Trump wants to make Medicaid recipients work to get benefits. That’s a very bad idea.

Seema Verma, the Trump-appointed administrator of the Centers for Medicare & Medicaid Services, has announced that the federal government will support state efforts to add work requirements to Medicaid, which provides health care and support services to poor and disabled Americans.

States will almost certainly take them up on this offer. Already, there is a state waiver request pending from Kentucky requesting permission to impose work requirements as a condition of Medicaid eligibility, and Arizona and Arkansas have expressed interest as well.

A Medicaid work requirement would be a huge departure from current practice, and a move that even many conservatives disagree with. It’s also likely to be ineffective, difficult to enforce, and maybe even illegal.

“I feel some sympathy for people who are saying, ‘This is my money, I’m working, I want to make sure other people who get support are working,’” says Gail Wilensky, who ran the Centers for Medicare and Medicaid Services in the George H.W. Bush administration. “But I’m not sure when you actually look where this has been tried, it’s had much effect.”

The majority of people benefiting from Medicaid are children, disabled, or elderly, and would be exempt from work requirements. If you exclude pregnant women and parents with young children, the number of affected people shrinks even more. The majority of the remaining non-disabled adults are working. And some of them can only work because they get Medicaid — such as people who have mental illnesses or struggle with substance abuse but who, with reliable health care, are healthy and stable enough to work. Making work a prerequisite for Medicaid could, perversely, wind up preventing such people from working.

“It would really harm people least able to hold and keep a job and hurt people who need health care to participate in the workforce,” Hannah Katch, a senior policy analyst at the Center on Budget and Policy Priorities and a former official in California's Medicaid program, says.

Work requirements aren’t appropriate for the vast majority of Medicaid beneficiaries, and the program actually encourages work

Center on Budget and Policy Priorities

Nearly two-thirds of Medicaid recipients are children, blind or otherwise disabled, or elderly. Work requirements would be plainly inappropriate in those cases, and so proposals typically exclude them. Kentucky’s proposed plan applies only to “‘able-bodied’ working age adults,” not including students and caregivers, per the Kaiser Family Foundation.

If you drill down into the third or so of Medicaid beneficiaries who are working-age adults and don’t get Supplemental Security Income for a disability, you find that the majority work, and an even larger majority live in households where someone works. Those who don’t have understandable reasons, as this chart from Kaiser Family Foundation president Drew Altman suggests — they’re sick, they’re in school, they’re retired, or they can’t find a job:

Axios / Drew Altman

“This is a solution in search of a problem,” Sara Rosenbaum, a professor of health policy at George Washington University who serves on a board advising Congress on Medicaid policy, says. “There’s just no evidence that too many people aren’t working who can work. If you say ‘able-bodied’ enough times, you give a sense that there are people just sitting around who could work, but that’s just not the case.”

For one thing, the definition of “able-bodied” under work requirements doesn’t always equal “able-bodied” the way most people would define. Typically, it means not getting Supplemental Security Income benefits through the government for having a disability.

The problem is that the SSI program doesn’t cover all or even most disabled people. Many people who fall outside its purview nonetheless have disabilities that can make working difficult. In an excellent piece on the likely effects of work requirements, the New York Times's Abby Goodnough interviewed a man named Jimmy Brunson, who has painful neuropathy in his feet due to diabetes. He relies on Medicaid for diabetes treatment and works when he can, but he told Goodnough, "Even though sometimes I can get a job, you’ve got to understand — sometimes I can’t even walk."

Katch notes that someone with diabetes typically wouldn’t qualify for SSI — but Brunson would still be hurt by a work requirement. “The fact that he can access treatment helps him work when he can do so,” she says. But if his disability meant that he could only put in, say, 16 hours a week, and the state is mandating 20 as part of its work requirement, he’d be out of luck. Worse, without Medicaid paying for his diabetes treatment, he might be unable to work at all.

The same issue arises for people with substance use or mental health issues, who may not be enrolled in disability programs but still need coverage for antidepressants or antipsychotics or mood stabilizers, or for medication-assisted addiction treatment like methadone or buprenorphine. Those treatments mitigate the symptoms of addiction or mental illness, and in doing so they help people continue working.

If you’re not working due to one of those conditions, getting Medicaid and gaining treatment could help you start working. But if a work requirement prevents you from signing up in the first place, you could be stuck.

“We all want to encourage people to work and to support them to work,” Joan Alker, a Medicaid expert who runs the Center for Children and Families at Georgetown, says. “If you take away their health care, people are less likely to be able to work, not more.”

The Medicaid expansion experience — which extended health insurance to millions of low-income adults — has produced evidence that Medicaid encourages, rather than deters, work. An evaluation of Michigan's expansion in the New England Journal of Medicine estimated that it expanded employment by more than 30,000 jobs. Ohio's evaluation of its expansion found that 74.8 percent of unemployed enrollees said getting covered made it easier to get and maintain employment.

Alker also notes that work requirements could wind up hurting children, even though they’re obviously exempt. Oftentimes children wind up enrolled in Medicaid after their parents are. If parents wind up not enrolling due to a work requirement, that could keep their kids out too.

And that’s just the direct effect. “A healthy parent is going to be a better parent,” Alker notes. “When a parent loses access to health services, that exposes the whole family to bankruptcy and economic insecurity. Because when one member is uninsured, they’re all at risk.”

Conservatives who support work requirements in other contexts don’t think it makes sense for Medicaid

Even some conservatives who support work requirements for other programs argue that they would be unenforceable, costly, and politically treacherous for Medicaid.

Robert Rector, a fellow at the Heritage Foundation and a longtime critic of programs for low-income people, opposes imposing work requirements on Medicaid. It just wouldn’t work, he concludes, because of the federal requirement that emergency rooms treat patients regardless of ability to pay. Rather than looking for a job, he argues, people could choose to go uninsured; when they eventually needed medical care, hospitals (including public hospitals funded via tax dollars just like Medicaid) would end up shouldering their emergency room bills.

And that’s not the only reason work requirements could wind up costing, rather than saving, money. States would have to build or improve systems to track whether recipients were working, Katch notes, which would entail more staff and other bureaucratic expenses.

Rector also expressed skepticism that work requirements for medical care would be politically viable. Making a work requirement effective means eventually penalizing people who don’t follow it. And, as he writes, while “it is politically challenging to restrict cash and food benefits to noncompliant recipients, denying medical care to sick, poor people is another problem entirely.”

Wilensky, the Medicare/Medicaid administrator under George H.W. Bush, shares Rector’s concerns, particularly on the emergency room issue. “Cutting off medical care from people who really need it doesn’t seem terrifically great idea,” she says. “They’ll just show up in the emergency room. … You’re going to get people who ignore or are forced to ignore their health problems until they get really sick, because that’s the law of the land. So then you have to decide, is this benefiting anyone?”

Work requirements are generally ineffective at fighting poverty

John Moore/Getty Images

As conservatives who worry about whether people who get government benefits are working, both Wilensky and Rector suggest work requirements for food stamps instead. Most people in poverty eligible for Medicaid also get food stamps, and the threat to cut off that aid might be more credible.

But the case for work requirements on any means-tested government aid program is pretty weak. The government has actually done numerous experiments to test out work requirements for cash welfare, as part of the state-level reforms leading up to the 1996 welfare reform law that imposed work requirements across the board. And the effects were meager at best, according to the Center on Budget and Policy Priorities’ LaDonna Pavetti, who reviewed this literature last year.

She found the premise of work requirements — the presumed existence of a large number of people on public assistance who should be working and aren’t, due to laziness or inertia or whatever — just isn’t true.

In most studies, work requirements led to an increase in the share of recipients working in the first year or two. But only three experiments out of 13 still found significant positive results after five years, and two experiments found that work requirements reduced one’s odds of working five years later. In the long term, the policy didn’t seem to promote work much at all.

Even when work requirements had the intended impact, the effects were modest. The program in Riverside, California, the most effective of the ones Pavetti analyzed, raised the share of welfare recipients working by 5 percentage points by year five. That's good, but it was ultimately an increase from 39.9 percent to 44.9 percent. Huge numbers were left not working and were also no longer eligible for benefits. Worse, subsequent research suggested that the Riverside program’s success at promoting work was a fluke, the product of an already stronger local economy.

In only two out of 13 experiments, in Atlanta and Portland, did work requirements significantly cut poverty, and even then, the effect size was small. In Portland, 83.4 percent of those not subject to work requirements were in poverty, but so were 79.4 percent of those subject to them.

The evidence, Pavetti concluded, was clear: In most cases, work requirements didn’t cause more people to work. When they did, the effect was small and didn’t typically lead to much poverty reduction.

“Too many disadvantaged individuals want to work but can’t find jobs for reasons that work requirements don’t solve,” Pavetti writes. “They lack the skills or work experience that employers want, they lack child care assistance, they lack the social connections that would help them identify job openings and get hired, or they have criminal records or have other personal challenges that keep employers from hiring them.”

All of that suggests that extending this approach to Medicaid would be a mistake.

And for Medicaid, work requirements might actually be illegal

Maury Phillips/Getty Images for Leigh Steinberg

But the Trump administration’s approach could be worse than a mistake — it could actually be illegal.

Work requirements would, barring an act of Congress, have to be authorized through what is called a Section 1115 waiver. The name refers to a section of the Medicaid law that allows the federal government to waive some legal requirements so that states can experiment with new policy approaches. The idea is that states get some flexibility in exchange for trying new ways of solving the problem that could, eventually, be adopted nationwide.

But that only makes sense if the approach being tested is actually new. Work requirements aren’t: The government experimented with them extensively during welfare reform. That led some Medicaid experts to express concern that waivers enabling work requirements wouldn’t just be ill-advised — without further action from Congress, they could be illegal.

“It’s a research and demonstration statute; it’s not carte blanche to waive whatever the secretary feels like waiving,” Rosenbaum says. “It’s not just a free pass to alter the program requirements for Medicaid.”

That’s what the Obama administration thought, at least. When states attempted to impose work requirements while President Obama was in office, his Department of Health and Human Services would deny them, arguing that the request was not justified. For example, they told Arizona that a work requirement (and other rejected provisions) would “undermine access to care and do not support the objectives of the program.”

“I don’t think there’s legal authority to grant a waiver, and the Obama administration essentially said as much,” Katch notes. If Trump’s administration attempts to approve provisions like that, litigants will take to court to make the same arguments.

For Republicans who want Medicaid work requirements, the safer approach, legally speaking, is for Congress to pass a new health care bill allowing states to impose them. But that isn’t likely to happen anytime soon.

Will you join us?

Our biggest supporters are our readers — and we’re so grateful to everyone who has made a contribution during our September campaign. We’re less than 1,000 contributions away from reaching our goal for the month, which in turn will allow us to say yes more often when our incredible journalists come to us with questions they want to answer and projects they want to pursue. Will you make a contribution before the month ends and support our policy coverage through 2024 and beyond?

One-Time Monthly Annual

$5/month

$10/month

$25/month

$50/month

Other

$

Yes, I'll give $5/month

Yes, I'll give $5/month

We accept credit card, Apple Pay, and Google Pay. You can also contribute via