Women are less likely to be named as authors on articles or as inventors on patents than are their male team mates, despite doing the same amount of work, according to an analysis of how research contributions are recognized. This is partly because women’s contributions to research are “often not known, not appreciated or ignored”, say the authors.
The results, published in Nature today1, hold true for almost all research fields and career stages in the United States. And although the study focused on women, the authors say they saw similar patterns for people from other groups that are marginalized in science.
The research is “innovative and important” because it partly explains why women publish less than men, says Virginia Valian, a psychologist at Hunter College in New York City. “This is a major wake-up call for scientists, and also for funding agencies,” she adds.
There is a well-documented gender-based productivity gap in science. On average, women publish fewer papers than men, secure fewer grants and fill fewer leadership positions. Previous research has suggested that women are less productive because scientific working environments are less welcoming to them, they hold different positions from men or they have greater family responsibilities2,3,4. But a 2020 study also hinted that women’s research is undervalued5.
Measuring what isn’t there, however, is challenging. To overcome this, Matthew Ross, an economist at Northeastern University in Boston, Massachusetts, and his colleagues used a large data set on almost 10,000 research teams in the United States to investigate who did and did not receive credit for work. The data set, hosted by the University of Michigan in Ann Arbor, includes information about almost 129,000 researchers working in those teams, including their job title, field of research, the grants they were employed on and how much of their time they charged to each grant between 2013 and 2016. The authors assigned gender to each person in the data set on the basis of their names, using two computer algorithms. The analysis did not take into account non-binary or gender-fluid researchers.
The authors used these data to group scientists working on the same projects into research teams — and then used bibliometric data to create a list of scientific outputs, such as published papers and patents, for each team between 2014 and 2016. The authors were then able to work out which researchers in a given team were and were not named on papers and patents, and to calculate differences by gender.
They found that the probability of a man ever being named as an author or inventor during the study period was 21%, compared with 12% for a woman. Even when men and women held the same position, women were 5% less likely to be named as an author or inventor than were men.
To estimate the potential authorships that women missed out on, the authors compared the team members employed a year before a paper’s publication date — the pool of potential authors — with the actual authors listed on the manuscript. They found that across all job titles and fields, men had double women’s chances of being named on any scientific document.
Valian, who studies gender bias in workplaces, describes the analysis as rigorous because it controls for factors that could dictate whether a researcher secured an authorship, such as the size of their role on the project.
She adds that previous work has shown that female senior authors are less likely to be cited than male ones, even if they publish in the same journals. “So we have a double whammy. Women are less likely to get authorship and they are less likely to be cited if they are the first or last author,” she says. She adds that funding agencies should care about how team leaders are crediting the work of their team members.
Ross and his colleagues also polled 2,660 researchers who had published a paper after 2014, to find out their experiences of authorship. They conducted short interviews with some respondents.
Both men and women said they had been excluded from papers to which they had contributed, but women were disproportionately affected. The most common reason researchers gave for not getting an author slot was that others underestimated their scientific contribution; 49% of women reported this, compared with 39% of men. Although respondents didn’t often mention feeling discriminated against, women were twice as likely to mention it as men.
The survey also asked respondents what they did to earn authorship on a recent paper, based on a list of contributions. They found that on average, women had to work harder than men get an authorship credit. Women did significantly more than men when it came to conceptualizing the research, curating data, writing, reviewing and editing. The only category in which men reported a greater contribution than women was developing software.
Although the survey and interviews revealed some egregious examples of gender discrimination, the vast majority of what was reported around authorship was people feeling ignored and not listened to, says co-author Bruce Weinberg, an economist at the Ohio State University in Columbus. He says that when researchers are making authorship decisions, they should be more mindful of people’s contributions.
Ludo Waltman, a bibliometrician at Leiden University in the Netherlands, says that the paper reflects “deeply problematic” research cultures. “There will be both women and men that suffer from this. I wonder whether we should shift the conversation from gender differences to culture problems, which I think would offer a way to discuss the deeper underlying issues,” he says.
Understanding how teams work is important for improving research culture, says study co-author Julia Lane, an economist at New York University in New York City. “Documents and grants don’t do science; people and teams do science. That is the fundamental unit of science and that is what we need to understand if we want to do science better,” she says.