HEALTH REFORM MONITO RING SURVEY Many Adults Targeted by Medicaid Work Requirements Face Barriers to Sustained Employment Michael Karpman May 2019 AT A GLANCE  Medicaid enrollees who would potentially be subject to work requirements (i.e., “nonexempt”) disproportionately face employment barriers, such as lack of a high school degree, health problems, limited transportation and internet access, criminal records, and residence in high- unemployment neighborhoods.  Likely reflecting these barriers, fewer than one in six potentially nonexempt Medicaid enrollees reported working at least 20 hours per week for all or nearly all weeks in the past year, compared with over three in five privately insured adults.  Half of nonexempt Medicaid enrollees reported issues related to the labor market or nature of employment (e.g., difficulty finding work, restricted work schedules) as reasons for not working more, and over one-quarter reported health reasons. Since January 2018, the Centers for Medicare & Medicaid Services (CMS) has approved Section 1115 demonstration waivers permitting nine states to adopt work requirements for Medicaid enrollees, with additional waivers pending approval in six other states.1 These waivers typically require nonelderly, nonpregnant adults who qualify for Medicaid on a basis other than disability to work or participate in work-related activities for 20 hours per week or 80 hours per month for most or all months to receive Medicaid, unless exempted from the requirement (Musumeci, Garfield, and Rudowitz 2018).2 In March 2019, a federal court vacated the approval of waivers in Arkansas and Kentucky, ruling that the approvals did not address how the waivers would promote Medicaid’s core objective of providing medical assistance.3 However, the administration has appealed both rulings and approved a new waiver that included work requirements as part of a partial Medicaid expansion in Utah shortly after the decisions.4 Despite states’ rapid adoption of work requirements and recent legal challenges, limited information has been available to assess the potential barriers current Medicaid enrollees would face in meeting these requirements on a sustained basis and the risk that these employment barriers will result in loss of health insurance coverage under work requirement policies. Using data from the September 2018 and March 2019 rounds of the Health Reform Monitoring Survey (HRMS) for a nationally representative sample of nonelderly adults, this study examines work patterns and potential barriers to employment among “nonexempt” Medicaid enrollees—those who would most likely be subject to work requirements, based on standards proposed in Kentucky—and compares this group with both Medicaid enrollees who would likely be exempt and privately insured adults. The analysis finds that both nonexempt and exempt Medicaid enrollees are more likely than privately insured adults to face potential employment barriers, such as lack of a high school degree, limited English proficiency, physical and mental health issues, and residence in a high-unemployment neighborhood. Nonexempt enrollees are also more likely than both exempt enrollees and privately insured adults to report multiple chronic health conditions, lack of household internet access, limited transportation access, and criminal records. These and other barriers likely contribute to the lower rates of sustained employment at or above 20 hours per week among nonexempt Medicaid enrollees relative to those with private coverage. Though most of these enrollees (61.6 percent) worked during the past year, fewer than one in six worked at least 20 hours per week for all or nearly all weeks, compared with over three in five privately insured adults. Issues related to the labor market or nature of employment (e.g., difficulty finding work, restricted work schedules) and health issues were the most common reasons nonexempt enrollees reported for not working more. These potential barriers raise concerns that introducing work requirements in Medicaid will result in loss of coverage for many enrollees without increasing their employment and underscore the importance of placing education, training, other employment services, and work supports at the center of policy initiatives to improve employment outcomes. What We Did This study used data from a sample of 15,031 nondisabled adults ages 19 to 64 who participated in the September 2018 and March 2019 rounds of the Health Reform Monitoring Survey,5 including 1,180 nondisabled adults with Medicaid and 12,097 with private coverage at the time of the survey.6 Though recent studies have drawn on federal surveys to explore the implications of work requirements (Bauer, Schanzenbach, and Shambaugh 2018; Gangopadhyaya and Kenney 2018; Garfield, Rudowitz, and Damico 2018; Wen, Saloner, and Cummings 2019),7 no single federal survey collects data on work experience over time, reasons for not working, and potential barriers to sustained employment among current Medicaid enrollees. The September 2018 and March 2019 HRMS collected detailed data on work patterns and potential individual-level employment barriers, which were grouped 2 ADULTS TARGETED BY M EDICAID WORK REQUIREMENTS FACE EMPLOYMEN T BARRIERS into the following categories: skill-related; health-related; search or participation–related (i.e., factors that make it difficult to look for or get to work); and hiring-related (i.e., factors other than skills or experience that may prevent hiring). Potential neighborhood-level employment barriers were assessed by merging 2013–17 American Community Survey census tract-level estimates of unemployment, poverty, and average commute times to the HRMS. 8 Employment barriers and work patterns are compared across three groups of nondisabled adults: 1. Medicaid enrollees potentially subject to work requirements (i.e., “nonexempt”) if implemented nationally, based on exemption criteria from Kentucky’s demonstration waiver, which was the first waiver with work requirements approved by CMS 2. Medicaid enrollees likely to be exempt 3. Privately insured adults, who constitute the majority of the working population and provide a standard for the self-sufficiency goals that work requirement policies are designed to promote Nearly 60 percent of nondisabled Medicaid enrollees in the sample would likely be exempt because they were pregnant in the past year, full-time students in the past year, or reported being primary caregivers of a dependent child or an adult family member who needs help with daily or household activities (table 1). States must also exempt adults deemed medically frail but have flexibility in defining these criteria, and Kentucky’s definition of medical frailty has been unclear. 9 A sensitivity test using health-related activity or functional limitations as a proxy for medical frailty exemption yielded little change in the study results. TABLE 1 Characteristics Used to Determine Exemptions from Work Requirements among Nondisabled Adults Ages 19–64, by Coverage Type, September 2018/March 2019 Medicaid Privately insured Reported being pregnant in past year 8.5% 2.4%*** Reported being a full-time student in past year 11.2% 8.9% Reported being a primary caregiver for a dependent child 45.1% 33.6%*** Reported being a primary caregiver for an adult family member 9.2% 5.7%*** Any of the above 59.6% 44.6%*** None of the above 40.4% 55.4%*** Sample size 1,180 12,097 Source: Health Reform Monitoring Survey, quarter 3 2018 and quarter 1 2019. Notes: Nondisabled is defined as those not receiving Supplemental Security Income or Social Security Disability Income in the past year or enrolled in Medicare at the time of the survey. */**/*** Estimate differs significantly from adults enrolled in Medicaid at the 0.10/0.05/0.01 level, using two-tailed tests. Survey responses on type of health insurance coverage, exemptions from work requirements, work patterns, and employment barriers are subject to measurement error.10 For instance, because respondents often have difficulty reporting their type of health insurance in surveys (Pascale 2008), a ADULTS TARGETED BY M EDICAID WORK REQUIREMENTS FACE EMPLOYMEN T BARRIERS 3 logical editing process is used to assign coverage type to adults participating in the HRMS (Blavin, Karpman, and Zuckerman 2016). Using Kentucky’s waiver to assign exemptions from work requirements provides a conservative estimate of the share of adults who would likely be subject to such requirements if they were implemented nationally, because other state waivers typically contain narrower exemptions (e.g., only exempting caregivers of children if one of their children is under age 6). Some limitations to the data likely result in further underestimation of the share of nonexempt adults. For instance, primary caregiver status is self-reported and may therefore be overreported in the HRMS; Kentucky only allows one caregiver exemption per household. Other limitations may lead to overestimating the share of Medicaid enrollees who may be subject to work requirements: the survey did not collect data on compliance with TANF or SNAP work requirements, which automatically establishes compliance with Medicaid work requirements, or on whether respondents younger than 26 aged out of foster care and would therefore be exempt. Previous research has found that respondents may underreport jobless spells when recalling past- year work experience (Horvath 1982). In addition, Medicaid enrollees who work 80 hours per month would meet work requirement standards even if they worked less than 20 hours in some weeks. A study using different exemption criteria found a higher share of enrollees consistently meeting these standards based on average monthly hours relative to the estimates in this brief (Bauer, Schanzenbach, and Shambaugh 2018). Finally, the HRMS did not collect data on participation in allowable nonwork activities, such as job training or community service, that Medicaid enrollees can use to meet work requirements. What We Found Medicaid enrollees who would potentially be subject to work requirements disproportionately face employment barriers. Relative to privately insured adults, Medicaid enrollees were more likely to have characteristics that place them at risk of facing employment barriers. For example, both nonexempt and exempt Medicaid enrollees were over five times as likely as privately insured adults to lack a high school degree and over four times as likely to have limited English proficiency, increasing their exposure to the unstable low-wage job market (figure 1). Over one-quarter (26.2 percent) of nonexempt Medicaid enrollees reported multiple chronic health conditions,11 compared with less than 20 percent of exempt enrollees and privately insured adults. Both groups of Medicaid enrollees were more likely than their privately insured peers to report activity or functional limitations (that have not qualified them for federal disability benefits) and having ever been diagnosed with a mental health disorder. 4 ADULTS TARGETED BY M EDICAID WORK REQUIREMENTS FACE EMPLOYMEN T BARRIERS FIGURE 1 Individual-Level Employment Barriers among Nondisabled Adults Ages 19–64, by Coverage Type and Potential Exposure to Work Requirements, September 2018/March 2019 Medicaid, potentially nonexempt Medicaid, likely exempt Privately insured from (i.e., subject to) work requirements from work requirements Skill barriers 21.6% Did not complete high school 26.6%** 4.1%***^^^ 7.2% Limited English proficiency 9.3% 1.7%**^^ Health barriers 10.6% Activity or functional limitation 9.8% 5.3%***^^ 26.2% Multiple chronic conditions 19.6%* 17.8%*** 27.7% Ever diagnosed with a mental health condition 26.7% 16.7%***^^^ Search/participation barriers 22.3% Lack of household internet access 10.9%*** 4.3%***^^ 14.6% Limited access to transportation 8.4%* 2.8%***^^^ Hiring barriers 15.2% Criminal record 10.5%* 5.1%***^^^ 7.8% Formerly incarcerated 6.6% 1.7%***^^^ 43.4% Black or Hispanic 49.1%*** 23.1%***^^^ URBAN INSTITUTE Source: Health Reform Monitoring Survey, quarter 3 2018 and quarter 1 2019. Notes: Mental health conditions include anxiety, depression, mood, bipolar, schizoaffective, and schizophrenic disorders. Limited access to transportation includes lack of a household vehicle and fair or poor ability to get around without driving. ADULTS TARGETED BY M EDICAID WORK REQUIREMENTS FACE EMPLOYMEN T BARRIERS 5 Criminal record includes those who were arrested in the last seven years or were ever convicted or found delinquent of, or pleaded guilty to, any charges. Formerly incarcerated includes those ever sentenced to time in prison, jail, or a juvenile detention center. Likely exempt Medicaid enrollees were full-time students in the past year, pregnant in the past year, or primary caregivers. */**/*** Estimate differs significantly from potentially nonexempt Medicaid enrollees at the 0.10/0.05/0.01 level, using two-tailed tests. ^/^^/^^^ Estimate differs significantly from likely exempt Medicaid enrollees at the 0.10/0.05/0.01 level, using two-tailed tests. More than 22 percent of nonexempt adults with Medicaid do not have internet access at home, compared with 10.9 percent of exempt enrollees and 4.3 percent of the privately insured. Though access may be available in libraries or other venues, lack of household access could hinder the ability to search for jobs or consistently report work hours. Nonexempt Medicaid enrollees were also more likely than both exempt enrollees and privately insured adults to have limited access to transportation (i.e., no vehicle and fair or poor ability to get around without driving). Over 15 percent of nonexempt Medicaid enrollees were likely to have a criminal record that would appear in a background check based on being arrested in the past seven years or ever being convicted or pleading guilty to any charges,12 compared with 10.5 percent of exempt enrollees and 5.1 percent of privately insured adults.13 Nonexempt enrollees were over four times more likely than the privately insured to report having ever been incarcerated. Medicaid enrollees are also disproportionately black or Hispanic, groups that face persistent hiring discrimination (Quillian et al. 2017). Overall, 82.9 percent of nonexempt enrollees reported at least one characteristic that may pose a barrier to employment, compared with 53.2 percent of privately insured adults (data not shown). Neighborhood characteristics may also affect employment opportunities. Both nonexempt and exempt Medicaid enrollees are more than twice as likely as privately insured adults to live in census tracts with unemployment that is over twice the national average (figure 2). Nearly 40 percent of nonexempt and 46.1 percent of exempt Medicaid enrollees live in neighborhoods with poverty rates of 20 percent or higher. These neighborhoods are often marked by weaker hiring networks and isolation from job opportunities. There were no statistically significant differences in the shares of adults in each group living in neighborhoods with long average commute times. 6 ADULTS TARGETED BY M EDICAID WORK REQUIREMENTS FACE EMPLOYMEN T BARRIERS FIGURE 2 Neighborhood-Level Employment Barriers among Nondisabled Adults Ages 19–64, by Coverage Type and Potential Exposure to Work Requirements, September 2018/March 2019 Medicaid, potentially nonexempt Medicaid, likely exempt Privately insured from (i.e., subject to) work requirements from work requirements 46.1% 39.8% 25.2% 21.8% 22.9% 18.7%***^^^ 12.8% 10.7% 5.1%***^^^ Unemployment rate over twice Poverty rate of 20 percent or more Highest quartile of average commute national average time URBAN INSTITUTE Source: Health Reform Monitoring Survey, quarter 3 2018 and quarter 1 2019; American Community Survey five-year estimates, 2013–17. Notes: Neighborhood unemployment rates are for adults ages 20–64 in the census tract and are based on tract-level data for 2013–17. The national average tract-level nonelderly adult unemployment rate for 2013–17 for adults in the sample was 6.3 percent. Neighborhood poverty rates and commute times are for all ages and are based on census tract-level data for 2013–17. Likely exempt Medicaid enrollees were full-time students in the past year, pregnant in the past year, or primary caregivers. */**/*** Estimate differs significantly from potentially nonexempt Medicaid enrollees at the 0.10/0.05/0.01 level, using two-tailed tests. ^/^^/^^^ Estimate differs significantly from likely exempt Medicaid enrollees at the 0.10/0.05/0.01 level, using two-tailed tests. Likely reflecting these barriers, fewer than one in six potentially nonexempt Medicaid enrollees reported working at least 20 hours per week for all or nearly all weeks in the past year, compared with over three in five privately insured adults. Given their greater prevalence of potential employment barriers, Medicaid enrollees were less likely than privately insured adults to achieve sustained employment at the levels required by work requirement policies being proposed in Medicaid. Though 61.6 percent of nonexempt enrollees worked in the past year, only 14.9 percent reported working at least 20 hours per week for all or nearly all weeks (figure 3). Another 19.3 percent reported usually working that many hours, 23.6 percent usually worked less, and 37.8 percent did not work at all, which also poses a barrier to future employment. Thus, without increased and more consistent employment, many nonexempt enrollees would be at risk of failing to comply or transitioning in and out of compliance if subject to work requirements. ADULTS TARGETED BY M EDICAID WORK REQUIREMENTS FACE EMPLOYMEN T BARRIERS 7 FIGURE 3 Work Experience in the Past Year of Nondisabled Adults Ages 19–64, by Coverage Type and Potential Exposure to Work Requirements, September 2018/March 2019 Medicaid, potentially nonexempt Medicaid, likely exempt Privately insured from (i.e., subject to) work requirements from work requirements 14.9% Worked 20 or more hours per week for 17.9% all or nearly all weeks in past year 61.9%***^^^ 19.3% Worked 20 or more hours per week for most weeks in past year 16.0% 12.9%** 23.6% Usually worked less than 20 hours per week or worked only some weeks in past year 28.6% 11.2%***^^^ 37.8% Did not work in past year 35.6% 12.7%***^^^ URBAN INSTITUTE Source: Health Reform Monitoring Survey, quarter 3 2018 and quarter 1 2019. Notes: Overall, 61.6 percent of potentially nonexempt Medicaid enrollees, 64.2 percent of likely exempt Medicaid enrollees, and 87.2 percent of privately insured adults worked in the past year. Estimates are not shown for the 4.3 percent of potentially nonexempt Medicaid enrollees, 2.0 percent of likely exempt Medicaid enrollees, and 1.3 percent of privately insured adults who did not report work experience, weeks worked, or usual hours worked in the past year. Likely exempt Medicaid enrollees were full-time students in the past year, pregnant in the past year, or primary caregivers. */**/*** Estimate differs significantly from potentially nonexempt Medicaid enrollees at the 0.10/0.05/0.01 level, using two-tailed tests. ^/^^/^^^ Estimates differs significantly from likely exempt Medicaid enrollees at the 0.10/0.05/0.01 level, using two-tailed tests. Half of nonexempt Medicaid enrollees reported issues related to the labor market or the nature of their employment as reasons for not working more, and over one-quarter reported health reasons. Among nonexempt Medicaid enrollees who did not always work 20 hours per week, about half (49.9 percent) reported not working or not working more hours for reasons related to the labor market or the nature of their employment, such as difficulty finding work, employer restrictions on their work schedule, employment in temporary positions, or reduced hours because business was slow (figure 4). 8 ADULTS TARGETED BY M EDICAID WORK REQUIREMENTS FACE EMPLOYMENT BARRIERS Over one-quarter (27.5 percent) reported health as the reason for not working more. Nonexempt Medicaid enrollees were 20 percentage points less likely than privately insured adults to report that they did not want to work more or were retired (15.6 percent versus 35.6 percent). Transportation was a problem for 10.1 percent of nonexempt Medicaid enrollees and 2.7 percent of the privately insured. Among exempt enrollees, the most common reasons for not working more owed to caregiving and school attendance, followed by factors related to the labor market or nature of employment. FIGURE 4 Reasons for Not Always Working at Least 20 Hours Per Week in Past Year among Nondisabled Adults Ages 19–64, by Coverage Type and Potential Exposure to Work Requirements, September 2018/March 2019 Medicaid, potentially nonexempt Medicaid, likely exempt Privately insured from (i.e., subject to) work requirements from work requirements 49.9% Labor market reasons/nature of employment 30.8%*** 27.2%*** 27.5% Health reasons 22.2% 13.5%***^^^ 15.6% Did not want to work or work more hours/retired 10.2% 35.6%***^^^ 10.1% Lack of transportation 7.4% 2.7%***^^ 19.0% Caregiving, school attendance, or other reasons 56.8%*** 37.2%***^^^ URBAN INSTITUTE Source: Health Reform Monitoring Survey, quarter 3 2018 and quarter 1 2019. Notes: Reasons related to the labor market or nature of employment include the following: could not find work or work with more hours; employer restrictions on work schedule; reduced hours because business was slow; and seasonal, temporary, or part-time work. Examples of "other reasons" include weather, moving, immigration status, and family-related reasons. Respondents could report multiple reasons. Likely exempt Medicaid enrollees were full-time students in the past year, pregnant in the past year, or primary caregivers. */**/*** Estimate differs significantly from potentially nonexempt Medicaid enrollees at the 0.10/0.05/0.01 level, using two-tailed tests. ^/^^/^^^ Estimate differs significantly from likely exempt Medicaid enrollees at the 0.10/0.05/0.01 level, using two-tailed tests. ADULTS TARGETED BY M EDICAID WORK REQUIREMENTS FACE EMPLOYMEN T BARRIERS 9 Linear probability models were used to estimate the association between employment barriers and the work patterns and reasons for not working reported by adults enrolled in Medicaid or private coverage.14 Holding demographic characteristics constant, low educational attainment, limited English proficiency, each health barrier, lack of household internet access, limited transportation access, and criminal justice system involvement were associated with a lower likelihood of always or almost always working 20 or more hours per week (table 2). Low educational attainment, health-related limitations, chronic conditions, lack of internet access, and criminal justice system involvement were positively associated with working below that threshold for reasons related to the labor market or nature of employment. TABLE 2 Association of Individual and Neighborhood Characteristics with Work Experience in Past Year among Nondisabled Adults Ages 19–64 with Medicaid or Private Coverage, September 2018/March 2019 Linear probability models DEPENDENT VARIABLES Sometimes or Always Worked 20 or Worked Less than 20 Sometimes or Always More Hours Per Hours Per Week in Past Worked Less than 20 Week for All or Year for Reasons Related Hours Per Week in Nearly All Weeks to Labor Market or Nature Past Year for Health in Past Year of Employment Reasons Coefficient Coefficient Coefficient Skill barriers Did not complete high schoola -0.343***^^^ 0.084***^^^ 0.136***^^^ Completed high school but not collegea -0.117***^^^ 0.024***^^^ 0.040***^^^ Limited English proficiency -0.126** 0.038 -0.022 Health barriers Activity or functional limitation -0.158*** 0.030* 0.258*** One chronic conditionb -0.017^ 0.014^ 0.037***^^^ Multiple chronic conditionsb -0.038**^ 0.025**^ 0.101***^^^ Ever diagnosed with a mental health condition -0.042** -0.007 0.054*** Search/participation barriers Lack of household internet access -0.103*** 0.101*** 0.015 Limited access to transportation -0.114*** 0.027 0.068*** Hiring barriers Criminal recordc -0.003^ 0.010^ 0.040 Formerly incarceratedc -0.079^ 0.084^ 0.009 Neighborhood-level barriers Unemployment rate between national average and twice the national averaged -0.011 -0.003 0.002 10 ADULTS TARGETED BY M EDICAID WORK REQUIREMENTS FACE EMPLOYMENT BARRIERS DEPENDENT VARIABLES Sometimes or Always Worked 20 or Worked Less than 20 Sometimes or Always More Hours Per Hours Per Week in Past Worked Less than 20 Week for All or Year for Reasons Related Hours Per Week in Nearly All Weeks to Labor Market or Nature Past Year for Health in Past Year of Employment Reasons Coefficient Coefficient Coefficient Unemployment rate over twice the national averaged -0.042 0.001 -0.011 Poverty rate of 10 percent to less than 20 percente 0.011 0.005 0.004 Poverty rate of 20 percent or moree -0.006 0.005 0.009 Middle quartiles of average commute timef -0.005 -0.009 0.002 Highest quartile of average commute timef -0.001 0.000 -0.004 Demographic characteristics Ages 35–49g 0.069***^^^ -0.030***^^^ 0.000^^^ Ages 50–64g 0.029*^^^ -0.050***^^^ -0.021**^^^ Female -0.167*** 0.036 0.013** Black, non-Hispanich 0.022^^^ 0.010^^^ 0.007 Hispanich -0.046***^^^ 0.031**^^^ 0.002 Other race or more than one race, non-Hispanich -0.070***^^^ -0.026**^^^ 0.010 Lives with a partneri 0.019 0.010^^^ 0.027**^^^ Singlei -0.006 0.043***^^^ 0.023***^^^ Potential exemptions Primary caregiver for a dependent childj -0.013 -0.021** -0.009* Primary caregiver for an adult family memberj -0.015 0.019 0.009 Full-time student in past yeark -0.309***^^^ 0.022^^ -0.036***^^^ Part-time student in past yeark -0.057*^^^ 0.074**^^ -0.005^^^ Pregnant in past year -0.108*** -0.026 0.052* Constant 0.809*** 0.084*** -0.029*** Sample size 13,089 13,089 13,089 Source: Health Reform Monitoring Survey, quarter 3 2018 and quarter 1 2019; American Community Survey five-year estimates, 2013–17. Notes: Mental health conditions include anxiety, depression, mood, bipolar, schizoaffective, and schizophrenic disorders. Limited access to transportation includes lack of a household vehicle and fair or poor ability to get around without driving. Criminal record includes those who were arrested in the last seven years or ever convicted or found delinquent of, or pleaded guilty to, any charges. Formerly incarcerated includes those ever sentenced to time in prison, jail, or a juvenile detention center. Neighborhood unemployment rates are for adults ages 20–64 and are based on census tract-level data for 2013–17. The national average nonelderly adult unemployment rate for 2013–17 for adults in the sample was 6.3 percent. Neighborhood poverty rates and commute times are for all ages and are based on census tract-level data for 2013–17. Adults with missing data on work experience or neighborhood characteristics are excluded from the sample. a–k ^/^^/^^^ Estimates for categories grouped by the same letter jointly differ from zero at the 0.10/0.05/0.01 level, using F-tests. */**/*** Estimate differs from zero at the 0.10/0.05/0.01 level, using two-tailed tests. ADULTS TARGETED BY M EDICAID WORK REQUIREMENTS FACE EMPLOYMEN T BARRIERS 11 What It Means Early experiences with work requirements have drawn attention to problems that Medicaid enrollees could face complying with reporting requirements as a key implementation challenge (Musumeci, Rudowitz, and Hall 2018).15 Yet beyond these logistical hurdles, many Medicaid enrollees who would potentially be subject to work requirements may struggle to comply because of barriers to employment. Over 80 percent of potentially nonexempt Medicaid enrollees do not consistently work the 20 hours per week typically required, primarily because of difficulty finding work, lack of control over work schedules, or health issues. Given the range of labor market and employment barriers facing Medicaid enrollees who could be subject to work requirements, substantial new investments would likely be needed to address challenges with the jobs that are available to enrollees and to provide targeted education and training and tailored work supports and employment services, such as job search and placement assistance, internet access, and transportation. However, CMS guidance prohibits use of Medicaid funding for these employment and work support services.16 Though the risk of losing Medicaid may lead some enrollees to increase their work effort, nonexempt enrollees who are unable to comply because of health challenges may experience worse employment outcomes because of loss of coverage and reduced health care access. High neighborhood unemployment rates and employer restrictions on work hours suggest that some adults could only comply through allowable nonwork activities or by supplementing their employment with these activities. Making work rules more flexible or establishing a statutory guarantee that work-related activities will be available and accessible could offer greater protection to adults with unsteady employment or variable schedules, and the latter would hold states accountable for ensuring the requirements can be fulfilled with reasonable effort. 17 These policy options could mitigate, but would not eliminate, the risk of coverage losses for those who have trouble complying with work requirements. States seeking to increase and support employment may want to consider workforce development strategies with a stronger evidence base than work requirements, such as postsecondary education, sectoral training programs, apprenticeships, subsidized jobs, and work supports for disadvantaged adults (US Department of Labor et al. 2014). About the Series This brief is part of a series drawing on the HRMS, a survey of the nonelderly population that explores the value of cutting-edge internet-based survey methods to monitor the Affordable Care Act before data from federal government surveys are available. Funding for the core HRMS is provided by the Robert Wood Johnson Foundation and the Urban Institute. For more information on the HRMS and for other briefs in this series, visit www.urban.org/hrms. 12 ADULTS TARGETED BY M EDICAID WORK REQUIREMENTS FACE EMPLOYMENT BARRIERS Notes 1 Maine has also received CMS approval to adopt work requirements but does not plan to implement the requirements at this time. See “Work Requirements Tracker,” Urban Institute, accessed May 13, 2019, https://www.urban.org/features/work-requirements-tracker. 2 Brian Neale, “RE: Opportunities to Promote Work and Community Engagement among Medicaid Beneficiaries,” the Centers for Medicare & Medicaid Services, January 11, 2018, https://www.medicaid.gov/federal-policy- guidance/downloads/smd18002.pdf. 3 Stewart v. Azar, Civil Action No. 18-152 (JEB) (D.D.C. Mar. 27, 2019). Gresham v. Azar, Civil Action No. 18-1900 (JEB) (D.D.C. Mar. 27, 2019). 4 Sara Rosenbaum and Alexander Somodevilla, “Inside the Latest Medicaid Work Experiment Decisions: Stewart v. Azar and Gresham v. Azar,” Health Affairs Blog, April 2, 2019, https://www.healthaffairs.org/do/10.1377/hblog20190402.282257/full/; Amy Goldstein, “Trump Administration Appeals Rulings That Blocked Medicaid Work Requirements,” Washington Post, April 10, 2019, https://www.washingtonpost.com/national/health-science/trump-administration-appeals-rulings-blocking- medicaid-work-requirements/2019/04/10/689024f6-5bb9-11e9-a00e- 050dc7b82693_story.html?utm_term=.a2ec9632117b. 5 Disability is based on reported Supplemental Security Income, Social Security Disability Income, or Medicare. CMS guidance requires states to exclude enrollees from work requirements if they qualify for Medicaid on the basis of disability. Recipients of Supplemental Security Income automatically qualify for Medicaid in most states, and the definition of disability used for this program is also used for determining eligibility for Social Security Disability Income, which can be used to obtain Medicare after a two-year waiting period (see “People with Disabilities,” Medicaid and CHIP Payment and Access Commission, accessed April 19, 2019, https://www.macpac.gov/subtopic/people-with-disabilities/). This study’s focus on nondisabled nonelderly adults excludes adults reporting Supplemental Security Income, Medicare, and/or Social Security Disability Income from the analysis. A limitation of this approach is that excluding adults with Social Security Disability Income or Medicare disproportionately excludes higher-income adults with longer work histories. Each round of the HRMS includes a sample of approximately 9,500 adults ages 18 to 64. The pooled September 2018 and March 2019 sample consists of 19,190 nonelderly adults, including 19,117 adults ages 19 to 64. Questions on work patterns and employment barriers were assigned to the full September 2018 sample and were randomly assigned to 75 percent of the March 2019 sample. 6 This analysis focuses on adults enrolled in Medicaidat the time of the survey but excludes uninsured adults, even though some of these people may be currently eligible for Medicaid or would be eligible if their states implemented the Affordable Care Act’s Medicaid expansion. Among current Medicaid enrollees who would be subject to work requirements under Kentucky’s waiver criteria, nearly 90 percent live in states that have expanded Medicaid, because eligibility in nonexpansion states typically depends on disability, pregnancy, or parental status. 7 Leighton Ku and Erin Brantley, “Medicaid Work Requirements: Who’s at Risk?” Health Affairs Blog, April 12, 2017, https://www.healthaffairs.org/do/10.1377/hblog20170412.059575/full/. 8 Research has focused on the implications of variation in county-level unemployment and poverty rates for work requirement policies (see Elaine Waxman and Nathan Joo, “Mississippi’s Work Requirements Don’t Account for a Varying Labor Market by Race and Geography,” Urban Wire (blog), Urban Institute, October 18, 2018, https://www.urban.org/urban-wire/mississippis-work-requirements-dont-account-varying-labor-market-race- and-geography). Though counties, metropolitan areas, or commuting zones may be more reflective of local labor markets, there are often pockets of high unemployment within these larger geographic areas that can be observed at the tract level. A sensitivity analysis using 2017 Local Area Unemployment Statistics data indicated that patterns of county-level unemployment were similar to patterns of tract-level unemployment for the three analytic groups in this study. 9 Affordable Care Act regulations define medical frailty under 42 CFR § 440.315. This definition includes people with disabling mental disorders, chronic substance abuse disorders, serious and complex medical conditions, and ADULTS TARGETED BY M EDICAID WORK REQUIREMENTS FACE EMPLOYMEN T BARRIERS 13 disabilities that impair ability to perform activities of daily living or qualify people for a federal or state disability determination; Deborah Yetter, “‘It’s a Mess’: Kentucky Medicaid Unclear on ‘Medically Fragile’ Meaning,” Courier Journal, February 25, 2019, https://www.courier-journal.com/story/news/2018/12/06/kentucky- medicaids-medically-fragile-meaning-unclear/2217346002/. 10 Measurement error may also result from excluding adults who are homeless, have low literacy, or do not speak English or Spanish from the HRMS sampling frame. Participants in the internet panel from which HRMS samples are drawn live in housing units and must be able to complete online surveys in English or Spanish. 11 Chronic conditions include those that have lasted or are expected to last a year or more (Hwang et al. 2001). 12 The Fair Credit Reporting Act prohibits commercial vendors that conduct background checks from releasing arrest records that are more than seven years old. See 15 U.S.C. § 1681. 13 An earlier validation study found significant underreporting of justice system involvement in surveys, but the findings from that study suggest that the self-administered survey mode used in the HRMS may reduce underreporting relative to interviewer-administered surveys (Preisendörfer and Wolter 2014). 14 Results were similar based on odds ratios from logistic regression models. 15 Jessica Greene. “Medicaid Recipients’ Early Experiences with the Arkansas Medicaid Work Requirement,” Health Affairs Blog, September 5, 2018, https://www.healthaffairs.org/do/10.1377/hblog20180904.979085/full/. 16 Brian Neale, “Re: Opportunities to Promote Work and Community Engagement among Medicaid Beneficiaries,” the Centers for Medicare & Medicaid Services. 17 For instance, some states receive additional SNAP Employment and Training funds if they commit to offering work activity slots to all able-bodied adults without dependents at risk of losing SNAP benefits because of time limits. References Bauer, Lauren, Diane Whitmore Schanzenbach, and Jay Shambaugh. 2018. Work Requirements and Safety Net Programs. Washington, DC: The Hamilton Project. Blavin, Fredric, Michael Karpman, and Stephen Zuckerman. 2016. “Understanding Characteristics of Likely Marketplace Enrollees and How They Choose Plans.” Health Affairs 35 (3): 535–9. Gangopadhyaya, Anuj, and Genevieve M. Kenney. 2018. “Updated: Who Could Be Affected by Kentucky’s Medicaid Work Requirement, and What Do We Know About Them?” Washington, DC: Urban Institute. Garfield, Rachel, Robin Rudowitz, and Anthony Damico. 2018. “Understanding the Intersection of Medicaid and Work.” Menlo Park, CA: Kaiser Family Foundation. Horvath, Francis W. 1982. “Forgotten Unemployment: Recall Bias in Retrospective Data.” Monthly Labor Review, March 1982: 40–4. Hwang, Wenke, Wendy Weller, Henry Ireys, and Gerard Anderson. 2001. “Out-of-Pocket Medical Spending for Care of Chronic Conditions.” Health Affairs 20 (6): 267–78. Musumeci, MaryBeth, Rachel Garfield, and Robin Rudowitz. 2018. “Medicaid Work Requirements: New Guidance, State Waiver Details and Key Issues.” Menlo Park, CA: Kaiser Family Foundation. Musumeci, MaryBeth, Robin Rudowitz, and Cornelia Hall. 2018. “An Early Look at Implementation of Medicaid Work Requirements in Arkansas.” Menlo Park, CA: Kaiser Family Foundation. Pascale, Joanne. 2008. “Measurement Error in Health Insurance Reporting.” Inquiry 45 (4): 422–37. Preisendörfer, Peter, and Felix Wolter. 2014. “Who Is Telling the Truth? A Validation Study on Determinants of Response Behavior in Surveys.” Public Opinion Quarterly 78 (1): 126–46. ADULTS TARGETED BY M EDICAID WORK REQUIREMENTS FACE EMPLOYMEN T BARRIERS 14 Quillian, Lincoln, Devah Pager, Ole Hexel, and Arnfinn H. Midtbøen. 2017. “Meta-Analysis of Field Experiments Shows No Change in Racial Discrimination in Hiring Over Time.” Proceedings of the National Academy of Sciences of the United States of America 114 (41): 10,870–5. US Department of Labor, US Department of Commerce, US Department of Education, and US Department of Health and Human Services. 2014. What Works in Job Training: A Synthesis of the Evidence. Washington, DC: US Department of Labor. Wen, Hefei, Brendan Saloner, and Janet R. Cummings. 2019. “Behavioral and Other Chronic Conditions among Adult Medicaid Enrollees: Implications for Work Requirements.” Health Affairs 38 (4): 660–7. ADULTS TARGETED BY M EDICAID WORK REQUIREMENTS FACE EMPLOYMEN T BARRIERS 15 About the Author Michael Karpman is a senior research associate in the Urban Institute’s Health Policy Center. Acknowledgments This brief was funded by the Robert Wood Johnson Foundation. The views expressed here do not necessarily reflect the views of the Foundation. The views expressed are those of the author and should not be attributed to the Urban Institute, its trustees, or its funders. Funders do not determine research findings or the insights and recommendations of Urban experts. Further information on the Urban Institute’s funding principles is available at urban.org/fundingprinciples. The author gratefully acknowledges the suggestions and assistance of Genevieve M. Kenney, Sharon K. Long, Anuj Gangopadhyaya, and Stephen Zuckerman, and research assistance from Dulce Gonzalez. ABOUT THE URBAN INST ITUTE The nonprofit Urban Institute is a leading research organization dedicated to developing evidence-based insights that improve people’s lives and strengthen communities. For 50 years, Urban has been the trusted source for rigorous analysis of complex social and economic issues; strategic advice to policymakers, philanthropists, and practitioners; and new, promising ideas that expand opportunities for all. Our work inspires effective decisions that advance fairness and 500 L’Enfant Plaza SW enhance the well-being of people and places. Washington, DC 20024 Copyright © May 2019. Urban Institute. Permission is granted for reproduction of www.urban.org this file, with attribution to the Urban Institute. 16 ADULTS TARGETED BY M EDICAID WORK REQUIREMENTS FACE EMPLOYMENT BARRIERS