Child human capital in developing countries

Abstract

This thesis contributes to three current topics in child human capital development in the developing world. First, we examine the role that child ability plays in parental investment decisions in Ethiopia. Second, using a different dataset, we investigate whether women’s empowerment could improve child nutritional status in Ethiopia. Third, we study the effect of sanitation on child cognitive ability in India, and explore heterogeneous effects by child endowments. In Chapter 2, we present a paper which exploits the longitudinal Young Lives data survey in Ethiopia to evaluate the causal effect of child cognitive skill on parental educational investment within the family. The study uses instrument variables combined with sibling fixed-effects to tackle the endogeneity in child ability and parental investment. We find that parents compensate the low-ability child among their primary school-age offspring through increased educational fees. We also find that this effect mainly holds for the high socioeconomic households. Chapter 3 investigates the association between maternal autonomy and child nutritional status among children aged 0-59 months in Ethiopia. Using the nationally representative Demographic and Health Survey data and an innovative empirical methodology- ‘post-double-selectionLasso’ (PDS-LASSO) - which helps to avoid dubious variable selection and to deal with omitted variablebias, we provide evidence that a child is less likely to be underweight if his/her mother has high autonomy. Notably, this correlation is strongest for children who are older than two years old. The last paper, shown in Chapter 4, explores the impact of shocks and policy during the fetal and infancy periods, along with their interaction, on later cognitive performance in rural India. Specifically, it studies the separate effect of in-utero rainfall fluctuations and a sanitation campaign at birth. It further discusses whether the effect of the sanitation campaign is differential by child endowments induced by previous rainfall shock. Using a difference-in-differences design, we do not find that the return to the sanitation programme is higher for advantaged children, as evidence shows that children who were exposed to positive rainfall in utero achieve similar scores at the age of 8-11 through the improved sanitary environment at birth.