Like many folks who work on Medicaid policy, we’ve been paying close attention to what is happening in Arkansas – the first state in the country to implement a work activities reporting requirement in Medicaid.[1] Earlier this week, the second round of data was released by the state documenting that another 4,109 Medicaid beneficiaries lost coverage retroactive to October 1, 2018 for failure to comply with the new reporting rules.

These unlucky people caught up in Arkansas’s new tangle of red tape join over four thousand former Medicaid beneficiaries who lost coverage as of September 1. All 8,462 of these folks are “locked out” of Arkansas’s Medicaid program until January 1, regardless of whether they come into compliance – an especially cruel twist. Most of them will likely become uninsured since the availability of affordable private insurance for those living below the poverty line is very limited in Arkansas and elsewhere.

Arkansas is phasing in its policy by age and by redetermination date, and we can expect a similar number of people to lose coverage next month. The state’s data show that 4,841 people have two strikes against them – this is the group who stands to lose on November 1 if they don’t report work hours or not enough work hours. So it is clear that coverage losses will continue.

While the direct coverage losses are relatively easy to discern from the data (subject to the caveat below regarding possible indirect losses), the failure of the new policy to incentivize new participation in work activities (its purported goal) is much harder to pick out. The second month of data tell the same story we called out last month. The policy is a complete failure at incentivizing behavioral changes (i.e. new work activities) by threatening to take people’s health insurance away.

Here’s why: 95% of those subject to the new rules in Arkansas have so far not changed their behavior and less than 1% are newly reporting work activities. The vast majority of Arkansas Works beneficiaries subject to the new rules (72% this month) are being exempted by the state’s computer because they are already working or are taking care of a child.  Of those who must do something (i.e. request an exemption or report work/volunteer hours), 81% did nothing. Together those two groups add up to 95% of those subject to the new rules.

Meanwhile, of the paltry 1,532 persons who are reported as having met the new Medicaid requirement, page 5 of the state’s data report shows that 1,025 are merely reporting that they meet the pre-existing SNAP work requirement. That leaves 507 folks (less than 1% of those subject to the new rules) newly reporting work or work related activities.

These numbers reflect the direct impact of Arkansas’s new policy, but there may be indirect effects as well. Enrollment in Arkansas Works (the state’s name for its Medicaid expansion population subject to the new rules) has declined by more than 30,000 since January 2018. While parents are supposed to be exempt from the new rules, it is worth noting that since the new rules went into effect on June 1, children’s enrollment has seen a net decline of 1,064 as of October 1. These numbers may be the result of an improving economy but they certainly raise red flags.

As HHS Secretary Azar stands on the brink of approving additional state Medicaid waiver requests to implement work requirements, these data make clear that this experiment is not working, and people are being harmed in the process. The truth is in the numbers.

[1] Kentucky’s Governor Bevin is eager to implement a similar approach, but a federal court ruling put a hold on implementation. New Hampshire and Indiana are scheduled to implement work requirements on January 1, 2019.