Production network disruption: Evidence from the civil-conflict in Ethiopia

This study investigates the role of trade networks in the dissemination and amplification of economic shocks using the 2020 Ethiopian civil war as a quasi-natural experiment. We use Value-added Tax (VAT) data from the Ethiopian Revenue and Customs Authority (ERCA) to construct a network of firm-level supplier–buyer connections. Using a difference-in-differences strategy, we find a significant negative impact of conflict exposure on firms located outside the conflict area. Firms located outside direct conflict zones but with pre-existing trade ties with affected areas experienced a 10% decrease in sales. Furthermore, we disaggregate the impact of buyer and supplier exposure to the conflict area. We find that both buyer and supplier connections to the conflict area have a negative impact, with buyer exposure having a stronger effect. The event-study analysis confirms the negative effect of the conflict on firms connected through trade networks. Our results are robust to various empirical checks including placebo tests with false treatment periods and regions, exclusion of outliers. Finally, we collected survey data from firms in non-conflict areas to provide further insights into how conflict affects their trade relationships with firms in conflict-affected areas. 82% of firms faced difficulties in communication or collaboration with their business partners, and 36% of firms experienced contract breaches or payment delays due to the conflict. Policymakers involved in post-conflict reconstruction should not only focus on directly affected regions but also on firms outside directly affected conflict areas, as they are also vulnerable to the ripple effects of conflict.