Why ACA Premiums Jumped When the Enhanced Tax Credits Expired

Imagine two married couples in their early sixties, living on the same street in a small town, both buying their own health insurance because they retired early or run a small business. One couple earns a dollar less than four times the federal poverty line. The other earns a dollar more. Under the Affordable Care Act as Congress first wrote it, the first couple got a tax credit that capped their premium at a share of their income. The second got nothing at all, and could face a premium bill several times larger, for the same plan, because of one dollar. That dollar is called the subsidy cliff. In March 2021 Congress filled it in. At the end of 2025 the fill washed away, and when people opened their bills for 2026, many of them found themselves standing at the edge again.

In short: The enhanced premium tax credits created in 2021 lowered what marketplace enrollees paid at every income level and removed the cutoff at 400 percent of the poverty line. Congress did not extend them past 2025, so the original, smaller credits returned in 2026, and with them the cliff. This explainer shows how the credit is calculated, who felt the change most, and what the fight over it was really about.

How the credit is built

The ACA's premium tax credit is not a flat coupon. It is a formula that works backward from what a household is expected to pay.

Start with the benchmark plan: in each area, the second-lowest-cost silver plan on the marketplace. Then take the household's income and look up its "applicable percentage," a share of income the law says the family should contribute toward that benchmark. The credit is the difference between the benchmark's full price and that expected contribution. If the benchmark costs $15,000 a year and the household is expected to pay $6,000, the credit is $9,000. The family can apply that credit to any metal level of plan; pick something cheaper than the benchmark and you pay less, pick something richer and you pay more.

The design has an important consequence. Because the credit fills the gap between a fixed contribution and the benchmark price, it rises when premiums rise. A family below the cutoff is partly shielded from premium increases: the government absorbs most of the change. That shield disappears for anyone who is not eligible for the credit at all.

Under the original law, the applicable percentages rose with income, from roughly 2 percent of income at the bottom to about 9.5 percent near the top, adjusted each year. And at 400 percent of the poverty line, eligibility simply stopped.

What the 2021 law changed

The American Rescue Plan Act, signed in March 2021, rewrote that table for 2021 and 2022. It did two things.

First, it lowered the expected contribution at every income level. For people with incomes up to 150 percent of the poverty line, the applicable percentage fell to zero, which meant a benchmark silver plan could cost nothing. At the higher end, it capped the expected contribution at 8.5 percent of income.

Second, and this is the cliff, it removed the 400 percent ceiling. Any household, at any income, could qualify for a credit if the benchmark plan cost more than 8.5 percent of its income. For a couple earning comfortably above the line, that might mean no credit at all, because 8.5 percent of a high income easily covers a benchmark plan. But for older enrollees, whose premiums are much higher because insurers are allowed to charge older adults up to three times what they charge the youngest adults, the cap made a large difference. A 62-year-old couple earning just over the old cutoff in a high-cost area could see thousands of dollars a year in help where before there had been none.

In August 2022 the Inflation Reduction Act extended the enhanced credits through the end of 2025. The extension had a deadline built in, which is how budget legislation often handles costs it does not want to score for the full ten-year window. That deadline is the reason the question came back.

What enrollment did while the credits lasted

When premiums fall, more people buy. Marketplace enrollment climbed steadily through the years of enhanced credits and reached a record of roughly 24 million sign-ups for 2025. The growth was not even across the country. It was especially large in Southern states that had not expanded Medicaid, including Texas, Florida and Georgia, where many low-income adults above the poverty line could now get a zero-premium silver plan.

That pattern matters for understanding the politics. A large share of the people who gained subsidized coverage in those years lived in states that voted Republican, and many of them were working adults: self-employed tradespeople, farmers, small-business owners, people at jobs that offer no insurance. The enhanced credits were not a coastal program. They reached deep into rural America.

There were real problems too, and they deserve naming. Zero-premium plans made the marketplaces attractive to fraudulent brokers, who in some cases enrolled people or switched their plans without consent to collect commissions. The federal government tightened broker rules in response. Critics of the credits pointed to that fraud as a reason to let them lapse. The better answer, I think, was to police the brokers rather than raise prices on the people they defrauded, but the problem was real, and progressives lose credibility when they wave it away.

The lapse and the shutdown fight

As the 2025 deadline approached, the credits became the center of a spending standoff. Democrats in Congress made an extension a condition of their votes on government funding, and the dispute helped produce a federal shutdown that began on October 1, 2025 and lasted into mid-November, the longest in the country's history. The shutdown ended without an extension. Republican leaders offered a later vote, and attempts in the Senate to extend the credits did not get the votes they needed. The enhanced credits expired on schedule, and the original ACA schedule came back for plan year 2026.

Fair-minded opponents of an extension made several arguments. The credits were expensive, with the Congressional Budget Office estimating a permanent extension would cost hundreds of billions of dollars over ten years. They sent some money to households with comfortable incomes. They had been sold as temporary pandemic relief, and temporary measures, the argument went, should be allowed to end. Some conservatives also argued that subsidies feed premium growth by making insurers less sensitive to price.

Those points are not frivolous. Here is the reply. The cost is real, but so is the cost of the alternative, which was paid by families instead of the Treasury. The help for higher earners was concentrated among older people buying their own coverage, a group with genuinely high premiums and few alternatives before Medicare. And "temporary" is a label Congress chose for scoring reasons, not a judgment about whether the policy worked. By the measure the law itself set, more people insured, it did.

What returned in 2026

When the enhanced credits ended, three things happened at once.

The expected contributions went back up. People at the bottom of the income range who had paid nothing for a benchmark silver plan were once again expected to contribute a share of their income. For a family that had grown used to a zero-dollar premium, even a modest new bill was a shock.

The cliff came back. Households above 400 percent of the poverty line lost eligibility entirely. For those older couples in high-premium areas, this was the steepest change. They went from a premium capped at 8.5 percent of income to the full sticker price.

Gross premiums rose too. Insurers filed some of their largest increases in years for 2026, pointing to medical costs, specialty drugs and, in some filings, the expected loss of healthier customers once the credits ended. That last factor feeds itself. When subsidies shrink, the people most likely to drop coverage are the healthier ones who can gamble on going without, which leaves a sicker pool and pushes premiums higher for those who stay.

The combined effect was large. KFF, the health-policy research group, estimated before the lapse that subsidized enrollees' annual premium payments would more than double on average if the enhanced credits expired. The experience of any single family varied with age, income, location and plan, but the direction was the same nearly everywhere.

Who felt it most

The change hit three groups especially hard.

Older adults between their fifties and 65 felt it first, because their premiums are the highest and because they cluster near the cliff after decades of earnings. A household that retires early, or a couple where one spouse runs a small business, can easily earn just above the line and face a premium that consumes a large share of income.

Low-income workers in states without Medicaid expansion felt it next. For them, the zero-premium silver plan had been the only affordable coverage available. Once a monthly contribution returned, some dropped coverage, and some moved to bronze plans with lower premiums and much higher deductibles, which turns insurance into something that covers disasters and little else.

Rural families felt it in a particular way. Premiums in rural areas tend to be higher because there is less competition among insurers and hospitals. The benchmark price there is steeper, so losing the cap meant a bigger bill.

Walking through the cliff with round numbers

Abstract formulas hide how sharp the edge is, so take an illustration built on deliberately simple, made-up figures. They are not any real family's numbers, and the actual benchmark price where you live may be far higher or lower.

Suppose a 63-year-old couple in a rural county, where the benchmark silver plan for two people their age costs $24,000 a year. Suppose their income sits just above four times the poverty line for a household of two, which in recent years has meant somewhere in the high five figures.

Under the enhanced rules, their expected contribution was capped at 8.5 percent of income. If they earned $85,000, that cap came to about $7,200 a year, and the credit covered the remaining $16,800 of the benchmark. Their premium was large but manageable.

Under the original rules that returned in 2026, a couple at that income is past the cutoff. Their credit is zero. They owe the full $24,000, about 28 percent of their income, for the same plan from the same insurer. Had they earned a little less, enough to slip under the line, they would again qualify for help, and their expected contribution would be back near a tenth of their income.

That is why economists dislike cliffs regardless of their politics. A small raise, a good year for a small business, or a required withdrawal from a retirement account can cost a family far more in lost credit than it gains in income. Some households respond by managing their income to stay under the line, deferring work or shifting retirement withdrawals, which is a strange thing for a health law to encourage. A smooth phase-out, like the one the 2021 law created, avoids that trap.

The real question underneath

The fight over the enhanced credits was framed as a fight about subsidies. I think the better frame is this: what share of income should a working family have to spend on health insurance before the country decides that is too much? The original ACA answered with a sliding scale and a hard stop at 400 percent of poverty. The 2021 law answered with a ceiling of 8.5 percent of income for everyone. The difference between those answers is not a technical tweak. It is a judgment about whether the middle class buying its own insurance deserves the same protection against runaway premiums that the lower-income buyers get.

There is also a lesson about how progressives write laws. Putting an expiration date on a popular benefit to make the ten-year cost look smaller creates a cliff of its own, in time instead of income. Every few years, the benefit becomes a hostage in a budget fight. If a policy is worth having, the case for writing it permanently, and paying for it honestly, is strong.

What you can do

If you buy your own coverage, check your numbers before the next open enrollment rather than letting a plan renew automatically. The size of your credit depends on your projected income, your age and the benchmark price in your area, all of which can change. A licensed navigator can walk you through it for free, and many navigator programs run through local nonprofits; you can find health and legal aid groups by state in our organizations directory. This is general information, not advice for your situation; the official source for your options is your state marketplace or HealthCare.gov.

Tell Congress what the lapse meant for you. The new Congress sworn in on January 3, 2027 will decide whether to revisit premium help. Look up your U.S. representative and senators with our representatives tool, which also shows their recent votes, and describe your household's change in premium in a sentence or two. Specific stories travel further in a congressional office than form letters.

Watch the votes. Our Congress tracker follows the bills and roll calls, so you can see whether a premium bill moves and who votes for it. For the bigger picture of what families are paying across housing, food and care, our topic page on affordability and the cost of living gathers the reporting.

The couples on that imaginary street did nothing different from one year to the next. The rules around them changed. Rules can change back, and the people who write them answer to the people who write to them.