Website CV Paper
Fields
Environmental & Energy, Development, Labor, Health
Job market paper
Shaken but Standing: Self-Reported Shocks, Physical Exposure and Household Coping
Governments and researchers rely on households’ self-reports to measure disaster impacts and target assistance. This paper asks whether households reporting disaster impacts were physically hit harder than those who don’t, and whether self-reports or hazard data can identify those most in need. I link data on 104,000 households in Haiti after the 2021 earthquake to ground-shaking intensity at each household’s location, self-reported earthquake impacts, and distress measured through coping strategies such as borrowing, withdrawing children from school, begging, and selling land. Stronger shaking increased distress. Yet households self-reporting impacts experienced only marginally stronger shaking than their neighbours, and differences in shaking explain just 3% of the difference in distress between them. Neither shaking intensity nor self-reports reliably identify the most distressed households. When assistance is allocated to 20% of households in each community, self-reports identify only 20-22% of those in distress, barely above random selection. A score based on other registry information performs better, identifying 26-30%, even when estimated on a small sample. These findings highlight the potential of administrative data to improve disaster assistance targeting.