Item

Abstract

Anticipatory humanitarian assistance is a novel approach to aid in the context of weather disasters, drawing on meteorological forecasts. Using a randomized study design, we analyze the impact of anticipatory cash transfers distributed to pastoralist households in Mongolia before an extreme winter event reached its peak. For recipients with smaller herds before the disaster, the receipt of anticipatory cash transfers increases their post-disaster herd size, herd-related investments, and home consumption of livestock. These effects, measured on average seven months after the event, are large in magnitude and consistent across a range of robustness tests. No significant effects of the intervention are detected when considering the full sample, including relatively wealthy pastoralists. Furthermore, no heterogeneous effects are found for different levels of disaster intensity. The paper concludes by highlighting practical challenges in evaluating (anticipatory) humanitarian interventions.