The Spillover Effects of Hospital Closures on the Efficiency and Quality of Other Hospitals

Abstract

Problem definition: The recent trend in the U.S. hospital closures can have important impacts on the healthcare sector by changing the operational efficiency and quality of care of the remaining hospitals. We investigate the impact of hospital closures on the surrounding hospitals’ operational efficiency and quality, and study how such hospitals respond to the closure of their neighboring hospital. Academic/Practical Relevance: Understanding how hospital closures impact the way the remaining hospitals operate can allow policymakers to utilize more effective policy levers in order to mitigate the negative consequences of hospital closures. Methodology: We analyze more than 14 million inpatient visits made during 11 years to over 3,000 hospitals in the U.S. (before and after various closures), and utilize causal methods to evaluate the spillover effect of hospital closures on the nearby hospitals. We also conduct counterfactual analyses to evaluate policy interventions that could have been used by policymakers. Results: Hospital closures have both positive and negative spillover effect. When a hospital closes, its nearby hospitals improve their operational efficiency. However, they do so via a speed-up response (i.e., by reducing their service durations to accommodate the increased demand) instead of an effort to lower their average bed idle time. This speed-up response negatively affects some important aspects of the care quality provided, including the 30-day mortality rate. The spillover effect of a hospital closure is highly heterogeneous: hospitals in markets where patients have limited choices of hospitals (e.g., less competition, fewer resources) and hospitals that are more desirable (e.g., high-quality, urban, teaching, and large) tend to experience greater spillover effects. Managerial Implications: Our analyses suggest two effective policy levers: (a) bailing out specific hospitals (e.g., rural or less desirable than neighbors) from potential closures, and (b) eliminating the speed-up response of specific hospitals (e.g., rural or more desirable hospitals). In addition to helping policymakers, our results help hospital administrators: our findings help them to better understand the consequences (or the absence) of their strategic responses to a neighboring hospital closure, and thereby, adopt more suitable management strategies.