Association for Public Policy Analysis & Management

Medicare reimbursement policy affects the access, amount, and quality of services provided to Medicare beneficiaries, and may induce changes in the care of Non-Medicare beneficiaries given that health care providers serve patients from different insurers. Despite the existence of spillover effects, policy debates and existent literature on the effects of Medicare's reimbursement policy tend to focus on Medicare beneficiaries only. This study aims at broadening our understanding of the interaction between Medicare and Medicaid using as context the nursing homes' market. The question is of particular interest in this setting as the U.S. nursing home market has experienced a substantial reduction in the utilization rate and a major shift in focus from long-term care, covered by Medicaid, toward short-term care, covered by Medicare.

Demographic and policy forces different from Medicare can contribute to the mentioned trends, in particular the increase in assisted living and the expansion of Medicaid Home and Community Services could also cause a decline in the utilization rate of nursing homes. Therefore, to isolate the independent effect of Medicare reimbursement policy I use as natural experiment the increase in Medicare fees mandated in the Medicare, Medicaid, and SCHIP Balanced Budget Refinement Act (BBRA) of 1999, and the Medicare, Medicaid, and SCHIP Benefits Improvement and Protection Act (BIPA) of 2000. The research hypothesis is that in a market with excess demand an increase in the relative generosity of the Medicare fee with respect to the Medicaid fee, will encourage facilities to substitute Medicaid with Medicare patients, as health care providers want to maximize their profits.

This study uses a difference--in--difference identification strategy to isolate the substitution effect. The treatment group is the facilities that previous to the price shock were working near full capacity, and therefore had a disproportional incentive to substitute Medicaid with Medicare patients. The findings suggest that, as expected, only For--Profit institutions responded to the change in financial incentives. In particular, I find that after Medicare increased its fee For--Profit nursing homes decreased on average the share of Medicaid admissions in 1.36 percentage points, from an average of 27.9% to 26.57%, decreasing the number of Medicaid admissions by about 3,385 patients in the treated facilities. I do not find an statistically significant change in the share of Medicaid admissions in Not--For--Profit facilities, which provides more evidence that the estimated effect for For--Profit is capturing supply--side responses rather than demand--side responses.