The Gender Gap in STEM

Professor Shulamit Kahn (University of Boston) sheds light on the underrepresentation of women in STEM

Like Comment

In view of International Women’s Day, I speak to Professor Shulamit Kahn regarding her research into gender differences in academic sciences.

In a recent paper you observed that the gender retention gap in engineering is largely down to women leaving the labor force entirely and that this exit is highly correlated with child-bearing. Do you believe that this is the case in all STEM disciplines?

I have also found this in biomedicine as well in previous published work based on earlier data [1]. I wondered whether it was true in general, so I just ran some analyses for different STEM fields, to investigate whether there were gender differences in either being out of the labor market or working part-time. All of this analysis controlled for years since PhD, sub-field, age at PhD, quality of PhD dept, race, etc.

In summary I just found: The effects of child-bearing on women with STEM PhDs leaving the labor market or working part time are very large in all fields (with PhD in computer science having the smallest difference.) And PhDs in psych (not clinical) have the largest differences.


Single women without children were NOT more likely than single men (no children) to be out of the labor force/working part time in: biomed, computers, math, physical sciences (incl. chem, phys, geo), economics, engineering and other soc sciences (except econ, psych) but surprisingly in psych (not clinical) even single women without children were 6% more likely to not be working or working part-time than single men.

Married females without children were either equally likely as single men to be out of the labor force/working part-time, in all of these STEM fields except psychology (again). Psychology had a really big gender gap (10 percentage points!) To a much lesser extent, married women in physical sciences and engineering might be slightly more likely to leave (1-3 percentage points).

But when you look at married women with children, that is where you get huge differences. Compared to single men:

  • In biomed: women are 15 ppt (ppt percentage points) more likely to leave the labor force/work part work).
  • In math women are 9 ppt more likely
  • In physical sciences women are 15 ppt
  • In psychology women are 25 ppt more likely. (What’s up with psychology, I wonder?)
  • In other social sciences women are 14 ppt more likely
  • In engineering women are 13 ppt more likely
  • In economics women are 12 ppt. more likely
  • In computers women are 5 ppt. more likely

In contrast married MEN with children are LESS likely to be out of the labor force or working part-time compared to single men (by anywhere between 2 to 9 ppt.)

Do you feel that, in general, academia/industry offers sufficient support to young mothers pursuing careers in STEM?

This is not a question that I have data on, but from talking to people etc. I would say: Academia is beginning to respond in to terms of regularizing rules on paid family leave, but I still hear of universities where family (maternity) leave is far too short and not systematic.

Many people (particularly men) will tell you that the women have chosen to go off the fast track and stop doing cutting edge science. And I personally do know some very highly educated women who realize they want to stay home and take care of the kids.

But I can also tell you of many cases where women complain that their family (maternity) leave is simply too short. I can also tell you of many cases where women would be more than happy to work more hours if only they had husbands who helped more with the childcare and household tasks. Or, women whose paid child care options are inadequate.

And I can tell you of stories both within and outside academia where colleagues frown on women who take maternity leave of more than a month or two.

The answer to your question is therefore, it depends. Different institutions/companies offer different maternity leave policy, on-site child-care, tenure-track stop the clock policies, etc. Also, different institutions/companies have different kinds of cultures and attitudes.

To what extent do you feel that the gendering of labor, such as childcare and elderly care, along traditional lines impacts women’s progression in STEM?

The best thing that a woman who wants children and a STEM career can do for her career is to marry well – i.e. marry a man who believes that child care should be shared, and that the woman’s career is just as important as his. Without that equality, women will continue to find any time-consuming career, whether in PhD-level STEM or in the c-suite (the upper executives) – extremely difficult to combine with children.

In general, evidence shows that women are more heavily underrepresented in the more math-intensive fields, such as engineering and computer science. Do you believe that uptake of these subjects among women is lower as a result of nature or nurture?

Based on my reading of the literature [2], I strongly believe that the answer is mostly – but not exclusively – nurture. By middle school, too many girls believe they can’t do math as well as men (even though in earlier grades they often do better at math than boys.) There are many studies showing that women and minorities who believe that math is a subject that can be learned do quite well. The culture of women not competing also sets in around puberty and yet there are societies where differences are smaller. On the other hand, there are some physical differences (e.g. more boys/men are on the autism spectrum) that are correlated with some math skills, and with some math/computer cultures.

The gender gap in STEM is particularly significant at more senior level positions. Why do you think this is?

This is not particular to STEM, I will note. Again – There is a combination of factors that work differently for different people. Many careers require constant pushing and self-promoting. Women have been taught not to self-promote and this keeps some of them from moving up. (For instance, we’ve found that women are less likely to reapply for NIH grants if they get turned down.) In other cases, women have other priorities and don’t want to spend the time and effort it takes to get to and stay at the top. In some cases, it is the fact that once a woman is slowed down for even a short time (e.g. in child-bearing or rearing), others treat her as a less serious researcher/employee. Or, in some cases, she does lose her networks and connections after a gap, and finds it hard to catch up.

What measures do you think need to be put in place to stem the “leaky pipeline”?

Start very early, in grade school and middle school. Get better math teachers – teachers who teach math intuitively, teachers who believe that math can be learned in the same way that reading can be learned. Get more female math teachers at all levels. Get girls involved in science fairs and math leagues. Universities need maternity leave and child-care resources for graduate students and postdocs as well as for faculty. Industry needs maternity leave and on-site child care. But institutions need to see these kinds of policies as ways to get more from their human resources and advance their institutions, not as ways to be kind to women.


[1] Ginther D, Kahn S. “Does Science Promote Women? Evidence from Academia 1973-2001". In: Science and Engineering Careers in the United States. Freeman R, Goroff D (Eds). University of Chicago Press for NBER, Chicago, IL, USA (2009).

[2] Ceci S, Ginther D, Kahn S, Williams W. Women in Academic Science: A Changing Landscape. Psychol. Sci. Public Interest. 15(3), 75-141 (2014).

Did you enjoy reading this interview? If so, you might wish to check out our interview with inclusivity and diversity champion, Tom Welton:

Plus our panel discussion on the Gender Equality in Science:

Both are free to access. Happy Women’s Day everyone!

Go to the profile of Hannah Coaker

Hannah Coaker

Contributor, Future Science Group

No comments yet.