Violent Crime and Political Participation

• In Event: Justice and Injustice: Political Science Perspectives on Crime and Punishment Mini-Conference: The Political Origins of State Violence

Abstract

Public safety is a central mandate of government, and theories of retrospective voting predict that
citizens may reward or punish incumbents when local crime improves or worsens. However,
prior research on the relationship between crime exposure and political participation has been
vulnerable to mismeasurement due to poor data quality and selection bias. In this paper, we
marshal evidence from highresolution
geocoded administrative data sets on violent crime,
gunfire, and individuallevel
voter turnout records in close to 20 US cities. We first report
descriptive patterns that shed light on which populations are most exposed to violent crime in
American cities. We exploit exogenous variation in exposure to violent crime close to election
day to test whether such exposure affects turnout. Using a regression discontinuity design, we
also test whether switching the party in power in local government has discernible effects on
crime rates.

Authors

• Jonathan Mummolo, Princeton University

• Christopher Lucas, Washington University in St. Louis