The Roberts Court is abdicating one of its most important roles, that of truth-teller.

This much was apparent on Monday in Trump v. California, when the Supreme Court’s Republican appointees let the Trump administration move forward with a plan to curb voting by mail. Via the Court’s shadow docket, the justices lifted an injunction against parts of an executive order that President Trump had issued to stop what he believes is the “massive cheating that’s gone on” in American elections, as he said at the Oval Office signing ceremony in March.

The Court’s order is troubling on many axes. It injects needless chaos into the midterm elections that are a little more than two months away, restoring, at least formally, one of the administration’s chosen tools for hampering a method that millions of Americans use to vote. It emboldens those already gunning for a more wholesale attack on democratic institutions, a base enthralled by the president’s election lies and White House officials’ willingness to use the machinery of the federal government to preserve their power.

The ruling is also troubling for what it says about the Court and its relationship to facts. Confronted with an imagined crisis of rampant voter fraud, the Court constructed an illusion of its own. It reimagined what the executive order did and glossed over the intentions of the man who signed it. This willingness to both ignore facts and conjure fantasies is not a one-off.

The mail-in-ballots case grew out of a conspiracy theory that Trump has long promoted: American elections are riddled with fraud on a grand scale, and voting by mail is a chief conduit. The March executive order was intended to target this supposed problem. Twenty-three states (along with the District of Columbia) sued, taking issue with three parts of the order.

Quinta Jurecic: What, exactly, is the Supreme Court thinking?

The first directed federal agencies to compile “state citizenship lists,” presumably as a precursor to purging voter-registration rolls. The second instructed the Postal Service to restrict delivery of mail-in ballots based on these makeshift lists. The third threatened prosecution of state officials who provided ballots to voters deemed ineligible.

The states’ lawsuit objected on both constitutional principle—especially the Constitution’s assignment of election administration to the states—and pragmatic grounds about the burdens of compliance with a muddled order that creates real problems by chasing phantom ones.

In late June, Judge Indira Talwani of the Federal District Court in Massachusetts agreed with them and issued a narrow, temporary block on parts of the order. Her opinion was not some dashed-off expression of her particular preferences. Rather, she undertook a granular evidentiary analysis—the trial court’s métier—scrutinizing how the order had already diverted election officials’ work assignments in Connecticut, affected ballot designs in Massachusetts, and collided with an approaching voter-registration deadline in New Mexico. A federal appeals court endorsed her factual analysis and declined to overturn her order.

The Trump administration then sought emergency intervention from the Supreme Court, which the conservative justices obliged this week. And yesterday, Judge Talwani lifted her injunction, saying the majority’s emergency opinion tied her hands.

Yet the majority’s opinion pays no mind to the undisputed facts. It contains nothing about what Judge Talwani probed, nothing about the burdens it found resulted from the executive order, nothing about a president with delusions of widespread fraud and ambitions of authoritarianism. Instead, the Court inhabits the antiseptic register of institutional injury, abstracting Trump into an ordinary president pursuing reasonable policies “regarding election integrity” and rendering the district court an overzealous judge frustrating the president’s good-faith-policy “goals” with a “string of speculations.”

The damage from the decision occurs at several levels. It includes the slow unraveling of constitutional structure when a rogue executive encroaches on territory assigned to others, in this case Congress and the states; the institutional signaling when the Court blesses that encroachment; the deepening tangle of the Court’s election-law jurisprudence, incoherent as doctrine but reliable in outcome, with the Republican appointees’ delivering on Republican interests; the diminishing of public trust in the Court; and the shunting of impositions and uncertainties onto the people who run elections and the people who vote in them.

For the states, these impositions and uncertainties are not theoretical. As Justice Ketanji Brown Jackson wrote in her dissent, the record and the district court established that Trump’s order had already placed an “incredible strain” on officials who must now spend time and money to “coordinate the implementation of the changes” created by the order on “compressed timelines” and with “limited resources” in a tense political climate.

Jackson launched a subtler and seemingly technical objection in a footnote. The strains on the states, she stressed, “by the way, are facts” found by a district court. Her italicized reference called attention to facts as they operate in law, and to the entitlement of such findings to “clear-error deference,” a procedural rule that bars a reviewing court from setting aside facts found by a lower court unless they are “clearly erroneous,” not merely contestable.

But the Court under Chief Justice John Roberts has made a practice of doing just that, evading facts found by district courts to facilitate ideological objectives, something I describe in a forthcoming article in the Michigan Law Review. This approach predates Trump 2.0, but it has been particularly pronounced in the second term, especially in shadow-docket orders reinstating his policies. We saw it last fall in Noem v. Vasquez Perdomo, in which a district court had documented the terrors and indignities of the immigration raids that the administration inflicted on Latino residents of Los Angeles. On the shadow docket, the Court’s Republican appointees green-lighted the raids, and Justice Brett Kavanaugh wrote a concurrence reimagining the sweeps’ brutalities by disregarding the clear-error standard and claiming that the prolonged and in some cases violent stops were “brief,” and that ICE agents “promptly” let people go. The order precipitated the spread of racialized ICE crackdowns across the country.

We saw it in McMahon v. New York, in which a district court had made concrete findings about the effects on vulnerable students from the administration’s gutting of the Department of Education. The Court, again on the shadow docket, and again along partisan lines, and again ignoring the district court’s factual work, cleared the dismantling—opening a pathway to further mass government firings and to the broader unmaking of the administrative state.

Adam Serwer: The see-no-evil Supreme Court

We saw it in Mullin v. Doe, in which a district court had documented Trump’s scurrilous racism toward Haitian immigrants that motivated the cancellation of Temporary Protected Status for 350,000 people. Granting a request from the Trump administration to hear the case early, the Court, along the same ideological lines, rewrote that plain animus out of the record to reinstate deportations.

What is lost when the Court disregards the facts that lower courts have established, and reasons instead from a record of its own, or Trump’s, choosing?

The traditional answer is that a Court that circumvents the messy factual texture of the world and that even declines to go through the motions of the clear-error standard is more likely to get things wrong. America’s constitutional and procedural order gives district courts special fact-finding tools and responsibilities. As a matter of this institutional design and historical custom, American law assumes that district judges are better positioned to find facts.

But there is a deeper reason why a society must care about courts, facts, and clear error: Reckoning with events, people, and actions grounds legal power and protects against the temptations of ideology. Disregarding facts and the procedural conventions that safeguard them does the opposite, risking the substitution of ideology for legal reasoning.

In California, the district court did the work of documenting facts about the world. It probed the evidence and took the world as it found it, unflinching about what the administration is and what it is doing. The highest court in the land ignored this, seeing the world as the justices wished it to be, not as it is.