← Back to Library

America’s self-inflicted doctor shortage

A Bottleneck Built by Design

The United States does not have a shortage of people who want to become doctors. It has a shortage of permission slips. PolyMatter's examination of America's physician crisis zeroes in on a structural chokepoint that most healthcare debates gloss over entirely: the residency system, frozen in place by a 1997 congressional cap that has barely budged in nearly three decades.

The numbers are stark. Eighty-three million Americans live in medically underserved regions. By 2036, the country could be short as many as 86,000 physicians. And yet, medical schools reject roughly as many applicants as they accept, with admission rates hovering between two and five percent. The instinct is to blame a lack of training capacity, but the real constraint sits one step further down the pipeline.

Last year, there were 47,000 applicants for 37,000 places. Meaning after 13 years of pretertiary education, 4 years of undergrad and 4 years of medical school, not to mention all the entrance exams in between, nearly 10,000 people find themselves blocked from progressing.

Those ten thousand are not underqualified washouts. They are medical school graduates, people who survived the top-half MCAT filter and four years of grueling coursework, now stranded with two hundred thousand dollars in debt and no legal pathway to practice medicine. The system manufactured them and then discarded them.

America’s self-inflicted doctor shortage

The Ghost of Roemer's Law

How did the country end up here? The answer traces back to a spectacularly wrong forecast. In the early 1980s, riding the wave of Reaganomics and armed with researcher Milton Roemer's observation that hospital beds built tend to be hospital beds filled, a government committee predicted a surplus of at least 100,000 physicians by the year 2000. The prescription was deliberate constriction: freeze medical school enrollment, require certificates of need for new hospitals, and eventually cap federal residency subsidies.

For the next 25 years, from 1980 to 2005, the entire US health care pipeline was deliberately frozen in place.

The prediction was catastrophically wrong. Demand for healthcare did not recede. It accelerated, driven by population growth, an aging demographic, and rising prosperity. Between 1980 and 2005, the United States added 68 million people while keeping its medical training infrastructure virtually unchanged. The supposed surplus never materialized; a deficit did.

There is a certain dark irony in Roemer's Law being used to justify restricting supply. The observation that available hospital beds get used is not inherently evidence of waste. It could just as easily indicate pent-up demand finally being met. As the piece notes, most people look for reasons not to go to the hospital, not reasons to check in. The Americans filling those postwar hospital beds were not malingerers. They were patients who had previously gone without care.

Who Holds the Keys

The 1997 Balanced Budget Act did more than cap residency funding. It effectively handed private hospitals a monopoly over physician production. Hospitals that happened to have large residency programs on December 31, 1996, were grandfathered into a permanent revenue stream. Everyone else was locked out.

Congress effectively created a bottleneck, an artificial limit on the number of subsidized residencies and then handed private hospitals control over who gets through.

This arrangement has consequences that ripple far beyond hospital administrators' balance sheets. The residency match system, in which graduates submit to a single annual lottery with no ability to negotiate or compare offers, would be considered unconscionable in almost any other profession. Imagine telling a newly minted engineer or lawyer that they must accept whatever employer and salary an algorithm assigns them, with no recourse and no counteroffer. Yet for physicians, this is simply how it works.

The power imbalance feeds a culture of exploitation. Sixty- to seventy-hour weeks are standard. Sleep deprivation is endemic. The emotional toll is immense. Hospitals have little incentive to improve conditions because the supply of desperate applicants vastly exceeds available slots. When ten thousand qualified graduates are turned away each year, every resident who complains knows there are dozens ready to take their place.

The Specialist Trap

Even among those who do make it through, the system's incentive structure pushes them away from where they are needed most. With average medical school debt exceeding $200,000 and some institutions sending graduates out with $300,000 in loans, the gravitational pull toward high-paying specialties is powerful.

The result is that while the United States has a perfectly normal number of specialists per capita, it has the lowest number of generalists among the OECD countries.

This is not a minor statistical footnote. It means the country excels at performing aortic valve replacements but struggles to provide routine checkups. The downstream effect is predictable and devastating: treatable conditions go undiagnosed until they become emergencies, at which point the patient faces both a health crisis and a financial one. The system is optimized for heroic interventions and starved of preventive care.

The geographic dimension compounds the problem. Because the 1997 cap froze not just the number but the distribution of residencies, fast-growing Sun Belt cities find themselves critically underserved. Las Vegas, Austin, and Orlando have doubled or tripled in population since the cap was set, but their residency allotments remain stuck in the mid-1990s. Since 56 percent of physicians end up practicing within a hundred miles of where they trained, these cities are structurally unable to grow their physician workforce.

Counterpoints Worth Considering

Critics of simply uncapping residency funding raise legitimate concerns. The United States already spends $21 billion annually on residency subsidies, more per capita on healthcare than any peer nation. Simply writing bigger checks without structural reform risks perpetuating the same misallocations. More residency slots concentrated in the same coastal academic medical centers would do little for Wakulla County, Florida.

There is also a reasonable argument that expanding the physician supply is not the only lever available. Nurse practitioners and physician assistants can deliver a significant share of primary care, particularly in underserved areas. Several states have already expanded scope-of-practice laws to allow this, with outcomes that are, by most measures, comparable to physician-delivered primary care for routine conditions. The piece mentions this option but does not dwell on it, perhaps because it challenges the physician-centric framing of the problem.

Additionally, the international comparison deserves nuance. Many countries with more generalists per capita achieve this through centralized workforce planning and salary structures that American physicians and their lobbying organizations would fiercely resist. The AMA has historically opposed most measures that would increase physician supply, a fact the piece alludes to but does not interrogate as deeply as it might.

Bottom Line

America's doctor shortage is not an accident of demographics or economics. It is the direct result of policy decisions made in the 1980s and 1990s based on a forecast that proved spectacularly wrong, compounded by nearly three decades of congressional inaction. The residency cap remains the binding constraint, and until it is meaningfully reformed, with attention to geographic distribution, primary care incentives, and the exploitative match system, no amount of new medical schools will produce a single additional practicing physician. The country has spent thirty years treating a self-inflicted wound with bureaucratic band-aids. The patients, quite literally, are still waiting.

Sources

America’s self-inflicted doctor shortage

by PolyMatter · PolyMatter · Watch video

Wakulla County, Florida is home to 37,000 people. There are three state parks, nearly a dozen schools, and even a small municipal airport. What there are not are doctors. Just 10 serve the entire county, which means per capita, there are fewer doctors here in the suburbs of Tallahassee than in Afghanistan or Haiti.

And Wakulla County is not alone. 83 million Americans live in a medically underserved region. With so little health care to go around, we drive long distances for basic care, regularly forego checkups, and wait weeks or even months for appointments. When we finally do see someone, they're often overwhelmed with too many patients to remember our names.

And things are only about to get worse. Over the next 10 years, the US population is expected to grow by 15 to 20 million people, and the population aged 75 and older, the largest consumers of health care, will grow by 54%. Meanwhile, many doctors are nearing retirement. By the end of that same decade, two out of five currently practicing physicians will themselves be at least 65.

Add all this up and we're looking at a shortage as large as 86,000 doctors by 2036 according to the Association of American Medical Colleges. In other words, Wakulla County and places like it may already have more access to health care today than they will for decades to come. The issue is not a lack of interest. Doctor is still one of the most coveted professions.

Each year, medical schools collectively reject roughly as many applicants as they ultimately enroll. The acceptance rate at most is between just 2 and 5%. Nor is the problem a shortage of schools. The truth is, you could double the number of schools tomorrow and you wouldn't create a single additional doctor.

That's because strictly speaking, medical schools don't create doctors. They create doctors in training. Before you can actually practice medicine independently, you have to complete a residency, a minimum of three and as many as 9 years of supervised on the job training. And there are more aspiring residents than there are residency positions.

Last year, there were 47,000 applicants for 37,000 places. Meaning after 13 years of pretertiary education, 4 years of undergrad and 4 years of medical school, not to mention all the entrance exams in between, nearly 10,000 people find themselves blocked from progressing. Denied the right ...