The doctor pipeline is still sized for 1996

Every doctor practicing in the United States trained somewhere on Medicare's dime.▸ The program doesn't just insure the elderly — it is the largest single funder of physician training in the country, underwriting residency slots at teaching hospitals nationwide. The catch: how many slots it will pay for at any given hospital was fixed by a number from three decades ago, and Congress has only nudged that number since.

Medicare's graduate medical education spending moves through two channels, both funded out of the Part A Hospital Insurance Trust Fund.▸ Direct GME (DGME) reimburses hospitals for the literal cost of running a residency program — resident stipends, supervising-physician time, administrative overhead. Indirect Medical Education (IME) is a separate add-on payment, recognizing that patients at teaching hospitals cost more to treat. In FY2023, per CRS analysis of CMS hospital cost report data, Medicare paid hospitals an estimated $21.2 billion combined: $6.1 billion in DGME for 116,431 funded training positions (FTEs), and $15.0 billion in IME for 119,328 FTEs.

That's real money, and it buys real training.▸ What it doesn't buy is growth. The number of FTEs Medicare will fund at any hospital is hospital-specific, and for institutions that were already training residents in 1996, that number was frozen in place by the Balanced Budget Act of 1997 (P.L. 105-33) — capped at whatever the hospital was training that year. Only hospitals that started residency programs after 1996 get to set a fresh cap. Everyone else has been training against a ceiling drawn before most of today's residents were born.

That freeze wasn't supposed to bind forever, but it has.▸ The same CRS report finds that in FY2023, 73% of teaching hospitals (788 of them) were training at or above their Medicare-funded FTE cap, and another 25% were within 15 FTEs of it — together, 98.5% of all teaching hospitals were at or pressed right up against a limit set by their 1996 headcount. Most hospitals respond by training residents beyond the Medicare cap anyway, covering the excess with state, hospital, or other federal funds — Medicare just won't pay for it.

Congress has moved the ceiling exactly twice since 1997.▸ The Consolidated Appropriations Act, 2021 (P.L. 116-260) created 1,000 new Medicare-funded FTEs nationwide, phased in at 200 per year through FY2027 and distributed by statutory priority to hospitals in states with low physician-to-population ratios. The Consolidated Appropriations Act, 2023 (P.L. 117-328) added 200 more, aimed at behavioral health. That's 1,200 new slots, total, against a base of roughly 236,000 already-funded FTEs — a 0.5% increase, spread across a system that hasn't otherwise grown its Medicare-funded capacity since Bill Clinton's first term.

The scale gap in that chart is the story.▸ AAMC's 2024 projections put the total U.S. physician shortage at 13,500 to 86,000 by 2036 — but that range already assumes GME funding keeps climbing beyond today's levels, driven by increases like the 1,200 slots above. AAMC says so explicitly: without continued funding growth, "future shortages will be much worse... closely resembling" its prior 2021 report, which projected a shortfall of up to 124,000 physicians by 2034. In other words, the more optimistic shortage number on the chart isn't a forecast of what will happen if nothing changes — it's a forecast of what happens if Congress keeps adding a few hundred slots a year, indefinitely. The pessimistic number is what happens if it doesn't.

Dollar and FTE figures are for FY2023, the most recent year with complete Medicare hospital cost report data available in CRS's analysis; shortage projections are AAMC's national totals across all physician specialties, not limited to Medicare-trained residents.