Racial Profiling in the United States: A Review of the Literature

This project was undertaken by an external, independent researcher to explore, and provide information about, an issue or topic. Opinions expressed are those of the author and do not necessarily reflect those of the Royal Canadian Mounted Police or the Government of Canada.

Executive Summary

One of the most important public policy and research issues facing police executives today is the concern that race is being used as a major determining factor in police decisions to make a traffic stop and, in some cases, to subsequently search the vehicle and its occupants. This type of behaviour is, increasingly referred to as 'racial profiling' by the public. There is a concern about racial prejudice and/or discrimination by police officers and their supervisors. The term racial profiling was originally used to refer to the use of race as an explicit criterion in developing more accurate profiles of offenders that some police organizations used to guide the decision-making of their officers. Today, there is a highly publicized debate as to whether racial discrimination exists and, if so, to what extent.

This report reviews empirical studies conducted in the United States to evaluate police-citizen contacts during traffic stops. The thrust of these studies has been to determine if racial profiling (in the popular sense) exists within an agency or jurisdiction, by determining if minorities are disproportionately represented in traffic stops, searches and/or citations based on their population in the study jurisdiction. Many of these studies concluded that racial disparities exist in the aggregate rates of such stops. The issue with these conclusions is the mere presence of a disparity in the aggregate rate of stops does not, in itself, demonstrate racial prejudice or racial discrimination by police officers any more than the racial disparity in prison populations demonstrates racial prejudice and/or discrimination by judges. As a result, all studies investigating racial profiling focus much of their attention upon methodological concerns, most notably the proper denominator or base rate - that is, the expected rate of stops of a racial minority group by police officers. Different studies have used different base rates, but to date, the studies of racial profiling are split in their conclusions about disparity in the aggregate rate of stops being adequately explained by individual police officers racial prejudice and/or discrimination. However, since there is no general agreement concerning the appropriate base rate, all such conclusions are premature.

One issue with these studies involves the actual definition of racial profiling. The term has developed without a universally agreed-upon definition, although what it involves has become clearer through public debate, research studies and legislation. Generally speaking, it appears that any definition of racial profiling should have two elements-the first element typically makes a statement that individuals are being stopped as the result of their belonging to a class of individuals, and the second element notes, that the police must have some level of suspicion beyond race to stop them.

The bulk of the research literature in this area focuses upon pretextual traffic stops. Pretextual stops occur when a police officer, who is suspicious about a vehicle, stops it under the pretext of a traffic violation to conduct an investigation of matters that may not be related to the specific traffic offence. The United States Supreme Court, in Whren v. U.S. (1996), cleared the way for pretextual stops in the United States . In this case, the Supreme Court stated that regardless of a police officer's actual motivation, a police officer is legally permitted to conduct an investigation into other possible criminal activities once a traffic violation has occurred.

The studies reporting the existence of racial profiling as evidenced by the disproportionate representation of racial minority groups in traffic stops, searches and/or citations are based on three assumptions that may be false. The first assumption is that 'baseline' data (i.e., observations of the driving population) as opposed to 'benchmark' data (i.e., census-based data) are the best source of data for traffic stop comparisons, i.e., the denominator. While most researchers today prefer to use baseline data, this is a costly and time-consuming approach. In addition, many courts in the United States are now requesting jurisdiction-wide data on police activities, a request best served by collecting benchmark data.

The second assumption is that the rates of traffic violations and offending are the same for all racial groups. Some researchers argue that this is so, while others argue that groups violate traffic laws at different rates.

The third assumption is that law enforcement is equally distributed across space and time. In fact, most police agencies distribute their officers across a jurisdiction according to the time of shifts and/or by management strategies, such as calls for service.

The issues noted above, such as the selection of a benchmark or baseline as the denominator, demonstrate that there is no coherent theory to guide the data collection or to interpret the results. Specifically, almost all of the studies reviewed in this report fail to measure any explanatory factors beyond the simple aggregate rate of stops. This is a significant variation compared to other research efforts in police activities that have been conducted since the 1970s, many of which have studied police behavior. Studies of racial profiling have, in contrast, failed to explain how and why police officers make their decisions; for example, is it the result of bias on their part or a problem with policy?

While there are problems with research to date on 'racial profiling', this does not mean that police agencies should discount the importance of data collection in this area. There could be a small number of officers who do engage in such activities, but because all traffic stops are analyzed, their activities do not influence the aggregate outcomes. Some policies now exist that attempt to uncover whether or not such officers do exist through the use of early warning or intervention systems. In addition, because the type of data requested by the courts is not typically collected by the courts, it is important to engage in pilot projects to identify the appropriate data. Finally, and perhaps most emphatically, it is important for police agencies to study the perceptions of racial minority groups concerning their treatment by the police. This would involve how unfair they perceive their treatment to be, whether or not they believe this unfairness to be the result of a bias and how this may have influenced their perception of the police. In addition, the police should explore what needs to be done to increase citizens' confidence in the fairness of the police.