Current Research Notes

Paper: Undocumented Immigration & Crime in the United States: A Micro-Level Analysis

One common criticism of my paper on migrant crime in NYC is that the arrest rates potentially reflect both criminality and enforcement. Per a referee's recommendation, I have recently re-estimated the results using crime complaint rates in place of arrest rates. The drawback is a non-trivial number of missing observations in the race/ethnicity breakdowns, which are omitted from the analysis. The advantages, however, are that this approach isolates criminality from enforcement and distinguishes between suspects and victims.

Coupled with the paper's current findings, these results overwhelmingly reject an enforcement mechanism. The Hispanic suspect crime complaint rate explains 84% of the corresponding arrest rate. Also, as shown in the paper, Hispanic stop-and-frisks account for only 2.5% of the baseline effect on Hispanic arrests.

Contrary to most media reports, the NYC migrant housing crisis increased the city's Hispanic arrest rates due to rising criminality, not increased enforcement. To be clear, as explained in the paper, this an economic story: it is not because they are migrants, but because their lack of legal work authorization generated rising petit and grand larceny rates. This is a textbook Beckerian story that applies to natives and immigrants alike.

If these migrants had access to legal work opportunities, which could be achieved through general amnesty (among many potential policy prescriptions), the crime effect would vanish.

This is the first paper of its kind to study hyper-localized effects in the United States using variation in actual undocumented immigrant rates induced by a natural experiment: the southern border busing scheme.