Women less likely than men to get authorship on scientific publications, analysis finds

Look at the list of authors on a scientific paper and you’ll typically see more men than women. It’s easy to notice that gap and blame it on productivity-limiting factors many female scientists face, like greater family obligations or unwelcome environments.

A new study published in Nature attempts to interpret this authorship discrepancy in a different way. It asked whether everyone who put in the same effort on the same project had an equal chance of being listed as an author.

Instead of focusing on names on papers and patents, researchers compared employment records with author lists to find out whose names were missing. Their findings are clear: Women are less likely to be credited than men for the same work.

advertisement

Specifically, women are 13% less likely to be credited with authorship than men on a paper and 58% less likely to receive credit on a patent. The discrepancy exists despite survey results showing women self-report contributing to a broader swath of types of scientific work that would merit authorship than men.

The study harkens back to one of the most egregious author oversights in scientific history. When the structure of DNA was discovered in the 1950s, James Watson and Francis Crick crafted their model of DNA off the images obtained by fellow researcher Rosalind Franklin. Yet, Watson and Crick leap-frogged over Franklin to publish their work in a 1953 Nature paper without including her as author.

advertisement

Franklin’s story may serve as a warning that rings true to young female scientists today. While the number of women in STEM fields has dramatically increased in the decades since her work, the gap in authorship between men and women on papers and patents has only widened.

The new study was led by senior author and labor economist Julia Lane, who more than a decade ago was put in charge of a program by the National Science Foundation to find a way to measure the return on scientific investment. Instead of merely counting publications and patents, Lane wanted to focus on the output of people.

She and her colleagues approached the problem using data, surveys, and stories. The bulk of their analysis drew on the administrative records of over 120,000 individuals working at 77 higher education institutions between 2013 and 2016. Individuals – with job titles of graduate students, post-docs, research staff, undergraduates, and faculty – were grouped into teams based on grants funding the research project. All of the data were anonymized, confidential, and the surveys were approved by an institutional review board.

Albert-László Barabási, a network scientist and senior author of a large-scale analysis of author lists, said the sheer size of the dataset and the “clever analysis” of output by teams were “the most valuable part of the study because that is somewhat objective data.”

Lane and her colleagues found women make up nearly 50% of the workforce, but the likelihood of a woman being credited on a paper is 13% lower than the men on their team. The gap between men and women grows significantly as the impact of a paper increases, as measured by how frequently that paper is cited in other articles.

Co-authors Raviv Murciano-Goroff and Matthew Ross, economists and social scientists, said high-impact papers are often rife with competition to get on the author list. As researchers jockey for credit on career-changing papers, the contributions of women are overlooked, Murciano-Goroff and Ross said.

That pattern can be seen even more clearly with patents, which tend to have fewer contributors. Their data showed women are 58% less likely to receive credit for their work on patents than men.

“That should give people pause…. There’s not just a disparity, but it really seems like there’s something systematic that needs to be investigated and to think about what the long-term impact is on young researchers who are trying to build their career,” Murciano-Goroff said.

“Effectively, what we found was not anything that we could say conclusively as discrimination, but the self-reported evidence sure looks a lot like it’s discriminatory,” Ross added.

The surveys and stories reinforce the quantitative data. Both men and women reported feeling excluded from authorship or feeling their work was underestimated, but women significantly more so. When men and women were asked to claim the types of work they did that merited authorship on a published article, women were more likely to contribute to conceptualization, data curation, analysis, writing, editing, reviewing, and project management. Men were more likely to contribute to software, the study found.

In interviews Murciano-Goroff and Ross conducted with six scientists, all six cited not knowing the guidelines of authorship. Murciano-Goroff said the frustration women felt was palpable as they told stories of times they faced backlash after asking for credit, or watching their career trajectory shift after being left off an important publication.

Lanu Kim, a social science and computer professor at Korea Advanced Institute of Science and Technology, said this study is a counterargument to the idea that science is a meritocracy, which she said stipulates scientific success be “neutrally evaluated, regardless of [the author’s] demographic information, such as gender, race and ethnicity.”

“The thing is, it actually matters,” said Kim, who was not part of the study. “This article adds one more piece of evidence for why [the evaluation of scientists] cannot be perfectly neutral.”

In Lane’s mind, if women – or scientists from any marginalized group — are less likely to get credit, they will be more likely to part ways with their scientific career. In order to increase diversity, “we need to send people the right signals,” Lane said. “It’s important for the country.”

The fix, Lane said, is to train principal investigators to be better managers. “Science over the past 20-30 years… has become bigger. It’s team science. And we basically suck at being managers.”

Ross said he hopes their team of economists can dig deeper into the data to look for trends among other marginalized gender, ethnic, or racial groups. Other higher education institutions are invited to use and add to the dataset.

Rosalind Franklin was not credited for her work during her lifetime. She died at the age of 37 in 1958, five years after the game-changing paper by Watson and Crick on DNA structure was published and four years before they were awarded the Nobel Prize in physiology or medicine for the discovery of the structure of DNA. At the time of her death, the Nobel Foundation had no official statute excluding the option of awarding a prize posthumously. She did not receive the Nobel Prize.