A Disturbing Trend of Hiding the Coverage Losses is Emerging in Medicaid Waivers

For those of us closely watching the action on new Section 1115 Medicaid research and demonstration waivers that the Trump Administration has been inviting and approving this year, the recent court decision putting a hold on Kentucky’s plans was welcome news.

The court ruled that the Secretary’s decision in granting the Medicaid waiver was “arbitrary and capricious.” A key point in coming to the court’s conclusion was the impact of taking health insurance away from thousands of Medicaid beneficiaries. The judge wrote:

For starters, the Secretary never once mentions the estimated 95,000 people who would lose coverage, which gives the Court little reason to think that he seriously grappled with the bottom- line impact on healthcare. (P. 38)

Throughout the oral arguments and the written decision, this estimate of 95,000 Kentuckians losing their Medicaid coverage was cited as a key piece of evidence in the judge’s determination that the decision was arbitrary and capricious.

Let’s take a minute to unpack where this 95,000 number came from. All states are required to submit budget neutrality projections for Section 1115 waivers. This policy is meant to ensure that the federal government doesn’t spend more money under the proposed waiver than it would have without the waiver. These cost projections are primarily built on two key factors – enrollment projections (represented as per member months) and cost growth (represented as monthly cost for different groups included in the waiver).

So the 95,000 estimate comes from doing some basic math to the numbers provided by the state itself in its budget neutrality projections negotiated with the federal government.1 This estimate is likely too low, as the court noted, an amicus brief filed in the Stewart v. Azar case by health care scholars suggests that this estimate of the number of people losing Medicaid coverage would more likely fall between 175,000 and 297,500. (P. 37) Regardless of which number winds up to be a more accurate estimate of how many people will lose coverage, the public needs to know that it will be substantial, and states should not hide the truth.

The state of Kentucky and the Trump Administration were not overtly forthcoming on how many people would lose coverage but the budget neutrality estimate provided that information to those who can read the spreadsheets. In fact, federal regulations explicitly require that waiver applications that are open for public comment at the state level provide “an estimate of the expected increase or decrease in annual enrollment” (Section 431.408(a)(1)((C)). If this information is not forthcoming in the budget neutrality assumptions then it should be provided elsewhere.

The state of Kentucky and the Trump Administration are now facing the reality of being hoisted on their own budget neutrality petard, so to speak. Other states seem to be taking note and may be trying to avoid the same fate by withholding even more information from the public.

In the middle of June, before the court ruled on the Kentucky waiver, I blogged about South Dakota’s Section 1115 waiver and noted that the waiver application did not include this critical piece of information. South Dakota has already finished its state public comment process.

Unfortunately, since I wrote the South Dakota blog, more states are considering applying these draconian proposals to very poor families and are obfuscating this critical point. Just last week Oklahoma posted its Section 1115 application work requirement proposal and also claimed that it will have no impact on enrollment projections used in the budget neutrality assumptions. The state also fails to provide any coverage loss estimates – even though the application itself says in numerous places that the intent of the project is to reduce Medicaid enrollment. South Carolina’s recent draft waiver proposal doesn’t include any budget neutrality estimates or enrollment projections either.

Perhaps most disturbing of all, a revised version of Mississippi’s proposed Medicaid work requirement waiver on very poor parents has already gone through the state and federal public comment process and is now posted on the CMS website with no budget neutrality spreadsheet at all. Mississippi previously had budget neutrality estimates that showed a decline in enrollment as we discussed in our report.

It is obvious why states may go down this path – especially since the Kentucky ruling underscores that the question of coverage losses is critical. What is less obvious is how this all will end.

  1. Addressing the question of whether these 95,000 persons are likely to become uninsured, the court noted: “Left with little else, the Secretary now argues that perhaps the 95,000 individuals would not lose coverage after all; instead, maybe they will simply transition to “employer-sponsored and commercial coverage.” CMS Br. at 11 (quoting AR 4, 7). It made no such finding below, however. While the agency spoke generally of “creating incentives for individuals to obtain and maintain coverage through private, employer-sponsored insurance,” AR 5, it cited no research or evidence that this would happen, nor did it make concrete estimates of how many beneficiaries might make that transition. And, of course, it is not obvious that the community-engagement requirement alone would help a person shift to private insurance. As the Secretary stresses, this is not a work requirement; individuals can meet it, for example, by volunteering in the community. While those unpaid activities may have long-term benefits, he never discussed how they will promote a “transition from Medicaid to commercial coverage.” (P. 40)
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.