A multi-agent model of a low income economy: simulating the distributional effects of natural disasters

This paper develops an agent-based model of a stylized low income region in order to study the impact of natural disasters on population displacement, income, prices, and consumption with a focus on distributions and coping strategies of low income groups. Key features of the model include the integration of decentralized markets into a full economy in a spatially explicit way and the analysis of short-run adjustment processes. The model is calibrated to a low income region of rural agrarian Pakistan that faced severe floods in 2010. Dynamic adaptation by agents in response to falling income includes migrating and running down savings. Despite these consumption smoothing strategies, some low income groups are vulnerable to starvation. The paper showcases two hypothetical policy scenarios, a cash and a food transfer program, and tracks their effects on the welfare of low income groups in the economy.