Essays in Labor Economics: Effects of Immigration Policy on Vulnerable Populations

This dissertation studies three questions in labor economics centering around immigration policy and its effects on vulnerable populations. I use administrative data from the United States (U.S.) and Germany to address these questions empirically.

In the first chapter, I investigate the effects of an increase in immigration enforcement through local police force on domestic violence victims’ help-seeking behavior. I use a fixed effects model, a generalized synthetic control method and novel administrative data on the use of services for domestic violence victims in North Carolina. In North Carolina, a large fraction of individuals of Hispanic origin are undocumented immigrants or connected to undocumented immigrants. I find that local immigration enforcement significantly reduces Hispanics’ use of domestic violence services. It does not affect service use by African Americans who are predominantly U.S.-born citizens. This suggests that the decrease in Hispanics' service use is directly related to their immigration status and not driven by a general effect on minorities. I do not find any robust evidence that local immigration enforcement affects intimate partner homicides of Hispanic women.

The second chapter studies the relationship between workforce demographics, workplace hazard, and worker complaints about hazardous or illegal working conditions. In joint work with Matthew Johnson, I examine if worker complaints are effective in directing inspections by the U.S. Occupational Safety and Health Administration to the most hazardous workplaces. We measure worker complaints and workplace injuries by inspections triggered by a worker complaint and a serious workplace injury, respectively. We find that the complaint rate in a county and industry is positively associated with its recent injury rate. This relationship changes for workplaces with a high share of Hispanic workers. Workplaces that employ larger shares of Hispanic workers have lower complaint rates, but higher injury rates. We use fixed effects regressions to estimate the effect of local immigration enforcement on Hispanic workers' willingness to complain about hazardous conditions. We find that stronger enforcement significantly reduces worker complaints, but not workplace injuries in workplaces with a high share of Hispanic workers. This provides evidence that stronger local immigration enforcement reduces Hispanic workers' willingness to complain about unsafe working conditions irrespective of the true workplace hazard.

In the third chapter, I investigate the effect of integrating refugee students in elementary schools on the academic performance of native students. Using administrative data from Germany, I exploit the variation in the percentage of refugee students within schools to account for endogenous sorting of refugees into schools. I do not find any evidence for negative effects of refugee students' integration on the academic performance of native elementary students. In contrast, exposure to refugee students reduces mandatory grade retention rates of German fourth graders. Effects on the percentage of students who receive a recommendation for the higher secondary track are very small and statistically insignificant. I also show that refugee students attend schools where German students' performance is lower.