Within a day of the Trump administration’s January 27 travel order, a U.S. district court ordered Customs and Border Protection (CBP) to permit lawyers access to certain detainees at Dulles Airport. There have been allegations that CBP failed to comply with the order, prompting the Virginia attorney general to move that CBP show cause why it shouldn’t be held in contempt; the judge has denied that motion without prejudice. These events have drawn attention to a foundational question in administrative law and separation of powers: If the federal government disobeys a judicial order, can the courts use their contempt power to force compliance? How willing have federal judges been to exercise this power, and what are its limits?
As it happens, I’ve been researching this issue with a team of student assistants for about the last two years. The paper, “The Endgame of Administrative Law: Governmental Disobedience and the Judicial Contempt Power,” forthcoming in the Harvard Law Review, provides the first general assessment of how federal courts handle the federal government’s disobedience. Conventional methods of legal research, focused on appellate case law, are hopeless for addressing the matter. There are no opinions of the Supreme Court on the subject, and while the courts of appeals have handled the issue many times over, they have done so in a manner calculated to avoid setting clear and general precedent. My team and I have assembled relevant cases through a wide-ranging search covering over 12,000 opinions (many of district courts) and 4,000 docket sheets. In these cases, I examined numerous unpublished opinions and orders, briefs, and other filings, and also conducted archival research and interviews. Here I offer a partial summary, focused on what’s of immediate interest for anyone litigating these issues.
In general, it’s not uncommon for federal agency compliance with court orders to be imperfect and fraught. This is especially true when a court orders an agency to take an affirmative act that is costly or complex. That kind of order can strain the agency’s funding and personnel, mess with the agency’s ability to fulfill and prioritize its many legally-required tasks, and press the agency to act without enough information to make its decisions rational and legally defensible. If an agency gets hit with such an order, it will often return to court warning the judge that it badly needs more latitude (especially more time) to comply. Judges often relent, cutting slack and extending deadlines while demanding progress reports and other assurances from the agency to satisfy themselves of the bureaucrats’ good faith. But the judge may become suspicious that the agency is delaying and demanding slack more than is reasonable, perhaps out of incompetence or a brute political aversion to doing what’s been ordered. In that situation, judges may consider making the order more detailed, making the reporting requirements more intrusive, and ultimately, finding the agency or responsible officials in contempt.
The potency of contempt findings against private defendants arises most obviously from the court’s power to attach sanctions of fine or imprisonment. But are contempt sanctions available against federal government defendants? There are three main possibilities to consider:
(1) Contempt Fines Against the Agency as an Institution? At first glance, the idea of federal courts fining federal agencies may seem meaningless. Wouldn’t the federal government just be paying money to itself? Actually, the idea is a lot more meaningful than this, and the reason why it virtually never happens is subtle.
The money would most likely come out of the agency’s appropriations, in accordance with opinions of the Justice Department and the Government Accountability Office (GAO) that an appropriation for a government program is available to pay any penalties the agency incurs in the course of implementing the program. The money would go into the registry of the court imposing the fine, and the judge could either keep it there or allocate it to other purposes, such as paying for compliance-related activities that the judge wants the agency to undertake. Effectively, the fine would be a means for the judge to take away part of the agency’s budget and either freeze it or use it for other purposes than the agency would. The notion of docking an agency’s budget to deter agency wrongdoing is hardly unfamiliar: it’s a feature of various statutes that seek to stop agencies from breaching contracts, retaliating against whistleblowers, or engaging in unjustified litigation.
That said, there could be complications. If Congress wanted to shield the agency, it could replenish the budget losses with new appropriations. Or if the President wanted to shield the agency, he could try to pressure the Treasury Department’s Bureau of Fiscal Service to pay the fine out of the Judgment Fund, which is an indefinite appropriation available to pay any judgments or settlements against the federal government. The GAO used to administer the Fund and said it could not be lawfully used to bail out agencies for penalties (including most types of contempt fines), but Treasury took over the Fund in 1996 and appears not to have public guidance on this question. In any event, the actors who could most likely shield the agency’s budget from a contempt fine are Congress and the President, and imposing the fine would have the effect of elevating the agency’s noncompliance to the attention of those actors and forcing them to take responsibility for using public money to protect the agency from the consequences of its disobedience.
But we’re getting ahead of ourselves. The question of who pays a contempt fine against a federal agency doesn’t even arise unless an agency incurs the fine to begin with. And there is an argument that no such fine can be incurred. The reason is sovereign immunity: the principle that no remedy can be had against the federal government without consent of Congress. Plaintiffs would argue that Congress has in fact consented, in the 1976 statute that waived sovereign immunity for any suit against a federal agency “seeking relief other than money damages.” 5 U.S.C. § 702. Are contempt fines “money damages” under this provision? You might argue they are, at least if the fines are of the compensatory variety, meant to compensate the plaintiff for loss suffered due to the defendant’s noncompliance. But it’s tougher to argue that contempt fines are “money damages” if they are of the coercive variety (a certain sum per day until defendant complies) or criminal variety (a flat sum imposed to punish past disobedience), since those fines are not measured by the plaintiff’s loss, and they don’t necessarily go into the plaintiff’s pocket. Further, § 702 has always been read to waive immunity to injunctions, and plaintiffs would contend that coercive and criminal contempt fines are necessary incidents to injunctions. Still, notwithstanding all this, federal defendants would point out that § 702 never mentions contempt fines — and waivers of sovereign immunity are supposed to be strictly construed in the federal government’s favor, especially when we’re talking about taking money from federal coffers. Then again, even if the government wins on this waiver question, plaintiffs could still raise a broader constitutional claim that, whatever Congress has done, sovereign immunity just doesn’t apply to contempt fines in the first place, because Article III’s grant of “judicial power” implies that federal courts must have the capacity to enforce judgments by sanction, even against the executive.
In fact, the judiciary has carefully avoided ever getting to the point where a federal agency incurs a contempt fine (except possibly in one or two low-profile cases of tiny fines imposed by district courts and not appealed). This isn’t to say that fining federal agencies is unthinkable. Several district judges have tried to do it in recent decades, and major agencies have come within days or even hours of starting to accumulate large per diem fines that would take away 4% or more of their daily budgets. Yet courts of appeals (or, sometimes, the MDL Panel) have unfailingly swooped in to stop any significant fine from actually being incurred. The judiciary could have achieved this fine-avoidance outcome by embracing the view — consistently propounded by the Justice Department across Republican and Democratic administrations for decades — that federal sovereign immunity bars contempt fines, but apparently it doesn’t want to do that. Instead the judiciary has nearly always found some more case-specific reason to block the fines, sometimes using stretched reasoning on other issues so as to avoid the fine question. The result is that explicit appellate holdings on federal sovereign immunity to contempt fines have been quite rare. They are nearly all in odd contexts to which § 702 isn’t applicable (like claims of government misconduct within criminal prosecutions), the one exception being a single case from the Eighth Circuit that upheld sovereign immunity — though even that case was narrow (covering only fines of the compensatory variety) and questionable (since the panel neglected to address § 702).
(2) Imprisonment of the Agency Official? If any corporate entity disobeys a judicial order, the corporate officer responsible for the entity’s disobedience can be sanctioned for contempt. This would seem to imply that the federal official responsible for an agency’s disobedience can be sanctioned, including by imprisonment. The Justice Department itself appears to accept this view. To be sure, it hasn’t always done so. Back in the 1950s, DOJ contended that individual federal officials are absolutely immune to contempt sanctions for official acts, similar to their absolute immunity to common-law tort damages. But in more recent briefing in 1997 and 2008, DOJ distanced itself from this blanket immunity claim.
That said, there is another line of argument — less categorical and more prudential — for the government to fall back on: that a court should not sanction a person who makes all reasonable efforts to comply. The concept of reasonable effort must be understood in light of the nature of the noncompliant organization and each official’s position within it. Recall that compliance problems are most common when the judge’s order tells the agency to do something costly or complex. That kind of order implicates the agency’s resource limitations, its competing legal mandates and priorities, and the question of what kind of technical and scientific information it needs to acquire to formulate its action correctly. A judge assessing how the agency handles these issues ends up having to make complicated judgments about agency management. Identifying an official as blameworthy for noncompliance entails an especially fraught judgment of this variety: the judge must figure out which official(s) should’ve acted, and what the official(s) should’ve done. To be sure, higher-level agency officials — up to and including the agency head — are formally responsible for whatever happens below them, but before a judge could imprison an agency head or other manager for noncompliance, she would probably need to find that actions promoting compliance were reasonably within that manager’s grasp and that the manager failed to take them.
In light of these prudential concerns, the judiciary has shown great reluctance to imprison officials. Imprisonment has occurred only twice, never for more than a few hours, and in both instances, the biggest losers proved to be the imprisoning judges, one of whom was thrown off the case for bias, while the other recused himself to avoid a similar fate. Cases coming near to imprisonment are almost equally rare. In the most spectacular of these — and the only federal agency contempt case of any kind to get near the Supreme Court — the Commerce Secretary in 1951 engaged in unusually clear disobedience when told to return the shares in a bailed-out company to their private owners. The D.C. Circuit ordered the Secretary jailed, only to have the Supreme Court stay the sanction at the eleventh hour — and grant certiorari in the case despite its previous refusal to do so. Justice Robert Jackson wrote separately to say the Court should’ve demanded immediate adversary briefing on the stay rather than simply grant it indefinitely — a move that, in his view, rewarded the government for its disobedience and signaled that the Justices had no stomach to use force against official lawbreaking. Soon after, the plaintiffs settled for a 50% discount, and the case basically disappeared from memory, setting no precedent. Doctrinally, the door remains open to imprisoning federal officials, yet judges’ prudential reluctance has proven extremely strong.
I should note there is no suggestion in the sources, for the 1951 case or any other, that the President would prevent the U.S. marshals (whose leadership he can hire and fire) from doing their clear statutory duty to carry out any judicial imprisonment order. For the President to do such a thing would violate the norm of presidential enforcement of court orders that has been unbroken since the Civil War.
(3) Fines Against the Agency Official? Just as a court could in principle imprison an official, so could it fine that official. Assuming the official must personally pay the fine, this sanction raises the same prudential concerns as imprisonment about identifying the blameworthy official and the actions that should’ve been taken. But in fact, the monetary liabilities that federal officials incur in doing their jobs are virtually always indemnified out of the appropriations of their agencies, and there is little reason to think contempt fines would be different. Thus, a contempt fine against an individual official — if indemnified — could be a mechanism for a court effectively to dock the agency’s budget without having to worry about sovereign immunity. Indeed, the Supreme Court blessed this kind of maneuver as a means to circumvent state sovereign immunity to contempt fines in Hutto v. Finney (1978). Still, not every point about state sovereign immunity carries over to its federal analogue. And doctrinal obstacles could arise. For example, if the fined official had no personal connection to the noncompliance (e.g., if the court tried to fine the agency head on a vicarious liability theory), contempt sanctions might be barred altogether, or compensable from the Judgment Fund.
In practice, courts have not much tested the viability of this maneuver as a means to circumvent federal sovereign immunity. In the cases we found of fines against federal agency officials — always for conduct pursuant to superiors’ orders or official policy — the fines were usually overturned on case-specific grounds or were pro-forma and not meant to be collected. In the three instances where this was not the case, the fines were compensatory or criminal in nature (meaning they were flat sums and didn’t accumulate continuously like coercive fines) and were modest in size, never more than $6,500 in 2017 dollars. Presumably the officials were indemnified, and the fines were mere pinpricks to the agencies’ finances.
* * *
The vanishing rarity of contempt sanctions against federal agencies and officials stands in contrast to contempt findings against such agencies and officials, which are more common. In successful contempt proceedings against federal defendants, the most typical outcome is a finding of contempt without fine or imprisonment — a “reprimand” with “no sanction,” as the D.C. Circuit said in Cobell v. Norton (2003). That contempt findings may come with no sanction does not render them powerless. As the Environmental Law Institute wrote in a report during the Reagan administration, after discussing a case against EPA with a sanctionless contempt finding: “Top management [at EPA] takes the threat of contempt quite seriously and personally, even though the threat is not real.” Consistent with this, our research found many examples of agencies shifting toward compliance on being faced with a mere contempt motion. Agency officials and their lawyers may sometimes be willing to push the envelope on compliance, but the prospect of a contempt finding can make them back off. In the social and media world that agency officials inhabit, being held in contempt can have shaming and reputation-harming effects. Attorneys and officials make decisions on the premise that such shame is a real force. The Justice Department, which is famously self-controlled in its decisions to appeal, has shown itself willing to do so in the case of sanctionless contempt findings, when officials’ reputation is the only thing at stake. And not just the Justice Department — when Interior Secretary Gale Norton was held in contempt by Judge Royce Lamberth in 2002, she sought to intervene in the appeal in her personal capacity, retaining a private attorney in addition to the Justice Department attorneys who were representing her officially. Another example involves the ACLU’s FOIA suit against various agencies to obtain disclosure of records of interrogations. After the CIA destroyed interrogation videotapes, the ACLU and Justice Department went to the mat regarding whether to find the CIA in contempt even though all agreed that no material sanctions turned on the finding. The battle, as the judge put it, was over “a pejorative term” and the opportunity to “write a newspaper headline.”
But while the judiciary relies mainly on the shaming and reputation-harming powers of contempt findings — and on the underlying norm of official compliance with judicial orders that those powers reflect — judges have nonetheless left the door open to contempt sanctions, hardly ever forswearing them in principle. Therefore, if some future official were to engage in clear enough disobedience to overcome the prudential obstacles to imposing sanctions, and if the official were shameless enough to be undeterred by a contempt finding in itself, sanctions could be wielded to try to force that official to comply. Of course, that kind of official behavior might reflect a more general breakdown in American political norms, and it is possible that a contempt sanction — by attracting publicity for official lawbreaking and thereby undermining the widespread perception that officials simply don’t disobey court orders — could aggravate the process of norm breakdown.
[Editor’s Note: You may also be interested in, Ambassador (ret.) Keith Harper’s “The Importance of Judicial Contempt Proceedings in a Trump Era”]
Image: Michael Zwahlen / EyeEm – Getty