The enhanced premium tax credit was one of the largest coverage expansions since the Affordable Care Act itself, and letting it lapse at the end of 2025 was a mistake that is now visible in the data. Paid enrollment is down by about 2.6 million people, premiums paid by enrollees are up by more than half, the people above the old income cliff are leaving in the largest numbers, and insurers are pricing in a sicker pool for a second straight year. Congress should revive the credit and make it permanent. At roughly $35 billion a year, it buys more coverage, lower full-price premiums and real relief for middle-income families, all at once.
What the enhanced credit did
The Affordable Care Act’s premium tax credit caps what a household pays for a benchmark “silver” plan at a share of its income. As originally written, that share rose to almost 10 percent, and the help stopped entirely at 400 percent of the poverty line, about $62,600 for a single person in 2026. Earn a dollar more and you paid full price.
The 2021 American Rescue Plan, later extended through 2025 by the Inflation Reduction Act, rewrote that schedule. Enrollees up to 150 percent of poverty paid nothing for a benchmark plan, the share of income everyone else paid fell, and above 400 percent of poverty no one paid more than 8.5 percent of income. The cliff was gone.
People responded. Open-enrollment sign-ups rose from 11.4 million for 2020 coverage to a record 24.3 million for 2025, and the share getting a subsidy climbed from 84 percent to 92 percent.
What expiration has already cost
Congress came close to acting and did not. The Senate blocked a three-year extension in December 2025, 51 to 48, short of the 60 votes needed to advance it. The House passed one 230 to 196 in January 2026, with 213 Democrats and 17 Republicans in favor, but the bill has sat on the Senate calendar since February. The IRS’s 2027 contribution table still stops at 400 percent of poverty, so 2027 coverage will be sold under the old rules again.
The damage shows up first in prices. Under the restored schedule, the average enrollee’s monthly payment after subsidies rose 58 percent, from $113 to $178. That understates the hit, because millions of people kept their payment down by trading down: bronze plans rose from 30 percent to 40 percent of sign-ups, and the average deductible jumped 37 percent to a record $3,786.
Then it shows up in coverage. Open-enrollment sign-ups fell to 23.1 million, the first drop since 2020. The better measure is people who actually paid their first premium. By that count, February enrollment fell 12 percent, from 21.8 million to 19.2 million. Paid enrollment fell in every state but one: New Mexico, the only state that fully replaced the lost federal help with its own money, grew 14 percent. Sign-ups, a looser measure shown in the state chart below, rose in a handful of states.
The cliff is the sharpest part of the change. For a 60-year-old earning just under 400 percent of poverty, a benchmark plan at the 2026 national average premium costs about $6,200 a year. At 405 percent, about $63,400 of income, the same plan costs $15,900, a quarter of income. With the enhanced credit, it would cost $5,400.
The enrollment data show people reacting to exactly that. Sign-ups from people with incomes above 400 percent of poverty fell from 1.64 million to 1.07 million, a 35 percent drop. They were 7 percent of 2025 enrollment but nearly half of the decline in sign-ups. Under the enhanced credit these households were expected to pay up to 8.5 percent of their income for a benchmark plan. They were the market’s paying customers, and they walked away when the bill rose by a third or more just above the cliff, and doubled or tripled for older enrollees.
The risk pool is getting sicker
Who leaves matters as much as how many. When prices rise, the healthy are the first to decide insurance is not worth it, and the people who stay are sicker and costlier to cover. That is the classic insurance death spiral, and the subsidy structure normally blocks it, because a subsidized enrollee’s payment is pegged to income rather than to the premium. The people above the cliff have no such protection, and neither do the healthy enrollees who can drop to bronze.
The Congressional Budget Office now says the average benchmark premium “increased sharply in 2026, largely driven by the disproportionate disenrollment of healthier marketplace enrollees” after the credit expired. The pattern is continuing. Insurers’ 2027 filings propose a median increase of another 15 percent, with many citing the expiration and a sicker pool, and 11 carriers have announced they will pull out of some or all of the states where they sell marketplace plans next year. CBO estimated in September 2025 that a permanent extension would make full-price benchmark premiums 7.6 percent lower on average through 2035, because the people it brings in are healthier than average.
That matters beyond the marketplaces. People who lose coverage still get sick, and the bills land on hospitals, doctors and other providers and, through them, on everyone else. Before expiration, the Urban Institute projected $7.7 billion more in uncompensated care in 2026 alone.
Who would benefit from bringing it back
Reviving the credit would help four groups most.
- Low- and moderate-income enrollees. About three-quarters of 2025 sign-ups reported incomes under 250 percent of poverty. The Urban Institute estimated that expiration would raise their average annual premium payment from $169 to $919.
- Older adults who are not yet on Medicare. About 8 million enrollees were 50 to 64 in 2023. Because premiums for a 64-year-old can be three times those for a 21-year-old, they face the tallest cliff, and about half of the enrollees above 400 percent of poverty who stood to lose eligibility are in this age group, according to KFF.
- The self-employed and small-business owners. The Treasury Department counted 3.3 million of them in the marketplaces in 2022, 28 percent of enrollees aged 21 to 64. Without the enhanced credit, buying your own insurance is again a large tax on leaving a job with benefits to start a business.
- The South and the states that voted for President Trump. 77 percent of 2025 enrollees, 18.7 million people, lived in states he won in 2024. Enrollment more than tripled from 2020 to 2025 in Texas, Mississippi, West Virginia, Louisiana, Georgia and Tennessee, mostly states that never expanded Medicaid, where the marketplace is often the only option for low-income adults.
The objections, and the price
The main argument against the enhanced credit is fraud. Zero-premium plans gave unscrupulous brokers an incentive to enroll people without their knowledge, and the Paragon Health Institute, a conservative think tank, estimates that 6.2 million 2026 sign-ups were improper. That estimate compares low-income sign-ups with Census counts of who could be eligible. Paragon also cites the high share of enrollees who filed no claims as evidence of “phantom” enrollment, an argument KFF’s Cynthia Cox has rejected: treating “anyone who’s not using health care” as a fraudulent enrollee, she said, is “not true.” But the problem is real, and it has a direct fix: verify enrollment, police brokers and require a small minimum payment. The Centers for Medicare & Medicaid Services has already canceled enrollments covering more than 760,000 people it confirmed were unauthorized. Cutting subsidies for 20 million people to stop fraud by brokers is like closing a bank because some checks bounce. The 35 percent drop among the paying customers above the cliff shows the cost of using the blunt tool.
The other objection is the price. CBO estimated in September 2025 that making the enhanced credit permanent would add $350 billion over 2026 to 2035, about $35 billion a year, and increase the number of people with insurance by 3.8 million in 2035. That is real money, and it should be paid for. But much of it goes to lower premiums for the 20 million subsidized people who already buy insurance, and to a cap on what a 60-year-old earning $65,000 can be charged. Few uses of $35 billion a year cover millions more people, stabilize an entire insurance market and cut what middle-income families pay at the same time.
Short-term extensions are a worse deal. The House’s three-year bill would cost $80.6 billion and leave insurers, and families, guessing again when it runs out. The last expiration showed what that uncertainty costs. Congress should restore the 2021 schedule permanently, pair it with real fraud controls, and end the cliff for good.