Are 4.8 Million People on Medicaid ‘Cheating the System’?

The Washington Post

15:11 JST, June 2, 2025

“You’ve got about 4.8 million people on Medicaid right now nationwide who are able-bodied workers, young men, for example, who are not working, who are taking advantage of the system. If you are able to work and you refuse to do so, you are defrauding the system. You’re cheating the system.”

– House Speaker Mike Johnson (R-Louisiana), interviewed on CBS’s “Face the Nation,” May 25

With the passage of the One Big Beautiful Bill in the House, which included new work requirements for Medicaid, the health-care program for the poor, Republicans led by Johnson made the case that the legislation was necessary because there were nearly 5 million freeloaders in the system – an example of what Johnson called “fraud, waste and abuse” that is draining federal coffers. Because the bill also includes tax cuts, the Congressional Budget Office says it will boost federal budget deficits by more than $2 trillion over 10 years, a finding that has irritated Johnson – but that’s another story.

We became interested in the origin of the 4.8 million figure. It didn’t track with a February study by KFF, a nonprofit health-policy organization, on the 26.1 million people who used Medicaid. Most people who did not work were caregivers, parents, students or had a disability, according to 2024 Current Population Survey data analyzed by KFF. Only 8 percent fell in a category of “not working due to retirement, inability to find work, or other reason” – which adds up to only about 2 million people. So Johnson’s figure is more than double.

The speaker’s office directed us to the House Energy and Commerce Committee as the source of the figure. The committee provided documentation that the number was generated by the CBO. And therein lies a tale.

The facts

The CBO is a nonpartisan committee staffed by dedicated professionals. But they are bound by the requests made to them by members of Congress.

This month, for instance, Democrats claimed that 13.7 million Americans would lose health insurance because of the House bill, citing the CBO. As the New York Times reported, Democrats came up with that figure by also asking the CBO for the impact of a related health-care policy that is due to expire at the end of the year (as originally designed by Democrats). The big number earned headlines – but it was misleading.

The figure of 4.8 million promoted by Johnson requires some interpretation from technical language. The committee provided the CBO language that was the source of the figure. In a May 12 email, a CBO official wrote: “We plan to communicate that on a stand-alone basis, work requirements reduces the deficit by $300.8 billion, reduces Medicaid coverage by 5.2 million, and increases uninsured by 4.8 million; however, effects on the deficit and coverage are smaller when interacted with other Medicaid policies.”

“Increases uninsured” then was crafted into the talking point of “adults who can work but are refusing to.” But that doesn’t necessarily mean the same thing.

Medicaid is funded by states and the federal government. Under the bill, states would be required to mandate that certain Medicaid recipients work at least 20 hours a week – or participate in “community engagement” activities like volunteering – to keep their coverage.

The CBO wrote of 5.2 million people losing Medicaid coverage because of the new rules. But apparently 400,000 would gain insurance in other ways, leaving 4.8 million without insurance. (It also suggested that the numbers would be smaller when combined with other policies in the bill.) Matt VanHyfte, the House Energy and Commerce Committee’s communications director, argued that the bar for work requirement is low enough that it’s clear these people refused to work.

VanHyfte provided a list of the exemptions to work requirements given to the CBO when doing its calculations. One was “under age 19 or age 65 or older (or a former foster youth under the age of 26).” But, in 2023, when the House GOP pushed a different Medicaid bill with work requirements, it also exempted people 55 to 64.

According to KFF’s research, less than half of the people in that age group have jobs. Jennifer Tolbert, deputy director of KFF’s program on Medicaid and the uninsured, said that people in this age range have higher rates of disability or may be retired. (Social Security benefits can be received as early as 62.)

Republicans haven’t emphasized this shift in policy – which we suspect could be a factor in the CBO’s estimate of a big increase in uninsured people – and instead have focused, as Johnson did on “Face the Nation,” on “young men” gaming the system. In April, Johnson colorfully put it: “You return the dignity of work to young men who need to be out working instead of playing video games all day.”

But that’s a relatively small universe. Jim Stimpson, a health economics professor at UT Southwestern Medical System, crunched the numbers in the 2023 American Community Survey for us and calculated that about 1.65 million men ages 19 to 25 are unmarried and unemployed while on Medicaid. (Almost 1.5 million men in that age group are employed.) So even if each of those men lost their access to Medicaid, there’s supposedly an additional 3.2 million people who also lost their coverage.

VanHyfte acknowledged adding work requirements for the 55-to-64 age group is big shift from 2023. But he said the policy is drafted to “more explicitly exempt people with disabilities on a broader scale.”

The CBO has not yet released a full report explaining its estimates or assumptions; in fact, the figure cited by Republicans is preliminary and does not reflect the final legislative language as passed by Congress. VanHyfte said, however, he did not expect the number to change.

The 2023 bill was modeled on an Arkansas work requirement program that, according to a 2020 Harvard study published in Health Affairs, led to a drop in the number of insured while not increasing the number of people who worked because of burdensome paperwork requirements. When the CBO evaluated the 2023 bill, it calculated 1.5 million people would lose health insurance, relying in part on the Arkansas experience.

By contrast, the Medicaid provisions of the new bill is modeled after a Georgia program called Pathways to Coverage and requires verification of work at the time of enrollment. VanHyfte said the CBO studied the effect of the Georgia program when doing its analysis. “The CBO score for the 2023 bill cannot be used to compare the two because they are fundamentally two different policies,” VanHyfte said.

One major change is that the 2023 bill allowed people to reenroll in Medicaid after a three-month lockout period, even if they still weren’t meeting the work requirements. In contrast, the 2025 House bill allows individuals to reenroll only if they do meet the requirements – which can include working at least 80 hours a month (or earning the equivalent of minimum wage for that time), volunteering or attending school part-time.

“The bar is low and if folks are uninsured, it’s because they chose not to comply,” VanHyfte said. “Given all the exemptions, notices and grace periods, the people who would not comply with the requirements would be those who refuse to work.”

The bottom line

All estimates depend on assumptions. The CBO determined that 4.8 million would lose Medicaid insurance if the House bill became law. The House GOP has concluded that this means these people refuse to work and are therefore “cheating the system.”

Without access to the full CBO analysis, that’s a big assumption. The CBO appeared to predict what might happen under the new law – not what people are doing now. But it’s ironic that Johnson is relying on the CBO for this estimate when he’s also attacking the agency for its deficit forecast for the same bill.