This dissertation consists of four essays on the economic causes and consequences of large floods. The first part of the dissertation investigates the socioeconomic and institutional determinants of large floods, and the second part analyzes the impact of large floods on armed conflict.
In the first essay, we build on previous research that reports that deforestation increases the frequency and severity of large floods. Unlike previous studies, we control for population, urbanization, income, and corruption in addition to forest cover and exploit the panel nature of the data to account for unobserved country and time heterogeneity. The essay shows that the link between floods and deforestation at the country level is not robust. It seems to be driven by sample selection and omitted variable bias.
The second essay analyzes the impact of large floods on armed conflict. Unlike previous studies that group all natural disasters indistinctly and treat the incidence of natural disasters exogenously, this essay separates floods from other natural disasters and uses an instrumental variable approach to correct for the endogeneity of floods. Flood could be endogenous (that is, determined simultaneously with the occurrence of conflict) if the presence of conflict reduces a country’s ability to effectively provide public services related to floodplain management and flood emergency management, thereby increasing the probability of severe flood events. Results show that
large floods increase the probability of conflict incidence (continuation of existing conflicts). The estimated impacts are substantially larger (8- to 10-fold) under specifications that control for the endogeneity of floods.
The third and fourth essays explore the potential transmission channels through which floods may affect armed conflict. The third essay shows that large floods, by displacing thousands of people, increase the probability of conflict incidence in the receiving areas. The effect is larger in developing countries and decays with time. The fourth essay finds that floods are a negative shock to short-run GDP growth and that the decline in short-run GDP growth increases the probability of conflict incidence.