Assessing the use of profiling in searches by law enforcement personnel

This article considers statistical means of evaluating profiling in searches of individuals by law enforcement agencies. It reviews examples of inappropriate methods that have been used to date. A relatively recent, more appropriate, and simple method is elaborated and its usefulness documented with data from Maryland State Police highway stops and searches. The method evaluates profiling for specified groups of interest by comparing the relative magnitudes of contraband find rates among those stopped and searched of each group. When find rates are approximately the same across groups, then profiling used in making the stops and searches is deemed effective and nondiscriminatory. With an increasing number of jurisdictions mandated to collect data on stops and searches, it is important that appropriate methods be used in their analyses so correct public policy conclusions can be drawn.