Effects of anticipatory humanitarian cash assistance to households forecasted to experience extreme flooding: evidence from Bangladesh

Abstract The 2020 monsoon floods in Bangladesh were among the most severe and protracted in decades. Instead of waiting for disaster to strike, the Bangladesh Red Crescent Society used impact-based forecast data to reach nearly 3,800 vulnerable households along the Jamuna River with a one-off unconditional cash transfer of BDT 4,500 (about $53) before peak flooding in July 2020. Anticipatory action to help at-risk populations avoid or mitigate extreme weather event impacts has become widely used by governments and humanitarian organisations worldwide. However, robust evaluations of the effectiveness of forecast-based assistance are limited. This assessment follows a quasi-experimental approach, drawing on survey data from a sample of cash recipients and equally vulnerable and flood-affected households that were not reached by BDRCS before the flood. Our analysis finds robust statistical evidence that the intervention was effective in helping households evacuate the flood-affected area, protecting personal health and well-being, and safeguarding people's productive assets and livestock. It was also effective in enabling beneficiaries to avoid taking on high-interest loans and selling valuable assets during and after the flood. The intervention does not appear to have helped cash recipients avoid food-based coping mechanisms or regain their productive capacity sooner after the flood.