Score contribution per author:
α: calibrated so average coauthorship-adjusted count equals average raw count
In Brazil, different employers report different racial classifications for the same worker. We use the variation in race across employers to estimate the relationship between race and wages. Workers whose reported race changes from nonwhite to white receive a wage increase; those who change from white to nonwhite realize a symmetric wage decrease. As much as 40 percent of the racial wage gap remains after controlling for all individual characteristics that do not change across jobs. We formally test, and reject, the hypothesis that our results are driven by misclassification. We also evaluate several mechanisms that could explain our findings.