Congress priced five million people out of health insurance.
The number comes from a new analysis by KFF, and it means that one in five people who had coverage through the Affordable Care Act marketplace will lose it this year. The mechanism is straightforward: the enhanced subsidies that Congress first enacted in the American Rescue Plan, and then extended through the end of 2025, expired on January 1. Those subsidies capped what any household paid for a benchmark plan at a percentage of income — no more than 8.5%, and considerably less for lower-income families. Without them, the cap rises, sometimes sharply. Deductibles are up by more than a thousand dollars. Monthly premiums are up by sixty-five. The KFF numbers do not measure a natural market correction; they measure a designed reversion to a price point that five million enrollees cannot afford.
It is worth being precise about what is happening, because the public discourse has the habit of treating an insurance subsidy as a budget line item rather than as the structural load-bearing element it turns out to be. The Affordable Care Act was designed before these enhanced subsidies existed. The original law capped premiums relative to income at a higher percentage and ended subsidy eligibility entirely above 400 percent of the poverty line. The American Rescue Plan filled those gaps temporarily, on the theory that a pandemic was not the moment to let people go uninsured. The fix worked: enrollment rose, the uninsured rate fell, the individual market stabilized. Congress then extended the fix twice, most recently through the Inflation Reduction Act. At the end of 2025, the fix was allowed to expire. No vote was taken to end it; the expiration was the default, built into the original legislation. Inaction is not the absence of a decision. Inaction is a decision, and it has consequences.
The trouble is that the patch was doing the only structural work the underlying system could not do, which is decoupling the premium from the actuarial value. When the credit expires, the premium reverts to the underlying cost curve. The underlying cost curve is administered by the same hospital systems and pharmacy-benefit managers that operate as rent-extraction cartels in almost every other regulated sector. The ACA marketplace is not an insurance product in the traditional sense; it is a publicly subsidized conduit for private risk-pricing. The subsidy was never a market subsidy. It was a rent subsidy. When it expires, the rent extraction resumes at the statutory pace, and the enrollment numbers collapse to the level that the extraction curve can sustain.
The mechanism is actuarial and administrative. The ACA’s premium tax credit caps an enrollee’s contribution as a fixed percentage of income. Below the cap, the credit subsidizes the carrier directly. When the enhanced credit lapses, the percentage-of-income cap tightens, the subsidy shrinks, and the carrier is free to load the residual cost onto the deductible. This is the same cost-shifting move that turns a broad risk-pool into a narrow one: healthy enrollees drop out first, the remaining pool ages, the carriers raise premiums to cover adverse selection, and the next tranche of enrollees drops out. The loop is recursive. It does not need a conspiracy to operate; it needs only the statutory expiration date and the actuarial reality that insurers are priced to profit on the people who stay, not the people who leave. The January drop of eight hundred thousand was the leading edge. The five-million projection is the full curve.
Marcia Angell’s work, from her tenure editing the New England Journal of Medicine through her exposés of pharma pricing, has shown repeatedly that the manufacturing cost of a drug or the marginal delivery cost of a service is a sliver of the charged amount; the rest is the carry cost of the private-administration apparatus. The statutory architecture, by design, delivered exactly what its 2010 drafters intended: coverage expansion funded through private-premium subsidies while leaving the administrative layer, provider consolidation, and deductible structures untouched. That there is now a five-million-person gap is not a bug; it is the architecture operating as built. The carriers are not operating outside the rules. They are operating exactly at the margin the rules permit, and the margin has widened into a gap that will swallow five million people.
The consequences are not distributed evenly. The people losing coverage are disproportionately in states that did not expand Medicaid, where the marketplace serves as the only option for low-income adults. They are, overwhelmingly, working — the kind of work that does not come with employer-sponsored insurance: home health aides, retail clerks, delivery drivers, the people whose labor the economy runs on and whose health the economy then declines to pay for. As MSI reported in March, enrollees are already cutting food and skipping care when faced with these costs.
There is a temptation to describe this as a cliff — a natural feature of the budgetary landscape, an unavoidable consequence of fiscal constraints. It is not. It is a policy choice. The same Congress that allowed these subsidies to lapse found the money to extend trillions in tax cuts heavily concentrated at the top of the income distribution. The cliff is man-made. The people who built it will not fall off.
The reflexive political response will be to demand a simple restoration of the enhanced credits. That is an arithmetic placebo. Restoring the credits does not touch the underlying cost structure; it merely re-inflates the balloon after the statutory leak has let the air out. A recurring leak that must be plugged with taxpayer money every legislative cycle is not a policy fix. It is a recurring appropriation to the rent-extraction sector.
The only structural remedy that closes the gap is to decouple coverage from the private-premium apparatus entirely. A public buy-in or a single-payer mechanism that sets a binding fee schedule and eliminates the administrative middleman is not a political preference; it is an engineering necessity for a network that cannot scale its premiums without collapsing its enrollment. The current architecture requires a continuous subsidy to keep the premium from outpacing the income cap. The architecture should replace the premium with the fee schedule. Until that substitution is made, the marketplace is a leak vessel, and the subsidy is just a bucket.
When a steel mill closes, the consequences are visible: the gates locked, the lunch pails carried home, the town knowing at a glance what has been taken from it. When a health-insurance subsidy expires, the consequences are quieter — until the bills arrive, until the skipped checkup becomes the late-stage diagnosis, until the county hospital’s charity-care ledger swells with the cost of a decision made in Washington by people who will never have to choose between a premium and a grocery bill. The mechanism is different. The extraction is the same. The actuaries have already priced the enrollment drop into their risk-adjustment models. The extraction curve does not wait for the rhetoric to catch up to it.