Surveillance camera prevalence and racial diversity in ten US cities

Surveillance cameras are a prevalent form of monitoring in modern cities, with their placement reflecting neighborhood dynamics and social control. Here, using computer vision and human verification on Google Street View images across ten densely populated US cities, we assess how surveillance camera presence varies by neighborhood racial composition. Contrary to theories predicting high surveillance in predominantly Black neighborhoods, we find that cameras are most prevalent where racial diversity is high. Furthermore, over time, increases in racial diversity when white residents move in are associated with increases in cameras. These patterns, persisting after accounting for crime rates, suggest that surveillance cameras in diverse neighborhoods may be used to monitor and control minority groups by white householders, potentially reinforcing spatial inequality and eroding social trust. Our findings illustrate the value of computational methods and visual data in understanding spatial inequality and highlight the presence of surveillance cameras in diverse and diversifying neighborhoods.