Does multiple categorization reduce intergroup bias?

Two experiments explored whether crossing social category memberships can reduce intergroup bias. Experiment 1 provided a precise comparison between discrimination against single outgroups, partial outgroups, and double outgroups. Intergroup bias and perceived intergroup similarity followed an additive pattern such that partial outgroups were discriminated against as much as single outgroups, whereas both were discriminated against to a lesser extent than double outgroups. In Experiment 2, a more realistic form of crossing was employed whereby five additional dimensions of categorization were considered by participants instead of the traditional two. In line with a decategorization perspective, intergroup bias was reduced in both multiple group conditions relative to the single categorization (baseline) condition. Participants perceived a weakened intergroup structure and displayed a greater tendency to see outgroup members as individuals in multiple group conditions; however, only perceived intergroup structure mediated the pattern of intergroup bias. The implications of these findings for conceptualizations of crossed categorization are discussed.