The impact of changing ambient population composition on crime during COVID-19 lockdowns

Governments around the world implemented social distancing measures and lockdowns to limit people's movement and stem the spread of the global COVID-19 pandemic. These restrictions have changed the ambient population and altered its racial composition. By analyzing trips between census block groups using data from SafeGraph, we calculate the ambient population and infer its racial makeup during pre-lockdown, lockdown, early-post lockdown, and late-post lockdown periods in Cincinnati, Ohio. We examine the relationship between the ambient population-based racial heterogeneity (H) index and assault, robbery, and theft across the four periods. Our findings indicate that the lockdown affected mobility differently across racial groups. Additionally, we observe a stable, statistically significant influence of the ambient population-based H index on street crimes, in contrast to traditional census-based and spatial lagged measurements. This study demonstrates the effectiveness of the ambient population-based H index in explaining street crimes, particularly when people's routine interactions are significantly altered. It also contributes to theories of social organization, crime mobility, and routine activities. • Estimated racial/ethnic composition and revealed mobility disparities during COVID-19 lockdowns. • Compared census, spatial lagged, and ambient population-based racial heterogeneity measures in relation to street crimes. • Demonstrated the effectiveness of the ambient population-based approach in modeling crime. • Discussed broader contributions and implications of the mobility-informed estimation for crime and urban studies.