Unintended Consequences of Reimbursement Schemes in Health Care

Abstract

The three essays in this dissertation explore the impact of various reimbursement structures and chain ownership in the context of U.S. health care. In the first essay, I study how the moral hazard created by Medicare part D reimbursement structure affects coverage choices made by participating insurers. Using data on formularies for the universe of U.S. health insurance plans, I find that after rival entry non-Medicare plans reduce coverage of reference products by 20-25% more than Medicare plans. If Medicare Part D plans emulated non-Medicare plans, the federal government and patients could save around 6% annually. Finally, a 2018 policy change had only a small effect on the take-up of low-cost entrants. In the second essay, co-authored with Paul J. Eliason, Benjamin Heebsh, Riley J. League, Ryan C. McDevitt and James W. Roberts, we study how health insurers game pay-for-performance reimbursement schemes. We use annual variation in Medicare's criteria for its Quality Incentive Program in dialysis to distinguish strategic patient dropping from the provision of higher-quality care. We find that patients whose characteristics would trigger penalties are 40% more likely to switch facilities, while along some dimensions facilities exert more effort to provide better care. In the third essay, co-authored with Paul J. Eliason, Ryan C. McDevitt, James W. Roberts, Vincenzo Villani, Gabriel Butler, Nicole DePasquale, Christine Park and Lisa M. McElroy, we try to identify sociodemographic and facility characteristics associated with not receiving transplant information. We find that non-informed patients are more likely to be older, female, and on Medicare. Patients at chain facilities and independent facilities acquired by chains are more likely to receive information, but this does not translate into higher waitlist or transplantation rates. The chains' objective of profit maximization may bring facilities to disregard patient outcomes.