Reply to Drago: Culture and history are important in understanding the low number of women

Drago (1) makes a number of good points about the need to consider historical trends and culture in understanding women's underrepresentation in math-intensive fields. We agree with some of these points and have no position on others. Nevertheless, we want to point out that the main point—that we should consider historical and cultural influences in sex segregation—is something that we have done consistently in our publications on this topic, including this one (2–4). In online SI Text accompanying our article, we reiterated a point we have made in numerous past articles, namely, that sex cannot be considered in isolation from culture or history, and we discussed a number of historical trends. In SI Text, we wrote: “Lest these data be assumed to reflect innate, biological superiority of males, it deserves noting that there is large variance in cross-cultural analyses, with females outperforming males at the right tail in some countries, and the best predictor of international sex differences is the degree to which its citizens exhibit implicit gender-science stereotypes.” This goes well beyond the Asian countries the writer mentions because international trends in the production of top female mathematical talent have shifted from Asian to non-Asian countries over the past few decades several times. For example, among the top 25 female medalists in the 1992–2007 Putnam Mathematical Competition, viewed by many as the top mathematics competition in the world, small Eastern European countries, such as Romania and Bulgaria, supplied a larger number of female medalists than did the United States or Asian countries, despite being less than or equal to 1/15 in size (5, table 1).

In our article (2), we argued that family formation and attendant difficulties are the single most important cause of women's underrepresentation in the fields of science and engineering examined. The number of women in graduate school who initially plan to pursue academic careers in science declines by 30% over the course of their doctoral study, and the number of men declines by 20% (6).* Thus, women are 50% more likely than men to opt out at this stage. Whereas men drop out for reasons having to do with their perceptions of the unrelenting work involved, women drop out primarily because of family concerns. We provided references to support this claim, and none of the writer's points alter this evidence.