You are here

States’ Experiences Confirming Harmful Effects of Medicaid Work Requirements

October 22, 2019

Arkansas’ and New Hampshire’s Medicaid waivers provide direct evidence of what happens when states take away Medicaid coverage for not meeting work requirements. Unfortunately, the data corroborate the most pessimistic projections.

  • In Arkansas, more than 18,000 people — nearly 1 in 4 of those subject to work requirements — lost coverage over the course of just seven months. In New Hampshire, almost 17,000 people, or about 40 percent of those who would have been subject to work requirements, would have lost coverage had state policymakers not put the policy on hold.
  • In both states, evidence suggests that people who were working and people with serious health needs who should have been eligible for exemptions lost coverage due to red tape. Large numbers of beneficiaries in both states reported that they didn’t know about the work requirement or whether it applied to them.
  • Arkansas’ work requirement increased uninsured rates without increasing employment, a study by Harvard researchers found. Consistent with that study, data show that only a small minority of those who lost coverage in Arkansas found jobs, and most of them would likely have found jobs with or without the work requirement policy.

The Arkansas and New Hampshire Work Requirement Policies

In January 2018 the Trump Administration began approving state work requirements in Medicaid, which the federal government had never previously permitted. Between June 2018 and March 2019, Arkansas required some Medicaid expansion enrollees to document that they worked or engaged in work-related activities (e.g., job training or volunteer work) for at least 80 hours per month, unless they reported that they qualified for limited exemptions. In June 2019, New Hampshire began implementing its own work requirement, requiring all expansion enrollees to work or engage in work-related activities for 100 hours each month or obtain an exemption.

In March 2019, a federal district court vacated the Department of Health and Human Services’ (HHS) approval of Arkansas’ waiver (and a similar waiver in Kentucky, which had not yet been implemented), thereby preventing Arkansas from continuing to implement work requirements. New Hampshire voluntarily suspended its work requirement in July 2019 and, a few weeks later, the same federal district court vacated HHS’ approval of its waiver. The Trump Administration is appealing these decisions and has continued to approve similar waivers in other states.

Arkansas is the only state to have taken coverage away from people for not meeting work requirements, but New Hampshire did not pause its requirement until it was on the brink of doing so. Thus, both states’ experiences provide important information about who loses coverage and why as a result of these policies.[1]

Before any state implemented a work requirement, we and others predicted that these policies would lead to large coverage losses — including among beneficiaries who are eligible but get tripped up by red tape — and would not meaningfully increase employment.[2] Since work requirements in Medicaid were unprecedented, we based our predictions on evidence from other types of eligibility restrictions and from work requirements in other federal programs. Here’s how those predictions stack up against the data from Arkansas and New Hampshire.

Data Show Work Requirements Lead to Large Coverage Losses

In Arkansas, over 18,000 beneficiaries, or about a quarter of those subject to the work requirement, lost coverage over the first seven months of implementation.[3] In New Hampshire, almost 17,000 beneficiaries, or about 40 percent of those subject to the work requirement, were set to lose Medicaid before the state suspended the requirement.[4] These coverage losses are even higher than the 6 to 17 percent coverage loss that Kaiser Family Foundation researchers forecasted could result from implementing work requirements nationwide and the 15 percent coverage loss that Kentucky projected would ultimately result from its waiver.[5]

If Arkansas had continued implementing its policy, the coverage losses would likely have grown. When the court halted the state’s waiver, Arkansas was in the process of extending the policy to beneficiaries aged 19-29 and those with incomes between 100 and 138 percent of the poverty line. In 2018, the policy applied only to beneficiaries aged 30-49 and to those below the poverty line.

Arkansas Uninsured Rate Rose Sharply Among Those Subject to Policy

A study by Harvard researchers finds that the uninsured rate among low-income Arkansans aged 30-49 rose from 10.5 percent in 2016 to 14.5 percent in 2018, after the work requirement took effect. There was no similar increase for low-income Arkansans of other ages or for low-income people aged 30-49 in other, similar states.[6] This finding refutes claims, for example from HHS Secretary Alex Azar, that most people leaving Medicaid due to the policy did so because they found jobs with health insurance.[7]

Focus groups with Arkansas beneficiaries have also found that people losing Medicaid lost health coverage — and with it, access to needed health care. Among focus groups of participants who lost Medicaid, the majority did not learn that their coverage had been terminated until they sought care or tried to fill a prescription. “I went to pick up my prescription and they said I couldn’t get it,” one beneficiary recounted. “It was a big shock. At first, I was upset and then I was worried because I need my medicine. I shouldn’t find out the day I need my medicine.” Another said, “I have to have my medication because I am epileptic, and I take three different seizure medicines. If I don’t take it, I can’t work. . . . I had to be stable and if I was not stable I was going to lose my job.”[8]

Number Losing Coverage Exceeded Policy’s Supposed Target Population: Beneficiaries Not Working or Eligible for Exemptions

Around 3 or 4 percent of those subject to the Arkansas work requirement were not working and did not qualify for exemptions, studies estimate.[9] Yet each month, 8 to 29 percent of those subject to the requirement failed to report sufficient work hours, many of them not reporting any hours. And over 75 percent of those required to report hours (that is, those not automatically exempted by the state) failed to do so each month.[10] Likewise, a study estimates that all but a small minority of Medicaid expansion beneficiaries in New Hampshire are either working or ill or disabled (and therefore should qualify for exemptions), yet 40 percent of those subject to the work requirement were set to lose coverage had the state not put the policy on hold.[11]

News accounts corroborate that eligible beneficiaries in Arkansas lost coverage and were at risk of losing coverage in New Hampshire. For example, one working Arkansas beneficiary with a chronic condition explained that he lost Medicaid and then could not afford medications, which in turn caused him to lose his job due to his deteriorating health. Another reported rationing her medication after failing to navigate the reporting requirement and losing coverage, despite working 25 to 35 hours each week — which equates to well over the state’s monthly minimum requirement. [12] And a New Hampshire woman described her struggle to obtain a “medical frailty” exemption, which failed because her primary care doctor and neurosurgeon each insisted that the other should fill out the necessary paperwork.[13]

Beneficiaries Faced Many Challenges Complying

Evidence from Arkansas and New Hampshire confirms earlier research on work requirements showing that red tape and paperwork requirements create serious hurdles for eligible beneficiaries, which then causes enrollment to decline.[14] Examples include:

  • Complex and confusing rules. Nearly half the population subject to Arkansas’ work requirement reported that they were unsure whether it applied to them, while another third said they had heard nothing about it, the Harvard researchers’ study found.[15] In New Hampshire, many beneficiaries reportedly didn’t know about the work requirement or received confusing and often contradictory notices about whether they were subject to it.[16]

    Some Arkansas beneficiaries apparently believed, incorrectly, they could maintain their coverage by reporting work hours just once, rather than every month. Also, some beneficiaries reported over 80 hours of job search each month, twice the number they are allowed to count toward the work requirement; they likely thought they were complying as they diligently looked for work and reported their hours but were not actually complying with the rigid policy.

  • Insufficient and ineffective outreach. Arkansas conducted much of its outreach through social media and online videos, which failed to reach much of the population without access to computers or the Internet. The state (along with health plans, providers, and advocates) also conducted other outreach efforts to educate beneficiaries about the new work requirement, but these efforts failed to reach many enrollees, often due to inaccurate phone numbers in case files.[17]

    New Hampshire attempted a more comprehensive outreach effort, undertaking multiple activities such as mail notifications, town halls, phone calls, and text messages, at significant expense. But officials acknowledged that “these considerable efforts” were ineffective.[18] For example, on only 500 out of the state’s 50,000 phone calls did a state official discuss the work requirement with the person who would be affected by it, and the state’s 2,011 home visits reached only 270 people who would be affected. These unsuccessful efforts cost the state $187,000.[19]

  • Complex reporting systems. Many rightly criticized Arkansas’ decision to initially restrict reporting to an online portal. More than 20 percent of Arkansas beneficiaries lack Internet access and another 20 percent have no broadband access.[20] In addition, Arkansas’ web portal was unavailable after 9 p.m. and before 7 a.m. each day, was not mobile friendly, was not accessible for those with certain disabilities, and required a complex login procedure. Participants in an Urban Institute-run focus group said that setting up an account proved extremely difficult.[21]

    While Arkansas later added a phone reporting option, that did not eliminate the hurdles for eligible beneficiaries.[22] Some, for example, reported waiting on hold for 45 minutes to an hour, after which they were sometimes hung up on.[23]

    New Hampshire beneficiaries also experienced online reporting problems. Just days before the reporting deadline, the state’s online system experienced a glitch that created problems for people trying to report their work hours.[24]

  • Lack of staff support. Neither state hired additional staff to answer questions, and, at least in Arkansas, there were no accommodations for individuals with disabilities. Arkansas Human Services Director Cindy Gillespie justified the lack of additional staff by saying, “If you implement [work requirements] in the old-fashioned way of ‘Come into our county office,’ we would have to hire so many people — and that just doesn’t make sense.”[25] But Arkansas’ failure to invest in staff and other resources to support enrollees who should be exempt or need help complying with the requirements created a bureaucratic maze that caused many eligible enrollees to lose coverage.

Work Requirements Endanger People With Disabilities

When the Centers for Medicare & Medicaid Services (CMS), in announcing the Administration’s support for Medicaid work requirements, notified states that the policies must comply with the Americans with Disabilities Act, we and others warned that protecting people with disabilities would prove impossible.[26] Indeed, Arkansas didn’t adequately explain beneficiaries’ rights under the Act, and it lacked a comprehensive system for providing reasonable modifications to protect people with disabilities, such as modifying the hourly requirement or providing support to help people meet the reporting requirement.[27] Due to the lack of protections and the design of the work requirement itself, individuals with disabilities lost coverage and may face serious harm as a result. In fact, “people with disabilities were particularly vulnerable to losing coverage under the Arkansas work and reporting requirements, despite remaining eligible,” a Kaiser Family Foundation study concluded.[28]

Data Don’t Show Employment Gains Due to Work Requirements

Nearly all of the beneficiaries who met Arkansas’ requirement by working were automatically deemed compliant because they were already working before the rules took effect or because they complied with work requirements already in place under SNAP (formerly food stamps). Only the group that had to report hours each month faced any new work incentive due to the policy. And of that group, only a few hundred each month met the requirement by reporting sufficient work hours, the state reports.[29] Moreover, these data don’t establish that any of those individuals found a job because of the work requirement. Low-income people move in and out of jobs frequently under any circumstances, and the state offered no evidence that more beneficiaries found jobs under the work requirement than in prior years.

Arkansas released data in March 2019 showing that of the 18,164 beneficiaries losing coverage in 2018 for not complying with the work requirement, just 1,981 had matches in the state’s New Hire Database, indicating they found new work. There is no evidence that any of the roughly 16,000 others losing coverage found new work. In addition, these data almost certainly overstate the number of those losing Medicaid who found steady employment, because of limitations of the database. For example, the New Hire Database includes individuals who were employed for only hours or days — it doesn’t show whether the new jobs are full or part time or if they’re seasonal or permanent, or if the new employee was previously unemployed or just changed jobs.[30]

Meanwhile, the Harvard researchers’ study cited above found no significant increases in employment, number of hours worked, or overall rates of community engagement activities (such as volunteering) among those subject to Arkansas’ work requirement.[31]

These data are consistent with focus group interviews showing that the Arkansas work requirement didn’t change most beneficiaries’ behavior. Beneficiaries already had enough reasons to work: they needed to pay their bills. But they often struggle with unstable work hours, live in rural areas with few jobs, or face other barriers to employment — and the state didn’t invest any new money in job training programs, services to address barriers, or supports like transportation to help beneficiaries connect to jobs.[32] The Arkansas data are also consistent with evidence that work requirements in other federal programs have had limited effects on employment.[33]

Few Beneficiaries in Arkansas Who Lost Coverage Re-Enrolled This Year

Individuals in Arkansas who lost coverage last year could have re-enrolled effective January 1, but only about 4,300 of the more than 18,000 who lost coverage had done so as of mid-May, according to the state.[34] Most losing coverage (many of whom didn’t even know they were subject to the work requirement) are likely unaware that they are again eligible or may struggle to navigate the application and verification process. Or, fearing another loss of coverage for non-compliance with the requirement, some residents may be waiting to enroll until they need coverage more.

Arkansas’ and New Hampshire’s experiences should serve as a warning to other states about the human toll of taking coverage away from people who can’t meet rigid work requirements. No matter how they’re implemented, work requirements have harmful and unintended consequences — most notably, taking coverage away from people who are already working or should be exempt. These fundamentally flawed policies can’t be fixed.[35]

End Notes

[1] Indiana is the only other state to have begun implementing work requirements, but it is not scheduled to take coverage away from those failing to meet them until January 1, 2020. Kentucky’s waiver was blocked by a court prior to implementation, and other states with approved waivers are not planning to implement them until January 2020 or later.

[2] Jennifer Wagner and Judith Solomon, “States’ Complex Medicaid Waivers Will Create Costly Bureaucracy and Harm Eligible Beneficiaries,” Center on Budget and Policy Priorities, May 23, 2018,; Hannah Katch, Jennifer Wagner, and Aviva Aron-Dine, “Taking Medicaid Coverage Away From People Not Meeting Work Requirements Will Reduce Low-Income Families’ Access to Care and Worsen Health Outcomes,” Center on Budget and Policy Priorities, updated August 13, 2018,

[3] For data availability reasons, we calculate the coverage loss percentage as the total cases terminated as of January 1, 2019 (18,164) as a share of the cases subject to the work requirement in September (after all four groups were phased in) plus the number of cases terminated effective September 1. This calculation approximates the reduction in January enrollment, compared to a counterfactual in which the work requirement had not been implemented. (Both the numerator and denominator likely include some people who would have left the program by January even absent the work requirement.) Arkansas Department of Human Services, “Arkansas Works Program: December 2018,” January 15, 2019,

[4] New Hampshire Department of Health and Human Services, “DHHS Community Engagement Report: June 2019,” July 8, 2019,

[5] Kentucky HEALTH §1115 Demonstration Modification Request, Commonwealth of Kentucky, July 3, 2017,; Rachel Garfield, Robin Rudowitz, and MaryBeth Musumeci, “Implications of a Medicaid Work Requirement: National Estimates of Potential Coverage Losses,” Kaiser Family Foundation, June 27, 2018,

[6] The other states included in the study are Kentucky, Louisiana, and Texas. Benjamin Sommers et al., “Medicaid Work Requirements — Results from the First Year in Arkansas,” New England Journal of Medicine, June 2019,

[7] Dan Diamond, “Alex Azar’s confusing claim on Medicaid work requirements,” Politico, March 15, 2019,

[8] Ian Hill and Emily Burroughs, “Lessons from Launching Medicaid Work Requirements in Arkansas,” Urban Institute, October 2019,

[9] Anuj Gangopadhyaya et al., “Medicaid Work Requirements in Arkansas,” Urban Institute, May 24, 2018,; see also Sommers, op cit.

[10] Arkansas Department of Human Services, “ARWorks Reports,”

[11] Rachel Garfield et al., “Understanding the Intersection of Medicaid and Work :What Does the Data Say?” Kaiser Family Foundation, August 8, 2019,

[12] “With new work requirement, thousands lose Medicaid coverage in Arkansas,” PBS NewsHour, November 19, 2018,; Benjamin Hardy, “Locked out of Medicaid — Arkansas’s work requirement strips insurance from thousands of working people,” Arkansas Times, November 19, 2018,

[13] Caitlin Andrews and Nathan DeWitt, “For ‘medically frail,’ judge’s decision on Medicaid expansion work requirement comes as a relief,” Concord Monitor, August 3, 2019,

[14] Margo Sanger-Katz, “Hate Paperwork? Medicaid Recipients Will Be Drowning in It,” New York Times, January 18, 2018,

[15] Sommers, op cit.

[16] Dave Solomon, “Spreading the word on Medicaid work requirement proves challenging,” Union Leader, June 16, 2019,; Jason Moon, “Confusing Letters, Frustrated Members: N.H.’s Medicaid Work Requirement Takes Effect,” New Hampshire Public Radio, June 18, 2019,

[17] MaryBeth Musumeci, Robin Rudowitz, and Cornelia Hall, “An Early Look at Implementation of Medicaid Work Requirements in Arkansas,” Kaiser Family Foundation, October 2018,

[18] New Hampshire Department of Health and Human Services, “Letter to Governor Sununu, President Soucy, and Speaker Shurtleff,” July 8, 2019,

[19] Ethan DeWitt, “Now-disbanded Medicaid work requirement costs New Hampshire $187,000,” Concord Monitor, September 23, 2019,$187-000-28745556.

[20] Anuj Gangopadhyaya et al., “Under Medicaid work requirements, limited internet access in Arkansas may put coverage at risk,” Urban Institute, October 29, 2018, Data are for non-elderly, non-disabled Medicaid recipients aged 19-49, the population subject to the work requirement.

[21] Hill and Burroughs, op cit.

[22] Arkansas Department of Human Services, “DHS Expanding Phone Reporting, Outreach for Arkansas Works Enrollees,” December 12, 2018,

[23] Hill and Burroughs, op cit.

[24] Jason Moon, “As Deadline Looms, Some Struggle to Submit Work Hours Online for New Medicaid Requirement,” New Hampshire Public Radio, July 3, 2019,

[25] Benjamin Hardy, “Medicaid advocate criticizes Arkansas Works’ email-only reporting for work requirements,” Arkansas Times, April 28, 2018,

[26] “Taking Away Medicaid for Not Meeting Work Requirements Harms People with Disabilities,” Center on Budget and Policy Priorities, updated December 10, 2018,

[27] Anna Bailey and Judith Solomon, “Medicaid Work Requirements Don’t Protect People With Disabilities,” Center on Budget and Policy Priorities, November 14, 2018,

[28] MaryBeth Musumeci, “Disability and Technical Issues Were Key Barriers to Meeting Arkansas’ Medicaid Work and Reporting Requirements in 2018,” Kaiser Family Foundation, June 11, 2019,

[29] Arkansas Department of Human Services, ARWorks Reports, op cit.

[30] Jennifer Wagner, “New Arkansas Data Contradict Claims That Most Who Lost Medicaid Found Jobs,” Center on Budget and Policy Priorities, March 19, 2019,

[31] Sommers, op cit.

[32] MaryBeth Musumeci, Robin Rudowitz, and Barbara Lyons, “Medicaid Work Requirements in Arkansas: Experience and Perspectives of Enrollees,” Kaiser Family Foundation, December 18, 2018,

[33] LaDonna Pavetti, “TANF Studies Show Work Requirement Proposals for Other Programs Would Harm Millions, Do Little to Increase Work,” Center on Budget and Policy Priorities, November 13, 2018,

[34] Harris Meyer, “More Arkansans uninsured, unemployed post-Medicaid work requirement,” Modern Healthcare, June 19, 2019,

[35] Judith Solomon, “Medicaid Work Requirements Can’t Be Fixed,” Center on Budget and Policy Priorities, January 10, 2019,