How do crime statistics in the USA compare when broken down between illegal immigrants, legal non-citizens, and citizens?
Executive summary
Research that directly compares citizens, legal non‑citizens, and undocumented crime-rates">immigrants—most comprehensively the Texas study using state arrest records—finds that immigrants overall, and undocumented immigrants in particular, have lower felony arrest and incarceration rates than native‑born U.S. citizens (e.g., felony arrest rates ≈1,000/100,000 for U.S.‑born, ≈800/100,000 for legal immigrants, ≈400/100,000 for undocumented in Texas) [1] [2]. Multiple outlets and policy researchers have summarized these findings: undocumented immigrants are arrested at substantially lower rates for violent, drug, and property felonies compared with U.S.‑born citizens [3] [4].
1. The Texas data: a rare direct comparison and what it shows
A uniquely detailed dataset from the Texas Department of Public Safety that records arrestees’ immigration status allowed researchers to calculate comparable arrest rates for native‑born citizens, legal immigrants, and undocumented immigrants between 2012–2018 and repeatedly found the same ranking: U.S.‑born citizens highest, legal immigrants intermediate, and undocumented immigrants lowest; for example, felony arrest rates were about 1,000, 800, and 400 per 100,000 respectively [1] [2]. The study’s sensitivity analyses and replication by federal research arms have reinforced that undocumented arrest rates are less than half those of native‑born citizens for violent and drug crimes and much lower for property crimes [3] [4].
2. Broader reviews and longitudinal work that back the headline
National and historical analyses complement the Texas result: reviews and long‑run studies find immigrants are not more crime‑prone and in recent decades are substantially less likely to be incarcerated than the U.S.‑born—one long‑run analysis reported immigrants 60% less likely to be incarcerated than native‑born in recent periods [5] [6]. Policy organizations like the Migration Policy Institute and the American Immigration Council summarize the literature as showing lower offending and incarceration rates for immigrant populations overall [7] [8].
3. Why researchers still hedge: measurement limits and possible biases
Caveats accompany the consistent findings: most analyses rely on arrests or incarcerations as proxies for crime commission, which can reflect policing patterns as well as offending, and Texas is unusually well‑instrumented because it records immigration status at arrest—other states lack that link, complicating national generalization [2] [1]. Researchers note potential under‑reporting or nonresponse in victimization surveys and acknowledge that legal‑status measurement and population denominators require assumptions; still, multiple robustness checks in the Texas work and corroboration from other studies mitigate some concerns [9] [10].
4. What the statistics don't say and common misreads in public debate
Aggregate arrest and conviction gaps do not imply immigrants never commit serious crimes—federal criminal alien conviction tallies and individual high‑profile cases generate political attention—but those counts often reflect enforcement focus (immigration violations) rather than a higher propensity for violent offending, and they do not change the comparative rates found in population‑based studies [11] [4]. Advocates and some journalists warn that selective use of federal conviction or apprehension tallies can create a misleading “migrant crime wave” narrative contrary to the broader empirical record [6].
5. Bottom line and where evidence remains thin
On the best available empirical footing—state arrest records, national reviews, and long‑run incarceration studies—undocumented immigrants have lower arrest and incarceration rates than legal immigrants and native‑born citizens, with legal immigrants typically intermediate [1] [3] [5]. That conclusion is robust in Texas and consistent with multiple national summaries, but national extrapolation is constrained by data gaps outside a few jurisdictions and by the difference between arrests, convictions, and true offending, which future data collection would need to resolve more completely [2] [9].