The Expansion of Home Care Benefits in Medicaid Does Not Save Money on Long-Term Care

By Mark J. Warshawsky of AEI.

"It is claimed by advocates that the provision and expansion of home-and-community-based services (HCBS) care by states through Medicaid to the eligible elderly and disabled populations will save the federal and state governments lots of money on nursing home care because the average nursing home daily cost is triple that of average home care. This rhetoric can be simply challenged by noting that for similarly situated severely disabled individuals, say those afflicted with Alzheimer’s disease needing full-time care, surely the economies of scale and scope of nursing home care produce a cost much lower than paying for 24/7/52 personal aides in the home.

But more nuanced and sophisticated claims are also being made, based on a new study by professors McGarry and Grabowski (MG) funded by the AARP, that, while not dollar for dollar, savings can be enjoyed through the expansion of HCBS. In particular, the study purports to find that a $1 increase in HCBS spending is associated with a $0.74 increase in total long-term services and supports (LTSS) spending, suggesting that each dollar directed to HCBS was offset by $0.26 in savings from decreased nursing home use. Upon closer examination, however, the assertions are doubtfully supported.

Piecing together state-year data for 1999 to 2017 from various government and survey sources, MG estimate regressions comparing outcomes in states that expanded Medicaid HCBS aggressively versus those that expanded more modestly, focusing on the aged and disabled populations. Some basic controls were included in the regressions by state and year, such as the size of the older population, average economic status, and population density, as well as state and year fixed effects. The data, however, has several severe problems.

Five states, including California, which has pushed HCBS expansion hard, are missing entirely from the regressions because their data quality was particularly poor, many anomalous variable readings had to be deleted and substitute data interpolated, and enrollment in and spending on managed LTSS and waiver programs was mostly missing or interpolated. Moreover, and more significantly, the possibility of several biases in the results is present. As MG themselves note, “. . . it is not random as to which states have expanded Medicaid HCBS more aggressively. The results we observed for these more aggressive states may not apply to the states that have yet to expand.” For example, states that expanded HCBS may have been reacting to shortages of nursing home spaces.

But even for the expanding states, there may be other biases that call the results into question. States may have been making policy or administrative changes in their Medicaid eligibility rules, perhaps for budget reasons, expanding or contracting claim allowances, which may be related intentionally or inadvertently with state policy changes in HCBS provision. Related, the Social Security Administration, which handles claims at the federal level for Supplemental Security Insurance, which in turn automatically qualifies disabled and aged individuals for Medicaid in most states, has loosened and tightened the eligibility standards for claims over time, and this impact may differ by state. MG use Medicare data to establish a base population, but, as people work longer, they increasingly delay claiming Medicare and this effect too may differ by state and over time.

Overall, the many problems and issues with the data and methodology of the MG study suggest that state and federal policymakers and analysts scoring the costs of HCBS policy changes, such as reprises of the poorly designed and ill-fated proposal of the Biden administration to infuse hundreds of billions dollars of federal resources into HCBS, should just assume that a dollar change costs a dollar."