In this dissertation, I explore the financial mechanisms available to individuals and governments after a disaster in developing countries, using Mexico as a case study. Specifically, I explore the role of fiscal federalism, remittances, migration, access to financial services, and catastrophe bonds. Across these dimensions, I focus on the effects of hurricanes and earthquakes, as these have been the costliest hazards for Mexico.

First, I estimate the effects of natural disasters on municipal revenues and expenditures. I find that the current fiscal rules disincentivize municipalities from collecting more taxes or reallocating funds to implement post-disaster policies. Additionally, state-level officials do not appear to follow federal criteria when distributing federal transfers to municipalities in the aftermath of a disaster.

Second, I test whether remittances to the affected region are filling the gap left by the lack of government response. Overall, the findings reveal that remittance levels decrease after a municipality has been hit by an earthquake or both earthquakes and hurricanes. I further test whether this decline is driven by access to financial services or migration. My results suggest that the affected agents migrate after a disaster to smooth consumption. Third, I propose a comprehensive policy evaluation framework to evaluate risk-financing tools. Then, I illustrate this policy framework using the Mexican CAT bonds program as a case study. The results reveal that the Mexican CAT bonds program performs well against the criteria of the input evaluation stage.

However, my results suggest that the program could be adjusted to improve its performance relative to the process/product evaluation criteria.

Based on this work, there are two potentially important future research areas. First, improve the understanding of the underlying mechanisms that explain changes in the spending priorities to identify leverage points for policy improvements in the fiscal federalism and disaster response systems. Second, identify damage drivers and use them to estimate both the marginal effects of disasters based on their intensity and to improve the triggering parameters of Mexican CAT bonds.

Further exploring these research streams is particularly relevant in a future where the frequency and magnitude of natural hazards are expected to increase.