Washington's National Guard "emergency" now has an end date: January 2029

On Oct. 14, judges on the U.S. Court of Appeals for the D.C. Circuit heard arguments over whether the National Guard troops patrolling Washington should finally go home, more than 14 months after they arrived. While that question sat with the court, the Pentagon had already answered it in practice: officials confirmed to NPR that the mission will run through Inauguration Day 2029, the last possible day of this presidential term, unless the president decides otherwise. That is not an emergency timeline. It is a lease.

Elizabeth Goitein, who directs the Brennan Center's Liberty and National Security Program, put it plainly when the extension became public.

"Announcing that an emergency will last for another 2 1/2 years means it's not an emergency."

The argument of this column is narrower than the broader debate over whether the deployment has cut crime, though that evidence matters too and is thinner than the White House claims. The deeper problem is structural: an occupation with a published end date tied to an election term, not to any measurable condition on the ground, has stopped being emergency response and become a standing feature of how the federal government governs the one American city it also directly controls. The courts reviewing it keep calling their own rulings "preliminary." At some point that stops being caution and starts being abdication.

Thirty days, then indefinitely

The occasion for all of this was a mugging. In early August 2025, a group of teenagers attempted to carjack Edward Coristine, a 19-year-old former Department of Government Efficiency staffer, in Washington's Logan Circle neighborhood, beating him in the process. President Trump cited the attack within days when he declared a "public safety emergency" in the capital on Aug. 11, 2025, invoking the Home Rule Act to direct the mayor to place the Metropolitan Police Department under federal control and mobilizing several hundred Guard members. Two weeks later, on Aug. 25, the White House issued a follow-up executive order creating a standing, specialized D.C. Guard unit and a nationwide "quick reaction force." Tellingly, that order set no end date and cited no statute beyond the president's general "constitutional obligation to restore order."

The formal emergency declaration governing the police takeover lapsed within roughly a month, as D.C.'s limited home-rule law contemplates. The troops did not leave. They were folded into what the administration now calls the D.C. Safe and Beautiful Task Force, and the numbers grew rather than shrank: a "summer surge" in 2026 pushed the Guard presence toward a peak of nearly 5,000 members from more than 20 states, timed in part to the country's 250th-anniversary celebrations. By the Oct. 14 hearing, about 2,900 remained, housed in part through a federal contract worth $292 million to lease apartments for Joint Task Force-D.C. personnel. The Congressional Budget Office has priced the operation at roughly $3 million a day; a Pentagon comptroller nominee told Sen. Elizabeth Warren's office this month that keeping troops in place through January 2029 will cost taxpayers another $1.4 billion on top of what has already been spent. Warren's response was blunt: "The National Guard are not pawns, and taxpayer dollars are not a piggy bank for Trump's political stunts."

A ruling that keeps calling itself temporary

The legal case has tracked the occupation's drift from emergency to fixture. On Nov. 20, 2025, U.S. District Judge Jia Cobb ruled that while the president may act to protect federal functions and property, he cannot unilaterally deploy the D.C. Guard for ordinary crime control or summon Guard units from other states to do it. She stayed her own order for 21 days to let the administration appeal. A three-judge D.C. Circuit panel obliged: after an administrative pause on Dec. 4, the same panel of Judges Patricia Millett, Gregory Katsas and Neomi Rao issued a unanimous, 32-page stay on Dec. 17 that let the deployment continue, reasoning that the District "has not identified any ongoing injury to its statutory interests" and that the government was likely to prevail on the merits. Judge Millett was careful to note the ruling was "preliminary," decided on an expedited record, and would not bind the panel that eventually hears the case on the merits.

That panel finally sat on Oct. 14 — roughly eleven months after Cobb's injunction and more than a year after the Guard arrived. Whichever side loses is expected to seek Supreme Court review, a court that has generally deferred to the administration in similar disputes. Each stage of this process has been framed as provisional, a placeholder pending a more considered ruling. But the practical effect of a year-plus of "preliminary" findings is that the occupation itself became the baseline condition, outlasting any single judge's capacity to say, with finality, whether it was ever lawful. The District of Columbia, which has no voting senators and no vote in the House, has no political remedy for this beyond litigation; Mayor Muriel Bowser and the D.C. Council never requested the deployment and have said so repeatedly.

What the deployment can actually claim credit for

The administration's defense rests on crime statistics, and the headline numbers are real. According to the Metropolitan Police Department's own tally, homicides fell 32% in 2025 versus 2024, and violent crime overall fell 29%. Defense Secretary Pete Hegseth has credited the Guard with cutting crime by "staggering amounts" and called the mission "righteous and beautiful."

Two independent analyses complicate that story considerably. The nonpartisan Council on Criminal Justice, which tracks crime across dozens of major U.S. cities, found that Washington's decline from its 2023 peak closely tracks a national trend; homicides fell in nearly every large city it studied over the same period, and the council's own researchers say there is "no single or easy explanation" for the broader drop. More specifically, a Niskanen Center study released in June 2026 found the Guard's presence was associated with roughly a 24% drop in opportunistic property crime, concentrated in the high-visibility public spaces where troops were actually stationed, but no measurable effect on violent crime, which is overwhelmingly concentrated in neighborhoods the Guard rarely patrols. The same analysis noted that a Guard member costs the government nearly twice what a D.C. police officer does, without carrying full police powers.

• Deployment began: August 2025, roughly 800 troops, under a since-lapsed emergency declaration
• Peak size: nearly 5,000 troops from more than 20 states, summer 2026
• Current size and extension: about 2,900 troops, mission extended to January 2029
• Projected remaining cost: about $1.4 billion through January 2029, on top of roughly $223 million already spent through 2025

None of this means the deployment did nothing, and a fair column has to concede the administration's strongest point: deterrence at the margins is still deterrence, and a visible troop presence plausibly suppressed some property crime in tourist corridors, whatever else it did or didn't do. But that is a far cry from the sweeping claim used to justify an open-ended, multibillion-dollar commitment extending three and a half years past the "emergency" that supposedly required it. If the administration genuinely believed the Guard, rather than the broader national decline already under way, was driving the result, it would have set a metric for success and an exit ramp tied to it. It has set neither. The mission's end date is tied to the inauguration calendar, not to any crime statistic it claims to be chasing.

An emergency that outlived the thing it answered

Emergency powers are supposed to be self-liquidating: they exist to bridge a crisis, not to become a permanent arrangement that simply gets re-justified, year after year, by pointing back to the crisis that started it. A deployment that both sides now openly discuss in terms of a 2029 end date is no longer being defended as an emergency measure at all; it is being defended as a policy choice the public never voted on, imposed on residents of the one American jurisdiction with no senators and no voting member of Congress to object to it on their behalf. The courts' answer so far — repeated, carefully hedged findings that nothing has been decided yet — has let that policy choice entrench itself for more than a year before a single appellate panel even heard argument on whether it was lawful in the first place. A legal process that takes that long to ask the question will always arrive after the answer no longer matters. That is the actual scandal here, independent of whatever the next crime report says.