Within-Family Resource Allocation Decisions And Human Capital Formation

Individuals' human capital accumulations,closely depending on within-family resource allocation decisions,affect not only individuals' earnings and happiness,but also the quality level of labor force and economic development.My dissertation focuses on the impact of within-family resource allocation decisions on individuals' human capital formation,and studies the following three questions.The first question examines how within-family resource allocation decisions affect individuals' human capital formation by utilizing an exogenous increase in the supply of educational resources.The second question turns to the externalities of individuals' human capital caused by within-family resource allocation decisions based on Becker and Lewis's(1973)Quantity-Quality theory.The third question focus on the externalities of individuals'human capital based on the perspective of the social network in a classroom.The findings of my dissertation are provided below.In the first question,I utilize China's college expansion campaign as a quasi-natural experiment and find that college expansion has increased the gender gap in high school enrollment among families with both sons and daughters.The conclusion remains unchanged even after the parallel trend assumption,a battery of robustness checks,placebo tests,the impacts of composition bias and some confounding events are taken into account.Combining the theoretical results with other empirical findings that girls perform better academically than boys and that college expansion has increased the female high school enrollment rate,the results further show that son preference is the major factor in explaining the rising gender gap in high school enrollment.Finally,in order to provide more direct empirical evidence,I divide the hukou cities into two groups,one with strong son preference(if the sex ratio>1.07)and the other one with weak son preference(otherwise).I find that the gender gap in high school enrollment is significantly increased by college expansion among the individuals from the strong son preference group but is not significantly affected in the weak son preference group.To address my second question,I incorporate externality into Becker and Lewis's(1973)Quantity-Quality model and utilize the 1979 one-child policy in China as an exogeneous shock to demonstrate the quantitative importance of externality impacts on high school enrollment as predicted by the extended model.The traditional Becker and Lewis's model only focuses on economic decisions within a household.However,household decisions collectively could affect aggregate outcomes which may generate feedback effects on education outcomes through external quantity and quality channels.My theoretical results show that rising fertility costs can decrease the quantity of children,however,the impact on the quality of children will depend on the relative importance between the price and income effects.Under the assumption of homogeneous families,I further show that the external quantity impact on high school enrollment is positive and it dominates the external quality impact.Using the 2000 and 2005 census data,the empirical results are consistent with the theoretical prediction such that rising fertility costs have a positive impact on high school enrollment,which can be explained by the external quantity channel.Furthermore,the conclusion remains unchanged after I conduct a series of robustness tests and placebo tests and employ the instrumental approach to correct for endogeneity bias.To examine my third question,I focus on the externality of individual noncognitive ability in the context of a classroom.Given that emotional characteristics are an important indicator of individuals' non-cognitive ability,I take the negative emotions as an example and study the externality of students' non-cognitive abilities within a classroom.Using data on students' self-reported negative emotions from the 2013-2014 China Education Panel Survey,the empirical results show that negative emotions are contagious among students within a classroom.The conclusion remains unchanged after conducting a battery of robustness tests.In addition,I find that female students or more socially active students are more responsive to negative emotions.