By: Brandon J. Romero, University of Michigan
The following blog post summarizes the research that was a co-winner of the Midwest Political Science Association’s Best Paper in Political Behavior Award for research presented at the 2024 MPSA Annual Conference.
In my paper “More COPS, Higher Turnout?” I investigate the causal effect of police on voter turnout across racial groups. I examine how police indirectly affect political participation through their effects on crime—an important but previously unexplored mechanism in political science literature. While extant studies have thoroughly documented how adversarial police contact can reduce civic engagement, I theorize that by reducing criminal violence, police may simultaneously enable greater political participation among those most vulnerable to victimization.
I leverage data from a natural experiment induced by the disbursement of the federal Community Oriented Policing Services (COPS) Hiring Program, which provided grants to municipal police agencies in 2009. Using an instrumented difference-in-differences approach, I analyze voter turnout data from 2004-2014 across more than 3,600 municipalities to estimate the causal effect of adding police officers on community-level political participation.
My findings reveal that hiring an additional police officer per 10,000 residents increases Black voter turnout by approximately 2 percentage points, thus narrowing the Black-White turnout gap. To identify the mechanism driving this effect, I use grant disbursement as an instrument for homicide rates and find that higher homicides significantly widen the Black-White turnout gap by reducing Black turnout. This provides evidence that police-induced reductions in violent crime translate to higher political engagement among Black voters.
Further analyses reveal important demographic heterogeneity in this effect. When examining municipalities closest to the funding cutoff (where treatment assignment is most plausibly random), the narrowing of the Black-White turnout gap is driven specifically by increased participation among Black men—the demographic group most at risk of criminal victimization. This finding is particularly significant given the disproportionate burden of violent crime borne by Black men, who are more than six times as likely to be murdered than their White counterparts according to CDC data.
My research complements policy feedback theories of criminal justice institutions by incorporating crime as a mediating factor in the relationship between policing and political behavior. Prior scholarship has documented how criminal justice contact can erode civic trust and capacity, but my findings suggest the protective function of police can simultaneously foster conditions conducive to political participation among those most affected by crime. Physical safety, like time, money, and civic skills, appears to be an important resource that potential voters draw upon when engaging in politics.
One limitation of my study is its focus on registered voters, who may have different experiences with police compared to non-registered individuals. It remains possible that justice-impacted individuals experience police presence differently from politically engaged citizens who benefit from crime reduction. Future research could explore this heterogeneity more explicitly.
These results help explain seemingly paradoxical public opinion data showing that many Black Americans simultaneously support police reform while opposing reductions in police presence in their neighborhoods. My findings indicate that both under-protection and over-policing can undermine democratic participation, with different subgroups within racial categories potentially experiencing these effects differently.