disasters, i.e., how severe is the initial economic impact and how quickly do firms recover, and ii) how much of the total damage comes from direct disruption versus propagation through supply chains? We address both questions using Ecuador’s administrative tax records, which capture the near-universe of formal firm-to-firm transactions between 2012 and 2022. This allows us to reconstruct Ecuador’s national production network at the firm level and at a monthly frequency. We exploit the April 2016 magnitude 7.8 earthquake as a natural experiment, applying a difference-in-differences event study to quantify the impacts on purchases and sales of directly affected firms and on their indirectly affected trading partners. We find strong direct effects for firms in the disaster zone, with a clear dose–response pattern across levels of local earthquake severity. Aggregate activity returns to pre-event levels within roughly two months, but a sectoral decomposition reveals that this rebound is driven by construction-related transactions consistent with reconstruction spending rather than a genuine recovery of pre-event productive activity. Indirect effects scale with firms’ pre-event trade exposure to the disaster zone, indicating that a substantial share of the earthquake’s economic impact materialized in the supply chains of directly affected firms. Together, these results show that time-resolved firm-level supply chain data are essential for capturing the true reach and persistence of disaster shocks, which aggregate statistics alone cannot reveal. Our findings also suggest that post-disaster aid should extend beyond firms physically inside the disaster zone to firms indirectly exposed through supply chain links, to address the full economic footprint of natural disasters.

Please note that this is an internal talk. If you want to participate, write an email to .