In response to legislation, lawsuits, community pressure, and internal concerns, over 4,000 police departments are collecting traffic enforcement data to determine whether their officers racially profile. In this context, racial profiling is defined as when police inappropriately use a motorist’s race as a factor in deciding which motorists to stop, cite, search, and arrest, where the appropriateness of the use of race is defined by the law, consent decrees, and department policies. Most studies have estimated the use of race at the department level, but there is a growing interest to estimate its use at the officer level and incorporate the results into Early Intervention (EI) systems, which identify potential problem officers. For either department- or officer-level studies, the dominant concern is to employ methods that accurately estimate the use of race and distinguish between appropriate and inappropriate uses.

This study summarizes the key implications of incorporating racial profiling analyses into an EI system and improves upon existing methods that estimate the use of race in stop, search, and DUI arrest decisions at both the department and officer levels. The methods are applied to the traffic stops of 16 Washington State Patrol troopers who patrolled South Seattle during 2003 to 2005 in order to estimate the troopers’ average use of race as well as each trooper’s relative use of race as compared to his 15 peers.

For the combined trooper analyses, the study does not find conclusive evidence that the troopers used race as a factor in their stop, search, and DUI arrest decisions. However, the study does find that the use of race significantly varied among the troopers, and a few troopers are identified as potential problem troopers who warrant further scrutiny to determine if they inappropriately used race against minority motorists.

The results lead to two key policy recommendations. First, police departments should strongly consider incorporating traffic enforcement racial profiling analyses into their EI systems in order to identify potential problem officers. Second, methods used to identify potential problem officers need to allow for the possibility that an officer’s peer group may change over time and that non-racial motorist characteristics differentially affect officers’ law enforcement decisions. These types of methods should also be used to identify potential problem officers for other EI system performance indicators.