Modeling loss-propagation in the global supply network: The dynamic agent-based model acclimate

• Published in: Journal of economic dynamics & control Vol. 83; pp. 232 - 269
• Main Authors: Otto, C., Willner, S.N., Wenz, L., Frieler, K., Levermann, A.
• Format: Journal Article
• Language: English
• Published: Elsevier B.V 01.10.2017
• Subjects: Agent-based modeling; Disaster impact analysis; Dynamic input-output model; Economic network; Higher-order effects; Resilience; Disaster impact analysis; Higher-order effects; F180; Economic network; Q540; Dynamic input-output model; Agent-based modeling; F100; Q560; Resilience
• ISSN: 0165-1889; 1879-1743
• DOI: 10.1016/j.jedc.2017.08.001

World markets are highly interlinked and local economies extensively rely on global supply and value chains. Consequently, local production disruptions, for instance caused by extreme weather events, are likely to induce indirect losses along supply chains with potentially global repercussions. These complex loss dynamics represent a challenge for comprehensive disaster risk assessments. Here, we introduce the numerical agent-based model acclimate designed to analyze the cascading of economic losses in the global supply network. Using national sectors as agents, we apply the model to study the global propagation of losses induced by stylized disasters. We find that indirect losses can become comparable in size to direct ones, but can be efficiently mitigated by warehousing and idle capacities. Consequently, a comprehensive risk assessment cannot focus solely on first-tier suppliers, but has to take the whole supply chain into account. To render the supply network climate-proof, national adaptation policies have to be complemented by international adaptation efforts. In that regard, our model can be employed to assess reasonable leverage points and to identify dynamic bottlenecks inaccessible to static analyses.