ESSAYS ON ENVIRONMENTAL SHOCKS IN DEVELOPING COUNTRIES: ADAPTATION OF SOCIAL PROTECTION, MIGRATION AND LABOR

Abstract

Environmental shocks, particularly high impact natural disasters, such as earthquakes, floods and droughts, test the boundaries of social resilience and vulnerability. According to the International Disaster Database (EM-DAT), from 1975-2011 the number of natural disasters reported worldwide, along with the number of households affected, gradually increased over this period (Natural Disaster Trends, Center for Research on the Epidemiology of Disasters (CRED)). Economic status of a country did not predict the number of disasters a country faced. However, findings indicate that countries with lower incomes (Kahn (2005), Stromberg (2007), Keefer et al. (2011)) and countries with greater income inequality (Anbarci et al. (2005)) encounter more casualties and greater economic damage. It is therefore important to understand ways in which communities in lower income countries can cope with such community level shocks, as this can then point to changes that can be made to help these countries better cope with environmental shocks. This dissertation is comprised of three applied essays focusing on identifying consequences of environmental shocks related to social protection, migration and labor in developing countries. Recent literature on environmental shocks in low-income countries have focused on improving the measurement of such shocks to avoid common identification issues. The essays in this dissertation provide empirical and methodological contributions to a growing literature on measuring and understanding the implications of environmental shocks. In the first essay I address a gap in the current analytical literature on the effectiveness of decentralized targeting of social safety nets (often delivered in the same way as humanitarian aid) in insuring households against disaster risk. I combine survey data from Indonesia with geological earthquake data to determine if village leaders change the pattern of distribution of a subsidized rice program intended for the poor in earthquake affected villages. My findings suggest that the central government targets more safety net resources to earthquake villages, but access to these resources declines for its intended poorest beneficiaries, and targeting is worse in communities with higher social capital. I discuss how these findings may be linked to bargaining power assigned by village leaders to poor and non-poor recipients, which can be a function of the leader's personal benefit, either electorally or through reciprocity expected from social contacts or family members to whom the leader provides access. The second essay examines migration as a key mode of adaptation to extreme floods and droughts, and investigates the impact of weather-driven internal migration on local labor markets in Nepal. In this essay the identification and methodology used by Boustan et al. (2010) is modified to a dynamic framework to fit the contextual setting of the study. We combine survey data from Nepal with 0.5x0.5 degree gridded satellite based weather data to identify weather anomalies and then create instruments for local migration in Nepal. Our analysis of the impacts of local migration on labor markets finds native wage losses are slightly larger than those observed in the U.S. and elsewhere. Labor substitution is imperfect in Nepal, as migrants appear more skilled than the average native worker in hosting communities. These results suggest imperfect substitution coupled with fixed labor demand in the formal sector may partially explain why wage losses are more pronounced here than in other settings. We also find strong negative effects of migration on wages of low-skilled workers and informal sector employment. This is consistent with a displacement of low-skilled workers out of the labor markets. Highly skilled migrants may have to accept lower-skilled jobs to integrate into the labor markets and therefore, push low-skilled natives out of the labor markets. The third essay identifies the detrimental impacts of crop shocks, predominantly floods and droughts, on secondary school aged youth (aged 14 to 19) in Tanzania. While a large body of literature has focused on the causes and consequences of child (aged 7 to 13) labor very little is known about the impact of transitory shocks on youth. I find that crop shocks may increase youth labor significantly, and be particularly detrimental to school attendance of youth enrolled in school. Youth enrolled in school increase unpaid labor to substitute for the paid labor taken up by others in the household. These results also indicate that female youth are disproportionately more likely to engage in unpaid labor and miss school compared to male youth. I also identify that while youth schooling outcomes are affected by shocks, child schooling is not affected. These research findings suggest that more attention needs to be paid to short and long term consequences of shocks for youth.