President Donald Trump is sending agents and the military storming into major U.S. cities like Los Angeles and Washington, D.C., while threatening other cities, like Chicago and New York, that they’re next. Trump administration officials claim a federal response is needed to fight crime, particularly in neighborhoods with larger immigrant populations. But new research suggests that Trump is looking for criminals in the wrong places, and it finds that immigrants bring less violence to their communities, not more.
Our new paper for the Cato Institute provides a new way to approach this problem: talk to the victims.
For decades, researchers have found evidence that immigrants are less likely to commit crimes. The Census Bureau surveys repeatedly showed that immigrants are dramatically underrepresented in jails and prisons compared with the U.S.-born. Surveys tracking immigrants over time found they were less likely to offend. Arrest and conviction data from the states supported this conclusion.
Despite all this evidence, skeptics focus on potential issues with surveying criminals or how states collect the data.
Our new paper for the Cato Institute provides a new way to approach this problem: talk to the victims. Criminals often victimize people in their neighborhoods, those they know and people of similar backgrounds.
If immigrants are committing lots of crimes, creating chaotic communities that need military intervention, we should find that immigrants are being victimized at higher rates. But according to the Census Bureau’s National Crime Victimization Survey, immigrants were 44% less likely than U.S.-born Americans to be victims of violent crimes from 2017 to 2023.
Confirming the hypothesis, the difference was even greater when we focus on people immigrants knew before the attack, who are obviously much likelier to be immigrants themselves. Immigrants were 64% less likely to be violently attacked by people they knew or had seen before and 65% less likely to be harmed by family members.
Noncitizens — half of whom are here illegally — were also 49% less likely to be violently victimized by people they knew or had seen before than U.S.-born citizens. Amazingly, when we look at prisons as recorded by the Census Bureau, we find immigrants overall were also 64% less likely to be incarcerated, and noncitizens 49% less likely, than the U.S.-born — a near exact match to their differences in victimization rates.
When we look at specific crimes, we find the same pattern. Immigrants are less likely to be victims of most types of violent crime, including sexual assault and aggravated assaults.
The one apparent exception was robberies, but it’s the exception that proves the rule. Robberies are the offense likeliest to be committed by people unknown to their victims. When we look at robberies committed by people known to their victims, the difference between immigrants and the U.S.-born becomes evident. Immigrants were half as likely to be robbed by people they knew or had seen before.
Not only are immigrants less likely to be victimized in general or by people they know — suggesting that they are less likely to be criminals — but the National Crime Victimization Survey also shows that before the current crackdown, they were helping police catch perpetrators at higher rates than the U.S.-born population.








