Empirical essays on development economics

This thesis investigates empirically three questions of key relevance for the life of disadvantaged people in developing countries. Using a sample of Ethiopian women and a regression discontinuity design exploiting age discontinuities in exposure to a law that raised the legal age of marriage for women, the first chapter documents for the first time (a) the effect of increasing the legal age of marriage for women on infant mortality and (b) the causal effect of early cohabitation on infant mortality. The analysis shows that, even though it was not perfectly enforced, the law that raised the legal age of marriage had a large effect on the infant mortality of the first born child. Furthermore, the estimates suggest that the effect of a one-year delay in women's age at cohabitation on the infant mortality of the ffrst born is comparable to the joint effect on child mortality of measles, BCG, DPT, Polio and Maternal Tetanus vaccinations. Using longitudinal data from northern Ghana, the second chapter shows that parents allocate more schooling to children that are more cognitively able. These results provide evidence for the main prediction of the model of intra-household allocation of resources developed in Becker (1981), which concludes that parents allocate human capital investments reinforcing cognitive differences between siblings. The third chapter uses the 8.8 Richter magnitude earthquake that struck Chile in February 2010 as a case study and employs a difference in difference strategy to investigate whether natural disasters have lasting effects on property crime. The results show that the earthquake reduced the prevalence of property crime the year of the earthquake and that this effect remained stable over the 4 post-earthquake years studied. The lasting drop in crime rates in affected areas seems to be linked to the earthquake strengthening community life in these municipalities.