By: Ernesto Tiburcio and Kara Ross Camarena
The following blog post is a summary of the research that co-won the Midwest Political Science Association’s Best Paper in Political Behavior Award for research presented at the 2023 MPSA Annual Conference. The award recognizes the best paper on the topic of political behavior.
We study the political, economic, and social effects of unauthorized migration at the county level in the United States. We find that increases in unauthorized migration shift voters and policy to the right—in favor of Republican candidates and more conservative local expenditure. Job and wage loss play a minimal role in explaining the conservative response of voters and policy. Meanwhile changes in socioeconomic status, out-migration, racial bias, and more general out-group bias help to explain the conservative response.
We leverage preexisting migrant networks and conditions in Mexico to identify the causal impact of unauthorized Mexican arrivals in the United States. We use a shift-share instrument, a method used to study historical migration across countries and dynamics in the labor market. Our application of this identification strategy is the first to study sub-national variation in migration in the US. We show the shift-share identification strategy is robust to recent critiques to the method.
Our strongest findings are on the impact on voting behavior and local fiscal policy. Larger increases in unauthorized Mexican migration shift voting to the right. This evidence for the conservative shift is strongest in House elections. In close House contests, it is plausible migration could alter the outcome. Larger migrant flows also shift policy toward more conservative platforms. More migrant arrivals decrease local spending on public education and increase relative spending on policing and the judiciary. With the spending on policing, we looked for changes in crime, but did not find any.
We also identify evidence of socioeconomic effects that may help to explain the change in voter behavior and local policy. Impacts on formal job loss and wages from unauthorized migration are minimal, a finding consistent with studies on refugees and low skilled migrants. We find no impact on jobs, wages, or unemployment in the labor market as a whole. We find some evidence of formal job loss in construction and in hospitality and leisure. In some places, job loss in these sectors is offset by gains in manufacturing, a sector where work authorization is often necessary.
The labor market consequences of unauthorized migration, while statistically significant, are small. They likely do not explain what seem like much larger political impacts. Instead evidence suggests that voting behavior is changing due to social and demographic impacts of unauthorized migration. Unauthorized migrant arrivals drive out-migration, differentially among Whites, increase poverty (while having little impact on GDP), and increase communal values, an indicator of lower support for universal welfare policy. These social and demographic changes help explain why voters shift to the right.
The change in voting behavior is surprisingly consistent across political and demographic characteristics of counties. We observe the shift to the right in swing counties, Republican strongholds, and Democratic strongholds. Neither the average age nor the prior racial and ethnic composition of the county influence the change in voting behavior. Only the size and scope of a county’s social safety net seems to alter the impact of unauthorized migration. We find suggestive evidence that counties with more equitable taxation structures and more social welfare use see muted impacts of unauthorized migrant arrivals.
Our work lays a foundation for future work on understanding local political responses to unauthorized migration. Scholars can leverage our novel application of the shift-share identification strategy. Future research can explore a wider range of social and political outcomes or examine the impact of recent arrivals from the Northern Triangle and South America.