The ACA Subsidy Cliff Was Supposed to Save Money. The Data Suggests Otherwise. 

When Congress let the ACA’s enhanced premium tax credits expire at the end of 2025, the logic was straightforward: fewer people would qualify for subsidies, so the federal government would spend less. CBO’s own scoring backed that up — letting the enhancements lapse was projected to reduce the deficit relative to extending them. 2026 open enrollment and effectuation data tells a different story.

What the Numbers Show

CMS’s 2026 Open Enrollment Period data shows plan selections down 4.9% nationally versus 2025. February 2026 effectuation data shows effectuations down 12% from February 2025. Under the old assumption, that should mean a smaller subsidy bill. When we leverage Advance Premium Tax Credit (APTC) amounts from the open enrollment files and utilize February effectuation, APTC dollars are estimated to decrease by 2.9%, and the average annual subsidy per enrollee climbed 10.3% — from roughly $6,050 to $6,675 a year. When we limit to the 10 largest effectuation states, total APTC increases by 4%, total premium increases by 10% and overall enrollment decreases by 9%. In many states the government is paying more overall dollars for less people to be insured. 

Why This is Happening

Two forces are compounding. First, the people most likely to drop coverage when premiums jump are younger and healthier — they simply aren’t sick enough to justify the new sticker price. That leaves a costlier risk pool behind, and insurers have priced 2026 plans accordingly. Second, everyone who lost eligibility as the income limit on tax credits snapped back to 400% of the federal poverty line is now paying full premium — non-APTC premium dollars jumped from 18.6% to 24.9% of total marketplace premium nationally, in some states by 50-90%. For the enrollees who still qualify for a subsidy, the government is now subsidizing a pricier plan.

The State-Level Pattern Confirms It

Florida and Texas, which leaned heavily on enhanced subsidies to sign up enrollees at little or no premium, saw plan selections stay relatively flat while per-person subsidy costs jumped over 20%. States with sharper enrollment collapse, like North Carolina and South Carolina, muted the total dollar increase simply by losing more people — but even there, per-enrollee subsidy costs still rose by 6% and 12% respectively. Of the top 23 states by plan selections, only 5 had lower subsidy amounts per member.

The Takeaway

This analysis uses open enrollment effectuated enrollment data, so the final number may shift as enrollment fluctuates throughout the year. However, the early signal is clear enough to act on: policy aimed at trimming a subsidy line item can raise the average cost of what’s left, simply by changing who stays in the pool. 

The data suggests consumers are making different coverage decisions in a post-enhanced-subsidy environment. Product selection, metal mix, affordability, and the underlying risk pool all appear to be shifting. 

For health plans, success in 2027 won’t depend solely on forecasting enrollment, it will depend on understanding who remains in the market, what coverage they choose, and how those decisions affect pricing, product design, and long-term profitability.rwise.