United States v. Texas (2023)/Opinion of Justice Gorsuch
SUPREME COURT OF THE UNITED STATES
No. 22–58
UNITED STATES, ET AL., PETITIONERS v. TEXAS, ET AL.
ON WRIT OF CERTIORARI BEFORE JUDGMENT TO THE UNITED STATES COURT OF APPEALS FOR THE FIFTH CIRCUIT
[June 23, 2023]
Justice Gorsuch, with whom Justice Thomas and Justice Barrett join, concurring in the judgment.
The Court holds that Texas and Louisiana lack Article III standing to challenge the Department of Homeland Security’s Guidelines for the Enforcement of Civil Immigration Law. I agree. But respectfully, I diagnose the jurisdictional defect differently. The problem here is redressability.
I
Article III vests federal courts with the power to decide “Cases” and “Controversies.” Standing doctrine honors the limitations inherent in this assignment by ensuring judges attend to actual harms rather than abstract grievances. “If individuals and groups could invoke the authority of a federal court to forbid what they dislike for no more reason than they dislike it, we would risk exceeding the judiciary’s limited constitutional mandate and infringing on powers committed to other branches of government.” American Legion v. American Humanist Assn., 588 U. S. ___, ___ (2019) (Gorsuch, J., concurring in judgment) (slip op., at 3).
To establish standing to sue in federal court, a plaintiff must show that it has suffered a concrete and particularized injury, one that is both traceable to the defendant and redressable by a court order. See Lujan v. Defenders of Wildlife, 504 U. S. 555, 560–561 (1992). If a plaintiff fails at any step, the court cannot reach the merits of the dispute. See Steel Co. v. Citizens for Better Environment, 523 U. S. 83, 102–104 (1998). This is true whether the plaintiff is a private person or a State. After all, standing doctrine derives from Article III, and nothing in that provision suggests a State may have standing when a similarly situated private party does not. See Massachusetts v. EPA, 549 U. S. 497, 536–538 (2007) (Roberts, C. J., dissenting).
The Court holds that Texas and Louisiana lack standing to challenge the Guidelines because “a party lacks a judicially cognizable interest in the prosecution … of another.” Ante, at 5 (internal quotation marks omitted). To be sure, the district court found that the Guidelines have led to an increase in the number of aliens with criminal convictions and final orders of removal who are released into the States. 606 F. Supp. 3d 437, 459–463, 467 (SD Tex. 2022). The district court also found that, thanks to this development, the States have spent, and continue to spend, more money on law enforcement, incarceration, and social services. Id., at 463–465, 467. Still, the Court insists, “[s]everal good reasons explain why” these harms are insufficient to afford the States standing to challenge the Guidelines. Ante, at 6.
I confess to having questions about each of the reasons the Court offers. Start with its observation that the States have not pointed to any “historical practice” of courts ordering the Executive Branch to change its arrest or prosecution policies. Ante, at 5, 6. The Court is right, of course, that “history and tradition offer a meaningful guide to the types of cases that Article III empowers federal courts to consider.” TransUnion LLC v. Ramirez, 594 U. S. ___, ___ (2021) (slip op., at 8) (internal quotation marks omitted). But, again, the district court found that the Guidelines impose “significant costs” on the States. 606 F. Supp. 3d, at 495. The Court today does not set aside this finding as clearly erroneous. Nor does anyone dispute that even one dollar’s worth of harm is traditionally enough to “qualify as concrete injur[y] under Article III.” TransUnion, 594 U. S., at ___ (slip op., at 9); see also Uzuegbunam v. Preczewski, 592 U. S. ___, ___ (2021) (slip op., at 11). Indeed, this Court has allowed other States to challenge other Executive Branch policies that indirectly caused them monetary harms. See, e.g., Department of Commerce v. New York, 588 U. S. ___, ___–___ (2019) (slip op., at 9–10). So why are these States now forbidden from doing the same?
Next, the Court contends that, “when the Executive Branch elects not to arrest or prosecute, it does not exercise coercive power over an individual’s liberty or property.” Ante, at 6. Here again, in principle, I agree. But if an exercise of coercive power matters so much to the Article III standing inquiry, how to explain decisions like Massachusetts v. EPA? There the Court held that Massachusetts had standing to challenge the federal government’s decision not to regulate greenhouse gas emissions from new motor vehicles. See 549 U. S., at 516–526. And what could be less coercive than a decision not to regulate? In Massachusetts v. EPA, the Court chose to overlook this difficulty in part because it thought the State’s claim of standing deserved “special solicitude.” Id., at 520. I have doubts about that move. Before Massachusetts v. EPA, the notion that States enjoy relaxed standing rules “ha[d] no basis in our jurisprudence.” Id., at 536 (Roberts, C. J., dissenting). Nor has “special solicitude” played a meaningful role in this Court’s decisions in the years since. Even so, it’s hard not to wonder why the Court says nothing about “special solicitude” in this case. And it’s hard not to think, too, that lower courts should just leave that idea on the shelf in future ones.
Finally, the Court points to the fact that Article II vests in the President considerable enforcement discretion. Ante, at 6–8. So much so that “courts generally lack meaningful standards for assessing the propriety of [the Executive Branch’s] enforcement choices.” Ante, at 7. But almost as soon as the Court announces this general rule, it adds a caveat, stressing that “[t]his case concerns only arrest and prosecution policies.” Ante, at 12, n. 5. It’s a curious qualification. Article II does not have an Arrest and Prosecution Clause. It endows the President with the “executive Power,” §1, cl. 1, and charges him with “tak[ing] Care” that federal laws are “faithfully executed,” §3. These provisions give the President a measure of discretion over the enforcement of all federal laws, not just those that can lead to arrest and prosecution. So if the Court means what it says about Article II, can it mean what it says about the narrowness of its holding? There’s another curious qualification in the Court’s opinion too. “[T]he standing calculus might change,” we are told, “if the Executive Branch wholly abandoned its statutory responsibilities to make arrests or bring prosecutions.” Ante, at 11. But the Court declines to say more than that because “the States have not advanced” such an argument. Ibid. Is that true, though? The States have pleaded a claim under the Take Care Clause. App. 106. Is that not an abdication argument? Did they fail to plead it properly? Or is the Court simply ignoring it?
II
As I see it, the jurisdictional problem the States face in this case isn’t the lack of a “judicially cognizable” interest or injury. Ante, at 5 (internal quotation marks omitted). The States proved that the Guidelines increase the number of aliens with criminal convictions and final orders of removal released into the States. They also proved that, as a result, they spend more money on everything from law enforcement to healthcare. The problem the States face concerns something else altogether—a lack of redressability.
To establish redressability, a plaintiff must show from the outset of its suit that its injuries are capable of being remedied “ ‘by a favorable decision.’ ” Lujan, 504 U. S., at 561; see also id., at 570, n. 5 (plurality opinion). Ordinarily, to remedy harms like those the States demonstrated in this suit, they would seek an injunction. The injunction would direct federal officials to detain aliens consistent with what the States say the immigration laws demand. But even assuming an injunction like that would redress the States’ injuries, that form of relief is not available to them.
It is not available because of 8 U. S. C. §1252(f)(1). There, Congress provided that “no court (other than the Supreme Court) shall have jurisdiction or authority to enjoin or restrain the operation of” certain immigration laws, including the very laws the States seek to have enforced in this case. If there were any doubt about how to construe this command, we resolved it in Garland v. Aleman Gonzalez, 596 U. S. ___ (2022). In that case, we held that §1252(f)(1) “prohibits lower courts from … order[ing] federal officials to take or to refrain from taking actions to enforce, implement, or otherwise carry out the specified statutory provisions.” Id., at ___ (slip op., at 5). Put simply, the remedy that would ordinarily have the best chance of redressing the States’ harms is a forbidden one in this case.
The district court thought it could sidestep §1252(f)(1). Instead of issuing an injunction, it purported to “vacate” the Guidelines pursuant to §706(2) of the Administrative Procedure Act (APA), 5 U. S. C. §706(2). 606 F. Supp. 3d, at 498–501, and n. 71. Vacatur, as the district court understood it, is a distinct form of relief that operates directly on agency action, depriving it of legal force or effect. See id., at 499–500. And vacatur, the district court reasoned, does not offend §1252(f)(1), because it does not entail an order directing any federal official to do anything. See id., at 501, n. 71. The States embrace this line of argument before us. Brief for Respondents 43–47; Tr. of Oral Arg. 75–82.
It’s a clever workaround, but it doesn’t succeed. Start with perhaps the simplest reason. Assume for the moment the district court was right that §1252(f)(1) does not bar vacatur orders and that §706(2) authorizes courts to issue them. Even so, a vacatur order still does nothing to redress the States’ injuries. The Guidelines merely advise federal officials about how to exercise their prosecutorial discretion when it comes to deciding which aliens to prioritize for arrest and removal. A judicial decree rendering the Guidelines a nullity does nothing to change the fact that federal officials possess the same underlying prosecutorial discretion. Nor does such a decree require federal officials to change how they exercise that discretion in the Guidelines’ absence. It’s a point even the States have acknowledged. Tr. of Oral Arg. 82–83; see also id., at 75–77, 125.
Faced with that difficulty, the States offer this reply. As a practical matter, they say, we can expect federal officials to alter their arrest and prosecution priorities in light of a judicial opinion reasoning that the Guidelines are unlawful. See id., at 80, 82–83. But this doesn’t work either. Whatever a court may say in an opinion does no more to compel federal officials to change how they exercise their prosecutorial discretion than an order vacating the Guidelines. Nor do we measure redressability by asking whether a court’s legal reasoning may inspire or shame others into acting differently. We measure redressability by asking whether a court’s judgment will remedy the plaintiff’s harms. As this Court recently put it: “It is a federal court’s judgment, not its opinion, that remedies an injury; thus it is the judgment, not the opinion, that demonstrates redressability.” Haaland v. Brackeen, 599 U. S. ___, ___ (2023) (slip op., at 32). If the rule were otherwise, and courts could “simply assume that everyone … will honor the legal rationales that underlie their decrees, then redressability [would] always exist.” Franklin v. Massachusetts, 505 U. S. 788, 825 (1992) (Scalia, J., concurring in part and concurring in judgment).
Perhaps sensing they have run into yet another roadblock, the States try one last way around it. Fleetingly, they direct us to the parenthetical in §1252(f)(1): “(other than the Supreme Court).” That language, they say, allows this Court to invoke the All Writs Act, 28 U. S. C. §1651, to fashion its own injunction. And the possibility that this Court might award them relief, the States suggest, makes their injuries redressable after all. See Brief for Respondents 47; cf. post, at 12 (Alito, J., dissenting).
It’s an argument that yields more questions than answers. The parenthetical the States cite is a “curious” provision, one that “does not appear to have an analogue elsewhere in the United States Code.” Biden v. Texas, 597 U. S. ___, ___ (2022) (Barrett, J., dissenting) (slip op., at 4). Even assuming it permits this Court to award an injunction when a case comes to us on review, it does not obviously solve the States’ redressability problem. Normally, after all, a plaintiff must establish redressability from the outset of the suit. See Lujan, 504 U. S., at 561; see also id., at 570, n. 5 (plurality opinion). Not only that, a plaintiff must show a favorable decision is “ ‘likely’ ” to provide effectual relief. Id., at 561. When the States filed this suit, however, the possibility that it might find its way to this Court was speculative at best. See id., at 570, n. 5 (plurality opinion) (rejecting an argument that redressability could depend on “the fortuity that [a] case has made its way to this Court”).
Nor is that the only complication. Ordinarily, to win an injunction from any court, a party must satisfy several factors. See eBay Inc. v. MercExchange, L. L. C., 547 U. S. 388, 391 (2006). The States relegate any mention of these factors to a short, formulaic paragraph tacked onto the end of their brief. See Brief for Respondents 48. Worse, the only injunction they seek is one barring “implementation and enforcement” of the Guidelines—essentially an injunction imitating a vacatur order. Id., at 47. And as we have seen, an order like that would leave officials with their prosecutorial discretion intact. See supra, at 6. So, even if this Court were to take the unusual step of issuing and superintending its own injunction, giving the States the very order they seek is hardly sure to redress the injuries they assert.
III
Beyond these redressability problems may lie still another. Recall the essential premise on which the district court proceeded—that the APA empowers courts to vacate agency action. The federal government vigorously disputes this premise, arguing that the law does not contemplate this form of relief. The reasons the government offers are plenty and serious enough to warrant careful consideration.
A
Traditionally, when a federal court finds a remedy merited, it provides party-specific relief, directing the defendant to take or not take some action relative to the plaintiff. If the court’s remedial order affects nonparties, it does so only incidentally. See, e.g., Doran v. Salem Inn, Inc., 422 U. S. 922, 931 (1975) (“[N]either declaratory nor injunctive relief can directly interfere with the enforcement of contested statutes or ordinances except with respect to the particular federal plaintiffs.”); Alemite Mfg. Corp. v. Staff, 42 F. 2d 832 (CA2 1930) (L. Hand, J.) (“[A] court of equity … cannot lawfully enjoin the world at large.”); see also Trump v. Hawaii, 585 U. S. ___, ___ (2018) (Thomas, J., concurring) (slip op., at 6). This tracks the founding-era understanding that courts “render a judgment or decree upon the rights of the litigant[s].” Rhode Island v. Massachusetts, 12 Pet. 657, 718 (1838). It also ensures that federal courts respect the limits of their Article III authority to decide cases and controversies and avoid trenching on the power of the elected branches to shape legal rights and duties more broadly. After all, the “judicial Power” is the power to “decide cases for parties, not questions for everyone.” S. Bray, Multiple Chancellors: Reforming the National Injunction, 131 Harv. L. Rev. 417, 421 (2017).
Despite these foundational principles, in recent years a number of lower courts have asserted the authority to issue decrees that purport to define the rights and duties of sometimes millions of people who are not parties before them. Three years ago, I reflected on the rise of the “universal injunctio[n]” and raised questions about its consistency with the separation of powers and our precedents. Department of Homeland Security v. New York, 589 U. S. ___, ___ (2020) (opinion concurring in grant of stay) (slip op., at 3). I observed, too, that “the routine issuance of universal injunctions” has proven “unworkable, sowing chaos for litigants, the government, courts, and all those affected by these [sometimes] conflicting” decrees. Ibid.
Matters have not improved with time. Universal injunctions continue to intrude on powers reserved for the elected branches. They continue to deprive other lower courts of the chance to weigh in on important questions before this Court has to decide them. They continue to encourage parties to engage in forum shopping and circumvent rules governing class-wide relief. Recent events have highlighted another problem too. Sometimes, the government may effectively submit to a universal decree running against it in order to avoid “the usual and important requirement, under the [APA], that a regulation originally promulgated using notice and comment … may only be repealed through notice and comment.” Arizona v. City and County of San Francisco, 596 U. S. ___, ___ (2022) (Roberts, C. J., concurring) (slip op., at 2). It is a strategy that amounts to little more than “ ‘rulemaking-by-collective-acquiescence.’ ” Ibid.; see also Danco Laboratories, LLC v. Alliance for Hippocratic Medicine, 598 U. S. ___, ___ (2023) (Alito, J., dissenting from grant of application for stays) (slip op., at 3); Arizona v. Mayorkas, 598 U. S. ___, ___–___ (2023) (statement of Gorsuch, J.) (slip op., at 1–4).
Today’s case presents a variation on the theme. The district court ordered “wholesale vacatur” of the Guidelines, rendering them inoperable with respect to any person anywhere. 606 F. Supp. 3d, at 499, 502. As authority for its course, the district court cited §706(2) of the APA. That provision does not say anything about “vacating” agency action (“wholesale” or otherwise). Instead, it authorizes a reviewing court to “set aside” agency action. Still, from those two words alone, the district court thought the power to nullify the Guidelines with respect to anyone anywhere surely follows. See 606 F. Supp. 3d, at 498–500.
Color me skeptical. If the Congress that unanimously passed the APA in 1946 meant to overthrow the “bedrock practice of case-by-case judgments with respect to the parties in each case” and vest courts with a “new and far-reaching” remedial power, it surely chose an obscure way to do it. Arizona v. Biden, 40 F. 4th 375, 396 (CA6 2022) (Sutton, C. J., concurring). At the very least, it is worth a closer look.
B
Begin with the words “set aside” in isolation. If they might suggest to some a power to “vacate” agency action in the sense of rendering it null and void, just as naturally they might mean something else altogether. They might simply describe what a court usually does when it finds a federal or state statute unconstitutional, or a state law preempted by a federal one. Routinely, a court will disregard offensive provisions like these and proceed to decide the parties’ dispute without respect to them. In Dennis v. United States, 341 U. S. 494 (1951), for example, Justice Frankfurter observed that “[w]e are to set aside the judgment of those whose duty it is to legislate only if” the Constitution requires it. Id., at 525 (concurring opinion). Justice Frankfurter hardly meant to suggest the Court had the power to erase statutes from the books. See id., at 525–526. Instead, he used the phrase to mean that a court should disregard—refuse to apply—an unconstitutional law. It is a usage that was common at the time of the APA’s adoption and that remains so today. See Webster’s New International Dictionary 2291 (2d ed. 1954) (defining “set aside” as “to put to one side; discard; dismiss” and “to reject from consideration; overrule”); Webster’s New World College Dictionary 1329 (5th ed. 2016) (defining “set aside” as “to set apart” and “to discard; dismiss; reject”).
There are many reasons to think §706(2) uses “set aside” to mean “disregard” rather than “vacate.” For one thing, at the time of the APA’s adoption, conventional wisdom regarded agency rules as “quasi-legislative” in nature. See Humphrey’s Executor v. United States, 295 U. S. 602, 624, 628 (1935); see also D. Currie & F. Goodman, Judicial Review of Federal Administrative Action: Quest for the Optimum Forum, 75 Colum. L. Rev. 1, 40 (1975). And federal courts have never enjoyed the power to “vacate” legislation. Instead, they possess “little more than the negative power to disregard an unconstitutional enactment.” Massachusetts v. Mellon, 262 U. S. 447, 488 (1923). Reading “set aside” to mean “disregard” ensures parallel judicial treatment of statutes and rules.
For another thing, the term “set aside” appears in §706 of the APA. That section is titled “Scope of review,” a title it has borne since the law’s enactment in 1946. 60 Stat. 243. And ordinarily, when we think about the scope of a court’s review, we do not think about the remedies the court may authorize after reaching its judgment on the merits. Instead, we think about the court’s decisional process leading up to that judgment. Understanding “set aside” as a command to disregard an unlawful rule in the decisional process fits perfectly within this design. Understanding the phrase as authorizing a remedy does not.
What follows in §706 appears to confirm the point. The statute begins by providing that, “[t]o the extent necessary to decision and when presented, the reviewing court shall decide all relevant questions of law, interpret constitutional and statutory provisions, and determine the meaning and applicability of the terms of an agency action.” Exactly as expected, we find an instruction about the decisional process—one requiring the court to apply “de novo review on questions of law” as it considers the parties’ arguments in the course of reaching its judgment. Kisor v. Wilkie, 588 U. S. ___, ___ (2019) (Gorsuch, J., concurring in judgment) (slip op., at 15) (internal quotation marks omitted). Nothing here speaks to remedies.
The remaining statutory language is more of the same. Section 706 goes on to instruct that “[t]he reviewing court shall … hold unlawful and set aside agency action, findings, and conclusions found to be,” among other things, “arbitrary,” “capricious,” “contrary to constitutional right,” “in excess of” statutory authority, or “unsupported by substantial evidence.” §706(2). Looking at the provision as a whole, rather than focusing on two words in isolation, we see further evidence that it governs a court’s scope of review or decisional process. The statute tells judges to resolve the cases that come to them without regard to deficient agency action, findings, or conclusions—an instruction entirely consistent with the usual “negative power” of courts “to disregard” that which is unlawful. Mellon, 262 U. S., at 488.
Other details are telling too. Consider the latter part of §706(2)’s directive to “set aside agency action, findings, and conclusions.” The APA defines “agency action” to include “the whole or a part of an agency rule, order, license, sanction, relief, or the equivalent or denial thereof, or failure to act.” 5 U. S. C. §551(13). A court can disregard any of those things. But what would it even mean to say a court must render null and void an agency’s failure to act? Notice, too, the language about “findings.” Often, judges disregard factual findings unsupported by record evidence and resolve the case at hand without respect to them. See Fed. Rule Civ. Proc. 52(a)(6) (“Findings of fact … must not be set aside unless clearly erroneous.”). None of that means we may pretend to rewrite history and scrub any trace of faulty findings from the record.
Consider as well the larger statutory context. Section 702 restricts judicial review to “person[s]” who have “suffer[ed] legal wrong because of agency action, or [been] adversely affected or aggrieved by agency action.” The provision also instructs that “any mandatory or injunctive decree shall specify the Federal officer or officers … personally responsible for compliance.” Here, it seems, Congress nodded to traditional standing rules and remedial principles. Yet under the district court’s reading, we must suppose Congress proceeded just a few paragraphs later to plow right through those rules and empower a single judge to award a novel form of relief affecting parties and nonparties alike.
Then there is §703. That is where the APA most clearly discusses remedies. Section 703 authorizes aggrieved persons to bring “any applicable form of legal action, including actions for declaratory judgments or writs of prohibitory or mandatory injunction or habeas corpus.” Conspicuously missing from the list is vacatur. And what exactly would a “form of legal action” seeking vacatur look like anyway? Would it be a creature called a “writ of vacatur”? Nobody knows (or bothers to tell us). Nor is it apparent why Congress would have listed most remedies in §703 only to bury another (and arguably the most powerful one) in a later section addressed to the scope of review. Cf. J. Harrison, Section 706 of the Administrative Procedure Act Does Not Call for Universal Injunctions or Other Universal Remedies, 37 Yale J. Reg. Bull. 37, 37, 45–46 (2020).
The district court’s reading of “set aside” invites still other anomalies. Section 706(2) governs all proceedings under the APA. Any interpretation of “set aside” therefore must make sense in the context of an enforcement proceeding, an action for a declaratory judgment, a suit for an injunction, or habeas. See §703. This poses a problem for the district court’s interpretation, for no one thinks a court adjudicating a declaratory action or a habeas petition “vacates” agency action along the way. See Brief for United States 41–42; Harrison, 37 Yale J. Reg. Bull., at 46. The anomaly dissipates, however, if we read §706(2) as instructing courts about when they must disregard agency action in the process of deciding a case.
Imagine what else it would mean if §706(2) really did authorize vacatur. Ordinary joinder and class-action procedures would become essentially irrelevant in administrative litigation. Why bother jumping through those hoops when a single plaintiff can secure a remedy that rules the world? See Bray, 131 Harv. L. Rev., at 464–465. Surely, too, it is odd that leading scholars who wrote extensively about the APA after its adoption apparently never noticed this supposed remedy. See J. Harrison, Vacatur of Rules Under the Administrative Procedure Act, 40 Yale J. Reg. Bull. 119, 127–128 (2023) (discussing scholarship of Professors Kenneth Culp Davis and Louis Jaffe); see also Department of Justice, Attorney General’s Manual on the Administrative Procedure Act 108 (1947) (offering the Executive Branch’s view that §706 simply “restates the present law as to the scope of judicial review”). These are not people who would have missed such a major development in their field.
C
As always, there are arguments on the other side of the ledger, and the States tee up several. They first reply that §706(2) must allow vacatur of agency action because the APA models judicial review of agency action on appellate review of judgments, and appellate courts sometimes vacate judgments. Brief for Respondents 40. But just because “Congress may sometimes refer to collateral judicial review of executive action as ‘an appeal’ … does not make it an ‘appeal’ akin to that taken from the district court to the court of appeals.” Garland v. Ming Dai, 593 U. S. ___, ___ (2021) (slip op., at 9). Nor does any of that tell us in which respects the APA models judicial review of agency action on appellate review of lower court judgments. According to one scholar, the “salient” similarities between appellate review and judicial review of agency action concern the standards of review—in both types of proceedings, a reviewing court engages in a more rigorous review of legal questions and a more deferential review of factual findings. T. Merrill, Article III, Agency Adjudication, and the Origins of the Appellate Review Model of Administrative Law, 111 Colum. L. Rev. 939, 940–941 (2011). None of that has to do with remedies; once again, it concerns a court’s scope of review or decisional process.
The States next invoke §706(1) and §705. The former provides that courts shall “compel agency action unlawfully withheld or unreasonably delayed.” The latter says courts “may issue all necessary and appropriate process to postpone the effective date of an agency action or to preserve status or rights pending conclusion of the review proceedings.” The States insist that “[i]t would be illogical” for the APA to authorize these remedies but not vacatur. Brief for Respondents 40. Is it so clear, though, that §706(1) and §705 authorize remedies? Section 706(1) does seem to contemplate a remedy. But it’s one §703 mentions—mandatory injunctions. So §706(1) might not authorize a remedy as much as confirm the availability of a traditional remedy to address agency inaction. The same could be said about §705; it might just confirm courts’ authority to issue traditional equitable relief pending judicial review. Cf. Sampson v. Murray, 415 U. S. 61, 69, n. 15 (1974) (explaining that §705 was “primarily intended to reflect existing law”).
The States also direct us to scholarship that in turn purports to identify a few instances of federal courts “setting aside” agency action in the years leading up to the APA. See Brief for Respondents 41; see also Brief for State of Florida as Amicus Curiae 17. It is not obvious, however, that these few cases stand for so much. In two of them, this Court upheld the agency action in question and thus had no occasion to opine on appropriate relief. See Houston v. St. Louis Independent Packing Co., 249 U. S. 479, 486–487 (1919); The Assigned Car Cases, 274 U. S. 564, 584 (1927). In a third case, the plaintiff sought “to enjoin enforcement of” an order of the Federal Communications Commission. Columbia Broadcasting System, Inc. v. United States, 316 U. S. 407, 408 (1942). That is a claim for traditional equitable relief, and indeed, the Court held that the complaint “state[d] a cause of action in equity” and remanded for further proceedings. Id., at 425. A fourth case, involving an order of the Interstate Commerce Commission, seems of a piece. There, a district court held the Commission’s order invalid and “restrain[ed] … enforcement” of it. Baltimore & Ohio R. Co. v. United States, 5 F. Supp. 929, 936 (ND Ohio 1933). This Court affirmed. See United States v. Baltimore & Ohio R. Co., 293 U. S. 454 (1935). True, this Court described the case as an “appeal from [a] decree … setting aside” the Commission’s order. Id., at 455. But the fact that the lower court had only restrained enforcement of the order goes to show that “set aside” did not then (and does not now) necessarily translate to “vacate.”
At the end of the day, the States fall back on other lower court decisions. “For more than 30 years,” they say, “vacatur has been the ordinary result when the D. C. Circuit determines that agency regulations are unlawful.” Brief for Respondents 42 (internal quotation marks omitted). Doubtless, to the extent those decisions are carefully reasoned, they merit respectful consideration. But, equally, they do not bind us. Cf. post, at 14, n. 7 (Alito, J., dissenting) (observing that this Court has only ever “assumed” that the APA authorizes vacatur).
In raising questions about the district court’s claim that §706(2) authorizes vacatur of agency action, I do not pretend that the matter is open and shut. Thoughtful arguments and scholarship exist on both sides of the debate. Nor do I mean to equate vacatur of agency action with universal injunctions. Despite some similarities, courts can at least arguably trace their authority to order vacatur to language in a statute and practice in some lower courts. But the questions here are serious ones. And given the volume of litigation under the APA, this Court will have to address them sooner or later. Until then, we would greatly benefit from the considered views of our lower court colleagues.
D
Suppose my doubts about vacatur are unfounded. Suppose the APA really does authorize both traditional forms of equitable relief (in §703) and a more expansive equitable power to vacate agency action (in §706). Even if that were true, a district court should “think twice—and perhaps twice again—before granting” such sweeping relief. Arizona v. Biden, 40 F. 4th, at 396 (Sutton, C. J., concurring).
After all, this Court has long instructed that equitable relief “must be limited to the inadequacy that produced [the] injury in fact.” Gill v. Whitford, 585 U. S. ___, ___ (2018) (slip op., at 14) (internal quotation marks omitted). Any remedy a judge authorizes must not be “more burdensome [to the defendant] than necessary to redress the complaining parties.” Califano v. Yamasaki, 442 U. S. 682, 702 (1979). And faithful application of those principles suggests that an extraordinary remedy like vacatur would demand truly extraordinary circumstances to justify it. Cf. S. Bray & P. Miller, Getting Into Equity, 97 N. D. L. Rev. 1763, 1797 (2022) (“[I]n equity it all connects—the broader and deeper the remedy the plaintiff wants, the stronger the plaintiff’s story needs to be.”).
The temptations a single district judge may face when invited to vacate agency rules are obvious. Often, plaintiffs argue that everyone deserves to benefit from their effort to litigate the case and the court’s effort to decide it. Judges may think efficiency and uniformity favor the broadest possible relief. But there are serious countervailing considerations. As with universal injunctions, vacatur can stymie the orderly review of important questions, lead to forum shopping, render meaningless rules about joinder and class actions, and facilitate efforts to evade the APA’s normal rulemaking processes. Vacatur can also sweep up nonparties who may not wish to receive the benefit of the court’s decision. Exactly that happened here. Dozens of States, counties, and cities tell us they did not seek and do not want the “benefit” of the district court’s vacatur order in this case. See Brief for New York et al. as Amici Curiae 1–2; Brief for 21 Cities, Counties, and Local Government Organizations as Amici Curiae 2–3.
More importantly still, universal relief, whether by way of injunction or vacatur, strains our separation of powers. It exaggerates the role of the Judiciary in our constitutional order, allowing individual judges to act more like a legislature by decreeing the rights and duties of people nationwide. This Court has warned that “[f]ew exercises of the judicial power are more likely to undermine public confidence in the neutrality and integrity of the Judiciary than one which casts [courts] in the role of a Council of Revision, conferring on [themselves] the power to invalidate laws at the behest of anyone who disagrees with them.” Arizona Christian School Tuition Organization v. Winn, 563 U. S. 125, 145–146 (2011). At a minimum, then, district courts must carefully consider all these things before doling out universal relief. And courts of appeals must do their part, too, asking whether party-specific relief can adequately protect the plaintiff’s interests. If so, an appellate court should not hesitate to hold that broader relief is an abuse of discretion. Cf. Kentucky v. Biden, 57 F. 4th 545, 556–557 (CA6 2023) (Larsen, J.). * In our system of government, federal courts play an important but limited role by resolving cases and controversies. Standing doctrine honors this limitation at the front end of every lawsuit. It preserves a forum for plaintiffs seeking relief for concrete and personal harms while filtering out those with generalized grievances that belong to a legislature to address. Traditional remedial rules do similar work at the back end of a case. They ensure successful plaintiffs obtain meaningful relief. But they also restrain courts from altering rights and obligations more broadly in ways that would interfere with the power reserved to the people’s elected representatives. In this case, standing and remedies intersect. The States lack standing because federal courts do not have authority to redress their injuries. Section 1252(f)(1) denies the States any coercive relief. A vacatur order under §706(2) supplies them no effectual relief. And such an order itself may not even be legally permissible. The States urge us to look past these problems, but I do not see how we might. The Constitution affords federal courts considerable power, but it does not establish “government by lawsuit.” R. Jackson, The Struggle for Judicial Supremacy 286–287 (1941).