Title of presentation: Effects of Immigration Enforcement on Hispanic Mobility and Avoidance Behavior: Evidence from Secure Communities and 287(g) Agreements
Abstract:
Over the last fifteen years, the United States has substantially increased interior immigration enforcement, aiming to decrease undocumented immigration. This study asks how increased interior immigration enforcement affects the mobility of Hispanic people, as increased enforcement may cause Hispanics to attempt to avoid interaction with law enforcement. We focus on Secure Communities, a group of data sharing agreements between local law enforcement and US Immigration and Customs Enforcement (ICE), and 287(g) agreements, which deputize members of local law enforcement agencies to act as immigration agents under certain circumstances. Direct measures of mobility linked to ethnicity and documented status do not exist, so we look at traffic fatalities in the Fatality Analysis Reporting System (FARS) dataset. In these data, we can observe Hispanic ethnicity, some detail of place of origin (Mexican, Central/South American, Puerto Rican, and Cuban), driver’s license status (a proxy for documentation in much of the US), and detailed information on the time, location, and other circumstances around the universe of fatal traffic accidents in the United States from 2005-2020.
Using a two-way fixed effects, staggered rollout design, we find that 287(g) agreements decrease Hispanic traffic fatalities by 15.5%, with this effect entirely driven by a 29.2% reduction in fatalities for Mexicans. We find precise null effects for Cubans and Puerto Ricans, two ethnic groups who are not subject to deportation due to refugee status and/or US citizenship. In addition, we find a 17.7% decrease in fatalities for unlicensed drivers, suggesting large effects for undocumented immigrants.
For Secure Communities, we use an intensity-of-treatment design, taking the number of deportations under the Secure Communities program from the Transactional Records Access Clearinghouse data as a measure of the intensity of immigration enforcement in a county/month pair. We find that a 1 SD increase in deportations per county/year pair (about 7 deportations) results in a 14.6% decrease in Hispanic fatalities, with larger effects for Mexicans and null effects for Cubans and Puerto Ricans. Taken together, our findings demonstrate that increased immigration enforcement substantially decreases Hispanic mobility, with these effects caused almost entirely by groups at high risk of deportation when interacting with law enforcement.
Contact Mary Kate Batisitich for information
Originally published at economics.conductor.nd.edu.