Racial Profiling, Statistical Discrimination, and the Effect of a Colorblind Policy on the Crime Rate

Using a model similar to labor market models of statistical discrimination, I de- scribe how and why racial profiling can arise even when law enforcement officers are racially unbiased. Specifically, if one racial group has a higher fraction of individuals who are at risk of committing the relevant type of crime than another, and if law enforcement officers can observe a noisy signal of guilt in addition to an individual's race, then it will be optimal for officers to treat observationally equivalent individu- als of different races differently. Moreover, this model can be used to show how the effect of a racially colorblind policy on the overall crime rate for a particular type of crime will depend on the racial make-up of the relevant jurisdiction, the relative proportions of each racial group that are at risk of choosing to commit that crime, the proportion of the relevant population that officers can observe, the magnitude of the punishment for that particular type of crime, and distribution of the benefits to committing that particular crime. The implications coming from this analysis are then applied and analyzed with respect to two specific contexts--highway patrol vehicle searches for drugs or weapons, and border patrol investigations of foreign entrants for terrorist connections.