Stewart v. Azar is the lawsuit that challenges U.S. Secretary of Health and Human Services (HHS) Alex Azar’s approval of Kentucky’s Medicaid waiver program, which, among other things, imposed work requirements on new enrollees (the decision is reviewed at greater length here). In the wake of Judge Boasberg’s June 29 decision overturning that approval, two questions arise. First, what key takeaways can be discerned from his ruling? Second, what might happen next?
1. Work requirements may not be unlawful per se, but any Medicaid waiver demonstration conducted under Section 1115 of the Social Security Act must be carefully assessed for its impact on people’s health care coverage. Providing health insurance, after all, is Medicaid’s reason for being.
A little background on Section 1115: it gives the HHS secretary special powers to undertake “experiments” that promote Medicaid’s goals. But this power comes with important responsibilities. Stewart’s central holding is that when the secretary fails to carry out his key responsibility — determining whether an experiment promotes Medicaid’s objective of insuring people — his action amounts to a violation of federal law. As Judge Boasberg notes, improving health may be a salutary effect of Medicaid, but by failing to evaluate whether a proposed demonstration meets the program’s basic purpose, the secretary acted outside the bounds of his powers in approving the project.
The decision, however, sidesteps the ultimate question: Does 1115, as a matter of law, permit the HHS secretary to approve Medicaid demonstrations that are designed to reduce coverage through a combination of new eligibility requirements, including work, premiums, lockouts, complex reporting rules, the elimination of retroactive eligibility, and high cost-sharing. As Judge Boasberg notes, it certainly is possible that, as an inevitable side effect, broadly conceived experiments of the type envisioned by 1115 could produce coverage losses or restrictions. This, of course, is the case with the Obama-era 1115 Medicaid expansion demonstrations. Even as these produced a net increase in coverage, they also allowed states to pursue restrictions not permitted under normal Medicaid standards, such as premiums coupled with exclusion for nonpayment.
Regardless of whether a Medicaid 1115 demonstration is aimed at expanding medical assistance or reducing it, its impact on coverage is the crucial question. That’s because the statute, by its very terms, is intended to extend insurance to those who need it and qualify for help. If stricter eligibility standards or a lessening of coverage are to be tested, then, as part of the approval process, their potential impact must be evaluated — and presumably weighed against — alternative experimental designs that could mitigate the impact while still granting states the flexibility to try new approaches.
2. Where Medicaid’s fundamental purpose and statutory protections are concerned, expansion beneficiaries stand on equal footing with “traditional” populations.
Judge Boasberg spends considerable time debunking what has become an especially contested aspect of the debate surrounding the Affordable Care Act (ACA)’s Medicaid expansion population: namely, that these individuals are somehow less worthy — and less deserving of program protections — than traditional populations. As drafted, the ACA’s Medicaid expansion to cover all low-income working-age adults protects the new population exactly in the same way that the law protects traditional beneficiaries who were eligible for coverage the day before the law’s enactment. As the judge emphasizes, Medicaid’s transformation extended to all poor Americans the same rights that had previously been conferred only to certain subgroups. And, as such, they are entitled to the same safeguards, including protection against the arbitrary loss of coverage.
3. Comprehensive, high-quality evidence is the heart of lawful administrative decision-making.
In promoting work demonstrations, federal officials until now have relied on vague generalizations not grounded in evidence: work promotes health; work promotes dignity and self-sufficiency; Medicaid should cover only the truly vulnerable; Medicaid must be cost-conscious. These justifications were not lost on Judge Boasberg; indeed, it was Secretary Azar’s failure to do his job — not the ultimate decision permitting Kentucky’s work demonstration to proceed — that emerged as the judge’s stated concern.
If thousands of Americans are to lose their health benefits in an experiment, then what Stewart says is that the HHS secretary has a duty to let the public know that he recognizes this reality, that he has considered it, that he has weighed it against alternatives, and that he has concluded Medicaid’s coverage mission somehow would be advanced regardless. He also must explain why. Moreover, he must acknowledge how many people will be affected and what underlying factors will create this loss. And he must consider modifications that might alter such results. In doing so, the secretary needs to rely on real evidence, not platitudes.
What Might Happen Next?
The Trump administration could decide to appeal. It could argue that 1115 gives the HHS secretary the absolute, nonreviewable power to decide which experiments further Medicaid’s objectives, what those objectives are, and the type of evidence to use in making his determination. Such a position would run counter not only to the terms of the Administrative Procedure Act but also to the numerous precedents regarding administrative law generally, and Medicaid demonstrations specifically, on which Stewart rests.
Alternatively, Secretary Azar could decide to reopen the record in the Kentucky case and perform the review that Stewart requires. This would seem to be the prudent course, since the decision leaves open the door to Medicaid work experiments, and since multiple state proposals are potentially waiting in the wings. This includes Virginia’s, which would expand coverage but only with work restrictions built in from the outset, thereby presenting the type of net-benefit situation yielded by previously sanctioned 1115 Medicaid experiments.
If the secretary chooses the latter approach, then his findings regarding the coverage impact on Kentucky’s residents would carry enormous import for other approved work demonstrations in Indiana, New Hampshire, and Arkansas. To be sure, Stewart’s holding applies only to Kentucky. But if the secretary nonetheless chooses to simply wait to see if beneficiaries in other work-requirement states mount similar challenges — knowing that the record in Kentucky was fundamentally lacking — that choice could increase the odds that another court will find his conduct arbitrary and capricious, since he would have chosen to deliberately ignore the central issue of a clear instruction by a court. A more sensible course would be to reopen the record in all work demonstrations. If this is the pathway chosen, then the importance of evidence-based comments can hardly be overstated.
Whatever happens next, the bottom line is that Stewart has brought into sharp focus Medicaid’s mission for tens of millions of Americans, regardless of whether their eligibility is tied to old rules or those added by the ACA. The decision also makes clear that the law requires that any attempt to curtail Medicaid’s mission in the name of experimentation must be grounded in real evidence.