As our colleague Edwin Park has explained, there is keen interest on the part of some Congressional Republican leaders in capping federal Medicaid payments to states, ideally by converting Medicaid from a health insurance program into a block grant.  Currently, the federal government matches state spending for the costs of health and long-term care services needed by their low-income populations—children and families, pregnant women, people with disabilities, the elderly, and others—on an open-ended basis.  The purpose of a Medicaid block grant is to limit the federal government’s spending on services for these populations, shifting more and more of the costs to the states over time. 

To date, the arguments for a Medicaid block grant have not prevailed.  But there is a new kid on the block (so to speak): capping federal Medicaid matching payments to states can create “shared savings” that allow states to replace state dollars with federal dollars, freeing up state funds for other purposes. The poster child for this argument is Tennessee’s section 1115 demonstration, TennCare III.  This past August, Governor Lee announced that the state will use some of its “shared savings” to provide 100 diapers per month for low-income children under 2. And in October, the Governor announced that “shared savings” from TennCare III would be used to fund a $100 million program of no-interest disaster relief loans for counties in East Tennessee to repair infrastructure damaged by flooding from Hurricane Helene.

For many Medicaid block grant advocates, the TennCare III “shared savings” model is proof of concept.  Let the federal government cap the total amount of federal Medicaid funds you can receive in a year, and if you keep your spending under the cap, you can use some or all of the federal government’s “savings” however you want.  If this sounds too good to be true, that’s because it is too good to be true. 

The TennCare III demonstration is not now—and never was—a Medicaid block grant.  As the CMS Administrator in the first Trump Administration explained in her January 8, 2021 approval letter, the “aggregate cap approach” in TennCare III “is not a ‘block grant’ but rather, gives the state flexibility in operating its program under a defined cap by putting expenditures at risk based on both cost and population growth….”  And again:  “CMS…is not approving the ‘block grant’ financing approach the state described in its original application.”  These caveats didn’t stop the state from calling it “a Medicaid block grant waiver.”  But whatever TennCare III was called  when it was first approved— a block grant, an aggregate cap, or a windfall—it has been fundamentally changed. Following a lengthy renegotiation, the Biden administration in August of 2023 approved a revised version of TennCare III that no longer includes any aggregate cap. 

The TennCare III demonstration is voluntary; a Medicaid block grant is not.  Tennessee applied for and negotiated its “aggregate cap approach” with the first Trump administration.  It was under no obligation to accept the aggregate cap in the initial version of TennCare III, and it could have withdrawn at any point and returned to Medicaid’s regular open-ended federal matching arrangement.  Under a Medicaid block grant there’s no exit ramp. The federal government caps the matching funds it will pay to each state each year at a fixed dollar amount.  Up to that cap, the federal government will share in the cost of the state’s Medicaid program.  Beyond that amount, each state is responsible for 100 percent of its costs.  The cap amount is set based on federal budget considerations, not the needs of the state’s low-income populations. If a state doesn’t like the cap, the answer is what kids hear in preschool:  you get what you get, and you don’t get upset.

The purpose of a Medicaid block grant is to produce large and growing savings for the federal government, not to share those savings with the states.  The whole point of a block grant is to significantly reduce federal spending on Medicaid. The federal government is going to use the savings it gets from a Medicaid block grant to pay for tax cuts, for spending on other priorities, for deficit reduction, or for some combination of the above; it is not going to share any savings with the states.  And the savings will be real, because under a real block grant, the federal government will pay the states less than it would pay the states under the current program. Much less.

Medicaid spending is driven largely by enrollment growth and medical price inflation. Capping federal Medicaid spending does nothing to address these drivers; it just shifts more and more of the resulting costs to the states over time.  States will be left with only bad choices to deal with the shortfall in federal funding: raising taxes, cutting other parts of their budget like education, or, as is far more likely, limiting eligibility, reducing benefits, cutting payments to providers and managed care plans, and/or raising administrative barriers that make it more difficult to enroll or access needed services.

So what are Tennessee’s “shared savings,” and where did they come from?  From January 8, 2021 until August 4, 2023, the Trump administration’s version of TennCare III was in effect.  In that version, the state accepted an annual “aggregate cap” on federal Medicaid matching payments; in exchange, the state was allowed to keep up to 55 percent of the federal government’s “savings” if its total (federal and state) spending through the demonstration fell below the aggregate cap and it met certain quality targets. The state was allowed to spend the federal “savings” on specified state-funded programs (Designated State Investment Programs, or DSIPs), that were not otherwise qualified to receive federal Medicaid matching funds.  In effect, the demonstration allowed the state to replace the state dollars spent on the DSIPs with federal Medicaid funds, freeing up those state funds for other purposes.

Not surprisingly, the aggregate cap to which the state agreed to subject itself proved to be exceedingly generous to the state. The cap in the first year of the demonstration (CY 2021) was $10.509 billion (federal and state); the state’s spending through the TennCare III demonstration that year (federal and state) was $8.612 billion. The difference was $1.897 billion, or 122 percent of the state’s demonstration spending.  In the second year (CY 2022), the cap was $11.360 billion, and the state’s demonstration spending was $9.855 billion.  The difference was $1.505 billion, or 115 percent of the state’s demonstration spending. 

Source:  CCF calculations based upon Tennessee Department of Finance & Administration, Division of TennCare, TennCare III Demonstration Annual Monitoring Report (June 30, 2024), Appendix M.

Aggregate caps that are set at levels 15 to 22 percent higher than actual spending are, in a word, illusory.  Except in Tennessee’s case, they were the gateway to a real windfall.  Under the formula in the demonstration, the state’s “shared savings” in the first year of the demonstration were 45 percent of the $1.897 billion difference between the cap and actual spending, or $854 million.  The state reported its spending on the specified DSIP programs for that year to be $457 million.  This was the amount of “shared savings” that the state claimed from the federal government.  In the second year of the demonstration (CY 2022), the state claimed “shared savings” from the federal government (in the form of its spending on DSIPs that year) of $473 million.

Over the two years, the amount of “shared savings” claimed by the state totaled $930 million.  Due to ambiguity in the approval documents, there’s a question as to whether the state received the entire $930 million from the federal government or whether that amount was matched by the federal government at the state’s matching rate of about 72 percent, for a gain to the state of approximately $6 million.  Either way, it’s a lot of “savings” for the federal government to “share.”  What there’s no ambiguity about is that the state is allowed to use the state funds it spent on the specified DSIP programs that were replaced by the federal Medicaid funds—whether $670 million or $930 million—for whatever purpose it wants. As indicated above, that is exactly what Governor Lee is doing.

So after paying this amount to Tennessee, how much, if anything, did the federal government actually save?  It’s hard to know.  One thing is clear, however.  For each of the calendar years (2021 and 2022) that the “shared savings” arrangement was in place, federal Medicaid payments to Tennessee increased.  In fact, according to MACPAC, federal Medicaid payments to Tennessee increased every year from FY 2017 through FY 2023, from $6.3 billion to $9.5 billion. (See table attached).  There are many factors that explain this 51 percent increase in federal spending over 7 years—including the $670 or $930 million in “shared savings” paid to the state in CY 2021 and CY 2022. But there’s no evidence that the “shared savings” arrangement reduced annual federal Medicaid matching payments to Tennessee.

As a false narrative on Medicaid block grants, the TennCare III “shared savings” model has much in common with Rhode Island’s “Global Medicaid Waiver,” which was approved by the Bush Administration in 2009. That 1115 demonstration also had an inflated aggregate cap, it also allowed the state to collect federal matching funds on state-only programs, and it was also cited by block grant proponents as a model for other states. (Rhode Island abandoned the cap in 2013).

Bottom line:  the TennCare III “shared savings” demonstration is not a Medicaid block grant. A real Medicaid block grant will not allow federal funds to increase by 51 percent over the next six years to Tennessee or any other state. (CBO projects that federal Medicaid spending will increase by 30 percent from FY 2026 to FY 2031; to save money, a cap would have to limit growth in federal Medicaid spending below 30 percent.)  A real Medicaid block grant will reduce federal funds to states, forcing states to choose between raising taxes, robbing other state programs, or cutting Medicaid eligibility, benefits, or provider payments.  In short, a real Medicaid block grant, with a real cap that produces real savings for the federal government, will leave states holding the bag—not sharing the savings.