Published 9/1/2014
Jared L. Harwood, MD; Andrew J. Pugely, MD

The Evolution of GME Funding

U.S. healthcare workforce reaches a crossroads

Although graduate medical education (GME) funding has directly affected the lives of most U.S. physicians, few understand its evolution. Before the 1940s, hospitals paid trainee stipends by building costs into patient charges. The Servicemen’s Readjustment Act of 1944, known informally as the GI Bill, provided subsidies for residency expenses and teaching hospitals. Hospitals were also able to add facility and technology acquisition costs to insurance charges. For the first time, postgraduate medical residents earned salaries.

The modern GME system was born in 1965 with the creation of Medicare. Resident education was thought to be a public good that warranted public investment. According to the rationale in House and Senate reports, “…educational activities enhance the quality of care in an institution, and it is intended, until the community undertakes to bear such educational costs in some other way, that part of the net cost of such activities (including stipends of trainees, as well as compensation of teachers and other costs) should be borne to an appropriate extent by the hospital insurance program.”

Financing for GME funding was provided either through reimbursements from private payers, based on allowable costs (including GME expenses), or through Medicare’s per-patient hospital reimbursements.

Under Medicare’s prospective payment system, direct graduate medical education (DGME) payments were made through a “per-resident payment,” based on costs negotiated in fiscal year 1983, adjusted for inflation. These DGME payments are meant to cover direct expenses such as office/classroom space, salaries, and benefits for residents, supervising physicians, and administrative and clerical staff.

In 1983, Medicare also started to cover “indirect medical education (IME) costs”—payments meant to compensate teaching hospitals for increased use of tests and ancillary services, greater severity of illness, increased inefficiencies in teaching, greater concentration of high technology, and differences in types of physicians and payments. The DGME and IME payment formulas established at that time still exist in a slightly different form (Table 1).

Despite caps in federally funded residency slots established in 1997, GME has been able to grow by roughly 1 percent per year since 2001. These “over cap” residency and fellowship positions are primarily funded by the sponsoring teaching hospital. A disproportionate amount of these increases have been in subspecialties, not primary care specialties.

In 2002, the American Association of Medical Colleges (AAMC) began to call on legislators to expand Medicare-funded residency positions by 15 percent or roughly 15,000 positions. Though some unused slots have been redistributed, calls to increase federally funded residency slots have not been heeded. Over that same period, medical schools have increased total enrollment by 30 percent.

Current GME funding
Most of the increases in residency positions over the last 5 years have been in primary care, including internal medicine and family practice. These programs filled more than 95 percent of their positions, although less than half were filled by U.S. medical school seniors.

The current median PGY-1 resident stipend is approximately $50,000, but outside of DGME expenses such as salaries, the true cost of the medical resident is an elusive figure. Annually, $13 billion is spent on residency education, which equates to about $100,000 for each of the country’s approximate 115,000 residents.

Medicare accounts for nearly three-fourths of federally financed GME funding, with Medicaid, the Department of Veterans Affairs, the Department of Defense, and the Bureau of Health Professions making up the bulk of the rest. The Medicare subsidy for resident training costs is proportional to the percentage of Medicare patients cared for by that particular teaching hospital.

A 2013 RAND report to the Medicare Payment Advisory Commission (MedPAC) found a median GME cost per full-time equivalent resident of $134,803. A University of Washington study calculated a mean expense per resident of $338,421 and mean revenue of $317,248 for a mean cost per resident of $27,260, based on data covering 10 years of family practice residencies.

Medicare funding varies widely from state to state, as does Medicaid funding. One study that analyzed Medicare cost reports from teaching hospitals found several large state-level differences (an average payment per resident of $63,811 in Louisiana versus $155,135 in Connecticut). In 2011, only 44 states funded GME through their Medicaid funds, while nine others had considered ending GME Medicaid funding.

Residents are a very productive group of highly educated/skilled individuals. One workforce study conducted in 1995 concluded that it would take three midlevel practitioners to replace one resident. Although hospitals are prohibited from charging Medicare for procedures performed by residents, billing private insurers for resident-performed procedures might be a way to subsidize the costs of resident education. One study calculated the value of unsupervised procedures performed by residents or fellows at approximately $232,726 annually.

Some of the main concerns facing GME involve the overall physician workforce, the lack of GME expansion in the face of the Affordable Care Act and an aging population, and an outdated, inflexible payment model. The AAMC predicts a physician shortage of 91,000 physicians by 2020, consisting of 45,000 primary care physicians and 46,000 specialists.

Despite the significant public investment in GME, some argue that a disconnect exists between current GME and the needs of the nation’s healthcare workforce. Physician specialty mix has been specifically targeted. The shortage of U.S. primary care physicians continues, despite consistent evidence that health systems with more primary care providers provide better population-based health outcomes.

In 2013, 20,164 U.S. medical students graduated versus 26,504 who matriculated. In that same year, 26,392 PGY-1 residency positions were available. Without significant increases in residency positions, the domestic production of medical school graduates will soon outstrip the number of GME positions available. This is a daunting predicament for medical students graduating with an average debt that now exceeds $170,000.

Several efforts to pass federal legislation have been attempted, but have not progressed past committee. Bills in the House and Senate have been introduced that call for an increase of 15,000 Medicare-supported residency positions over 5 years at a cost of roughly $9 billion dollars over 10 years. With constrained federal and state budgets, relying solely on Medicare and Medicaid funds is proving vulnerable at best.

Current recommendations
Several advisory bodies have offered recommendations for improving the status of GME. For example, the AAMC recommends an increase of at least 4,000 new positions each year. Even so, this increased physician training capacity would address less than half of the expected shortages.

The 2010 MedPAC report offered nine recommendations, including changing GME funding to a performance-based allocation system, based on the achievement of established standards for practice-based learning and improvement, interpersonal and communication skills, professionalism, and systems-based practice, such as integration of community-based care with hospital care. It also called for a workforce analysis to determine the number of residency positions needed in the United States—in total and by specialty—as well as the development of strategies for increasing the diversity of the health professional workforce.

The Council on Graduate Medical Education, a division of the Health Resources and Service Administration in the Department of Health and Human Services, has also made recommendations. Their latest report (August 2013) called for increased funding from a wide range of funding sources, as well as revisions in both the criteria for recruiting medical students and GME training requirements to align with development of a physician workforce that meets the healthcare needs of the populations served.

Finally, the Institute of Medicine report released this year recommended shifting GME payments to a performance-based platform, creating a GME policy and financing infrastructure within the federal government, establishing subsidiary funds to support both current operations and innovative programs, modernizing the GME payment methodology, and mandating similar levels of transparency and accountability in Medicaid GME funding at the state level.

What is the future of GME funding?
An all-payer system that would more evenly distribute the cost of GME may be key to fixing the GME funding pipeline. In addition to federal attempts, some states have tried to move in that direction. A bill introduced in California in 2013 suggested assessing third-party payers to create grants that would help augment GME funding, but it failed to pass.

Demonstration projects under the Centers for Medicare & Medicaid Services (CMS) can potentially offer alternative and more effective ways of distributing Medicare funds. In 1997, for example, Utah applied for a CMS waiver and established the Utah Medical Education Council, which performs ongoing state workforce analysis and distributes GME funds based on the need of the state.

Another route would be establishing an alternative GME funding scheme with a limit on Medicare/Medicaid contributions, requiring new funding streams to make up the difference. This approach could be used for new residency slots and serve as a pilot program to lessen GME’s reliance on Medicare and Medicaid funding.

Several groups have suggested moving to an outcomes- or performance-based model for funding GME. The Accreditation Council for Graduate Medical Education has been moving to an outcomes-based evaluation system. These data could be used to reward programs based on performance.

Currently, private payers are not being charged for most unsupervised procedures performed by residents and fellows. Although insurance companies may subsidize GME funding by negotiating higher reimbursement rates for teaching hospitals, this process is not mandated or transparent. By assessing a resident charge (payable directly to the institution’s GME office) on patients cared for by residents, hospitals would increase transparency and accountability.

Starting new programs at nonteaching hospitals is permitted under current law and has been done in some states, but the 5-year lag in accreditation and reimbursements is problematic.

Given the importance of and the public good provided by the physician workforce, it is imperative that we take substantive steps toward improving GME by finding ways to increase stability and secure its future. Developing better ways to invest in residents and fellows today will pay big dividends tomorrow.

Jared L. Harwood, MD, is a fourth-year orthopaedic surgery resident at Ohio State University and Andrew J. Pugely, MD, is a fifth-year orthopaedic surgery resident at the University of Iowa. They are the current AAOS Washington Health Policy Fellows.


  1. House Report, Number 213, 89th Congress, 1st session 32 (1965) and Senate Report, Number 404 Pt. 1 89th Congress 1st Session 36 (1965)
  2. Advance Data Tables 2014 Main Residency Match. March 21, 2014 NRMP
  3. Survey of Resident/Fellow Stipends and Benefits Report 2013–2014. AAMC
  4. Improving Value in Graduate Medical Education. Council on Graduate Medical Education Twenty-First Report. August 2013
  5. Metzler I, Ganjawalla K, Kaups KL, Meara JG: The critical state of graduate medical education funding. Bull Am Coll Surg 2012;97(11):9-18.
  6. Mullan F, Chen C, Steinmetz E: The geography of graduate medical education: Imbalances signal need for distribution policies. Health Aff (Millwood) November 2013;32(11):1914-1921. doi: 10.1377/hlthaff.2013.0545
  7. Feinstein AJ, Deckelbaum DL, Madan AK, McKenney MG: Unsupervised procedures by surgical trainees: A windfall for private insurance at the expense of graduate medical education. J Trauma 2011 Jan;70(1):136-9; discussion 139-40. doi: 10.1097/TA.0b013e3182014caf
  8. Lesko S, Fitch W, Pauwels J: Ten-year trends in the financing of family medicine training programs: Considerations for planning and policy. Fam Med 2011;43(8):543-550.
  9. 2012 Commonwealth Fund International Survey of Primary Care Doctors. Published November 15, 2012
  10. MedPAC Report to Congress: Aligning Incentives in Medicare. June 2010
  11. Twenty-First Report: Improving Value in Graduate Medical Education. COGME. August 2013
  12. Retrieved from http://www.leginfo.ca.gov/pub/13-14/bill/asm/ab_1151-1200/ab_1176_bill_20130321_amended_asm_v98.html . Accessed on July 25, 2014