By Ruth Steinhardt
Female journalists, and even casual Twitter users, may be unsurprised to learn that their male colleagues dominate political discourse on the platform. But the starkness of the numbers might take even seasoned observers aback.
According to a recent study by George Washington University journalism Professor Nikki Usher, male journalists retweet other men at three times the rate at which they retweet women. And they reply to other men 91.5 percent of the time.
Dr. Usher called her findings “stunning and upsetting”—but not exactly unexpected.
“It’s not something we would necessarily like to admit, but it’s something any woman who has been on Twitter in any professional capacity has noticed,” she said.
“Twitter Makes It Worse: Political Journalists, Gendered Echo Chambers, and the Amplification of Gender Bias” was published in the International Journal of Press/Politics this month, but Dr. Usher and her research team didn’t set out to study gender disparities. They had assembled and begun to analyze tweets from 2,292 credentialed journalists living and working around Washington, D.C., as part of a larger project on the culture of news production in Washington.
“I’m particularly interested in influential and elite journalists—what they do, where they do it, and what that means more generally for how news gets made,” Dr. Usher said.
As Twitter has become increasingly “ambient,” she said, it has become an indispensible element in any analysis of Beltway journalism—particularly when it comes to understanding elite journalists’ relationships with their peers.
“The dynamics within this group are particularly important because they affect the way news is made and amplified,” Dr. Usher said.
So Dr. Usher and her team, including co-author Justin Littman, a software developer at GW Libraries and Academic Innovation, collected almost 2 million tweets from Washington, D.C.-based journalists credentialed to cover Congress. To mine the accounts they used Gelman’s Social Feed Manager, an open-source software developed as a prototype in 2012 and improved since 2014 with a grant from the National Historical Publications & Records Commission. The software archives not only the text of a post on a social media platform like Twitter, Flickr or Tumblr but also the metadata—time of posting, number of likes or retweets—associated with that post.
Analyzing those tweets by gender was just one of the initial “cuts” that Dr. Usher and her team did as a way to establish a quick data overview. It was never intended to be the final focus. But the numbers were too glaring to be ignored, she said.
Dr. Usher and her colleagues were hardly unaware of the pervasive gender inequity in media. Even assembling their data set was an illustration of some of the problems. Journalists receive congressional credentials from their peers, she said, meaning they have to be fairly well known even to get on the list. And men tend to be better known, partly because Twitter has become an increasingly valuable way to establish legitimacy.
“Just being [male] means you get a follower advantage that is highly important for setting the tone of how people respect you,” Dr. Usher said. “And more people are more likely to follow you if they see you already have more followers.”
In fact, of the 25 Beltway journalists most followed by other journalists, the study found that only four were women.
Twitter is a notoriously hostile environment for women, which Dr. Usher said is a barrier to its effective use by female journalists. But she said it’s not just the platform that disadvantages women. It’s also the way their male peers use it: retweeting and replying to other men almost exclusively.
“If male journalists had a better understanding of how their own behavior affects their peers, that might make a difference,” Dr. Usher said. “It would be great to build a tool where journalists could check their own tweet patterns for balance. But that’s above and beyond what we could do right now.”
For Dr. Usher, one of the study’s most exciting aspects was collaborating with the university’s libraries. While she said she wouldn’t describe herself as a quantitative researcher or a specialist in big data, the Social Media Feed Manager and other tools made it possible for her and her team to analyze a lot of quantitative information in an intuitive way.
“This was really special, because it wasn’t just cross-departmental research, it was actually cross-unit,” she said. The digital resources available “speak to how GW libraries are aggressively moving into the future in a really powerful way and can help students and researchers in ways libraries aren’t always expected to.”
But despite the vastness of the information available, Dr. Usher said she believes the study has attracted attention because it is “fundamentally very basic.”
“We’re looking at very concrete behaviors on Twitter that most people are familiar with: retweet patterns, following patterns, reply patterns, original tweet patterns,” she said. “These are very basic, crude, aggregate ways of looking at how this group engages on the platform. There’s a simplicity in the research and the way we cut the research that make the findings so easily understood.
“In the context of the #MeToo movement, in D.C. and elsewhere, questions of gender have been at the forefront, and there’s a lt of palpable frustration for women,” Dr. Usher said. “Hopefully this study gave women a vocabulary to quantify the challenges they face just because they are …women.”