Heterogeneous Returns to University Education in Urban China

This paper employs the China General Social Survey (CGSS2003) data and uses the method proposed by Heckman, Urzua, and Vytlacil (2006) to estimate the return to university education in urban China. The estimation shows that there is severe gender difference, regional difference in the opportunity to enter university. Male students and those from the east provinces have more chance to enter university. Parental education, especially father's education has a positive effect on children's chance to enter university. The return to university education for randomly selected person, i.e. average treatment effect (ATE), is 128.63%, which is tremendously higher than OLS estimate, 82.34%. The estimated marginal treatment effect for a person who is indifferent whether enter university or not is a decreasing function of unobservable heterogeneity, which shows there is a tremendous unobservable heterogeneity in China's education.