A Response to The Lost History of the “Universal” Injunction, by Samuel Bray
Professor Mila Sohoni is the author of a string of significant articles on administrative law, and two of her forthcoming articles are about national injunctions. One of these is The Lost History of the “Universal” Injunction, which is forthcoming in the Harvard Law Review. The article is a vigorous argument that national injunctions are in continuity with the long-established practices of federal courts. The article identifies what can be taken to be the earliest known national injunction—from 1939—and it uncovers a rich history of federal injunctions against the enforcement of state laws at the beginning of the twentieth century. The article also pushes legal scholars working in this area to consider normative questions about the judicial power, the relevance of equity’s history for the present, and the distinction between federal and state statutes for purposes of equity.
Nevertheless, in this piece I want to give critical attention to the argument. In my view, there are four serious problems with The Lost History. In short form: (1) the 1939 injunction that the article identifies was reversed by the Supreme Court and it actually undermines the article’s argument; (2) the article does not adequately contextualize the cases it discusses; (3) equity’s tradition of representative suits is misunderstood; and (4) the argument cannot be squared with central features of American legal history in the 1920s and 30s (i.e., the declaratory judgment controversy and the legal challenges to the New Deal). In my view, these problems together render the article an unreliable guide to the history of equity in the federal courts.
First, it is to Professor Sohoni’s credit that she has uncovered Lukens Steel Co. v. Perkins, which can be understood as pushing the date for the the first known national injunction back to 1939.* The short version of the case is that the D.C. Circuit issued a preliminary injunction prohibiting the enforcement of the Secretary of Labor’s minimum-wage determination for the steel and iron industry across the country; that injunction was reversed by the Supreme Court.
But the article misreads the significance of the case. The article says that the Supreme Court reversed because of a lack of standing and was silent on the scope of the injunction (pp. 5, 66-67, 70—the last cited page of the article says “Perkins did not implicitly limit or reject the injunction that reached beyond the plaintiffs”). But that is not so, and the article’s separation of “standing” from “remedies” is artificial. The reason there is an Article III problem with the national injunction is because one individual plaintiff doesn’t have standing to seek a remedy on behalf of others not in some way before the Court. (This point is made unmistakably plain in Gill v. Whitford, a case The Lost History does not cite.) Indeed, in the opinion of the Court in Perkins, Justice Black explicitly says that the lower court’s “action”—which is its injunction—”goes beyond any controversy that might have existed between the complaining companies and the Government officials” (quoted on p. 65–all cites are to the draft of September 23, 2019). When Justice Black’s paragraph is set out in full, it is hard to miss the Court’s emphasis on the remarkable breadth of the injunction and its sense of surprise at the lower court’s temerity:
In our judgment the action of the Court of Appeals for the District of Columbia goes beyond any controversy that might have existed between the complaining companies and the Government officials. The benefits of its injunction, and of that ordered by it, were not limited to the potential bidders in the ‘locality’, however construed, in which the respondents do business. All Government officials with duties to perform under the Public Contracts Act have been restrained from applying the wage determination of the Secretary to bidders throughout the Nation who were not parties to any proceeding, who were not before the court and who had sought no relief.
310 U.S. at 123. All of that was worthy of remark to the Court as it reversed the injunction 8-1. Indeed, at the start of the opinion of the Court, Justice Black framed the entire question in terms of whether the injunction was inappropriately overbroad:
We must, therefore, decide whether a Federal court, upon complaint of individual iron and steel manufacturers, may restrain the Secretary and officials who do the Government’s purchasing from carrying out an administrative wage determination by the Secretary, not merely as applied to parties before the Court, but as to all other manufacturers in this entire nation-wide industry.
Id. at 118; see also id. at 120-121 (noting yet again that the plaintiffs “did not merely pray relief for themselves against the Secretary’s wage determination but insisted that all these Government officials be restrained from requiring the statutory stipulation as to minimum wages in contracts with any other steel and iron manufacturers throughout the United States,” and that the district court “declined to interfere so sweepingly with the administration of the Act”). This is all in plain sight in the opinion of the Court. In still another part of the opinion Black even quotes from Justice Sutherland’s opinion in Massachusetts v. Mellon (companion case to Frothingham v. Mellon), to the effect that “a suit of this character cannot be maintained.” It is true that Black doesn’t put this under the rubric of “scope of the injunction,” but rather the fundamental relationship of judicial power to the political branches. So did Frothingham. We cannot understand the problem of the national injunction, and we cannot understand older cases like Frothingham and Perkins, if we treat “standing,” “judicial power,” “equity,” and “remedies” as separate boxes. They are reinforcing and interrelated concepts (again, as the unanimous Court spells out in Gill v. Whitford). ’Tis “murder to dissect.”
In fact, the striking thing about Perkins is how much it confirms the previously recognized timeline for the development of the national injunction. In 1940 the Supreme Court reversed an injunction that went beyond the “controversy” of the parties. The wide-ranging historical investigation that is reflected in the article has failed to show a single non-reversed national injunction before 1963.
Second, there is a pattern of incomplete contextualization for the cases that are presented. Now I have not yet explored the history of each case presented in the article, because doing so requires scrutiny not only of the opinions but of the record. There can be a stray line in a case involving an injunction—a line that looks to us now like vindication of a national injunction—yet such a reading will become impossible when set against the larger procedural background. An example of this that I discuss in Multiple Chancellors is Panama Refining Co. v. Ryan (see pp. 433-434 & n.87). But where I am familiar with the background of the cases discussed, I find a tendency to highlight the stray line and not to place it in context.
Let me give one example. The article relies on West Virginia v. Barnette as a case where the Supreme Court endorsed an injunction that prohibited the enforcement of a law not merely against what Sohoni calls “the plaintiff class.” The article relies on this case as “establish[ing]” that the Supreme Court did not have any qualms about the scope of the injunction in Perkins (p. 70; see also pp. 5, 62, 71). And it is true that the injunction affirmed in Barnette prohibited enforcement of the state law against “the children of the plaintiffs, or any other children having religious scruples against such action” (i.e., the flag salute). See Final Decree, in Transcript of Re cord, West Virginia State Board of Education v. Barnette, No. 591, Supreme Court of the United States, October Term, 1942, at 46. But what the plaintiffs sought was an injunction protecting the children of Jehovah’s Witnesses. See First Amended Complaint, in Transcript of Record, at 15. It is not at all clear that the three-judge district court granting the injunction was thinking of “children having religious scruples against the flag salute” as a broader set than “children of Jehovah’s Witnesses”—and that group would have a claim analogous to the bill of peace. See id. at 10 (invoking as heads of equitable jurisdiction “[no] adequate remedy at law” and “multiplicity of suits”). (Side note: in the case Barnette overruled, the injunction requested from and given by the district court was plaintiff-protective. See Gobitis v. Minersville School District, 24 F. Supp. 271, 272, 275 (E.D. Pa. 1938).)) In short, set against the complaint, the bill of peace, and equity’s tradition of representative suits (discussed more below), this looks like what we would now call a class action for all Jehovah’s Witness school children in West Virginia. That is more plausible than saying the court rendered a “universal injunction.” Moreover, the injunction ran only as to a defined group, not universally to everyone in the state. But none of this is presented in the article, and the reader is given the impression that between the “any other children” phrase and the Supreme Court’s silence about remedies, the case establishes that as of 1943 the Supreme Court had no problem with universal injunctions.
Another illustration: the article seems to characterize Lewis Publishing, in 1913, as the first known national injunction (p. 23), a case in which “the plaintiffs . . . were able to win an injunction barring the enforcement of the new federal law against anyone until the merits had been decided—much as a nationwide universal preliminary injunction does today” (p. 26). What is the foundation on which this striking claim rests? It is based on what the article refers to as “a brief per curiam opinion” (p. 26). How brief? Quite brief. Here is the entirety of the “opinion”: “Granted.” And the striking singularity of this reading becomes all the more clear when this one-word non-opinion is read in context: first the publisher plaintiffs did not seek an injunction protecting other newspaper publishers, then the government (according to the plaintiffs) agreed that it would not apply the challenged provision to any publishers during the pendency of the suit, then the government (according to the plaintiffs) went back on its word, and finally the SG made no response to these claims by the plaintiffs. (The preceding points are all in the article, but they are given no interpretive significance.) Now it could be that lurking in the white spaces between the letters in the word “Granted” we have an intuition by the Supreme Court that national injunctions are A-OK. Alternatively, and more plausibly, the Court could have been making no new law and treating this as a matter of estoppel: the government didn’t need to make the representation that it wouldn’t enforce the provision against newspapers during the pendency of the suit, but once it did, it should be held to it. If the article placed Lewis Publishing in its legal and historical setting, these are the kinds of ambiguities and irresolutions that would need to be surfaced. In short, what does Lewis Publishing show about whether “in the period from 1890 to 1943, the law-declaration model animated and guided the actions of federal courts of all stripes” (p. 8)? Exactly nothing.
Third, there is a misreading of equity’s tradition of representative suits—a misreading that impairs the conclusions that the article tries to draw from the state-law cases as well as from the federal-law cases. It is widely recognized that the class action was a creature of equity, a descendant of the bill of peace and more generally of representative suits in equity. Here is Wright & Miller:
Class suits long have been a part of American jurisprudence. From an early date they were authorized in federal courts by the equity rules for suits involving members of a class so numerous that it was impracticable to join them all as parties. Justice Story, building upon the doctrines developed by Lord Eldon for the English courts, generally is credited with having formulated the standards in this country. His analysis was adopted in Smith v. Swormstedt, in which the Supreme Court allowed a representative suit to be brought on behalf of all the preachers in the Methodist Episcopal Church South seeking a declaration of the respective rights of each sectional group of the Methodist Episcopal Church of the United States to funds originally belonging to the entire church.
1751 History and Purpose of the Class Action, 7A Fed. Prac. & Proc. Civ. § 1751 (3d ed.). The Lost History concedes, as it must, that Equity Rule 38 was at least a “progenitor” of the modern class action (p. 42). That rule, adopted in 1912 and based on the previous federal equity rule on the subject, said: “When the question is one of common or general interest to many persons constituting a class so numerous as to make it impracticable to bring them all before the court, one or more may sue or defend for the whole.” But the article tries to escape this close connection between the equity representative suit (growing out of the bill of peace) and the class action. The article correctly identifies differences between the tradition of representative suits in equity and the modern class action—no talk of “certification” in the old cases, no requirement to go beyond certain allegations in the complaint. The article doesn’t see the tradition of representative suits in equity as represented now by the class action; it calls them simply “cases in which plaintiffs obtained relief on behalf of others well before the modern-day class certification procedure came to be adopted” (p. 14; see also p. 7).
What exactly is the point of all these differentiations? It is the article’s conclusion that “these cases enjoining the enforcement of state law look a lot more like modern-day universal injunction cases than they do like modern-day class actions” (p. 55). But noting a few requirements in modern day class actions that don’t exist in their older equity antecedents only works if the argument is something like this: the class action is a newly built thing, circa 1966, and it offers one way plaintiffs may proceed, but they may also continue to proceed with a non-FRCP-based equitable suit on behalf of all similarly situated parties, and that is what the recent outpouring of national injunction cases really is. If true, very interesting. We should expect courts to start putting flesh on the bones of this newly resurrected non-FRCP tradition of equitable representative suits; we could bring back all of Joseph Story’s writing on necessary parties in equity; we could cite Smith v. Swormstedt and Supreme Tribe of Ben Hur to show that if a suit for a national injunction fails, everyone who it would have benefitted is bound by the defeat. I suspect for some reason—actually, for many reasons—that the public interest litigators and state AGs who are currently seeking and obtaining national injunctions will not see this as a rosy opportunity.
Let me give a specific instance of how the article fails to appreciate fully the equity tradition of representative suits. One of the article’s central examples is Hill v. Wallace, about which the article says: “Though it was not formally nationwide, the Hill interim injunction appears to have effectively hit ‘pause’ on the coercive enforcement of the new federal statute throughout the country—exactly as many modern nationwide universal preliminary injunctions do.” On no understanding is this remotely a national injunction—as the article says, the injunction ran against the Chicago Board of Trade, the local federal collector of internal revenue, and the U.S. attorney for the Northern District of Illinois (p. 29). But the more basic reason this case is not an antecedent for “universal” or “national” injunctions is that this is a thoroughly unremarkable instance of an American bill of peace brought by some shareholders of a corporation on behalf of all of them. Justice Sutherland’s opinion for the Court in Frothingham v. Mellon explicitly singled out such suits as within equity’s power, even as he was distinguishing them from the relief requested by the plaintiff in that case (which I consider a national injunction). The article gives the following account of why Hill v. Wallace shouldn’t be considered a garden variety representative suit: “The eight members’ suit against the CBoT can be (and was) conceptualized as a stockholder’s suit, but with respect to the federal government, that frame is obviously inapt. On the latter score, the suit was simply an action to protect each of the 1,600 members of the CBoT from the record-keeping, tax, and criminal provisions of a bothersome new federal statute through a representative suit by some of their number.” I cannot follow the argument. “Obviously inapt” is conclusory. “Simply” seems to suggest some kind of idea that it can’t be bill of peace if—if what exactly? There is nothing at all “odd” (p. 30, n. 81) about how Hill fits with the reading of Frothingham offered in Multiple Chancellors. It is as simple as this: Hill is a representative suit by some of the shareholders on behalf of all of the shareholders; Frothingham says equity can do that.
If we think about Hill v. Wallace not with anachronism, but with an internal perspective on contemporaneous equity, it is completely unsurprising as an exercise of judicial power. That is why the government did not object (p. 27). There were representative suits in equity; there was a line of authority (not uncontradicted) that those suits could give decrees that settled the rights of people who were not in the courtroom; but the theory was one of representation in equity in which the rights and interests of those not present were represented by those who were. We have a modern incarnation of that tradition: it is the class action. Failure to recognize that continuity of organic development is what leads the article to find “universal injunctions” against state laws—but these are better seen as ordinary representative suits in equity, analogous to class actions today. They were uncontroversial: not because “universal injunctions” were uncontroversial, but because representative suits in equity were uncontroversial.
Fourth, the article fails an important test of any legal intellectual history: take the proposed conclusions, assume they’re true, and ask how the relevant actors would therefore have behaved. Do they act like they should have acted with these views? Two of the big developments of the 1920s and 1930s would be quite astounding if the article’s account were sound. One is the enormous controversy over the declaratory judgment, much of which centered on the effect of a court’s judgment on non-parties. This is why the assimilation of the declaratory judgment to Article III required such a strong focus on party-specific resolution (e.g., Haworth). The other is the huge number of suits challenging New Deal legislation, many successful, apparently without any national injunctions being granted. (See Multiple Chancellors.) The article’s explanation for why there weren’t more national injunctions in the first half of the twentieth century is the government’s willingness not to enforce a law against anyone during the pendency of a suit challenging the law (pp. 35-36). But that explanation will not work for the New Deal, where the Roosevelt administration strenuously resisted the suits and by official policy continued to apply challenged laws in other districts. (Again, see Multiple Chancellors.) If the article is right, the declaratory judgment should have been uncontroversial and the New Deal challenges should have been a hotbed of “universal injunctions.”
The Lost History explores the fascinating past of federal injunctions against the enforcement of state laws. And there are a number of important questions the article pushes us to consider. One is federal and state similarity or difference (perhaps too state and municipal similarity or difference). Another is the normative question of whether the class action with its developments over the last fifty years strikes a better balance than the old representative suits of equity. Yet another is important questions about how the past of the equity tradition should relate to the present, especially in U.S. public law. The article also serves as a valuable reminder to look to federal injunction practice with respect to state courts. As a work of historical scholarship, however, the article’s failure to understand the continuity between the representative suit in equity and the class action, its insufficient contextualization of the cases, and its inversion of the significance of Perkins represent a missed opportunity. I say this with regret, because there is a great need for sound historical work on equity in the federal courts.
Samuel L. Bray is a Professor of Law at Notre Dame Law School.
* There is a reasonable argument that the injunction in Perkins was actually not a national injunction, or at least not one that was self-consciously intended by the court to protect non-parties. Its scope was broad, but it can be thought of as entirely for the protection of the plaintiffs themselves. Critically, the plaintiffs’ attorneys argued that only an injunction covering the entire industry would protect them: If the injunction were to cover only the plaintiffs themselves, allowing the Secretary of Labor to apply the minimum-wage requirement to all competitors, there would be short run competitive advantage for the plaintiff firms, but they feared that any government contracts they procured would later be challengeable as illegally granted under the competitive-bidding rules. (See pp. 350-351 of the Record in the Supreme Court.) If the argument is accepted, as it seems to have been by the court of appeals, the injunction would be broad but it would actually comply with the proposed standard in Multiple Chancellors: “An equitable remedy is flexible, and its scope is not automatic. . . . But there should be no term or breadth that is for the protection of nonparties rather than for the protection of the plaintiff” (Multiple Chancellors, at 471). The injunction was still too broad—which explains the Court’s surprise at its scope—but it was too broad in its attempt to protect the plaintiffs, not too broad in an attempt to protect nonparties. It was in this respect unlike the national injunctions that have dominated executive branch-judicial branch relations for the last five years.