Your web browser is out of date. Update your browser for more security, speed and the best experience on this site.

Update your browser

Blog November 13, 2026

Medicaid Work Requirements Are Starting to Look a Lot Like the Unwinding

Adam Auerbach
Author
Adam Auerbach

Similar to the unwinding in 2023, the bigger risk for Medicaid members in 2027 isn't that they’ll fail to meet the work requirement. It's that administrative verification and communication breakdowns will cost eligible people their coverage.

Ten weeks out from national implementation, the first real data looks familiar. Ten weeks out from federally mandated implementation, the first real data looks familiar. Here's what states can still do before January 1.

Nebraska started enforcing Medicaid work requirements on May 1, eight months ahead of the rest of the country, and at first glance the results look fine. Of the roughly 7,300 people who went through a renewal under the new rules this summer, 93% kept their coverage. The more telling number sits beneath the waterline. Of the people who lost coverage, 92% never responded to the state's request for information. Very few were found to be not working. Most of them simply never answered the mail.

Anyone who worked in Medicaid during 2023 and 2024 will recognize that pattern. During the unwinding, about 69% of the people who lost coverage were disenrolled for procedural reasons rather than because a state found them ineligible. Nebraska is already running ahead of that rate — on a small population, early in the process, with self-attestation still allowed. Its numbers are more likely to be the best case than the worst.

Most of what states learned during the unwinding carries over to work requirements. But I'd argue the unwinding was the easier version of this problem, and states that plan as if 2027 will look like 2023 are going to be caught short.

Why This Time Is Harder

The unwinding was a single pass through the caseload. Work requirements are checked at application and at every renewal, and expansion adults move to six-month renewals in 2027. Centers for Medicare and Medicaid Services (CMS) estimates that about 8.8 million people won't be verifiable from data and will have to submit something at each check. The procedural-loss pattern we saw once is about to become a standing feature of the calendar. Churn comes with it. In states that have the capacity to track the data, 25% to 40% of people disenrolled during the unwinding came back onto Medicaid, and each of those round trips is work the state ends up doing twice.

The safety valves are also mostly removed. During the unwinding, states could pause procedural disenrollments, and many did. The work requirement timeline is set in statute. The (e)(14) waivers that let states renew people using SNAP data and MCO-supplied addresses expired in June 2025, and the effect showed up right away: California's ex parte renewal rate fell from 78% to 36% the following quarter. Someone who loses Medicaid for not meeting the requirement can't fall back on subsidized Marketplace coverage, and retroactive coverage for expansion adults shrinks from three months to one on the same day the requirements start.

Whatever your view of the policy, the practical consequence is the same. A procedural loss in 2027 costs the member and the state more than the same loss did in 2023.

Where to Expect Things to Break

The unwinding lesson that rarely makes the summaries is about eligibility logic. In 2023, 29 states turned out to be running renewals at the household level instead of the individual level, and about 420,000 people, many of them children, lost coverage before anyone caught it. The work requirement rules carry far more branching than that. Exclusions have to be checked person by person before activity is considered, medical frailty depends on a 12-month claims lookback, and the parent of a 13-year-old is treated differently from the parent of a 14-year-old. All of it is being built quickly, against a rule that's still being litigated. Expect a defect of that kind to surface somewhere in 2027, and the states that find their own will be in a far better position than the ones that hear about it from CMS. 

The second weak point is the mailbox. Nebraska's losses are mostly non-response, and non-response usually starts with returned mail, outdated addresses and notices nobody opens. Paper is still the legal default, and CMS assumes 75% of enrollees will never opt into electronic notices. The rule does give states an opening, though. Every outreach notice has to go out through a minimum of two channels, which can include text, phone or the member's online account. Most states will treat that second channel as a compliance requirement but in fact, it's one of the most useful tools states have for proactive communication.

States have some experience to draw on here. At least 15 states texted enrollees during the unwinding, several for the first time, and the results were mixed. Florida's texts caused panic, partly because of poor translations. Digital outreach works when it gets the same care as the notice itself: plain language, real translation, testing with actual members. It also works better as a sequence than as a single message—a text before the letter arrives, a reminder before the deadline, and a call from the health plan when neither gets a response.

Timing matters as well. Current enrollees are first checked at their first renewal on or after January 1, so once notice periods and disenrollment timing play out, the first real wave of losses lands between February and April. States are heading into that window with less capacity than they had in 2023. The longest average peak call wait time in any state during the unwinding was 72 minutes. In the first quarter of 2026, Hawaii's was 193.

What States Can Still Do

No one is replacing an eligibility system before New Year's, but there is meaningful work that fits in the next ten weeks, and most of it proved itself during the unwinding.

1. Test 

Test the exclusion logic before members do. Build synthetic cases where the rules are most likely to break - mixed-age households, frailty that shows up only on a denied claim, an exclusion that lapses mid-period - and run them all the way through notice generation.

2. Verify

Use the data you already have. SNAP's work rules overlap heavily with Medicaid's, and every exclusion confirmed from data is one less notice someone has to answer.

3. Outreach

Fix contact data and start outreach earlier. Run address updates now, collect cell numbers and e-notice consent at every contact, and begin outreach well before the renewal deadline across text, email, phone, and mail. Then share the non-responder list with the health plans, health centers and community partners who can actually reach people.

4. Prepare

Plan for March, not January. Size call center and processing capacity for the first renewal wave, track non-response by cohort, channel and language, and decide now how you would reinstate a group of members if a defect turns up. During the unwinding, at least 35 states had to do exactly that.

Looking Ahead

The unwinding was an expensive lesson in how procedural loss happens and what reduces it. Nebraska’s early data suggests that those lessons apply again. The difference is that states now know where the system is likely to break. Testing the rules, using the data they already have, reaching members earlier, and preparing for the first renewal wave can help prevent the same problems from repeating themselves.

There’s still time to act on those lessons, but the window is getting shorter.

Adam Auerbach

Director

Adam is passionate about drawing new insights from health data and building digital experiences that simplify access to meaningful healthcare. Clients rely on his 10+ years of expertise to guide and oversee strategic initiatives ensuring that systems are efficient, scalable, and aligned with stakeholder objectives. A life-long learner and Health Systems Subject Matter Expert (SME), Adam is dedicated to solving complex problems through tech-forward solutions.

LinkedIn Envelope