Feature|Articles|August 6, 2026

Medicaid work requirements and how states may apply them: An in-depth discussion with Eric Levine, M.P.H., of Avalere Health

Listen
0:00 / 0:00

Key Takeaways

  • National Medicaid expansion work requirements start in 2027 and hinge on 80 hours of qualifying activity monthly, reframed as “community engagement,” with broad implications for ~20 million beneficiaries.
  • Medical frailty exemption is functionally tightened by requiring ADL impairment in addition to a diagnosis, potentially capturing neurodevelopmental and chronic conditions that do not limit basic self-care.
SHOW MORE

Levine, a principal at Avalere Health, explains some of the details of Medicaid work requirements, including how activities of daily living may factor in and some of the leeway states have in how they are applied.

On June 1, CMS released the interim final rule on the much-talked about new national-level Medicaid work requirements. The requirements — which CMS is calling “community engagement requirements” — go into effect in 2027. They apply to people who became eligible for Medicaid through the Affordable Care Act state-by-state Medicaid expansions, a group that includes approximately 20 million people. To stay eligible for coverage, people in the expansion population will need to show 80 hours of employment, community service or participation in a training. Part-time enrollment in school also will satisfy the requirement.

But there are exemptions, including one for people who are “medically frail.” Eric Levine, M.P.H., principal at Avalere Health, recently wrote a piece analyzing the medical frailty exemption. As Levine explained, the rule includes a provision that narrows the frailty exemption. It is not enough to have a qualifying medical condition. The person must also be impaired in at least one activity of daily living (ADL), which includes mobility, bathing, dressing and other basic activities.

Peter Wehrwein, managing editor of Managed Healthcare Executive, recently spoke to Levine about his analysis of the work requirement interim final rule and the medical frailty exemption.

This transcript has been edited for clarity and length.

The interim final rule that came out on June 1. I'm a little bit confused. Interim final — that seems like almost an oxymoron.

The interim final rule invites public comments. It’s like, “Here's the rule. We invite comments on it,” and then I guess there's an implication maybe they will make changes, but here's the rule. Whereas when you have a proposed rule through the notice of proposed rulemaking, that is sort of a draft. We probably wouldn't see change barring any lawsuits.

So, this interim final rule sets rules for Medicaid work requirements, but there were other changes to Medicaid in the One Big Beautiful Bill, such as more frequent redeterminations.

This one is specifically around the work requirements, which they're calling “community engagement requirements” because they can be fulfilled by things other than work, but they’ve been known as work requirements.

I don't know if there's going to be a separate rule. I think CMS laid the groundwork of requiring redeterminations, and it's more up to the states to figure out, do they meet the minimum? Do they add more frequent checks?

Same with this rule, up to a point. This is the final rule from the federal level, but we need to see how individual states will implement it because there's some flexibility for them.

Where are those flexibilities? Medicaid, because it's state-run, tends to have a fair amount of state-to-state variety. What's the latitude here for state differences with respect to work requirements

I'll preface this all by saying I'm not a lawyer, so take everything I say with a grain of salt. But my understanding is one area is in the exclusions because of medical frailty. CMS made it quite broad and said, “We're not going to be defining specific conditions.” They gave examples, but then they said we're not defining a set condition list. But states can, and likely, will say, “When we say a serious medical condition, here are the 40 conditions that we're talking about.” So that would be one area of state flexibility — how specific or broad are those lists, and I feel like that can vary depending on a state’s philosophy about whether they want to keep this narrow or broad.

But the real liberty for the states, in my mind, is how they run the verification process. How frequently, and then what data sources do they use? How do they collect the data? That's probably where the most variation would be.

If a state has a health information exchange, it will use that as a data source. Some states may create databases for plans or providers to submit data. Maybe they'll create certain data feeds to make it easier for the states to collect this information. They’ll be getting Medicaid encounters, but that's only one piece of the puzzle, especially when you think about the activity of daily living requirement — that’s not an encounter. The question remains how are they going to get proof of the ADL [impairments]?

That brings us to the piece you posted on the Avalere Health website. The piece was nice and concise! Your main point seemed to be that people will not be exempt from the work requirements just by virtue of having a medical condition. The condition has to affect at least one ADL; for example, getting dressed is one of them. That was a major fork in the road, the fact that the federal government says it's not sufficient that you have a condition. The medical condition has to affect activities of daily living for a person to be exempt.

I think that's where the surprise was, in the narrowing of it. So, for example, if you're an individual with an intellectual developmental disability, it's not enough to just have a diagnosis of, let's say, autism, because under this rule, if you have, if you're on the autism spectrum that doesn't affect your ability to, say, bathe or shower or dress or eat yourself, you would then be subject to the work requirements.

The intent is not for the burden to be on the member. This is up to the states to verify. But unfortunately, if the state can't get the data, they're going to go to the members at some point potentially.

There are certain conditions they lay out, such as blindness, there’s no ADL requirement and for substance use disorder, unless you've been in stable recovery of five or more years.

But then there's this like whole swath of gray area. A state could potentially say all COPD [chronic obstructive pulmonary disease] is exempt.

When you dig into the rule itself, ADL [impairment requirements] are attached to the line around individuals with a physical, intellectual or developmental disability, and then a separate line for individuals with a serious or complex medical condition. There's some overlap of what is a disability and what is a medical condition, and I think that's the particular area where we may see state leeway.

I think some states that want an expansive Medicaid program will probably try to comb through the language of the bill to find where canweuse the language of the bill to allow us to have broad lists, so that more people would be exempt from the requirements.

You’ll start to see varying condition lists. Patient A and patient B. One lives in one state, the other in another. Same condition, same acuity. One could be exempt, and one can be nonexempt.

There are states already filed suit against this. I would imagine as the requirements get implemented, you will see lawsuits from patient advocacy organizations, maybe provider groups, maybe plan groups.

In your piece you suggest that because of the layering on ADL impairment to be eligible for exemption, there’s a new role for life science companies, for drug companies, for health plans, and for providers. Could you talk about that?

My working assumption or philosophy is that assuming work requirements go through, I think there's a role that each stakeholder can play, within regulatory boundaries, to make sure the number of people who lose their Medicaid is zero. It's sort of lofty. It’s a little bit provocative.

The CBO [Congressional Budget Office] estimated that 15% of the Medicaid expansion population people would lose enrollment as a result of work requirements, with 6% to 7% losing coverage because of procedural reasons. They didn't follow the process; they didn’t get a notification; they didn't submit their documents for exemption; or they didn't follow the right procedures to prove that they are working. And then the other half of that 15% is just noncompliance. They have to do work requirements, and they decide not to work.

There's no reason why anyone, in theory, should lose coverage. I think that there are a few things that each stakeholder can do. I think that there’s work around the exemptions population. Providers can make sure that they are documenting robustly a member's condition. Maybe they start including ADL assessments or questions in their evaluations of patients and documenting severity or making sure that they’re using the right codes matching the severity of the condition.

Health plans can be making sure they're getting the ADL assessments for their members who are at risk, so they have that documentation. They can also make sure members know about these requirements coming up — not to intervene in the process itself, but for education, because in past iterations of work requirements, a lot of people didn't know this was coming or didn't understand what it meant. This rule specifically says MCOs [managed care organizations] cannot do the verification.

A pharma company can do above-brand communications campaigns and provider education around these, and so I think that's where they can play the role.

For those who would be subject to work requirements, then it's about the education and the resources so that those people have every opportunity to fulfill those requirements. Whether it's investing and supporting and or hosting job trainings that meet the requirements, connecting with community-based organizations that connect members with volunteer opportunities that would meet the requirements, transition-to-work programs, things like that. I think that it's an opportunity for all stakeholders to invest around these activities to try to make [the disenrollment] zero.

There may be reasons why someone can't comply with these, and I think that's a whole separate discussion. But, in a general sense, we can support members wherever they fall on the spectrum so that no one loses coverage because they didn't know about this requirement to begin with, or didn't understand how it applied to them, or the documents weren't there—some of the things that I feel are avoidable.

When you talk about job fairs and referrals to community organizations, do you see drugmakers doing that or MCOs?

MCOs can do that, and they already have. They also often connect members with those resources in the community as part of their Medicaid offerings.

But I personally see a world in which life science companies, as part of corporate giving, maybe give grants to community-based organizations. In a market they serve, they can offer volunteer days or courses.

I see funding opportunities for life science companies as part of their corporate giving programs. Providers and hospitals are in the communities, so whether it's hosting the social workers in the hospitals to be reaching out to these members or booths in hospitals — things like that. I think there's something everyone can do to help people meet the requirements.

An MCO cannot, however, set up a volunteer program for its members in order for them to meet the requirement, correct?

An MCO cannot be responsible or be delegated to do the verification. So, a state can't say to the MCO, "You tell us whether the members are frail.” But my interpretation of the rule would be that an MCO can absolutely develop its own volunteer program within its community. I think that would probably need to be open to anyone in the community, so they couldn't just like wall it off to their members, but they could do a volunteer program.

Is there a danger of conflict of interest and fraud and abuse in that? Because it would seem then that the MCO might set up a volunteer program, and it could be a shell operation. It just seems like it might invite abuse.

They would definitely need to get their legal counsel involved in planning it. I know, for example, in Medicare Advantage, whenever plans are doing those types of community activities, they need to be very careful that they're not sales activities. There are a lot of regulations, but I think it's achievable.


Latest CME