CBO Likely Underestimates Medicaid Coverage losses
Looking at the impact of Mike Johnson and the House proposal of cutting Medicaid will have an impact. Meanwhile, Trump is expecting cuts so as to keep the 2017 tax cuts of which a major portion went to the upper one-percent in income.
Estimate “seems low” — KFF’s Larry Levitt, Brookings‘ Matt Fiedler, and Georgetown’s Edwin Park agree.
“CBO likely underestimates coverage losses if Republicans defund the ACA Medicaid expansion”
– by Andrew Sprung
As Republicans in the House and Senate continue to tussle over various options for cutting hundreds of billions of dollars in federal Medicaid spending, the Congressional Budget Office has dropped a gift for those who favor defunding the ACA Medicaid expansion.
The ACA expansion, which the Supreme Court made optional for states, opens Medicaid to adults with income below 138% of the Federal Poverty Level. To make that coverage affordable for states, the ACA established a permanent 90% Federal Medical Assistance Percentage (FMAP) for those rendered eligible by the expansion. For other Medicaid eligibility categories — children, Aged, Blind and Disabled, and low-income seniors, the federal FMAP ranges by state from 50-77%, varying according to state per capita income.
In a report released yesterday estimating the fiscal and coverage effects of five Medicaid cost-cutting options Republicans are considering,* CBO produced surprisingly low estimates of the coverage losses that would result from Republicans defunding the ACA Medicaid expansion.
Of the five cost-cutting options CMS considered,* I want to focus first and foremost on two of them:
- Option 1 would repeal the ACA expansion’s 90% Federal Medical Assistance Percentage (FMAP), paying only each state’s match rate for all other populations (ranging from 50% to 77%) for the expansion population (adults with income below 138% FPL).
- Option 4 would impose a per-enrollee inflation cap on federal spending for the ACA expansion population alone. KFF estimates that such a “per capita cap, using the most likely inflation measure (CPI-U plus 0.4%), would ratchet the expansion FMAP down to 69% by 2034.
According to CMS’s most recent tally (June 2024), the federal government pays the 90% FMAP for 20.9 million adults rendered eligible by expansion criteria (income under 138% of the Federal Poverty Level).**
CBO estimates that repealing the expansion’s 90% FMAP outright would reduce Medicaid enrollment by 5.5 million and increase the uninsured population by 2.4 million. Imposing per capita caps on the expansion population would reduce enrollment by 3.3 million and increase the ranks of the uninsured by 1.5 million.
Those estimates assume that most states will not end eligibility for the expansion population — whether quickly, if the 90% FMAP is repealed outright, or somewhat more slowly, if it’s eroded over time by per capita caps. That’s surprising.
The estimates “do seem low,” Larry Levitt of KFF told me in an email. Matt Fiedler of the Brookings Institute agreed. “I think they are on the low side. I think I would bet on states cutting back more aggressively, although I think it’s also fair to say that there is a lot of uncertainty here.” Edwin Park of Georgetown University wrote, “CBO is likely overly conservative here because it’s unlikely that states in reality would be able to replace half of the cost shift in the face of all of these cuts.”
Park’s last point is key. Explaining the basis of its estimates, CBO spells out its assumption that “on average, states would replace roughly half of the reduced funds with their own resources.”
That assumption raises expert eyebrows. By CBO’s estimate, eliminating the ACA’s 90% match rate for the expansion population would reduce federal payment to states for expansion enrollment by $860 billion over ten years. That total includes the effects of states reducing benefits and provider payments. The ten-year federal payment reduction before state response is estimated at $516 billion.
All states but Alaska are required by their Constitutions or state law to balance their budgets. By KFF’s estimate, the 40 states (plus D.C.) that have enacted the expansion collectively would have to spend $626 billion over ten years to make up the federal shortfall resulting from repeal of the ACA’s 90% FMAP. New Jersey has estimated the annual cost of FMAP to the state budget at $2.2 billion (in an annual budget of $56 billion). As anyone who’s followed any state’s annual budget deliberations can testify, states don’t raise billions in new revenue lightly. It is highly questionable whether any state would be able to maintain the expansion in its entirety at the state’s own FMAP (especially wealthy states, where the FMAP is just 50% or slightly higher). On May 1, prior to the CBO report release, Park assessed the odds as follows:
Faced with such massive cost shifts under these proposals alone [90% FMAP repeal or per capital caps for the expansion group], states would either have to dramatically raise taxes, cut other parts of their budget like education, deeply cut the rest of their Medicaid program, or as is most likely, eventually drop the expansion.
It’s possible that a handful of determined Democratic-led states would maintain the expansion for enrollees with income up to 100% FPL (as Wisconsin does at present, without the 90% FMAP), which would place those in the 100-138% FPL bracket into marketplace coverage, for which the federal government pays 100% of the cost. But those blue states are also mostly wealthy states with low FMAPs — e.g., 50% for California, New York, Illinois, and others.
Under the assumption that no state would maintain coverage to 138% FPL, enrollment declines would likely top 15 million.
In fact, 12 states have “trigger” laws requiring them to end the expansion if the expansion FMAP is reduced. While there is some ambiguity and flexibility in how some states might interpret those triggers, Georgetown’s Adam Searing notes that nine of them would all but certainly have to terminate coverage immediately if the expansion’s 90% FMAP is repealed. Should all of them do so, the Center for American Progress has tallied the enrollment losses at 3.6 million. (Subtract the three states with more flexible triggers — Idaho, Iowa and New Mexico — and the total is 3.1 million.) Under CBO’s estimate, that would suggest coverage losses in 28 non-triggered states of just 2-odd million.
It seems unlikely that losses would be that low. CBO’s estimates virtually assume that the expansion would be maintained in high-population states such as California (which has 5 million expansion enrollees), New York (2.1 million), Illinois (843,000), and Pennsylvania (832,000).
Per capita caps
Capping payments for the expansion population would ratchet down the expansion FMAP annually rather than in one fell swoop. As mentioned above, KFF estimates that the FMAP would drop to 69% over ten years — and, I would add, would continue to drop in a second decade, assuming that any policy in the U.S. remains stable for that long. Because the damage is incremental, the ten-year cost deficit reduction estimate is far lower ($225 billion vs. $710 billion) — but the ultimate effect on coverage should be more or less the same (or worse in future decades, if the caps are not removed). It’s hard to imagine states maintaining coverage to 138% FPL with an FMAP under 70%.
Other options
CBO Option 3, imposing per capita caps on federal spending for the entire Medicaid population, appears to be off the table. As to Option 5, repeal of an array of Biden administration administrative measures designed to reduce friction in the application process and churn in enrollment, I have no comment, except that this repeal would be (will be, sigh) very unfortunate. Reducing administrative barriers to enrollment is a slow boring of hard boards.
Provider tax wipeout?
Option 2, limiting state taxes on health care providers (or rather, limiting payback to those providers) bears some consideration.
Provider taxes are a financial maneuver through which states plus up their federal Medicaid contribution. If a tax on a provider class does not exceed 6% of the provider’s net revenues, the state can essentially give the tax dollars back to the provider group in the form of higher payments — and receive its federal FMAP (ranging by state from 50-77%) for the extra payments.
Right-wing Paragon Health Institute head Brain Blase, deploying Gingrichian rhetoric, calls these maneuvers “money laundering.” In fact they’re the kind of kludge that state-federal partnerships and U.S. political sclerosis routinely generate: a workaround to compensate for chronic underfinancing of Medicaid, which keeps payment rates below Medicare rates and below cost for providers in many categories and places. States must propose such arrangements to CMS before implementing them, and CMS must assess the proposal and approve it. These taxes are legal, and states depend on them.
Repealing states’ ability to hold the taxed entities harmless — that is, to essentially pay the tax back, largely with federal dollars — would be very expensive for states. CBO estimates the deficit reduction effect of complete repeal of the ‘hold harmless’ option at $668 billion over ten years — nearly the same as savings from repealing the ACA FMAP ($710 billion). The gross reduction in federal outlays is slightly higher than for ACA FMAP repeal. The coverage loss estimate also is higher — 8.8 million.
The catch is that a complete wipeout of the provider payment option is highly unlikely. A more realistic option is to reduce the threshold (“safe harbor”) under which the taxed entities can be held harmless from the current 6% of revenues. A prior CBO estimate pegs the 10-year federal savings from reducing the safe harbor threshold to 2.5% at $241 billion. There would be large variation in how this measure would affect states, as the number of provider taxes and thresholds states use varies widely.
In connection with the provider taxes, Park’s analysis of the likely effects of Republican proposals makes two important points. First, a dozen states by Park’s count fund their 10% share of the ACA expansion cost via provider taxes. “Restricting provider taxes,” Park writes, “could by itself prevent some expansion states from continuing to directly rely on this state financing source for the expansion. If such states were unable to identify other revenues, they would have no choice but to eliminate their expansion.”
Second, as the last point illustrates, any combination of cuts will have a cumulative effect. As Park put it in an email to me, “With multiple cost-shifts, it’s hard to see how states can compensate for any of them in combination.” CBO perhaps had to consider the five options for which Wyden and Pallone requested analysis (see note at bottom) in isolation, because they are not part of an actual bill, and no one knows which, if any, Republicans will put into legislation. But that limitation will likely limit stakeholders’ perception of the damage these cuts may inflict as they consider CBO’s projections.
* CMS considered these five options at the request of Democrats Ron Wyden, ranking member of the Senate Finance Committee, and Frank Pallone, ranking member of the House Energy & Commerce Committee.
** Of those, 4.3 million would have been eligible under pre-ACA eligibility criteria in a handful of states that obtained waivers to expand coverage, with the largest numbers in New York (1.8 million, Puerto Rico (634,000), Massachusetts (393,000) and Louisiana (216,000).