Date of Award
Spring 2024
Document Type
Dissertation
Degree Name
Doctor of Philosophy (PhD)
Department
Economics
First Advisor
Rosenzweig, Mark
Abstract
A substantial fraction of the population in the developing world experiences the death of a parent during childhood. This dissertation analyzes the consequences of early parental death on schooling in India, where one in ten individuals faces such a loss before turning 18. By examining the mechanisms through which parental death affects schooling, this study sheds light on how educational choices are shaped by household resources, children's opportunity cost of time, parental preferences, and intrahousehold bargaining. The study disentangles and quantifies these mechanisms by using the death shock, time-use data, and itemized expenditure data to estimate a structural model of household consumption and time allocation. First, I estimate the causal effect of parental death on school enrollment. To address endogeneity concerns associated with comparing orphans and non-orphans, identification relies on variation in the timing of parental loss among nearly 10,000 orphans. Specifically, I compare the enrollment choice of a child before and after parental death, relative to children of the same age observed at the same time, but who will experience the shock in a later wave. I find that among children aged 12 to 18, both paternal and maternal deaths lead to a 7-percentage point decline in the probability of school enrollment. Second, this study analyzes the mechanisms underlying the effects of parental loss on schooling. Specifically, I evaluate the importance of three channels: (1) changes in household resources, (2) loss of parental inputs into human capital production, and (3) shifts in preferences due to the transition from two-parent to single-parent households. Given the strict gendered division of labor in India, the resource mechanism delivers tight predictions. Paternal death reduces household income and therefore should mostly affect sons, while maternal death constitutes a loss in home production and therefore should mostly affect daughters. Analysis of time-use data confirms that when a father dies, sons drop out of school to enter the labor force, and when a mother dies, daughters drop out of school to manage domestic responsibilities. However, I also find that maternal loss drives sons into the labor force, despite the fact that the shock does not affect household income or the probability of hiring domestic help. This result indicates that the resource mechanism cannot fully explain the observed effects and that changes in household preferences or loss in parental inputs play an important role. To quantify the relative importance of each mechanism, I estimate a structural model of household consumption and time allocation that incorporates household production, public goods, and parental bargaining. On the production side, households combine time and raw materials to produce a domestic good. Each parent has distinct preferences over market goods, the domestic good, children's education, and the leisure of each household member. The parents bargain over the intra-household allocation of resources in a collective household framework. The model is estimated using shifts in time allocation and itemized expenditure in response to the death shock. The estimates show that compared to fathers, mothers have a stronger preference for schooling and a lower bargaining weight in intrahousehold decisions. Paternal death affects schooling decisions by reducing household income, while maternal death affects schooling decisions by changing household preferences and removing her contribution to home production. Counterfactual simulations show that providing unconditional cash transfers of Rs. 4,000 per month following the death of either parent is sufficient to keep male orphans in school. In contrast, unconditional transfers prove ineffective in keeping female orphans in school. To eliminate the schooling effects of maternal death on daughters, a cash transfer conditional on enrollment of Rs. 3,000 per month is an effective policy tool.
Recommended Citation
Guerrero Castaneda, Rodrigo, "Parental Death and Schooling: Gendered Spheres of Production and Parental Preferences" (2024). Yale Graduate School of Arts and Sciences Dissertations. 1291.
https://elischolar.library.yale.edu/gsas_dissertations/1291
