# US universal healthcare affordability evidence dossier

Research checked 6 October 2026. Prepared for Getting It Right, a general-audience evidence website.

## Editorial conclusion

The owner's proposed claim, “Yes, we can afford universal healthcare,” is supportable as a qualified judgment about national economic capacity. It is not a settled quantitative finding about an unspecified program. Universal coverage can take several forms, and credible models produce both savings and additional costs. A proposal can lower national medical spending while requiring much more federal revenue. Neither national savings nor higher taxes alone establishes what a particular household would pay.

Recommended framing: **“Universal coverage is financially feasible. Savings depend on the design.”** Label the first sentence an evidence-based assessment, not the result of a single definitive experiment. A defensible narrower factual headline is **“Universal coverage does not have to increase total health spending.”** It describes credible modeled possibilities, not a guarantee.

A page that insists on the owner's exact wording needs an immediately visible qualifier: **“The U.S. has the resources to cover everyone, but costs, taxes, benefits, and access depend on how the system is built.”** Do not hide that sentence behind an accordion.

## Suggested short explanation

Universal coverage can use public or regulated private insurance. Models show possible U.S. savings, not a guarantee. Simpler billing and lower prices save money; additional care costs money. Tax-funded plans shift much of today’s private spending to government. The questions are the total bill, who pays, and access to care. [CBO overview](https://www.cbo.gov/publication/56898), [OECD coverage](https://www.oecd.org/en/publications/2025/11/health-at-a-glance-2025_a894f72e/full-report/population-coverage-for-healthcare_5c968b73.html).

## What the claim must define

- **Universal coverage:** who is eligible and enrolled, and for which services? Coverage for all residents differs from coverage limited to lawfully present residents. Insurance coverage is not the same as zero bills or timely access.
- **Single payer:** a financing architecture, not a synonym for universal coverage. Do not conflate either with government ownership of hospitals.
- **Affordable to whom:** the nation, the federal budget, states, employers, individual households, or providers?
- **Compared with what:** the same-year current-law baseline, not today's spending compared with a ten-year future total.
- **What benefits:** standard medical coverage differs materially from including comprehensive dental, vision, hearing, or long-term services and supports.
- **When:** startup/transition, first full year, and mature system outcomes are different questions.

CBO identifies multiple routes to near-universal coverage, including approaches retaining employer insurance and private plans. Its 2020 definition was close to 99% of citizens and lawfully present noncitizens; it did not claim 100% of all residents. All four approaches required additional federal receipts to be deficit-neutral. Reusing existing federal health subsidies alone would not fully cover their additional federal cost. [CBO, Policies to Achieve Near-Universal Health Insurance Coverage, October 2020](https://www.cbo.gov/publication/56666).

## Strongest supporting evidence

### The existing resource base is large

The latest historical CMS series reports U.S. health spending of **$5.3 trillion in 2024, $15,474 per person, or 18.0% of GDP**. This is evidence that the country already devotes enormous resources to health care, not proof that those funds can be effortlessly redirected or that any proposed benefit package fits within that amount. Use these historical figures rather than presenting projections as completed expenditure. [CMS historical National Health Expenditure Accounts](https://www.cms.gov/data-research/statistics-trends-and-reports/national-health-expenditure-data/historical).

### Real countries achieve broad coverage with different systems

OECD reports that most member countries provide universal or near-universal core coverage. The Netherlands and Switzerland use compulsory private insurance with subsidies and strong regulation. That refutes the premise that universal coverage requires one government insurer. These are real functioning systems, but they are not controlled experiments predicting U.S. transition costs. Benefit scope differs, and insurance eligibility by itself does not guarantee access. [OECD 2025 coverage chapter](https://www.oecd.org/en/publications/2025/11/health-at-a-glance-2025_a894f72e/full-report/population-coverage-for-healthcare_5c968b73.html).

For internationally comparable spending, OECD's U.S. note reports **$14,885 per person in purchasing-power-adjusted dollars versus a $5,967 OECD average**, and 17.2% versus 9.3% of GDP. Do not interchange these with CMS's 18.0% figure: the statistical series have different expenditure scope/methodology. Use one consistent series in a chart and retain its year/definition. [OECD Health at a Glance 2025, United States](https://www.oecd.org/en/publications/health-at-a-glance-2025_15a55280-en/united-states_3517f35e-en.html).

### CBO finds feasible lower-spending designs, with major tradeoffs

CBO's five illustrative single-payer options projected changes in **annual national health spending in 2030 from about $0.7 trillion lower to $0.3 trillion higher**, alongside **$1.5–$3.0 trillion more in annual federal health subsidies**. Those subsidies include relevant tax preferences, not just program outlays. Lower provider payments and payer administration are key savings mechanisms; increased use pushes costs up. CBO also projects more care delivered but even greater demand, resulting in more unmet demand and congestion. These findings are conditional scenario estimates, not a score of every universal plan or a promise of savings. [CBO director's explanation, December 2020](https://www.cbo.gov/publication/56898).

### Peer-reviewed synthesis favors savings but is not proof

Cai and colleagues reviewed 22 single-payer plans from 18 studies published in 1991–2018, including eight national and fourteen state-level plans. Nineteen projected first-year savings; the median was 3.5%. This supports the existence of plausible saving designs. Important limits: the studies differ in rigor, assumptions, benefits, and political/funding contexts; there was no formal meta-analysis or study-quality scoring; U.S.-wide implementation had no domestic empirical test. A count of favorable models is not 19 independent experimental replications, and it excludes later CBO/RAND/Urban work. The paper discloses four small research grants from Physicians for a National Health Program and one author’s board role in its student organization; use transparent methods and independent corroboration rather than either dismissing or canonizing the review. [Cai et al., PLOS Medicine, 2020, full article and limitations](https://journals.plos.org/plosmedicine/article?id=10.1371/journal.pmed.1003013).

## Strongest counterevidence and qualifications

### Broader coverage and benefits can increase national spending

Urban Institute's 2019 comparison projected its enhanced single-payer option, including all residents and extensive benefits without cost-sharing, would add approximately **$720 billion in national spending in 2020**, with **$34 trillion in additional federal spending over 2020–2029**. Its narrower single-payer option reduced national spending by 6%, but retained income-related cost-sharing and excluded undocumented residents. This is a valuable same-research-team demonstration of design dependence, not a contest between inherently pro- and anti-universal researchers. The lower-cost option must not be described as covering literally everyone. [Urban Institute report summary and linked complete study](https://www.urban.org/research/publication/incremental-comprehensive-health-reform-how-various-reform-options-compare-coverage-and-costs).

For the enhanced option, Urban explains the ten-year accounting as **$34T more federal spending offset by $27T less household/employer/state spending, leaving $7T more total national spending**. That old scenario is not a current bill's score. Its lesson is accounting: shifting who writes the check is not the same as changing the total bill. [Urban's federal-versus-national spending explanation](https://www.urban.org/urban-wire/dont-confuse-changes-federal-health-spending-national-health-spending).

### Lower modeled spending can partly reflect unavailable care

RAND's 2019 estimate for comprehensive national single payer with long-term care and no cost sharing was **1.8% higher national spending**. Crucially, that scenario assumed half the new demand was unmet because of supply constraints. Without that constraint, RAND estimated **9.8% higher spending**. The report used aggregate extrapolation from previous research rather than a new national microsimulation. This is a particularly useful caution against treating every low expenditure estimate as successful delivery of all desired care. [Liu and Eibner, RAND, National Health Spending Estimates Under Medicare for All, April 2019](https://www.rand.org/pubs/research_reports/RR3106.html).

### Startup costs are not included in the headline CBO range

CBO's 2020 technical report uses its 2020 baseline and focuses on 2030 after assumed implementation. Section 17, printed pages 165–168, explicitly excludes substantial startup expenditures: payment/claims systems, eligibility and enrollment, winding down programs, and possible support for displaced administrative workers. It also discusses implementation risks and phasing payment changes. Sections 9 and 19 address uncertainty and differences across estimates. Do not multiply the 2030 annual saving by ten and label the result a ten-year reform score. Its full transition and financing cannot be inferred from the headline annual comparison. [CBO Working Paper 2020-08, full PDF](https://www.cbo.gov/system/files/2020-12/56811-Single-Payer.pdf).

### Financing has real economic effects

CBO's 2022 working paper separates the effects of single payer from its financing. Favorable GDP results before financing must not be presented as the complete economic outcome. Its approximate combination of selected payroll/income-tax financing scenarios with the lowest/highest subsidy options yields GDP **about 1%–10% below baseline in 2030**. The paper emphasizes that those tax designs are only a subset of alternatives; other funding choices could differ substantially. Private nonhealth consumption outcomes also vary. These are model-dependent comparisons, not evidence universal coverage is categorically unaffordable, nor a complete measure of health benefits or welfare. [CBO, Economic Effects of Five Illustrative Single-Payer Health Care Systems, 2022, financing discussion and Table 9](https://www.cbo.gov/system/files/2022-02/57637-Single-Payer-Systems.pdf).

### Household savings require a tax specification

A 2024 JAMA Health Forum microsimulation counts premiums, out-of-pocket spending, and taxes. Under its single-payer income/payroll-tax scenario, mean health payments fell from about 27% to 15% of compensation in the lowest-income group and rose from about 27% to 31% in the highest-income group. Within every income group some households paid more and some less. Assumptions include employer premium incidence on workers; different tax schedules produce different results. This supports saying financing can reduce lower-income burdens, not promising every family a saving. [Liu et al., Household Health Care Payments Under Rate Setting, Spending Growth Target, and Single-Payer Policies, 2024](https://pmc.ncbi.nlm.nih.gov/articles/PMC11215555/).

## How to explain the disagreement honestly

The important disputes are about attainable administrative savings, how providers respond to lower prices, how much utilization grows, which services are added, drug-pricing leverage, and how the transition and taxes work. They are not resolved merely by choosing the lowest headline estimate.

Some additional use meets previously unmet needs; calling all of it waste prejudges the benefits. Conversely, lowering provider payments is not free in economic terms: monitor whether safety-net, rural, and other vulnerable providers can maintain services. Longer waits and financial barriers are both relevant access problems; neither should be omitted because it complicates a preferred narrative. CBO's design review recommends attention to phase-in, provider capacity, and disruption to insurance workers. [CBO 2019 design report](https://www.cbo.gov/publication/55150).

## Publication guardrails

1. Do not say “CBO proves Medicare for All saves money.” Specify illustrative designs, baseline vintage, year, and range.
2. Do not describe a $30T-plus increase in federal spending as an equal increase in national medical costs.
3. Do not claim that removing premiums makes care free or guarantees everyone lower total payments.
4. Do not equate insurance coverage with immediate appointments, all services, or no cost sharing.
5. Do not use overseas spending gaps as an exact savings calculator for the United States.
6. Do not treat older 2019–2020 dollar estimates as current legislative scores. Keep them labeled as scenarios and update for an actual proposal.
7. Do not say prevention necessarily pays for the whole expansion. A health benefit can be valuable without reducing spending.
8. Do not infer that increased medical spending necessarily makes a policy bad: additional services may have value. Conversely, savings alone do not establish quality or fairness.

## Editorial confidence

High confidence: universal coverage and single payer differ; national and federal spending differ; benefit/payment/financing design changes costs; a universal system need not cost more than the current system in credible models.

Moderate confidence: a carefully designed U.S. reform could attain broader coverage with similar or lower national spending. This rests on model assumptions, international experience, and mechanisms, rather than a completed U.S. national transition.

Not established without a concrete proposal: exact savings, tax rates, household winners and losers, startup budget, provider viability, waits, long-run innovation effects, and net macroeconomic/welfare outcome.

The strongest page invites the reader to inspect the conditions under which the claim holds. It should not present affordability as settled for every version of universal healthcare.