Medicaid beneficiaries will start working in droves. Few sick Americans will lose coverage. The government will generate huge savings.
This is how the CMS expects new Medicaid work requirements to play out. But those assumptions, outlined in a rule this summer that created implementation guidelines for states, are fundamentally defective, according to a new analysis from nonpartisan think tank the Brookings Institution.
The Trump administration’s analysis of the rule’s impact on beneficiaries and taxpayers isn’t based on actual data, according to the report. Instead, it relies on hypotheticals meant to make the rule look like it’ll benefit Medicaid beneficiaries and the federal government, with only a minor tradeoff on enrollment losses — and ignores evidence that the rule could do more harm than good.
The CMS made “complete random assumptions. They didn’t provide any — zero — justification for their assumptions,” Sherry Glied, a senior fellow with Brookings’ Center on Health Policy and an author of the report, said in an interview.
“It was patently obvious,” Richard Frank, the center’s director, said. “If you’ve been around health economics for a while and read that literature, it was just totally clear that these guys were way off the reservation.”
Shaky assumptions
The GOP’s One Big Beautiful Bill passed last summer created a first-of-its-kind national mandate requiring Medicaid beneficiaries to work, volunteer or attend school for 80 hours a month in order to stay enrolled in the safety-net insurance program. The policy, which will kick in at the start of next year, applies to the roughly 21 million people who gained Medicaid eligibility through the Affordable Care Act expansion.
Conservatives believe that tying Medicaid eligibility to employment will help prevent fraud and get more Americans into jobs, boosting their wellbeing and that of their commmunities. The neediest and sickest Americans shouldn’t be affected by work requirements, given medically frail Americans, caregivers and some other groups are meant to be exempt, supporters argue.
But the work requirements have sparked concern and condemnation from Democrats, patient advocates, clinical groups and health policy researchers, given evidence that work requirements don’t increase employment and cause many eligible individuals to lose coverage due to issues documenting compliance.
Those concerns intensified in June, after the CMS released a rule that created unexpectedly high hurdles for sick beneficiaries to prove the mandate doesn’t apply to them. That could drive up the number of eligible Medicaid beneficiaries who improperly lose coverage, according to experts.
But CMS actuaries offered surprisingly rosy assumptions about the policy’s impact — including that 29% of affected beneficiaries will move into qualifying activities, like becoming employed.
That assumption is more than six times larger than the largest estimate in any line of research the CMS could have used to estimate work requirements’ impact — including from studies of welfare-to-work programs, studies of the ACA’s Medicaid expansion and research on other work mandates, including experiments in the states, according to the new Brookings report.
“This is an old trick of the actuaries. When in doubt, when there is no information, just make stuff up. But in this case, there was information,” Frank said.
The assumption that beneficiaries would gain jobs had major implications on other estimates in the rule, including the CMS’ prediction that only 3.1 million to 3.3 million people will lose Medicaid over the next decade as a result of the policy.
That’s well below estimates from nonpartisan congressional scorekeeper the Congressional Budget Office, which forecasts that roughly 5.7 million people will lose Medicaid by 2034.
And despite CMS’ lower disenrollment estimates, the agency forecasts higher federal savings: The agency expects the government to save $350.3 billion over the next decade, 10% higher than the CBO’s best guess.
Those figures don’t make sense, according to Glied and Frank. If a relatively small group of people lose coverage, savings should also be smaller. In addition, research shows that people who leave Medicaid when administrative requirements increase are generally healthier, and therefore drive less spending — another factor that should tamp down projected savings.
“What they're implicitly saying is that the people who are going to be kicked off Medicaid because of this rule are actually going to be the really sickest people on the program, which says something about their exemption structure,” Glied said. “And it’s not even clear that there are enough people for them to kick off the program to make the money that they want to make.”
“Complete nonsense”
It’s not unusual for presidential administrations to toy with data to support their political goals. But the Trump administration has been particularly criticized for twisting or ignoring research, including in health and scientific policy.
Glied and Frank pointed to a report from the HHS published earlier this summer that reviewed existing literature to argue for the benefits of work requirements. That report was also faulted as having unrealistic assumptions.
It was “complete nonsense,” Glied said. “And then they went even further.”
The CMS declined to comment for this story. But agency officials have defended the rule as a commonsense implementation of the Big Beautiful Bill’s work requirements.
Individuals who are meant to be exempt should easily be able to prove it, and states have options to make that happen, Medicaid Director Dan Brillman told reporters in June.
More than two dozen states are suing the Trump administration over the rule, arguing it’s significantly less flexible than the CMS’ earlier guidance and will make it harder for medically frail people to prove they’re exempt.
A judge ruled in July that the regulation would remain in effect as the case moves forward.
Meanwhile, some Republican-led states are enforcing even stricter Medicaid work stipulations than required. According to Politico, at least eight states are eschewing a one-year grace period and asking beneficiaries to prove their eligibility for an exemption earlier than the rule asks.
