The most commonly cited statistic in the debate over equal pay—that American women on average make about 80 cents for every dollar American men make—has come to signify a lot more than just a financial disparity between genders. These days, it also connotes inequity in the workplace, the motherhood penalty, gender-based discrimination, a more limited path to top-paying and leadership positions. It encapsulates all the difficulties of being a working woman.

In fact, the issue is even more complicated than it lets on. A common rebuttal to the 80-cents figure is that, when controlling for the career choices that men and women each tend to make, the pay disparity shrinks. In other words, men and women are being paid differently for doing different work.

This concept—economists call it “gender-based occupational segregation”—is something that researchers identified as a contributing factor to pay disparities as early as the 1960s. Economists have estimated that adjusting for these occupational differences brings the gap down to about 9 percent, and recent data on white-collar professionals puts the adjusted gap at even lower than that.

This interpretation of occupational segregation, though, is somewhat controversial, since it puts the onus on women to “choose” the jobs that pay more. Efforts to alter this dynamic have been made, for instance, in STEM fields, where a lack of female representation has resulted in initiatives to get young women interested in science and coding as well as to create pathways for them to attain those high-paying jobs.

Despite these efforts and the fact that more women are seeking higher education than ever, gender-based occupational segregation has been dissolving at a glacial rate in recent decades: In the U.S., gender segregation declined by over 10 percent in the 1970s and 1980s. In the 1990s and 2000s, it has only decreased by 3 percent. According to researchers, more than half of men or women working today would need to change occupations for the distribution to be equal.

Perhaps there would be more parity in the professional world if more was known about why and how men and women come to choose the career paths they do. A new NBER working paper addresses this question from the angle of considering that men and women have different aptitudes and skills, and to an extent, this shapes their careers. The results are surprising, partly because it’s commonly assumed that women’s skills are not valued and lead to low-paying work.

“There is unlikely to be a single account of gender-based occupational segregation,” caution the authors of the paper, Michael Baker and Kirsten Cornelson, an economics professor and a PhD student at the University of Toronto, respectively. But Baker and Cornelson note that male and female workers tend to have different strengths and weaknesses. For example, research from various fields has shown that, on average, women have better hearing abilities and motor dexterity, while men have better hand-eye coordination and faster reaction times. To see whether these differences have any association with labor market outcomes, Baker and Cornelson mapped differences like those (as documented by copious previous research done by others) onto the attributes required by different occupations as defined by the Dictionary of Occupational Titles, a publication of 13,000 job descriptions put together by the Department of Labor

They found that these differences in abilities play a significant role in why women and men are each likelier to choose certain professions over others. For example, men tend to have a much higher tolerance for noise than women, which may make them more suited for, say, a construction job. Women are documented to have better finger dexterity, which might make them better dentists or bank tellers.

“In all of these cases, we found that there was a higher probability of female employment in occupations that relied more on skills for which women have a documented advantage, and vice versa for men,” say Baker and Cornelson. They say that the male-female differences they looked into accounted for about a quarter of the gender-based occupational segregation in 1970 as well as in 2012.

There are countless cultural reasons why, say, women are more reluctant to enter blue-collar professions or why men are more reluctant to enter predominantly female occupations like nursing. But, Baker and Cornelson note, the inclination to take a job that plays to one’s inborn strengths and preferences matters too. To a degree, they say, it can help explain some of that reluctance.

Interestingly, though, the job-sorting that results from these gender differences has an unexpected effect on the pay gap. For instance, the inborn edge that women tend to have in manual dexterity, to a degree, can nudge them toward higher-paying jobs, such as those in dentistry and medicine. That’s just one example, but adding up all of those tendencies across every male and female worker in the U.S., gender-based occupational segregation has actually reduced the pay disparity between men and women. “The result,” says Baker, “could be viewed as surprising because the debate about gender-based occupational segregation is often framed in terms of low-paying ‘female jobs’ and high-paying ‘male jobs.’”