Learning in Recovery from Disruption: Empirical Evidence from the U.S. Drug Shortages

U.S. hospitals continue to face chronic shortages of essential drugs, jeopardizing patient care and driving up costs. Although prior work has identified the structural and economic drivers of these shortages, little is known about whether pharmaceutical plants accumulate operational knowledge from their recovery efforts—tasks that are often irregular and ill-defined. We assemble a unique data set linking 4,741 drug shortages (2015–2020) to 136 finished dosage form plants to examine whether—and under what conditions—plants learn from past recoveries. Using an instrumental variables approach, we find that each additional shortage a plant has previously resolved significantly reduces its time to recovery for a current shortage, indicating cumulative learning. We further show that learning is concentrated in shortages caused by internal plant issues, with no such effect for externally induced disruptions. Breadth of experience also matters; learning gains are larger when prior recoveries span diverse active ingredients or causal categories. Post hoc analyses reveal that the knowledge gained is confined largely to the focal plant and does not readily transfer across facilities within the same firm. These findings suggest that structured knowledge-capture processes, cross-site learning mechanisms, and incentives to diversify recovery experience can strengthen supply chain resilience.

This paper was accepted by Stefan Scholtes, healthcare management.

Supplemental Material: The online appendix and data files are available at https://doi.org/10.1287/mnsc.2024.08337.