Malicious Compliance: When Governments Obey Court Orders by Making Things Worse

The Pentagon issued a statement. It was cooperative in tone, measured in language, and it acknowledged the federal court ruling that its press access restrictions were unconstitutional. Senior U.S. District Judge Paul L. Friedman had found, on March 20, 2026, that the Defense Department's media policy violated both the First and the Fifth Amendments, that it was designed "to weed out disfavored journalists" and replace them with reporters "favorable to or spoon-fed by department leadership." The Pentagon expressed respect for the judicial process. And then, in what might be described as the same breath, it published a new set of rules governing media access to the building, rules that journalists and press freedom organizations immediately recognized as accomplishing the same thing the court had just struck down. The old rules barred reporters from moving freely; the new rules required them to obtain authorization from Pentagon staffers before entering the building. The old rules restricted who could enter; the new rules shuttered the Correspondents' Corridor where journalists had worked for decades and announced their relocation to an external annex on Pentagon grounds. The Pentagon said it was complying. The Pentagon Press Association called the changes "a clear violation of the letter and spirit" of the judge's ruling.

There is a term for this in workplaces and bureaucracies: malicious compliance. Following the rules so literally, so technically, so meticulously that the rules cease to serve their purpose. It is the employee who, told to document every task, produces a three-hundred-page daily log that no one can read. It is the contractor who builds exactly to the flawed specification, knowing the building will leak. But when a government does it to a court order, the stakes are rather different than a leaking roof.

A Very Old Trick

The deepest American example is not about press passes or the Pentagon. It is about schools.

When the Supreme Court ruled in Brown v. Board of Education in 1954 that racial segregation in public schools was unconstitutional, the decision left a question dangling: how should desegregation actually happen? The Court's follow-up ruling in 1955 offered the phrase "with all deliberate speed," which turned out to be an invitation that many Southern states accepted with enthusiasm, the kind of enthusiasm reserved for things one intends to do as slowly as possible.

Several states adopted what they called "freedom of choice" plans. The name itself was a small masterpiece. Who could oppose freedom? Who could argue against choice? Under these plans, students and parents could choose which school to attend, theoretically allowing Black students to enroll in formerly all-white schools. In practice, Black families who attempted to exercise this "freedom" faced harassment, economic retaliation, and the unmistakable message that their choice was unwelcome. In New Kent County, Virginia, no white student ever chose to attend the Black school, and only 15 percent of Black students enrolled in the white one. The plans produced almost no actual integration. They were designed not to.

Prince Edward County, Virginia chose a more direct route. Rather than integrate its schools, the county's Board of Supervisors refused to appropriate any money for the school system in 1959, forcing every public school in the county to close. They stayed closed for five years. White students attended private academies funded by state tuition grants and tax credits. Black students had nothing, and some went to live with relatives in other counties, some attended makeshift schools organized by civil rights groups, and some simply went without formal education for half a decade. The county was technically in compliance. It had not segregated anyone. It had simply stopped offering schools.

The Supreme Court finally addressed this pattern head-on in Green v. County School Board in 1968, ruling that "freedom of choice" plans that did not actually produce integration were constitutionally insufficient. The Court did something it had hesitated to do in Brown: it shifted from telling states what they could not do to telling them what they must do. Not merely stop discriminating, but take affirmative steps to dismantle the dual school system. The Court had learned that forbidding a behavior and ending it are two entirely different things.

The Grammar of Compliance

What makes malicious compliance so effective as a strategy is that it exploits a structural weakness in how law works. A court order is a text. It says certain specific things. And like any text, it has gaps between what it says and what it means.

A court can hold a party in contempt for failing to comply with its orders. But proving contempt becomes remarkably difficult when the party in question can point to the text of the order and demonstrate, line by line, that it has done exactly what was asked. The order said restore press access. Access has been restored. The order did not specify unescorted access. It did not specify access to the same building. It did not say the restored access must be functionally equivalent to what existed before.

Legal scholars draw a distinction between compliance and implementation. You can comply without implementing. You can satisfy the letter of a mandate without achieving any of its objectives. This distinction is not a loophole in some obscure regulation; it is a fundamental problem in how legal orders work. Every order is finite in its specificity. Every institution that wishes to resist has the entire universe of unspecified details to work with.

And here is where the asymmetry bites hardest: when a court strikes down one set of rules and the government responds with a new set that achieves the same end through different means, the burden shifts back to the plaintiffs. They must file a new challenge, demonstrate that the new rules violate the same principles, and do all of this while the new rules are already in effect, reshaping the landscape they are trying to protect.

The Environmental Detour

If this pattern existed only in the dramatic arena of desegregation, one might attribute it to the unique pressures of racial politics in America. But it does not.

Environmental regulation offers a quieter, less dramatic, but equally instructive set of examples. When the Clean Air Act's New Source Review provisions required power plants to install modern pollution controls before undertaking major modifications, utilities found a vocabulary of evasion. The Act never defined what counted as "routine maintenance," and that undefined term became a doorway. Major renovations became "routine maintenance." Projects that replaced entire boiler systems were classified as repairs rather than upgrades. The companies complied with the requirement to file certain paperwork while avoiding the trigger that would have required them to install expensive equipment. When the EPA launched a wave of enforcement actions in 1999, it faced the same problem courts face with malicious compliance everywhere: the companies could point to the specific regulatory text and argue, with considerable technical sophistication, that they had followed it.

Courts responded by issuing more specific orders. Regulators wrote more detailed rules. And regulated entities found new gaps in the new specificity. The pattern is sometimes called "creative compliance" in the regulatory literature, or "regulatory arbitrage" when it involves moving activities from one jurisdiction or category to another to avoid oversight. The language is clinical, almost admiring. But the effect is the same as Prince Edward County closing its schools: the rule exists, the behavior continues, and the people the rule was supposed to protect remain unprotected.

There is, however, a difference when governments do this rather than private companies. A company that creatively complies with a regulation is exploiting gaps that someone else wrote. A government that maliciously complies with a court order can write the replacement rules itself. It controls the vocabulary. It defines the categories. It decides what "access" means, what "escort" means, what "authorized personnel" means. The gamekeeper has become the poacher, and the poacher writes the hunting regulations.

Who Bears the Cost

Consider the economics of this game. The New York Times sued the Department of Defense. The case required lawyers, filings, court time, and the institutional commitment to see it through. The Times won. And then the Pentagon issued new rules, and the clock reset.

If the Times or another press organization wants to challenge the new rules, it must begin again with new filings, new arguments about standing and specificity, and new months of litigation. Meanwhile, the rules are in effect. Journalists who once walked the Pentagon's corridors freely now require authorization to enter the building outside of scheduled briefings. Sources who once spoke to reporters in passing now see those reporters in controlled settings only. The information environment has already changed, regardless of what a future court might say.

This asymmetry is not accidental. It is the mechanism through which malicious compliance operates. The institution playing the game has permanence, budgets, and in-house counsel on salary. The party challenging it must fund each round separately, must maintain organizational will across years, and must persuade its stakeholders that the fight is worth continuing.

The chilling effect extends well beyond the named parties. Other news organizations watch. Smaller outlets without the Times's legal resources calculate whether they can afford to challenge similar restrictions. The precedent is not just legal; it is psychological. The message is not "you will lose in court" but something more insidious: "even when you win, you will have to keep winning, indefinitely, and we can keep changing the rules each time."

The Architecture of Delay

At its core, malicious compliance is a strategy of delay. It transforms a single definitive ruling into an indefinite series of challenges. And delay, in matters that involve ongoing harm, is itself a form of damage.

Consider the timeline of desegregation. Brown was decided in 1954. The "freedom of choice" plans that hollowed out its mandate were not definitively rejected until Green in 1968. The busing orders that followed in Swann v. Charlotte-Mecklenburg in 1971 were themselves challenged and modified for decades. By the time courts achieved something approaching actual integration in many districts, a generation of students had passed through the system or been excluded from it entirely.

The Pentagon press access dispute is still in its early chapters, but the structure is recognizable. Defense Secretary Pete Hegseth imposed new reporting restrictions in 2025. On October 16, reporters from more than 30 major news organizations forfeited their press credentials rather than accept the new terms, a collective act of refusal that included outlets from ABC and Fox News to the Associated Press and NPR. The lawsuit followed. The ruling came down. Three days later, new restrictions arrived. The next step, almost certainly, is a new legal challenge. Then perhaps new rules again. The question is not whether the press will eventually prevail on the constitutional principle. The question is what happens to defense journalism during the years it takes to resolve each iteration.

Time is not neutral in these disputes. While the legal process runs its course, the facts on the ground shift. Journalists lose source relationships that took years to build. Institutional knowledge atrophies. The reporters who knew the building, who knew which office to visit for which story, who could read the body language of a colonel walking to a meeting, those reporters are now somewhere else, covering something else. Even if access is fully restored, the network of relationships that made that access meaningful does not regenerate automatically.

The Voting Rights Act's Section 5 preclearance provision was itself an acknowledgment that the cycle of discriminatory law, court challenge, new discriminatory law was unsustainable. Congress concluded that requiring advance approval was the only way to break the pattern of post-hoc compliance games. When the Supreme Court in Shelby County v. Holder in 2013 struck down the coverage formula that determined which jurisdictions needed preclearance, it rendered that protection inoperable, re-exposing the system to exactly the compliance-and-rewrite cycle that preclearance was designed to prevent.

When Courts Lose Patience

The judiciary is not powerless against malicious compliance, but its tools are blunt and their use is contested.

When a court concludes that an institution has exhausted its good faith, that repeated cycles of order, compliance, and evasion demonstrate a pattern rather than an honest effort, it can escalate its interventions dramatically. Structural injunctions allow courts to embed themselves in the operations of the institutions they oversee. Special masters and compliance monitors become permanent fixtures. Courts have taken over prison systems, school districts, and police departments, dictating staffing levels, training procedures, and operational details that would normally be left to administrators.

In Swann v. Charlotte-Mecklenburg in 1971, the Supreme Court unanimously authorized sweeping remedial powers including mandatory busing, a remedy far more intrusive than anything the Brown court had envisioned. The escalation was not the Court's first choice. It was a response to seventeen years of creative noncompliance.

These interventions work, in the sense that they force behavioral changes that voluntary compliance did not produce. But they come with costs. They generate accusations of judicial overreach. They strain the separation of powers. They require judges to become administrators, a role for which they have neither training nor democratic accountability. And they can persist for decades, creating their own institutional pathologies.

The question of whether the federal courts will eventually escalate their intervention in the Pentagon press access case depends on factors that are as much political as legal. How many cycles of order-and-evasion will the judiciary tolerate? How specific will courts be willing to make their orders? Will a court appoint a monitor to oversee press access at a military facility?

These are not hypothetical questions. They are the questions that emerge, reliably, at the end of every malicious compliance cycle. The institution bets that the court will not go that far. Sometimes it wins that bet.

The Quiet Part

What makes all of this worth examining is not the Pentagon's press rules, specifically. Press access policy is important, but it is also a single instance of a larger phenomenon. The thing worth examining is the gap itself, the structural vulnerability in democratic governance that malicious compliance exploits.

Democratic systems depend on an assumption that is almost never stated because stating it would reveal how fragile it is: that institutions will implement judicial decisions in good faith. Not merely comply with them, but implement them. That when a court says "restore access," the institution will understand this to mean "restore meaningful access" and act accordingly. The entire architecture of separated powers rests on this assumption of good faith, and good faith is precisely the thing that cannot be legislated.

Political scientists have a term for what happens when this assumption frays: norm erosion. The gradual degradation of unwritten expectations that make formal rules function. Norms are the lubricant of governance. They are the reason a system designed in the eighteenth century can operate in the twenty-first, not because the written rules anticipated every situation, but because the people operating within those rules shared an understanding of what the rules were for.

The Pentagon's response to the press access ruling arrives in a context where these norms are under visible, acknowledged strain across multiple domains. Courts issue orders and wonder whether they will be implemented. Legislatures pass laws and wonder whether the executive will enforce them. Citizens win legal victories and wonder whether anything will change.

None of this is unique to one administration or one party. The impulse to comply without obeying is as old as authority itself. But there are periods when the gap between compliance and obedience widens, when institutions test how much evasion the system will tolerate, when the unwritten rules that hold the written ones together begin to thin.

The Pentagon's statement expressed respect for the judicial process. Then it closed the Correspondents' Corridor. The words said yes. Everything around them said something else. And in the space between those two messages, in that carefully maintained gap between what was said and what was done, a question sits that no court order can answer: what happens to a system built on good faith when the faith runs out?