Police serve more than residents: Examining police enforcement compared to photo radar cameras and residential population estimates

Background: Residential population benchmarking is often used for identifying racial disparities in police enforcement. The present study constructs an at-risk estimate from non-discretionary photo radar violations to compare against residential population estimates to identify how racial disparities conclusions may vary.Methods: Automated red-light camera photos (n=3,542, 7.8%) from Boulder, Colorado were randomly sampled and coded to contrast with police tickets, police stops, and residential population estimates in 2024.Results: About 40% of crime suspects’ addresses, and 78% of red-light violators’ vehicle registrations are not within Boulder, despite their physical presence in the city. The at-risk population is more diverse demographically than the residential population. Using White (Non-Hispanic) as the reference category, risk ratios were lower for nearly every race category across both tickets and stops when using the alternative benchmark.Discussion: Like other studies, using a non-residential population benchmark yielded different conclusions about racial disparities in police enforcement. With the at-risk population, for most categories of individuals, the relative risks in Boulder of a police stop or traffic ticket are lower than the risks experienced by non-Hispanic White individuals.