Katalin Karikó’s work on messenger RNA helped enable the development of the Pfizer–BioNTech and Moderna COVID-19 vaccines. But early in her career, the biochemist says, she suffered a setback when her name was taken off a manuscript that was in press. She explains that her supervisor paid the Proceedings of the National Academy of Sciences $150 to remove her name after learning that Karikó had tried to get a job at another lab. “This is why I do not have a PNAS paper,” says Karikó, who is currently at the University of Pennsylvania.
Karikó suspects a lot of scientists have been in similar situations, and the results of a survey published in Science Advances on 1 September suggest she is right. More than half of the survey participants reported having authorship disagreements. The survey also found that women scientists are more likely to be involved in authorship disputes than men.
Study coauthor Cassidy Sugimoto, an information scientist at the Georgia Institute of Technology, attributes the results in part to the fact that men are more likely to be principal investigators and thus can make unilateral authorship decisions, which may lead to women’s work being devalued. “What we see is women want to have more clear-cut guidelines,” Sugimoto says. “I think that’s because it protects them, and it gives them tools to use in conversations” with the PI.
The global survey questioned 5575 researchers (3566 men and 2009 women) in the social sciences, medical sciences, natural sciences, and engineering about how often they have had disputes about author lists on manuscripts. Such disputes can be over whether an author should be listed for a minor contribution, or whether an author is being unfairly omitted despite having done a lot of work, or what order the authors are listed in—particularly for the coveted first and last spots on the list. Fifty-three percent of the participants said they had encountered disagreements as to who should be included as an author or the order of names.
Women were more likely than men to have been part of disputes about whether a researcher should be named on a paper (roughly 53%, in contrast to 45%) as well as disagreements over the order of author lists (43%, in contrast to 36%). Overall, more than 23% of women participants said they received less credit than they deserved, compared with just over 18% of men. And the men surveyed (11%) were more likely than the women (8%) to say they received more credit than they should have.
Authorship disputes have led to retractions of papers and even lawsuits. One problem is that different disciplines have different norms about what contributions warrant authorship. Another is that the research community lacks consensus on how authorship lists can be constructed objectively. Some academics have developed standardized criteria and processes, including point-based systems, or have adopted arbitrary ways of breaking impasses. For instance, virologist Sizun Jiang of Beth Israel Deaconess Medical Center in Boston says he persuaded the coauthors listed first on a July 2021 Frontiers in Immunology paper to agree to play three rounds of the Super Smash Bros. video game to decide the order of their names.
According to the survey, just under 40% of women, compared with about 33% of men, discuss authorship expectations—such as who will be an author and how author lists will be arranged—at the start of a project. More men (19%) than women (15%) said that they discuss authorship at the end of a project, usually with only the most senior collaborators.
Both approaches have their shortcomings, Sugimoto says. Having discussions at the start may discount the actual labor that’s performed, whereas leaving it to the end may mean that those who did the latest work are prioritized over earlier contributors.
When asked about the importance of different types of contributions to papers, both men and women rated writing manuscripts and analyzing data as the most important tasks. Women, however, are more likely than men to be involved in technical work such as conducting experiments, collecting data, and carrying out statistical tests. Even so, men and women rated technical work as the least important.
Sugimoto says she suspects that authorship practices would become more equitable if more women occupied senior research positions. But getting to those higher-ranking roles in the first place requires publishing papers with their names in leading positions of author lists. Previous studies have shown that women are less likely than men to be listed as lead authors early in their careers, which often contributes to a decision to drop out of science. “We’re still finding that women are leaving science at greater rates” compared with men, Sugimoto says. “They lack the dominant author positions that are necessary currency for moving through an academic trajectory.”
Ludo Waltman, deputy director of the Centre for Science and Technology Studies (CWTS) at Leiden University in the Netherlands, says it’s important to highlight the prevalence of such disputes in science. “This is something we should be concerned about, because it suggests there is a lot of unfairness in authorship,” he says. However, he notes that the gender differences identified in the study tend to be small.
Lisette Jong, who studies research funding also at CWTS at Leiden, points out that the new study doesn’t include a nonbinary gender assignment. “It fails to acknowledge the existence of others who don’t identify as either men or women,” she says.
Issues related to authorship are among the most common topics of complaints filed with the Committee on Publication Ethics (COPE), says Daniel Kulp, the organization’s chair. Current COPE advice suggests that an agreement should be made on author lists before a paper is submitted to a journal, and if a dispute occurs when a paper is going through peer review, the reviewing process should be halted. If the work is already published, COPE advises, a correction should be issued.
Sugimoto suggests that PIs should be required by their institutions to produce mentoring plans outlining their authorship practices and that journal editors should ask all study authors to confirm exactly what they did and whether they agree with the author list. In addition, scientists and publishers could adopt a systematic way of reporting each researcher’s contributions using tools like the CRediT taxonomy, which defines 14 distinct roles such as conceptualization, data curation, and supervision.
Editor’s note, 2 September: The first sentence of the fifth paragraph was updated to correct the comparisons between how often men and women experience author disputes.
Editor’s note, 2 September: The first sentence of the fourth paragraph was updated to correct the number of men and women who responded to the survey.