The effect of schooling on income in Japan

Abstract

This paper uses cross-sectional data from the 1955, 1965, and 1975 Social Stratification and Mobility Surveys to investigate the effect of schooling on personal income in the Japanese male labor force. For each survey, log-income regressions are estimated which include (in addition to controls for years of work experience) two variables to indicate educational attainment: (1) years of schooling completed, and (2) percentile ranking in the distribution of years of schooling for one's age-cohort. The effect of the first educational variable may be interpreted as the human capital effect of schooling. The effect of the second educational variable may be interpreted as the screening or credentialing effect of schooling. The results indicate that controlling for the credentialing effect of schooling significantly reduces the net effect of schooling as human capital. Regression decomposition is then used to ascertain the components of the growth in mean log-income between 1955 and 1975. The contribution of years of schooling to the increase in mean log-income across these decades is significantly reduced after controlling for the credentialing effect.