Study says Medicaid work requirement increased uninsured rate for Arkansans but did not boost employment

Governor Hutchinson

A study published Wednesday in the New England Journal of Medicine found that Arkansas’s Medicaid work requirement led to lower levels of health insurance coverage* among 30- to 49-year-olds in 2018, the group targeted by the first-of-its-kind work rule last year. The researchers also concluded that the policy did not lead to a rise in employment among this target population.

Dr. Benjamin Sommers, the study’s lead author, said researchers compared 30- to 49-year-old low-income adults in Arkansas with other age groups, as well as their counterparts in Kentucky, Louisiana and Texas. The study was limited to households that earned less than 138 percent of the federal poverty line — the eligibility cutoff for Medicaid expansion — and was conducted by telephone survey in November and December. Respondents were asked about their work status and insurance status at that time and during the same month a year earlier, in 2017. In total, 1,500 Arkansans and a similar number of people from the other three states were interviewed.

“The share of people who are working when you compare across the groups by age and by state seems to be pretty similar,” Sommers, a health economist and physician at the Harvard T.H. Chan School of Public Health, said in a phone interview Monday. “So if the goal of the policy is to get 30- to 49-year-olds in Arkansas into new jobs, that doesn’t seem to have happened, at least through the end of 2018.”

However, the researchers did find one significant difference between the groups: “Uninsured rates increased among 30-49 year-old Arkansans from 10.5% in 2016 to 14.5% in 2018, with smaller or no changes in the other groups,” the study says. (The researchers didn’t find a significant reduction in access to medical care, though Sommers said such “downstream effects” would not be measurable until later.)

The findings run counter to arguments made by Governor Hutchinson, the Arkansas Department of Human Services and the Trump administration. State and federal officials are asking the U.S. Court of Appeals for the D.C. Circuit to overturn a federal judge’s order that halted the Arkansas work requirement in March.

Some 18,000 beneficiaries were stripped of Medicaid coverage in 2018 due to not meeting the terms of the work requirement, according to DHS. But it has not been clear what happened to those people after they were kicked off the program. Hutchinson has downplayed concerns about coverage losses, suggesting many former beneficiaries found insurance through an employer or elsewhere. DHS officials have also pointed out that most of those who lost coverage in 2018 have not re-enrolled, even though they became eligible to do so as of Jan. 1. As of mid-April, fewer than 3,000 of the 18,000 had re-enrolled in Medicaid.

DHS spokeswoman Amy Webb said by email on Wednesday that the new study “does not provide a meaningful or thorough evaluation of the Arkansas Medicaid demonstration project [the work requirement]. It does not tell us why, for example, most people who left the program have not returned this year, even though they were allowed to do so. It does not provide information about the health status of individuals or how they view the importance of insurance.”

Sommers acknowledged that some people who lost Medicaid coverage may have chosen to not re-enroll. “That’s one explanation,” he said. “Another explanation is that people don’t understand what happened in the first place, so they think they’re not eligible anymore. So we want to evaluate that and do a survey later in the year to see how many of them did come back.”

The study did find a slight increase in employer-sponsored insurance among the 30-49 age group, Sommers added, but it was within the margin of error. “So, we don’t know if something’s actually happening there. … We can’t say that definitely none of these people got employer coverage — it’s possible some did — but we don’t see any clear evidence of that,” he said.

He emphasized the study found no change in job status when comparing the target group to the other groups. Hutchinson has claimed the work rule has incentivized thousands of people to find jobs. National health advocacy groups have disputed the governor’s figures.

The study found that beneficiaries were confused about the work requirement and its complex rules, Sommers said. “A third of the people in this target population said they’d never even heard of it,” he said. At the time the survey was performed, the policy had been in place for some six months.

In fact, the study concluded, the main reason people lost coverage was confusion over the specifics of the policy. The overwhelming majority of beneficiaries surveyed — over 95 percent — were meeting the terms of the work requirement or should have been exempt from reporting, according to the survey. The policy mandated beneficiaries to report at least 80 hours per month of work, school, volunteer or certain other activities, or else claim one of a wide variety of exemptions, including caring for a dependent child or having a medical condition or disability that prevented them from working. (Those who did not meet the requirement for three months were locked out of Medicaid coverage for the rest of 2018.)

Some 40 percent of respondents told the surveyors they were disabled or too sick to work. They were not asked to provide any proof of a medical condition, but Sommers said the question mirrored the state’s questions for determining a medical exemption. “If most of the people in our survey told the state the same thing they told us, they would have been exempt or satisfied the requirement,” he said.

Among beneficiaries who said they had not reported their information to the state as required, Sommers said, “the most common answer [why] was, ‘Well, I don’t think I qualify.’ ” Some 40 percent of people who didn’t report fell into that category. “And, from our other questions, we actually know that [most of] those people did,” he said. “They were either working or they met one of the exemptions.”

Another 32 percent in the “required to report but did not do so” category said they were stymied by a lack of internet access. Until late in 2018, Arkansas required that beneficiaries log their work hours on a web portal, though it also allowed third parties (such as insurance agents) to assist beneficiaries and relay information by phone. In December, DHS began operating a “helpline” to allow phone reporting directly to the agency.

Webb, the DHS spokeswoman, criticized this portion of the study for its small number of respondents — only 46. “Two of the key topics regarding reporting are based on responses from fewer than 50 people from the thousands who were required to report,” she wrote.

The study also notes that the work requirement generally introduced even more red tape into an already confusing system, increasing the likelihood that people lose coverage due to mistakes or oversights.

“While Medicaid has always struggled with high turnover due in part to legally-required annual eligibility redeterminations, our findings suggest work requirements have substantially exacerbated administrative hurdles to maintaining coverage,” it says.

Webb said the conclusions were premature.

“Most importantly, this report is based on less than a year’s worth of data as the court intervened before even a full year of the demonstration project was complete,” she wrote. “So you cannot describe this as the robust evaluation that we want and expect of a demonstration project that truly has national significance. The best way to get answers to everyone’s questions about the impact of work and community engagement requirements would be to let Arkansas continue what was started and conduct a true evaluation that follows people over time.”

However, the state has shown little urgency in initiating such an evaluation. As part of the terms of the federal waiver that allowed Arkansas to implement the work requirement — which was granted by the Trump administration in March 2018 — the state was required to design a plan for an independent third-party to evaluate the experimental policy. But last November, federal Medicaid authorities rejected Arkansas’s draft evaluation design, saying its proposed outcomes were “not well defined and outcome measures are not specified,” among other flaws. The state’s design “should be better articulated and strengthened,” the federal Center for Medicare and Medicaid Services said in a letter to Arkansas DHS.

Sommers acknowledged the limitations of the new study. “If we had a longer follow-up and a bigger survey, maybe we would have seen some small changes [in employment],” he said. But he said the main reason the work requirement didn’t appear to affect job status was that most people subject to the rule were already working or meeting one of the exemptions. “There’s just a small number of people whose behavior you could modify with this sort of policy,” he said.

Yet he concurred with Webb on the importance of doing a more thorough evaluation of policies like Arkansas’s. Several other states are now in the process of rolling out Medicaid work requirements.

“These are large scale social experiments we’re conducting, and we ought to evaluate them the same way we try to do with new medical treatments that affect thousands of people,” Sommers said.

*Correction, June 20: This sentence has been modified for clarity since this story's first publication. The sentence originally said the study found the Arkansas work requirement "... led to lower insurance rates among 30- to 49-year-olds in 2018 ..."

This reporting is courtesy of the Arkansas Nonprofit News Network, an independent, nonpartisan news project dedicated to producing journalism that matters to Arkansans. Find out more at

The Arkansas Nonprofit News Network is an independent, nonpartisan news organization dedicated to producing journalism that matters to Arkansans. Our work is re-published by partner newsrooms across the state.