The Hospital Bill: America’s Most Extreme Inflation Story

In the previous episodes of this series, we tracked the price of groceries, gasoline, and rent — the everyday costs that Americans notice because they pay them with cash, a card swipe, or a monthly check. Medical care is different. Most Americans encounter its true cost only in moments of crisis: a trip to the emergency room, a surgery, a hospitalization. The bill arrives weeks later, often incomprehensible, frequently staggering, and almost always far larger than the patient expected.

The Bureau of Labor Statistics has been tracking the price of medical care since 1947 as part of the Consumer Price Index. The data tells a story that is, by the numbers, more dramatic than any other category in the CPI. In 1947, the medical care index stood at 13.2. By January 2026, it reached 589.6 — an increase of 4,367%. Over the same period, the all-items CPI rose from 21.5 to 326.6, an increase of 1,420%. Medical care has risen roughly three times faster than everything else.

But the aggregate number conceals a more alarming story. Within the medical care category, one subcategory has pulled away from all the others with a ferocity that defies comparison: hospital services. And that is where this episode begins.

For the first three decades of the CPI’s modern history, medical care prices and overall prices moved in rough parallel. In 1947, the medical care index sat at 13.2 while all items sat at 21.5 — medical care was actually lower on the index because it had risen less than the average good during the war years. By 1965, the two lines were converging: medical care at 24.8, all items at 31.3. Doctors made house calls. Hospital stays were short and relatively inexpensive. Health insurance, where it existed, was a simple indemnity plan.

Then two things happened in quick succession. In 1965, Congress created Medicare and Medicaid, instantly expanding the pool of insured patients — and the pool of money flowing into the healthcare system. Through the 1970s, employer-sponsored insurance became the norm rather than the exception, further insulating patients from the direct cost of care. The result was predictable to economists and invisible to everyone else: prices began to rise without constraint.

By 1980, the medical care index had reached 71.4 while all items stood at 78.0 — medical care had nearly caught up. By 1990, the crossover was complete and decisive: medical care at 156.0, all items at 127.5. By 2000, the gap had widened to 86 points. By 2010, to 165 points. By 2026, the medical care index stands at 589.6 and the all-items index at 326.6 — a gap of 263 points and growing.

The chart above captures what may be the most consequential price divergence in American economic life. Two lines that were virtually on top of each other in 1965 now differ by a factor of nearly two. Every American who has opened a medical bill and felt a jolt of disbelief is looking at the real-world consequence of this divergence.

Medical care is not a monolith. The CPI breaks it into subcategories — hospital services, physicians’ services, prescription drugs, medical care commodities — and the differences among them are almost as dramatic as the difference between medical care and everything else.

Hospital services are the outlier within the outlier. The BLS began tracking hospital services as a separate index in the late 1970s, with the earliest available data point at 52.3 in January 1978. By January 2026, that index had reached 1,185.9 — an increase of 2,168%. That works out to an annualized rate of approximately 6.7% per year, sustained for nearly half a century. No other major CPI category comes close.

Physicians’ services, by contrast, rose from 73.9 in 1980 to 448.2 in 2026 — an increase of 506%. Still far above overall inflation, but less than a quarter of the hospital increase in percentage terms. Prescription drugs went from 69.4 in 1980 to 569.6 in 2026 — up 721%. Again, extraordinary by any normal standard, but moderate compared to the hospital juggernaut.

The chart below puts all three lines on the same canvas. The visual is striking: physicians’ fees, which most patients would consider expensive, look almost flat compared to the vertical ascent of hospital costs.

One of the most persistent features of medical inflation is that it exceeded overall CPI in every full decade from the 1950s through the 2010s — seven decades running. The pace varied — the 1970s and 1980s were the worst, the 2010s the mildest — but the direction never reversed. In the partial 2020s, a general inflationary surge driven by food and energy temporarily pushed overall CPI ahead of the medical care index, but hospital services specifically continued to outpace everything.

In the 1950s, medical care rose 47% while overall CPI rose 25%. In the 1960s, medical care rose 50% while CPI rose 29%. The 1970s brought general inflation to everyone, but medical care still outpaced it: 118% versus 106%. Then came the 1980s, when overall inflation cooled under Volcker’s Fed — all items rose 64% for the decade — but medical care surged 119%, nearly double the pace. The 1990s were similar: 64% for medical care versus 33% for all items.

The 2000s and 2010s brought narrowing. The Affordable Care Act, managed care pressure, and slower overall healthcare utilization growth helped moderate the gap. Medical care still outpaced overall CPI, but by smaller margins: 50% versus 29% in the 2000s, and 34% versus 19% in the 2010s. The 2020s, so far covering just six years, are the first period in which a broad inflationary wave pushed overall CPI growth (26%) above the medical care index (15%) — though hospital services, at +30%, continued to outpace both.

The structural explanation for medical inflation is, at its core, deceptively simple: when the person receiving a service does not pay for it directly, the price of that service rises without constraint. This is not a controversial observation among economists. It is the predictable consequence of a system in which patients consume, insurers negotiate, employers fund, and the government subsidizes — and no single party bears the full cost of any decision.

Third-party payment is the foundation. In 1960, patients paid about 48% of their healthcare costs out of pocket. By 2020, that figure had fallen to roughly 11%. The remaining 89% flowed through private insurance, Medicare, Medicaid, and other programs — each adding layers of administrative complexity and each removing the patient one step further from the price of the service.

Administrative costs are the second pillar. A 2019 study in the Journal of the American Medical Association estimated that administrative costs account for 34% of total U.S. healthcare expenditure — roughly $1 trillion per year — compared to 17% in Canada and 12% in the Netherlands. Every hospital employs armies of billing specialists, coders, compliance officers, and insurance liaisons. Every insurance company employs armies of claims processors, network negotiators, and utilization reviewers. The two armies face each other across a bureaucratic battlefield, and the cost of the battle is passed through to patients and premiums.

Technology is the third factor, and the most counterintuitive. In most industries, new technology drives prices down. In healthcare, it drives prices up. An MRI machine costs $1–3 million. A da Vinci surgical robot costs $2 million plus $150,000 per year in maintenance. These technologies improve outcomes, but they do so at enormous cost — and because patients rarely see the price, there is no market pressure to deliver the same capability at a lower cost. The technology treadmill runs in one direction.

Certificate-of-need laws, which exist in 35 states, add a final structural constraint. These laws require hospitals to obtain government approval before expanding capacity or adding new services — effectively limiting competition. The result is regional hospital monopolies that face little pricing pressure from potential entrants.

There was a moment, around 2012–2018, when it appeared that the long arc of medical inflation might be bending. Annual increases in the medical care CPI slowed to 2–3%, barely above overall CPI. Health economists pointed to several causes: the Affordable Care Act’s cost-containment provisions, the spread of high-deductible health plans that reintroduced some price sensitivity, and a secular slowdown in healthcare utilization as the population aged past its peak consumption years.

The hope was premature. Hospital services went from an index of 910.4 in January 2020 to 1,185.9 in January 2026 — an increase of 30.3% in just six years. That is 4.5% per year, well above the 3.8% annualized rate for overall CPI during the same period. The pandemic did not create hospital inflation, but it removed the forces that had been temporarily restraining it.

Nursing shortages drove wage increases of 15–25% at many hospitals between 2020 and 2023. Travel nurse costs, which had been a modest line item before the pandemic, surged to consume 5–10% of some hospitals’ labor budgets. Supply chain disruptions raised the cost of medical devices, pharmaceuticals, and even basic supplies like gloves and gowns. And as patients returned for deferred procedures in 2022 and 2023, hospitals found themselves operating at capacity with a workforce that demanded — and received — substantially higher compensation.

Prescription drugs offer one of the few bright spots: the index peaked around 572.7 in 2025 and actually declined slightly to 569.6 in January 2026. This is exceedingly rare in medical care — actual deflation in a major subcategory. The cause is a combination of patent expirations on several blockbuster drugs, the growth of biosimilars, and the Medicare drug price negotiation provisions of the Inflation Reduction Act, which took effect in 2026 for the first ten drugs.

The table below shows every major medical subcategory tracked by the CPI, with its earliest available index value, its January 2026 value, and the total percentage change. The range is enormous: hospital services lead at +2,168%, while physicians’ services — the subcategory most patients interact with directly — trail at +506%.