Assessing the economic ripple effect of flood disasters in light of the recovery process: Insights from an agent‐based model

Abstract To assess the economic ripple effect, this study integrates agent‐based modeling (ABM) with a multiregional input–output (MRIO) table to develop an assessment model that considers capacity recovery process. The intermediate and final demands in the MRIO table are used to describe the agents’ interdependence. Survival analysis is used to construct capacity rate curves. By defining the first‐ and second‐order ripple effects, ABM is used to capture the ripple process in days. To conduct a case study, the service and retail sectors in Enshi in Hubei, China, are selected as disaster‐affected sectors (they were severely affected by the July 17, 2020 flood disaster). The main findings are as follows: (1) With the first‐order ripple effect, the losses caused by service and retail are concentrated within Enshi. Enshi's final demand, construction, and raw materials manufacturing sectors as well as Wuhan's construction sector are seriously affected. (2) With the second‐order ripple effect, the losses caused by the service and retail sectors expand, forming a prominent industrial ripple chain: “service (retail)–raw materials manufacturing–construction.” (3) The direct and indirect losses caused by the service sector are more significant than those caused by the retail sector. However, the loss ratio of the service sector is smaller than that of the retail sector because of its sound industrial structure and strong resilience. Hence, the indirect losses caused by different sectors are not entirely determined by their direct losses; instead, they are also related to the degree of perfection of the structures of different sectors.