Male-Female Wage and Productivity Differentials: A Structural Approach Using Japanese Firm-Level Panel Data

In an attempt to explain male-female wage differential, I estimated the relative marginal productivity and relative wage of female workers compared to those of male workers using panel data of Japanese firms. The relative wage of female workers is also estimated from the same data. Cross-sectional estimates that neglect firm-level, fixed effects indicate that the marginal productivity of female workers is 44 percent of that of male workers, while female wage is 31 percent of that of male workers. These estimates indicate that part of the wage differential cannot be explained by the productivity differential. However, the IV estimates that allow for firm-level, fixed effects indicate that both female marginal productivity and wage are about 50 percent of those of male workers. Thus we cannot reject the null hypothesis of no discrimination against female workers once the selection of workers into productive and high paying firms is accounted for. Evidence found in this study is consistent with the existence of employer sex discrimination at the point of job entry, but not afterward.