A person fills out paperwork to document compliance with the new Medicaid work requirements — a process that begins January 1, 2027 where states must verify at application and periodically thereafter to maintain benefit eligibility.
getty
On June 1, the Centers for Medicare and Medicaid Services (CMS) published the interim final rule that converts the Medicaid work requirements provision of 2025’s “One Big Beautiful Bill Act” from statute into policy. Starting January 1, 2027, roughly 20 million adults in the Affordable Care Act’s Medicaid expansion population — non-pregnant adults ages 19 to 64, across 43 states and the District of Columbia — will have to show 80 hours a month of qualifying activity to keep their coverage. Employment counts, as does half-time school, job training, an approved work program, community service or any combination. Earning at least $580 in a month, which is 80 hours at the federal minimum wage, satisfies it as well.
Many are excluded: pregnant and postpartum women, the medically frail, parents and caregivers of children 13 or younger, American Indians and Alaska Natives, former foster youth, disabled veterans, people in addiction treatment and those already meeting food-stamp or cash-assistance work rules.
Documentation submission to meet the requirement will be a regular process. States must verify compliance at application and at least every six months at renewal, checking work data they already have. If people miss a check, the state will send notice and allows 30 days to respond before coverage ends. Enrollees may self-attest in the first year, but from 2028 documentation is generally required. A handful of states plan to switch on early.
This is a big change in policy that will affect millions of Americans. What happens next is a $326 billion question.
One account of how this will play out comes from the House Energy and Commerce Committee staffers Katie West and Matthew VanHyte, who work for the Committee that wrote the provision.
A different perspective comes from Dr. Adam Brown, an emergency physician and managing partner of the advisory firm ABIG Health and his colleague Dr. Steve Farmer, a cardiologist and former CMS Chief Strategy Officer, also at ABIG Health.
Same Policy, Two Stories About Its Purpose And Effects
In West’s telling, Medicaid work requirements will re-engage able-bodied adults through jobs, training or school. Employment and income are expected to rise and some workers will move to employer coverage. This, according to West, will free Medicaid to concentrate on the disabled, children, and fragile adults it was built for. The provision will bend an unsustainable federal spending curve for the Medicaid program.
West framed Medicaid work requirements as restoring a Clinton-era principle: benefits for able-bodied adults without young children should carry an expectation of work. The intent, she said, echoing Chairman Brett Guthrie, was never to police the program with paperwork but to move people toward independence. An HHS brief projects the requirements could lift 1.6 to 2.9 million people out of poverty.
She also insists the rollout of the work requirements will not resemble the Arkansas rollout, whose 2018 Medicaid work requirement program dropped about 18,000 people from coverage in months with no measured employment gain. Instead, there is a new implementation framework that consists of data-first verification, broader exemptions, mandatory outreach, first-year self attestation, and a roster of vendors that have pledged more than $600 million in discounted technology, with CMS working alongside red and blue states. Most enrollees who already work, West argued, will be confirmed automatically and never file a form. Additionally, nearly eight in ten Americans back the policy in conservative polling. And a $50 billion rural health fund will cushion hospitals.
Brown’s view differs. He leaned into the Arkansas experience as the only real-world test. He noted that researchers found it produced no employment gain over eighteen months, a null result a 2025 Urban Institute study confirmed.
Brown also went on to describe what it did produce among those who lost coverage: medical debt for about half and delayed care or skipped medications for roughly six in ten. The reason, Brown argued, was a design flaw could be impossible to fix. A Health Affairs analysis found more than 95 percent of Arkansas’s target group already met the requirement or qualified for exemption but couldn’t prove it.
Because about 64% of Medicaid enrollees work and another 15% can’t because of illness or disability, he sees a system that will spend its capacity re-confirming the roughly 80% who already comply while dropping eligible people who fall a month short. And the risk is double because renewals will occur every six months.
West’s answer is that data-first verification exists precisely to protect the already-working. Attention will be focused on the smaller share who are genuinely disengaged. Defenders of the HHS analysis add that the literature does point to real, if more modest, employment gains.
Georgia is where the two readings collide most directly. Its Pathways program, the only work requirement running today, spent more than twice as much on administration as on care: $54 million versus $26 million, per a 2025 GAO review. West and VanHyte, however, argue Georgia is a poor representation of what’s coming nationally: it conditions new coverage on proving work before enrollment, a higher bar than keeping coverage one already holds.
Brown calls it the preview. Nearly 90% of Georgia’s administrative spending was federal money, he noted. So federal dollars were never the missing ingredient. Additionally, the same data-matching now promised nationally is what Georgia has used since 2023 . GAO has also traced the lopsided costs to building eligibility systems and start-up delays. These are structural features each state must replicate.
The $326 Billion Question: Where Will The Savings Come From?
When asked how the policy saves money, West described a chain: get more people working, many gain employer insurance, and Medicaid is relieved of costs it should not carry. The requirement is the Trump’s Beautiful Bill’s largest saver, scored by CBO at about $326 billion over a decade against spending the committee forecasts will climb from $708 billion this year to $981 billion by 2036. Bending that curve, West states, is the goal.
Brown and Farmer call the figure a cost shift, not savings. CBO expects most of it to come from people leaving the program rather than finding cheaper coverage. This will be money moved off the federal ledger onto people who still get sick, the hospitals that treat them, and the states that run the system.
A George Washington analysis estimates up to 5.2 million losing coverage in a single year and some $59 billion in lost state GDP, and about 449,000 eliminated jobs. An AP review of state filings found implementation could cost states more than $1 billion against a $200 million federal allotment.
Farmer retorted that the people who lose Medicaid for failing the work requirement are treated, by the law’s own terms, as still Medicaid-eligible. This makes them ineligible for the premium tax credits that lower marketplace premiums. Most Medicaid workers aren’t offered an employer plan either.
West sees that as consistent with steering able-bodied adults toward work rather than subsidized coverage. Brown and Farmer see a foreclosed exit.
Who Will Own The Results?
So which story comes true? No one knows yet. But we will know the answer soon.
West and VanHyte expect early-adopter states like Nebraska, Montana, and Arkansas, which switch on ahead of schedule, to show the design working.
Brown expects the opposite, and points to the metrics that will settle it fast: whether the roughly 92% already working or exempt rises once reporting starts.
What is not uncertain is who answers for the result. It’s the Energy and Commerce Committee who wrote this provision, set its timeline and defended it on the merits, . If it pulls people into work and lifts them from poverty, its authors will have earned the win. If eligible people instead lose coverage over paperwork and turn up sicker in emergency departments, that will be their results to own too.
Guthrie’s committee built the test. In January, 20 million people begin taking it — and whatever the answer, it will carry exactly one set of names.