NCJ Number
199116
Journal
Justice Research and Policy Volume: 4 Dated: Fall 2002 Pages: 131-142
Date Published
2002
Length
12 pages
Annotation
This article assesses the strengths and weaknesses of using "outcome tests" to assess racial disparities in police practices.
Abstract
The basic concept of the "outcome test" is to analyze whether the outcomes, about which the decisionmaker cares, are systematically different for minorities and nonminorities. Because outcome testing is an especially useful tool for assessing allegations of racial profiling by police, this article focuses on "police search" outcome tests as an example of the strengths and weaknesses of this methodology. The ex-post probability that a police search will discover contraband or evidence of illegality is strong evidence of the average level of probable cause that police require before undertaking a search. A finding that minority searches are systematically less productive than white searches is thus evidence that police require less probable cause when searching minorities. Although the outcome test methodology has significant strengths, it also has limitations. Outcome tests can be underinclusive as tests of disparate treatment because they are not well structured to capture disparate racial treatment motivated by rational statistical inference (so-called statistical discrimination). Certain forms of the outcome test may also be overinclusive as a test of disparate treatment, particularly regarding problems of "infra-marginality" and "subgroup validity." Regarding the "infra-marginality" problem, outcome tests of disparate treatment may only measure the average outcome and not the outcomes associated with the marginal decision. Observing that the average search success rate for minorities was lower than for whites does not necessarily prove that the threshold (or marginal) expected success rate was lower for minorities than for whites. A second limitation on the use of outcome tests as evidence of disparate racial treatment concerns the "subgroup validity" problem. When a particular observable characteristic is valid for some races but not for others, it is possible that a decisionmaker who conditions decisions on this characteristic generally might induce racially disparate outcomes. A decisionmaker's unwillingness to engage in disparate racial treatment may induce just the racial disparities in outcomes that are generally a concern. The author discusses adequate responses to each of the noted limitations, concluding that outcome tests can provide credible evidence of racial profiling, especially when combined with other more traditional types of evidence that decisionmaking subjects minorities to an unjustified disparate impact. 12 references