Ethnic Discrimination in the Lab : Evidence of Statistical and Taste-Based Discrimination

Using a lab experiment which simulates a labor market and does not use deception, we investigate racial discrimination for employee selection. We find that discrimination against Blacks persists even when information about candidate abilities is known. The experiment design allows us to observe within-subject variation in discrimination based on different information available about candidates, which helps distinguish between statistical and taste-based discrimination. We find evidence of discrimination by Non-Blacks (mainly Whites) against Blacks whenever race is salient. Some discrimination against Blacks is statistical based on distrust of self-reported abilities. But discrimination by Non-Blacks against Blacks is taste-based as it exists even when accurate information about abilities is known and even when such discrimination is costly. This study is one of the few experimental studies that finds racial discrimination persists in the lab even with good information. We suggest that the institutional and social climate may shape the prevalence of this type of discrimination