As I noted in an article earlier this month, the Congressional Budget Office (CBO) estimates that the federal government will save $911 billion over 10 years and that 10 million people will lose health insurance as a result of Medicaid changes in H.R.1, aka the One Big Beautiful Bill Act (OBBBA). This bill requires near-poor, able-bodied Medicaid beneficiaries with incomes currently above the poverty line but under thresholds that depend on family size — $22,025 for an individual or $45,360 for a family of four — to navigate a maze of red tape and reporting requirements twice a year to remain enrolled. Disabled people on Medicaid, including those with cancer or end stage renal disease, will have to prove that they are too disabled to work in order to waive the work requirement.

An analysis by health policy organization KFF found that in 2023, 64 percent of adults 19 to 64 years of age enrolled in Medicaid were working either full or part-time. Nearly 30 percent were not working because of caregiving responsibilities, disability or school attendance — barriers to employment that typically exempt people from work requirements in programs that mandate them. Most of the 8 percent who were not working were retired or unemployed. The result is that only a small share of Medicaid enrollees will lose access to the program because they do not meet work requirements. However, many more will lose coverage because of the difficulty of navigating the reporting requirements and the administrative red tape built into the OBBBA for this purpose.  

In June 2026, HHS released interim regulations for the changes to Medicaid that states must implement by January 1, 2027. The regulations are much more severe than the requirements in H.R.1, and have upended work that states had already done to be ready to comply by the January 1 deadline. This has forced states to scramble to be ready in just six months, an extraordinarily short window to prepare for the large changes in reporting requirements. Disenrolling people by mistake may mean the difference between life and death for Medicaid enrollees. The rollout of the new work and massively burdensome reporting requirements is likely to be bumpy.

ASPE Publishes Very Brief Fantastical Briefing Report 

In conjunction with the release of the new eligibility and reporting requirements, the Department of Health and Human Services released a brief, prepared by the Office of the Assistant Secretary for Planning and Evaluation (ASPE), with the Orwellian title, “Medicaid Work Requirements Incentivize Employment and Are Estimated to Reduce Poverty.” The brief’s conclusion that pushing people off the Medicaid rolls will increase employment and reduce poverty is contradicted by an earlier assessment by the Congressional Budget Office (CBO). CBO found that mandating work requirements would decrease what the federal government pays for Medicaid, increase the number of people without health insurance, raise costs for states, and wouldn’t increase employment.

The four-and-a half page ASPE brief has few citations for its assertions about how cutting access to health insurance for poor people will magically produce jobs for those disenrolled from Medicaid. While the authors say they have reviewed three decades of literature on work requirements in social programs, and they have an online bibliography with more than 130 references, they cite only 10 papers in the brief. As NOTUS observes, the researchers that produced the 10 papers cited in the brief are crying foul, claiming that their studies are being misused, and in some cases are being used to support conclusions that are the opposite of what the studies show. In other cases, researchers complain that while the findings of their studies may not be distorted, crucial job search supports for job seekers in their studies that are not available to people facing disenrollment from Medicaid in 2027 are ignored in the ASPE brief.

As noted in a technical analysis of the ASPE brief published in Health Affairs, the brief assumes totally unrealistic employment effects. It considers two scenarios. In the first, all 5.8 million people on Medicaid — who the brief assumes will not meet the new work requirements and are not exempt from them — will increase their work effort. The report does not spell out where the 5.8 million figure comes from, but a back-of-the-envelope calculation suggests it includes about 4.2 million people enrolled in Medicaid who are working part-time and 1.6 million who are retired, unemployed or otherwise not working. The brief assumes that these Medicaid recipients will increase their hours of work at their current wage or at the average wage among similarly educated Medicaid-eligible individuals. 

Part-time workers are assumed to increase their hours and those who are not working are assumed to find jobs. Unbelievably, the change in employment participation is assumed to be 100 percent in the first scenario. This is more than 23 times higher than the 4.2 percent increase the ASPE brief authors report for welfare-to-work experiments that, unlike Medicaid work requirements, involved cash assistance and provided people leaving welfare programs with substantial help for finding work.The help included employment services such as transportation and job search assistance and case management services to assist people in meeting work requirements. In the second scenario, the brief’s authors assume that 80 percent of the 5.8 million people they believe will not meet the work requirements will succeed in finding employment. This is a 57 percent increase in participation and is 13 times the welfare-to-work effect. 

The reductions in poverty cited in the ASPE brief are not calculated independently, but follow from the entirely unrealistic assumptions about the increases in work participation. The earnings increase on average per family is $16,780 — a result of the built-in employment and wage assumptions. Net of the loss of benefits as their incomes rose, families of Medicaid recipients whose work participation rises experience an increase in resources or $12,034. In the first scenario, this reduces poverty by 2.9 million people; in the second, by 1.6 million. 

Arkansas’ Experience with Work Requirements

The ASPE brief omits any mention of the one instance where a state — Arkansas — implemented work requirements in 2018. This is a glaring omission in a report on the effects of work requirements in Medicaid, and suggests the brief’s authors may not have wanted to report the disastrous results of this introduction of work requirements. 

The application of work requirements to Medicaid is a new development and there are just a  few empirical studies of the Arkansas case. A cutting edge 2026 analysis by Harvard Medical School researcher Yuji Mizushima found that total Medicaid participation declined sharply during the months in which the mandated work requirements were in effect. Mizushima examined two groups of people — those on Medicaid who were disenrolled because they didn’t meet the work requirement, and those in the general population who failed to enroll in Medicaid because of the work requirement. The author found that participation in Medicaid fell by about 28,810 adults — 18,164 who were disenrolled and more than 10,500 who did not apply or reapply for Medicaid. There was no discernible change in the number of hours worked and no increase in employment or in participation in the labor force of the Medicaid population in the months leading up to, during, or following the enforcement of the work mandate. This supports similar findings in research on the Arkansas experience in other studies. A year after Arkansas implemented work requirements, the poor results led a federal judge to pause and later to strike down the work requirement. 

Conclusion

The introduction of work requirements in Medicaid in Arkansas during the first Trump administration led to a disastrous decline in access to health insurance for people who had relied on Medicaid for health care, but did not increase hours of work or employment. This stands in direct contradiction to the wishful thinking and wildly unrealistic assumptions embodied in the two scenarios in the ASPE brief, and raises unavoidable questions about why the authors failed to examine the Arkansas experience.