After Arkansas established a Medicaid work requirement last year, significantly fewer low-income adults subject to the mandate reported having health insurance than the year before, and fewer had a job or were engaged in other qualifying community engagement activities, a new study found.
That contrasted with the experience of low-income adults in Arkansas who were not subject to the work requirement, as well as with comparable people in Kentucky, Louisiana and Texas, three states that did not impose a Medicaid work requirement. Those comparison groups all reported little or no change in their insurance rate from 2017 to 2018.
But the percentage of people in the comparison groups reporting that they worked at least 20 hours a week also dipped.
The study, based on a phone survey of about 6,000 people late last year and published in the New England Journal of Medicine on Wednesday, cast doubt on the effectiveness of the CMS-approved Arkansas demonstration to get more non-disabled Medicaid enrollees into health-enhancing jobs, training or volunteer activities. Nine Republican-led states have gotten similar waivers and hope to shift this population to private health insurance.
"We didn't find any employment changes and instead we see Medicaid coverage rates dropping and more people without health insurance," said Dr. Benjamin Sommers, a Harvard University health policy professor who is lead author of the study. "Based on our results so far in Arkansas, it doesn't appear that this particular policy is accomplishing its goals."
By the end of December, Arkansas had disenrolled more than 18,000 people for noncompliance. Approximately 250,000 are enrolled in the state's Medicaid expansion program called Arkansas Works.
In March, a federal judge halted the work requirement programs in Arkansas and Kentucky, ruling that the HHS secretary did not adequately consider the impact of the waivers on providing coverage, which is one of Medicaid's central objectives. The same judge is set to rule on New Hampshire's work requirement waiver.
But J.R. Davis, a spokesman for Arkansas Gov. Asa Hutchinson, said nearly 15,000 Medicaid expansion beneficiaries found jobs from June 2018 through April 2019, though the state doesn't know how many are the same people who lost coverage due to non-compliance with the work requirement.
In addition, he said the fact that only about 4,300 of the 18,000 people who lost coverage in 2018 reapplied for Medicaid this year when they became eligible again suggests they found other sources of coverage.
"It goes to what the governor said all along, these individuals either have found work, moved onto their spouse's insurance, or moved out of state," Davis said.
But Arkansas has shelved plans to conduct a formal evaluation of the work requirement demonstration after a federal judge halted the program. Given the lack of data, Davis acknowledged that "there are a lot of scenarios" for why people dropped off Medicaid.
Sommers said Davis' claims are unsupported.
"So far we find no evidence to support the notion that most people got jobs or other coverage," he said.
Starting last July, Arkansas required non-disabled Medicaid expansion enrollees ages 30-49 to report monthly on whether they had spent at least 80 hours during the previous month working, taking classes or volunteering. They could only report through an online portal. If they failed to meet the reporting and community engagement requirements for three months, they lost coverage for the rest of the calendar year.
The Arkansas Hospital Association and other provider and advocacy groups opposed the work requirement, arguing it would lead to people losing access and more uncompensated care.
The new study found the self-reported uninsured rate for the Arkansas group subject to the work requirement rose from 10.5% to 14.5% from 2016 to 2018, and the share of people saying they were working at least 20 hours a week dropped from 42.4% to 38.9%. In the three control states, the uninsured rate for comparable low-income adults held steady at 16.2%, while their employment rate dropped from 45.2% to 39.5%.
The study also found significant lack of knowledge and confusion among Arkansans subject to the work requirement. Only 21.8% thought the requirement applied to them, while 44.2% were unsure. Nearly one-third said they did not comply with the reporting requirement because they lacked Internet access.
The authors said the lack of awareness and mass confusion about the reporting requirement may explain why thousands of people lost coverage even though nearly 97% were already meeting the community engagement requirement or qualified for one of the exemptions, such as having a disability.
Officials in other states that are moving forward with a Medicaid work requirement, such as Michigan, say they're trying to avoid the mistakes Arkansas made, and seek to make it as easy as possible for beneficiaries to comply.
But policy experts and patient advocates say it's likely that other states will experience the same or worse coverage losses as in Arkansas because many Medicaid beneficiaries won't know about or understand the new rules or be able to navigate the reporting process.
"Our study should raise notes of alarm because there is a real potential for harm," Sommers said. "We should be pumping the brakes on how many people we are willing to see lose coverage as we try out this very large policy experiment."