Medicare graduate medical education funding cuts are moving from budget scenario to operational planning at U.S. academic medical centers. The two payment streams that keep residency programs staffed, direct graduate medical education and indirect medical education, are under the same fiscal pressure that has already reached federal research grants and Medicaid supplemental payments.
Roughly $16 billion a year goes from Medicare to teaching hospitals through graduate medical education. That money pays resident stipends and benefits, teaching physician time, a share of administrative overhead, and part of the credentialing and accreditation work that keeps a program running. Cuts of even a few percentage points would not be absorbed quietly. They would show up in how many residents a program can carry and which clinical services stay covered overnight.
Two payments, one workforce pipeline
Direct GME covers the cost of training residents: salaries and benefits, faculty teaching time, administrative overhead, and the supervision that teaching demands. IME is an add-on to Medicare inpatient payments based on a hospital's resident-to-bed ratio. The IME formula exists because teaching hospitals order more diagnostic tests, request more consults, manage longer admissions, and absorb the pauses that training requires.
DGME per resident is not a single national rate. Each hospital has a base-year per-resident amount updated annually, then multiplied by its Medicare share of inpatient days. That means two teaching hospitals in the same city can receive very different direct GME payments for the same resident. IME is a percentage add-on that rises with a hospital's resident-to-bed ratio, but it is also the part that MedPAC has targeted for years because it is not tied directly to the cost of running a residency.
The constraint underneath all of this is the Balanced Budget Act of 1997. It capped how many residents Medicare will fund at each hospital, using 1996 counts as the baseline. A hospital over its cap can still train more residents, but Medicare will not pay DGME or IME for them. That one provision is why new teaching hospitals, rural training tracks, high-growth health systems, and some safety-net networks have spent years without full Medicare GME support.
Photo by Piron Guillaume on Unsplash
Why the cuts land unevenly
Safety-net teaching hospitals have the most to lose. They run high resident-to-bed ratios not because they are rich, but because their patients are complex and the work is training-intensive. A reduction in IME changes Medicare payment on every inpatient discharge, not just on the education line. That is a much deeper cut than a line-item reduction sounds.
Rural tracks face a different version. They often operate with a small number of Medicare-funded positions, and losing three or four of them can make a track too small to sustain. The American Association of Medical Colleges tracks those formulas because the caps and payments determine where new programs can even open. The AAMC's Medicare GME explainer walks through how the payment pieces fit together.
MedPAC has argued for years that IME payments exceed the actual added cost of training. That keeps graduate medical education on the deficit-reduction menu, because the recommendation is already written and scored, and it sits ready for a committee to pick up. MedPAC's June 2024 report outlines the case for recalibrating those payments.
The physician shortage math is already bad
The country does not have room to shrink funded residency slots while facing a projected shortage of up to 86,000 physicians by 2036. The AAMC's workforce analysis made that number a policy benchmark, and it assumed current training levels would continue. AAMC's physician shortage data shows the gap concentrated in primary care and rural areas, the same tracks that depend most on Medicare GME.
Medical school enrollment has grown. Medicare-funded residency positions have not kept up. The 1,000 new positions Congress approved in 2021 were a start, but they are distributed in small annual waves and do not offset the 1997 cap. The Resident Physician Shortage Reduction Act would add 14,000 Medicare-funded positions over seven years, and its funding question is exactly where the current budget fight gets hard.
Photo by Piron Guillaume on Unsplash
What administrators can do before the formula changes
Start with your own Medicare GME cost report. Know your hospital's direct GME per-resident amount, your resident cap, your current resident count billed against it, and which programs are nearest the cap. That snapshot is the baseline for every scenario after a cut.
The pattern will feel familiar to anyone who managed the NIH indirect cost reset: a formula change announced quietly in Washington becomes a multiyear payroll adjustment at home. The same federal pullback that hit NIH indirect cost recovery is now reaching teaching-hospital budgets. Many research medical centers then moved toward broader faculty hiring freezes as the uncertainty spread. Academic medical center hiring budgets are often the next line item to freeze.
The annual Inpatient Prospective Payment System rule from the Centers for Medicare & Medicaid Services is where IME adjustments often surface. If you don't read it until after comments close, you've missed the chance to influence the formula. The CMS IME page tracks the current adjustment.
- Pull your hospital's Medicare GME cost report and confirm your direct GME per-resident amount before the next budget cycle closes.
- Model IME as a variable in every five-year financial plan; do not keep it flat by default.
- Protect rural and primary care slots first, because they are the hardest to rebuild once a track closes.
- File comments on the annual CMS proposed rule through your state hospital association and the AAMC before the comment window shuts.
One number will tell you how much room you have before a cut becomes a service reduction: your Medicare-funded resident cap. Ask the GME office for it this week, along with the count of residents currently billed against it. You can't plan for a cut you haven't measured.
