By Marcus Hart
Somebody in Washington needs to explain what “temporary” means anymore, because the National Guard deployment that started as an emergency response to what President Trump called a crime crisis in the nation’s capital is now scheduled to run through Inauguration Day 2029. Not the end of this year. Not the end of this term. The next one.
The Pentagon confirmed it this week: more than 5,000 troops remain stationed in D.C., with numbers spiking again around the July 4th holiday, and the mission stays live “unless the president deems otherwise.” That last clause is doing a lot of work. It means there is no built-in expiration, no legislative checkpoint, no moment where Congress or the D.C. government gets to say the emergency is over. The emergency ends when one man says it ends.
I want to be careful here, because this is exactly the kind of story that gets flattened into a partisan cage match the moment it leaves the runway. One side will tell you this is martial law creeping into an American city. The other will tell you anyone raising a question about it is soft on crime. Both of those framings let the actual policy question walk right out the door untouched: is this working, and what is it costing us to find out?
The number that should stop everyone
The Congressional Budget Office — nonpartisan, not exactly known for hot takes — puts the current cost of this deployment at more than $3 million a day. Run that out over the stretch already logged since the August 2025 executive order that started this, and you’re deep into hundreds of millions of dollars spent putting soldiers on street corners in a city where soldiers don’t have arrest powers, don’t have local law enforcement training, and answer to a chain of command that isn’t the D.C. police chief.
That’s not a talking point. That’s a budget line. And budget lines deserve the same scrutiny whether they’re funding a housing program or a military occupation of an American city, which — let’s use the accurate word — is what a National Guard deployment inside U.S. borders, aimed at civilians, without a natural disaster or insurrection driving it, actually is.
Does it work? The research says no
Here’s where this stops being a vibes argument. Two separate studies — including one released this week by the Center for American Progress — both found the National Guard’s presence in D.C. has had little to no measurable effect on violent crime. The drop in crime numbers that the administration points to as justification appears to be happening on a trajectory that was already underway before troops arrived, and would likely have continued regardless.
I’ve spent time around military deployments. I know what a mission with clear objectives and a measurable outcome looks like, because I’ve served in one. This isn’t that. This is a presence without a metric. Nobody in this administration has said “we will know the mission succeeded when X happens.” There’s no X. There’s just troops, indefinitely, in a city that didn’t ask for them and whose own elected leadership — Mayor and Council alike — has said clearly and repeatedly that they don’t want them.
That last part matters more than it’s getting credit for. Washington, D.C. is a majority-Black city that has never had full congressional representation and doesn’t control its own National Guard the way every state does. The D.C. Guard answers directly to the President, not a governor, which is precisely why this city was chosen as the test case for what an open-ended domestic deployment looks like. If this were happening to a state with full autonomy over its own Guard, this would already be a constitutional fight in federal court. In D.C., it’s just policy.
The precedent is the story
Set aside whether you think crime in D.C. warranted a federal response in August 2025 — reasonable people can disagree on that, and I’m not interested in relitigating it here. The real story in July 2026 is what gets normalized once a “temporary emergency” quietly becomes a fixture with a 2029 end date attached before anyone voted on it.
If this can happen in Washington, it can happen anywhere a president decides a city has a crime problem serious enough to justify troops — and it will happen with even less oversight in cities that, unlike D.C., at least have a governor who could theoretically say no. Milwaukee doesn’t get to opt out of that precedent just because it isn’t D.C. Neither does Chicago, Baltimore, or Memphis. Once “the president deems otherwise” becomes the only check on an urban military deployment, that check applies everywhere the next administration — of either party — decides it wants to apply it.
What accountability actually looks like here
This isn’t a call to abolish the National Guard’s domestic role — the Guard does real, necessary work in disasters, and I’d never diminish that. It’s a call for the basic accounting that should apply to any four-year, nine-figure use of federal troops on American soil: a published cost-benefit analysis, a legislative reauthorization requirement instead of unilateral extension, and an independent crime-impact review conducted by people with no political incentive to find what they’re looking for.
Congress has the power to demand all three. It hasn’t, because doing so would require members of the president’s own party to publicly question a policy branded as tough on crime, and would require the opposition to risk being painted as anti-public-safety. That’s the actual dysfunction here — not the troops themselves, but a legislative branch that would rather let an open-ended military presence sit in the capital for four years than take the political risk of asking whether it works.
I’d rather see the number: how many arrests, how many prosecutions, how much of the reported crime decline is attributable to the Guard versus everything else already trending downward before boots hit the ground. Give me that, and I’ll change my mind if the data says I should. What I won’t do is treat “trust me” as an acceptable answer for $3 million a day with no legislative reset button, in a city that never got a vote on any of it.