An agent-based model for the assessment of the ripple effect of disaster economic losses considering firms’ adaptive behaviors

Abstract

Assessing the spatial and temporal ripple effect of disaster economic loss is essential for integrated disaster risk management considering disaster countermeasures dynamically and systematically. Traditional loss estimation models based on general equilibrium theory hardly take the ripple process into account, as they just focus on the equilibrium status of economic system. However, the debate is that system will not reach equilibrium at one stroke and firms' adaptive behaviors will also change the disaster ripple losses. This paper proposes an agent-based model to investigate this phenomenon: the economic linkages among micro-agents are established based on regional input–output table and firms' production behaviors are defined according to nested production functions. The 2020 flood disaster in Hubei Province, China, is taken as an application case which demonstrated the feasibility of agent-based model for assessing the economic loss ripple effect considering firms' adaptive behaviors. Several interesting findings from the application are also drawn. (1) The speed of loss ripple is influenced by the distance to the disaster-stricken region, while the intensity of loss ripple depends on the strength of interregional industry linkages. (2) Round by round ripple effects could be found in non-disaster regions due to the lagged impacts caused by damage of production capacities in other non-disaster regions. (3) Firms' adaptive behavior reduced the ripple losses nearly by 60 % and saved CNY 5.72 billion in this case study. Notably, the gains from overproduction behavior in non-disaster regions are critical, with reducing the loss of CNY 5.8 for every CNY 1 of overproduction.