Medicare Home Health Reimbursement and Informal Home Care: Evidence from the Balanced Budget Act of 1997

Home health care spending in the US has more than doubled during the past decade. Medicare has been the primary payer for home health services. Although the 2015 Medicare Trustees Report reveals a somewhat optimistic financial future for the Medicare program, the program continues to face long-term budgetary challenges due to population aging. In recent years, lawmakers have recognized the role of family caregivers in the long-term care system and have regarded informal home care as a potential way to reduce public long-term care expenditures. Many states have taken actions to assist family caregivers, such as providing tax credits or deductions to caregivers and expanding respite care program. It is most certain that changes in government-provided home health benefits would affect how the elderly utilizes formal as well as informal home care.

The goal of this paper is to estimate the effect of Medicare home health benefits on the use of formal and informal home care. The estimation suffers an econometric problem: Medicare home health benefit is likely to correlate with unobserved factors that associate with individuals’ decision on home care use. To identify the causal effect, I use an instrumental variables strategy based on the Balanced Budget Act of 1997 (BBA97). BBA97 imposed a more restrictive reimbursement policy on home health agencies. It implemented the Interim Payment System (IPS) to replace the fee-for-service system in place to reimburse home health agencies. The IPS took effect in October 1997 and imposed a per-patient cost limit on agencies in addition to a per-visit cost limit. The per-patient cost limit was calculated as a weighted average of the 75% of the agency’s 1994 average per-patient costs and the 25% of the 1994 average per-patient cost in the agency's census division. Accordingly, the IPS created an exogenous variation in Medicare home health reimbursement rates across time and location.

Using a restricted data set linking the Second Longitudinal Study of Aging with the Medicare enrollment and claims records, I find that a 10% decrease in Medicare home health benefits reduces the probability of using formal home care by 9% (23% for formal home health services covered by Medicare) and reduces Medicare home health expenditures at the individual level by 50%. Informal home care provided by children is quite responsive to home health care benefits especially among the low-educated (the elasticity is -0.6 for the full sample and -1.2 for the low-educated sample). The findings also reveal that many elderly individuals replace formal with informal home care in response to a tighter home health policy. The increase in informal home care use, however, is not large enough to compensate the reduction in the proportion of using formal home care. Accordingly, a less generous policy may create access barriers among Medicare beneficiaries. For the low-educated, however, I find that a tighter policy may increase the proportion of individuals using any type of home care as some children would provide parental care after the policy change.