On January 27, 2017, President Donald Trump issued an executive order barring millions of people from entering the country. The order went into effect immediately.1×1. Exec. Order No. 13,769, 82 Fed. Reg. 8977 (Jan. 27, 2017). It stranded hundreds of people in transit and led to the cancellation of 60,000 valid visas.2×2. Jaweed Kaleem, Nearly 60,000 Visas Revoked Since Trump’s Immigration Order, L.A. Times (Feb. 3, 2017), http://www.latimes.com/politics/washington/la-na-essential-washington-updates-more-than-100-000-visas-revoked-since-1486148132-htmlstory.html. People were being put back on planes and about to be deported, and only a small fraction could make it into court in time. But within days, two judges issued nationwide injunctions blocking the order.3×3. Washington v. Trump, No. 17-141, 2017 WL 462040 (W.D. Wash. Feb. 3, 2017); Darweesh v. Trump, No. 17-480, 2017 WL 388504 (E.D.N.Y. Jan. 28, 2017). Without them, most affected people would have been deported, lost their visas, and been kept from work, family, and study, all pursuant to a policy that the government ultimately chose not to defend.4×4. See Exec. Order 13,780, 82 Fed. Reg. 13,209 (Mar. 6, 2017) (revoking and replacing the original ban). When a policy threatens widespread irreparable harm, most people will never be fully protected without broad preliminary relief.
Professor Samuel L. Bray takes up the propriety of nationwide injunctions in Multiple Chancellors: Reforming the National Injunction.5×5. Samuel L. Bray, Multiple Chancellors: Reforming the National Injunction, 131 Harv. L. Rev. 417 (2017). His Article is a major contribution to an issue that deserves more reflection than it has received.6×6. We have litigated on both sides of nationwide injunction cases. Our office has sought and obtained several nationwide injunctions since January 2017. And one of us served in the Civil Division at the Department of Justice in 2016, when a number of nationwide injunctions were issued against federal agencies. He traces injunctive relief from its origins in the English Chancery, which had a single chancellor, to its use in the federal courts, which are of course divided into circuits and districts. He explains how courts in the second half of the twentieth century began issuing nationwide injunctions with increasing frequency.
In Bray’s account, that trend has bred a host of problems. He focuses on four in particular: First, nationwide injunctions encourage plaintiffs to shop for a favorable forum in order to obtain an injunction that applies everywhere.7×7. Bray, supra note 5, at 457–61. Second, the nationwide injunction prevents legal questions from percolating up through multiple circuits.8×8. Id. at 461–62. Third, nationwide remedies make conflicting injunctions more likely.9×9. Id. at 462–64. And fourth, they seem in tension with other doctrines in the law.10×10. Id. at 464–65. Bray’s proposed solution is simple: courts should grant only the relief necessary to protect the plaintiffs from the defendant, and should not constrain the defendant’s conduct toward others.11×11. Id. at 469. As he puts it, in all cases, “injunctions should not protect nonparties.”12×12. Id.
Many of Bray’s criticisms of nationwide injunctions are well founded, if debatable, as we discuss below. But we found two parts of his analysis incomplete. First, in weighing the costs and benefits of nationwide injunctions, Bray gives short shrift to their role in preventing widespread harm, even though that is probably their most important function. Indeed, the Supreme Court has recently suggested that sometimes “the equitable balance” in a case will favor extending injunctive relief to “parties similarly situated to” the plaintiffs.13×13. Trump v. Int’l Refugee Assistance Project, 137 S. Ct. 2080, 2087 (2017) (per curiam). It remains for others to explore where that value outweighs the downsides Bray highlights. Second, Bray’s plaintiffs-only proposal extends far beyond the problems it is meant to solve. Bray does not grapple with that mismatch, nor does he account for the amount of existing judicial practice that his proposal would upend. Future work should consider less radical alternatives, which would allow broad injunctions when necessary to prevent real-world injuries, but would otherwise preserve opportunities for percolation across multiple chancellors.
The Importance of Irreparable Harm
Bray’s call to end nationwide injunctions implies that their vices always outweigh their virtues. But while he documents their detriments in some detail, he largely passes over one of their core purposes: preventing irreparable harm. When a court issues interim relief — a stay, preliminary injunction, or temporary restraining order — the same equities that require protection for the plaintiff often support protection for those who are similarly situated.
Some government policies, like President Trump’s travel ban, threaten immediate and lasting damage. They go into effect quickly, and their impact cannot be reversed at the end of a lawsuit. Anyone who does not or cannot bring her own case can only be protected if a court concludes the policy is illegal and fully enjoins it. Preventing widespread and illegal injuries is a good thing, especially when the government and others would not be much harmed in the process. This is the kind of balancing courts do all the time when asked to issue interim equitable remedies.14×14. Winter v. Nat. Res. Def. Council, 555 U.S. 7, 24 (2008) (“In each case, courts ‘must balance the competing claims of injury and must consider the effect on each party of the granting or withholding of the requested relief.’” (quoting Amoco Prod. Co. v. Village of Gambell, 480 U.S. 531, 542 (1987))).
Other times, there is no pressing need for a broad injunction. When harm is remote or reversible, there is ample time for issues to percolate up through multiple cases in multiple circuits. The challenges to the Affordable Care Act are a good example.15×15. See, e.g., Seven-Sky v. Holder, 661 F.3d 1 (D.C. Cir. 2011); Liberty Univ., Inc. v. Geithner, 671 F.3d 391 (4th Cir. 2011); Florida ex rel. Att’y Gen. v. U.S. Dep’t of Health & Human Servs., 648 F.3d 1235 (11th Cir. 2011); Thomas More Law Ctr. v. Obama, 651 F.3d 529 (6th Cir. 2011). The law passed in 2010,16×16. Patient Protection and Affordable Care Act, Pub. L. No. 111-148, 124 Stat. 119 (2010) (codified as amended in scattered sections of the U.S. Code). but its individual mandate did not take effect until 2014.17×17. 26 U.S.C. § 5000A (2012). A broad injunction would have served little purpose. Instead, the courts that ruled against the government stayed their decisions, and eventually the cases reached the Supreme Court in 2012, two years before anyone would feel the policy’s impact. Bray describes the lower courts’ forbearance as “acts of judicial self-restraint, not judicial necessity.”18×18. Bray, supra note 5, at 461. But in reality, no one faced any irreparable harm, so there would have been little basis for interim relief, even as to the plaintiffs.
Of course it is often a judgment call whether the plaintiffs — or anyone else — actually face harm that cannot be reversed later. Not all cases involve businesses shuttering or people being deported in real time. But this is a question that judges face every time a litigant asks for interim relief, regardless of its scope. Broader remedies certainly might demand a more careful assessment of the harm. But when many people face the same genuinely irreparable injury as the plaintiff, a complete injunction of the illegal policy serves an important purpose.
We do not mean to suggest that a nationwide injunction should always issue once a judge finds a policy illegal. Their propriety will vary depending on a number of factors, the balancing of which is beyond the scope of this Response.19×19. Equitable relief has long required courts “to ‘balance the equities’ — to explore the relative harms to applicant and respondent, as well as the interests of the public at large.” Barnes v. E-Systems, Inc. Grp. Hosp. Med. & Surgical Ins. Plan, 501 U.S. 1301, 1305 (1991) (Scalia, J., opinion in chambers) (quoting Rostker v. Goldberg, 448 U.S. 1306, 1308 (1980) (Brennan, J., opinion in chambers)). Our point is simply that what Bray calls the “epistemic advantages”20×20. Bray, supra note 5, at 424. of narrow relief — slower deliberation, more forums and judges weighing in — do not always outweigh the need to prevent real-world harm. Courts do not exist simply to refine legal principles.
Nor are Bray’s concerns self-evidently dispositive. For example, as he acknowledges, the risk of conflicting injunctions is vanishingly low.21×21. See id. at 462–63. Such an unlikely scenario does not justify depriving injured people of protection. And it could be avoided with a simple strengthening of the existing comity doctrine, which typically prevents conflicting injunctions already.22×22. Under the principle of comity, “federal district courts — courts of coordinate jurisdiction and equal rank — . . . exercise care to avoid interference with each other’s affairs.” W. Gulf Mar. Ass’n v. ILA Deep Sea Local 24, 751 F.2d 721, 728 (5th Cir. 1985) (citing Kerotest Mfg. Co. v. C-O-Two Fire Equip. Co., 342 U.S. 180 (1952)). Such care normally counsels against issuing a conflicting injunction.
Or take percolation. There is a widely held belief that it is useful, which we share.23×23. Judge Leventhal hailed the “value in percolation among the circuits, with room for a healthful difference that may balance the final result.” Harold Leventhal, Eleventh Annual Mooers Lecture, A Modest Proposal for a Multi-Circuit Court of Appeals, 24 Am. U. L. Rev. 881, 907 (1975). And the Supreme Court has mentioned “the benefit it receives from permitting several courts of appeals to explore a difficult question” before granting certiorari. United States v. Mendoza, 464 U.S. 154, 160 (1984) (holding that even after losing in one circuit, the government may adhere to its position in other circuits). But intuitions aside, what is the evidence that percolation among the circuits yields better-reasoned decisions? The question is ultimately an empirical one, and we are not aware of persuasive evidence on either side.24×24. See Patricia M. Wald, Upstairs/Downstairs at the Supreme Court: Implications of the 1991 Term for the Constitutional Work of the Lower Courts, 61 U. Cin. L. Rev. 771, 793 (1993) (“It is a pity we do not have a better empirical fix on how important a role percolation plays in Supreme Court constitutional decisionmaking . . . .”). Even Chief Justice Rehnquist — the author of United States v. Mendoza,25×25. 464 U.S. 154. the Supreme Court’s most explicit pro-percolation case — subsequently criticized the notion of percolation for percolation’s sake.26×26. See William H. Rehnquist, The Changing Role of the Supreme Court, 14 Fla. St. U. L. Rev. 1, 11 (1986) (“If we were talking about laboratory cultures or seedlings, the concept of issues ‘percolating’ in the courts of appeals for many years before they are really ready to be decided by the Supreme Court might make some sense. But it makes very little sense in the legal world in which we live. . . . What we need is not the ‘correct’ answer in the philosophical or mathematical sense, but the ‘definitive’ answer, and the ‘definitive’ answer can be given under our system only by the court of last resort.”). And in practice, nationwide injunctions do not always foreclose percolation.27×27. For instance, both the Seventh and Third Circuits will likely review the Department of Justice’s immigration-related spending conditions, despite a nationwide injunction in the former. See City of Chicago v. Sessions, No. 17 C 5720, 2017 WL 4081821 (N.D. Ill. Sept. 15, 2017), appeal docketed, No. 17-2991 (7th Cir. Sept. 26, 2017); City of Philadelphia v. Sessions, No. 17-3894, 2017 WL 5489476 (E.D. Pa. Nov. 15, 2017). Both the Fourth and Ninth Circuits have reviewed the travel bans, despite nationwide injunctions in both. See Int’l Refugee Assistance Project v. Trump, 857 F.3d 554 (4th Cir.) (en banc), vacated as moot, 86 U.S.L.W. 3175 (U.S. Oct. 10, 2017); Hawaii v. Trump, 859 F.3d 741 (9th Cir.), vacated as moot and appeal dismissed, 874 F.3d 1112 (9th Cir. 2017) (mem.).
Our point is not that multi-circuit review is unimportant, or that conflicting injunctions should be allowed. It is that those concerns do not automatically outweigh the importance of preventing concrete injuries when a nationwide policy threatens many people with the same harm.
An Overbroad Response to an Overbroad Remedy
Bray would not merely outlaw nationwide injunctions. He would go much further, and have courts tailor relief such that it never blocks the challenged policy as to anyone beyond the plaintiffs. This far outstrips the three main problems Bray identifies — forum shopping, hurried decisionmaking, and conflicting injunctions. Elsewhere, he suggests that our system already requires plaintiffs-only relief, because Article III does not allow “remedies for those who are not parties,”28×28. Bray, supra note 5, at 471. and because nationwide injunctions did not exist in traditional equity.29×29. Id. at 472–73. We are not historians, but Bray’s historical account seemed to raise a few questions. For instance, given that traditional equity allowed “no injunctions against the Crown,” id. at 425 & n.32 (emphasis added), it is hard to see why the absence of nationwide injunctions against the Crown tells us anything about the allowable scope of modern remedies. What’s more, Bray explains that traditional equity sometimes extended relief to “nonplaintiffs” whose claims were “identical” to the plaintiffs’, id. at 426, just as broad injunctions do today. It thus appears that traditional equity’s relevance to nationwide injunctions might not be as tidy as Bray suggests. See id. at 425; see also id. at 423, 473 (acknowledging that some “translation” needs to be made either way).
Whatever his principle’s possible justifications, it stretches well beyond any concern with nationwide injunctions. Imagine, for example, a challenge to a policy confined to one district or circuit — at a single prison, agency field office, or national park. Bray’s rule would allow an injunction only as to the plaintiffs. But because the policy is confined to a single district or circuit, a broader injunction that enjoined the policy in full would do little to encourage forum shopping, hamper percolation, or threaten conflicting injunctions. Bray does not explain this mismatch.
Nor does he grapple with how much existing practice his proposal would disrupt. Courts regularly “set aside” (or remand) agency policies wholesale under the Administrative Procedure Act;30×30. 5 U.S.C. § 706(2); see also Humane Soc’y of United States v. Zinke, 865 F.3d 585, 614 (D.C. Cir. 2017) (describing remand without vacatur). Bray questions this practice’s pedigree, see Bray, supra note 5, at 438 n.121, but not its regularity. “the ordinary result is that the rules are vacated — not that their application to the individual petitioners is proscribed.”31×31. Nat’l Mining Ass’n v. U.S. Army Corps of Eng’rs, 145 F.3d 1399, 1409 (D.C. Cir. 1998) (quoting Harmon v. Thornburgh, 878 F.2d 484, 495 n.21 (D.C. Cir. 1989)). And courts generally enjoin state and local policies in their entirety, even in a case brought by a limited number of plaintiffs.32×32. See, e.g., N.C. State Conference of the NAACP v. McCrory, 831 F.3d 204 (4th Cir. 2016) (enjoining state voting laws); Villas at Parkside Partners v. City of Farmers Branch, 726 F.3d 524 (5th Cir. 2013) (en banc) (enjoining local immigration ordinance). Bray acknowledges this practice. See Bray, supra note 5, at 444 n.161, 454 n.220. None of this would be possible if injunctions could never “protect nonparties.”33×33. Bray, supra note 5, at 469.
We would have a very different system without these remedies. No one would be protected from an illegal policy without bringing their own challenge. The number of lawsuits over some policies might have to increase dramatically. And while the eventual development of appellate precedent might ultimately provide broader protection, the government would have far less incentive to appeal, because appellate precedent is the only thing that could shut down an illegal policy in full.
Bray claims that his proposal is only for federal defendants, and therefore avoids grappling with examples of broad injunctions against state and local policies.34×34. Id. at 424. Bray does say in passing that perhaps his rule could be applied to state laws as well. See id. at 424 n.29. But statewide injunctions are a problem for many of his rationales. They belie the notion that Article III somehow already requires plaintiffs-only injunctions.35×35. Contra id. at 421, 471–72. Article III has never required courts to meticulously ensure that no relief reaches anyone beyond the plaintiff. See Lujan v. Nat’l Wildlife Fed’n, 497 U.S. 871, 913 (1990) (Blackmun, J., dissenting) (expressing the views of all the Justices on the relevant point); see also id. at 890 n.2 (majority opinion). 7A Charles Alan Wright et al., Federal Practice and Procedure § 1771 (3d ed. 2005) (“[T]he requested relief generally will benefit not only the claimant but all other persons subject to the practice or rule under attack.”). As long as a plaintiff has standing to challenge a policy, Article III is no barrier to enjoining it in full. They raise all but one of the “doctrinal inconsistencies” Bray attributes to nationwide injunctions.36×36. Bray, supra note 5, at 464. The exception is the rule that the United States is not subject to nonmutual collateral estoppel — it may adhere to its position in other circuits even after losing in one circuit. United States v. Mendoza, 464 U.S. 154 (1984). It is worth noting, however, that this rule merely means that percolation can happen after the federal government loses. It does not suggest that percolation is the paramount concern in all cases. And they call into question whether Bray’s three main criticisms of nationwide injunctions are really so damning. A state law might usefully percolate to the circuit through different district courts, and to the Supreme Court through both state and federal courts. A federal and a state judge could issue conflicting injunctions. Plaintiffs challenging state policies can easily shop between different districts, different divisions, and between federal and state courts. And so on. It is thus unclear why there should be such a sharp difference in how equity treats federal and state actors.
Bray’s Article remains a useful guide to the history and pitfalls of nationwide injunctions. Future work, however, should explore less drastic solutions than outlawing all nonparty relief, and should examine the values that broad injunctions might sometimes serve, including the importance of stepping in when the government threatens irreparable harm. That is what the courts did in the immediate aftermath of the travel ban, and the plaintiffs understood the significance of the relief. “In any other country, when the president wants something, he gets it,” one of the plaintiffs told the New York Times.37×37. Vivian Yee, Meet the Everyday People Who Have Sued Trump. So Far, They’ve Won., N.Y. Times (Mar. 29, 2017), https://www.nytimes.com/2017/03/29/us/trump-travel-ban.html [https://perma.cc/X7AL-5RXW]. “The fact that a lowly judge somewhere can basically stop the most powerful man on earth with a simple ruling is gratifying, and it shows what this country’s all about.”38×38. Id.
* We are attorneys at the American Civil Liberties Union Immigrants’ Rights Project. We write in our personal capacities.
Recommended Citation: Spencer E. Amdur & David Hausman, Response, Nationwide Injunctions and Nationwide Harm, 131 Harv. L. Rev. F. 49 (2017), https://harvardlawreview.org/2017/12/nationwide-injunctions-nationwide-harm/.