Siblings, public facilities and education returns in China

This paper investigates the intrahousehold resource allocation on children’s education and its earnings consequence in Chinese labour market. In order to overcome the endogeneity problem of schooling, we consider the siblings structure and the available public facilities as instrumental variables. Females’ education is negatively affected by siblings (brothers or sisters) number, while males’ education is also negatively affected by their brothers but much less by their sisters. For the youngest cohort born after 1980, the education of a girl would be heavily impeded by her sisters, reflecting strong distortion of “One-Child Policy” on intrahousehold resource allocation. Comparing the OLS and GIV estimations for returns to schooling, we find that there are downwards biases of OLS estimations for males in all cohorts and in all years. However, for females, downwards biases of OLS estimation are only for data before 2004, as females in the old cohorts actually have upwards biases after 2004. Education returns of the youngest cohort are much higher than old cohorts supporting the argument of heterogeneous human capital accumulation during transition.