Arkansas’ Medicaid Work Reporting Rules Lead to Staggering Health Coverage Losses

Earlier this week, the state of Arkansas released its final round of data for 2018 for its controversial work reporting requirement policy. The data shows that an additional 1,232 adults lost their Medicaid coverage at the end of December, bringing the total Medicaid coverage loss since August to 18,164.

The news motivated me to take stock of where things stand with the Trump Administration’s signature Medicaid policy. Let’s take a look back to the beginning of 2018 when the Trump Administration turned its attention in earnest to using executive authority to promote policies in the Medicaid program through Section 1115 waivers that would cut Medicaid coverage after it failed to do so through legislative efforts.

On January 11th 2018, the Centers for Medicare and Medicaid Services issued guidance encouraging states to request Section 1115 waivers that were, allegedly, designed to promote work. In particular, the guidance said:

“CMS will support state efforts to test incentives that make participation in work or other community engagement a   requirement for continued Medicaid eligibility or coverage for certain adult Medicaid beneficiaries…”

To state the obvious, Governors are free to work with Medicaid beneficiaries and other residents of their states to promote employment, support job training programs, launch affordable child care initiatives, and engage in other such activities without permission from CMS through a waiver. So the essence of the new guidance boils down to the Secretary of Health and Human Services Azar telling states that he would give permission for them to punish people and cut them off of their Medicaid coverage if they don’t comply with new work reporting rules that a state and CMS agree to in a waiver.

Whether the Secretary has authority to allow states to do this is another question, and as readers of SayAhhh! know, the first federal court weighed in on this question last summer. As a result, Kentucky’s complex and punitive waiver is on hold.

Fast forward to the end of 2018, and CMS has approved seven states (KY, AR, IN, NH, ME, MI, WI) to impose work reporting requirements (and more). In one of the more bizarre Medicaid related tweets I have ever seen, CMS Administrator Verma announced a rushed set of Medicaid waiver approvals right before Christmas — including to outgoing Republican Governors in Maine and Michigan.

Maine’s waiver almost certainly won’t be implemented, and the status of Michigan and Wisconsin are unclear – all three states have new Democratic Governors. With Kentucky on hold, that leaves three states (AR, IN, NH) moving forward with or continuing to implement work reporting requirements.  

In Arkansas, which rushed to become the first and only state to implement the new work reporting rules in 2018, the final numbers are now in. During the final four months of last year, 18,164 adults lost Medicaid coverage as a result of “noncompliance” with work reporting rules. The vast majority of those folks are likely to become uninsured for at least some period of time – exposing them to financial peril and worsening health. Arkansas is only applying the new rules to those living below the poverty line – for these folks there are scant options for other affordable coverage sources.

Arkansas’ final tally amounts to a staggering coverage loss ratio of 23 percent for those subject to the new rules — considerably higher than the national estimates (6 to 17 percent) that the Kaiser Family Foundation used in its recent study estimating the impact of work requirements.

Meanwhile, virtually no one in Arkansas is newly reporting work as a result of the new “incentives” (i.e. you lose your Medicaid if you don’t comply). As is true in general with the Medicaid expansion population nationwide, many are already working and are being exempted automatically by the state’s computer. But as I have blogged about before, less than one percent of those subject to the new rules are newly reporting work hours. The December data shows the same thing, with 1,311 clients actually reporting “work activities” (as opposed to the computer taking care of it), but 849 of  them merely meeting the SNAP requirement. This leaves just 462 people out of the 60,680 subject to the work requirement newly reporting work or job search activities this month.

The adults losing coverage were between the ages of 30 to 49 with incomes below the poverty line. All parents are supposed to be exempt. In 2019, the state will start to apply the same punitive rules to younger adults 19 to 29 – a smaller but still significant cohort. The clock resets for the “three strikes you are out” policy on January 1st – in Arkansas there is a lockout period for the remainder of the calendar year when an individual is kicked out of Medicaid. The state reports that less than 1,000 of those individuals who were kicked out have reapplied for Medicaid. The first round of 2019 coverage losses will begin at the end of March for those with 3 months of non-compliance.

While much has been made of the obvious flaws with Arkansas’ implementation – the only way to report for much of the year was online despite the state’s low levels of internet connectivity – in some respects the state’s underlying policy is relatively forgiving compared to what may lie ahead in other states. Arkansas requires reporting of 80 hours per month and does not kick people off until they have three months of non compliance, and everyone over 50 is exempt.

By contrast, the next state up in 2019 for implementation, New Hampshire, plans to require a steeper 100 hours a month, affects 19 to 64 year olds, and allows just a one-month grace period to make up for non compliance with the tougher rules after just one month on noncompliance. If New Hampshire’s waiver is implemented as approved by CMS,  we expect to see much greater coverage losses.

There is still no Arkansas waiver evaluation in place. Despite the heavy coverage losses, and the fact that this is CMS’ signature Medicaid initiative for Section 1115 research and demonstration waivers, there is still no approved evaluation in place for the Arkansas waiver. I have said for years that policy making in Washington can often be a “fact-free” environment, but this is a new low — when thousands appear to be losing their health insurance in service of goals that are touted by proponents but with no independent evaluation.

Why are so many people losing Medicaid expansion coverage more broadly in Arkansas? Also worth noting is that the enrollment for adults in Arkansas’ Works, which is the state’s Medicaid expansion program, has declined by a substantial 50,000 over the course of 2018. This obviously includes the 18,000 we know about, but a topline look at the data given for the rest of the closures raises some red flags. Since the work reporting rule has been in effect, a similar or larger percent of cases (between 25 to 36 percent) have been closed for “failure to return requested information.” And between 18 to 24 percent of case closures fall into a mysterious “Other” category. Only a minority of case closures have been for clearly legitimate reasons such as increased income or client requests for closure. These are the kinds of questions a thorough evaluation would explore in depth.

Children’s Medicaid/CHIP enrollment declined in Arkansas in 2018 as well by approximately 8,000 – let’s hope this is the sign of an improved economy, but our recent report documenting  the growth in uninsured kids nationally gives one pause….

What work requirements waivers are pending at CMS? Both Ohio and Arizona (expansion states) may be up next for approval, and then CMS would likely move next to requests from non-expansion states such as Mississippi, Alabama, and South Dakota — which underwent the federal comment process a while ago. Oklahoma and Tennessee are currently in the federal comment process, and South Carolina is taking state public comments through January 22.

As readers of SayAhhh! know well, we have done a series of reports on these punitive proposals in non-expansion states that are targeted squarely and exclusively on parents in the poorest families. As parents become uninsured, the entire family is at greater risk of falling further into poverty because of medical debt or bankruptcy. Research shows that when a parent is uninsured, their children are much more likely to be uninsured.

We’ll be following closely where this waiver path leads.

Joan Alker is the Executive Director of the Center for Children and Families and a Research Professor at the Georgetown McCourt School of Public Policy.