Continue to publish less than male colleagues

Women in academia publish 21 per cent less than their male colleagues, and this figure has been stable for nearly 20 years. A new master’s thesis takes a look behind the numbers.

Women in academia continue to publish less than their male colleagues. (Illustration photo: iStockphoto)

The difference in men’s and women’s publication rates is as great today as in the early 1990s. A Norwegian study from 1996, based on numbers from 1991-1993, showed that women published 20 per cent less than their male colleagues. Today the difference is 21 per cent. These are the findings presented in Peter Bentley’s master’s thesis Gender differences in research productivity - A comparative analysis of Norway and Australia.

When looking at these numbers, it is easy to conclude that working conditions for women have not changed in 20 years. But Bentley suggests a different reason:

“Norway has an increasingly family friendly working life. This may attract women to positions they previously stayed away from in academia because they considered them difficult to combine with having children. If there are more women with small children in academia today, one could reasonably expect these women to publicize less for a period of time,” he says.

In Australia, where working life is less family friendly, the gender difference in publication rates are higher, with women publishing 26 per cent less than men.

Large-scale survey

The thesis is based on results from the Changing Nature of the Academic Profession (CAP) project. This survey-based project was undertaken in 16 countries across five continents, and its goal was to examine changes in the academic profession in recent years. Bentley’s thesis looks solely at publication rates in Norway and Australia.

The types of publications counted in the survey are books authored, books edited and articles authored. A point system was used in which books authored were given five points, books edited two points and journal articles one point.

Not institutional reasons

“I looked for institutional reasons, like for instance research funding and teaching responsibilities, for differences in publication. In Australia especially, such factors can differ greatly between institutions. But I found that they had little effect after controlling for other factors; the differences came from individual achievements,” Bentley says.

Individual achievement factors refer to individual choices and engagement with the academic profession, such as the time spent on research, collaboration, whether the person holds a doctoral degree, and academic rank.

“I found that the most important variable that affects publication rates is academic rank. Higher ranked researchers publish more than lower ranked researchers for both men and women. It is not very informative, however, to compare men and women with similar ranks,” Bentley says.

He explains that while many studies show that publication output is the same for males and females with full professorships, academic rank acts as both a cause and an effect of research output. Academics with high publication output may be rewarded with promotion, and higher level positions may increase their visibility and access to resources. This is especially true in Australia. If high levels of publishing are a requirement for full professorship, it is not surprising that female professors have an output equal to that of their male colleagues.

International collaboration important

Researchers who collaborate with others publish more than those who do not, especially when the collaboration is international.

“While higher ranking academics publish more than the lower ranking, Norwegian associate professors who partake in international collaboration publish at equal levels to full professors who do not,” Bentley explains.

In Norway, there are negligible differences in the percentage of men and women who take part in international research collaboration, but the effect is somewhat greater for men. Compared to academics who did not have international collaborators, productivity was 92 per cent higher for women and 116 per cent higher for men. Australia has less international collaboration, but the effect for those who did cooperate with colleagues abroad was even greater, again more so for men.

Marriage and children

Family responsibilities are often mentioned as a hindrance to women’s academic careers, even in Norway’s comparably family friendly and gender equality oriented working life. Bentley’s study found that married men are generally more productive than those who are unmarried. For women, the effect is a little more complicated.

Peter Bentley. (Photo: Heidi Elisabeth Sandnes)

“Marriage has some effect for women. Norwegian women whose husbands hold university-level education are more productive than single women, but those with husbands who do not hold university education publish less,” Bentley says.

The fact that single women publish less may be explained by several factors. Younger academics tend to publish less than those who are older, and they are also likely to be unmarried. When controlling for age and other family characteristics, the positive effects of marriage is significant only for Australian men.

Unfortunately, the CAP survey did not adequately pose questions to give reliable evidence for the effect of children. While the survey asked whether the respondents had children living at home, it did not ask for the age of the children, nor about children no longer living at home.

“There is a tendency in my material for both men and women in Norway to be less productive when they have children at home, but the effect is small. However, it is difficult to draw any conclusions from this due to the way the question was asked,” Bentley comments.

The male breadwinner

For Australia the effect of children is equally small, but men tend to be slightly more productive when they have children at home, while women are less so. But Bentley believes that the issue of children plays a more significant role in Australia than in Norway.

“If I were to have a child while working at an Australian university, as a man I would get maybe five days of paid parental leave. This shows that the idea of the male breadwinner is going strong in Australia,” he says.

Just under half of all Norwegian women had interrupted their careers to care for children or elderly family members, compared to 19 per cent of men. These gender differences were even greater in Australia, where 44 per cent of women had interrupted their careers, compared to 10 per cent of men. Such career interruptions were also far longer for women. Bentley does not think it is possible to draw conclusions about the effect of children on women’s career based on this type of study.

“Only people who have chosen an academic career responded to this survey, and the same is true for other studies that look at the effect of children. Thus we have no way of knowing how many choose not to enter the profession because they see it as incompatible with having a family,” he points out.

Working hours

Working hours are typically long – about 48 hours per week for full-time academics in Norway. Men work one extra hour per week, but this factor does not explain Norwegian women’s lower publication output.

“There is a tiny difference shown in that female associate professors work one hour less than similarly ranked men, but the opposite is true of full professors where women work one more hour per week,” Bentley explains.

Time spent on research and teaching also fails to account for the differences between men and women in Norway, as men and women spend a similar number of hours on these activities. However, Australian men spend two more hours per week on research than women, and this may help to explain the large gender differences in Australia.

“Hours spent on research are a significant predictor of research output, even after controlling for other factors. Whereas all academics engage in teaching and research equally in Norway, lower and middle ranked academics in Australia spend far more hours on teaching and fewer hours on research compared to full professors,” Bentley explains.

A few high-publishing individuals

A factor that can explain quite a lot of the difference between women and men are the high-publishing individuals. It is a well known fact in academia that a small percentage of researchers publish well beyond the average.

“In the group of individuals that publish 10 or more articles per year, 89 per cent are male. 7.5 per cent of all males in our survey fall within this group, but only 2.5 per cent of females. The relative contribution of this small group of predominantly male academics has increased since the early 1980s, which goes to explain why gender-based differences have remained stable despite an overall increase in female publishing,” Bentley says.

The fact that Norwegian academia is divided into typically male and female disciplines goes a long way in explaining this difference.

“Extremely highly publishing individuals are often in the medical profession, where co-authorship is very common. The CAP survey does not control for co-authorship, so a high ranking medical professor who co-authors articles with several doctoral students gets the same credit as a social scientist who is the sole author of an article,” Bentley explains.

Men in the medical profession are 70 per cent more productive than men the social sciences, whereas there is little difference between women in the two fields.

“This is because there are more men in the higher ranks in medicine who can benefit from co-authorship with lots of people working beneath them. Women and men of similar ranks publish at the same rates in medicine, but the proportion of male professors is far greater in medicine than in other academic fields,” Bentley says.

Master’s thesis

Bentley, Peter: Gender differences in research productivity. A comparative analysis of Norway and Australia. European Master in Higher Education, Institute for Educational Research, Faculty of Education, University of Oslo 2009.

See