Thoughts on the universal injunction argument (Updated)

Transcript here. Lee Kovarsky live-Blueskied the argument and gave a quick end-of-argument summary. I agree with Lee that Justice Gorsuch seemed more open to some universal relief in some cases than expected and that he and Barrett might join Sotomayor, Kagan, and Jackson to allow universal relief in some situations. But it really was impossible to predict. Lee is also adamant that Rule 23 cannot bear the weight non-universalists would put on it.

Other thoughts after the jump:

• A lot of discussion of how to get cases to SCOTUS as quickly as possible and how to allow SCOTUS to decide the issues as soon as possible. That took on a special focus when the SG suggested that the federal government would abide by SCOTUS precedent but not necessarily circuit precedent. It was part of a broader assumption and acceptance of judicial supremacy--and really SCOTUS supremacy. Lower courts make temporary predictive decisions that do not deserve as much deference from the executive, while the SG disclaimed power to ignore SCOTUS precedent.* Some questions (notably from Alito) suggested district courts are out of control and arrogating power to themselves.

[*] The SG could not pursue full-throated departmentalism in this case. The Court will not have it generally. It certainly will not have it in a case such as this, where the full range of executive power is laid bare.

I wonder if this case is the wrong vehicle to resolve universality because the substantive constitutional question is so obvious. Every lower court reaches the same conclusion and no one seriously believes any lower court will come out differently (or any district court that did would be quickly reversed). That gives the federal government an incentive to never seek cert. Instead, the government takes repeated party-specific losses in the lower courts but ignores circuit precedent (as the SG suggested the executive could and historically has done*) and continues to pursue non-party enforcement until a new order tells it to stop as to that party. Members of the Civ Pro Listserv suggested work-arounds,--prevailing plaintiffs seeking SCOTUS review and courts of appeals certifying questions to SCOTUS--while recognizing that these are rare and disfavored.

[*] To Barrett's surprise.

* The argument revealed competing (and irreconcilable) assumptions about  the system of constitutional litigation. The SG emphasized percolation, a necessarily slow process grounded in the dispute-resolution model of courts declaring law while resolving discrete disputes between parties. Jackson derogated this position as turning the justice system into a "catch me if you can" regime in which everyone must have a lawyer and file a lawsuit. Respondents and the justices questioning the SG emphasized the need for SCOTUS to decide constitutional questions  and make constitutional law (preferably sooner rather than later), the unquestionably binding-on-everyone nature of those SCOTUS decisions, and the harm that occurs pending slow, atomized litigation. Constitutional litigation must move fast and reach final resolution quickly; anything else leaves too many people exposed.

• Barrett and Jackson asked questions that illustrated confusion (on everyone's part) about what courts do in issuing injunctions and how that affects remedies. Does a court order the government not to do something unlawful ("Don't do X") or does a court order the government not to do something unlawful to the challenging party(ies) to the action ("Don't do X to this person"). Jackson and Barrett suggest that the court does the former. But they then suggest that only parties can enforce the injunction through an enforcement stopping conduct inconsistent with the injunction or by seeking contempt.

This disaggregates the "scope" of an injunction from who can enforce it. An injunction in A v. X prohibits enforcement against A; were X to attempt enforcement against A, A could ask the court for an order enforcing the injunction barring X from doing so and holding X in contempt. But a non-universal injunction does not affect X's attempt to enforce against B; B could not ask the court for an order enforcing the injunction and barring X from doing so and holding X in contempt. This also covers non-universal injunctions that incidentally affect non-parties--B cannot demand that X do a better job of cleaning the raw sewage out of the prison or of creating appropriate legislative districts; only A can ask the court to hold that X is out of compliance. An injunction becomes universal because of who it protects and because of who can enforce those protections.

Barrett and Jackson suggest another option: An injunction prohibits all enforcement ("Don't do X as to anyone"), but only the named party can enforce. That is, X cannot enforce the challenged law against B, although only A can ask the court to stop X from doing that or to hold X in contempt.

• No one seems to think about this problem in light of parallels between offensive pre-enforcement litigation (such as these cases) and defensive litigation. If the government attempts to enforce the EO against A (e.g., in a removal action) and A successfully raises the EO's invalidity as a defense to removal, the court will resolve the removal proceeding in A's favor; no one believes the order or judgment in that proceeding protects anyone other than A from deportation. The same should follow when A pursues offensive litigation to prevent that enforcement action, subject to devices available in equity (such as Rule 23). But if the judgment in United States v. A is not universal, there is no reason the judgment in A v. Trump should be universal.

• A lot of people in the argument attributed the rise in universal injunctions on the rise in unilateral executive action. I do not see why this matters for this case.  I suppose executive unilateral action produces more regulations, more APA challenges, and more universal vacatur, which raises similar-but-distinct issues. Perhaps unilateral executive action via EO produces more laws, more litigation, and more injunctions than where law only comes from the longer legislative process. But this is not a separation-of-powers issue; it is a government activity issue. As Sam Bray has pointed out, the New Deal produced a wave of new laws from Congress producing (pre-1937) a wave of injunctions stopping enforcement, none of them universal. A similar flurry of congressional action in 2022 or 2025 would produce the same universal injunctions as we now see against Trump's EOs. We would be having the same universality conversation had Congress eliminated birthright citizenship by statute. 

• Both respondents offered limiting principles. Counsel for the states  identified three "buckets" in which broader remedies are possible (I need to review the transcript). Counsel for individuals said when necessary for complete relief and in exceptional circumstances. The first is baked into the existing remedial standard, although I am not sure how guaranteeing A's citizenship is incomplete if B's citizenship is not protected. The second would swallow the rule. Even if "exceptional circumstances" incorporates the merits and applies to really unconstitutional actions, every plaintiff believes that to be their case. In any event, I do not know why that should matter. The logic of universality is that B should not be subject to enforcement if A is not subject to enforcement. That is true regardless of whether the challenged law is a little bit unconstitutional or a lot unconstitutional.

Update: Finally had a chance to see the transcript. Counsel for the states offered three: 1) Alternative ways to remedy harms to parties are practically or legally unworkable; 2) Congress authorized universality (e.g., APA vacatur); and 3) Alternative forms of non-party relief are practically or legally unworkable (e.g, the plaintiff will be deported before the court can consider class certification at the TRO stage).

• There were concern for organizational plaintiffs and how they may not fit within the Rule 23 solution. Of course, organizational plaintiffs raise their own problems. They create (permissibly or otherwise) another work-around to Rule 23. And they raise the prospect of universality through an injunction protecting the organization and thus everyone one of its members, with the burden on the court to identify those members in subsequent efforts to enforce the injunction.