Differences In Debt Among Postgraduate Medical Residents By Self-Designated Race And Ethnicity, 2014-19

Louisa W. Holaday, Jasmine M. Weiss, Sire D. Sow, Hector R. Perez, Joseph S. Ross, Inginia Genao

Research output: Contribution to journalArticlepeer-review

7 Scopus citations


The US physician workforce does not reflect the racial and ethnic makeup of the country’s population, despite efforts to promote diversity. Becoming a physician requires significant time and financial investment, and populations that are underrepresented in medicine have also been excluded from building wealth. Understanding the differential burden of debt by race and ethnicity may inform strategies to improve workforce diversity. We used 2014-19 data on postgraduate resident trainees from the Association of American Medical Colleges to examine the association between race and ethnicity and debt independent of other demographics and residency characteristics. Black trainees were significantly more likely to have every type of debt than the overall sample and all other racial and ethnic groups (96 percent of Black trainees had any debt versus 83 percent overall; 60 percent had premedical education loans versus 35 percent overall, and 50 percent had consumer debt versus 25 percent overall). American Indian/Alaska Native, Hispanic, and Native Hawaiian/Pacific Islander trainees were more likely to have debt compared with White and Asian trainees. Overall, debt prevalence decreased over time and varied by specialty; however, for Black trainees, debt decreased minimally over time and was stable across specialties. Scholarships, debt relief, and financial guidance should be explored to improve diversity and inclusion in medicine across specialties.

Original languageEnglish
Pages (from-to)63-73
Number of pages11
JournalHealth Affairs
Issue number1
StatePublished - Jan 2023


Dive into the research topics of 'Differences In Debt Among Postgraduate Medical Residents By Self-Designated Race And Ethnicity, 2014-19'. Together they form a unique fingerprint.

Cite this