Troops in American cities, for emergencies that were not there.

The military is not supposed to police Americans.

Start with the baseline, because it is old and deliberate. The Posse Comitatus Act of 1878 makes it a crime to use the federal military to do civilian law enforcement, the arrests, patrols, and crowd control that police do, unless Congress has specifically authorized it. It reflects a founding-era conviction that soldiers policing citizens is a danger to liberty itself. (Brennan Center)

The National Guard is the hinge. Under state command (Title 32), the Guard answers to the governor and can assist with disasters and unrest. That is the normal mode, and it is not controversial. But when a president federalizes the Guard (Title 10), the troops become federal forces, and Posse Comitatus kicks in: they generally cannot police. The one major exception is the Insurrection Act, which a president can invoke to use the military for law enforcement in a true breakdown of order. It has not been invoked once in these deployments. (Brennan Center; Lawfare)

That detail is the legal spine of everything below. Without the Insurrection Act, the federalized troops sent into these cities were barred from policing. When they policed anyway, they crossed a line a court could name.

Every deployment began with a claim that did not hold up.

Each operation was sold with a specific, vivid description of a city in crisis. Set those descriptions next to the record, the crime data and the judges who reviewed the evidence, and the same gap opens every time. This page uses "false" and "not supported" rather than "lie," because intent is hard to prove. The quotes are verbatim. The records are sourced.

There is also a standing claim about who the threat is. In October 2024 Trump described "the enemy from within," the "radical left lunatics," and said the problem could be "handled" "by National Guard, or if really necessary, by the military." A year later, speaking to a gathering of generals at Quantico, he said the country should "use some of these dangerous cities as training grounds" for the military. The cities he named were run by his opponents. (PBS; Military Times)

The picture on television was not the picture on the ground.

A claim that a city is in chaos needs pictures of chaos. Where the streets did not supply them, some coverage and some official accounts reached for footage that was old, filmed somewhere else, or altered. A few documented examples, each debunked by reporting outside the outlet that ran it.

To be fair, some of the dramatic Los Angeles images were entirely real: protesters did set several Waymo robotaxis on fire, and that happened. The point is not that nothing occurred. It is that a contained, mostly peaceful set of protests was dressed up, with old film, borrowed film, and the occasional outright fake, into the war zone the deployments were sold as answering.

Why the governors said no, and why the courts agreed.

What made this a break from American practice was not troops on a street. It was federalizing a state's Guard over its governor's objection, something no president had done since 1965, when Lyndon Johnson did it to protect civil-rights marchers from a segregationist governor. In each city the elected leaders objected, sued, and won, at least on the law. The chips mark where each case ended up: Ruled unlawful / blocked Unlawful, but allowed to continue Governor invited

The dividing line was consent. Where Republican governors invited the Guard, as in Memphis and New Orleans, the troops stayed under state command, in support roles, and the courts did not strike those deployments down on these grounds. The cases that drew the lawsuits were specifically the ones imposed on states against their will. (NPR)

It was expensive, and it did almost nothing.

Set the law aside for a moment and ask the practical question the deployments were sold on: did they make anyone safer, and what did they cost?

In Washington, where the troops stayed longest, an independent study of the surge found no measurable effect on violent crime. It found a roughly 24 percent drop in opportunistic property crime in the busy commercial areas where troops stood as a visible deterrent, but the violence the emergency was declared over did not move, because it happens elsewhere. The Guard had no arrest power; the arrests that did rise came from a separate federal task force, not from the soldiers. Each guardsman cost about $607 per day, against $384 for a city police officer. (NPR)

Los Angeles is the clearest measure of usefulness. Roughly 4,700 troops were sent in at a cost of around $130 million, and for the most part they stood guard at two federal buildings. By one accounting they made a single temporary detainment in the first 40 days, of a man hurrying to an appointment, not a protester. Mayor Bass said the buildings they were guarding frankly did not need to be guarded. (CalMatters; reporting by The Intercept)

The troops were aimed at protest.

This is the part that should trouble anyone regardless of party. The cities where the Guard was federalized over objection, Los Angeles, Portland, Chicago, were not facing some collapse of order. They were sites of protest, mostly against immigration raids. Sending soldiers to confront people exercising their First Amendment rights is not a side effect here. It was the occasion.

The same judge warned where the logic leads: that reading the law the administration's way "would permit a president to create a perpetual police force comprised of state troops." A standing domestic army, pointed at whichever city is protesting, is not a hypothetical the courts invented. It is the power the administration claimed.

Researchers who study failing democracies recognize this.

The alarm here is not loose talk. The particular combination, declaring emergencies the facts do not support, turning the military on domestic protest, and overriding opposition-party local officials, is one that scholars of democratic decline say they have seen before, and they name cases.

The historian Ruth Ben-Ghiat, who studies authoritarian leaders, argues the deployments echo "the playbook followed by El Salvadoran President Nayib Bukele... and other despots." In December 2022 Bukele sent 10,000 soldiers to seal off the city of Soyapango, searching homes and stopping people in the street. The scale and the violence there were far greater than anything in the United States, and that difference is real, but the move is the same one: the army turned inward, on a city, in the name of public safety. A closer parallel for force aimed at an opposition-run city came in March 2025, when Turkey arrested the opposition mayor of Istanbul and met the resulting protests with tear gas and mass detentions. (Ben-Ghiat; NPR on Soyapango)

This is not only a few scholars' opinion. When Bright Line Watch surveyed several hundred political scientists, 77 percent rated the Los Angeles deployment an "extraordinary or serious threat to democracy." In its 2026 report, the V-Dem Institute, which measures the health of governments worldwide, downgraded the United States from a liberal democracy to an "electoral democracy" for the first time in more than 50 years, the country falling from the 20th to the 51st most democratic. (Bright Line Watch; V-Dem)

The sharpest contrast is with America's own history. Every prior time a president federalized a state's Guard over a governor's objection, it was to enforce rights against a defiant state: Eisenhower at Little Rock, Kennedy at Ole Miss, Johnson at Selma, troops sent to protect Black Americans and uphold federal civil-rights law. This is the mirror image: soldiers sent against protesters and against local governments, in cities run by the president's opponents.

Polling suggests most Americans were uneasy: a Quinnipiac survey found voters opposed sending the Guard into DC by 56 to 41 percent, and an NPR/Ipsos poll found a majority said the president had gone too far. (Quinnipiac; NPR/Ipsos)

What happened after the courts ruled.

By early 2026 the legal picture had resolved, mostly against the administration, city by city.

On December 31 Trump said he was dropping the push for troops in Chicago, Los Angeles, and Portland, but promised to "come back, perhaps in a much different and stronger form." Two weeks later he threatened to invoke the Insurrection Act over Minneapolis, then backed off the next day. That act, the one law that could have made domestic policing by these troops lawful, was never invoked in any of it. He had said earlier he would use it if "courts were holding us up, or governors or mayors were holding us up." (ABC News)

The strongest case for the deployments.

An honest account has to grant the strongest version of the other side. Here it is, point by point, and what each one leaves unanswered.

Granting the strongest case leaves the core intact. The deployments that drew the lawsuits were the ones imposed over a governor's objection, without invoking the Insurrection Act, against protest, for emergencies the courts found were not real. That is the set this page is about.