Job tasks and wages in the Japanese labor market: Evidence from wage functions

Based on the microdata from the Basic Survey on Wage Structure and the original survey, we estimate the Mincerian wage function, incorporating job tasks, to examine the importance of job task wage premiums as well as long-term changes in the Japanese labor market. In the estimation, we found that the association between abstract tasks and wages is positive and that a one standard deviation increase in the abstract task scores was associated with a 21.2% wage premium, while that of routine and manual task wage premiums are negative. We also found that the total explanatory power of three task scores (routine, abstract, and manual tasks) is higher than that of the education dummies or the major occupation group dummies. We also confirmed two testable implications from the Roy model regarding the workers’ self-selection into occupations in the Japanese labor market. These findings are similar to those obtained by Autor and Handel (2013). On the other hand, we found no major changes between 2005 and 2016 in the coefficients of routine, abstract, or manual task scores as well as their explanatory powers in the wage function. We then observed that demand for labor increased in many occupations involving many non-routine or manual tasks, but at the same time, the supply of labor to those occupations also increased. Therefore, we discussed that the change in labor demand and supply may be one of the reasons for the stable relationship between job tasks and wages.