The Policy Reversal Happened Faster Than Anyone Expected
In March 2025, the Centers for Medicare and Medicaid Services issued new guidance that essentially opened a door that had been locked for four years. States could now apply for waivers to impose work requirements on Medicaid recipients. Within weeks, the applications started arriving. By February 2026, at least 12 states had either submitted waiver requests or publicly announced they were preparing to do so. Texas. Florida. South Carolina. The list keeps growing.
What’s striking isn’t just the speed. It’s how the information travels. I’ve been monitoring these applications the way I used to track zoning permits, calling state Medicaid directors and asking the same three questions: What exactly are you proposing? How many people will this affect? Where is the cost-benefit analysis? The answers vary wildly depending on whether you’re talking to a state health official or reading the actual waiver language posted on the CMS website.
The Georgia case offers the clearest picture we have so far of what implementation actually looks like when the political will exists to move forward. The Trump administration approved Georgia’s Pathways to Coverage program in early 2025, making it the first active Medicaid work requirement program in the country. By mid-2025, approximately 5,400 people had enrolled. The administrative cost to set up and run the program reached $26 million.
One Small Program. One Huge Question About Scale.
Here’s where a journalist’s instinct kicks in. Take that Georgia number and multiply it. Five thousand four hundred people enrolled. Twenty-six million in administrative costs. That’s roughly $4,800 per person just to manage the paperwork and verification systems. Now imagine that across Texas, which has nearly 4.5 million Medicaid beneficiaries, or Florida with 3 million. The administrative machinery required becomes almost incomprehensible.
I called three sources in state government last month to ask how they calculated their projected administrative costs for similar programs. One was honest: they hadn’t done detailed calculations yet. Another sent me a generic spreadsheet. The third pointed me to a consultant’s report that cost the state seventy-five thousand dollars to produce and was deemed too preliminary to share.
This is where the data problem becomes urgent. We don’t have clean numbers on what these programs actually cost to administer at scale. We have Georgia. We have projections. We have historical data from failed experiments. But we don’t have evidence from a large-scale, sustained work requirement program operating in a modern Medicaid system with digital verification requirements and continuous compliance monitoring.
The Arkansas Warning That Keeps Getting Overlooked
Let me take you back to 2018. Arkansas implemented a Medicaid work requirement program with much fanfare and minimal opposition. Within a year, 18,000 people had lost their coverage. Not because they didn’t work. The research published in the New England Journal of Medicine showed most of them actually had jobs. They lost coverage because they missed a single reporting deadline, didn’t understand the requirement, or couldn’t navigate the verification process.
Courts eventually blocked the Arkansas program. The damage was already done. Eighteen thousand people without health insurance. The ripple effects lasted for years.
When I ask state officials about Arkansas now, I get the same response repeatedly: “We’ve learned from that. Our program will be different. We’ll do education. We’ll be flexible with reporting.” This is where source criticism matters. These aren’t lies. They’re aspirations dressed up as plans. The officials genuinely believe they’ll execute better than Arkansas did. But they’re also operating under political pressure to implement quickly, with budget constraints, and without the infrastructure that would make compliance actually manageable for low-income workers juggling multiple jobs and unstable schedules.
What the Research Actually Shows About Coverage Loss
The Urban Institute conducted the most comprehensive pre-implementation analysis we have. Published in 2023, the study projected that if work requirements were imposed nationally across all states, between 1.4 and 4 million people could lose Medicaid coverage. That range is enormous. The difference between the low and high end represents millions of people.
The variation exists because predicting behavioral responses is difficult. Will people increase work hours to comply? Will they leave the workforce because compliance becomes impossible while working unstable jobs? Will administrative errors cause coverage loss among people who actually meet requirements? All three happen simultaneously. The researchers constructed different scenarios. The optimistic scenario assumes good program design and minimal administrative friction. The pessimistic scenario assumes poor implementation and high error rates.
Given what we know about Arkansas, the pessimistic scenario probably tracks closer to reality. But state officials making waiver applications are naturally citing the optimistic projections to justify their programs.
Tracking the Waiver Applications: What to Watch For
If you want to follow this closely, the KFF Medicaid Work Requirements Policy Tracker maintains the most up-to-date public record of which states have applied and what they’re proposing. The CMS Medicaid Waiver Guidance 2025 explains the process states must follow and the documentation they need to provide.
But here’s what those resources won’t tell you: Which states are actually collecting baseline data about their current Medicaid population before implementation? Which states have hired enough staff to handle the verification process without overwhelming workers with confusing requirements? Which states are building in feedback mechanisms so they can actually see if people are losing coverage due to administrative errors versus work status?
These questions matter because they determine whether we’re going to repeat Arkansas or actually learn from it. I’m still making those calls. Still asking for the documentation. Still pushing back when officials offer me vague assurances instead of specific implementation plans. If you’re watching this unfold in your state, do the same. Ask your state legislators. Request the waiver applications. Look for the baseline data. The information is out there. You just have to know where to look and how to read it critically.