# Jason Buxbaum Assistant Professor, Department of Health Services, Policy, and Practice, Brown University School of Public Health *Dissertation Title*: “Hospital Responses to Changes in Incentives, Resources, and Demand” At $1.4 trillion, US hospital spending is comparable to spending on national defense and K-12 education combined. The more populous EU spends less than half this sum. Improved understanding of our exceptional hospital industry can help US policymakers better align spending with policy priorities. These papers have been written to that end. Chapter 1, Money—What Is It Good For? The Impact of Enhanced Covid-19 Provider Relief on Hospital Finances, Capacity, Utilization, and Equity Congress appropriated $178 billion in emergency Covid-19 relief for US healthcare providers in 2020. Nearly every hospital received at least some funding. However, a $35 billion subset of hospital-directed relief was awarded using all-or-nothing qualification criteria. I exploit these thresholds to evaluate the impact of extra relief among 518 hospitals near a cut-off for enhanced funding. Using a novel dataset in combination with a difference-in-discontinuities design, I find that enhanced relief not only held hospitals harmless but differentially improved finances. While enhanced relief significantly increased hospital revenues by about two percent—and possibly increased the complexity of patients served—neither overall capacity nor overall spending detectably increased. The result was an increase in margins of about two percent and comparatively limited increases in liabilities. However, I find some evidence the relief had greater effects for hospitals serving a relatively high share of Black, Hispanic, or dually eligible patients. Adjusted approaches to relief distribution may best support hospitals in future disasters. Chapter 2, Need and Determinants of Hospital Resource Use, 2010-2019 Joint with Michael E. Chernew, Zirui Song, and Laura A. Hatfield It is intuitive that changes in demand should lead to proportionate changes in resource use. However, this assumption is often questioned in healthcare markets. We use region-level data in combination with a long differences strategy to study the relationship between change in area population, on the one hand, and hospital capacity, utilization, and spending on the other. We focus on 2010 to 2019. We report several trends. First, overall inpatient utilization increased (decreased) faster than did population growth (loss) in most regions. However, increases in outpatient care tended to counterbalance declines in inpatient care. Second, a majority of regions lost beds in the decade preceding the outbreak of Covid-19. However, rural and low-volume designations likely attenuated the relationship between population and beds. Third, growth in total hospital costs increased about 20 percentage points faster than inflation. The subset of administrative costs increased even faster. We further observed that Medicare’s rural/low-volume policies were associated with differentially faster cost growth. Insurer market power had the opposite effect. Moving from the 25th to 75th percentile of region-level concentration in the 2010 insurance market was associated with 2 to 3 percent slower growth in costs. Taken together, the disconnects between change in resource use and change in population suggest suboptimal resource allocation within the hospital industry. Chapter 3, Hospital-Skilled Nursing Facility Integration and Participation in a Medicare Bundled Payment Initiative Joint with François de Brantes, David C. Grabowski, Laura A. Hatfield, Daniel Koppel, Robert E. Mechanic, Jennifer Perloff, and Michael E. Chernew Reduced use of institutional post-acute care has been the most important source of savings in alternative payment models. However, organizations may avoid participation in voluntary alternative payment models when participation jeopardizes their own revenue. This paper studied the relationship between hospital-skilled nursing facility (SNF) integration and hospital participation in the first year of Medicare’s Bundled Payments for Care Improvement (BPCI) Advanced initiative. We began by defining a hospital as SNF-integrated based on the presence of a common referral pattern and shared legal ownership. Using a matching strategy, we estimated the relative risk of participation across integration. We found that integration was associated with reduced participation in the joint replacement episode (relative risk = 0.25, 95% CI: 0.10, 0.62). Integration was not detectably associated with differences in participation for the hip and femur, sepsis, and stroke episodes. We conclude that other factors may be more consistent determinants of selection into voluntary payment reform. - ## Dissertation Committee Member [Laura Hatfield](/dissertation-committee-member/laura-hatfield) [Michael Chernew](/dissertation-committee-member/michael-chernew) [Zirui Song](/dissertation-committee-member/zirui-song) - ## Concentration [Methods for Policy Research](/conclabel/methodsforpolicyresearch) - ## Graduation Year [2024](/graduation-year/2024) - ## Role [Alumni](/people/alumni)