Are OLS Estimates of the Return to Schooling Biased Downward? Another Look.

The authors examine evidence on bias in OLS estimates of the economic return to schooling. To study omitted-ability bias, they use test scores available in the National Longitudinal Survey of Youth as proxies for ability allowing for measurement error in these test scores. The authors also explore biases from the endogeneity of schooling or experience, or measurement error in these variables. In their data, OLS estimation including test scores appears to be appropriate and indicates an upward bias of roughly 40 percent in the OLS estimate ignoring ability. This contrasts with evidence from other recent research using different statistical experiments to purge schooling of its correlation with the wage equation error. Copyright 1995 by MIT Press.