UC Irvine

Research on the immigration-crime nexus has grown substantially, yet few studies differentiate among immigrants of different legal statuses, including undocumented immigrants. Of those that do, analyses are conducted at state and metropolitan area levels, leaving unanswered the question of how the presence of undocumented immigrants is associated with neighborhood crime rates—our focus. To construct neighborhood-level estimates of the undocumented population, we access restricted-use individual-level Census data, where tract of residence can be identified, and apply an imputation procedure. We combine these estimates with a dataset containing crime information for roughly 11,500 neighborhoods across U.S. cities over a decade. Controlling for change in foreign-born citizens and lawful permanent residents as well as other correlates of community crime rates, we estimate neighborhood-level fixed effects models to determine how change in undocumented residents is associated with change in violent and property crime levels between 2010 and 2018. We find that neighborhoods that experienced greater increases in the share of undocumented residents saw greater reductions in property crime and no detectable change in violent crime. We also document distinct patterns in which increasing undocumented residents is associated with decreasing aggravated assaults but increasing robberies, consistent with a possible victimization explanation.