Congress is debating significant Medicaid policy changes and funding cuts, including the establishment of a federal work requirement. Proponents argue work requirements promote employment. States such as Arizona and Ohio also intend to pursue this policy through Medicaid demonstration projects.

In a recent Health Services Research journal article, we present new evidence from Arkansas’s 2018–19 Medicaid work requirement program showing that the policy reduced the number of adults with health insurance coverage and had no effect on employment—failing to achieve its intended outcome. Our findings suggest the adoption of a similar work requirement in other states or nationally would decrease health insurance coverage rates without achieving the policy goal of promoting work.

What do we know about the effects of Medicaid work requirements?

Efforts to impose work requirements are motivated by the assumption that many adults with Medicaid are healthy enough to work but choose not to and that conditioning Medicaid eligibility on employment would provide them with incentives to join the labor force. Cosponsors of recent legislative proposals assert that Medicaid work requirements will “get healthy adults back into the labor market” and help enrollees “find employment that leads to better health coverage.”

In 2018, Arkansas became the first state to establish a work requirement for adults enrolled through its Medicaid expansion program, starting with 30-to-49-year-olds with incomes below the federal poverty level (FPL). Automatic exemptions were granted for parents living with dependent children and certain other groups. The requirement led to the disenrollment of more than 18,000 adults over four months before a 2019 federal court ruling halted further implementation.

Using data from telephone surveys of adults with low incomes in Arkansas and three comparison states in the South, Harvard University researchers found the work requirements were associated with an increase in uninsurance among 30-to-49-year-olds in Arkansas between late 2016 and late 2018, but they found no evidence of a positive effect on employment.

National data confirm Arkansas’s work requirements didn’t have the intended employment effects

Our study was designed to build the evidence base on the effects of Medicaid work requirements by reexamining experiences in Arkansas using national survey data. We drew on the US Census Bureau’s American Community Survey, leveraging its strengths, which include large sample sizes in all 50 states, a multiyear trend, high response rates, and detailed data for identifying the specific population targeted by work requirements.

We compared coverage and employment trends among 30-to-49-year-olds with low and moderate incomes and without dependent children in Arkansas before and after the state’s work requirement policy was implemented in 2018 with trends for adults in the same age group in 27 states that, like Arkansas, expanded Medicaid eligibility in 2014 or earlier.

Despite using different data sources and methods, our findings are consistent with those of the Harvard study. In the years leading up to Arkansas’s work requirement demonstration waiver, uninsurance rates in Arkansas and other states moved roughly in parallel. After the waiver was implemented, uninsurance spiked for 30-to-49-year-olds in Arkansas while remaining stable for their peers in other states and for adults in unaffected age groups (i.e., 19 to 29 and 50 to 64) both in Arkansas and other states. The increase in uninsurance coincided with a decline in reported Medicaid/private nongroup coverage (Arkansas’s expansion uses Medicaid funding to help enrollees purchase private Marketplace health plans) and no change in employer-sponsored insurance.

Controlling for other factors that could affect coverage, we found Arkansas’s work requirement was associated with a 4.4 percentage point increase (18.7 percent) in the uninsurance rate among 30-to-49-year-olds with incomes below 300 percent of the FPL and a 7.4 percentage point increase (27.5 percent) among 30-to-49-year-olds with incomes below the FPL, who were most likely to be affected by the policy. Coverage losses disproportionately affected adults without home internet access, who could have experienced difficulty accessing the state’s online portal for requesting exemptions and reporting work activities.

In contrast to our findings on coverage, we did not observe any changes in employment associated with Arkansas’s work requirements. The policy’s effect on employment among the affected age group was negative, small, and statistically insignificant. We also found no significant effects on other measures of work, such as hours worked or on employer-based coverage.

National Medicaid work requirements are unlikely to increase employment but would worsen health outcomes and strains on medical systems

As federal and state lawmakers consider work requirements for Medicaid enrollees, it’s important to assess whether these policies would achieve their stated goal of increasing employment alongside the negative effect of disenrollment for adults who neither obtain an exemption nor satisfy the requirements. Public support for this policy is grounded in deeply held values that “those who can work should work” and the belief that “for people who are able to work, being on Medicaid makes them less interested in working.”

Previous studies have found mixed evidence on Medicaid’s effect on employment. Medicaid expansions under the Affordable Care Act had little measurable impact on employment and hours worked, and some earlier state-specific studies found negative effects or no effects on labor supply. A few studies found Medicaid may increase employment for some groups. Our study contributes to this evidence base by assessing policies that directly target work incentives in Medicaid.

The finding that Medicaid work requirements are ineffective in promoting work is consistent with other data suggesting opportunities to increase employment are limited, as nearly all Medicaid enrollees are already working, looking for a job, enrolled in school, caring for a family member, or have health conditions or disabilities that prevent them from working. It also reflects challenges with the design of work requirements, in which people are often disenrolled because they are unaware or confused about the policy or have trouble navigating state reporting systems.

The negative consequences of work requirements could be severe. Even though Arkansas’s program was in effect for only a short time, enrollment declines were substantial, and the Harvard study found adults who lost Medicaid in Arkansas were more likely to delay care and incur medical debt.

A recent analysis estimates that about 5 million adults ages 19 to 55 who are enrolled in the Affordable Care Act’s Medicaid expansion could lose coverage under a federal work requirement, assuming state implementation processes are similar to those applied in Arkansas and New Hampshire, which also attempted to impose work requirements in its Medicaid expansion program under a previous waiver.

Our study of Arkansas’s experience suggests nearly all adults losing Medicaid would become uninsured, leading to worse health outcomes and increased financial strain on health care providers facing higher uncompensated care costs.

Alternative policies could increase employment among Medicaid enrollees without taking away their health care coverage. These include voluntary programs, such as Montana’s Health and Economic Livelihood Partnership, that address barriers to work, such as lack of transportation, child care, and the need for occupational training matched to available job opportunities. The growing evidence that previous Medicaid work requirements in Arkansas were ineffective in promoting employment despite significantly increasing the uninsurance rate among affected adults could generate more interest in these alternative voluntary approaches.