Inconsistent work hours threaten Medicaid coverage for one in five eligible adults

Nearly 1 in 5 Medicaid Enrollees Could Lose Coverage—Not for Refusing to Work, but for Falling Short of 80 Hours a Month

Starting January 1, 2027, millions of Americans enrolled in Medicaid will be required to prove, month after month, that they spent at least 80 hours working, volunteering, studying, or training. A new study suggests that nearly one in five of them may be unable to clear that bar—not because they are unwilling to work, but because the hours available to them are too few, too unpredictable, or both. The research, published on August 28, 2026, in JAMA Health Forum by scientists at Boston University School of Public Health and the University of South Carolina’s Arnold School of Public Health, found that 19.8 percent of Medicaid-eligible adults living in states that expanded the program are at risk of noncompliance with the incoming federal work-reporting requirements, putting their health coverage directly in jeopardy. The finding lands as federal projections suggest up to 10 million people could lose Medicaid coverage under the new law.

The mandates originate in the One Big Beautiful Bill Act, the 2025 legislation that directs a reduction of nearly $1 trillion in Medicaid spending over the next decade. Under the law, adults enrolled through the Affordable Care Act’s Medicaid expansion must document at least 80 hours per month of work, volunteering, school enrollment, or job training, with certain groups exempt. The Congressional Budget Office has projected that up to 10 million people may lose Medicaid coverage as a consequence of the new work requirements and eligibility redeterminations. Much of the attention so far has centered on administrative burden—the paperwork, documentation, and reporting hurdles expected to be the primary driver of disenrollment. But the new analysis, led by Dr. Paul Shafer of Boston University and Dr. David Anderson of the Arnold School, redirects focus toward a more fundamental obstacle: a substantial share of the Medicaid population may simply be unable to accumulate the required work hours, no matter how streamlined the reporting process becomes.

To identify who is most exposed, the team analyzed federal social and economic data from 2023 through 2025, focusing on non-elderly adults plausibly eligible for Medicaid in the 40 states that had expanded the program, plus Washington, D.C. The researchers deliberately moved beyond a simple count of the unemployed. An adult was classified as at risk of noncompliance if they worked close to the 20-hour-per-week threshold that underpins the 80-hour monthly mandate, or if they reported variable hours—fluctuating by five hours or more in either direction within the last week compared with their normal weekly hours. The approach captures a group that unemployment statistics miss entirely: people who are technically employed but structurally unable to reach a fixed monthly total. By that definition, 19.8 percent of eligible Medicaid recipients stood in danger of failing the requirement. Roughly 13.6 percent hovered near the minimum-hours threshold, where a single slow week could tip them out of compliance, while 7.6 percent carried highly inconsistent schedules that made any monthly target a moving goalpost.

The findings collide head-on with a durable political narrative. Contrary to claims from work-requirement proponents that able-bodied Medicaid recipients simply do not work, the study showed that more than 66 percent of applicable recipients are already in the labor force—and among that group, over 85 percent were working more than 35 hours per week on average. The vulnerability lies elsewhere: in the structure of low-wage work. Many Medicaid recipients hold jobs in retail, hospitality, food service, healthcare, and agriculture, industries known for unpredictable scheduling and unstable positions. Gig and part-time arrangements intensify the problem, since workers in them often have little control over how many hours they are assigned from one week to the next. Beneath that volatility sit practical barriers that quietly erode monthly totals, including unreliable transportation, the absence of affordable and dependable child care, and chronic health conditions or unrecognized disabilities that never rise to the level of formal documentation yet make steady week-after-week work genuinely hard to sustain.

The risk was far from evenly distributed. Women faced a 22 percent higher probability of having insufficient or inconsistent work hours than men. Married Medicaid enrollees had an 18 percent lower probability of noncompliance than unmarried enrollees. Education carved its own dividing line: adults with higher education levels faced lower risk than those without a high school diploma. Race and ethnicity shaped exposure as well, with eligible Black and Hispanic Medicaid recipients showing a lower risk of noncompliance than eligible White recipients—a pattern that runs against the stereotypes often invoked in debates over safety-net programs. Taken together, the demographic map of potential coverage loss skews toward women, unmarried adults, people with limited formal education, and White enrollees, meaning the burden of the new rules will not fall uniformly across the expansion population.

“These results show that the ideological narrative of Medicaid recipients just needing to pull themselves up by their bootstraps does not consider the precarious employment and other real-world barriers that often prevent them from maintaining stable work over time,” said Dr. Paul Shafer, the study’s lead and corresponding author and an associate professor of health law, policy and management at Boston University School of Public Health, where he co-directs the Boston University Medicaid Policy Lab. The new requirements, he explained, will be toughest to fulfill for people with less education, for workers piecing together part-time jobs or gig work with little control over their hours, and for those carrying substantial health concerns that may not rise to the level of a documentable disability. Shafer also pointed to ample evidence that Medicaid coverage itself supports stable employment, because better-managed health makes dependable work more achievable. Without coverage, he warned, enrollees are unlikely to be able to meet both basic and critical health needs.

Dr. David Anderson, the study’s senior author and an assistant professor in the Department of Health Services Policy and Management at the Arnold School, framed the problem as a mismatch between policy design and labor-market reality. “Work and community engagement requirements assume Medicaid expansion eligible individuals have predictable schedules that translate into the same hours worked as scheduled,” he said. “Our study shows that this is not the case. People who are trying to meet these requirements are at higher risk if they are limited to short look-back periods, and limited hardship exemptions.” The point cuts to the technical core of the policy. Work-reporting systems measure compliance across fixed windows, and when a worker’s hours swing from 25 one week to 14 the next, a single bad week inside the measurement window can erase a month of steady effort—even if the worker’s long-term average comfortably clears the federal threshold.

Geography deepened the inequity. The prevalence of potential noncompliance varied substantially across expansion states, spanning multiple regions, and the shape of the risk differed depending on which form of instability dominated. Connecticut, Ohio, Arkansas, and Nevada recorded among the highest proportions of at-risk populations due to insufficient work hours—workers whose average hours simply fall short of the 80-hour monthly bar. Massachusetts, Wisconsin, Colorado, and Washington, by contrast, ranked highest for at-risk populations driven by variable work hours, where average employment may look adequate on paper while week-to-week fluctuation turns compliance into a gamble. That distinction carries real implementation weight: states dominated by hour-short workers confront a different challenge than states where scheduling volatility prevails, and an identical federal threshold will land very differently across these distinct labor markets.

If anything, the study’s estimates may be conservative. Federal rules require states to look back to the prior month for compliance before a Medicaid application is filed, but states can stretch that look-back to three months, and several are expected to determine eligibility based on multiple months of preemptive compliance before enrollment even begins. “That means that, in some states, individuals must demonstrate work over several months to qualify for benefits,” said Dr. Timothy Callaghan, study coauthor and associate professor of health law, policy and management at Boston University. “This will slow the access of newly employed individuals to Medicaid benefits in these states, and could lead to disparities in Medicaid access for those with inconsistent work.” Single mothers with high-school-aged children may be among the most exposed, Shafer noted, because parents are exempt only if their children have a recognized disability or are under age 14—leaving mothers of teenagers to meet and report the full requirement on their own.

As the January 1, 2027, deadline approaches, the study delivers a quantitative warning about what happens when a fixed federal threshold collides with an unstable low-wage labor market. With administrative burden already expected to drive much of the projected disenrollment, the analysis suggests that even a perfectly frictionless reporting system would still leave roughly one in five eligible adults in expansion states struggling to document hours they may never be able to secure. Anchored in real federal data from 2023 to 2025, the findings offer one of the clearest previews yet of who will be swept up when the rules arrive. For millions of enrollees whose paychecks depend on scheduling decisions made far above their pay grade, the difference between keeping and losing health coverage may come down to forces no work-requirement statute can control—cancelled shifts, seasonal slowdowns, or a sick child at home.