Insurance Coverage Does Not Guarantee Timely, Effective Care

Scott Atlas frames the central error in health-policy debate as a confusion between coverage and care. The goal, he argues, should be access to high-quality health care for everyone—not merely the designation that someone is insured.

That distinction anchors his critique of single-payer proposals. Atlas acknowledges that the United States spends much more on health care than peer countries: an OECD comparison he cites puts U.S. current health spending at 16.9% of GDP in 2018, compared with 12.2% in Switzerland, 11.2% in Germany and France, 10.9% in Sweden and Japan, and 9.8% in the United Kingdom. But spending alone, he says, is not an adequate measure of a system’s performance. The relevant question is what patients receive for that money: speed, capacity, treatment availability, and outcomes.

Atlas describes American health care as the global standard of excellence on those measures. He points to what he says are superior U.S. cancer-survival rates, chronic-disease outcomes, screening use, access to new drugs, diagnostic technology, surgery, and specialist physicians. He also emphasizes the country’s role as the leading source of medical innovation.

The contrast is directed at claims that universal public insurance itself guarantees care. Atlas points to the 2023 Medicare for All Act, S. 1655/H.R. 3421, which would expand Medicare to all U.S. residents, eliminate premiums, deductibles, and copays, and prohibit private insurers and employers from offering coverage that duplicates the program’s benefits. He characterizes the proposal as government control over every phase of care and stresses its explicit ban on duplicative private insurance.

Atlas invokes Senator Bernie Sanders’s call for single-payer health care to end what Sanders calls the United States’ failure to “guarantee healthcare,” and Senator Elizabeth Warren’s statement that Medicare for All would guarantee care “when they need it.” In his telling, those claims rest on a policy myth: that insurance coverage and access are interchangeable. He uses René Magritte’s Ceci n’est pas une pipe—a painting of a pipe rather than a pipe itself—as the analogy. The label is not the underlying thing.

Atlas’s explanation for the lower costs of single-payer systems is straightforward: they limit the use of medical care. In practice, he says, that means constraints on access to doctors, treatments, drugs, and technology.

Wait times are his principal example. The U.S. and Canadian numbers he presents do not measure identical points in a patient’s path: U.S. figures measure waits for a new-patient appointment, while the Canadian figures run from GP referral to treatment. Atlas’s comparison is therefore not a like-for-like measurement. His point is that formal coverage does not establish whether a patient can obtain care on a medically useful timetable.

Across Canadian specialties, Atlas cites a weighted median of 27.8 weeks from referral to treatment, including 37.4 weeks for gynecology and 43.5 weeks for neurosurgery. He argues that even routine U.S. appointments are available more quickly than treatment for many seriously ill patients in Canada.

He makes the same argument about England’s National Health Service. Atlas cites NHS England figures from March and April 2024 showing 7.6 million people on waiting lists, more than 3 million waiting over four months after referral for treatment, and more than 314,000 waiting over a year. He says 57.1% of general-surgery patients, 76.2% of psychiatry patients, and 57.5% of neurosurgery patients had waited longer than four months. He also cites a figure that 41% of patients referred for urgent cancer treatment waited more than two months to begin treatment.

Atlas objects not only to the queues but to the standards used to judge them. He says the NHS has treated a wait of no more than two months for first treatment after an urgent cancer referral as its target, with an 85% performance benchmark. In his view, that is already a low bar for cancer care. Nor does he accept the characterization of the NHS as free: he cites a £152.6 billion budget for 2022–23, mainly financed through general taxation.

Capacity, detection, and timely treatment form the chain Atlas uses to explain the outcomes he credits to the United States. He says U.S. critical-care capacity per 100,000 people, including per 100,000 people over 65, exceeds that of the United Kingdom, France, Italy, and Germany. He makes a similar point about CT and MRI availability, highlighting the United Kingdom and Canada as having less access to equipment he regards as essential for serious illness and cancer care.

He uses hypertension to distinguish whether people are diagnosed, treated, and successfully controlled. An article reporting that 53% of diagnosed Americans receive treatment was portrayed, he says, as evidence of a dysfunctional system. But he says treatment rates in England, Sweden, Germany, Spain, Italy, and Canada are also around half. For Atlas, control among treated patients is the consequential measure, and he says the United States performs better than those countries. He attributes that result to more aggressive diagnosis and drug treatment.

Cancer is his larger test case. Atlas says the United States has statistically significantly better outcomes than Western Europe across common cancer types, alongside more extensive screening than Canada. He cites a claim that more than 44,000 additional Canadian women died over 16 years while waiting for health care. He also argues that innovation matters because cancer treatment cannot be deferred until a drug becomes available years later: of 54 new cancer drugs in the comparison he presents, U.S. patients had access to 51, compared with 39 in Germany, 38 in the United Kingdom, 23 in France, and 14 in South Korea.

For Atlas, insurance expansion is not meaningless, but it is not the endpoint. A health system has to be judged by whether people can see clinicians, obtain tests and treatment, and receive effective care before delay changes the result.

Atlas treats the collapse in trust during COVID as relevant to current health-policy claims because, in his account, public institutions asked for deference while misdescribing both evidence and outcomes. Trust in the CDC and FDA, he says, fell from roughly two-thirds of people rating them excellent or good before the pandemic to 35% to 40% afterward. Trust in doctors and hospitals fell, he says, from 71.5% saying they trusted them “a lot” at the beginning of the pandemic to 40% today.

Scott Atlas also says confidence in science itself has become newly partisan. For decades, he argues, there was no significant partisan difference in the share of Americans expressing a great deal of confidence in the scientific community. Now there is a large split. Atlas regards that loss of confidence as dangerous, while warning against the opposite error: dismissing legitimate expertise and treating unsupported theories as equivalent to scientific evidence.

His explanation for the decline in trust is unequivocal. Atlas says claims used to justify COVID restrictions—including claims about universal risk, asymptomatic spread, lockdowns, school closures, travel restrictions, masks, and vaccine-only immunity—were false and known to be false by spring 2020.

He says prior pandemic planning supported a different strategy: ask sick people to isolate while allowing others to continue working, rather than broadly closing schools, canceling large gatherings, quarantining groups, or restricting travel. His preferred alternative was “targeted protection”: increase protection for people at high risk of severe outcomes while reopening society and avoiding the wider harms imposed on lower-risk people. Atlas identifies Stanford epidemiologist John Ioannidis, Yale public-health figure David Katz, Harvard epidemiologist Martin Kulldorff, and himself as advocates of versions of that position in 2020.

The line responds to the charge that opposing lockdowns meant choosing the economy over lives. Atlas argues instead that missed care, unemployment, isolation, and school closures carried their own health consequences, especially for poorer people. He says he and colleagues calculated in spring 2020 that missed health care and unemployment caused solely by lockdowns were producing at least 700,000 lost years of life each month in the United States, more than the virus in their calculation.

Atlas cites a series of studies as evidence that lockdowns did not achieve their stated purpose. Christian Bjørnskov’s analysis of weekly mortality in 24 European countries found, in the quoted conclusion on Atlas’s slide, that more severe lockdown policies were not associated with lower mortality. Eran Bendavid and colleagues found no significant benefits on case growth from more restrictive non-pharmaceutical interventions in the countries they evaluated. Atlas also cites an NBER working paper by Agrawal and colleagues that found shelter-in-place policies associated with increased excess mortality, and a 2022 meta-analysis by Jonas Herby and colleagues concluding that lockdowns had little to no effect on COVID-19 mortality while imposing substantial economic and social costs.

Sweden is Atlas’s international counterexample. He says the country was widely criticized for avoiding lockdown orders, yet an analysis by Ioannidis found Sweden had the lowest cumulative excess mortality among OECD countries. Atlas also cites Ioannidis’s retrospective estimate that the United States would have had 1.6 million fewer deaths by early 2023 if it had matched Sweden’s performance.

He argues that censorship compounded the policy failure. Censorship, Atlas says, does not merely silence a speaker; it prevents others from hearing alternatives and creates a false appearance of expert consensus. He says he received hundreds of messages from people around the country, including at Stanford and the NIH, who agreed with his position privately but feared professional or personal consequences if they spoke publicly.

Atlas’s case against lockdowns rests not only on their claimed lack of effect on mortality and transmission, but on what he describes as their unequal collateral damage. Scott Atlas cites an NBER paper by Francesco Bianchi and colleagues estimating that unemployment shocks alone would cause between 0.89 million and 1.37 million additional U.S. deaths over 15 years, disproportionately affecting Black Americans and women.

He says healthy children faced a minuscule risk of serious illness from COVID, while school closures cut them off from instruction, meals, social development, physical activity, and opportunities to identify hearing and vision problems. Online education was not merely a question of whether a household had a computer, he argues. Atlas points to learning losses in the Netherlands despite widespread household broadband access and maintains that school supplies social, nutritional, and developmental functions that a remote lesson cannot replicate.

Atlas also cites CDC findings, as he describes them, that mental-health conditions among young adults had risen sharply by June 2020, including anxiety, depression, and suicidal ideation. By that point, he says, 25% of college-age people had thought about suicide, which he attributes to isolation rather than the virus.

A global-learning update from UNESCO, the World Bank, UNICEF, FCDO, USAID, and the Bill & Melinda Gates Foundation found that learning losses were especially large where schools were closed the longest and disproportionately concentrated among disadvantaged students. Atlas calls lockdowns “a luxury of the rich,” contrasting the ability of affluent families to create neighborhood learning groups with the circumstances of families who could not.

Atlas’s accounting extends beyond infections and deaths directly attributed to COVID. It includes missed medical care, unemployment, isolation, lost learning, and the distribution of those harms. The legacy he identifies is avoidable death among vulnerable people, destruction of low-income families, ongoing damage to children, and diminished trust in public health and science.