April Grady, Manatt Health
During the first half of 2018, the Centers for Medicare & Medicaid Services (CMS) granted four states—Arkansas, Indiana, Kentucky and New Hampshire—Section 1115 waivers permitting them to condition Medicaid coverage for certain enrollees on work or community engagement requirements. An additional 11 states have similar waiver applications under CMS review.1 In June 2018, Arkansas was the first state to implement its work requirement, under which Medicaid enrollees ages 19 to 49 who are part of the expansion group must complete and electronically report at least 80 hours of work or community engagement activities per month if they do not qualify for an exemption.2 Enrollees who fail to meet this standard for three months during a calendar year are disenrolled and locked out of coverage until the following January. In August 2018, three Arkansas Medicaid enrollees filed suit against Secretary of Health and Human Services Alex Azar, CMS Administrator Seema Verma and their respective federal agencies, alleging that their approval of Arkansas’s work requirements exceeded their authority under the Medicaid statute by imposing limits on eligibility for Medicaid that are not permitted. Soon after, in September 2018, Arkansas disenrolled its first cohort of individuals for failure to meet the requirement, resulting in 4,353 Arkansans losing coverage.
Stewart v. Azar Litigation
While Arkansas is the first state to terminate Medicaid coverage for individuals not meeting work requirements, it is the second state to have that requirement tested in federal court. In June, Judge James E. Boasberg of the federal district court in Washington, D.C., in Stewart v. Azar, concluded that CMS had not provided an adequate explanation for its approval of Kentucky’s work requirements and other aspects of its Medicaid waiver and remanded the waiver to CMS for further action. Stewart concluded that Medicaid waivers should further the objectives of the Medicaid program, which include providing health coverage, and that CMS had not adequately explained how Kentucky’s waiver would provide health coverage.
In light of Judge Boasberg’s remand to CMS of the Kentucky waiver, CMS opened a new public comment period on that waiver, with comments due August 18. At the parties’ request, Judge Boasberg modified his opinion in Stewart by striking the portion of the opinion in which he ordered that final judgment be entered. This means the court retains jurisdiction of the case while CMS reconsiders the waiver, and any challenge to CMS’s reapproval of the waiver could be brought by amending the existing Stewart complaint and would continue to be heard by Judge Boasberg. CMS also can now wait until Judge Boasberg has an opportunity to review the new waiver approval, if any, before CMS decides whether to appeal his decision. CMS has not yet issued a new decision on the Kentucky waiver.
Gresham v. Azar Litigation
Meanwhile, Arkansas Medicaid enrollees sued CMS on August 14 to challenge that state’s work requirements, and the state of Arkansas has intervened in the litigation to defend the waiver. As in the Kentucky case, the complaint in the Arkansas case, Gresham v. Azar, alleges that CMS focused on goals other than providing health coverage when approving the waiver, including improving health outcomes and moving enrollees out of poverty. While the outcome in Stewart is not necessarily binding in the Gresham case, Judge Boasberg has ruled that both cases have “common factual and legal issues, many of which are quite complex,” so “in the interests of judicial economy,” he should hear Gresham, despite CMS’s request that the case be assigned randomly to any judge on his court.
Coverage Impact of Arkansas’s Work Requirements
Recently released data from Arkansas provide early insight into the potential coverage impacts of Medicaid work requirements. As noted above, 4,353 individuals were terminated from Arkansas’s Medicaid program in early September for failure to report that they had completed 80 hours of work or community engagement activities in June, July and August. These individuals represent at least 17% of the 25,815 individuals who were phased in to the work requirement in June, and at least 55% of the 7,909 nonexempt individuals who were required to report their work and community engagement activities as of June.3
In its public reporting, Arkansas does not provide any information on the demographic characteristics of the population disenrolled from coverage in September; however, it provides more detailed data on the full population of 60,012 individuals targeted by the work requirement in August. Of these 60,012 enrollees:
- Approximately two-thirds, or 40,190 individuals, did not have a reporting obligation because state administrative data indicated that they were employed for at least 80 hours per month based on their income or that they met one of the exemption criteria.
- Approximately one-third, or 19,822 individuals, had an obligation to log in to Arkansas’s electronic portal and report compliance with or an exemption from the work requirement.
While the state’s use of administrative data has lifted the reporting obligation for many enrollees, the state’s data also suggest that a substantial share of those who continue to have a reporting requirement are at high risk for loss of coverage. Of the 19,822 individuals with a reporting obligation in August, an overwhelming majority (82.5%) either did not report any hours of work or community engagement activities (16,132) or reported less than 80 hours (225). The remaining 17.5% either reported that they completed at least 80 hours per month of work or community engagement activities (1,218) or reported that they met one of the criteria for exemption from the work requirement (2,247).
If current trends continue, the expansion population in Arkansas could face substantial coverage losses. Among those who have been phased in to the work requirement thus far, there are currently 5,076 individuals who have not met the requirement for two months—in addition to the 4,353 who were already disenrolled. If these individuals do not comply with the work requirement or secure an exemption in September, they will be subject to disenrollment and lockout from coverage in early October.
Moving Forward
Nationally, stakeholders are closely watching Arkansas’s experience with work requirements and how early coverage losses will shape legal decisions, state policies and CMS actions going forward. Arkansas’s governor, Asa Hutchinson, has defended the first round of coverage terminations, commenting, “While many fully complied by taking advantage of work opportunities under the work requirement, there were some that either found work, moved onto other insurance, or moved out of state without notifying [the Arkansas Department of Human Services]…Some simply chose not to comply.” Without additional data, it is unclear whether the governor is correct in assuming that many individuals who were disenrolled from coverage no longer needed it; however, it is likely that many are experiencing barriers to reporting compliance, such as insufficient education regarding the requirement or lack of Internet access. A series of recent interviews with 18 Arkansas Medicaid enrollees published in Health Affairs found that two-thirds were unaware of the new work requirement, and Urban Institute estimates that between 21% and 31% of the population targeted by the work requirement have no Internet access at home. These findings have the potential to impact the judicial and administrative review of work requirements in Medicaid waivers in Arkansas, Kentucky and other states.
1One additional state has posted for public comment a waiver application seeking work requirements, but has not yet submitted it to CMS.
2In 2018, only enrollees ages 30 to 49 are subject to the work requirement. The state will phase enrollees ages 19 to 29 in to the work requirement in 2019.
3Arkansas does not provide the number of individuals potentially affected by the work requirement (i.e., expansion adults ages 30 to 49 for whom the requirement is being phased in based on their eligibility date) for all three months—June, July and August. But in light of normal attrition from the program, it is likely that the number of expansion adults potentially affected for all three months was less than 25,815; similarly, it is likely that additional individuals secured exemptions. As a result, the 17% (4,353 out of 25,815 potentially affected individuals) and 55% (4,353 out of 7,909 nonexempt individuals with a reporting obligation) disenrollment figures are likely underestimates.