The Fund supports networks of state health policy decision makers to help identify, inspire, and inform policy leaders.
The Milbank Memorial Fund supports two state leadership programs for legislative and executive branch state government officials committed to improving population health.
The Fund identifies and shares policy ideas and analysis to advance state health leadership, strong primary care, and sustainable health care costs.
Keep up with news and updates from the Milbank Memorial Fund. And read the latest blogs from our thought leaders, including Fund President Christopher F. Koller.
The Fund publishes The Milbank Quarterly, as well as reports, issues briefs, and case studies on topics important to health policy leaders.
The Milbank Memorial Fund is is a foundation that works to improve population health and health equity.
February 18, 2025
Report
Yalda Jabbarpour
Anuradha Jetty
Hoon Byun
Anam Siddiqi
Jeongyoung Park
Publication
Feb 18, 2025
Back to Publications
Rebuilding the primary care workforce will require training the workforce of the future. Unfortunately, over the past decade, we have not seen a proportionate rise in the number of students choosing a primary care specialty compared to the number choosing non-primary care specialties (Figure 8). This has led to a widening of the gap between the density of primary care residents and residents in all other specialties and further exacerbation of the primary care workforce crisis.
Figure 8. Disparity Widens in Residents Per Capita Growth Between Primary Care and All Other Specialties (2012—2022)
Data Sources: Analyses of American Medical Association Masterfile (2016—2022), Centers for Medicare and Medicaid Services Medicare Provider Enrollment, Chain, and Ownership System data (2016—2022), National Plan and Provider Enumeration System data (2016—2022), Centers for Medicare and Medicaid Services Physician and Other Practitioners data (2016—2022), and the American Community Survey Five-Year Summary Files (2016—2022).
Medical students rotating through outpatient primary care are likely to see dysfunctional outpatient settings with underresourced practices and stressed clinicians.65 This experience, coupled with the fact that PCPs are relatively underpaid compared to their subspecialty colleagues, is a well-documented reason that medical students do not choose primary care careers.46, 65
Even students who do enter primary care residencies do not end up staying in the primary care workforce.7, 66, 67 Many of the primary care residents represented in Figure 8 will end up going into subspecialty fellowships after residency, or working in an inpatient, hospital-only setting as hospitalists. In fact, only 19.8% of all physician residents (excluding hospitalists in 2022) end up working in primary care, whereas 24.4% of all residents (including hospitalists in 2022) are in primary care residencies (Figure 9).
Figure 9. Percentage of New Physicians Entering Primary Care Drops to Lowest Rate in a Decade (Hospitalists vs Non-Hospitalists) (2012—2022)
Data Sources: Analyses of the 2024 American Medical Association Historical Residency File, the 2024 American Medical Association Masterfile, and the 2012-2022 Center for Medicare and Medicaid Services Physician and Other Practitioners data
Notes: Primary care specialties included family medicine, general practices, internal medicine, geriatrics, and pediatrics. Specialty for doctors of osteopathy (DOs) are not always included in the American Medical Association Masterfile, so these data may be an underestimation of the true workforce. (See limitations in Appendix for more details.)
While unfavorable views of the primary care lifestyle and underpayment of PCPs contribute to this attrition, the way that the US graduate medical education (GME) system is financed also plays a role. A majority of the approximately $24 billion spent on GME comes from Medicare ($17.8 billion in 2021),68 and a large portion comes from Medicaid ($7.38 billion).69 Although there are other sources of GME funding, including the US Department of Veterans Affairs, the Department of Defense, private payers, and the Health Resources and Services Administration (HRSA), these sources are dwarfed by CMS. CMS GME funding is intended to develop a workforce to effectively and efficiently care for Medicare and Medicaid patients, but that goal is not being met. Despite the reality that most patient care occurs in primary care settings,70 Medicare and Medicaid funding is not effectively building a primary care workforce. Our state-level analysis shows a negative association between Medicare GME funding and the percentage of physicians entering primary care three years later: the more funding a state receives, the fewer physicians join primary care after three years (Figure 10). A similar, though weaker, association exists for Medicaid (Figure 11). For example, New York received the highest level of Medicare funding for GME in 2019 at $15.5 million, but had very low primary care production, with only 15% of its physicians entering primary care three years later. New York was also an outlier on the Medicaid side, receiving $8.3 million in Medicaid funding for GME in 2015, with only 17% of new doctors entering primary care three years later. By comparison, North Dakota received only $2.1 million in Medicare GME funding, but 43% of new physicians in the state entered primary care.
Figure 10. More Medicare GME Funding (2019) in a State Is Associated with Lower Percentage of New PCPs (2022)
Data Sources: Analyses of the 2019 RGC’s Graduate Medical Education For Teaching Hospitals; the 2019 Area Health Resource File; the 2024 American Medical Association Historical Residency File; the 2024 American Medical Association Masterfile; and the 2012-2022 Center for Medicare and Medicaid Services Physician and Other Practitioners data
Notes: DC was excluded. Primary care specialties included family medicine, general practices, internal medicine, geriatrics, and pediatrics. Hospitalists were excluded. Specialty for doctors of osteopathy (DOs) are not always included in the American Medical Association Masterfile, so these data may be an underestimation of the true workforce. (See limitations in Appendix for more details.)
Figure 11. More Medicaid GME Funding (2015) Leads to Lower Percentage of New PCPs (2018)
Data Sources: Analyses of the 2016 AAMC’s Medicaid Graduate Medical Education Payments: A 50-State Survey; the 2015 Area Health Resource File; the 2024 American Medical Association Historical Residency File; the 2024 American Medical Association Masterfile; and the 2012-2022 Center for Medicare and Medicaid Services Physician and Other Practitioners data.
Notes:DC and states with $0 Medicaid GME (AK, CA, MA, NC, ND, NH, RI, and WY) were excluded. Primary care specialties included family medicine, general practices, internal medicine, geriatrics, and pediatrics. Hospitalists were excluded. Specialty for doctors of osteopathy (DOs) are not always included in the American Medical Association Masterfile, so these data may be an underestimation of the true workforce. (See limitations in Appendix for more details.)
State Standout: Training
As of 2022, three western states stand out on Scorecard training measures: Montana, North Dakota, and Idaho. Montana is the highest-ranked state for community-based training, providing 65% of its training to primary care residents in community settings (broad definition); the national average is just 16% (Figure 12). North Dakota has the greatest share of new entrants in primary care, at 46.6% of all new physicians (or 43.1% when excluding hospitalists), nearly double the national average (Figure 10). Idaho has the largest share of total primary care clinicians, with the percentage of physicians, NPs, and PAs in primary care at 31%, 42%, and 41%, respectively. By comparison, the national averages are 27% of all physicians, 30% of all NPs, and 24% of all PAs. See the data dashboard for more state data.
Why is it that traditional GME funding from CMS is not contributing to the creation of a robust primary care workforce? One answer lies in where the money goes. Data suggest that how and where a physician trains dictates how and where they end up practicing.71–75 At present, most GME funding is provided by Medicare and funneled directly to academic hospitals.76, 77 Consequently, the majority of CMS-funded residency training occurs in large academic hospitals rather than community-based settings, leading many physicians to pursue hospital-based specialties as opposed to community-based primary care.78
The Teaching Health Center (THC) program is one example of a training program that is primarily community-based. Residents in these programs end up staying in primary care and working with more vulnerable patient populations.79 Unfortunately, funding for community-based programs such as THCs is low ($155 million in 2022)80 and unstable.81, 82 As a result, most residents in the United States do not receive community-based training, with funding favoring traditional GME programs (Medicare-funded) in large academic hospitals over community-based programs like THCs. When examining the percentage of residents training in community-based settings — defined broadly as programs outside academic medical centers or hospitals affiliated with medical schools, or narrowly as rural training tracks or HRSA THC GME grant programs — we found that only 5.1% of primary care residents spent the majority of their training in outpatient settings with underserved populations (Figure 12). By not funding more community-based training, the GME system is failing to produce a workforce that meets the needs of the US population.
Figure 12. Increase in Percentage of Primary Care Residents Training in Community-Based Settings (2013—2022)
Data Sources: Analyses of Accredited Council of Graduate Medical Education program-level data to get counts for medical residents; FREIDATM American Medical Association Residency & Fellowship Program Database, a rural residency program list from the RTT Collaborative, and Health Resources and Services Administration Teaching Health Center Graduate Medical Education program dashboards to identify community-based training programs, 2013-2022.
Notes: Community-based training was identified if the majority of training does not take place in a university academic medical center or a hospital with a medical school affiliation (broad) or programs with rural training track or Health Resources and Services Administration Teaching Health Center Graduate Medical Education grant (narrow). Primary care specialties included family medicine, internal medicine, geriatrics, and pediatrics.
ACTION 3.1: Health care organizations should strive to diversify the primary care workforce and customize teams to meet the needs of the populations they serve. Government agencies should expand educational pipeline models and improve economic incentives.
ACTION 3.2: CMS, [Veterans Affairs], HRSA, and states should redeploy or augment Title VII, Title VIII, and GME funding to support interprofessional training in community-based primary care practice environments.
Florida Brings Transparency to the Distribution of Medicaid GME Funding
By Christine Haran
The majority of graduate medical education for physician residents and fellows is taxpayer-funded. Given that Medicaid is the second-largest source of this funding behind Medicare, some states are taking steps to bring more transparency to the process of allocating these funds in ways that help to improve population health.
In 2024, Florida passed legislation to better track Medicaid GME dollars.i “We have between $1 billion and $2 billion a year going to residency programs or hospitals and the handful of outpatient clinics that have resident slots,” said Christopher Cogle, MD, chief medical officer at the Florida Agency for Health Care Administration. “We get very little information back from the training sites except for bills.”
The new legislation requires that the sites report on how the money is being used and account for the number and specialty type of their graduating physicians. Dr. Cogle explained that his agency will analyze these data to help a newly established state GME committee shape policy that will align the distribution of Medicaid GME dollars with the state’s needs by geography, specialty, and setting. The new data will help the state track, for the first time, the level of need for primary care residents as compared to other specialties.
The first accountability reports are due to the legislature in December 2025. “We’re doing this in a stepwise fashion,” Cogle said. “Everyone knows that we want to make data-informed decisions for improving Florida’s health.”
i. The 2024 Florida statutes. Online Sunshine website. http://www.leg.state.fl.us/Statutes/index.cfm?App_ mode=Display_Statute&Search_String=&URL=0400-0499/0409/Sections/0409.909.html. Accessed January 13, 2025.
Drawing More Medical Students to Primary Care Requires More than Free Tuition
By Mary Louise Gilburg
Given the urgent need to increase the primary care workforce, some medical schools are taking action through creative financing and admission processes to encourage more medical students to pursue primary care.
Major donations have allowed some medical schools to offer free tuition to all students.i Many hoped that reducing the financial burden of medical school would lead students to choose primary care residencies with lower compensation. Early residency match data from these schools does not support this theory. In the years following the free tuition program at NYU Grossman School of Medicine, for example, the residency match rates for primary care decreased,ii falling below the national average. The data suggest that making medical school tuition free led to a more competitive applicant pool and a more selective admissions process.iii This trend favors high-income applicants who are more likely to specialize, while lower-income students from backgrounds underrepresented in medicine are more likely to go into primary care.
The new NYU Grossman Long Island School of Medicine, however, is one free tuition program that does appear to be increasing the number of primary care residents. Three years into offering free tuition, the primary care residency (including internal medicine, pediatrics, obstetrics and gynecology, and general surgery) match rate is over 60%.iv This high percentage is due to the school’s focus on primary care and community-based primary care, said Dean Gladys M. Ayala, MD, MPH. The school recruits students who have experience working with primary care doctors and working in underserved areas and who are interested in patient advocacy.
In addition to offering free tuition, the medical school is accredited as a three-year program. “When we originally started this, one less year of medical school was already a way of reducing educational debt,” Dr. Ayala said. “The fact that we now have our own endowment to foster our tuition-free status is a great thing for us to continue to work on primary care shortage issues.”
Around 40 medical schools across the country offer a three-year medical education through an accelerated pathway, resulting in much lower lifetime loan repayment. The University of California– Davis School of Medicine offers an Accelerated Competency-based Education in Primary Care (ACE-PC) program that trains students who often have backgrounds that are traditionally underrepresented in medicine, such as students with low socioeconomic status, students from racial and ethnic minority groups, and students from rural areas. The school uses an internal scale that quantifies socioeconomic disadvantage and complies with the state’s 1997 affirmative action ban. Graduates from the program are more likely to practice primary care and serve in communities with high proportions of Medi-Cal patients.
The track was developed to address the primary care shortage, recognizing that one of the biggest barriers to medical education is finances, said Mark C. Henderson, MD, MACP, the associate dean for admissions. UC Davis has four other tracks focused on rural health, urban underserved community health, tribal health, and providing care to California’s Central Valley communities.v
Both the NYU Grossman Long Island School of Medicine and UC Davis select for students who are poised to help alleviate the primary care shortage. “More and more we’ve emphasized the lived experiences of the students — where they grew up, what kind of community that was, [and] what kind of barriers were faced in that community — because these are many of the barriers that our patients face every day,” Dr. Henderson said.
i. Henderson MC, Fancher TL, Murin S. Holistic admissions at UC Davis—journey toward equity. JAMA. 2023;330(11):1037-1038.
ii. Emanuel EJ, Guido M. Free med school tuition won’t solve the shortage of primary care physicians. STAT First Opinion. https://www.statnews.com/2024/04/22/free-medical-school-tuition-primary-care-doctor-shortage. Published April 22, 2024. Accessed January 13, 2025.
iii. Horowitch R. The perverse consequences of tuition-free medical school. The Atlantic. https://www.theatlantic.com/ideas/archive/2024/10/perverse-consequences-tuition-free-medical-school/680321. Published October 22, 2024. Accessed January 13, 2025.
iv. Residency match day results for MD students. NYU Grossman Long Island School of Medicine website. https://medli.nyu.edu/education/md-degree/md-admissions/match-day-results. Accessed January 13, 2025.
v. REACH. US Davis School of Medicine website. https://health.ucdavis.edu/mdprogram/REACH-PRIME/about.html. Accessed January 13, 2025.