In a new study published in JAMA, my colleagues and I found that even after accounting for productivity, women working as physician researchers at American Medical Schools are paid $13,000 less per year than their male colleagues, a difference that amounts to hundreds of thousands of dollars over the course of their careers.

But does this difference stand as evidence of discrimination?

Many claims of gender inequity in pay have suffered from an apples vs. oranges problem.  For example, consider gender disparities across different careers.  Many traditional male careers, like construction work, pay better than traditionally female careers, like nursing and teaching.  It’s plausible that these disparities result, at least in part, from societal bias about how relatively important it is for men and women to make enough money to provide for their families.  However, these disparities could also result from more justifiable factors.  Maybe the physical demands of the work differ in important ways, or perhaps the marketplace is simply responding to supply and demand.

Medical experts have long noticed gender disparities in physician pay.  Traditionally male fields like neurosurgery pay substantially more than fields preferred by more women, such as general pediatrics.  If women are voluntarily choosing lower paying fields—perhaps for lifestyle reasons or maybe because they don’t value money as much as men do—then it’s arguable that we should not fret over pay disparities.  It’s America, after all, where people have the right to choose.

In our new study however, led by Reshma Jagsi a colleague of mine at the University of Michigan, we found disparities even after accounting for the different career choices and trajectories of male and female physicians.  We focused our research on mid-career physicians who had chosen to pursue primarily research careers.  All of the people we surveyed were physicians who had received training grants from the National Institutes of Health (the NIH) to pursue research careers.  These NIH grants (called K awards) are extremely competitive, meaning we were surveying the best of the best.  It also means we were surveying people who would largely be paid according to their research accomplishments, not according to their clinical work productivity.

By the time we surveyed these folks, it had been almost ten years since they had received their K awards.  Most of them had remained in academic medical centers, seeing patients part-time and conducting research the majority of the time.  Our initial analysis showed that men were making around $32,000 more per year than women.

But wait!  Remember, men are more likely to be surgeons than women.  Did that account for these differences?  We took account of this fact through a process called “statistical adjustment”, and found that the salary disparity, while smaller, still persisted even after accounting for their medical specialty.

What about productivity, you ask?  Did the men work harder?  Accomplish more?  Well, we asked these physicians about their publication success and about their ability obtaining additional research grants; we asked them whether they had been promoted or had taken on any kind of leadership role at their medical school; we even asked them how many hours per week they worked.  In other words, we did our best to measure the kinds of things that ought to influence how much a boss pays an employee.

And we still found that women made substantially less than men.

Is this a matter of sex discrimination?  Not necessarily.  Research suggests that women negotiate for themselves less actively than men. (See Linda Babcock’s wonderful book—Women Don’t Ask—for examples of this research.)  In academic medicine, in fact, division chiefs and department chairs typically give raises to their faculty not based purely on measurable accomplishments but, just as often, based on the need to compete with an outside university trying to lure the young faculty member away from them.  Perhaps women were simply less willing or able to entertain such outside offers.

Whatever the cause of these disparities, we should strive to pay people according to their performance, not according to their willingness to play hardball.  No one’s income should depend on whether they have a Y chromosome.

Peter Ubel is a physician and behavioral scientist who blogs at his self-titled site, Peter Ubel and can be reached on Twitter @PeterUbel.  He is the author of Critical Decisions: How You and Your Doctor Can Make the Right Medical Choices Together.