Cascading economic impacts of critical infrastructure failures on supply chains

Transportation networks, such as roads and bridges, have a fundamental role in the daily economic and social activities, enabling the access to jobs, goods delivery, and social services. However, due to their interconnected nature, these infrastructures are particularly vulnerable to external shocks, especially natural hazards, and therefore it is essential to assess their risk through a systemic approach. Yet, traditional risk assessment methods typically focus only on direct physical damage to infrastructure, often overlooking the cascading effects, especially how this affects firms and households, a crucial limitation for understanding the large-scale economic costs of extreme weather events. In order to address these gaps, we present a comprehensive framework that integrates the novel DisruptSC model, a spatially explicit agent-based model that captures the propagation of infrastructure failures through supply chains, with standard direct impact modeling approaches. By explicitly representing synthetic firms, households, and transport networks within a unified system, the framework allows to quantify both the direct infrastructure damage and the ripple effects that spread through interconnected economic systems can be quantified thanks to this integrated approach.

We applied this framework to Cambodia in order to not only quantify the indirect impacts of critical infrastructure failures but also the effects of sequential or spatially distributed cascading hazards on supply chains. Cambodia is in fact affected by extended heavy rainfall during the wet season leading to multiple flooding phenomena occurring in close succession. The results show that transport disruptions generate substantial indirect economic losses that extend well beyond the directly affected areas. In particular, the model highlights two distinct but interacting mechanisms, namely increases in prices driven by costly rerouting and shortages arising from complete network blockages. While inventories initially buffer these shocks, their depletion over time leads to strongly nonlinear increases in losses, underscoring the importance of disruption duration. Moreover, the analysis reveals that a limited number of critical road segments disproportionately drive aggregate impacts, with relatively small additional disruptions triggering sharp increases in economy wide losses. Overall, the results demonstrate that indirect losses can equal or exceed direct infrastructure damages, and that ignoring cascading effects leads to a systematic underestimation of flood related risks. These findings underline the need for integrated assessment frameworks that explicitly link hazard processes, infrastructure vulnerability, and supply chain dynamics in order to support more effective resilience oriented investment and policy decisions.