Economic Competition and Police-caused Killings

Blacks, Latinx, and American Indians are killed by police at a disproportionately higher rate than Whites and Asians, but whether racial discrimination accounts for these killings remains disputed. We contribute to this debate by examining structural conditions in U.S. metropolitan areas that are associated with the expected count of police-caused killings. Using an economic competition model, we find that the size of the metropolitan Black population (relative to the White population) positively predicts the expected count of police-caused killings for Blacks. Moreover, the size of the Latinx population (relative to Whites) predicts the expected count of police-caused killings of Latinx civilians. Furthermore, we find that metropolitan areas with more mixed-race neighborhoods experience higher expected counts of police-caused killings, specifically, for all, Black and White civilians. Finally, we find that overall population size also predicts the expected number of people killed by the police but violent crime does not, calling into question accounts that deaths are a function of crime. Our findings suggest, first, that the underlying conditions that lead to the deaths of Black and Latinx people at the hands of police are different than police-caused deaths of people of other races. Second, in developing solutions to the serious social problem of police-caused deaths, we need to look beyond the proximal causes of these deaths (i.e., the police) to the distal factors operating at the metropolitan level that promote White supremacy.

Automated summary

This study finds that in U.S. metropolitan areas, the expected count of police-caused killings for Blacks, Latinx, and American Indians is predicted by the size of their respective populations relative to Whites, indicating a possible link between racial proportions and police-related deaths. Additionally, mixed-race neighborhoods are more prone to experiencing police-caused killings, and the overall population size correlates with the expected number of deaths by police, challenging the assumption that crime rates determine police-related deaths.