There is a version of the American health spending story that almost everyone has heard. The United States spends far more than other rich countries because Americans go to the doctor more, demand more tests, get more procedures and expect more of everything. Cut the overuse, the argument goes, and the bill comes down.
It is an intuitive story. It fits a certain picture of American consumer culture. And as an explanation for the gap between the US and its peers, it is mostly wrong.
This matters well beyond the United States. Health ministries everywhere are told that controlling costs means controlling demand: fewer visits, fewer admissions, stricter gatekeeping. Sometimes that is true. But the best-known international comparison suggests that the biggest lever in the most expensive system on earth was never volume. It was price.
The myth
The claim: US health spending is high mainly because Americans consume more health care than people in other wealthy countries.
The fact
In March 2018, JAMA published a study by Irene Papanicolas, Liana Woskie and Ashish Jha comparing the United States with ten other high-income countries: the United Kingdom, Canada, Germany, Australia, Japan, Sweden, France, the Netherlands, Switzerland and Denmark. Using data mainly from 2013 to 2016, largely drawn from the OECD, they looked at the usual suspects one by one.
The US spent 17.8% of its GDP on health care. The other ten countries spent between 9.6% and 12.4%.
If the utilisation story were right, you would expect Americans to see doctors far more often and spend far more days in hospital. They did not. The study found that US rates of utilisation were broadly similar to the other countries. Americans did not have unusually frequent physician visits or hospital stays. The authors concluded that the spending gap was driven mainly by prices: for labour and goods, including physician and hospital services, pharmaceuticals, diagnostic tests and devices, plus administrative costs.
Two numbers from that study make the point concrete. The US spent $1,443 per person on pharmaceuticals, against an average of $749 across the comparison countries. And planning, regulating and managing the health system, essentially administration, consumed about 8% of US national health expenditure, compared with an average of around 3% elsewhere.
The study was not the first to reach this conclusion. In 2003, Gerard Anderson and colleagues published a paper in Health Affairs whose title said it bluntly: “It’s the prices, stupid.” Fifteen years later, with better data, the answer had not changed.
Why the myth survives
If the evidence has been consistent for two decades, why does the utilisation story persist?
Part of the reason is that it is partly true in specific places. The US does perform some procedures and imaging at higher rates than peers. Anyone who has worked in an American hospital can point to tests that did not need doing. Those examples are real, visible and easy to talk about. Prices are abstract and buried in contracts nobody outside the finance department reads.
Another reason is that utilisation is a more comfortable target. Telling patients they use too much care puts the burden on individuals. Telling hospitals, drug manufacturers and physicians that they are paid more per unit of care than their counterparts in Germany or Japan puts the burden on some of the most powerful interests in any health system.
There is also a subtler point. High prices and high utilisation can coexist in narrow areas while averaging out across the whole system. Myths are often built from true anecdotes that do not add up to a true aggregate.
What this means outside the US
For health systems elsewhere, the lesson is not “the US is expensive”, which everyone knows. It is that when costs are rising, the reflex to squeeze demand first may be aimed at the wrong variable.
Consider what the comparison countries have in common. Most have mechanisms that set or strongly negotiate prices: national fee schedules, single or dominant purchasers, health technology assessment bodies that tie coverage to value, reference pricing for drugs. They are not buying less care per person, in the aggregate. They are paying less for each unit of it.
That has practical implications for any system trying to contain costs:
- Before cutting access, look at what you pay per unit: per drug, per procedure, per bed-day, per consultation.
- Measure administrative spending honestly. Fragmented payers and complex billing rules generate real costs that do not improve care.
- Treat purchasing capacity as core infrastructure. A strong purchaser that can negotiate on price is one of the most powerful cost controls a system can have.
- Be sceptical of cost-control strategies that rely mainly on patients using less, especially where cost-sharing already deters people from care they need.
Why prices end up so different
If the gap is mostly price, the next question is why American prices are so much higher. The answer is less about any single villain than about who holds bargaining power.
In most of the comparison countries, a single public payer or a small number of tightly regulated insurers buy care on behalf of nearly everyone. When one buyer represents most of a market, a hospital or manufacturer cannot easily walk away from the table. Prices are negotiated, or simply set, with that leverage in the room.
The US splits purchasing across Medicare, Medicaid, and many private insurers and employer plans, each negotiating separately. Private insurers pay hospitals and physicians substantially more than Medicare does for the same service, and the gap varies widely from one market to the next. Where hospitals have consolidated into large local systems, insurers have little choice but to include them in their networks, and the hospitals can charge accordingly. A buyer that cannot credibly say no is not really negotiating.
Fragmentation also explains part of the administrative bill. Every separate payer has its own rules, forms, prior authorisation requirements and billing codes. Providers employ large teams to navigate them, and insurers employ large teams on the other side. Much of that 8% administrative share is the cost of many buyers and sellers arguing with each other about money, an activity that barely exists in systems with one set of rules.
Anderson and colleagues returned to the question in Health Affairs in 2019 with a follow-up paper titled “It’s still the prices, stupid.” The title tells you what they found. Sixteen years on, the diagnosis had not changed.
A worked illustration
Consider how the same outpatient scan can turn into very different bills. The figures below are illustrative, not real prices.
Suppose two countries each perform 100 MRI scans per 1,000 people a year. Country X pays a fixed national fee of 300 per scan. Country Y has no single fee; the price depends on which insurer the patient has and which facility they use, and averages 1,000 once the facility charge, the radiologist’s fee and the billing overhead are added up.
Both countries use exactly the same amount of imaging. Country Y spends more than three times as much on it. If Country Y’s government responds by trying to cut the number of scans by 10%, it saves 10% of its imaging bill and probably denies some patients a scan they needed. If it moves the average price halfway towards Country X’s, it saves 35%, without anyone going without.
That is the whole argument of the JAMA study in miniature. Volume is a lever. Price is a much bigger one.
The strongest case for the myth
The fairest version of the utilisation argument is not “Americans see doctors too much”. It is that Americans receive more intensive care per encounter: more imaging, more specialist involvement, more expensive technology used earlier. The Papanicolas study did find that the US used some diagnostic imaging and some procedures at higher rates than peers. If you define “use” as intensity rather than visits, the myth has a stronger leg to stand on.
There is also a feedback loop worth taking seriously. High prices can create high intensity. If a hospital earns a large margin on an imaging scan or a procedure, it has a reason to buy the equipment and to fill its schedule. Price and volume are not fully independent, and some of the excess use Americans do have is plausibly driven by the prices themselves.
But this reinforces the conclusion rather than overturning it. If prices partly drive intensity, then controlling prices also dampens overuse. Starting with volume leaves the underlying incentive in place.
Second-order effects
Getting this wrong has consequences beyond the budget.
When systems blame demand, they tend to reach for cost-sharing: higher deductibles, co-payments, charges at the point of care. Cost-sharing does reduce use, but it is a blunt instrument. People cut back on needed care as well as unnecessary care, and the burden falls hardest on those with the least money and the most illness. A policy built on the wrong diagnosis can reduce spending while making health worse.
When systems focus on price, the costs land elsewhere. Providers and manufacturers lose revenue. They warn, sometimes correctly, that lower prices will reduce investment, shrink the workforce or slow innovation. Those are real trade-offs to be argued over. But they are the right argument to have, because they are about the variable that actually explains the gap.
What to watch
For readers outside the US, the practical signal to watch is whether your own system’s cost-control debates start with unit costs or with patient behaviour. Watch, too, how new high-cost medicines are priced as they arrive. The countries in the JAMA comparison kept spending lower mainly by paying less per unit, and the pressure on that model tends to show up first in drug pricing negotiations.
The limits of the fact
Myth-busting can overshoot, so it is worth being clear about what the study did not show. It did not show that American health care has no overuse, or that low-value care is not a problem worth tackling. It did not show that prices are the only driver of spending growth over time; new technology, chronic disease and ageing all matter. And its data comes from the mid-2010s, so the exact figures have moved since.
What it showed is narrower and more useful: the enormous gap between the US and other rich countries is not primarily explained by Americans using more care. It is explained by what each unit of care costs.
That distinction changes where you look for savings. It changes who has to give something up. And it is one of the most consistently misunderstood facts in health policy.
One more caution. A finding about the gap between countries is not the same as a finding about what drives growth within one. A system can have fair prices today and still see spending climb quickly because its population is ageing or because expensive new treatments keep arriving. Price control buys a lower baseline. It does not end the need for hard choices about what to cover, and anyone who sells it as a permanent solution to cost pressure is overselling it.
Your turn
Which health-policy “fact” do you hear repeated most often that the evidence does not support? Tell us in the comments, along with the source you would use to set the record straight. We may take on the best ones in a future myth vs fact.
