ISSUE BRIEF SEPTEMBER 2019 Work Requirements in Kentucky Medicaid: A Policy in Limbo Joy Madubuonwu Lucy Chen Benjamin D. Sommers M.P.H. Candidate M.D.–Ph.D. Candidate in Health Policy Professor of Health Policy and Economics Harvard T.H. Chan School of Public Health Harvard University Harvard T.H. Chan School of Public Health ABSTRACT TOPLINES ISSUE: Kentucky was the first state approved to implement a work Nearly half of Kentuckians requirement for adult Medicaid beneficiaries. A federal judge blocked enrolled in Medicaid said they had not heard anything implementation right before it was scheduled to take effect, but the about the state’s plans for program may be reinstated on appeal. a new work requirement; GOAL: To examine several aspects of Kentucky’s Medicaid work minorities and those with less requirements, including awareness and current work activities among education were less likely to Medicaid beneficiaries, and the potential impact of the policy on have heard about the policy. employment and health insurance. About 97 percent of adult METHODS: Findings from telephone surveys of 500 low-income Medicaid beneficiaries surveyed Kentuckians and 1,501 low-income Arkansans, ages 19 to 64, conducted in in Kentucky already meet the late 2018. proposed requirements — either by working 20 hours a week, KEY FINDINGS AND CONCLUSIONS: Nearly half (46%) of Kentuckians conducting other activities like enrolled in Medicaid said they had not heard anything about the state’s job search or training, or being plans for a new work requirement. Minorities and those with less eligible for an exemption like a education were less likely to have heard about the policy. Eighty percent disability. of low-income Kentuckians reported they were unsure whether the requirement was in effect, and only 8 percent knew the policy was not in effect. Nearly all adults in Kentucky Medicaid surveyed were already working or otherwise meeting the state’s proposed work requirements. The survey found similar results in Arkansas, which implemented work requirements in 2018 and experienced higher uninsured rates, no increase in employment, and substantial confusion. Work Requirements in Kentucky Medicaid: A Policy in Limbo 2 BACKGROUND FINDINGS In 2014, Kentucky expanded Medicaid under an executive order made by the governor. Kentucky’s uninsured rate Kentuckians’ Awareness of Medicaid Work among low-income adults dropped from 40 percent Requirements to 7 percent between 2013 and 2016. This expansion Many low-income Kentuckians had not heard about the of coverage was one of the largest increases in health state’s proposed work requirements. When asked whether insurance experienced by any state.1 In early 2018, they had heard about the new requirements, 45 percent Kentucky became the first state to receive approval from of low-income Kentuckians overall and 54 percent of the federal government to implement a “community those with Medicaid said they had heard “a lot” or “a little” engagement requirement” (also known as a “work about work requirements. The remaining 55 percent requirement”) for its Medicaid enrollees.2 The program, of low-income Kentuckians and 46 percent of those in part of the state’s “Kentucky Works” Medicaid waiver, Medicaid reported they had heard “nothing at all” about would have required nonelderly, nonexempt adults to the policy. Not all Kentucky Medicaid beneficiaries were participate in community engagement for 80 hours a subject to the new policy, which likely affected the rates of month. Employment, job searching, volunteering, school, how many had heard about it.5 caretaking of family members, and participating in Awareness of work requirements was characterized by treatment for substance use disorder are all examples of racial and educational disparities (Exhibit 1). Among community engagement activities. self-identified racial minority Medicaid beneficiaries, In June 2018, U.S. District Judge James Boasberg ruled that 39 percent had heard something about the state’s work Kentucky’s waiver was inconsistent with Medicaid’s intent requirements, compared with 57 percent of white and issued a stay that halted the rollout of the Kentucky Medicaid beneficiaries. Forty-eight percent of Medicaid Works program two days before it was supposed to take beneficiaries with less than a high school diploma had effect.3 Meanwhile, Arkansas instituted a similar program heard about the work requirements, compared with 62 for 30-to 49-year-olds that was in effect from June 2018 percent of those with some college education. Awareness until March 2019, until the same judge ordered it to stop. also varied by age group: 41 percent of Medicaid beneficiaries ages 19 to 29 had heard of the policy, In this issue brief, we report findings from a November– compared with 58 percent of those ages 30 to 64 (data not December 2018 survey of 500 low-income Kentuckians shown). A multivariate regression model confirmed that (including 297 who reported having Medicaid coverage) minorities (p=0.04), people with lower levels of education that asked about their understanding of and experiences (p=0.05), and younger adults (p=0.04) were significantly with the state’s work requirements and their current work less likely to know about the policy. There were no or other community engagement activities. We compare significant differences by gender, marital status, or urban the results of the survey to similar estimates from our vs. rural residence. survey of 1,501 low-income adults in Arkansas, the first state to implement work requirements. Details on the There was also substantial confusion about the legal survey methodology have been published previously.4 status of the work requirement program. As of late 2018, 80 percent of low-income Kentuckians were not sure whether the state’s work requirements were in effect. Twelve percent incorrectly thought they were in effect; only 8 percent knew the policy was not in effect (data not shown). This high level of confusion may have dissuaded some eligible individuals from enrolling in Medicaid. commonwealthfund.org Issue Brief, September 2019 Work Requirements in Kentucky Medicaid: A Policy in Limbo 3 Exhibit 1 Exhibit 1. Share of KentuckiansMedicaid Who Have Heard About Work Requirements, Share of Kentuckians in in Medicaid Who Have Heard About Work Requirements, by Race and Education and Education by Race Percent 62 57 48 39 Racial minority White Less than a high Any college school diploma Data: Authors’ analysis of data from a telephone survey conducted in late 2018 of 297 low-income Kentuckians on Medicaid (ages 19–64). Communityof data from a telephone survey conducted in late 2018 of 297 low-income Kentuckians men (p=0.09). Meanwhile, there were no significant Data: Authors’ analysis Engagement Activities in Kentucky’s and on Medicaid (ages 19–64). Medicaid Population differences in rates of community engagement activities Source: Joy Madubuonwu, Lucy Chen, and Benjamin D. Sommers, Work Requirements in Kentucky Medicaid: A Policy in Limbo Our survey assessed whether adults who were enrolled (Commonwealth Fund, Sept. 2019). based on age, marital status, or urban vs. rural residence. in Kentucky’s Medicaid program were currently working, Work requirements in the Kentucky Medicaid program were engaging in other community activities that would may struggle to meaningfully increase employment since satisfy the state’s requirement, or were disabled or qualified the vast majority of Kentuckians appear to be exempt or for another exemption. Most adults were either working are already be participating in actions that would fulfill or had a disability: 34 percent reported working at least the requirements. 20 hours a week, and 44 percent reported that they were unable to work because of a disability (Exhibit 2). Among the remainder, 11 percent were spending at least 20 hours a How Do Kentucky and Arkansas Compare? week in school, caring for a family member, or performing Asking several similar questions in Arkansas, we found community service; 8 percent were spending at least 20 that 67 percent of Arkansans targeted by that state’s hours a week looking for work or in job training. This left work requirements (Medicare beneficiaries ages 30 to 49) only 3 percent of Kentuckians in Medicaid who would not had heard of the policy as of late 2018, compared with have satisfied the state’s proposed community engagement 54 percent in the Kentucky Medicaid sample (ages 19 to requirements. These results are similar to another recent 64), a statistically significant difference (p=0.007, Exhibit 3). In Arkansas, rates of awareness were significantly analysis of Medicaid beneficiaries in the state.6 lower among younger adults, men, and people with A multivariate regression model found that the groups less education.7 Similar to Kentucky, Arkansas engaged least likely to satisfy the state’s requirements were those in a significant education campaign, with mailings, who had not attended college (p=0.04), whites (p=0.07), information fliers, and other efforts designed to inform commonwealthfund.org Issue Brief, September 2019 Work Requirements in Kentucky Medicaid: A Policy in Limbo 4 Exhibit 2 Exhibit 2. All Kentucky Medicaid Recipients Report Activities That as Community Nearly Nearly All Kentucky Medicaid Recipients Report Activities That Qualify Qualify as Engagement Engagement Community Percent 3 8 Working at least 20 hours a week 11 34 Unable to work because of a disability Spending at least 20 hours a week in school, caring for family members, or participating in community service Spending at least 20 hours a week looking for work or in job training Would likely not satisfy the state’s proposed community engagement requirements 44 Data: Authors’ analysis of data from a telephone survey conducted in late 2018 of 297 low-income Kentuckians on Medicaid (ages 19–64). Outcomes were assessed in a mutually exclusive hierarchy, such that if someone reported working more than 20 hours per week, then we did not assess whether the person was disabled or fulfilled another category of community engagement. Data: Authors’ analysis of data from a telephone survey conducted in late 2018 of 297 low-income Kentuckians on Medicaid (ages 19–64). Outcomes were assessed in a mutually exclusive hierarchy, such that if someone reported working more than 20 hours per week, then we did not assess whether the person was disabled or fulfilled another category of community engagement. Exhibit 3 Exhibit 3. Share of Medicaid BeneficiariesPotentially Affected byRequirements Who Have Heard Share of Medicaid Beneficiaries Potentially Affected by Work Work Requirements Source: Joy Madubuonwu, Lucy Chen, and Benjamin D. Sommers, Work Requirements in Kentucky Medicaid: A Policy in Limbo (Commonwealth Fund, Sept. 2019). About the Policy — Kentucky vs. Arkansas — Kentucky vs. Arkansas Who Have Heard About the Policy Percent 67 54 Kentucky Arkansas Data: Authors’ analysis of data from a telephone survey conducted in late 2018 of 297 low-income Kentuckians on Medicaid (ages 19–64) and 429 low-income Arkansans subject to that state’s work requirements (ages 30–49 enrolled in Medicaid or marketplace coverage in the past year). commonwealthfund.org a telephone survey conducted in late 2018 of 297 low-income Kentuckians on Medicaid (ages 19–64) and 429 low-income Arkansans subject to that state’s work Data: Authors’ analysis of data from Issue Brief, September 2019 requirements (ages 30–49 enrolled in Medicaid or marketplace coverage in the past year). Work Requirements in Kentucky Medicaid: A Policy in Limbo 5 beneficiaries about the new policy.8 One critical difference CONCLUSION is that the court blocked Kentucky’s policy roughly six The Kentucky work requirement has been blocked by months before the survey was conducted. In contrast, the courts, but the state’s leaders remain committed to Arkansas’s policy had been in effect for six months at the the policy. The decision is under appeal and a ruling is time of our survey and was only halted by the courts after possible later in the year.12 Despite this, many low-income survey completion. This undoubtedly had an impact on adults in Kentucky are not aware of the requirements, and the extent to which Kentucky residents were aware of the there is substantial confusion about whether the policy policy, although the state had been conducting outreach is currently in effect. Further, approximately 97 percent and education efforts up until the program was blocked of adults surveyed were already meeting the proposed by the courts on June 29, 2018, just two days before the requirements — either by working 20 hours a week, requirements were set to become effective.9 These results conducting other community engagement activities like suggest that informational barriers to Kentucky’s work job search or training, or being eligible for an exemption requirements program may be an important challenge if like a disability. These results indicate there may be the policy goes into effect in the future. relatively few Medicaid beneficiaries who might be Similar to Kentucky, we found that more than 95 percent compelled to work by the new policy. of low-income adults in Arkansas were already meeting Arkansas implemented work requirements in 2018, and the state’s community engagement requirements. Despite our survey showed similar patterns as in Kentucky — this, there were significant losses of coverage in Arkansas significant confusion, lack of awareness about the after work requirements were implemented with no policy, and most adults already satisfying the proposed change in employment. The uninsured rate among requirements. Further, the first six months of the policy 30-to-49-year-olds (i.e., the age group targeted by the were associated with a significant loss in coverage, eroding policy) increased from 10.5 percent to 14.5 percent while some of the improvements documented under the state’s employment did not change significantly. These results Medicaid expansion,13 but no employment gains. appear partly because of substantial confusion about the new policy and difficulty navigating the system’s reporting These results highlight the potential effects of a Medicaid requirements. Only 22 percent of Arkansans who were work requirement and challenges that policymakers in likely subject to work requirements thought that they Kentucky and other states must address if they implement were affected by it; 44 percent were unsure if the new rules a similar policy. To reduce potential coverage losses, applied to them. Only half of those required to report 10 states should increase awareness of work requirements, information to the state were doing so, citing confusion especially among minority groups and people with lower and lack of Internet access as barriers. In his March 2019 levels of education, and reduce the administrative burden ruling, Judge Boasberg expressed concern that the state’s of reporting. Research to evaluate these ongoing state “outreach efforts may well be falling severely short.” 11 efforts will be essential. commonwealthfund.org Issue Brief, September 2019 Work Requirements in Kentucky Medicaid: A Policy in Limbo 6 NOTES 1. “Health Insurance Coverage of Adults 19–64 Living 7. Sommers et al., “Medicaid Work Requirements,” 2019. in Poverty (under 100% FPL),” Henry J. Kaiser Family Foundation, 2013; and Benjamin D. Sommers et al., “Three- 8. Arkansas Department of Human Services. “Healthcare Year Impacts of the Affordable Care Act: Improved Medical Notice Summary: You Are Subject to the Work Requirement,” Care and Health Among Low-Income Adults,” Health accessed Mar. 29, 2019 (no longer available online). Affairs Web First, published online May 17, 2017. 9. Amy Goldstein, “The Nation’s First Medicaid Work Rules 2. Phil Galewitz, “Kentucky Is First State Granted Approval Loom, and Many Fear Losing Health Coverage,” Washington Post, Jan. 19, 2018; and Lena H. Sun and Amy Goldstein, for Medicaid Work Requirements,” Kaiser Health News, Jan. “Kentucky’s Requirement That Medicaid Recipients Must 12, 2018. Work Is Blocked by a Federal Judge,” Washington Post, June 29, 2018. 3. Amy Goldstein, “Federal Judge Blocks Medicaid Work Requirements in Kentucky and Arkansas,” Washington Post, 10. Sommers et al., “Medicaid Work Requirements,” 2019. Mar. 27, 2019. 11. U.S. District Judge James E. Boasberg, “Memorandum 4. Benjamin D. Sommers et al., “Medicaid Work Opinion,” Gresham v Azar, U.S. District Court for the Requirements — Results from the First Year in Arkansas,” District of Columbia, Mar. 27, 2019. New England Journal of Medicine, published online June 19, 2019. 12. Dylan Scott, “The Legal Fight over the Trump Administration’s Most Aggressive Play to Cut Medicaid, 5. Atheendar S. Venkataramani et al., “Assessment Explained,” Vox, Mar. 14, 2019. of Medicaid Beneficiaries Included in Community Engagement Requirements in Kentucky,” JAMA Network 13. Benjamin D. Sommers et al., “Changes in Utilization Open, July 17, 2019. and Health Among Low-Income Adults After Medicaid Expansion or Expanded Private Insurance,” JAMA Internal 6. Venkataramani et al., “Assessment of Medicaid,” 2019. Medicine, published online Aug. 8, 2016. commonwealthfund.org Issue Brief, September 2019 Work Requirements in Kentucky Medicaid: A Policy in Limbo 7 ABOUT THE AUTHORS ACKNOWLEDGMENTS Joy Madubuonwu is an M.P.H. candidate in health policy Support for this research was provided in part by the and management at the Harvard T.H. Chan School of Robert Wood Johnson Foundation’s Policies for Action Public Health. Her research interests are in U.S. health care program, and Baylor Scott & White. The views expressed delivery reform, maternal and child health care, and access here do not necessarily reflect the views of these to affordable health care for vulnerable populations. organizations. Before attending Harvard, Ms. Madubuonwu was an Urban Fellow for New York City’s First Deputy Mayor Tony Shorris and then a health policy analyst for Deputy Mayor for Health and Human Services Herminia Palacio. She For more information about this brief, please contact: graduated from New York University with a B.A. in global Benjamin D. Sommers, M.D., Ph.D. public health and medical anthropology. Professor of Health Policy and Economics Department of Health Policy and Management Lucy Chen is an M.D.–Ph.D. candidate in Health Policy at Harvard T.H. Chan School of Public Health Harvard University. Her research interests are in physician bsommershsph.harvard.edu behavior, U.S. health care delivery reform, and access to care for vulnerable populations. Ms. Chen graduated from the University of Pennsylvania’s Life Sciences and Management Program with a B.S. in economics and B.A. in biology. Benjamin D. Sommers, M.D., Ph.D., is professor of health policy and economics in the Department of Health Policy and Management at the Harvard T.H. Chan School of Public Health. His main research interests are health policy for vulnerable populations, the uninsured, and the health care safety net. Sommers has received numerous awards for his research, including the Outstanding Dissertation Award, Alice Hersh New Investigator Award, and Article of the Year Award from AcademyHealth. He is a practicing primary care internist and an associate professor of medicine at Brigham & Women’s Hospital and Harvard Medical School. From 2011–2012, Sommers served as a senior advisor in the Office of the Assistant Secretary for Planning and Evaluation, U.S. Department of Health and Human Services. He received his medical degree from Harvard Medical School and his doctorate in health policy from Harvard University. Editorial support was provided by Deborah Lorber. commonwealthfund.org Issue Brief, September 2019 About the Commonwealth Fund The mission of the Commonwealth Fund is to promote a high-performing health care system that achieves better access, improved quality, and greater efficiency, particularly for society’s most vulnerable, including low-income people, the uninsured, and people of color. Support for this research was provided by the Commonwealth Fund. The views presented here are those of the authors and not necessarily those of the Commonwealth Fund or its directors, officers, or staff.