How dark is dark? Bright lights, big city, racial profiling

Grogger and Ridgeway (2006) use the daylight saving time shift to develop a police racial profiling test that is based on differences in driver race visibility and (hence) the race distribution of traffic stops across daylight and darkness. However, urban environments may be well lit at night, eroding the power of their test. We refine their test using streetlight location data in Syracuse, New York, and the results change in the direction of finding profiling of black drivers. Our preferred specification suggests that the odds of a black driver being stopped (relative to nonblack drivers) increase 15% in daylight compared to darkness.