Advocates of the "outcome test" assume that if it can be shown that police stops/searches of drivers/pedestrians yield a disproportionately low rate of evidence "hits" in the resulting searches, then this is evidence that the police have made the stop for other than valid probable cause reasons, most likely because of the race/ethnicity of the suspect. This assumption fails because the outcome test has no way of controlling for all of the legitimate factors that might cause minority drivers to be searched at a higher rate than White drivers. It is therefore difficult to determine the true motivation behind a stop/search, regardless of whether evidence of a crime is found in the search. The inability of the outcome test (or any other statistical test currently used on traffic stop data) to determine the motivation of individual officers is related to Tomaskovic-Devey et al.'s (2004) description of the various types of racial bias. General equilibrium models are unable to distinguish between prejudices that motivate disproportionate stop/search rates based on various driver characteristics. For example, if police are stopping and searching higher proportions of minorities with low hit rates, it could be due to holding a stereotypical view that minorities are more likely to carry contraband rather than to racial bias. In fact, this belief may be true if measured under certain conditions, but untrue if measured under other conditions. 1 table and 28 references