To examine factors that affected the pay and career advances of physicians, Robert Wood Johnson Foundation (RWJF) Physician Faculty Scholar Reshma Jagsi (’09) wanted to study as homogeneous a sample as possible. So Jagsi, MD, DPhil, looked at data on a group she termed “the cream of the crop”: physicians who in 2000-2003 received prestigious, early-career research grants from the National Institutes of Health (NIH).
In 2009 and 2010, Jagsi and colleagues sent surveys to these physicians—now mid-career—seeking details about salary, age, race, medical specialty, work hours, publications, additional grants and degrees, promotions, marital and parental status, and more. After adjusting for all factors, they found males in the sample were still paid $12,000 more a year than females.
“People point to a lot of possible reasons for pay disparities, so we examined a population in which you would be least likely to pick up gender differences in salary that were due to differences in the abilities or motivation of the men and women studied,” says Jagsi, associate professor of radiation oncology at the University of Michigan Medical School. “After we adjusted for a host of factors that could explain differences, there was still a pay disparity of $12,000 a year. That’s more than $350,000 over the course of a career which, where I live, can buy a house or send your kids to college.”
Results of the study, “Gender Differences in the Salaries of Physician Researchers,” appear in the June 13 edition of JAMA, the Journal of the American Medical Association. Funding for the study came from RWJF, the University of Michigan Office of the Vice President for Research, and the National Institutes of Health.
When Jagsi and colleagues first looked at what had become of the 1,853 recipients of 2000-2003 NIH career development grants, they noticed that more men than women had received independent research grants after their federal grants. “Women in this population were highly select, highly motivated, and should have been largely successful at getting those independent grants, but were less likely to get them than their male peers,” Jagsi says. “That’s what sent us down the path of doing the survey: We wanted to see what kinds of things not available in publicly-recorded data were potentially influencing success and how outcomes other than grant attainment, such as salary, might differ as well.”
Of the original recipients still alive, Jagsi and colleagues found mailing addresses for 1,729. The recipients were sent a survey with 39 questions about factors that could influence salary, including productivity measures such as how much time they spent on research, how much they had published in peer-reviewed publications, and what academic rank and leadership roles they held at their institutions. “We were very grateful” that 71 percent of the recipients answered the survey, Jagsi says. Limiting their analysis to physicians who reported their salary and who were practicing at U.S. academic institutions, Jagsi and colleagues wound up with a study population of 800 physicians, of which 247 were women and 553 men.
Among those 800, 76 percent were white, and mean age was 45 years old. “Nearly half held an additional graduate degree, and the majority held the rank of associate professor, with no statistically significant differences by gender,” according to the study.
Men in the cohort, however, were more likely than women to be married and to have children. But “in contrast to some other studies, we did not observe any interaction between gender and parental status,” the study says. “Even women without children had lower pay than men.”
Gender differences were also significant in terms of medical specialties. Among women in the cohort, 34 percent were in the lowest-paying category of specialties compared to 22 percent of men, while three percent of women were in the highest-paying category compared to 11 percent of men. When study authors compared men’s and women’s mean salaries without controlling for specialty, the difference was more than $32,000, says Jagsi—a salary of $167,669 for women and $200,433 for men.
“When we controlled for specialty, the difference dropped to about $17,000,” she says. “But even when we controlled for all other factors, it still was about $12,000. We really didn’t expect to find such a substantial unexplained difference.”
Jagsi and colleagues also processed the survey data using Peters-Belson analysis, an approach common in labor economics. Using this approach, the researchers looked only at the men in the group and created a model for what predicted the salary among them, Jagsi says. “Then we used that model developed for the men and applied it to the women to see what the women would be expected to earn if they had all of their own characteristics—like the number of hours they worked and specialty—but were men. And with this method, we again found about a $12,000 unexplained difference.”
“Given that we still found a difference after controlling for so many other factors, we think it’s important to consider the role of unconscious biases” in the pay disparity, Jagsi says. “There is considerable evidence from psychology that women with identical accomplishments are rated less highly than men.”
The study authors also considered the implications of academic advancement in the pay gap. Physicians in the study cohort “all got the NIH award at roughly the same point in their careers, so they started off at the same point—and then the women seemed not to advance quite as fast as the men,” Jagsi says. “If differences in pay are the result of gender biases resulting in women not getting promoted at the same rate as their male peers, then perhaps saying they earn less because they’re less likely to have leadership positions—‘Oh, well, men earn more than women because they’re more likely to be department chairs’—isn’t necessarily a justification. There may be women who ought to have been department chair but didn’t get the position in part due to bias. So by controlling for things like academic rank and leadership positions, our study was designed to be conservative in estimating the size of the disparity.”
Both Jagsi and senior author Peter Ubel, MD, former associate director and co-director of the Michigan RWJF Clinical Scholars program, caution against attributing the salary difference to conscious discrimination. “For all we know, women are paid less in part because they don’t negotiate as assertively as men, or because their spouses’ jobs make it hard for them to entertain competing offers,” Ubel says. But whatever the reason for the pay disparity, Ubel says, “academic medical centers should work to pay more fairly. A person’s salary should not depend upon whether they have a Y chromosome.”
To Jagsi, discovering “that men and women doing similar work are paid quite differently in this cream-of-the-crop sample is both surprising and disturbing. I hope these findings will help inform policy discussions on how to address these disparities. Some people have said to me, ‘Physicians are so well paid, is $12,000 a year really that big a deal?’ But I think you’d be hard pressed to find anyone who would say that $350,000 over a career is trivial.”
Results of the study, “Gender Differences in the Salaries of Physician Researchers,” appear in the June 13 edition of JAMA. Co-authors with Jagsi on the study were: Ubel, MD, of Duke University, Durham, N.C.; Kent A. Griffith, MS; Abigail Stewart, PhD; Dana Sambuco, MPPA; and Rochelle DeCastro, MS, of the University of Michigan.