Why the Court Once Again Struck Down Federal Approval of Medicaid Work Experiments
The two new decisions handed down by a federal court this week (Stewart v. Azar and Gresham v. Azar) are the latest developments in the ongoing debate over whether employment should be a basic condition of eligibility of Medicaid for most working-age adults. The new decisions vacate U.S. Department of Health and Human Services (HHS) approval of Section 1115 Medicaid demonstrations that allow both Kentucky and Arkansas to impose work as a requirement of coverage. Notably, the court halted the Arkansas experiment midstream, before any more people lose coverage.
Despite this legal development, on Friday the administration announced approval of Utah’s partial Medicaid expansion coupled with a work requirement.
The basic premise of work requirements is questionable, since most Medicaid enrollees are already working. Nearly eight in 10 nondisabled working-age adults with Medicaid live in working families, and the majority work full-time. Those who are not working report that they are caring for family members, struggling with their own health issues, or have been unsuccessful in finding employment.
Ironically, Medicaid expansion was incorporated into the Affordable Care Act precisely because low-income workers are the least likely to have jobs with health insurance. Evidence shows that Medicaid has, in fact, helped people return to work, because individuals finally can get the health care they need to make working possible.
Nonetheless, the Trump administration has promoted and authorized Medicaid work requirements. It has propelled its policy forward by using special powers reserved, as the court’s decision emphasizes, for experiments that promote Medicaid’s objective of insuring people who are eligible. To get around the chasm between Medicaid’s purpose and experiments aimed at lowering costs by reducing enrollment, the administration invented a new purpose for Medicaid: to promote health by incentivizing work. On that basis, the administration has approved eight work-requirement demonstrations, with more in the pipeline.
Meanwhile, news accounts continue to report on the toll these experiments are taking on the poorest communities, where there is no work to be had and people are unable to navigate the system for reporting their work status — or even to understand the reporting rules.
The court’s decision to not only strike down the Arkansas rule but also halt the experiment comes as little surprise, given the harm it has caused. Both HHS and Arkansas argued for the demonstration to continue, so that the state could collect data and educate people about their obligations. The court rejected the arguments, stating:
Arkansas’s own numbers confirm that in 2018, more than 16,000 people have lost their Medicaid. Defendants offer no reason to think the numbers will be different in 2019; indeed . . . they seem likely to rise. Weighting the harms these persons will suffer . . . against the disruptions to . . . data collection and education efforts . . .renders a clear answer: . . . Arkansas Works . . . cannot stand.
The court’s Kentucky decision represents the state’s second time through the waiver review process; the judge vacated HHS’s first approval of the waiver in June 2018. The judge concluded that the first approval was arbitrary and capricious as a matter of law, since because the HHS secretary had failed to offer evidence that an experiment designed to remove thousands of people promotes the Medicaid’s purpose.
Like the court’s earlier decision, the latest two don’t explicitly conclude that work experiments can never be approved under Section 1115 of the Social Security Act, which allows the HHS secretary to waive federal Medicaid requirements. But the decision does, implicitly at least, raise the central issue: Is it possible that an experiment calibrated to remove people who don’t meet work and reporting requirements from Medicaid can ever be shown to further Medicaid’s purpose? The HHS secretary’s inability to demonstrate this was, for the court, the “signal omission” in his initial approval. The administration obviously failed to correct this basic problem the second time through.
To overcome the inability to make the basic connection between work requirements and the purpose of Medicaid, the administration essentially pursued two strategies. In Arkansas’s case, the HHS secretary argued that since the state had offered no estimates of its own, he did not need to either. For Kentucky, the administration argued that while 100,000 would lose coverage (the state’s estimate), another 450,000 would get to keep it, because the governor had threatened to drop the expansion entirely if the work requirements were not maintained. (Current law does not permit him to end the expansion unilaterally.)
The court rejected the argument that 1115 allows HHS to dispense with its own impact estimates, as well as the argument that 1115 frees the secretary to allow states that want to spend less money to “refashion the program Congress designed any way they choose.” According to the court, although cost savings may be a defensible consideration in 1115 experimental design, the secretary failed to explain how the state’s approach to cost cutting made sense.
In vacating both approvals, the court remanded the two decisions back to HHS. We don’t yet know the next move, but a federal appeal is likely. The administration could again attempt to persuade the court to stay its order halting the Arkansas experiment while the appeal proceeds, but the court presumably will not change its mind. It’s possible the appeals court could allow Arkansas to continue its experiment while the appeal is pending, although the equities argue against it.
In the meantime, a third case challenging implementation of New Hampshire’s approved work experiment is moving forward before the same judge. Based on his decisions to date, it’s probably safe to assume a similar result.