The Myth of Meritocracy in Academic Publishing
In the first week of October 2018, the physics community experienced a great deal of attention in the press. On October 3, the Nobel committee awarded the Prize for Physics to only the third woman in history, Donna Strickland. An associate professor of physics at Waterloo University in Canada, Strickland and collaborator Gerard Mourou, shared the prize with Arthur Ashkin for their combined work on refining lasers for biomedical usage. At the time of her award, Strickland was not a household name and did not even have a Wikipedia page.
In the same week that Strickland became the third woman awarded this honor, another physics headline broke. At a workshop on gender diversity in physics held at CERN, researcher Allesandro Strumia told a room of physicists that their field had been built by men and that female physicists had been unfairly “invited” into the field based on politics and not merit. Strumia presented a set of slides that detailed his research into this problem, showing that women’s work in physics was cited less compared to men’s. According to him, this data showed definitively that men’s work is more integral to physics than women’s. Strumia has been dismissed from CERN but stands by his presentation.
Strickland’s rise from obscurity and Strumia’s insistence on using “hard data” to prove lack of discrimination are two parts to a larger concern in science today. Peer-reviewed publishing is considered a meritocratic system. The system is based on the theory that if a researcher is good, then they will be published and cited, and, in turn, the level of publications and citations of the work should be indicative of merit.
But new studies in a wide range of fields show that this is a myth of meritocracy. Recent research suggests that both conscious and unconscious biases continue to hamper women’s contributions to their fields of study starting as early as the first year of graduate school. Our continued belief in this meritocracy has very real outcomes for women in science; without change to this system, efforts to recruit and retain women in the sciences will fail. And this paucity of diverse voices will continue to result in an academy that operates at a diminished capacity by rewarding suboptimal work by the upper-class men by whom and around this myth is built.
Publication pathways taken by men and women in the same field is one area where researchers have tried to pinpoint gender bias in academic publishing. In theory, there are not supposed to be differences in pathways for publication because journal submissions are open to anyone who meets the requirements. A first-year graduate student submits their research through the same online portal that a tenured professor does. Gender should not be a factor. But throughout the process, researchers have found unconscious and conscious gender bias in the publication process.
Several large studies looking at the gender gap in publishing have found that women in science have fewer publications, fewer lead author publications, and they are less likely to place those papers in high prestige journals than their male counterparts. This seems to speak to Strumia’s claims of female inferiority, but there are alternative explanations for this trend.
Two separate meta-analyses published in 2018 found that many women have ample publications, but that extended career paths mean that women spend more time as second authors. In other words, because women often take longer for advancement through their field, because of a number of factors, they are often listed as secondary authors on papers instead of lead authors.
The difference in publishing records can start early. First year female PhD students in “bench” biology programs spend more time in the laboratory than their male counterparts but are 15 percent less likely to publish their research. Most explanations for this gap point not to the age-old myth that women are less confident but to the combination of fewer female lab leaders, a difference in the type of projects undertaken, and a subconscious valuing of male over female labor.
In a 2018 study by Holman et al., researchers found that women scholars were systematically overlooked for invitations for “invited” papers in prestige journals, suggesting “gender bias by journal editors.” Holman et al.’s study found that the gender gap in publishing in prestige journals, and especially in invited papers, was most noticeable in physics. So in essence, Holman et al. found that Strumia was right: physics does send out invitations not based on merit. But it is men, not women, are usually the recipients of those invites.
In addition to an imbalance of publications in top tier journals, it is possible that the peer review process looks different for women scholars. Erin Hengel’s research found that women economists took six months longer to publish papers than their male peers. Much of this process involved fine-tuning the “readability” of these papers. Hengel shows that women are producing higher quality writing, both at the start of the process and throughout, but is unable to pinpoint the reasons for the delays. What she can say is that this added time doesn’t result in higher citations or more accolades — papers with higher “readability” by women are still cited less often than those of less readability by men.
In astronomy, male authored papers are cited about 10 percent more often than female authored papers; controlling for a multitude of variables, these imbalances point to a combination of unconscious bias and network construction. According to a paper by Neven Caplar, scientists often cite based on who they know and this networking informs how they perceive work as intersecting with their own. In fields that are still heavily male, scholars in a wide variety of fields remain more likely to cite men over women, controlling for all other factors.
Additionally, men are more likely to cite themselves. In a 2016 paper, sociologist Molly King found that male scientists are 56 percent more likely to cite themselves than women and that this trend has risen significantly in the last 20 years. Researchers have several theories about gender and self-citation, some suggesting that women lack confidence in their own work while others suggest that self-citation is considered unacceptable in certain circles and a habit primarily learned from mentors.
The combination of fewer publications in high impact journals and fewer citations for those publications leads to something that historian Margaret Rossiter calls “the Matilda effect.” Rossiter’s work build’s off Thomas Merton and Harriet Zuckerman’s definition of the Matthew effect in science. Merton and Zuckerman use the parable of the talents to describe how well-known scientists often get more credit than unknown researchers. Rossiter extends this analysis to gender in science and names the phenomenon after Matilda Gage, an American suffragist who originally articulated this phenomenon regarding women generally. The Matilda effect is simple but powerful: women are disadvantaged going into scientific work and that initial disadvantage results in their work being overlooked or co-opted by men.
The Matilda effect is visible in the publication process: women are perceived as less competent to publish at the offset and are given less leeway in what they can and should publish throughout their career. Once they do publish, their work is cited less and seen as less important — not because it is any different than men’s work but because it is subconsciously perceived as less important. The Matilda effect is especially strong post publication because the perception that work is less important leads to fewer citations, which then leads to fewer invited journal articles. “Advanced” career options often rely on the perception that a researcher has attained significant visibility in their field via citations, invited lectures, and invited articles. And that advancement equals pay, prestige, and graduate students. This is a recursive cycle that keeps women out of every level of the academy.
Allesandro Strumia’s reliance on publication statistics — both the amount of publications and the citations resulting from them — to identify the important minds in his field is all too common. In academia, publications and resulting citations equal a person’s net worth. Researchers need them to get a job, keep a job, and get promoted. The academic community, and Strumia, believe that citations reflect value unencumbered by bias. This belief that citations are a pure analysis of scientific contribution is at the heart of how scientific merit is attributed. And this is a problem.
As long as citations are believed to be unbiased data representing scientific merit, we will continue to see women struggle to find footing in the science community. We will continue to see promotion to Full Professorships denied without Nobel Prizes. And Nobel Prizes (and Wikipedia pages) will be surprises, when they should be standard.
Ellen Mayock, Gender shrapnel in the academic workplace (Pallgrave Macmillan US, 2016).
Sue V. Rosser, The science glass ceiling: Academic women scientist and the struggle to succeed (Routledge, 2004).
Image credit: Donna Strickland as a graduate student at the University of Rochester , 1985 via Wikimedia Commons