Statistical Discrimination, Occupational Sorting, and Career Opportunities

Statistical discrimination is an important determinant of earning gaps. Yet, little is known about its consequences on wage determination. This paper studies the effects of statistical discrimination against female workers due to their weaker labor force attachment. We first analyze the equilibrium wage outcome using a screening model. We then test whether statistical discrimination exists by extending the employer learning model using Japanese data. We find evidence for statistical discrimination against women based on average labor force attachment. Last, we calibrate the model and results, which indicate that decreasing child-care subsidies might improve total welfare in a cost effective manner. JEL Classification: J24, J31, J63, J71