Using data from the U.S. Decennial Census and the National Longitudinal Surveys, we find little
evidence of differences in the return to schooling across racial and ethnic groups, even with attempts
to control for ability and measurement error biases. While our point estimates are relatively similar
across racial and ethnic groups, our conclusion is driven in part by relatively large standard errors.
That said, we find no evidence that returns to schooling are lower for African Americans or
Hispanics than for non-minorities. As a result, policies that increase education among the lowskilled
have a good possibility of increasing economic well-being and reducing inequality. More
generally, our analysis suggests further research is needed to better understand the nature of
measurement error and ability bias across subgroups in order to fully understand potential
heterogeneity in the return to schooling across the population.