The Economics of Inequality and Human Capital Development: Evidence from Nepal

Abstract

This thesis has three pieces of empirical studies that analyse economic inequality across social groups (castes and gender) and its impact on human capital endowments in developing countries with particular reference to Nepal. Three aspects of inequalities have been examined: disincentive in educational attainment in female arising from labour market discrimination, disproportional representation of low-caste workers in better jobs and inequity in health care utilisation and health outcomes across castes. This study contributes to the literature of economics by developing a new theory and extending existing econometric models in analysing economic inequality across social groups. The rst piece of research examines the impact of marital anticipation on female education in the presence of labour market discrimination. It develops a theoretical model for jointly determining the age at marriage and female education. The model hypothesizes that as females are not rewarded in the labour market as much as men are; married women are encouraged to engage in household work as a result of the intra-household division of labour in their marital union. Thus, parental anticipation of this e ect a ects their daughter's age at marriage and can in uence investment in girls' schooling. It then estimates the causal e ect of age at marriage on education in light of the theoretical model using household data from Nepal. In order to control for potential reverse causality this study uses variation in cultural norms regarding dowry and di erences in average age of female marriage among ethnicities and regions as instrumental variables. The econometric results con rm that the gender gap in education is signi cantly a ected by cultural practices that favour early marriage and that increasing girls' marriage age by one year would produce on average :4 year increment in women's schooling. The second study examines the sources of wage di erentials across castes in Nepal by employing an extended form of Oaxaca decomposition methodology. This study shows that, in countries such as Nepal which have imperfect labour market for both goods and services, the conventional Oaxaca decomposition methodology fails to estimate the source of wage di erential precisely.