Women publish more than people think

We hear constantly that women researchers publish less than men.

In 2018, woman researchers in Norway averaged 1.15 publication points, while men garnered an average of 1.67 points each. But what do these numbers actually tell us?

Averages reveal little

“Not everyone is aware that the production of scientific literature is very skewed. The 10 per cent most productive researchers account for more than 40 per cent of publication points,” says Dag W. Aksnes, a research professor at the Nordic Institute for Studies in Innovation, Research and Education (NIFU).

“This is not unique to Norway but is a pattern that holds true internationally,” he says.

In connection with this year’s edition of the Report on Science and Technology Indicators for Norway (more in fact box), Aksnes looked behind average publication figures to learn what they may conceal. All scientific publications registered from the higher education sector, the research institutes and the health trusts in the Current Research Information System in Norway (Cristin system) in 2019 were examined.

Of the approximately 26 700 researchers whose names appeared on at least one scientific publication in 2019, 11 700 were women – almost 44 per cent. But the women accounted for only 36 per cent of publication points. While that is four percentage points higher than in 2011, it is still lower than the ratio of women to men researchers would lead one to expect. The figures thus validate the international pattern of women systematically publishing less than men do.

However, much of the gender gap stems from the fact that the small cohort of super-productive researchers is predominantly male. Since this group accounts for such a large portion of all publication points, the gender distribution within it skews the average.

The research directors are men

The factor of super-producers came to light when the researchers collected the names of everyone credited on a scientific publication in 2019, then divided them into 10-percentile groups, with an equal number of researchers in each of the 10 groups, based on their publication point totals.

Men were strongly overrepresented among the most productive 10 per cent of researchers, who generated 43 per cent of total publication points in 2019. Women comprised the majority in the lowest percentile groups – those that publish little. But in the intermediate groups, where “normal researchers” tend to be, the gender difference in publishing is very small.

“One likely explanation for this phenomenon is career stage,” says Aksnes.

“Our data don’t show who those super-productive researchers in the highest 10-percentile are. We can only speculate – but I would think part of the reason men dominate in this group is that it includes established professors who are research directors and have their name on a great many publications. These are still largely men. So the dynamics of how the research system is organized play a role here.”

It is also reasonable to assume that many super-productive researchers spend an extraordinary amount of time on research at the expense of other things, and that men are perhaps more willing, or able, to prioritize this way than women.

Women not cited less

At the other end are researchers with looser ties to the research system and who hold other types of positions. The data studied encompass everyone with their name on a publication, from students who have participated in a study to lecturers and professors.

“We know, for example, that women make up a majority of college and university lecturers, who get less research time. If these differences were taken into account, the gender differences would be much smaller,” says Aksnes.

The analysis also looked at Level 1 and Level 2 publishing and found no gender differences.

Nor is there much of a gender gap in citations. Men are cited only marginally more often than women.

“I think this is interesting, because international studies indicate that men are systematically cited more than women. But our study does not support that,” says Aksnes.

“The same goes for international research collaboration in the form of co-publications: we only find small gender differences there, after adjusting for the fact that fewer women work in the subjects where such collaboration is most widespread.”

This is consistent with previous research on Norwegian publication figures, which showed average gender differences almost disappearing after adjusting for such factors as staff category, leaves of absence and research methodology in the publications.

“What is perhaps most surprising, given what we know from previous studies, is that the gender differences are not larger. There are relatively small differences between women and men in all the aspects we measured except publication volume. And when it comes to volume, as I said, the picture is more nuanced than one might have thought,” says Aksnes.

Major differences between fields

It is well known that gender distribution in the Norwegian research system varies greatly from discipline to discipline, with the variation becoming clearer the closer one looks. Relatively speaking, there are far more women in the humanities and social sciences than in natural sciences and technology.

But even within these large groups the differences are significant: in linguistics, 62 per cent of researchers who published in 2019 were women, while in philosophy and history of ideas only 24 per cent were women. In “interdisciplinary natural sciences/medicine”, 50 per cent of researchers are women, but in electronics and cybernetics the figure is only 15 per cent.

“Norway is unique because we have the Cristin information registration system, which provides very solid data for the different disciplines. International databases like the Web of Science and Scopus aren’t as complete, especially for the humanities and social sciences,” explains Aksnes.

In principle, the Norwegian publication indicator used for distributing publication points is subject-neutral. Some years ago, the method of calculating publication points was adjusted to improve the balance between subject areas with different publishing practices.

For example, natural sciences researchers tend to publish many articles, but the publication points are divided among many co-authors. In the humanities, publishing fewer articles is the norm, but because authors tend to publish alone or with few co-authors, each one obtains more publication points.

Women moving up

Aksnes also heads the research project Balansekunst (“The Art of Balance”), where the objective is to provide a thorough analysis of factors that explain the lack of gender balance in academia. On the one hand, major changes are underway: from 2011 to 2019, the number of women researchers publishing articles rose 62 per cent, compared to 34 per cent for men. In other words, the growth rate for women is almost twice that of men.

On the other hand, this growth is occurring mainly in younger age groups, which publish less. The highly productive older age groups, professors in particular, are still male-dominated and changing very slowly. The researchers are now working to analyse in more detail why the male-female productivity gap arises and how it affects career development.

“In the Balansekunst project we’re taking a closer look at things like recruitment bases, international recruitment and individual variables that can help to explain gender gaps at different career stages. For now, women make up the majority of doctoral students, and the associate professor category is well balanced. But these totals hide the fact that gender balance differs from subject to subject,” says Aksnes.

Must secure sufficient research time

It is too early, however, to conclude that things will even out on their own.

“The number of women publishing is increasing faster than the number of men publishing, but we know that the length of your publication list counts in hiring and other matters. We must ensure that women, too, have enough time to conduct research,” says Aksnes.

“We often hear that women are assigned major administrative burdens, and of course it’s important that they are well represented on committees, but we have to make sure they also get research time.”

Going forward, he believes, it will be important to keep an eye on the differences between disciplines, which are difficult to discern when looking at total gender distribution. To achieve gender balance in academia, we need to know more about why the gender breakdown varies so much between disciplines. Only then can effective measures be designed.

“Historically, men have been in the majority everywhere in the research system. Now we’re seeing disciplines with a preponderance of women, while women remain underrepresented elsewhere, as in technology disciplines. This is hard to change because the recruitment base is so skewed,” he says.

Translated by Walter Gibbs and Darren McKellep.