Charity hazard in climate adaptation : the effect of anticipatory cash transfers on demand for weather insurance

This study examines whether receiving anticipatory cash transfers during an extreme winter affects households’ demand for index-based livestock insurance. We exploit a randomized field experiment conducted during the 2020/21 winter disaster in western Mongolia and combine household panel survey data with administrative insurance records. We do not find evidence of charity hazard: the estimated effect of anticipatory cash transfers on insurance uptake is small and statistically indistinguishable from zero. The 95% confidence interval rules out large crowding-out effects but remains consistent with small negative effects of up to 2 percentage points. Treatment effects are heterogeneous: among households with prior insurance experience, estimated effects are positive and statistically significant, while effects among previously uninsured households are statistically indistinguishable from zero. These findings suggest that, in contexts where assistance is incomplete and index insurance is well-established, anticipatory assistance does not need to undermine insurance demand.