Is growth at risk from natural disasters? Evidence from quantile local projections

This paper investigates the heterogeneous and state-dependent effects of natural disasters on GDP per capita growth in developing countries. Using a panel of 67 countries, I apply the Quantile Local Projection (QLP) method to estimate Quantile Impulse Response Functions (QIRFs) following different types of disasters—floods, droughts, storms, and earthquakes. By focusing on the conditional distribution of growth rather than its average, the analysis captures how disasters distort growth distribution at the tails. Results reveal that natural disasters represent a growing downside risk, significantly amplifying the likelihood and depth of recessions at the lower tail of the distribution. In contrast, responses at the upper tail are often expansionary in some contexts, suggesting asymmetric effects. These patterns are shaped by countries' income levels, institutional quality, and sectoral structure. Overall, the findings highlight the importance of distribution-sensitive approaches to better understand and manage the macroeconomic consequences of climate and disaster shocks.