When Scientists Choose Motherhood
A single factor goes a long way in explaining the dearth of women in math-intensive fields. How can we address it?
Three major sets of factors have been offered to explain the dearth of women in math-intensive fields: ability differences; occupational and lifestyle preferences; and sex discrimination. Elsewhere we have extensively reviewed the evidence for several of these issues; here we provide a summary.
Ability differences: Scholars once thought sex differences that favored men on tests of quantitative ability were responsible for the shortage of women in math-based fields. Although there are no systematic sex differences in average scores on tests such as the SAT mathematics (SAT-M), many more men score in the very top range. For example, among the top 0.01 percent (1 in 10,000) of scorers in nonrandom samples, there are about 4 men for every woman. College admissions committees may admit more men because of their higher SAT-M scores, setting in motion a cycle of women’s attrition.
Yet this cannot be the whole story or even a large part of it. Starting about two decades ago, women began gaining on men in math ability and participation. By 2005, men and women were almost equally represented among college math majors, and women tend to get better grades in math courses. But far fewer women than men enroll in math-based Ph.D. programs. The GRE-Quantitative scores of graduate students in math-intensive fields at our own university are very high across the board. These students come from the top end of the ability distribution, and the test actually underestimates their ability due to ceiling problems (there are not enough very hard questions to distinguish the truly exceptional from the merely talented). So, since fewer women score in the top range on tests used for admission to Ph.D. programs, this fact may be responsible for some of the shortage of women in math-intensive careers.
As compelling as this argument seems, however, there are several problems with it. Females outperform males in math classes throughout schooling, including in college. Surely graduate-admissions decision makers take women’s higher GPAs into account. Also, even if among the top 1 percent of scorers there were 2 males for every female, there should still be more women in math-based disciplines, because we do not see anything close to a 2-to-1 ratio of men to women in these careers. Recall that 12 percent or less of full professors in these fields are women. Something more than scoring at the right tail is responsible.
Some researchers argue that those who succeed in mathematical fields come from more rarefied strata than the top 1 percent. David Lubinski, Camilla Benbow and their associates at Vanderbilt University have shown that those in the top quarter of the top 1 percent of scorers outperform those in the bottom quarter of the top 1 percent in obtaining tenure-track jobs, publishing articles and patents, and other indicators of excellence and success. If extremely high math ability is important to success, and if many more men possess this extreme ability than do women, perhaps it is responsible for observed ratios of men to women in math-based fields. Although further investigation might show a correlation between success and scores in the far-right tail of the range, no one has demonstrated a causal relation between the two factors. We do know that math-talented women are less likely than equivalently math-talented men to enter mathematical professions. In other words, many fewer women than men choose mathematical fields—even when they have comparable math scores.
There is no direct evidence that men’s math-score advantage explains this shortage of women. Forty-five percent of undergraduate degrees in mathematics go to women, as do 29 percent of Ph.D.s, suggesting that whatever these women’s math skills, they are compatible with very high levels of achievement. Thus, we believe mathematical differences between the sexes are not primary factors in women’s underrepresentation in math-heavy fields.
Career preferences and lifestyle differences: If cognitive differences cannot explain most of the shortage of women in these careers, what about sex differences in career preferences and lifestyle choices? Surveys have documented that females, starting at a young age, are more interested in careers that involve living things—such as medicine, biology, animal science and psychology—than fields such as computer science, mathematics, physics and engineering. Adolescent girls seldom name engineering and computer science as desired careers, whereas nearly a quarter of adolescent boys do. Unlike some researchers, we are not overly worried by these findings, because careers in biology, medicine and veterinary science seem as valuable and satisfying as those in math-based fields. It would be troubling if adolescents declined to try such professions on the basis of faulty information about what is possible, but as long as they are doing so to pursue careers they perceive as more rewarding, society still benefits from these young people’s talents.
A related factor concerns life-course differences between the sexes. In surveys of graduate students, Lubinski, Benbow and their students found that female graduate students viewed a full-time career as “important” or “extremely important” about as often as did their male counterparts (77 percent versus 81 percent, respectively). However, when it came to the importance of temporarily having a part-time career, significant sex differences emerged (31 percent versus 9 percent, respectively), as well as for always having a part-time career (19 percent versus 9 percent, respectively).
Such life-course preferences can lead to differences in research productivity and hours spent at the office, reflecting differing priorities in optimal life-work balance. Lubinski studied the amount of time that nearly 2,000 33-year-olds, who were in the top 1 percent of quantitative ability during their adolescence, spent on career-related work. He found that roughly twice as many high-aptitude men as women reported working at their jobs more than 50 hours per week. Other surveys underscore this male advantage in working very long hours.
Sex discrimination in publishing, funding and hiring: A frequent claim is that women are derailed by sex discrimination in publishing their work, obtaining grant funding and being hired. However, although these forms of discrimination may have played important roles historically, none of these causes can explain today’s underrepresentation. In an article in the Proceedings of the National Academy of Sciences of the U.S.A., we reviewed the evidence and concluded that such discrimination is not responsible for the current dearth of women. Consider one example of the research we synthesized. In 2004 and 2005, a National Research Council committee surveyed U.S. university departments and faculty in a number of math-intensive fields about their interviewing, hiring and promotion records. It found that women applicants were actually more likely to be interviewed and offered tenure-track jobs than were their male competitors, and that there were no differences in tenure and promotion rates for women and men. A number of other analyses have reached the same conclusions. In mathematics, only 20 percent of applicants for tenure-track posts were women, but 28 percent of those invited to interview were women, as were 32 percent of those offered positions.
The picture is much the same for funding and publishing. Women scientists are as successful as men at publishing work and at earning grant funding, according to analyses of hundreds of thousands of grant applications submitted throughout the United States, Canada, the United Kingdom and Australia.
Thus, the shortage of women in mathematical fields is not the result of discriminatory hiring, publishing and funding, nor can it be explained away on the basis of ability differences. Some portion of the dearth of women in math fields can clearly be traced to differences between the sexes in career preferences. Women’s greater desire for lifestyle flexibility, reflecting differing ideas about work-life balance and different expectations regarding responsibility for raising children and working in the home, also plays a role. This latter point leads to what we see as the single most important factor in explaining women’s underrepresentation: a desire for children and family life.