to resolve them. There will almost certainly be other plaintiffs with uncontested standing when the cases again reach the Court, and it’s likely the Court will simply decide the merits of those plaintiffs’ claims without having to grapple with the thorny issues of States’ Article III standing.[1]

DOJ’s more interesting standing argument in the citizenship cases invokes the Court’s prudential (i.e., not constitutional) “third-party standing” doctrine that even when a plaintiff has alleged injury sufficient to satisfy Article III’s “case or controversy” requirement, it “generally must assert [its] own legal rights and interests, and cannot rest [its] claim to relief on the legal rights or interests of third parties” (Warth v. Seldin (1975)). (The Court has recognized exceptions to this prudential rule in cases where the defendant applies its policy against the “third party” plaintiff itself or where the plaintiff has a close relationship with the person who possesses the right and there’s a “hindrance to the possessor’s ability to protect his own interests” (Kowalski v. Tesmer (2004)).) According to DOJ, this prudential doctrine precludes the States’ standing in the Massachusetts and Washington cases even if the States do meet the Article III injury thresholds, because the rights and interests at stake are those of the putative individual citizens against whom the Government would apply the Citizenship Order, and those individuals can protect their own interests in court.

In his concurring opinion in CASA, joined by Justice Thomas, Justice Alito indicated support for this argument: “So long as third-party standing doctrine remains good law,” he wrote, “federal courts should take care to apply these limitations conscientiously, including against state plaintiffs” (citing a recent article by Curt Bradley and Ernie Young arguing against third-party standing in cases of such “collateral” injuries).

If the Supreme Court has occasion to address the issue, however, there’s reason to think a majority of the Justices wouldn’t apply the Warth/Kowalski rule here. For one thing, no Justice even mentioned the issue in Diamond Alternative Energy v. EPA, a case decided a few days before CASA in which the Court affirmed a “third” party’s standing because it would suffer injury from the state’s application of a rule to another private party elsewhere in a supply chain. (California didn’t raise the third-party standing argument directly in Diamond Alternative, so it was waived—but the failure of any party or Justice even to flag it might be telling.)

More importantly (perhaps), in the 2014 Lexmark case, the Court suggested that such prudential standing limits are in “tension with … the principle that a federal court’s obligation to hear and decide cases within its jurisdiction is virtually unflagging”; and although the Court in Lexmark reserved the question of whether the third-party doctrine is such a prudential rule, a plurality of the Court in June Medical (2020) affirmed Kowalski’s assumption that the doctrine is, in fact, prudential, and only Justice Thomas took issue with that view. If the Court were to adhere to Justice Scalia’s view in Lexmark that the Court can’t deny plaintiffs their day in Court on “prudential” grounds, then it should turn aside the Government’s third-party standing argument.[2]

3.

As for the associational plaintiffs in the Maryland case—CASA, Inc. and the Asylum Seeker Advocacy Project (ASAP)—the Solicitor General argued in the Supreme Court that they have Article III standing only to represent those of their members who have identified themselves as having been injured by the Citizenship Order. Citing Justice Thomas’ sole concurrence in FDA v. Alliance for Hippocratic Medicine (2024), the SG contended that the Court’s established associational standing doctrine goes too far by permitting an association to seek relief for its entire membership on the basis of injuries to one or more of its members “even if the association has tens of millions of other, non-injured members.”

It is uncertain whether this issue will remain relevant in the CASA case itself on remand.  If the courts there grant the individual plaintiffs’ motion for classwide relief (something I discuss in Point 10, below), there might not be much of a need for CASA and ASAP to continue to seek a separate injunction to protect their members.  The issue remains a live one, however, in the New Hampshire case that will be argued in the First Circuit on Friday, which involves a preliminary injunction protecting all the members of the New Hampshire Indonesian Community Support association.

If the Supreme Court ever reaches the issue in one or more of these cases—which is unlikely—the Government’s argument probably won’t get any traction, for two reasons.

First, as Justice Thomas acknowledged in his Alliance for Hippocratic Medicine concurrence, the precedent is clear: “the Court consistently applies” associational standing principles to grant relief to all of an organization’s members if any one of them would have standing. Therefore, the Court would have to overturn numerous precedents  to abandon the doctrine—an unlikely result.

Second, there’s no practical reason to overturn those precedents—certainly not in the birthright citizenship cases, anyway. If many or most of an association’s members aren’t injured by the Citizenship Order—for example, because their parents were U.S. citizens or legal permanent residents (in which case the Order doesn’t deny them citizenship) or because they weren’t born in the United States (in which case they wouldn’t be birthright citizens on any reading of the Fourteenth Amendment)—then a judgment in their favor won’t benefit them in any way or harm the Government in any way. Thus, nothing of substance turns on the breadth of an association’s standing, i.e., on how many of its members are nominally covered by an injunction and whether anything in the Constitution requires them to identify themselves before the court issues an injunction.

4.

In his CASA applications, the Solicitor General asked the Court to hold that Article III of the Constitution generally forbids federal courts from issuing “universal” injunctions that provide relief to nonparties—even if Congress has authorized such relief. A constitutional holding to that effect would have been very significant. Among other things, it presumably would mean that the Administrative Procedure Act (APA) provision empowering reviewing courts to “hold unlawful and set aside agency action” could not be construed, as it long has been, to authorize courts to vacate unlawful agency rules. It also would have called into question various other statutes, such as the Hobbs Act and the Clean Water Act, in which Congress has given courts the power to categorically enjoin or vacate agency rules.

In footnote 4 of CASA, the Court declined the SG’s request for a constitutional ruling: “Our decision rests solely on the statutory authority that federal courts possess under the Judiciary Act of 1789,” wrote Justice Barrett. “We express no view on the Government’s argument that Article III forecloses universal relief.” And in footnote 10, the Court added that “[n]othing we say today resolves the distinct question whether the Administrative Procedure Act authorizes federal courts to vacate federal agency action.”

I think it is fairly safe to assume there aren’t five votes on the Court to hold that Congress can’t authorize courts to issue relief to nonparties. Indeed, as I note in Point 7 below, the Court itself issued such a universal injunction in the final sentence of its CASA opinion, something it couldn’t have done if Article III barred such relief. Nor, I suspect, would a majority of Justices read the APA not to authorize vacatur—a holding that would, in Chief Justice John Roberts’ words, “be fairly radical and inconsistent with” what D.C. Circuit judges do in APA cases “five times before breakfast.” See also Justice Brett Kavanaugh’s CASA concurrence (p.2): “[I]n cases under the Administrative Procedure Act, plaintiffs may ask a court to preliminarily ‘set aside’ a new agency rule.”

5.

The Court granted the Solicitor General’s stay application in one minor, discrete respect: To the extent any of the three preliminary injunctions under review could be read to preclude agencies from developing and issuing public guidance about their plans to implement the Citizenship Order, the Court lifted that restriction. Accordingly, after the decision in CASA, Section 3 of the Order required the agencies to promulgate such guidance by July 27. Several agencies have now done so. (See the documents linked at the top of this piece.)

Apart from that, did the Court grant any other aspect of the stay application?

It’s widely assumed that the Court stayed the effect of the three preliminary injunctions as applied to nonparties. That’s not quite right, however.

To be sure, in the final paragraph of her opinion for the Court, Justice Barrett wrote that “[t]he Government’s applications to partially stay the preliminary injunctions are granted.” She immediately added an important qualifier, however: “but only to the extent that the injunctions are broader than necessary to provide complete relief to each plaintiff with standing to sue.” In other words, the Court’s response to the SG’s principal application for a stay was: “It depends.”

What many observers of these cases didn’t appreciate was that the plaintiffs didn’t request universal injunctions primarily in order to provide relief to nonparties, as such; instead, they requested injunctions of that breadth in order to secure “complete relief” to the parties themselves. That was the overwhelming emphasis of the plaintiffs’ briefs in the Supreme Court, even if it wasn’t the question that most interested the Justices.

Significantly, the Court agreed with the respondents on the “complete relief” issue. It held that Congress has authorized federal courts to issue not only (a) certain forms of relief that incidentally benefit nonparties—Justice Barrett invoked the “archetypal” case (p.16) in which a plaintiff sues a neighbor for blasting loud music at all hours of the night and the court orders the defendant to turn it down—but also (b) an injunction that expressly directs a defendant to treat some or all nonparties in a certain way—such as “a blanket ban on the enforcement of the Executive Order” (p.18)—where that’s necessary in order to “offer complete relief to the plaintiffs before the court” (p.17). That’s the very theory of “universal” relief that the plaintiffs have emphasized in these very cases.

The Court itself did not decide whether a need for “complete relief” justifies the universal injunctions in these cases; instead, it remanded that inquiry back to the lower courts “to determine whether a narrower injunction is appropriate” (p.19). The Court directed those courts (p.26) to “move expeditiously to ensure that, with respect to each plaintiff, the injunctions comport with” the Court’s understanding of federal courts’ remedial authorities (including the courts’ power to issue universal injunctions if necessary to secure plaintiffs’ complete relief) “and otherwise comply with principles of equity” (p.26).

Accordingly, the Court did not grant the Solicitor General’s applications for partial stays of the preliminary injunctions—or, to be more precise, it purported to “grant” those applications contingent upon further findings by the lower courts.

6.

On pages 17-19 of her opinion, Justice Barrett discussed in detail the States’ arguments about why a universal injunction is necessary to ensure they wouldn’t be harmed by implementation of the Citizenship Order, but the Court declined to opine on whether those arguments pass muster.

In its opinion last week in the Washington case, the Ninth Circuit held (see pp. 48-49) that the district court did not abuse its discretion in issuing a universal injunction because such a remedy is necessary in order to provide complete relief to the plaintiff States there. (The appellees in the Circuit also included, in addition to the State plaintiffs, three individuals whose children would be covered by the Citizenship Order. The court of appeals purported to “dismiss” their claims (see p.21) because their children were recently included in a class certified in the District of New Hampshire. That dismissal was wrong for reasons I discuss in a footnote,[3] but the absence of the individual plaintiffs in the Washington case shouldn’t make much of a practical difference going forward.) Likewise, in the Massachusetts case, District Court Judge Leo Sorokin held on Friday that the 18 plaintiff States (and two localities) there met their burden of showing—on an elaborate factual record that the Government did not contest—that no narrower form of relief “would feasibly and adequately protect the plaintiffs from the injuries they have shown they are likely to suffer if the unlawful policy announced in the Executive Order takes effect during the pendency of this lawsuit.”

As a result of these two decisions last week, there remain in place at least two “universal” injunctions against implementation of the Citizenship Order. Even if the Solicitor General chooses to ask the Supreme Court to review these “complete relief” holdings, I think it’s unlikely the Supreme Court will adjudicate that question because (i) it would require the Court first to decide the State standing questions; (ii) the issue might not be of practical significance because one or more other, classwide injunctions will be in place that cover many or all of the same affected persons; and (iii) the Court will probably hear and resolve the substantive merits of the cases, which would pretermit the need to opine on the “complete relief” question.

Whether the Court left open a “complete relief” argument for a universal injunction on behalf of the associational plaintiffs in the Maryland case is a bit less clear.[4] However, in light of the fact that CASA’s and ASAP’s members will be covered by the classwide relief that Judge Boardman has indicated she will soon issue, and that they’re also protected by universal and classwide injunctions currently in place in other cases, there’s a good chance the courts in CASA won’t need to adjudicate CASA’s and ASAP’s argument that a universal injunction is appropriate in order to ensure their members aren’t injured. For that reason, I’ll relegate my discussion of that question to a footnote.

7.

Therefore, the Supreme Court did not hold that the Government may now implement the Citizenship Order with respect to nonparties.  To the contrary, in the final sentence of its opinion the Court added its own universal injunction, one that applies to all persons covered by the Trump Citizenship Order: “Consistent with the Solicitor General’s representation, § 2 of the Executive Order shall not take effect until 30 days after the date of this opinion.” If the Government were correct that Article III precludes such universal relief, the Court could not have issued that decree. What’s more, and as I discuss below, this SCOTUS injunction is difficult to square with the Court’s understanding about whether Congress has authorized federal courts to issue such relief to nonparties.

III. The Relative Insignificance of CASA’s (Tentative) Holding About Universal Injunctions

Of course, the question that dominated the Court’s opinion in CASA, and that grabbed all the headlines, is whether federal courts have statutory authority to issue “universal” injunctions other than where necessary to provide plaintiffs with complete relief.

8.

The Court declared its view on that question unequivocally early in Justice Barrett’s opinion: “Congress has granted federal courts no such power” (p.5). Technically, however, the Court hasn’t quite held as much. Because of the preliminary posture of the case at the preliminary injunction stage, all the Court decided was that Congress likely hasn’t provided federal courts with any such authority. Justice Barrett flagged the tentative nature of the Court’s conclusion at least three times: “These injunctions—known as ‘universal injunctions’—likely exceed the equitable authority that Congress has granted to federal courts” (pp. 1-2). “The Government is likely to succeed on the merits of its argument regarding the scope of relief” (p.5). “The question before us is whether the Government is likely to suffer irreparable harm from the District Courts’ entry of injunctions that likely exceed the authority conferred by the Judiciary Act. The answer to that question is yes” (p.25).

Justice Barrett’s caveats were especially appropriate in CASA for two reasons. First, the parties barely briefed the question, not only because the Court didn’t order full briefing on the Government’s stay application, but also because the parties emphasized the “Is it necessary for complete relief?” question discussed above. Second, because the Court acknowledged that courts can issue such universal relief to provide the parties with complete relief—and remanded the cases for the lower courts to adjudicate that question—the Court’s analysis of the broader question technically was dicta.

That said, there’s no doubt that six Justices believe that federal courts lack a general statutory power to issue “universal” injunctions (though the Court bracketed the question whether some statutes, such as the APA, might authorize such relief in certain circumstances). According to the Court’s Delphic decision last week in Trump v. Boyle, that conditional conclusion apparently will “squarely control[]” what lower courts should do when they’re considering requests for interim relief, even though it’s “not conclusive as to the merits.” I don’t think it ought to be “squarely controlling” when lower courts are deciding whether to issue ultimate relief against the federal Government, but that’s a formality:  Lower courts surely won’t disregard the Court’s view about the “likely” scope of their equitable authority simply because the Court hasn’t quite closed the door all the way. Therefore, as a practical matter, the tentative, conditional nature of the Court’s statutory holding will only be relevant if and when the Court includes five or more  Justices inclined to abandon it (in which case stare decisis wouldn’t stand as an obstacle).

That’s not happening anytime soon. So as a practical matter, the Court has resolved the question:  Federal courts generally can’t issue universal injunctions other than in order to secure complete relief to the plaintiffs.

How important is that resolution?

One might assume that it’s very significant, given the volume of heated debate about the question in recent years in the federal courts, in the pages of law reviews, and online. The Solicitor General’s Office, in administrations of both parties, has repeatedly implored the Court to declare that a lone district judge lacks the power to singlehandedly halt the operation of a federal program, particularly where that judge’s view might be an outlier. That argument clearly has found favor with several Justices—including Justices Sotomayor and Kagan, both of whom have suggested that district judges shouldn’t issue such “universal” injunctions in at least some cases. (See, for example, Justice Sotomayor’s dissent at p.23: “There may be good reasons not to issue universal injunctions in the typical case, when the merits are open to reasonable disagreement and there is no claim of extraordinary and imminent irreparable harm.”) Several scholars have likewise enthusiastically welcomed the Court’s decision. Nick Bagley, for example, wrote that “[n]ationwide injunctions are equal-opportunity offenders, thwarting Republican and Democratic initiatives alike. Today, it’s Trump’s birthright-citizenship order and USAID spending freezes. Yesterday it was mifepristone, the cancellation of student debt, and a COVID-vaccine mandate. Why should one federal judge—perhaps a very extreme judge, on either side—have the power to dictate government policy for the entire country? Good riddance.” Others condemned it. Mila Sohoni, for example, wrote: “I disagree intensely with the decision, which rests on an amalgam of bad history, fiat, and ipse dixit.” And in her dissenting opinion, Justice Jackson went so far as to warn that the Court’s decision might be “an existential threat to the rule of law.” Several prominent commentators expressed similar alarm.

For reasons I explain in Part IV below, I think that Professor Sohoni’s views on the merits of the question are closer to the mark. That said, I also think the Court’s decision isn’t as momentous as some scholars, Justices, and Executive branch officials have suggested, for a couple of reasons.

9.

First, the practical costs to the federal Government of such universal injunctions aren’t as severe as advertised. Like Sam Bagenstos, I was an Executive Branch attorney with “a front-row seat to myriad abuses” when “judges hand-picked by ideological plaintiffs for their hostility to the administration” issued universal injunctions that stymied administration initiatives right out of the gate with very little justification. Even so, I’m not sure that Sam is correct that such injunctions “created enormous … obstacles to governance”—any more than judicial vacaturs of agency rules under the APA have greatly impeded the federal Government from operating.

As I wrote back in 2018 (in connection with the universal injunction issued in the “Travel Ban” case), the cases where universal injunctions have raised the most controversy are those in which plaintiffs immediately challenge a new, controversial federal government policy. In such cases, it’s reasonable to assume the Supreme Court will adjudicate the legality of the policy in short order, probably within a term or two (or will at a minimum do so for purposes of preliminary relief). And once the Court rules, its decision will as a practical matter have the same effect as a universal injunction: either the Court will declare that the policy is lawful, in which case the Government will be able to implement it universally, or every affected person (not only the plaintiffs themselves) will benefit from a ruling against the Government. (Justice Kavanaugh echoes this point in his concurrence in CASA: “[T]he Court’s disposition of applications for interim relief often will effectively settle, de jure or de facto, the interim legal status of [contested] statutes or executive actions nationwide.”)

A district court’s universal injunction therefore merely establishes a nationwide status quo for the (usually) short interim period before the SCOTUS’s ultimate decision. And if DOJ thinks the Supreme Court is likely to disagree with the district court on the merits, it can expeditiously get the case before the Court and secure a reversal.

10.

On the flip side, denying courts the power to issue such injunctions shouldn’t have much of a practical impact, either—in most cases, anyway. For one thing, as Justice Barrett noted, district courts rarely issued such injunctions until the Obama Administration. This certainly made things more inefficient and costly for plaintiffs when the Government initiated an unlawful policy.  Sam Bray reports, for example, that between 1935 and 1937, courts issued approximately 1600 injunctions against implementation of the processing tax in the Agricultural Adjustment Act. That’s hardly an ideal way to run a railroad, in my book. But it wasn’t an intolerable system, either.

More importantly, the Court’s CASA decision is unlikely to revert us back to that sort of world, in which courts are issuing thousands of identical injunctions. As Justice Kavanaugh and several commentators have emphasized, there are at least three ways in which district courts can effectively provide “universal” relief against patently unlawful government action, even after CASA.

(i)

As discussed above, in many cases a universal injunction may be necessary to fully remedy a plaintiffs’ own injuries, in which case the district court can issue it. The Court in CASA specifically held that courts have the power to do so.

(ii)

Once a federal government policy is instantiated in an agency rule (as Trump’s birthplace citizenship policy is now, in the agency guidance documents), it can be challenged under the APA and a court of appeals can vacate (5 U.S.C. § 706(2)) or stay (5 U.S.C. § 705) that rule if it determines that it’s unlawful.

(iii)

And in many cases, plaintiffs ought to be able to bring a class action that protects all or most of the persons injured by a new Government policy. As Adam Liptak reports, class actions were mentioned at least 35 times in the CASA oral argument, and Justice Kavanaugh signaled that they could provide “a mechanism to do what’s needed here in terms of getting relief to people.” The Supreme Court even recently affirmed, in its A.A.R.P. decision, that district courts can issue a preliminary injunction protecting a “putative” class that hasn’t yet been certified. Notwithstanding CASA, that’s an injunction that expressly protects nonparties … which only goes to show that the Court’s categorical declarations in CASA about the absence of any such remedial power are questionable and difficult to square with the Court’s own practices when it comes to issuing equitable relief.

In the wake of the Court’s decision in CASA, plaintiffs in at least two birthright citizenship cases immediately moved to certify broad classes of plaintiffs and asked the district courts to issue preliminary injunctions protecting those classes.

On July 10, Judge Laplante in the District of New Hampshire did just that, with respect to a certified class defined as “[a]ll current and future persons who are born on or after February 20, 2025” and are covered by the two categories of the Trump Citizenship Order. He also preliminary enjoined implementation of the Trump Citizenship Order as to such children. Judge Laplante stayed that order for seven days to give the Government an opportunity to appeal, but for some reason the Government didn’t do so, and therefore the classwide injunction took effect on July 17.

In the Maryland case, the individual plaintiffs asked the district court to certify a class consisting of “all children who have been born or will be born in the United States on or after February 19, 2025, who are designated by Executive Order 14,160 to be ineligible for birthright citizenship, and their parents,” and also asked the court to issue a new preliminary injunction to protect the members of that class. In response, Judge Boardman ruled that she lacked jurisdiction over the preliminary injunction motion “because nearly every aspect of it is on appeal before the United States Court of Appeals for the Fourth Circuit”; but Judge Boardman further opined, as “an indicative ruling pursuant to Federal Rule of Civil Procedure 62.1(a)(3),” that “[i]f the Fourth Circuit were to remand for the limited purpose of allowing this Court to decide the motion for a class-wide preliminary injunction, the Court would grant the motion.” Judge Boardman also indicated that she intends to rule on the motion to certify the class “expeditiously.” The court of appeals remanded the case to the district court yesterday. It’s therefore likely that Judge Boardman will soon certify a class and issue a preliminary injunction protecting all of its class members.

Essentially for the reasons discussed in David Marcus’ excellent law review article, these (and perhaps other) class certifications, and injunctions, are very likely to survive any appeals by the Government.

One caveat is worth flagging, however, about the pending class actions: Those classes cover only persons born on or after Feb. 19, 2025. The plaintiffs presumably chose that date because Section 2 of the Trump Citizenship Order—the section prohibiting federal officials from issuing or accepting certain citizenship documents—by its terms applies only to such infants. Section 3 of the Order, however, has no such temporal limitation. It directs federal officials to “take all appropriate measures to ensure that the regulations and policies of their respective departments and agencies are consistent with this order, and that no officers, employees, or agents of their respective departments and agencies act, or forbear from acting, in any manner inconsistent with this order.” “This order,” of course, includes Section 1, in which President Trump has declared that no persons described therein are entitled to the privileges of birthright citizenship, regardless of their date of birth. As I read Section 3, therefore, it prohibits federal officials from treating such persons as U.S. citizens, full stop, whether they were born before February 20 or not.

In the Ninth Circuit oral argument, however, the attorney for DOJ represented that the Administration will apply the Order as a whole only to persons born after February 19, and thus that the Government will not, for example, try to remove from the United States someone described in Section 1 of the Order who has lived in the United States for many years. The agency guidance documents issued over the past few days—linked at the top of this article—similarly indicate that for persons born before February 20 (and their parents), the agencies will accept documents indicating birth in the United States as proof of their citizenship without more, i.e., without any further proof that their parents were citizens or lawful permanent residents on the person’s date of birth. (The SSA Guidance, for example, provides that for purposes of issuing Social Security numbers or cards, “a birth certificate showing a U.S. place of birth will not be sufficient documentary evidence of U.S. citizenship for persons born after the EO takes effect.” And the State Department Guidance specifies that an agent “must request” that an applicant for a passport “complete the parental information on the form DS-11 for individuals born in the United States on or after the effective date.”)

It’s not obvious to me why the Administration considers this to be a proper means of implementing Section 3 of the Citizenship Order. Indeed, if President Trump were correct that persons described in the two categories of Section 1 of the Order are not covered by the Citizenship Clause or the federal citizenship statute—a legal view that federal officials presumably would have to accept in the absence of an injunction—then it’s not apparent what authority executive officials would have to confer upon such persons any of the benefits and duties that are unique to citizens, unless such persons have been naturalized. (In their Supreme Court brief, the CASA plaintiffs warned of this possible application of the Citizenship Order to persons born before February 19, and in his reply brief the Solicitor General conspicuously didn’t offer any reason to question such a comprehensive reading of the reach of the Order.)

I suppose the agencies might have concluded that because it would be an administrative nightmare to determine whether the parents of persons born before February 20 are, or are not, described by Section 1 of the Order, it is reasonable for officials to apply a presumption that documentation of birth in the United States suffices to establish citizenship. But what about a case in which an agency learns of evidence that the parents of a person born in the United States before February 20 do fall within the confines of Section 1, such as proof that their father was a foreign national and they were born to a woman who was in the U.S. on a student visa? What then? Could the State Department deem that individual not to be a citizen and thus remove them from the United States? Could (or must) HHS, SSA, and other agencies cease providing them benefits that are statutorily limited to citizens and LPRs? The DOJ attorney represented to the Ninth Circuit panel that these sorts of things would not happen, but it is not at all obvious why they wouldn’t. Let’s hope the Citizenship Order never goes into effect so that such nightmare scenarios don’t arise.

(iv)

In addition to the steps described above that district courts might take to prevent the executive branch from applying an unlawful policy to nonplaintiffs (broad injunctions necessary to provide complete relief to the plaintiffs; class actions (in which case the class members are parties); and staying and vacating agency rules), Justice Kavanaugh opined in his separate concurring opinion in CASA that the Supreme Court itself has a responsibility to step in at an early stage to address the legality of a contested “major” executive policy in order to establish a uniform national treatment of that policy while its legality is being litigated (pp. 5, 8):

[I]n my view, there often (perhaps not always, but often) should be a nationally uniform answer on whether a major new federal statute, rule, or executive order can be enforced throughout the United States during the several-year interim period until its legality is finally decided on the merits. …

Often, it is not especially workable or sustainable or desirable to have a patchwork scheme, potentially for several years, in which a major new federal statute or executive action of that kind applies to some people or organizations in certain States or regions, but not to others. The national reach of many businesses and government programs, as well as the regular movement of the American people into and out of different States and regions, would make it difficult to sensibly maintain such a scattershot system of federal law. …

That temporary geographic, organizational, and individual variation in federal law might not warrant this Court’s intervention in run-of-the-mill cases—which is why it makes sense that this Court denies applications for interim relief when the Court is unlikely to later grant certiorari. See Does 1–3 v. Mills, 595 U. S. ___, ___ (2021) (BARRETT, J., concurring in denial of application for injunctive relief). But in cases involving major new federal statutes or executive actions, uniformity is often essential or at least sensible and prudent. In those kinds of cases, disuniformity—even if only for a few years or less—can be chaotic. And such chaos is not good for the law or the country.

It’s difficult to overstate the tension between Justice Kavanaugh’s perspective here and that of many opponents of universal injunctions—and of the Court itself in United States v. Mendoza (1984), where it unanimously held that nonmutual collateral estoppel does not apply against the federal Government. In the eyes of those opponents (and the Mendoza Court), the “patchwork scheme” described by Kavanaugh is a virtue rather than (in Kavanaugh’s words) “unworkable or intolerable.”

To the extent a majority of the Court now shares Kavanaugh’s view that, for the sake of national uniformity, the Supreme Court “should not and cannot hide in the tall grass” (p.7), then even in a situation where lower courts disagree about whether to grant preliminary relief to individual plaintiffs (or to small groups of plaintiffs), the Court can and should expeditiously grant a certiorari petition filed by a losing party in order to establish a uniform (if perhaps temporary) national rule. (Or if a losing plaintiff applies for an injunction from the Supreme Court itself, the Court should, at least according to Justice Kavanaugh, rule exactly as it would do if the Government were applying for a stay of an injunction: “The standards should mesh so that this Court can ensure uniformity without regard to the happenstance of how various courts of appeals and district courts ruled.”)

If that becomes the new norm, then that’s yet one more reason—together with “complete relief” injunctions, classwide injunctions, and APA vacatur—that the Court’s conditional holding in CASA about courts’ equitable authority shouldn’t be of vast practical importance. In light of all these alternative avenues for broad relief, the suggestion that the Court’s decision in CASA “amounts to a revolution in the remedial practices of the lower federal courts” will, I suspect, prove to be greatly overstated. In particular, and notwithstanding CASA, I think it is very unlikely the Trump Administration will be able to implement the Citizenship Order any time before the Supreme Court rules on the merits of the President’s new definition of citizenship.

IV. Why the Court was Probably Mistaken About Federal Courts’ Statutory Authority to Issue Universal Injunctions

As explained in Part III above, I don’t think the answer to the “merits” of the universal injunction question is as important as many have suggested. And as I discuss in Part V below, I think other aspects of the majority’s opinion are much more troubling. For what it’s worth, however, I have significant doubts about Justice Barrett’s analysis of the statutory authority question. Others undoubtedly will write at much greater length about this topic. For the time being, here are some initial impressions about why I don’t find the Court’s opinion very convincing.

11.

For starters, Jack Goldsmith argues that the Court asked and answered the wrong question.

According to the Justice Barrett (p.4), “[t]he issue before us is … whether, under the Judiciary Act of 1789, federal courts have equitable authority to issue universal injunctions.” The 1789 Act, she explained (p.5), “endowed federal courts with jurisdiction over ‘all suits . . . in equity,’ and still today, this statute ‘is what authorizes the federal courts to issue equitable remedies.” Barrett then invoked the Court’s decision in Grupo Mexicano de Desarrollo, S.A. v. Alliance Bond Fund, Inc. (1999), which held that the 1789 statute’s reference to “equity” “encompasses only those sorts of equitable remedies ‘traditionally accorded by courts of equity’ at our country’s inception.” “We must therefore ask,” wrote Barrett (p.6, again quoting Grupo), “whether universal injunctions are sufficiently ‘analogous’ to the relief issued “‘by the High Court of Chancery in England at the time of the adoption of the Constitution and the enactment of the original Judiciary Act.’” (For what it’s worth, Justice Barrett clerked for Justice Scalia the Term he wrote Grupo Mexicano.) The Court then concluded (id.) that “[n]either the universal injunction nor any analogous form of relief was available in the High Court of Chancery in England at the time of the founding.”

As Jack Goldsmith points out, however, whereas the 1789 Act established federal court jurisdiction for diversity cases, such as Grupo Mexicano itself, CASA instead involved suits brought against federal officers pursuant to the court’s federal question jurisdiction, which Congress conferred on lower federal courts in Section 1 of the Jurisdiction and Removal Act of 1875. (Congress first conferred such federal question jurisdiction in 1801, in Section 11 of the “Midnight Judges Act,” but repealed that authority the next year, and it wasn’t revived until 1875.) Like the 1789 Act, the 1875 Act referred to certain sorts of suits—including both diversity and federal question cases—“of a civil nature at common law or in equity.” It follows, according to Goldsmith, that if it’s appropriate to apply a Guapo Mexicano-like “incorporating extant law” reasoning to “suits in equity” for federal-question jurisdiction cases, the Court “should have looked to the state of equity in 1875, not at the founding.” And, as Goldsmith notes, “[p]rinciples of equity and equitable remedies changed in nontrivial ways between 1789 and 1875, and were in flux in the post-Civil War period.”

Goldsmith is right about that much: the Court asked the wrong question, about the wrong federal statute. If it’s proper to apply Grupo Mexicano’s reasoning, then the Court ought to have analyzed the remedial practices of equity courts as of 1875, rather than simply the sorts of relief the High Court of Chancery in England issued in 1789. Moreover, if the Court had done so, that might have made a difference in the outcome of the inquiry because, as Justice Barrett herself pointed out (pp. 14-15), some state courts in the Nineteenth Century had begun to enjoin governments’ collection of unlawful taxes “against an ‘entire community,’ even when a ‘single taxpayer su[ed] on his own account’” (quoting Pomeroy’s Equity Jurisprudence); see also Justice Sotomayor’s dissent at pp. 19-20). Justice Barrett dismissed the relevance of such “taxpayer suits” (p.15)—characterizing them as an “inadequate historical analogy”—largely because “the practice of extending relief ‘with respect to any taxpayer’ was not adopted by state courts until the mid-19th century.” If, however, 1875 equity jurisprudence is the appropriate touchstone for federal-question cases, as Goldsmith suggests it should be, then the Court would not have been able to so easily dismiss the compelling taxpayer suit analogy.

There’s a catch, however. Rightly or wrongly, in at least a couple of old cases the Court held that the inquiry about British practices in 1789, of the sort approved in Grupo Mexicano, is appropriate not only in diversity cases but in federal question cases, as well.

In 1911, Congress codified the language of the 1875 Act, which referred to suits “in equity” for both diversity and federal question jurisdiction, in section 24 of the Judicial Code. (It was then found in 28 U.S.C. § 41(1); today it’s in 28 U.S.C. § 1331.) In 1932, in Matthews v. Rodgers—a federal question case—the Court opined that “[t]he equity jurisdiction conferred on inferior courts of the United States by section 11 of the Judiciary Act of 1789, and continued by section 24 of the Judicial Code (28 USCA § 41), is that of the English court of chancery at the time of the separation of the two countries.” In other words, the Court appears to have concluded—albeit without explanation—that when Congress expanded federal jurisdiction to cover federal question cases in 1875 (and then codified that authorization in 1911), it incorporated the “Court of Chancery as of 1789” gloss on the reference to “suits in equity” that the 1789 Act had established.

It’s reasonable to think the Court’s unexplained conclusion in Matthews was much too quick. (Is it really fair to attribute to the 1875 and 1911 Congresses an intent that courts must disregard the development of equity after 1789?) Nevertheless, it is precedent, and assuming the current Court would follow it, then presumably it would have come to the same conclusion that it announced in CASA, even if it had acknowledged that Congress established the relevant jurisdiction in 1875 rather than in 1789. In particular, I assume the Court majority would have felt free to disregard the development of equitable remedies, such as in taxpayer suits, that had become common in the years between 1789 and 1875.

This is further demonstration why the “What would British equity courts have done in 1789?” analysis the Court has invoked in cases such as Grupo Mexicano and CASA is fundamentally misguided. That, however, is a broader problem that predated CASA. Rightly or (more likely) wrongly, the Court has long been committed, at least on paper, to the view that Congress has confined federal courts’ authority to issue remedies that aren’t closely analogous to those issued by the Court of Chancery in the Eighteenth Century, whether the court is exercising its diversity jurisdiction or (as here) its federal-question jurisdiction.[5]

12.

OK, so now let’s turn to why I have doubts about the Court’s substantive analysis of federal courts’ equitable remedial powers, even assuming the propriety of the Grupo Mexicano inquiry.

Justice Barrett concedes (pp.10-11) that, even under Grupo Mexicano, “equity is flexible,” and that a “modern device need not have an exact historical match” with something the Court of Chancery did in 1789. She nevertheless insists (p.11) that universal injunctions do not have a “sufficiently comparable predecessor” found “at the time of our country’s inception.” Justice Barrett reaches that result, however, only because she adopts an exceedingly narrow, and selective, view of what counts as a “sufficiently comparable” remedy. The primary candidate for a founding-era analogy, invoked by the plaintiffs and by amicus Mila Sohoni, is the bill of peace, a form of litigation recognized in Eighteenth Century English courts in which a ruling for the plaintiff redounded to the benefit of similarly situated injured parties. Justice Barrett devotes a paragraph of her opinion (pp. 12-13) to the bill of peace.  She concludes that it is not sufficiently analogous to the modern “universal injunction.” Her reasoning, however, isn’t terribly convincing.

Justice Barrett begins by characterizing a bill of peace as involving “‘group [that] was small and cohesive.’” She doesn’t, however, cite any English legal authority as support for the proposition that the protected group had to be either small or cohesive (the only thing she cites is the Sam Bray article in which that quotation appears), nor does she respond to Justice Sotomayor’s assertion (p.24) that “[t]here is no equitable principle that caps the number of parties in interest.” More to the point, Justice Barrett doesn’t explain why the group of persons who benefit from the universal injunctions in the birthright citizenship cases wouldn’t be sufficiently “small and cohesive,” even if that were a necessary condition: After all, that group consists only of those persons covered by Trump’s executive order, and no one else. (The group’s cohesiveness is demonstrated by the straightforward exercise of the recent class certifications.)

That brings us to Justice Barrett’s principal ground of distinction: She insists (p.13) that whereas a modern court’s denial of a request for a universal injunction would bind only the parties to the suit, a denial of a bill of peace in 1789 “would bind all members of the group, whether they were present in the action or not.” I don’t know whether, in fact, such “bindingness of an adverse judgment” was an unyielding requirement of a suit seeking a bill of peace. Justice Barrett’s own account suggests that it might not have been: She quotes Justice Story as writing that there was such a binding effect “in most, if not in all, cases of this sort.” Nor is it clear how frequently the issue even arose in bill of peace litigation—i.e., how often “covered” nonparties were precluded from bringing a subsequent suit after a court had denied a plaintiff’s request for a bill of peace.

More importantly, even if Justice Barrett’s distinction turns out to be historically accurate, she offers no reason why it’s a distinction that ought to make a conclusive difference for purposes of identifying federal courts’ equitable powers. In response to Justice Sotomayor’s accusation (p.29) that the Court was treating equity as if it were “fr[ozen] in amber . . . at the time of the Judiciary Act,” Justice Barrett insisted (p.10) that “[w]e said it before, … and say it again: ‘[E]quity is flexible.’” (quoting Grupo Mexicano). All that is necessary, Justice Barrett acknowledged, is that the remedy in question be “within the broad boundaries of traditional equitable relief.” But if that’s so, why don’t universal injunctions of the sort at issue in these cases fall well within the “broad boundaries” established by bills of peace, even if they aren’t precisely coterminous? Justice Barrett doesn’t say. She simply quotes Judge Sutton to the effect that “[t]he domesticated animal known as a bill of peace looks nothing like the dragon of nationwide injunctions.” That’s certainly an evocative metaphor; but Justice Barrett doesn’t explain why one creature is any more fire-breathing than the other.

Nor does she convincingly distinguish the plaintiffs’ proposed bill-of-peace analogy here from the analogy on which the Court has relied in its recognition of Ex parte Young injunctions. In footnote 9 of her opinion (p.11), Justice Barrett notes that there was some precedent in equity for suits similar to those seeking Ex parte Young injunctions—namely, that “a court of equity could issue an antisuit injunction to prevent an officer from engaging in tortious conduct.” That much is true. Yet the only such equity cases Justice Barrett cites in support of that proposition were from the Nineteenth Century—just like the taxpayer suits that she dismisses as proper precedents for the universal injunction because they appeared too long after 1789. Moreover, although Ex parte Young actions are analogous in some respects to the antisuit injunctions on which Justice Barrett relies, in other respects they’re far broader (or apply to very different sorts of government action), as Professor Sohoni has explained. In other words, the antisuit injunctions that Barrett invokes are partly analogous to Ex parte Young injunctions, just as bills of peace are analogous to universal injunctions. Indeed, if anything, there’s a closer correspondence in the latter case. Thus, if it’s right, as Justice Barrett agrees, that Ex parte Young injunctions fall “within the broad boundaries of traditional equitable relief,” universal injunctions of the sort at issue in these cases should, too. Yet the Court decrees, without much explanation, that they don’t.

13.

The Court in CASA also overlooked Grupo Mexicano’s recognition (527 U.S. at 326) that courts of equity will ‘“go much farther both to give and withhold relief in furtherance of the public interest than they are accustomed to go when [as in Grupo Mexicano itself] only private interests are involved’” (quoting and distinguishing United States v. First Nat’l City Bank (1965), which in turn quoted Virginian R. Co. v. Railway Employees (1937)). The Court has frequently relied upon this understanding about the broader scope of equitable relief in cases implicating the public interest—see, e.g., the cases cited in Virginian Railway, 300 U.S. at 552. It did so again in 2015, in Kansas v. Nebraska; there, the Court reaffirmed that “[w]hen federal law is at issue and ‘the public interest is involved,’ a federal court’s ‘equitable powers assume an even broader and more flexible character than when only a private controversy is at stake’” (quoting Porter v. Warner Holding Co. (1946)). [It’s noteworthy, I think, that the Court hasn’t bothered to cite any Eighteenth Century analogies in Chancery when it has recognized this equitable power to advance the “public interest”—just as it didn’t inquire in CASA itself about whether the Court of Chancery ever protected nonparties to ensure “complete relief” for a plaintiff. The Court’s reliance on the equitable state of play circa 1789, in other words, has been suspiciously selective.]

Similarly, the Court has long held that in deciding whether to stay a lower-court ruling—a context in which the court is also acting in its equitable capacity—an appellate court must consider “‘where the public interest lies.’” Nken v. Holder (2009) (quoting Hilton v. Braunskill (1987)). (It’s noteworthy, I think, that the Court has not, to my knowledge, ever pointed to any practice in the Court of Chancery in 1789 to justify this consideration of the “public interest.”) As authority for the “public interest” consideration, Nken cited Hilton, in which the Court relied upon Virginia Petroleum Jobbers Ass’n v. Federal Power Comm’n (D.C. Cir. 1958), a decision in which the court of appeals explained that because “[p]arties aggrieved by administrative agency orders act as representatives of the public interest in seeking judicial review,” “no artificial restrictions of the court’s power to grant equitable relief in the furtherance of that interest can be acknowledged.” Virginia Petroleum in turn cited Scripps-Howard Radio v. FCC (1942), in which the Supreme Court referred back to the proposition from Virginian Railway, quoted above, that “[c]ourts of equity may, and frequently do, go much farther both to give and withhold relief in furtherance of the public interest than they are accustomed to go when only private interests are involved.”

The Court in CASA didn’t so much as mention these venerable equitable principles requiring consideration of the public interest, even though the birthright citizenship cases certainly implicate public interests of the highest order.

14.

Finally, and perhaps most conspicuously, Justice Barrett didn’t persuasively account for the Supreme Court’s own practices involving equitable relief, including at least three recent instances in which it has issued or affirmed injunctions protecting nonparties.

For starters, in footnote 7 of her opinion Justice Barrett rejected plaintiffs’ reliance on a handful of early Twentieth Century decisions, including most famously Pierce v. Society of Sisters (1925), in which the Court approved or even issued universal injunctions. Those decisions, she explained, came too late—more than a century after 1789—and in none of them did the Court discuss the propriety of the universal relief. The respondents invoked those cases, however, not because they have “precedential effect” (p.9 n.7, quoting Steel Co.), but instead precisely because it appears that none of the parties or the Justices in those cases thought there was anything legally dubious about such injunctions—a “dog that didn’t bark” argument. The fact that the issue went unremarked—even by the federal Government in Lewis Publishing Co. v. Morgan (1913), with respect to an injunction against any implementation of a federal statute—ought to have at least given the Court pause about its fundamental assumption that federal courts aren’t authorized to grant relief that deviates in any respect from forms of relief recognized in Chancery in 1789. The Justices, the Solicitor General, and other parties in those Twentieth Century cases appeared to assume that equity is far more flexible than that.

Even more problematically, Justice Barrett’s opinion doesn’t come to terms with at least three recent injunctions of the Court’s own making. First, in Trump v. International Refugee Assistance Project (2017), the Court cut back somewhat on “nationwide” injunctions lower courts had issued against implementation of President Trump’s “travel ban” executive order, but specifically approved the application of those injunctions to any and all foreign nationals with a bona fide relationship with a person or entity in the United States, whether or not such U.S. persons were among the plaintiffs. The Court drew this line “as an exercise of discretion and judgment,” applying “the equities of [the] given case,” including the “interests of the public at large.” 582 U.S. at 579. And it did so over a three-Justice partial dissent that objected that the Court’s relief extended beyond what was necessary to provide complete relief to the plaintiffs. Id. at 585 (Thomas, J., joined by Alito and Gorsuch, JJ.). It wasn’t, in other words, a “drive-by” ruling.

Second, just two weeks after the oral argument in CASA itself, in A.A.R.P. v. Trump, No. 24A1007, the Court not only issued temporary relief to a “putative” (i.e., not-certified) class of persons in danger of removal to El Salvador, but also explained that because federal courts (including a district court) “may issue temporary relief to a putative class, … we need not decide whether a class should be certified as to the detainees’ due process claims in order to temporarily enjoin the Government from removing putative class members while the question of what notice is due is adjudicated.” As Professor Sohoni wrote, a preliminary injunction that shields a putative injunctive class “is essentially just the same thing as a universal preliminary injunction—both types of order constrain the defendant’s conduct with respect to non-parties even when that relief is not indivisible and even when that relief is not necessary to secure complete relief for the named plaintiff. Neither type of order creates preclusive effects on non-parties. And both types of order have thus been treated as equally suspect by those who contend that Article III courts may never extend protection to non-parties.” Unsurprisingly, then, in his dissent in A.A.R.P., Justice Alito, joined by Justice Thomas, questioned the Court’s authority to confer such relief upon nonparties; he argued that such relief is appropriate only after the class is certified and the class members have been made parties to the suit. Yet the Court issued its order in A.A.R.P. nevertheless. Obviously, it did so with full knowledge of the “relief to nonparties” issue—indeed, presumably after the Justices had voted at conference on the resolution of CASA.

Last but hardly least, in the final sentence of its opinion in CASA itself, the Court decreed that “§ 2 of the Executive Order shall not take effect until 30 days after the date of this opinion,” i.e., until this past Sunday (July 27). That injunction would have constrained Executive branch officials even if no lower courts had, in that span, certified a class or reaffirmed existing universal injunctions on grounds that they’re necessary to afford the parties complete relief. The Court did not explain how its own injunction might be reconciled with its tentative understanding that Congress hasn’t authorized federal courts to award such relief.

V. The More Serious Problems with the Court’s Ruling

Regardless of the merits of the Court’s analysis of federal courts’ statutory authority to issue universal injunctions, the Court’s decision to (nominally or provisionally) grant the Solicitor General’s application for partial stays of the lower-court injunctions in the three birthright citizenship cases was problematic for several reasons.

15.

For starters, I share the reaction of many observers (see, e.g., Sam Bagenstos here) that the Court exercised poor judgment in choosing to opine about the universal injunction question in these cases.

As Justice Barrett noted (p.4), in recent years Solicitors General in administrations of both parties, as well as several members of the Court, have repeatedly urged the Court to “review the question whether Congress has granted federal courts the authority to universally enjoin the enforcement of an executive or legislative policy.” In virtually all of those previous cases, a district court had issued a universal injunction based upon its assessment of a sharply disputed and unresolved question of the legality of the Government action—or, in some cases (such as Judge Kacsmaryk’s mifepristone decision), based upon the court’s idiosyncratic, outlier doubts about a longstanding Government practice that the vast majority of courts would have considered to be lawful.

As Justices Sotomayor and Jackson stressed in their dissenting opinions, however, the birthright citizenship cases aren’t remotely similar to those earlier ones. In this instance, the President directed Executive branch officials to act in a way that would be inconsistent with the settled constitutional understanding of all three branches for over a century. Indeed, as I’ve discussed elsewhere, the Department of Justice had long been of the view that it’s not even a close question. Therefore, as Justice Kagan noted at argument, this was a case in which “there [was] not going to be a great deal of disagreement among the lower courts.” To the contrary. As Justice Sotomayor wrote (p.23), “this is not a scenario where granting universal relief will encourage forum shopping or give plaintiffs the upper hand. Quite the opposite: By awarding universal relief below, the District Courts just ordered the Government to do everywhere what any reasonable jurist would order the Government to do anywhere.”

Yet when those district courts unsurprisingly began to issue injunctions preventing such patently unlawful conduct, the Solicitor General asked the Court to stay the universal effect of such injunctions without challenging the substantive basis for the injunctions—that is to say, without even offering any argument that the Government is likely to prevail on the merits. Worse still, the Solicitor General chose to tender such a bespoke application for what appears to be a manifestly illegitimate reason—namely, to enable the Administration be able to treat many U.S.-born individuals as noncitizens for an extended period, knowing that such treatment would be enjoined in virtually every such case if only the injured party had the knowledge, courage and resources to file suit. In other words, as Justice Sotomayor noted in her dissent (p.42, n.11), “the Government has adopted a plainly unconstitutional policy in defiance of this Court’s precedent and then gamed the system to stymie this Court’s consideration of the policy’s merits.” The intended result of this stratagem, if the Government were successful, would be (as Justice Kagan put it at oral argument) that for months or years there would be an “untold number of people who, according to all the law that this Court has ever made, ought to be citizens who are not being treated as such.”

The Court shouldn’t have rewarded the Administration’s gamesmanship. To be sure, many of the Justices were obviously eager to push back on the scope of district court injunctions in cases challenging Executive branch initiatives. Yet the Court chose not to do so in any number of earlier cases involving challenges to Biden Administration initiatives, where the Executive branch had a much stronger basis for acting than the Trump Administration does here. The Court could have continued that forbearance a bit longer, until the issue was raised in a less momentous Trump case where the underlying basis for a court’s universal injunction was more dubious, or at least contested. Alternatively, the Court could have addressed the universal injunction issue in conjunction with a later decision on the merits of the legality of the Citizenship Order itself. Had it done so, it could have accomplished the same effect on district court practices, but in a context in which the “top line” result would be to preclude the Executive branch from acting in a patently unlawful way.

16.

The Court passed up yet another option, as well—an alternative that it should have been eager to embrace, even if it insisted on using CASA as the vehicle for opining on the universal injunction question. The Court could have held that generally it is not “equitable” for a district court to issue a universal injunction against the Executive branch, but that the birthright citizenship cases present a rare exception in which such relief is equitable precisely because the Government is seeking to act in contravention of a longstanding consensus view of the three branches about a fundamental constitutional guarantee. Justice Sotomayor’s dissenting opinion indicates that such a ruling likely would have secured the votes of at least eight, and possibly all nine, Justices: “There may be good reasons,” she wrote (p.23), not to issue universal injunctions in the typical case, when the merits are open to reasonable disagreement and there is no claim of extraordinary and imminent irreparable harm. … The universal injunctions in these cases, however, are more than appropriate.”

The majority’s failure to accept Justice Sotomayor’s invitation for an 8-1 or 9-0 opinion along those lines was an unfortunate missed opportunity.

* * * *

Which brings me, finally, to what I consider to be the most alarming errors in the majority opinion. They appear not in its analysis of the statutory question involving universal injunctions, but instead in the final section of the opinion, Part IV, where the Court turned to the more case-specific question of whether it was appropriate to grant the Solicitor General’s application for a partial stay of the injunctions in the three cases before it.

In Nken, the Court explained that such a stay “is not a matter of right,” even if the party seeking the stay is likely to prevail at the end of litigation and, indeed, “‘even if irreparable injury might otherwise result’” (quoting Virginian R. Co.). The propriety of a stay depends upon “‘the circumstances of the particular case,’” and the party requesting the stay bears the burden of showing that those circumstances justify an exercise of the court’s discretion.

The Court should have held that the Government failed to meet that burden here even if it was correct about the limits of federal courts’ statutorily authorized equitable remedial powers.

17.

As I noted above in Point 13, when deciding whether to stay a lower-court ruling, one of the factors the Court is supposed to consider is “‘where the public interest lies,” Nken v. Holder (2009) (quoting Hilton v. Braunskill (1987)), just as federal courts are supposed to take the “public interest” into account in deciding whether to issue preliminary relief in the first instance (as the lower courts did in the birthright citizenship cases). So, for example, in Trump v. IRAP (2017), discussed in Point 14, above, the Supreme Court tailored the preliminary injunction to protect some nonparties but not others—and to protect the First Amendment associational interests of countless numbers of American “individual[s]” and “entit[ies]”—“as an exercise of discretion and judgment,” applying “the equities of [the] given case,” including the “interests of the public at large.” 582 U.S. at 579.

The Court in CASA referenced the Nken standards for a stay three times (pp. 5, 24, 25). Yet, in stark contrast with Trump v. IRAP, it never even mentioned the “public interest,” let alone assessed how it might bear on whether the Court should grant the stay applications. That oversight is particularly unfortunate in the context of the birthright citizenship cases, where the public interest points so squarely against the Government’s effort to enforce, for as long as possible, a policy that all three branches have long viewed as unconstitutional. (For more details on the “public interest” factor, see pages 20-22 of CASA’s brief in the Supreme Court; see also New Jersey’s brief at 17:  “[I]t is never in the public interest for any applicant, let alone the Executive Branch, to implement policies directly contrary to this Court’s decisions—without even trying to show that this Court’s decisions and the Executive’s actions can cohere. Indeed, the States know of no case, and applicants have identified none, in which this Court granted relief to allow such an undisputed conflict with its precedents. Taking that step here would contravene the principles of caution and prudence that traditionally guide emergency requests.”)

18.

As the Court explained in Nken, an appellate court considering a stay application must also consider “whether the stay applicant has made a strong showing that he is likely to succeed on the merits” and “whether the applicant will be irreparably injured absent a stay.”

In part IV of CASA, the Court concluded that “the Government” would suffer irreparable harm if it were enjoined from applying the Citizenship Order to parties not before the court, regardless of whether it would be constitutional to implement that order and even, apparently, regardless of whether the Supreme Court itself is likely to effectively prohibit the Government from implementing the order when it eventually adjudicates the merits. In support of that conclusion, Justice Barrett invoked a variation on a proposition Chief Justice Rehnquist and Chief Justice Roberts have occasionally articulated, which is that when a government is enjoined from enforcing a preferred policy, that creates a harm that’s necessarily irreparable if the government eventually prevails in the suit. The apparent logic behind this view is that in such a case the government will have been temporarily denied the ability to govern as it saw fit (or, in this case, as the Constitution allegedly requires)—a bell that can’t be unrung for purposes of the period when the injunction was in place.

Others have sharpy criticized this theory of governmental irreparable harm. But even if it makes some sense in a case where the government is ultimately likely to prevail in the litigation—or even where the underlying legal question is a close call—the birthright citizenship cases aren’t that. Here, the Government policy at issue is one that all three branches have long considered unconstitutional. And the odds are overwhelming that the Supreme Court will conclude that the Citizenship Order is unlawful when it finally addresses the question. Thus, as Justice Sotomayor writes in dissent (p.16), “by enjoining the Government from violating settled law, the District Courts’ orders do not cause the Government any harm” at all, let alone irreparable harm.

Why did the majority conclude otherwise? Its fundamental mistake is the way in which it treated the first Nken factor, namely, whether the defendants have “made a strong showing that [they are] likely to succeed on the merits.” Of course the defendants can’t do so in these cases. So how can Justice Barrett possibly assert that they’ve satisfied the “likelihood of success” inquiry?

First, Justice Barrett treats the ultimate constitutional issue here—the legality of Trump’s new definition of citizenship—as being, in effect, up for grabs (p.25): “[W]e take no position on whether the dissent’s analysis is right.” That statement is either a very disturbing harbinger of what some of the Justices think about the merits or a shockingly tone-deaf formalism that purports to deny what everyone knows to be true.

Either way, Justice Barrett then proceeds to suggest (p.25) that, in any event, the first Nken factor doesn’t require the stay applicant to make any showing about whether it’s likely to prevail in the suit itself: “That is not how the Nken factors work.” Instead, the Nken “likelihood of success” factor, according to Justice Barrett, requires the party seeking a stay merely to demonstrate that it’s likely to prevail on the discrete question it has put before the Court (here, the permissible scope of the preliminary relief the district court could issue), regardless of whether the party is ultimately likely to prevail in the litigation. And, the Court continues, the Government has made such a showing in these cases because any ultimate judgments against it likely won’t extend to nonparties. It follows, according to the Court, that “the Government is likely to suffer irreparable harm” from the universal preliminary injunctions because they extend further than any valid final judgments could do.

The first problem with this analytical framing is that it takes an overly narrow view of who constitutes “the Government” in a case such as this. Donald Trump and his agency officers are not the entirety of “the Government.” It includes Congress and the Supreme Court, too. And if and when the Administration were to implement the Citizenship Order, that will work to the detriment of the other two branches: The Executive branch will violate Congress’ enactments, and disregard the Court’s controlling precedent (Wong Kim Ark). Precluding a universal injunction therefore doesn’t serve the interests of those coordinate branches (nor the interests of the States that ratified the Fourteenth Amendment). More importantly, the defendants here are representatives of the United States, i.e., of the People as a whole, who undoubtedly are injured when their executive representatives violate the law, whether those violations are in the context of individuals who’ve already sued or those who haven’t.

Nor does the Court’s reasoning make a whole lot of sense, at least in a case such as this one. Justice Barrett reasoned that, in applying Nken, the Court asks whether the applicant “is likely to prevail on the merits of the issue before us, not whether he is likely to prevail on the merits of the underlying suit.” That proposition is in tension, however, with the Court’s statement one year earlier, in Ohio v. EPA (2024), that this first Nken factor concerns “who is likely to prevail at the end of th[e] litigation.”[6] In the birthright citizenship litigation, the defendants are virtually certain to lose at the end of the litigation.

Indeed, the Trump Administration is not likely to “prevail at the end of the litigation” in any meaningful sense, even with respect to nonparties. In footnote 18 of her opinion, Justice Barrett takes Solicitor General Sauer at his word that not only would he petition for certiorari when the Government lost “one” of the cases currently in the court of appeals (something that’s now happened), but also “that the Government will respect both the judgments and the opinions of [the Supreme] Court.” Therefore, if the Court holds, as it should, that the Citizenship Order is unlawful, the Administration will never enforce it against anyone.

What’s more, as Justice Sotomayor noted (p.16), Justice Barrett’s reasoning leads to absurd results, at least in a case where, as here, the party moving for a stay clearly lacks legal authority to do the thing the district court injunction purportedly can’t reach: “Suppose an executive order barred women from receiving unemployment benefits or black citizens from voting. Is the Government irreparably harmed, and entitled to emergency relief, by a district court order universally enjoining such policies? The majority, apparently, would say yes.” Yet “it strains credulity” even to treat the Executive branch “as irreparably harmed by injunctions that direct it to continue following settled law” (dissent at p.17). As the Ninth Circuit put the point in its opinion last week, the federal Government “cannot reasonably assert that it is harmed in any legally cognizable sense by being enjoined from constitutional violations” (internal citation omitted).

Accordingly, the Court’s conclusion as to irreparable harm can’t possibly be equitable (which is, after all, the touchstone for deciding whether to grant a stay)—unless the Court might actually rule in the Government’s favor on the merits of the legality of the Citizenship Order, i.e., unless Justice Barrett’s statement that “we take no position” on the underlying legal question means that the issue is actually up for grabs.

I doubt that’s the case.  Or, at least, I sure hope that at least two of the Justices in the CASA majority agree with the three dissenters that Trump’s definition of birthright citizenship is indefensible. (In my follow-up essays, I’ll explain why the Government’s principal arguments in support of the Order are mistaken.) If that’s correct, then that’s all the more reason the Court shouldn’t have chosen this moment—this case—to issue an opinion about the limits of district courts’ authority to enjoin unlawful executive actions.

– – – – – – – – – –

[1] In the unlikely event the Court reaches the question, it might be difficult for the Court to distinguish its approval of State standing in similar circumstances in the recent Biden v. Nebraska student-loan case and the apparent willingness of at least four Justices to do the same in the case challenging the Obama Administration’s Deferred Action for Parents of Americans and Lawful Permanent Residents (DAPA) program. As long as the Court is open to generous standing rules for “red” States to challenge Democratic presidents’ policies, it at least ought to be reluctant to make different rules for “blue” States when they challenge Republican presidents’ initiatives. That said, in one recent case involving a “red” State challenge, a majority of Justices signaled that perhaps they are open to cut back on the scope of State standing.  In United States v. Texas, 599 U.S. 670 (2023)—a case in which the Court held that Texas and Louisiana lacked Article III standing to sue Biden Administration officials for failure to enforce the immigration laws—the Court noted that “in our system of dual federal and state sovereignty, federal policies frequently generate indirect effects on state revenues or state spending,” and that when a State asserts that a federal law has produced only those kinds of indirect effects, “the State’s claim for standing can become more attenuated.” Id. at 680-81 n.3. It’s therefore possible the Court might begin to apply more searching standing scrutiny to red and blue State plaintiffs alike in future cases, even if it doesn’t reach the question in the birthright citizenship cases.

[2] Moreover, because the Trump Citizenship Order expressly directs federal officials not to accept States’ citizenship documents, the States’ suits arguably fall within an established exception to the Warth/Kowalski third-party doctrine. The Ninth Circuit accepted that argument last week (see pp. 20-21), as did an earlier First Circuit panel (see pp. 128a-131a). The Solicitor General, however, argued in response that because the Citizenship Order doesn’t require States to do or refrain from doing anything, nor subject States to potential sanctions, the exception in question is inapposite. I haven’t studied this question sufficiently to have a view on how the Supreme Court might resolve it—something I think the Court is unlikely to do in the context of the birthright citizenship cases.

[3] The individual plaintiffs sought and obtained a preliminary injunction in the District Court for the Western District of Washington, and the Government appealed that injunction in their favor only on the merits. The Government did not argue that the court of appeals should reverse the injunction, let alone that the individuals’ claims should be dismissed, merely because their children are now class members in a case in New Hampshire. And for good reason: Judge Laplante’s certification of the class in Barbara v. Trump (D.N.H.), discussed below, isn’t a valid basis for dismissing the individuals’ suits in Washington. Those individuals have already secured preliminary relief in the suit they filed before the Barbara class was certified. The fact that their children also benefit from a later-issued injunction on the other coast, and might ultimately prevail in that other case being litigated by other class representatives, is no reason to strip the plaintiff parents of their right to continue the suit in which they’ve already secured preliminary relief.

[4] Justice Barrett indicated (p.17) that such an argument should not be available for an individual plaintiff: “[P]rohibiting enforcement of the Executive Order against the child of an individual pregnant plaintiff will give that plaintiff complete relief: Her child will not be denied citizenship. Extending the injunction to cover all other similarly situated individuals would not render her relief any more complete.” CASA and ASAP, however, have argued that a universal injunction is necessary in order to ensure that their members obtain “complete relief.” It wouldn’t be feasible, they contend, to limit an injunction to cover only persons with a CASA or ASAP membership card because many parents would be deterred from proffering such a card due to a reasonable fear that it would signal that they might be unlawfully present and thereby subject them to immigration enforcement proceedings.

Justice Thomas, in his concurring opinion in CASA (p.5), read the majority opinion to have shut off this “complete relief” argument for CASA and ASAP: “[T]he Court today readily dispatches with the individual and associational respondents’ position that they require a universal injunction,” he writes, “notwithstanding their argument that a ‘plaintiff-specific injunction’ would be difficult to administer and would subject the associations’ members to the burden of having ‘to identify and disclose to the government’ their membership” (quoting Tr. of Oral Arg. 141–142). In her dissent (p.37), Justice Sotomayor characterized the Barrett opinion similarly: “Inexplicably, … the Court declares that, for the associational and individual respondents, injunctions enjoining the Government from enforcing the Citizenship Order against them (and only them) would have sufficed.”

Justice Barrett’s opinion for the Court, however, did not “declare” that CASA and ASAP’s “necessary for complete relief” theory was foreclosed, nor did Justice Barrett even refer to, let alone discuss the merits of, the associations’ argument that Justice Thomas identified. Therefore, as I read the Court’s opinion, that argument remains open on remand in the Maryland case, at least in theory. That said, the Thomas and Sotomayor opinions probably reflect internal discussions among the Justices (or earlier drafts of the Barrett opinion), and therefore they may suggest that a majority of the Court has significant skepticism of that argument. As I note in the text, however, this question might fall to the wayside in the Maryland case.

[5] As many observers have explained, there’s a good case to be made that Congress never intended to limit federal courts to the relatively static notion of equity reflected in Grupo Mexicano (and now CASA, too). After all, even the Court of Chancery itself implemented the system of equitable remedies in an accretive manner. As Owen Gallogly recently wrote:

[T]he remedial system administered by the Founding-Era Court of Chancery was not the dynamic, flexible, and discretionary form of justice that some modern commentators have advocated. But neither was it frozen in time; the Chancellor was not categorically limited to granting only those exact remedies that his forebears had issued. Reality lay somewhere between these two extremes. At the Founding, … the Chancery was governed by—and did not depart from—a core set of rules. But it could still develop, elaborate, and modestly update the law of equity by accretion of precedent—that is, by applying those core rules to new factual and legal contexts. Only avulsive changes to equity jurisprudence required legislative approval from Parliament.

Moreover, courts in the U.K. itself continued to fashion new equitable remedies throughout the nineteenth and twentieth centuries. Nevertheless, for purposes of this post I am accepting the Court’s view that Grupo Mexicano, Matthews and Atlas Life are binding and appropriate precedents.

[6] Justice Barrett cited two recent cases in which the Court (she suggests) focused only on a discrete disputed issue, rather than on which party was likely to win the suit, in applying the “likelihood of success” Nken factor. Those cases are not analogous to CASA, though. In OPM v. AFGE, No. 24A904 (Apr. 8, 2025), the Court granted a stay of a preliminary injunction because it found that the plaintiffs’ allegations were “insufficient to support [their] standing.” But if the AFGE plaintiffs lacked standing, their suit would have been dismissed, i.e., they wouldn’t have prevailed at the end of the litigation—whereas in CASA the plaintiffs are almost certain to prevail. In Gutierrez v. Saenz, No. 23-7809 (July 16, 2024), the Court granted a capital defendant’s application for a stay of execution while it was considering whether to grant his certiorari petition in a federal habeas suit seeking access to the State’s DNA evidence. The Court didn’t explain why it granted the stay, but presumably the Justices concluded (assuming they applied Nken) that Gutierrez was likely to succeed in his habeas petition (which he eventually did, the day before the Court issued its decision in CASA), in which the question was merely whether the Due Process Clause required the State to turn over the DNA evidence before putting Gutierrez to death, not whether the State would subsequently be legally entitled to execute him. In both cases Justice Barrett cited, in other words, the Court assessed whether the party applying for a stay was likely to prevail at the end of the litigation at hand.