ACA Enrollment Falls as Premiums Rise and Subsidies End

Enrollment in Affordable Care Act marketplace plans has entered a sustained contraction after the enhanced premium subsidies expired at the start of the year, and the arithmetic is unforgiving.

Roughly 19.2 million people held coverage as of February, a decline of 2.6 million from the prior year, and federal data and state reports indicate the slide continued through the summer and autumn.

The mechanism is straightforward: when subsidies lapse, net premiums rise, and households with thin margins respond by exiting the risk pool rather than absorbing the cost.

The human ledger behind those figures is severe. Holly Hendricks, 49, of Sarasota, Florida, lives with lupus and rheumatoid arthritis; when her monthly premium climbed to nearly $1,100, she dropped her plan in April and became uninsured.

She lost access to specialists, blood work, infusions and wound care, then developed sepsis in September and spent a month hospitalized, with a bill she expects may exceed $500,000. Her case illustrates how premium shocks convert into deferred care, acute crises and uncompensated hospital costs.

This analysis examines the enrollment data, the state-level variation and the political economy now shaping the individual market. It weighs the evidence from KFF, the State Marketplace Network and state insurance regulators, and it separates what the numbers demonstrate from what partisans on both sides assert.

The central finding is that the subsidy expiration did not distribute its effects evenly: declines concentrated in lower-income regions and in Republican-held districts that account for 55% of enrollees but 67% of the losses.

TL;DR Enhanced ACA premium subsidies expired at the start of the year, and enrollment has fallen sharply since: about 19.2 million people were covered in February, 2.6 million fewer than a year earlier, with KFF estimating the individual market shrank by more than 4.6 million by June. Eighteen state-run exchanges reported accelerating declines, with disenrollments up nearly 34% as of May year over year. Mississippi's enrollment collapsed from 338,000 at the end of 2025 to roughly 125,000 in September. Losses concentrated in Republican-held districts, which hold 55% of enrollees but 67% of the decline, and affordability now dominates midterm election messaging.

The Subsidy Cliff and the Mechanics of Coverage Loss

The enhanced subsidies were never a permanent feature of the Affordable Care Act; they were a pandemic-era augmentation to the advance premium tax credit, and their expiration restored the older subsidy schedule.

That restoration raised net premiums most sharply for middle-income households who earned too much for Medicaid but too little to absorb full-price coverage. The Congressional Budget Office has consistently estimated that premium tax credits reduce uninsurance, so the direction of the enrollment effect was predictable even before the data arrived.

What distinguishes this contraction is its velocity. Cynthia Cox, director of the Program on the ACA at KFF, observed that households which stretched to keep coverage early in the year may simply be unable to sustain it for twelve months.

That pattern produces a staggered attrition rather than a single cliff: people renew in January, ration through spring, and disenroll by summer. The result is a risk pool that deteriorates as healthier and poorer members exit first.

Premium Sensitivity and Household Budgets

Health insurance demand is price-elastic at the lower end of the income distribution, and the evidence here confirms it. A monthly premium near $1,100 consumes roughly a third of median gross pay for a full-time caretaker, which places coverage in direct competition with rent, utilities and food.

When a household must choose, insurance loses. Hendricks stated plainly that basic needs come first, and that ordering is rational rather than reckless.

The downstream costs do not disappear; they migrate. A single sepsis hospitalization projected above $500,000 is absorbed by hospital charity care, state uncompensated-care pools and, ultimately, premium increases for everyone still insured. Economists describe this as cost-shifting, and it creates a feedback loop: disenrollment raises premiums, which drives further disenrollment. The individual market is small enough that this spiral can move aggregate rates quickly.

State Exchanges as the Leading Indicator

The eighteen states operating their own marketplaces function as a laboratory, because they publish monthly enrollment data that the federal platform does not. The State Marketplace Network reported disenrollments up nearly 34% as of May compared with the same period a year earlier, with more people dropping coverage and fewer entering through special enrollment periods. That combination is the signature of affordability-driven exit rather than administrative churn.

KFF's estimate that the individual market shed more than 4.6 million people by June is larger than the February exchange figure because it captures off-exchange plans and later attrition. The two datasets are consistent in direction and roughly consistent in magnitude once timing is accounted for.

Independent verification from state regulators strengthens confidence that this is a real demand shock, not a reporting artifact.

Why the February Figure Understates the Damage

February is an early reading, and early readings systematically understate attrition when subsidies expire in January. Households that renewed at the start of the year had already paid for coverage and face switching costs, so they hold on until a premium notice or a medical event forces a decision.

Cox's observation that early retainers may not last a full year is therefore an analytical warning, not a hedge.

The June KFF estimate of a 4.6 million decline is roughly 77% larger than the February exchange drop, which is a plausible gap once off-exchange plans and mid-year attrition are included. Analysts should treat the February number as a floor and the June number as closer to the true demand response. Subsequent state reports, including Mississippi's autumn stabilization near 135,000, suggest the curve is flattening rather than reversing.

Geographic Concentration and the Political Economy of Premiums

The distributional pattern is the most politically consequential finding in the dataset. KFF determined that Republican-held congressional districts account for 55% of enrollees but 67% of the enrollment declines between 2025 and 2026.

That asymmetry is not evidence of intent; it reflects the underlying income geography of the ACA risk pool, since those districts contain more lower-income residents for whom a premium increase is decisive.

Cox's explanation is economic rather than partisan: households with less disposable income find any premium increase harder to absorb, and the enhanced subsidies were structured to compress that burden. Remove them, and the burden returns in proportion to income.

The result is that the policy's withdrawal is felt most acutely in the constituencies whose representatives voted for it, a dynamic that carries obvious electoral risk.

Mississippi and Massachusetts: Two Divergent Cases

Mississippi offers the clearest illustration of scale. The state had 338,000 residents enrolled at the end of 2025; that fell below 192,000 by February and to roughly 125,000 in September, before recovering to about 135,000 in late October, according to Insurance Commissioner Mike Chaney.

A decline of that magnitude in a single state implies a collapse of the individual market's risk pool, not merely a trimming of marginal enrollees.

Massachusetts provides the contrasting signal. The state recorded its first in-year decline outside the Covid-19 pandemic period, with about 8,000 fewer people enrolled in September than in February, according to exchange executive director Audrey Morse Gasteier.

Massachusetts has a comparatively high-income, heavily subsidized and historically stable market, so a decline there indicates that the shock is broad rather than confined to low-income states.

Congressional Districts and the Subsidy Vote

The district-level asymmetry carries a specific legislative history. The enhanced subsidies lapsed, and the One Big Beautiful Bill Act cut federal Medicaid funding by roughly $1 trillion, according to Democratic critics who are campaigning on both facts.

Whether voters assign responsibility to the subsidy expiration, the Medicaid reduction or broader inflation is an open empirical question that November will resolve.

Healthcare affordability has become a leading midterm issue precisely because it is legible to households in a way that abstract fiscal debates are not. A premium notice is a concrete, dated document.

The political risk for incumbents is that the 67% concentration of losses falls inside their own districts, and the counterargument that subsidies were fiscally unsustainable requires voters to weigh future savings against present medical bills.

Open Enrollment, Pricing Signals and the Road Ahead

Open enrollment begins Sunday in most states, and insurers have already filed rates reflecting a sicker, smaller risk pool. Premiums are expected to rise again, which means the affordability pressure that drove the 2026 contraction will persist into 2027 rather than ease.

Actuarial logic is unforgiving here: when lower-cost members exit, the remaining pool's per-member claims rise, and those costs are recovered through rates.

The policy menu is narrow and well understood. Congress could restore the enhanced subsidies, extend a transitional reinsurance program, or allow states to expand public options; each carries a measurable fiscal cost and a measurable coverage gain. The CBO's prior scoring of premium tax credits provides the template for estimating the trade-off. Absent intervention, the baseline is continued attrition toward a smaller, more expensive individual market.

Reading the Rate Filings

Rate filings are the most reliable forward indicator available before enrollment data arrives. Carriers that assumed a stable pool in 2026 will have repriced for adverse selection in 2027, and the magnitude of those requests reveals how much morbidity deterioration they observed. Large requested increases in states with steep declines, such as Mississippi, would confirm the spiral hypothesis.

Analysts should also watch special enrollment period volumes, which the State Marketplace Network tracks. Special enrollment is where newly uninsured people re-enter after a health shock, so a persistent decline in those applications signals that people are not merely delaying coverage but abandoning it. The May data already showed fewer sign-ups alongside more disenrollments, which is the more troubling combination.

What the 2027 Enrollment Season Will Reveal

The coming open enrollment period is the cleanest natural experiment available. If enrollment stabilizes near current levels, the contraction was a one-time repricing adjustment; if it falls again, the adverse-selection spiral is active.

The distinction matters enormously for policy, because a one-time adjustment can be addressed with transitional support while a spiral requires structural intervention.

Three indicators will settle the question: net enrollment change during open enrollment, the share of returning versus new enrollees, and the average age of the pool. Rising average age alongside falling enrollment is the definitive signature of adverse selection.

KFF and the state exchanges will publish these figures on a lag, so the first reliable reading will likely arrive in the spring.

Assessing the Evidence and the Cost of Inaction

The evidence assembled here supports a narrow but firm conclusion: the expiration of enhanced subsidies produced a large, rapid and geographically concentrated decline in individual-market coverage. The February figure of 19.2 million, the 34% rise in disenrollments across eighteen state exchanges, and KFF's 4.6 million individual-market decline all point the same direction. Independent sources with different methodologies rarely converge this cleanly by accident.

The cost of inaction is measurable in both fiscal and human terms. Uncompensated hospital care for cases like Hendricks's projected $500,000 bill is ultimately financed through cost-shifting, public programs and charity care, so the savings from letting subsidies lapse are partly illusory.

The Congressional Budget Office's scoring conventions make this trade-off explicit: premium tax credits carry a budget cost and a coverage benefit, and the two must be weighed together rather than separately.

Limits of the Current Data

Several caveats deserve emphasis. The February exchange count and the June KFF estimate measure different populations, so they should not be subtracted from one another. State reports vary in frequency and definition, and Mississippi's autumn rebound from 125,000 to 135,000 shows that enrollment is not monotonic. Single anecdotes, however vivid, cannot establish aggregate causation on their own.

What the data cannot yet show is the health consequence distribution: how many of the newly uninsured will experience a preventable hospitalization, and at what public cost. Hendricks's sepsis case is illustrative rather than representative, but it is consistent with established research linking uninsurance to delayed care and worse outcomes.

That literature is the strongest reason to treat enrollment decline as a health outcome, not merely an insurance statistic.

Implications for Households and Policymakers

For households, the practical implication is that coverage decisions made in January may not survive the year, and that lapse carries real medical risk. For policymakers, the implication is that subsidy design determines risk-pool composition, and that a lapse is not a neutral budget event. The 2026 data provide a quantified baseline against which any future intervention can be judged.

The broader lesson concerns how health insurance markets behave at the margin. A premium increase of a few hundred dollars is trivial to a high-income household and decisive for a caretaker earning modest wages, so aggregate policy changes produce sharply unequal effects.

That asymmetry, visible in the 55% versus 67% district split, is the central fact of this enrollment cycle and the one most likely to shape the political response.