Historical Slavery Predicts Contemporary Violent Crime

This study investigates the long-term relationship between slavery and violent crime in the USA. Although qualitative evidence suggests that slavery perpetuated violence, there has been no large-N study supporting this claim. Using county-level data, we find that the percentage of slaves in the population in 1860 is linked with violent crime in 2000. This result is specific to violent crime, robust to instrumenting for slavery and varying the approach to missing crime data, and not driven by biased crime reporting. Investigating the theoretical mechanisms driving these results, we find that historical slavery affects inequality (like Bertocchi and Dimico, 2014), white Americans’ political attitudes towards race (like Acharya et al., 2016b) and black American’s political attitudes – in opposite directions. Results suggest that inequality and black American’s political attitudes mediate the observed effect on violent crime in general, but that white American’s political attitudes mediate the effect on interracial violence.