How do policy responses (deportation, sanctuary policies) affect reported crime statistics for undocumented immigrants?
Executive summary
A broad body of empirical research finds that aggressive deportation programs and local “sanctuary” policies have had surprisingly small or nil effects on measured violent-crime rates: ramped-up removals did not lower violent crime, and sanctuary policies did not raise it—and in some analyses are associated with safer communities or more reporting of crimes [1] [2] [3]. That pattern reflects both the underlying behavior of undocumented populations and the way policing, reporting, and data linkages change when enforcement priorities shift [4] [5].
1. Deportation campaigns increase removals but rarely move violent‑crime statistics
Large federal‑local enforcement efforts such as Secure Communities substantially increased deportations of people with criminal records but produced no detectable decline in violent crime across counties, a finding that recurs in multiple comprehensive studies [1] [6]. Migration Policy Institute and other reviews conclude that expanded removals have not produced the promised public‑safety gains—indeed, large‑scale deportations often do not translate into lower crime rates and can have damaging social and economic side effects [3] [7]. One explanation emphasized in the literature is selection: many undocumented immigrants are less likely to commit crimes than native‑born residents because of strong incentives to avoid contact with police [1] [4].
2. Sanctuary policies do not produce spikes in crime—and can change what gets measured
Contrary to political rhetoric, the best‑known multi‑jurisdictional analyses find no evidence that sanctuary policies increase crime; some studies even show lower crime or strengthened negative correlation between immigrant concentration and violence in sanctuary jurisdictions [2] [8] [9]. Crucially, sanctuary policies shift incentives: by limiting cooperation with federal immigration enforcement, they can reduce fear of deportation and increase victims’ and witnesses’ willingness to report crimes, which may temporarily raise recorded incident counts without reflecting an underlying rise in offending [5] [10]. Researchers warn that interpreting raw UCR/FBI figures without accounting for reporting behavior and policing practices can mislead policymakers.
3. Measurement, data linkages and reporting behavior complicate causal inference
How policy affects “reported crime” is a mix of true changes in offending and changes in detection, reporting and classification; when police fingerprint arrestees and share data with ICE, enforcement regimes alter policing incentives and community cooperation, which feeds back into official statistics [2] [8]. Studies exploiting natural experiments—DACA eligibility, sanctuary adoption dates, or Secure Communities rollouts—show that legalization or sanctuary-like protections tend to increase reporting and cooperation with police, while enforcement actions can suppress reporting among immigrant victims and witnesses, biasing crime trends downward even if safety worsens [5] [10].
4. Mechanisms that explain null or counterintuitive aggregate effects
Multiple mechanisms explain why deportations don’t neatly reduce crime: undocumented populations are heterogeneous and often less crime‑prone; enforcement may remove lower‑level offenders whose absence has little effect on violent‑crime trends; and enforcement can fracture households and neighborhoods, producing social disruption that counters any marginal isolation of criminal actors [1] [3]. Sanctuary policies, by contrast, can strengthen social ties, trust and police cooperation—factors long associated with safer neighborhoods—so the net effect on measured crime is neutral or positive from a public‑safety standpoint [8] [10].
5. Politics, selective evidence, and what the studies don’t settle
The policy debate is heavily politicized: advocacy and think‑tank pieces emphasize different slices of the evidence—some spotlight isolated high‑profile crimes to argue for stricter enforcement, while others catalogue broad empirical studies showing no aggregate crime increases in sanctuary locales [11] [12]. Existing research is robust on aggregate crime measures (violent and property crime) but less definitive about narrow categories, local short‑term displacements, or how enforcement affects specific victim populations; published work repeatedly cautions against overinterpreting raw arrest or federal‑arrest counts as measures of community safety [6] [3] [2].