Do Immigrants Commit More Crime?

In 2025 federal policy took as its stated premise that many undocumented immigrants are a threat to public safety. The Trump administration’s January 20, 2025 executive order “Protecting the American People Against Invasion” framed immigration enforcement as a public-safety priority.

The Laken Riley Act, signed January 29, 2025, requires the Department of Homeland Security to detain noncitizens who meet two conditions. They entered without permission, by fraud or without valid papers. And they are charged with, arrested for, convicted of or admit committing burglary, theft, larceny, shoplifting, assault on a law enforcement officer or any crime causing death or serious injury.

The opposing claim (that immigrants are less crime-prone than the native-born) is backed by most of the modern U.S. empirical literature.

Public perception is divided. The Pew Research Center found in a January 2024 survey that 57% of Americans said the large number of migrants seeking to enter the country leads to more crime. In the same survey, 39% said it has little impact on crime. But in the 2022 American Values Survey of the Public Religion Research Institute (PRRI), 61% said the statement that immigrants increase crime in local communities did not describe immigrants well.

Legal immigrants undergo background checks and admissibility screening that may select against criminality. Undocumented immigrants face deportation risk as an additional deterrent but lack systematic national tracking in most crime databases.

Data availability favors the split. Residual methods applied to American Community Survey and census data estimate the legal and undocumented populations, and Texas records status at arrest.

The most recent national estimate comes from Cato Institute analysis of 2024 American Community Survey data (Nowrasteh & Landgrave, Briefing Paper No. 198). The Cato Institute is a public policy research organization in Washington, D.C. Its analysis uses a modified residual method for ages 18–54. It counts immigrants as legal when their survey answers show citizenship, military service, arrival before 1982 or another sign of legal status, and the remaining foreign-born as undocumented:

The gap has narrowed lately. Cato's own series shows the undocumented rate at a low of 538 per 100,000 in 2022. It then rose 25% to the 674 of 2024. Even so, in every year from 2010 to 2024 it stayed between 31% and 56% below the native-born rate.

Cato itself notes that the American Community Survey counts people held in Immigration and Customs Enforcement (ICE) detention for immigration violations rather than for violent or property crimes. Removing the 37,684 people in ICE detention at the end of fiscal 2024 lowers the undocumented incarceration rate to 356 per 100,000. That is 17% above the legal-immigrant rate rather than more than double it.

The gap is not new. Butcher and Piehl, in National Bureau of Economic Research (NBER) Working Paper 13229, tracked men ages 18–40 across Census decades. Their samples held 1.9 to 2.0 million native-born men and 127,000 to 353,000 immigrants. In 1980, native-born institutionalization (mostly incarceration) was 1.35%, against 0.42% for immigrants. The gap widened by 2000, to 3.45% against 0.68%.

Abramitzky and colleagues (NBER/American Economic Review: Insights 2024) cover men ages 18–40 from 1870 to 2020. Their series finds immigrant men incarcerated at about the native-born rate in 1870, slightly below it from 1880 to 1950 and far below it since 1960. The gap has widened from 1960 to about 60% less likely today. Mexican and Central American immigrants were above the native-born rate before 1960. Adjusting for age, marital status, state and race leaves this raw gap mostly unchanged, and adjusting for education widens it.

Texas is uniquely informative. The Texas Department of Public Safety (DPS) sends every arrestee’s fingerprints to the Department of Homeland Security (DHS) under Secure Communities. Texas law requires every arresting agency to report arrests for class B misdemeanors and above to the DPS, which gives 96% compliance for 2011 to 2015. Light and colleagues (2020), in the Proceedings of the National Academy of Sciences (PNAS), say the fingerprint check is common practice elsewhere, but Texas alone keeps the status DHS returns. They analyzed all felony arrests 2012–2018:

Undocumented immigrants had the lowest felony arrest rates across every category shown. Legal immigrants were usually in the middle, with three exceptions. Their sexual assault rate was considerably higher than the native-born rate (31.2 against 18.2). So was their rate for felony traffic offences such as drunk driving (86.7 against 68.3). Their homicide rate was roughly equal (5.0 against 4.8).

Undocumented violent-crime rates would match the native-born rate only if the actual undocumented population in Texas were 45% of the current estimate. Property-crime rates would match only if that population were 23% of the estimate. Light and colleagues call these scenarios highly implausible.

The DPS also publishes running raw counts, the numbers that circulate in this debate. As of 31 July 2026, 335,000+ people classified as illegal noncitizens had been charged with 601,000+ offenses since June 2011. The charges included 1,132 for homicide and 7,691 for sexual assault. There were 229,000+ convictions, 625 of them for homicide.

These numbers are real. But the DPS itself says the figures do not attempt to allege that illegal noncitizens commit more crimes than other groups. They are counts without population denominators. When Light and colleagues and Cato applied denominators to the DPS data, both found per-capita rates lower for undocumented immigrants.

The broad U.S. literature that compares areas such as counties (the ecological literature) most often finds no association between immigration and local crime. Negative associations, where more immigrants go with less crime, outnumber positive ones. The Ousey and Kubrin meta-analysis of 51 studies from 1994 to 2014 found the average association was negative but very weak, with wide variation across studies.

The National Academies’ brief on its 2015 report found higher immigrant concentrations associated with less crime. The Federal Reserve Bank of Dallas (Dallas Fed) review concluded that the research overwhelmingly indicates immigrants are less likely than similar native-born Americans to commit violent and property crimes. It added that the limited research on undocumented immigrants points the same way, although possibly with a higher propensity than legal immigrants.

A county with more immigrants and less crime does not automatically prove each immigrant is individually less crime-prone (the ecological fallacy). That is why individual-level data (incarceration and arrest) belong alongside area-level data, not in place of them.

Download this conclusion: Texas

This section as one image: Texas’s arrest rates by status, and the running counts that circulate without a denominator.

Download the card (PNG, 1,080 × 1,350)

First-generation immigrants have far lower incarceration rates than U.S.-born men of the same origin, and the advantage shrinks by the second generation.

Butcher and Piehl’s evidence suggests that the widening of immigrants’ advantage from 1980 to 2000 is not the work of deportation. It suggests instead that migration selects people less crime-prone or more responsive to deterrence. Rumbaut and Ewing put the second generation’s rise down to assimilation into the family breakdown and addiction that raise crime among other native-born Americans.

Immigrants, especially undocumented immigrants, are disproportionately young working-age males, the peak offending demographic. This composition should raise their aggregate crime rates, yet they still show lower rates than the native-born. The American comparisons on this page already restrict to an age band or control for age; only the Texas rates are unadjusted.

Germany’s Federal Criminal Police Office classes asylum seekers, recognised refugees, people with tolerated status and people staying unlawfully as Zuwanderer. Its 2024 report on crime in the context of migration gives the group’s share of suspects. The group’s size cannot be reliably counted, so the report gives no rate. As an approximation of the group’s size, it gives the number of refugees living in Germany:

The group is not uniform. Ukrainians were over a third of the refugees and 12.8% of the group’s suspects. People from the Maghreb states were 0.5% of the refugees and 9.1% of the group’s suspects.

Figures from the Swedish National Council for Crime Prevention (Brå) cover residents aged 15 or over. Of those born abroad, 8.0% were registered as crime suspects in 2015 to 2018, against 3.2% of Swedish-born people with two Swedish-born parents. That is 2.5 times that group’s rate, higher for violent offences, and 1.8 times after adjustment for age, sex, education, income and municipality type.

In Greece, a study of the islands facing the Turkish coast found that a 1-percentage-point increase in the refugee share raised crime incidents by 1.7 to 2.5 percentage points. The rise was driven by crimes committed by refugees, with no change in crime by natives.

U.S. immigration combines family, labor, and screened refugee migration. Europe’s recent debates, by contrast, center on asylum seekers in legal limbo with weak work access. In Italy, Pinotti (American Economic Review, 2017) found that obtaining legal status cut the crime rate of legalized immigrants by 0.6 percentage points on a baseline of 1.1%. That is a cut of roughly half, which he puts down to blocked access to legal work.

Download this conclusion: Europe

This section as one image: what Germany, Sweden, Greece and Italy show, and why their figures describe a different system.

Download the card (PNG, 1,080 × 1,350)

• Undocumented offending outside Texas
• Offense categories across states
• Refugee-specific U.S. crime effects
• Recidivism after deportation for non-immigration crimes
• Whether police behavior changes arrest rates by status

A national dataset linking arrests, convictions and immigration status by state would matter most, and none exists.

No. Every American dataset with a denominator puts immigrants’ overall rates below the native-born. Cato’s estimates from the 2024 American Community Survey give 1,195 incarcerated per 100,000 native-born Americans aged 18 to 54. Undocumented immigrants are at 674 and legal immigrants at 303. The gap runs back to 1980 in Butcher and Piehl’s census work and to 1880 in Abramitzky and colleagues’ series.

Texas is the only state that keeps immigration status in its arrest records. There, from 2012 to 2018, Light and colleagues found undocumented immigrants with the lowest felony arrest rate in every category. For violent felonies, the undocumented rate was 96.2 arrests per 100,000, against 213 for the native-born.

Findings in Texas and in Europe cut the other way. In Texas, legal immigrants’ sexual assault arrest rate was above the native-born rate. The Center for Immigration Studies also reads the state’s conviction data as putting undocumented immigrants above the state average for homicide and sexual assault. Cato’s later data dispute that reading for homicide.

In Europe, Germany records asylum seekers and refugees, and Sweden the foreign-born, among crime suspects out of proportion to their numbers, while refugee arrivals raised crime on the Greek islands. Those figures describe a different system, and nothing here shows them carrying over to American immigration.

Download the conclusion

The verdict above as one image, with the incarceration rates by immigration status as its chart. It is 1,080 by 1,350 pixels, and anyone may download and post it.

Download the card (PNG, 1,080 × 1,350)