91av

EVEN in 2007, science is still a profession dominated by men. Although there has been progress, the disparity in numbers between male and female scientists, particularly at senior levels, is remarkable. The difference is most noticeable in the physical sciences where, according to UK government figures, women make up less than 20 per cent of postdoctoral physicists and fewer than 5 per cent of professors. In the more woman friendly biological sciences, the proportion of women is better – more than 40 per cent – at postdoc level, but this has fallen to almost 1 in 10 at professorial level. The story is pretty much the same worldwide.

This under-representation of women in science is probably connected to the disparity between the rate at which men and women publish scientific papers. For many years, studies have shown that men publish more scientific papers than women. The difference is not slight, either. In a comprehensive review across science () Yu Xie and Kimberlee Shauman showed that on average men produced 33 per cent more papers over a two-year period than women, and almost twice as many over an entire career. This difference is crucial because kudos – and that all-important funding – are awarded mainly on the basis of productivity. We all know it’s a case of publish or perish, so what causes this “productivity puzzle”?

The most contentious explanation is simply that women are not as good at science as men. In a at the National Bureau of Economic Research in January 2005, Lawrence Summers, then president of Harvard University, said that he “sensed the unfortunate truth” was that women were not prepared to put in the hard work needed to reach the highest levels in science. He also argued that, while the sexes do not differ in their scientific ability on average, men exhibit a wider variation in ability than women. Thus there are relatively more men with exceptional scientific ability than women.

Predictably, many critics jumped on Summers, pointing out that women scientists face, if not outright discrimination, then certainly many obstacles men do not. Ben Barres, a neurobiologist from Stanford University in California, highlighted how women may be both judged more harshly, and not given the same encouragement and support as their male colleagues (). Barres speaks from the rare perspective of being a transgendered person who was born female but is now male. He tells an amusing, but revealing, story of how shortly after he changed sex, he overheard a member of his faculty praising the seminar he had given, and reflecting on how much better Barres’s science was than his sister’s.

If Barres is right, and women don’t receive the same support and encouragement as men, this may also affect productivity, particularly early on in their careers when they may not be pushed sufficiently to publish. A recent review by Dario Sambunjak and his colleagues (Journal of the American Medical Association, vol 296, p 1103) showed that early-career mentoring in academic medicine has positive effects not only on career satisfaction and progression, but on subsequent productivity. Tellingly, female medics typically report less mentoring than men.

Apart from mentoring, the most obvious explanation is that lack of productivity reflects time having and bringing up children. Yet there is little evidence that having children has a negative effect on research output, with one study in Research in Higher Education (vol 45, p 891) showing a positive correlation between women’s research productivity and the number of their children. In a study of British and Australian biologists by my colleagues and I (), we found that the difference in productivity between men and women occurs remarkably early in their careers, within two years of their first publication, which is probably before most scientists have started having families.

Superficially, these results suggest men and women produce their science in different ways. Perhaps men are more cavalier in the way that they pump out science, while women are more careful to ensure their output is of high quality? Of course, such a strategy would be characteristic of a situation where women felt their work was judged more harshly.

“Perhaps men are more cavalier in the way they pump out science”

Science may also favour other attributes more typically exhibited by men, such as self-promotion and competitiveness. One of the few top women physicists, Deb Kane, professor of physics at Macquarie University in Sydney, Australia, feels that “women’s reduced success may be an indicator of an environment where only the top ‘competitors’ succeed”. But, she adds, “the best competitors are not necessarily the best scientists”. Kane also challenges notions of productivity. “If multi-authored papers…provide[d] an accurate fractional breakdown of…contribution and intellectual input…by each author, and total papers…was related to amount of funding support received, would there be significant gender differences in…output and papers per dollar?”

The fact that women’s productivity falls behind early in their careers means that even if they subsequently increase publication rates (which they do) they will, at any time, appear less productive than men of equal experience. Because assessing the ability of a scientist focuses so greatly on quantity of publications, this “catch-up” scenario has all sorts of vicious-circle implications. Once you start to fall behind, it is harder to get funding, and more difficult to produce research and so on. While this quantity-based focus persists, women are bound to struggle, to publish less, and to be under-represented at the top levels.

This leads us to a crucial question: why do we think quantity of research is a good indicator of ability? This view seems peculiar to science. Would you think that newspaper cartoonists are better artists than Van Gogh because they produce art every day? Young scientists are brought up on mantras of “smallest publishable units” and “don’t worry if it’s not perfect, get it published”, so why not similar urgings regarding quality?

There are crude measures we can use to assess quality, such as the impact factor of a journal, or the citation rate of a paper. But assessing whether women produce higher quality science is a murky business because quantity has an unexpectedly pervasive influence: scientists who publish more also get cited more. Publishing more papers may work like an ad for your research: other people remember your name and refer to your work.

But in our study, we found that men and women with equal numbers of papers showed clear differences in quality, with women’s papers on average being cited about 20 per cent more than men’s. Again, the possibility that women focus on the quality of their science rather than on quantity is inescapable.

In the long term, solving the productivity puzzle will not come just from analysing publication trends. Instead, we must examine why women are slower at publishing at the outset of their careers. This means exploring everything from women’s support systems to such complex issues as whether research output reflects the sexes’ different behavioural strategies. At the same time, if we are to increase the number of women in science without socially divisive measures such as affirmative action, we must shift our thinking about what makes a good scientist.