Think open source is a meritocracy? It is, but only if no one knows you're a woman

If you're a female programmer, you can probably boost the chances of your code being accepted by picking a gender-neutral username.
Written by Liam Tung, Contributing Writer

The study finds a bias against open-source code written by women.

Image: iStock

Of course, it would be ridiculous to suggest that women coders should hide their gender to achieve greater acceptance.

Yet a study looking at gender bias in open-source software projects on GitHub has found that code written by women is accepted more often than that written by men, but only if the contributor can't be identified as a woman.

The findings, which are discussed in a new paper released this week, highlight flaws in the notion that open-source software is a meritocracy that rewards those who work hardest and perform the best.

The study focused on differences in the acceptance rates of pull requests written by women and men. Pull requests, or suggested changes to code in a repository, are first reviewed by collaborators who then decide whether to accept the change and merge it or reject it.

See also: Google: We're not biased against women and these details on pay prove it | Women's share of Microsoft jobs falls again, so it's tying top pay to diversity | Women in technology: Advice from Intel and Biogen | Women in the workplace: A $12 trillion opportunity | TechRepublic: Why women work in the gig economy

The researchers analyzed three million pull requests from approximately 330,000 GitHub users, which included 21,000 women.

They found that 78.7 percent of women's pull requests were accepted, compared with a figure of 74.6 percent for men.

However, if a name or profile picture easily reveals that a contributor is a woman, their rate of acceptance fell to 58 percent, while identifiable men had an acceptance rate of 61 percent. Women with a gender-neutral profile had 70 percent of their pull requests accepted, compared with 65 percent for men with gender-neutral profiles.

The paper released this week is peer-reviewed and updates a paper released last year that wasn't.

"Our results indicate that gender bias does exist in open-source programming," said Emerson Murphy-Hill, one of the authors of a paper on the study and an associate professor of computer science at North Carolina State University.

"The study also tells us that, in general, women on GitHub are strong programmers. We don't think that's because gender affects one's programming skills, but likely stems from strong self-selection among women who submit pull requests on the site."

As The Guardian reported following the paper's initial release, the researchers began exploring gender bias in pull requests after finding that code written by women was accepted more often than code from men, despite women being a minority among programmers.

Looking for an explanation, they questioned whether women were benefiting from male developers promoting the work of their female peers. However, they found the troubling answer that code by women is only more accepted if their gender isn't known.

Editorial standards