Essays on Wage Determination Factors and Individual Heterogeneity : Evidence from Japanese Panel Data

Abstract

This dissertation presents three essays related to factors determining wage. Verifying factors influencing wage levels and changes in wage rate has remained a key issue in labor economics. Since Mincer (1974) proposed his famous Mincer earnings function on the basis of human capital theory in his book “Schooling, Experience, and Earnings,” many empirical studies have shown that these years of schooling and labor market experience are major factors determining wages in the United States and several other countries (Burusku 2006). Studies analyzing the role of returns of education, work experience, and job tenure on wage apply the Mincer equation to tackle the problems of omitted variables and individual heterogeneity of biased estimates (Altonji and Shakotko 1987; Abraham and Farber 1987; Marshall and Zarkin 1987). These problems also apply to the case of Japan. This dissertation investigates factors affecting wages by conducting a Japanese panel data analysis. As Hsiao (2003) and Baltagi (2013) note, panel data analysis can remove potential sources of time-invariant unobservable individual heterogeneity, which cannot be controlled for using time-series or cross-sectional data. There are several potential factors influencing wages, including family and social environment and habituation. Recent economists have investigated the influence of behavioral factors on economic decision-making. In this thesis, we investigate the effects of two behavioral factors, smoking and marriage, on wages. We use two methods to measure personality traits as potential unobservable individual heterogeneity to assess the validity of these unstable variables in an economic empirical analysis. In fact, several recent studies, particularly in the field of economic behavior, use unstable variables such as time discount rate, risk aversion rate, and a locus of control or other personality traits as proxy variables of individual unobservable heterogeneity. The analysis results reveal that individual unobservable heterogeneity affects wage determination through two behavioral factors: smoking and