Why People Who Still Qualify Keep Losing Their Medicaid

Starting in September, states must begin notifying Medicaid enrollees about a set of new requirements, including work rules and twice-yearly eligibility checks, that were created by the 2025 tax-and-spending law and take effect in 2027. In a recent STAT essay, three health-policy researchers laid out evidence on how states should handle that outreach, and the reason their work matters points to something easy to miss about how public benefits actually function. The central danger in the months ahead is not that ineligible people will be removed from Medicaid. It is that people who still qualify will lose their coverage simply because they cannot get through the new paperwork.

That distinction, between losing a benefit because you no longer qualify and losing it because you could not navigate the process of proving you still do, sits at the heart of the coming changes, and it deserves more attention than the louder political fight around them usually allows.

The gap between qualifying and keeping coverage

There is a difference between who is eligible for Medicaid and who actually manages to keep it, and the space between those two groups is governed by administrative friction, the sheer difficulty of proving and maintaining eligibility over time. That space is not small. In a national survey of people who lost Medicaid coverage in 2023, more than 17% believed they had lost it not because they became ineligible or chose to leave, but because they could not complete the renewal process.

This is what researchers call procedural, or administrative, disenrollment: eligible people losing coverage to red tape rather than to any change in their circumstances. It has long been a feature of Medicaid renewals, and the new rules are expected to make it worse, because they layer on more frequent eligibility checks, every six months instead of once a year, plus the added burden of documenting work hours. Each additional requirement is another point at which an eligible person can fall out of the system without ever becoming ineligible.

Administrative friction is a hidden eligibility rule

Once you see coverage this way, an important truth follows: the real rule about who gets covered is not just the law's eligibility criteria. It is those criteria plus the administrative burden of complying with them. Add paperwork, and you reduce actual enrollment even among people who remain fully eligible, because some share of them will not manage the added steps, whether from a missed notice, a confusing form, or a deadline that slipped past during a hard month.

So administrative friction operates as a kind of hidden eligibility rule, a policy lever that changes who actually holds coverage, separate from and layered on top of the formal criteria everyone argues about. A debate framed entirely as "who should qualify?" can quietly produce an outcome that is really "how many eligible people will lose coverage because of paperwork?" Those are different questions, and they have different answers. The second one tends to go undiscussed precisely because it operates in the background, in the machinery of notices and renewals rather than in the language of the statute.

The rare point both sides should share

Whether Medicaid should carry work requirements at all is a genuine and heated values dispute. Supporters see them as encouraging work and reserving public assistance for those who most need it; critics see them as a mechanism for cutting coverage. This analysis takes no side in that argument, which turns on values on which reasonable people disagree.

The procedural problem, though, sits underneath that dispute rather than inside it, and it is something both sides have reason to care about. Whatever one believes about work requirements, an eligible person losing coverage because of red tape is a failure on anyone's terms. It does not advance the stated goal of the policy, which is to move able-bodied non-workers off the rolls, not to remove eligible workers and exempt people through administrative error. It does not save money in the way the policy intends, because it cuts people the policy meant to keep. And it plainly harms someone who was supposed to remain covered. So reducing procedural disenrollment is a rare piece of common ground: supporters should want it because it protects the policy's legitimacy and its actual aim, and critics should want it because it protects vulnerable people, two paths to the same conclusion from opposite starting points.

The friction falls hardest on those with the least

Administrative burden does not land evenly, and this is where the human stakes come into focus. The researchers found that the benefit of extra outreach was greatest for the most vulnerable enrollees, Native American tribal members, children, and people with chronic conditions. Among tribal members in their study, more than one in three lost coverage for procedural reasons, a striking measure of how unevenly the paperwork falls. And written-only outreach, they note, tends to help healthier and higher-income enrollees, who face less friction in acting on a letter, which means that leaning on written notices alone systematically misses the people who depend on Medicaid the most.

Seen plainly, administrative burden is an invisible and regressive tax. It is paid not in dollars but in time, attention, a stable mailing address, a current phone number, and the sheer capacity to fight one's way through a bureaucracy, and those are exactly the resources the most vulnerable have in shortest supply. How the paperwork is administered therefore determines who falls through it, and it falls hardest on the people who can least afford to lose their coverage and who will suffer the most from losing it.

The encouraging part: this is cheap to prevent

The genuinely hopeful finding is that procedural disenrollment is not inevitable. It is a design choice, and better design can prevent much of it for very little money. The researchers tested simple, low-cost outreach methods and found that they work. Text messages proved just as effective as mailed postcards at a fraction of the cost, roughly nine cents versus thirty-eight per contact. Adding a prerecorded phone call that offered help from a navigator measurably raised renewal rates and cut procedural denials, at about fifteen cents per contact and roughly fifteen dollars per additional renewal preserved, far below the value of the coverage itself.

Keeping contact information current turns out to matter enormously as well: the study found that at least one in five phone numbers or addresses on file no longer reached the right person, meaning even well-designed outreach can fail before it begins. The encouraging implication is that the tools to prevent avoidable coverage loss already exist and cost almost nothing. The open question is simply whether states choose to use them.

The timing, unfortunately, runs the wrong way

The complication is that the burden is rising at the same moment the help is being cut. The new rules add friction, more frequent renewals and work-hour documentation, on a short implementation timetable, while the navigator programs best equipped to guide enrollees through the process just absorbed a roughly 90% cut in federal funding. Assistance capacity is shrinking precisely as the paperwork grows, and unless states actively deploy the cheap, tech-enabled outreach the research points to, the predictable result is that more eligible people slip through.

If that happens, the coverage those people lose will not be a verdict on their work ethic or their eligibility. It will be a function of how much friction was added and how little help remained to navigate it, a distinction worth holding onto when the enrollment numbers eventually come in and are read as evidence for one political story or another.

The fight over Medicaid work requirements is a real argument about who should be covered, and it deserves to be had on its merits, which this piece does not try to settle. But running beneath that argument is a quieter and more universal fact: a benefit's real reach is set not only by who qualifies on paper, but by how hard it is to prove and hold onto that qualification, and administrative friction is a lever that works in the dark, trimming coverage among the eligible and pressing hardest on those least able to withstand it. The person who loses Medicaid over a letter that never arrived or a phone number that changed is not making a statement about work. They are falling through a gap that cheaper, smarter design could close. Whatever one believes about who ought to be covered, no version of that goal is served by removing people who qualify simply because the paperwork defeated them, and the evidence says preventing it is both possible and very nearly free. That is worth doing on anyone's terms.

Primary sources

1. A STAT First Opinion essay by Rebecca Myerson (Emory University), Allison Espeseth (Covering Wisconsin), and Laura Dague (Texas A&M University) for the September 2026 requirement that states begin notifying affected enrollees, the 2027 effective date of the work requirements and twice-yearly eligibility redeterminations created by the One Big Beautiful Bill Act of 2025, the requirement that states use at least two outreach methods, the finding from a national survey that about 17.4% of people who lost Medicaid in 2023 attributed it to inability to complete the renewal process, and the authors' randomized study with Covering Wisconsin, including that text messages matched postcards at far lower cost, roughly nine versus thirty-eight cents; that supplementing written outreach with a prerecorded navigator-assistance call increased renewals by about one percentage point and reduced procedural denials, at roughly fifteen cents per contact and about fifteen dollars per additional renewal; that effects were largest for tribal members, children, and people with chronic conditions, with more than one in three tribal members in the sample losing coverage for procedural reasons; that written-only outreach tends to favor healthier and higher-income enrollees; that at least one in five phone numbers or addresses on file were outdated; and that federal navigator funding was recently cut by about 90%.
2. The underlying research the essay cites, including studies in JAMA and JAMA Internal Medicine on procedural disenrollment, Urban Institute analysis of the six-month redetermination's projected enrollment effects, and the authors' NBER working paper.