Intersectional Discrimination and Candidate Selection: Evidence from a Lab Experiment in Kenya
Discrimination occurs when similarly qualified people are treated differently because of their identity. Discrimination is intersectional if discrimination on one dimension affects discrimination on another. The conditions under which intersectional discrimination arises, and the nature of intersectional discrimination when it arises, are not fully understood. We offer a formalization of when intersectional discrimination is likely to occur. We conducted a candidate selection lab experiment in Kenya to understand how three institutional conditions affect intersectional discrimination: (1) the presence of gender quotas; (2) objective information on candidate performance; (3) provisions for collective decision making in diverse groups. We find evidence of discrimination against both women and outgroups in candidate selection choices. We also find some evidence that patterns of discrimination are sensitive to these institutional conditions. However we find no strong evidence of intersectional discrimination in any of the eight institutional conditions that are formed from our experimental variations. Our results confirm the presence of multidimensional discrimination, but suggest limited gains from focusing on interactive aspects of identity in this setting.
