Pretextual Traffic Stops and Racial Disparities in their Use

Moving-violation traffic stops are pretextual when motivated by suspicion of unrelated crimes. Despite concerns that they are subject to racial bias, and recent reforms hoping to curb the practice; we lack empirical evidence to inform our understanding of pretextual stops. Using a decade's worth of traffic citation data from Louisville, KY, I provide evidence suggesting that pretextual stops predicated on a particular violation---failure to signal---were reasonably common. While arrest rates range from 0.01 to 0.09 in stops citing similarly common moving violations, stops citing failure-to-signal yield an arrest rate of 0.42. Importantly, pretext for a stop requires just one infraction. The arrest rate is 0.53 when failure-to-signal is the only cited traffic violation, and 0.21 otherwise. Prior to departmental deployment of body-worn cameras (BWCs), Black motorists account for a disproportionately high share of these likely pretextual stops (compared against observably similar conventional stops), but are arrested in them at significantly lower rates than other motorists. Both disparities are substantially larger during daylight, when driver race is more easily observed. The latter disparity dissipates following BWC deployment, which is found to initially reduce the frequency of these stops that fail to find contraband. Departmental prohibition of vehicle search based on a subject's nervousness was abruptly announced in May 2019, and immediately followed by a sharp 58% relative decrease in likely pretextual stop frequency.