But the study has a major flaw: A significant part of the data the finding depends on didn't identify whether the doctors were primary care doctors or specialists. Physicians in specialties make more money than primary care doctors, and black doctors are underrepresented in specialties.
A big caveat to any study of doctors' incomes is that neurosurgeons, pediatricians and interventional cardiologists are all doctors, but they are not paid the same. Comparing doctor pay without knowing which specialty a physician works in opens the possibility of attributing a pay gap to race or gender that might actually reflect differences in specialization. For context, consider this: Physicians in specialties make, on average, $284,000 a year, while primary care doctors make $195,000, according to the Medscape Physician Compensation Report 2015.
William Weeks, a health economist and physician at Dartmouth College's Geisel School of Medicine, roundly dismissed the study's findings for that reason. His research has found no evidence of a racial pay gap among doctors.
"This is really not good research," Weeks said. "It's a poor design. It's a poor study. It should be ignored."
Anupam Jena, a physician and health economist at Harvard Medical School who oversaw the study, acknowledged that one of the analyses couldn't account for specialties but said his team found the same racial disparities in a smaller database that did. He defended the research, saying his team looked at a larger population than Weeks's research.
"The reason why prior work has not found much [difference], we believe, is it was simply underpowered; there simply weren’t enough observations on black physicians to get anything meaningful out of it," Jena said.
To increase the number of black physicians in the sample, the researchers used two data sets -- the bulk of it data on physician incomes taken from census data. For 40 percent of the doctors in the sample — including almost half of the white male doctors — the researchers had to estimate their incomes by using the median income above an upper threshold in each state. This meant they were eyeballing physicians' salaries by using a mix of data that would have included highly paid people who were not doctors. The census data also contained no information on whether the doctors were primary care doctors or specialists, which could lead to an apples-to-oranges comparison of specialists with primary care doctors.
The second data set, from a survey that tracked more than 17,000 physicians, did include specialty information, but there were a small number of black physicians overall. Jena's team found the same patterns as in the larger data set, even after taking into account specialties, but Weeks critiqued some of the methods used.
Jena said that the second analysis finds the same racial gaps as the first one — white men out-earn black men; women — of both races — earn less. He also argues that despite any shortcomings, this kind of study is valuable because it reveals the broad landscape of physician pay by race. The gender pay gap has already been shown.
"It’s extraordinarily important to know what the differences in income are between these two groups. Some of these things you’d otherwise want to account for are things that mediate the problem," Jena said.
Salary disparities might stem from doctors of different races seeing more low-income patients — which may be something they do by choice, or might have more to do with different opportunities. Black doctors might choose not to go into highly paying specialties because they aren't drawn to them or because they want to start paying off their medical school debt back more quickly — or because of discrimination.
On this point, Weeks agrees.
"The key question is, do women or black or minority [doctors] have access — can they get into these higher-paid sub-specialties?" Weeks said. "That's a really different question and a really important one."