from the stopping-the-stopping dept
There is a disquieting update to my last post about a district court victory in Does 1-26 v. Musk that had enjoined DOGE, given its likely unconstitutional exercise of power, particularly in the context of its dismemberment of USAID. The Fourth Circuit has now stayed enforcement of that injunction, which raises a few issues worth commenting on.
First, on substance, the most alarming issue with its decision is that two judges on the Fourth Circuit found the arguments that Musk and DOGE were acting with an unconstitutional power unpersuasive. For example:
And as to Musk, the evidence before us creates a strong likelihood that he functioned as an advisor to the President, carrying out the President’s policies of shrinking government and reducing spending, not as an Officer who required constitutional appointment. The current evidence in the record indicates that Musk’s actions did not involve the exercise of authority of an office granted by law but rather the implementation of Executive policies. In order to be an Officer, he must (1) be “exercising significant authority pursuant to the laws of the United States” and (2) be “occupying a continuing position established by law.” Lucia v. SEC, 585 U.S. 237, 245 (2018) (cleaned up). Based on the current record, it appears that Musk’s role satisfies neither criterion. [p. 8]
As discussed in the last post, the district court judge carefully walked through each of those elements pertinent to determining whether Musk, as well as DOGE, were behaving within constitutional parameters before concluding that the answer was likely not. For the appeals court to simply cast all that analysis to the side was itself troubling, and the dissenting concurrence called it out:
The Majority questions this characterization of the results of Defendants’ actions. Maj. Order at 7. But, in doing so, it improperly meddles in factual determination properly left to the discretion of the district court. Second guessing factual findings of a district court in an emergency stay posture where the applicant has not even asserted that such findings were erroneous, let alone clearly so, is entirely inappropriate. As we have long held, a party seeking reversal of a district court’s injunction order must “overcome a deferential standard of appellate review.” Pierce v. N.C. State Bd. of Elections, 97 F.4th 194, 210 (4th Cir. 2024). We evaluate a district court’s decision to grant a preliminary injunction under an abuse of discretion standard and only review the district court’s factual findings for clear error. Id.; see also Roe v. Dep’t of Def., 947 F.3d 207, 219 (4th Cir. 2020). Indeed, “[w]here there are two permissible views of the evidence, the [district court’s] choice between them cannot be clearly erroneous.” Id. (citing Anderson v. City of Bessemer City, 470 U.S. 564, 574 (1985)). [p. 17]
For its part the majority defends its dubious read of the record:
Our concurring colleague claims the Majority “improperly meddles in factual determination properly left to the discretion of the district court,” in violation of our standard of review. Conc. Op. at 17–18 n.1. He is right to emphasis the importance of those standards. We are duty bound to apply them. But we apply our standard of review here in the context of our precedent that instructs that preliminary injunctions should be granted “sparingly and in limited circumstances” because of their implication of “very far-reaching [judicial] power,” Direx Israel, Ltd., 952 F.2d at 816, and that “ambiguity is simply insufficient to support a finding that success on the merits is ‘likely’ rather than merely ‘possible’ . . . .” Di Base v. SPX Corp., 872 F.3d 224, 235 (4th Cir. 2017). In that context, our conclusion is consistent with our standard of review. [p. 8]
It’s a flimsy defense of the majority’s analysis of the government’s “likelihood of success” on the merits, but for the moment the larger issue with this stay is what has been an issue in this and other cases: the difficulty in being able to obtain usable and immediate injunctive relief.
Because even the “dissenting” judge on this appeals court panel, who spent 24 pages on the “likelihood of success” factor explaining why Musk and DOGE were likely behaving with unconstitutional power [p. 17-40], found himself having to concur with the result in setting aside the injunction “because the proper defendants were not before” the courts in this case. [p. 40].
After the district court found that Musk and DOGE were responsible for dismantling USAID, the question then became: what’s next? The Majority correctly identifies, the constitutional violation and the remedy cannot go both ways: if Musk and DOGE lack the authority to order USAID officials to do anything, enjoining Musk to direct USAID officials to restore employment, building access, and security services, among other relief, also cannot stand. Plaintiffs did not name defendants with the proper authority to effectuate the relief that Plaintiffs seek. Suing Musk and DOGE constrained the court’s ability to redress their constitutional claims, which I have noted above, and as the district court found as a matter of fact, were quite strong. [p. 41-42]
So one issue emerging from this case is that DOGE challenges, at least those seeking injunctions, do need to be hybrid, with the agency and its officials running it named as co-defendants, because even if the argument is successfully made that what DOGE is doing is unconstitutional, an injunction telling them to “stay out” also needs to include an order on the agency to “keep them out.”
And one reason it is important that the injunction be so broad is because, now that Trump has had a chance to install his own cronies in agency leadership, the DOGE and “official” leadership are blurring together. In this case they explicitly blurred together, because Trump tried playing musical chairs, moving out the “official” person and replacing him with someone who had himself been DOGE. The district court called shenanigans, because it would mean that the new person, now “official,” could just ratify everything that DOGE had “recommended.” The injunction appeared to assume that by removing DOGE only independent agency judgment would be left, and that judgment would be exercised consistent with whatever lawful authority it had. But such a view appears naïve, because the blurring is inevitable, as a Trump toady in a DOGE hat may be no different than a toady in an official agency hat—they are both committed to the same constitutionally destructive mission, and both need to be stopped from completing it, under whichever auspices they claim to be using as they do it.
Also, the question of agency ratification is proving to be a bigger issue that needs to be dealt with directly. In part to address it itself and ensure that agency officials cannot somehow launder illegitimate DOGE “recommendations” into legitimate agency action, and in part because what irretrievably taints these “recommendations” is also what shows how what DOGE was doing is so unconstitutional—and also why there is no possibility of irreparable harm to the government if its own actions, including implementation of DOGE’s “recommendations,” is stopped.
In a universe where there were no DOGE, no agency official in their right mind would try to do what DOGE is “recommending” and cancel these contracts, fire these people, freeze this funding, or close down the agency altogether. Perhaps via some lawful path the agency could do some of these things but it could not do such things via this path because taking these drastic actions would, at minimum, seem to violate the APA’s prohibition against arbitrary and capricious decisions by the agency. It’s a big reason why these cases seeking injunctions need to be hybrid, so that even if agency officials claimed to be doing these things via their own independent judgment they still could be enjoined because of how that judgment violates the law. But also, if these decisions would have been wrong in a universe without DOGE telling them to make them, they are no less wrong when made in the wake of DOGE “advising” them to be made. It would be absurd if agency officials could “ratify” and somehow legitimize a DOGE “recommendation” that the agency could not make itself, and especially not when DOGE arrived at the “recommendation” via its own unlawful behavior. Two wrongs cannot make a right.
Because even if agency officials could make such decisions in a vacuum, no such vacuum exists here. It goes back to the idea raised earlier that DOGE’s hack is the harm, as well as the evidence of the harm. In other words, it was DOGE’s initial invasion of the computer systems that is both evidence of the abuse of power—that it was able to occur—and the vehicle used to cause the damage, as any “recommendation” they made stemmed from that illicit access. As a result that illicit access irredeemably taints any “recommendation” they made and thus makes it impossible for any official to ratify any of what they “proposed.” That usurpation of power, to force access to those systems, also undermines any claim that its “recommendations” were just “suggestions” that agency officials were free to refuse, and not the unconstitutional demands, like the demand to be given their illicit access to the computer systems in the first place, that they were. It would also independently offend the APA to ratify any action DOGE demanded—sorry, suggested—DOGE had made because it had made them, especially given how unlawfully it arrived at the instruction.
In fact, the more the government claims that it would suffer “irreparable harm” if these DOGE recommendations are not adopted, or that any further DOGE action within the agency be stopped, the more it shows how unconstitutional the entire enterprise is. The government’s only possible defense in these challenges is that the agency could somehow lawfully do what it is doing—cancel these contracts, fire these people, freeze this funding, shut down the agency—independent of DOGE. As discussed above it’s a dubious argument given how drastic these effects are, which laws like the APA would probably prohibit, but if the government can show that the agency could lawfully do any of the things that DOGE is charged with causing to happen according to their own independent judgment and authority then the challenges would fail.
But that argument, that the agency could do these things on its own, would mean that it could do them on its own, without DOGE. And the more the government claims that it would be harmed if DOGE were benched, the more it undermines any claim to the agencies’ independence. Removing DOGE from the equation should just leave the agency back where it was supposed to be, and if that’s now a problem for the government, then it’s a problem for the Constitution that these unappointed people now have such power over agency action. DOGE cannot cause something to be done that lawfully could not have been done without it, but the more the government argues that it is being harmed by excluding DOGE from agency activities, the more it shows how unlawful its activities are.
And it is why the irreparable harm factor needs to carry a lot more weight for challengers to anything DOGE has done (or Trump generally, even in instances without DOGE involvement). If it turns out that everything DOGE is doing is lawful, it can still be done via lawful means. Perhaps it can’t be done as quickly, but the government has yet to make a non-frivolous argument for why there is any true exigency requiring fast action—if there is any emergency it is the one it itself created, a fact that cuts the other way for why there should be relief afforded to challenging plaintiffs immediately in order to truly preserve the status quo.
Because if it turns out, as many district courts are finding, as well as the dissenting judge here, that DOGE is acting unlawfully and indeed unconstitutionally, then then the harm to the challengers, and the public beyond them, if relief isn’t afforded is immense. The majority in this case here dismisses the harm potential as potentially being easily remediated by money later, [p. 12], but their math seriously underestimates the amount of harm being done even in financial terms, the incalculable harm that comes from such a severe Constitutional violation being tolerated, which no amount of money can ever compensate, and the fact that it will be taxpayers—or, in other words, the plaintiff victims themselves—who will have to pay whatever compensation might eventually be ordered. Everything DOGE has been doing amounts to irreparable harm, and far beyond any sort of harm the Executive could experience if his and Musk’s power were carefully tested before letting it loose on the country.
Filed Under: 4th circuit, doge, elon musk
