Be prepared for a deluge of discussion in the websiteosphere. There’s a new paper in The Proceedings of the National Academy of Sciences (PNAS) by Stephen Ceci and Wendy Williams (you can haz free access from the link) that’s going to be controversial, for it tackles a subject that has become a minefield in academia: gender bias.
It has long been the conventional wisdom in science that women are discriminated against in publication (female-authored manuscripts are said to be rejected more often than those written by males), in funding (female-authored grants are less successful), and in hiring. This, it is said, accounts for the underrepresentation of women in some areas of science, particularly math-related areas like physics. As Ceci and Williams note, “among the top 100 US universities, only 8.8-15% of tenure-track positions in many math-intensive fields (combined across ranks) are held by women, and female full professors number <10%”.
The authors simply surveyed a number of studies (and there are many) addressing these claims. It turns out that, since at least the late 1970s, the claims of biases in publication, funding, and hiring aren’t borne out by the data: the usual citations of biases are based on only a handful of studies whose results have not been replicated by other work. In general, the authors show that in the three areas of “bias” mentioned above, women are on par with men. An occasional study will show inequities, but these seem to go against men as often as against women. (Their conclusions apply not just to math-related areas, but to science in general.)
The authors are perfectly aware of the political implications of their findings. As Steve Pinker has pointed out repeatedly, the academy does not welcome results like this, and researchers who produce them can be subject to accusations of sexism (although one of the authors of this paper is a woman) and of justifying or enabling further discrimination against women. So when you read the paper, notice that it is loaded to the gunwales with caveats—so many of them that they become almost annoying, like repeatedly hearing “We’re so sorry to have to tell you our findings.” But it’s important to know what they’re saying. First, they’re not saying that discrimination against women in funding, hiring, and publication never existed. It almost certainly did thirty years ago; the authors are talking about more recent studies. And discrimination back then can still affect inequities in the representation of females who began their academic careers, as I did, in the 1970s.
Most important, the authors are not saying that there are no factors that hold women back from achieving parity with men. While ruling out discrimination in hiring, funding, and publication, they note that the lack of parity results from other factors, some of them voluntary, others not. Much of it, as you might expect, comes down to reproduction:
Despite frequent assertions that women’s current underrepresentation in math-intensive fields is caused by sex discrimination by grant agencies, journal reviewers, and search committees, the evidence shows women fare as well as men in hiring, funding, and publishing (given comparable resources). That women tend to occupy positions offering fewer resources is not due to women being bypassed in interviewing and hiring or being denied grants and journal publications because of their sex. It is due primarily to factors surrounding family formation and childrearing, gendered expectations, lifestyle choices, and career preferences—some originating before or during adolescence (3, 50, 54, 58) (SI Text, S9)—and secondarily to sex differences at the extreme right tail of mathematics performance on tests used as gateways to graduate school admission (SI Text, S10).
Ceci and Williams don’t talk much about differences in performance on grad-school exams in math (you’ll remember that when Larry Summers claimed that the lower success of women in math and physics reflected innate biological differences, it helped bring him down as president of Harvard). But the authors do note that while this difference in test performance is statistically significant, it can’t account for the disparity in the number of women in math-related fields. That’s because while there is a lower proportion of women in the upper tails of the distribution of test scores, the proportion of female faculty in math-related fields is much lower than even that proportion.
The authors posit that the lower representation of women among math-related faculty, then, is due to women’s lack of those resources necessary for professional success (“resources” include positions at research-oriented colleges rather than teaching colleges and two-year institutions). The reproductive component is seen as important:
Given equivalent resources, men and women do equally well in publishing. A key issue, separable from sex discrimination in manuscript evaluation, is why women occupy positions providing fewer resources and what can be done about this situation. This situation is caused mainly by women’s choices, both freely made and constrained by biology and society, such as choices to defer careers to raise children, follow spouses’ career moves, care for elderly parents, limit job searches geographically, and enhance work-home balance. Some of these choices are freely made; others are constrained and could be changed (3).
I’m not a sociologist, and hardly an expert in this area, so all I can say is that their data seem sound, regardless of the conclusions and prescriptions. And why, exactly, do the authors see their findings as important? Because, they claim, we can’t fix the problem of gender inequity unless we correctly identify its cause. This is what they say in the analysis of “publication”, but it applies to grants and hiring as well:
However, a secondary issue is whether resources themselves are, in fact, evenly distributed between the sexes. The answer is that they are not, for a complex constellation of reasons, such as women being more apt to occupy teaching-intensive positions, part-time positions, etc. Thus, the attention devoted to righting perceptions of sex discrimination in reviewing of manuscripts, which as we show, does not in fact exist (SI Text, S2), focuses on a spurious issue and detracts from the very real problem that does plague women in publishing—the fact that women more often than men lack resources necessary to produce high-quality work.
At the end of the paper, Ceci and Williams suggest a number of solutions for giving women equal access to resources, including changes in the tenure system to allow for reproductive differences, the provision of research help during leave for childbirth, and changes in the policy of funding agencies. These may remedy some of the inequities that arise from even the voluntary choices women make that lead to their underrepresentation in academia. And, of course, the issue remains, politically volatile as it is, whether we should be fostering equality of opportunity (both sexes get equal access to resources) or equality of outcome, i.e., will we be satisfied only when half of all faculty positions in every field are occupied by women and half by men? If we do not achieve this parity across the board, does that prove discrimination?
The prospect of unequal desires and abilities based on biological rather than cultural differences is one of the biggest minefields in academia—so much so that it’s become nearly taboo to raise the issue. This is the subject of Monday’s New York Times piece by John Tierney discussing Ceci and Williams’s paper. Tierney describes the claim of John Haidt, a social psychologist from The University of Virginia, that there is discrimination against conservatives in academia, and that they are woefully underrepresented on faculties.
It [the “bias”] was identified by Jonathan Haidt, a social psychologist at the University of Virginia who studies the intuitive foundations of morality and ideology. He polled his audience at the San Antonio Convention Center, starting by asking how many considered themselves politically liberal. A sea of hands appeared, and Dr. Haidt estimated that liberals made up 80 percent of the 1,000 psychologists in the ballroom. When he asked for centrists and libertarians, he spotted fewer than three dozen hands. And then, when he asked for conservatives, he counted a grand total of three.
“This is a statistically impossible lack of diversity,” Dr. Haidt concluded, noting polls showing that 40 percent of Americans are conservative and 20 percent are liberal. In his speech and in an interview, Dr. Haidt argued that social psychologists are a “tribal-moral community” united by “sacred values” that hinder research and damage their credibility — and blind them to the hostile climate they’ve created for non-liberals.
I don’t have a lot of sympathy for Haidt’s equation of conservatives with other underrepresented minorities, and his calls for “affirmative action for conservatives.” (I note in passing that Haidt garnered a Templeton Prize in Positive Psychology.) After all, your political views are something that you choose, but you can’t choose your gender or ethnicity. And conservatives have hardly suffered “oppression” in the same way as women and African-Americans.
But I do think that scientific claims which are perceived as conservative or hereditarian are often dismissed on political grounds alone, and that’s not a good thing. We’re supposed to discuss ideas freely, and adjudicate them on evidence, not on how much they appeal to us ideologically. It’s not good if there is an informal ideological ban on discussing, for example, biological differences between genders. After all, we’re supposed to examine ideas freely regardless of their perceived political or religious consequences: that’s one of the pillars of humanism. We may like the prospect of eternal life, but as good atheists we know that we shouldn’t mistake what want to be true for what is true. A priori suppression of discussion can only inhibit finding the truth.
So discuss the PNAS paper freely, though I adjure you to read it before you discuss it. (Since it’s free, you have no excuse not to read it!) And please don’t take the few paragraphs I’ve extracted as representing the entire contents of the paper.

