Problem Definition: Natural disasters disrupt retailer operations, leading to supply and demand shocks. Purchasing becomes particularly challenging due to the rise in transportation costs and stores that remain open witness an increase in demand due to displacements and stockpiling. This paper studies how sales in retailers evolve after a natural disaster. Methodology/
Results: We run a triple difference regression using product-level scanner data with detailed supplier locations and major disaster declarations from three North Atlantic hurricanes to compare the sales evolution of products sourced from areas close to each store (referred to as “local”) against products sourced from farther away (referred to as “non-local”). We find that in operational stores located in affected areas, there is a significant increase in the sale of products from local suppliers compared to products from non-local suppliers one month after the disasters. We show that the former have a lower stockout probability compared to the latter. We also find that product-level prices increase for non-local products while they remain at similar levels for local products. Finally, we do not find evidence for assortment changes in both groups. Managerial Implications: Our study highlights the importance of supply chain management when facing a disruption from natural disasters. We provide evidence that local suppliers might be an important part of the recovery stage after a natural disaster when retailers face a large number of stockouts.
