Estimating the Human Capital and Screening Effects of Schooling on Productivity in U.S. Manufacturing Industries, 1979-1996

Arthur Sakamoto, University of Texas at Austin
Changhwan Kim, University of Texas at Austin

A classic issue in labor market studies is whether the correlation between an individual’s education and her socioeconomic rewards reflects increased productivity (i.e., human capital), labor market screening or credentialism. While previous research has usually avoided the challenge of empirically testing these three competing explanations, we provide some relevant findings using productivity data for U.S. manufacturing industries from 1976 to 1996. Our results indicate that, contrary to the credentialist hypothesis, mean years of schooling has a strong and robust net effect on industrial productivity. In contrast to the screening explanation, a measure of relative educational among the workers in an industry has no positive net effect on industrial productivity. Our findings most strongly support the human capital interpretation in that years of schooling has a large net effect on industrial productivity even after controlling for relative educational attainment.