The federal government has graciously decided that states can wait until July before kicking poor people off Medicaid for insufficient paperwork. Six months. A half-year reprieve, delivered via letter from Dan Brillman of CMS, who acknowledges — hold your applause — that states are struggling to build the Byzantine software systems needed to verify whether a home health aide earning $11 an hour has documented her 80 monthly hours of labor.
Let’s savor the irony here. The work requirement was the centerpiece of the 2025 tax-and-domestic-policy megabill, sold as a triumph of personal responsibility economics. The intellectual case was always thin — most Medicaid adults who can work already do work; the ones who don’t are typically disabled in ways the forms don’t capture, caring for family, or navigating labor markets that don’t exactly roll out the red carpet for the uninsured. But never mind the data. The policy was philosophically satisfying, and philosophically satisfying policies have a long, proud history of colliding catastrophically with administrative reality.
And collide it has. A recent court loss over a key element of the requirement, states begging for timeline relief, hospitals staging “massive education campaigns” — a euphemism for spending millions teaching patients to fill out forms so they don’t lose their insurance while undergoing chemotherapy. What an efficient allocation of healthcare resources. Medicaid Directors now have until July before coverage terminations begin, though Brillman insists states must still start checking in January. So the bureaucracy will dutifully verify employment while withholding the guillotine — a six-month limbo in which millions discover whether their gig-work pay stubs satisfy a federal algorithm.
The historical precedent here should embarrass anyone still touting this as workable policy. Arkansas’s 2018 experiment — the only one actually implemented — terminated coverage for roughly 18,000 people in months, with studies showing no measurable employment gains, only losses of insurance. Kentucky’s version collapsed in court. New Hampshire’s melted down. This isn’t a bug of work requirements; it’s the product. They don’t move people into jobs. They move people off rolls, which is the point, though we’re asked to pretend otherwise. The Congressional Budget Office projected substantial coverage losses from this provision and modest, if any, employment effects. Everyone voting for it knew the arithmetic. The paperwork burden is the mechanism, not an accident.
Which raises the obvious question: if the states administering this policy can’t get their systems running by the deadline, why did anyone believe enrollees — disproportionately poor, transient, low-wage workers with irregular schedules — would be able to document 80 hours a month with ease? The administration’s answer, implicitly, is that the friction is the feature. Coverage losses reduce federal outlays. That’s the fiscal math underneath the moral rhetoric, and the delay changes none of it. It merely moves the cliff from January to July, conveniently past any news cycle that might connect insurance terminations to the politicians who ordered them.
Notice, too, what this episode reveals about governance by deadline. Congress passed a mandate knowing states lacked the infrastructure to execute it, then expressed surprise when states said — as Brillman summarizes — they faced “impediments to successful launches at scale.” You don’t say. Building eligibility-tracking software that interfaces with state employment data has been a graveyard for state IT projects for decades. Arkansas’s system cost millions and still wrongly terminated enrollees. This time the scale is national, the timeline was tighter, and the stakes are larger.
For investors watching healthcare and hospital names, the delay is a six-month reprieve for providers already girding for uncompensated-care spikes — hospitals have been prepping those education campaigns precisely because every disenrolled patient who shows up in the ER becomes their balance sheet’s problem. Rural hospitals, operating on margins thinner than a philosopher’s conscience, stand to absorb the brunt. The July date doesn’t change the destination; it just gives administrators time to triple-check their code before the terminations flow.
The deeper lesson should be familiar to anyone who’s watched central bankers insist their models capture the economy: policymakers designing incentives on whiteboards rarely account for the humans actually living inside them. Work requirements sound rigorous in a policy paper. In practice, they’re a paperwork tax on the poor, administered by software that doesn’t work, enforced by states that asked for more time, defended by a government that just admitted it wasn’t ready.
January was supposed to be the starting gun. Now it’s July. And when July arrives and the terminations begin — as they will, because the savings were always the point — no one in Washington will have to explain why they delayed. They’ll just point to the extra six months of careful, responsible implementation. History suggests the people losing coverage won’t get the same courtesy of a smooth transition.
