Score contribution per author:
α: calibrated so average coauthorship-adjusted count equals average raw count
The extent to which discrimination can explain racial wage gaps is one of the most divisive issues in the social sciences. Using a newly available data set, this paper develops a simple empirical test that, under plausible (but not innocuous) conditions, provides a lower bound on the extent of discrimination in the labor market. Taken at face value, our estimates imply that differential treatment accounts for at least one-third of the black-white wage gap. We argue that the patterns in our data are most naturally rationalized through a search-matching model in which employers statistically discriminate on the basis of race when hiring unemployed workers but learn about their marginal product over time.