Birthright Citizenship and Class Actions after CASA

David Marcus has a terrific essay explaining why class certification in the birthright citizenship cases is an easy yes under existing law: The Class Action after Trump v. CASA.

Marcus is extremely knowledgable about class actions, and his prior scholarship includes a classic piece on civil rights cases and the development of Rule 23. That piece is Flawed but Noble: Desegregation Litigation and its Implications for the Modern Class Action (cited 4x in Multiple Chancellors).

I agree almost in toto with The Class Action after Trump v. CASA, and I think it demonstrates why we should not expect the unconstitutional executive order on birthright citizenship to ever go into effect. But I would nuance or call attention to three points.

First, there are two ways to read Marcus’s discussion of criticism by Michael Morley and others of nationwide class actions challenging federal rules. One is to say that Marcus is objecting to any bright-line rule against nationwide class actions, whether on the basis of percolation policy arguments or structural constitutional principles. The second is to say that percolation is not the sort of prudential consideration that leads a judge to certify a class that maps on to a state, a federal circuit, or some other geographic division that is less than the entirety of the United States. A lot depends on whether Marcus’s reference to “generic” percolation arguments means boilerplate, unreasoned ones, or any appeal to the general value of percolation.

If it’s the first reading, I agree, but if it’s the second, I do not. To encourage percolation of a question, there will be cases where it is sound judicial practice to limit the geographic breadth of the class action. I don’t think the birthright citizenship cases are a good illustration, though I differ with Marcus on why. Marcus suggests it would be too challenging on plaintiffs in these cases “who face serious structural and resource limitations that make piecemeal litigation a particular hardship.” I don’t find that persuasive. There were already three different cases consolidated in Trump v. CASA, and one of those started out as a suit by individual plaintiffs who sought a class action geographically delimited by the state of Washington (as noted in my adversarial collaboration with Jim Pfander). Were they wrong? Was that somehow unfeasible? The better argument for why statewide or (better) circuitwide class actions are unnecessary here is that the merits are so clear: President Trump’s executive order on birthright citizenship is unconstitutional, every court to consider the question has said that, and it is not in doubt that the Supreme Court will eventually say the same if squarely presented with the merits. In the abstract percolation is always good, but in the particular, percolation is pointless.

I say that with some hesitation, though. Not because of these cases, but because “percolation is pointless” is precisely the kind of inquiry that is inadvisable because of the heady mix of judicial polarization, aggressive forum-shopping, and merits-focused PI analysis that have marked many challenges to executive action over the last decade. We don’t yet know whether all the factors in that heady mix are going to continue unabated after Trump v. CASA. So here percolation is pointless, but should courts really be asking that? But how do they consider percolation as a prudential consideration if they don’t ask that?

There’s also a question of judicial restraint and realism. In other words, lower courts should question whether short-term maximization (or to put it differently, single-minded minimization of litigation costs for challengers) is the wisest approach. If every suit challenging a federal policy becomes a nationwide class action, then the chance that class action doctrine is revised in a restrictive direction increases. If some of the challenges are brought as multiple class actions, with multiple courts weighing in, the policy for argument for greater restrictiveness loses much of its force.

Second, in footnote 57 Marcus recognizes the possibility of subclasses for the two groups distinguished in the executive order. In the oral argument in Trump v. CASA, the Solicitor General pointed to the difference between the two groups when pressed to come up with some reason—any reason—why class certification would not be obvious. Subclasses seem unnecessary to me, but I note this point in case they are sensible from an abundance of caution.

Third, Marcus has a sentence saying that vacatur is “a power that survives Trump v. CASA.” It survives in the sense that it will continue to be used and continue to be contested. But one of the most important lines of argument for vacatur of rules—that the APA took a preexisting practice of granting universal injunctions and codified it—is now decisively rejected. That line of argument is option 2 in the taxonomy of The Truth of the Truth of Erasure. So vacatur lives to fight another day, but it does so on distinctly less favorable terrain, and the result should not be assumed.

A few brief notes in conclusion:

• Before CASA and afterwards, remember that preliminary injunctions are not supposed to be a quick trial on the merits. Their job is to preserve the court’s ability to give effective remedies at the conclusion of the litigation (see The Purpose of the Preliminary Injunction). There is going to be a temptation in the lower courts not to use them this way, especially in cases where the merits are so clear, like the birthright citizenship. Resist temptation.

• Watch out for a tacit collaboration between the Trump administration and its challengers in overstating the legal effect of an executive order. We should not lightly assume that an executive order is law.

• One of the puzzles of the birthright citizenship cases is whether the appropriate remedy is really an injunction at all. There is a long history of declaratory relief for citizenship. (Some of that is recounted in Preventive Adjudication.) Is a declaration of citizenship—rather than the non-enforcement of an executive order that is not itself law—what the plaintiffs really need?

• Finally, I won’t be able to write at length at the moment on Jack Goldsmith’s identification of a possible “legal mistake” in Trump v. CASA at the always trenchant Executive Functions Substack. Suffice it to say that I disagree and think there is no mistake at all in Trump v. CASA’s discussion of the Judiciary Act of 1789. By its own terms that Act covers cases in which the federal government is a petitioner or an alien is a party. More fundamentally, a number of post-Erie cases have read the Judiciary Act of 1789 as authorizing/recognizing federal court access to the body of equitable doctrine in all cases, not just diversity cases. See, e.g., the Stainback footnote quoting Atlas. On this point, Trump v. CASA is just restating what the Court has done for at least 85 years. Neither Grupo nor its antecedents distinguished diversity jurisdiction and federal question jurisdiction for these purposes. Jack may disagree with these cases, but Trump v. CASA can hardly be faulted for following them. Finally, as Jack suggests (more tentatively than I will say here), it would make zero difference. Assume we did say that in federal question cases we were supposed to start with equitable doctrine in 1874—a position to reject for numerous reasons, including the historical fixedness it might imply, as well as the division and profusion of equitable doctrines it would portend. That shift would have no effect on the outcome of Trump v. CASA, because there were no universal injunctions in 1789 or 1874. Nevertheless, the questions Jack raises in his final paragraphs about equity and the federal courts are important ones, and I expect they will be the subject of renewed attention by Jack, me, and many others. Of making many articles there is no end.