In 2026, ACA plan selections declined 5% compared to 2025. This period saw the expiration of subsidy enhancements introduced in 2021 and the removal of improper enrollments, accompanied by changes to state platforms. The Commonwealth Fund reports a CBO projection showing that the number of uninsured people would be 16 million higher in 2034. Coverage disruptions are associated with difficulties accessing care, but it is not established that worsening health conditions make returning to coverage itself less likely.
The essentials
- The United States is actively reducing its health coverage through legislative choice, at a time when international comparisons show that universal coverage costs less per capita.
- The Commonwealth Fund projects 17 million additional uninsured Americans by 2034; ACA selections fell 5% in 2026 according to federal agency data.
- Canada spends approximately half as much per capita as the United States for universal coverage, according to the same source.
- Uninsurance is associated with more delayed care and the absence of a usual source of care, which can worsen health and increase long-term costs.
- No comparable country has followed this trajectory: the United States constitutes an unintended experiment whose results will be measurable by 2030.
17% less in one year, without economic crisis
In the United States in 2026, Congress did not renew the subsidy enhancements introduced in 2021, and several states tightened enrollment conditions for Medicaid. Millions of households with modest incomes, eligible under the old rules, thus fell out of the system. The observed 5% decline in ACA selections in 2026 was accompanied by removal of improper enrollments and changes to state platforms.
This point deserves to be stated clearly because it changes the frame of reference. Health coverage in the United States is often discussed as an old structural problem, almost geological in nature. But the 2026 trajectory is not geological. It is recent, deliberate, and reversible if political will changes. What makes it difficult to analyze calmly is that it occurs in a country where the health care debate has been an ideological battleground for decades, not a terrain for evaluation.
The Commonwealth Fund, an independent organization that has been comparing rich countries’ health systems for thirty years, documents this trajectory in its May 2026 report. Its figures are based on enrollment data from federal agencies and established demographic trends. The CBO then projected 16 million additional uninsured people in 2034, a variation in the number of uninsured people at that date, not a cumulative total over nine years.
International comparisons
The Commonwealth Fund publishes an annual dashboard of health systems in rich countries. Since 2004, the United States has been ranked last in each edition of Mirror, Mirror; the 2024 edition ranks it last overall among 10 countries across five domains, including administrative efficiency. This persistent position says something specific: the United States spends the most and achieves the worst results in terms of access, equity, and health outcomes for the entire population.
The most direct comparison point is Canada, which has a set of provincial and territorial single-payer public regimes for medically necessary care, covering eligible residents. The cost per capita is approximately half as high as in the United States. The spending gaps cannot be attributed simply to lower quality of Canadian care. Canadian life expectancy exceeds American life expectancy, and infant mortality is lower. The structure of financing and organization is an important factor among several determinants of these gaps: a unified system negotiates prices, eliminates management costs between multiple insurers, and excludes no one.
This result is counterintuitive for those who think competition between private insurers should produce efficiency. In practice, fragmentation of the American market generates administrative costs that absorb between 25% and 30% of total spending, according to estimates from the New England Journal of Medicine. An insurer covering an entire population does not need to manage processes for selecting enrollees, contesting reimbursements, or competitive marketing. These are real costs that the universal system does not incur.
The United Kingdom, Germany, and France, each with different architectures—public for one, Bismarckian for the others—converge on the same finding: universal coverage costs less per capita than the American system, with better aggregate results. The question of what non-coverage really costs also arises in France, in a different register: when supplementary coverage becomes a discriminatory fixed cost, it creates access barriers that the basic regime does not entirely correct.
Uninsurance is not a stable state
Losing health coverage does not merely create a financial risk in case of illness. It measurably changes health behaviors. Data from the Centers for Disease Control show that uninsured Americans systematically delay preventive consultations, screenings, and monitoring of chronic diseases. They enter the health care system later, with more advanced conditions.
The mechanism is documented: untreated diabetes for two years costs three to five times more to treat than diabetes managed from diagnosis. Untreated hypertension increases the probability of a stroke, whose treatment costs are incommensurate with those of preventive medication. These effects accumulate. A population that goes several years without coverage develops a heavier disease burden than one that never had access: they experienced care, they know what they need, they can no longer access it.
This mechanism is what makes uninsurance structurally different from initial non-insurance. When a country like the United States reduces coverage after expanding it, it does not return to a previous state. It creates a new population: people who benefited from regular care, who have diagnosed and stabilized chronic conditions, and who find themselves without access to the treatments that enabled that stabilization. This group is both health-wise and economically more fragile than one that never had access.
Mental health illustrates this point particularly starkly: an interruption of psychiatric care produces relapses whose treatment is longer and more costly than the continuity of care that would have needed to be funded. The discontinuity itself has a price.
Who governs this choice, and with what interests
The reduction in American health coverage is not a budgetary accident. It results from a precise political architecture. Private insurers have a direct interest in selecting the least risky populations. Health care providers in a fragmented system charge at rates that only an unregulated market tolerates. States that refuse Medicaid expansion—several in the South and Midwest—do so on fiscal doctrine grounds, not for lack of financial capacity.
This picture does not reduce to a coordinated conspiracy. It describes a coalition of reinforcing interests that finds in ideological debate over the role of the state a political wrapper. The concrete result: millions of people with modest incomes, often employed part-time or in sectors without employer coverage, find themselves exposed to catastrophic health care costs as soon as a serious medical problem occurs.
KFF reported in 2014, based on 2007 data, that medical bills constituted a major cause of personal bankruptcy. This figure has no equivalent in countries with universal coverage. In Canada, the United Kingdom, Germany, and France, serious illness can put a family in financial difficulty. It does not make them homeless because they saw a doctor.
Actors working against this trend exist. Organizations like Community Catalyst and Families USA document the effects of uninsurance and advocate for extension policies at the state level. Several states—California, Colorado, Massachusetts—maintain or expand coverage programs that go beyond the federal floor. These state experiments are useful laboratories. They show that political will is sufficient where resources are not lacking.
The unintended experiment the world will watch
As of 2026, the United States constitutes a natural experiment whose protocol no one commissioned. The United States is an important case of coverage decline after the ACA, but not a historically singular case of deliberate coverage reduction. The health, economic, and political effects will be measurable by 2030-2034 with the tools epidemiologists and health economists have at their disposal.
What is already observable gives an idea of the trajectory. The years following implementation of the ACA in 2014 saw preventable mortality rates fall in states that adopted Medicaid expansion, compared to states that refused it. This divergence is documented in several studies published in the New England Journal of Medicine and JAMA. It provides a basis for comparison: the same methods will allow measuring what uninsurance produces in the opposite direction.
Pandemic preparedness offers another angle of analysis: fragmented health systems showed reduced capacity to mobilize a coordinated response during the Covid-19 crisis. The correlation between universal coverage and epidemic resilience is a dimension that post-2020 data are beginning to document seriously.
The Commonwealth Fund’s 2026 report raises a question that extends beyond the American case: at what point does the cost of uninsurance—delayed care, worsened conditions, overwhelmed emergency rooms, individual bankruptcies, lost productivity—exceed the cost of extending coverage? The answer for the United States will be readable in data that federal agencies will produce between 2028 and 2034. It will feed debates on health coverage in countries seeking to contain their health spending by reducing their systems.
What universal coverage actually requires
The American debate on health coverage is often presented as a debate about values: solidarity against individual responsibility. This presentation masks a more prosaic reality. Countries that chose universal coverage did so for economic reasons as much as ethical ones.
A healthy workforce is more productive. Better coverage promotes certain screenings, diagnoses, and access to care, but it does not allow one to assert in general that diseases will be treated at lower cost. A system that does not select enrollees based on risk reduces the massive administrative costs that such selection generates. These effects are measurable, and they have been measured.
Transition to a universal system requires difficult political trade-offs. It calls into question powerful organized interests—insurers, providers, segments of the pharmaceutical industry whose pricing model relies on market fragmentation. These obstacles are real. They explain in part why the United States has not taken this path. But they do not explain the magnitude of the cost that the current trajectory imposes on millions of people whose health depends on choices in which they do not participate.
From 2027 onward, Commonwealth Fund data, updated annually, will allow measuring the level of health degradation that states, insurers, and Congress accept before political rebalancing becomes necessary.
Sources
- Commonwealth Fund, U.S. Health Care from a Global Perspective, 2026
- Centers for Disease Control and Prevention, National Health Interview Survey, data on access to care (without coverage)
- Kaiser Family Foundation, Employer Health Benefits Survey; data on medical bankruptcies
- New England Journal of Medicine, studies on the impact of Medicaid expansion (2014-2019) on preventable mortality
- New England Journal of Medicine, Woolhandler & Himmelstein, estimates of administrative costs of the American health care system
