Medicaid's new work requirement comes with an exemption that sounds humane and straightforward: people who are "medically frail," too sick to work, do not have to meet it. The trouble is in how a person proves they qualify. Under the rule the Centers for Medicare and Medicaid Services issued in June, having a serious illness is no longer enough. Beneficiaries now have to demonstrate that their sickness is severe enough that it significantly impairs their ability to perform the required work activities, with the functional impairment well-documented. Across more than two dozen states, chronically ill people, legal advocates, and state agencies are now scrambling to figure out which conditions qualify and how to prove disease severity, against a January 2027 deadline and amid a court fight over the rule itself.

Underneath the confusion is a design problem worth stating plainly, because it explains why this exemption is likely to fail many of the people it was written to protect. The exemption asks the seriously ill to prove functional incapacity by completing a functionally demanding task. Assembling medical records, obtaining provider documentation, interpreting an unclear standard, and meeting deadlines is exactly the kind of sustained administrative effort that serious illness makes hardest. The people most entitled to the exemption on the merits are, by the nature of their condition, among those least able to navigate the process required to claim it.

From a category you have to a case you must make

The change that created this problem is a shift in kind, not degree, and it is worth being precise about it. Congress, in writing the law, identified categories of conditions that make a person medically frail, a largely categorical approach: you have a qualifying condition, you are exempt. CMS's interim final rule added a second gate on top. It is not enough to have a qualifying condition; the individual must also show that the condition impairs their capacity to work the required hours. The rule adopts a restrictive definition that ties frailty to functional work capacity in a way many states did not expect.

That converts a categorical protection into an individualized burden of proof. A categorical exemption is something a system can often confirm about you: your diagnosis is in the records, the box is checked. A functional-impairment exemption is something you have to build a documented case for: not just that you have multiple sclerosis or severe heart failure or a serious mental illness, but that it functionally prevents you from working twenty hours a week, evidenced to a standard the rule leaves unclear. The difference sounds technical and is in fact the whole story, because it moves the work of qualifying from the state, which can check a category, onto the individual, who must assemble a case, and it moves it precisely onto the individuals least equipped to do that work.

The paradox at the center

Here is the self-defeating structure, and it is not rhetorical, it is mechanical. The exemption exists to protect people whose illness is serious enough to prevent them from working. Claiming it requires performing a set of tasks, gathering records across providers, securing documentation of functional impairment, understanding and meeting the requirements, responding within tight deadlines, that themselves demand time, energy, organization, and executive function.

Those are the very capacities that serious chronic illness depletes. A person managing severe illness is already spending their limited reserves on treatment, symptoms, and daily survival, which is much of what "impaired ability to work" means in practice. Asking them to additionally run a documentation gauntlet to prove that impairment is asking them to spend the capacity they are trying to establish they lack. The relationship is inverse: the more genuinely impaired a person is, the more the exemption is meant for them, and the harder the proof burden falls, because impairment and the administrative capacity to document impairment move in opposite directions. An exemption structured this way does not fail randomly. It fails hardest at exactly the people it was designed to catch, and it does so by design rather than accident, because the design tests for incapacity by requiring a demonstration of capacity.

The data-matching hinge, and why the functional test undercuts it

There is a more hopeful path built into the law, and understanding why it is compromised is the key to the whole dispute. The statute directs states to use available information where possible to verify compliance or exemption without requiring additional documentation from individuals. In principle, a state could identify many medically frail people automatically, from claims data, diagnoses, and service-utilization records already in its systems, and exempt them without asking them to prove anything. Done well, that would defuse much of the burden, because the sickest people are often the most visible in medical claims data precisely because they use a lot of care.

The functional-impairment overlay is what sabotages this. Claims data are good at showing diagnoses and services; they are poor at showing functional capacity to work, which is not a billable event and appears nowhere cleanly in a claims record. CMS's rule compounds the gap by stating that diagnosis data alone is insufficient and cannot be older than twelve months. So the exemption now turns on a fact, functional work impairment, that the automated data channel largely cannot establish, which pushes people out of the channel the state can verify on its own and into the channel where the individual must supply documentation. The narrower and more functional the definition, the fewer people can be protected by quiet data-matching and the more are thrown onto manual proof. That is the mechanism by which eligible, genuinely frail people will lose coverage: not because a system judged them able to work, but because neither the system nor they could assemble the specific proof the rule now demands.

The case for the rule, fairly stated

The administration's rationale is coherent and deserves to be represented at its strongest, because the tradeoff here is real rather than one-sided. If the medical-frailty exemption were purely categorical, covering anyone with a listed diagnosis regardless of whether they can actually work, it could become a loophole large enough to swallow the work requirement itself. Many people with a qualifying diagnosis do work, and a diagnosis-only exemption would exempt them along with those who genuinely cannot, which from the standpoint of a policy premised on encouraging work would drain the requirement of meaning. Federal officials frame the functional test as a way to remove waste, fraud, and abuse and to target the exemption to those truly unable to work rather than anyone who carries a diagnosis.

That is a real argument. If one accepts work requirements as legitimate policy at all, the medical exemption has to be bounded somehow, or it becomes an automatic off-ramp that renders the whole requirement optional. A functional test is one logical way to bound it, tying the protection to actual inability to work rather than to a label. The disagreement is not over whether the exemption should have limits, but over whether this particular limit, in this particular form, protects the right people or mostly just erects a barrier that the deserving cannot clear. Reasonable people who accept the premise of work requirements can still disagree about whether the functional-proof design strikes that balance well, and that is the honest shape of the dispute.

The downstream the design tends to produce

The critics' case is equally grounded, and its most important point is about consequences rather than intentions. Independent estimates of the broader policy expect large coverage losses driven not by people who can work being asked to, but by eligible people, including exempt ones, failing to navigate the process; the Congressional Budget Office has estimated that 5.3 million people could become uninsured because they cannot meet or navigate the requirements. And the harm compounds, because a person disenrolled from Medicaid for non-compliance is also barred from subsidized Marketplace coverage, so a paperwork failure can cascade into being fully uninsured rather than merely shifting to another plan.

There is a bitter circularity in where this lands. The population the exemption targets is the seriously chronically ill, and losing coverage is worse for them than for almost anyone, because it interrupts ongoing treatment for conditions that require it continuously. Interrupted treatment tends to worsen health, and worsening health makes a person even less able to work, which is the opposite of the policy's stated aim. An exemption meant to shield the sickest from a work requirement, if it instead strips their coverage through a proof burden they cannot meet, does not just fail neutrally; it actively harms the people it named as deserving protection, and in a way that can make their underlying situation worse. That downstream is not a certainty, but it is the predictable tendency of a design that routes the most impaired into the most demanding proof process.

How to read it

The clear-eyed way to hold this is that the medical-frailty exemption embodies a genuine tradeoff that its design resolves in a specific and consequential direction. A broad categorical exemption is easy for eligible people to obtain and easy to over-claim; a narrow functional one is hard to over-claim and hard for eligible people to obtain. CMS chose the second, for reasons that are defensible if you accept the underlying policy, and the predictable cost of that choice is that some meaningful number of genuinely frail, genuinely eligible people will lose coverage not because they were found able to work but because the proof of their inability was more than their illness left them able to assemble. Whether that cost is worth the tighter targeting is a real political judgment, and this analysis takes no position on it.

What is not really a matter of judgment is the structural observation underneath, which holds regardless of where one stands on work requirements: an exemption that tests for incapacity by demanding a demonstration of capacity is in tension with its own purpose, and that tension falls hardest on the most impaired. How much damage it does in practice depends almost entirely on implementation, specifically on how aggressively states use the data they already hold to identify frail people automatically, rather than pushing them onto individual documentation, since the automated path is the one thing that can protect people the functional test would otherwise strand. For the chronically ill enrollees in the affected states, the practical reality in the meantime is that state outreach is occurring now through late summer, notices of non-compliance carry a thirty-day window to respond, and coverage lost this way can be reapplied for, so attention to the mail and early engagement with a state Medicaid office or a benefits counselor matters more than usual. The exemption is real, and it was meant well. Whether it reaches the people it names will be decided less by the words of the rule than by whether the systems around it are built to find the sick, or to wait for the sick to prove themselves.

Primary sources

  1. STAT's D.C. Diagnosis for the July 30 reporting that chronically ill people must now prove their illness is severe enough to prevent working roughly 20 hours a week, that functional impairments must be well-documented, that Democratic attorneys general challenged the rule as differing from prior state guidance, that the case went before a district judge and could determine the exemption's future in over two dozen states, and that states, advocates, and patients are scrambling to determine qualifying conditions and proof of severity.
  2. Foley Hoag for the interim final rule's addition of a functional-capacity requirement beyond the five statutory categories of qualifying conditions and the July 31 comment deadline.
  3. KFF's analyses for the rule's restrictive, work-capacity-tied definition of medical frailty differing from state expectations, the statutory direction to use available information where possible without additional individual documentation, and the operational and coverage risks.
  4. The Center for Health Care Strategies for the requirement that medically frail individuals demonstrate impaired ability to conduct work activities in addition to a qualifying condition, the rule that diagnosis data alone is insufficient and claims data cannot exceed 12 months, the June 30-August 31 outreach window, the 30-day compliance period, and the CBO estimates of $344 billion in reduced spending and 11.8 million losing coverage.
  5. The American Society of Addiction Medicine brief for the CBO estimate that 5.3 million could become uninsured unable to meet or navigate the requirements and the bar on subsidized Marketplace coverage after non-compliance termination.
  6. The Harvard T.H. Chan School of Public Health for the framing of the narrowed definition, federal officials' "waste, fraud, and abuse" rationale, and the January 1, 2027 implementation deadline.
  7. The Center on Budget and Policy Priorities and govtschemes.org for the medically-frail exemption as the rollout's central flashpoint and the state-policy choices that can reduce or increase coverage loss.