Local Immigration Enforcement and Crime in the United States: A Nationwide Analysis of the 287(g) Program

The 1996 Illegal Immigration Reform and Immigrant Responsibility Act (IIRIRA) created the 287(g) program, which authorizes Immigration and Customs Enforcement (ICE) to delegate immigration enforcement to local law enforcement agencies. Under the program, city police departments, county sheriffs' offices, and state law enforcement bodies voluntarily apply to the 287(g) program. If selected, ICE provides training in immigration enforcement practices to employees of the law enforcement agency, who are authorized to investigate individuals' immigration statuses and detain undocumented individuals prior to immigration proceedings. The earliest 287(g) agreements between ICE and local law enforcement agencies were signed in 2002, and the program continues to this day. Proponents of 287(g) agreements argue that participating law enforcement agencies identify and detain dangerous "criminal aliens," improving public safety and quality of life. Opponents claim that the 287(g) program undermines trust between immigrant communities and law enforcement, is costly for local law enforcement agencies, and contributes to the phenomenon of "crimmigration," in which criminal and immigration law increasingly overlap. In order to evaluate the effectiveness of 287(g) programs, I examine whether these programs reduce violent crime. In particular, I use propensity-score matching to pair counties which implemented 287(g) agreements with comparable counties that did not. I then use a differences-in-differences design to compare changes in county crime rates in the treatment and control groups. I find that after adjusting for potential confounding variables, 287(g) agreements have no effect on the trend in violent crime rates. I then repeat my analysis using only murder and nonnegligent manslaughter to assess whether 287(g) agreements cause underreporting of other violent crimes. I observe an increase in murders in 287(g) jurisdictions relative to other counties, suggesting that underreporting of other violent crimes may have occurred. This analysis suggests that 287(g) agreements may undercut public safety by damaging trust between immigrant communities and law enforcement. This study suggests that the evidence does not support the adoption of 287(g) agreements as a public safety measure and that more lenient policing practices towards immigrants may have a neutral or positive effect on public safety.