Physician Workforce Diversity by Race and Ethnicity
Introduction
Racial and ethnic disparities in health outcomes remain persistent in the United States, driven by inequitable access to and utilization of health care services and broader social and economic factors that reflect historical and ongoing racism. One factor that can mitigate these disparities is racial concordance between physicians and patients, that is, when providers and patients share the same racial or ethnic background. Research suggests that patient and provider racial concordance may be linked to increased visits for preventative care, greater treatment adherence, and lower emergency department use. One study found that greater representation of Black primary care physicians was associated with increased life expectancy and lower mortality among Black people. KFF 2023 survey data show Black, Hispanic, and Asian adults who have more health care visits with providers who share their racial and ethnic background more frequently report having positive and respectful interactions. Despite these benefits, many people of color face challenges accessing racially concordant providers. KFF survey data show that most Hispanic, Black, Asian, and American Indian and Alaska Native (AIAN) adults say that fewer than half of their health care visits in the past three years were with a provider who shared their racial or ethnic background. While data show that adults of color are more likely than White adults to prefer a provider of the same race or ethnicity, they are significantly less likely to have one and are more likely to have difficulty finding one.
Recent policy changes may exacerbate challenges to obtaining racially concordant care for those who value it most. In 2023, the U.S. Supreme Court effectively ended race-conscious admissions in higher education, overturning decades of precedent supporting affirmative action. Additionally, as one of his initial actions in office, President Trump issued executive orders eliminating federal diversity, equity, inclusion, and accessibility (DEIA) related programs and actions in the federal government and among federal contractors and grantees. These actions may reduce matriculation rates for students of color and reverse progress in diversifying the health care workforce to reflect the population it serves. For instance, research indicates that the Supreme Court’s 2023 ruling has already led to a decline in the number of Black, Hispanic, and AIAN students entering medical school. Among Black, Hispanic, AIAN, and Native Hawaiian and Pacific Islander (NHPI) medical school enrollees, there was a decline in matriculants between 2023 and 2024. Despite there being more Black and Hispanic applicants in 2024 compared to the previous year, the number of matriculants from historically underrepresented groups declined. It’s unclear how much impact the Supreme Court ruling has already had on medical school matriculation. However, the ruling has the potential to further decrease the diversity of the future physician workforce.
This brief provides an overview of the racial and ethnic composition of physicians compared to the total population at the national and state level based on KFF analysis of 2023 Association of American Medical Colleges (AAMC) Physician Workforce data and American Community Survey data. This analysis shows that Hispanic, Black, AIAN, and NHPI people were underrepresented among physicians relative to their share of the population, with the widest gap observed among Hispanic people who comprised 20% of the total population but only 7% of the total physician workforce. The pattern of underrepresentation held true across most states for Hispanic and Black people, with particularly large gaps for Hispanic people. Asian people accounted for a higher share of physicians than their share of the total population nationally and in all states; in most states, the share of White physicians was similar to or higher than their share of the total population. Data were not available for AIAN and NHPI populations in many states due to small numbers.
Racial and Ethnic Distribution of Physicians
Nationally, Hispanic, Black, AIAN, and NHPI people were underrepresented among physicians relative to their share of the population with the widest gap observed among Hispanic people. While one in five people in the U.S. population was Hispanic, they accounted for just 7% of the physician workforce (Figure 1). Similarly, 12% of the population was Black compared to 6% of the physician workforce. AIAN and NHPI individuals were also underrepresented among physicians compared to their share of the overall population. However, the absolute differences were small, as both groups make up less than one percent of the total population and the physician workforce. In contrast, White people accounted for similar shares of the total population and physician workforce, and Asian people accounted for a larger share of physicians (19%) than their share of the population (6%).
Across most states, Hispanic and Black people made up a smaller share of physicians compared to their share of the total population, with particularly large gaps for Hispanic people. In 45 states and DC, Hispanic people accounted for a smaller share of providers than their share of the population, as measured by a more than one percentage point difference (Figure 2). In seven states, the difference between the share of the population and providers who are Hispanic was over 15 percentage points. These tended to be states with the highest shares of Hispanic people, including New Mexico, California, and Texas. In New Mexico, Hispanic people made up nearly half (48%) the population but accounted for 17% of providers (a difference of 31 percentage points). In California and Texas, about 40% of the population was Hispanic compared to 7% and 13% of physicians, respectively. In only five states (West Virginia, Vermont, Maine, Mississippi, and Ohio) did Hispanic people account for a similar share of the population and the physician workforce, with a difference of less than one percentage point.
Similarly, in 35 of 50 states and DC, Black people made up a smaller share of physicians compared to their share of the population. In five states and DC, the difference between the share of the population and providers who are Black was 15 percentage points or higher, with the largest gap of 26 percentage points in DC and Mississippi (Figure 3). In DC, 17% of providers were Black compared with 43% of the population, and in Mississippi, Black people made up 11% of physicians compared with 37% of the population.
Data were insufficient for comparisons for AIAN and NHPI people in most states, but where data were available, AIAN and NHPI people were underrepresented among physicians relative to their share of the population. This pattern was most prominent in states where the majority of AIAN and NHPI people live. For example, In Alaska, AIAN people accounted for 14% of the population, seven times higher than their share of physicians (2%). Additionally, in New Mexico, AIAN people made up 9% of the population compared to 2% of providers, and in South Dakota they comprised 7% of the population compared with 1% of providers. Similarly, in Hawaii, NHPI people made up 10% of the population compared to 2% of physicians. In most remaining states, AIAN and NHPI people accounted for both a small share of the population and providers, so absolute differences in their shares were small.
Asian people accounted for a higher share of physicians than their share of the total population in all states. The largest differences were in Illinois, California, Delaware, and Texas. In Illinois, Asian people made up more than one in four physicians but accounted for only about 5% of the population (Figure 4). There were sizeable differences in the states with the highest shares of Asian people. In Hawaii, 37% of the population was Asian compared with 43% of providers, and in California, 15% of the population was Asian versus 32% of providers. The differences were smaller in other states, like Alaska, Montana, and Vermont, where Asian people accounted for a small share of both providers and the population.
In most states, the share of White physicians was similar to or higher than their share of the total population. States where White people were most underrepresented among physicians compared to their share of the population included West Virginia (67% vs. 90%), North Dakota (67% vs. 83%), and Michigan (58% vs. 73%) (Figure 5). The pattern in these states reflected higher representation of Asian people among physicians. States where White people were most overrepresented among physicians compared to their share of the population included New Mexico (56% vs. 37%), Alaska (77% vs. 58%), and Mississippi (70% vs. 55%).
A similar pattern was observed across medical specialties, with Hispanic and Black individuals underrepresented in more specialties than other racial and ethnic groups. Specialties providing primary care—such as family medicine, pediatrics, and obstetrics/gynecology—tended to better reflect the racial and ethnic distribution of the population. Among all specialties, providers in nephrology and interventional cardiology had the least racial and ethnic representation compared to the national population. The Hispanic population was the most underrepresented group in more than half of the specialties analyzed (27 out of 51 specialties). The Black population was the most underrepresented group in 9 out of 51 specialties.
Methods |
KFF collected U.S. Physician Workforce data from the Association of American Medical Colleges (AAMC) from the dashboard for 2023, aggregated by specialty and location. Although race/ethnicity data were provided as “alone or in combination,” categories were treated as exclusive in the analysis since a “multiracial” option was provided. Respondents with unknown race/ethnicity were proportionally distributed within each state to all other categories. Due to values being masked due to small cell sizes, only the Asian, Black, Hispanic, and White categories were used in the provider specialty analysis, where only specialties with more than 5,000 active physicians were included. For comparison to the population at large, 5-year estimates from the 2023 American Community Survey were used to determine the racial and ethnic distribution of the total population for each state, using single-response values for not Hispanic or Latino, and any response for Hispanic or Latino. To determine how well the active physician population matches the national population for each specialty, an index of dissimilarity was used, proportional to the total percentage point difference in the race/ethnicity share of providers and the country overall. To assess the robustness of this method, Kullback–Leibler divergence was calculated for each specialty, a measure of statistical distance between the race/ethnicity distribution among physicians and that of the national population. Rank order among the specialties was similar for both methods. |