On June 22, Senate leadership released its proposed substitute for the House-passed American Health Care Act (AHCA), the Better Care Reconciliation Act of 2017 (BCRA), as a discussion draft. Like the House-passed AHCA, the Senate bill imposes a per capita cap for virtually all Medicaid beneficiaries, but it includes a new exclusion for children who qualify for Medicaid based on disability. Notably, the Senate bill tightens the cap relative to the House-passed bill, meaning that states will have to further reduce their Medicaid spending for all populations in order to keep spending below the cap. The other major new feature of the Senate’s per capita cap allows the federal government to raise caps for the lowest-spending states by lowering caps for the highest-spending states, with a requirement that the adjustment not result in a net increase in federal spending.

The new redistribution provision likely was added to the Senate bill to mitigate the concern among traditionally low-spending states that their spending levels are “baked” into the formula for future Medicaid spending under a per capita cap, putting them at a permanent disadvantage relative to their higher-spending counterparts. Freezing in historical spending is typically a design feature of capped funding. In practice, though, our preliminary analysis indicates that this new provision may not work as intended and, in fact, could harm some of the states it is designed to help, hurt states with higher healthcare costs or that serve a higher proportion of people with significant healthcare needs (e.g., a state with an older population) and create even more uncertainty for states.

Redistribution Provision

Under BCRA, if a state’s per capita Medicaid spending is above the national mean by 25 percent or more, its per capita cap will be decreased the following year by between 0.5 percent and 2 percent. If a state is below the mean by 25 percent or more, its cap will be increased between 0.5 percent and 2 percent. The Secretary of Health and Human Services (HHS) makes the adjustments and in doing so must assure that the redistribution does not cost the federal government any additional money. Certain states are exempt from this provision; BCRA specifies that it does not apply to states with a population density of less than 15 individuals per square mile, which according to the latest Census Bureau data means that Alaska, Montana, North Dakota, South Dakota and Wyoming would be exempt. In FY 2020 and FY 2021, the adjustment will be based on a state’s average per capita spending across all eligibility groups, but, in FY 2022 and beyond, the adjustments are made separately for each eligibility group. As a result, a state could find that its per capita cap is adjusted downward for one eligibility group, but upward for another.

Preliminary Analysis

Using the Manatt Medicaid Financing Model, we examined what the new policy might mean for states, identifying which ones could get “caught” by the new policy that would lower their cap and which ones would potentially receive an increase that lowers the per capita cap cuts they are likely to face. In practice, though, a key feature of the new policy is that each state will be judged annually based on its actual spending relative to the spending of all other states. It is quite possible, indeed likely, that a different group of states than we identify here will be affected by the new policy. This is because state spending will evolve for any number of reasons, but also because this provision creates a strong incentive for all states to “out cut” their counterparts in the years ahead to avoid this new penalty.

With the caveat that this analysis does not take into account that states’ current spending patterns will change due to the per capita cap and the redistribution policy (indeed no analysis could make those projections with any certainty), our preliminary analysis indicates the following:

  • Nearly twice as many states would be losers under the redistribution provision as would gain from it. As shown in Table 1, if states were to maintain their current per capita spending patterns through FY 2019 (the first year of data used in redistribution calculations), 20 states would be hit by a tighter cap either because their average per capita spending is high (which is the metric used in FY 2020 and FY 2021) or their per capita spending in one or more individual eligibility groups is high (which applies in FY 2022 and beyond). In comparison, only 12 states would experience some relief from the per capita cap cuts because they have expected spending that is low on average for their overall populations or for any of the individual eligibility groups. Under the test applied in FY 2022 and beyond, five states—Alabama, Colorado, Iowa, Kentucky and Wisconsin—meet the threshold for both “high” and “low” spending state, making it difficult to say whether they would be harmed or helped by the policy.
  • Many of the affected states are not the “usual suspects.” States such as Alabama, Maine, Indiana, Iowa, Kansas, Kentucky, Maine, Minnesota, Missouri and Texas could end up losers under the provision. Although often thought of as low-spending, many of these states have relatively high spending in at least one eligibility category. This may well be because they have opted only to cover a narrow set of beneficiaries with intensive needs (e.g., only people in need of long-term care rather than a broader group of low-income seniors).
  • Even states that are not high spenders for any one eligibility group could be losers under the policy. In Maine, the state actually spends below or near the national average for each eligibility group, but its average per capita spending across all eligibility groups exceeds the national average by more than 25 percent, potentially subjecting it to the penalty in FY 2021 and 2022. This is because, relative to other states, its caseload is weighted more heavily to seniors and people with disabilities than to children and low-income adults, not because its spending would be considered “high” in a traditional sense.
  • No allowance for appropriate variation in per capita spending. The provision makes no allowance or adjustment for the many facts that might drive a state to spend considerably more than the national average. States may end up classified as high spending simply because they have a higher cost of living, provide a more robust benefit package to beneficiaries, have an older population, cover a higher need group of beneficiaries, or reimburse at higher rates to address access issues.
  • All states—even those that face a lower cut as a result of the new policy—will face major new uncertainty. Even states that might gain some relief from the per capita cap cuts under the new provision will face considerable new operational challenges. States won’t know whether they are winners or losers for a given fiscal year until after that fiscal year starts. This is because the adjustment up or down is based on a state’s actual spending in the preceding year relative to that of other states, which HHS will not be able to determine for many months. Even without this new provision, states will not know under a per capita cap what their federal Medicaid funds will be until after a federal fiscal year ends, but this new provision will make it even more difficult for state leaders to make decisions about Medicaid.

As states and other stakeholders evaluate the draft version of BCRA, it will be important to include an in-depth assessment of the new redistribution policy. Our early look suggests it may not work as intended; indeed, it could harm a number of states that are usually thought to be low spending; introduce major new operational issues; and create a “race to the bottom” as states seek to avoid the extra penalty it imposes by cutting more than others.

Table 1. “High” and “Low” States Potentially Affected by BCRA Redistribution Based on FY 2019 Projected Medicaid Spending Per Enrollee (Sorted by State Difference from U.S. Average)

Table 2. “High” and “Low” States Potentially Affected by BCRA Redistribution Based on FY 2019 Projected Medicaid Spending Per Enrollee (Sorted by State Name)