Are the Libs Really Leaving?

Twitter would have to become functionally worse, and something else would have to be obviously better.

A photo-illustration of birds flying in the shape of the Twitter logo
Getty / The Atlantic

Ever since Elon Musk announced his intention to acquire Twitter, take the company private, and introduce a new era of free-speech absolutism, users have been threatening to leave the platform—a threat that has been likened to those made by liberals during the 2016 presidential election to move to Canada.

Musk officially took over the company last week and has already been stomping through it in a chaotic fashion. Many Twitter users (including the actress Téa Leoni) have announced that they will be abandoning the site specifically because of their disdain for Musk and his values. Meanwhile, various outlets have run guides detailing available Twitter alternatives, and high-profile Twitter users have announced (on Twitter) their intention to move to Mastodon, an open-source, decentralized Twitter look-alike. That platform has been around for six years, but it’s never quite taken off. Regardless, it is openly capitalizing on the moment and reportedly signed up about 100,000 new users over the weekend.

This moment calls back to 2018, when #DeleteFacebook trended in response to the Cambridge Analytica data-collection scandal. Although people’s frustration was clearly real, it wasn’t obvious whether they would actually follow through on their commitment to abandon a major social-media platform they’d spent years using. As someone who has spent at least several hours on Twitter nearly every day for the past eight years, I find it hard to imagine myself leaving the site unless things devolve into total chaos or worse—unless everyone else leaves, and only brands and Musk fans remain.

But the history of social-media platforms shows that they do go through major moments of transition, and that sizable communities are sometimes pushed to migrate from one platform to another en masse. This could be such a turning point for Twitter: Back in April, when Musk made his plans public, the news was met with a flurry of right-wing former Twitter users returning to the site and a bunch of left-leaning users leaving. Florida Governor Ron DeSantis, for instance, gained more than 96,000 followers in just a couple of days; Barack Obama lost more than 300,000. At the time, Twitter told NBC News that the behavior was indeed “organic”—that actual people were closing their accounts. I had my suspicions about the gains, DeSantis’s in particular, but when I downloaded a spreadsheet of his new followers and clicked through them at random, I found that they were obviously real people—people who were posting things such as “Now that free speech is here, so am I #FreeSpeech #ElonMuskOwnsTwitter” and “Thank you @elonmusk!!! You are the sole reason myself and millions of others choose to redownload this platform.” (A new account, ostensibly belonging to an older woman, had the bio “Love our wonderful country. I like polite people. Love Elon for buying Twitter!!! Conservatives can post Again,” followed by three red-heart emoji.)

This swap of left-leaning users for right-leaning ones could shift the site. “I think there is a certain number of people who are going to leave immediately out of assumptions about how things are going to change,” Casey Fiesler, an associate professor at the University of Colorado at Boulder who has studied platform migration, told me. “So, as a lot of the current people leave and a lot of new people who are there because of what they think Elon is going to do join, that’s going to significantly change the culture of the platform.”

In 2020, Fiesler and her colleague Brianna Dym, now a lecturer at the University of Maine, published a study looking at why fandom communities migrated from platforms such as LiveJournal and FanFiction.Net to Tumblr in the early 2010s. They highlighted that—as with real-world migration—there are both “push” and “pull” factors at play when people decide to leap from one platform to another. Take, for example, the case of LiveJournal and Tumblr: LiveJournal implemented policy changes that were hostile to fan fiction and was then acquired by a Russian company that banned various political topics. People were pushed to leave and pulled to a place more suited to their needs: Tumblr was a bit of a blank slate, lacking the prohibitions and offering useful features such as customization and better social tools. (Tumblr has also been crucial to fandom because it allows users to remain anonymous, make multimedia posts, and archive material with a meticulous tagging system).

“In some ways, fandom is a unique online community due to its longevity and its lack of ties to any particular platform,” Fiesler and Dym wrote. “In other ways, it is the prototype example.”

When I spoke with Dym this week, she distinguished Twitter’s current situation from what happened to LiveJournal—a one-two punch of policy changes and the Russian acquisition. “The death knell for LiveJournal had been rung already,” she said. “With Twitter, it’s different in that, you know, this is the first major upset in Twitter’s history.” For people to really start leaving on a large scale, she predicted, Musk would have to make significant changes that “erode trust in the platform.”

The other requirement for mass migration is a place to go. “There has to be a compelling reason to leave and a viable alternative option,” Fiesler told me. “An immediate viable alternative option … People are impatient.” She evoked the example of Tumblr’s ban on NSFW content in 2018, which resulted in a catastrophic loss of users for the site. Many of them tried to get their communities to relocate to the newer platform Pillowfort, but it lacked the infrastructure to handle such a huge influx. Other Tumblr users, inspired by the fan-owned fiction-hosting site Archive of Our Own, attempted to start their own social-media platform and collaborated in a massive Google doc with support from the Pinboard founder Maciej Cegłowski—but that effort petered out. Now the people who once used Tumblr for sexual content have either dispersed to smaller platforms (or Twitter) or trudged back to Tumblr and accepted its limitations.

At the moment, Mastodon is not really a compelling alternative to Twitter—in part because it’s hard to instantly set up. If you navigate to the site and click the “Create Account” button today, you’ll be greeted by a pop-up: “Creating an account on is currently not possible, but please keep in mind that you do not need an account specifically on to use Mastodon … Since Mastodon is decentralized, you can create an account on another server and still interact with this one.” The platform’s federated structure—a collection of servers that can communicate but have no central hub—means that there is a significant learning curve for new users, and that it will never serve the same broadcasting function as a site known to drive national discourse. It is better for communities that want to talk among themselves, which could also be said of alternatives such as Discord and Reddit. (Fiesler noted that “Academic Twitter” could possibly make the move to Mastodon, as could smaller subcultures that want to use something that looks and feels like Twitter.)

Perhaps most important, Fiesler and Dym’s paper notes the “pull” factor of network effects. People will go where their friends and the people they’re interested in go, and they won’t go without them (which is exactly why centralized platforms such as Facebook and Twitter have stayed so powerful for so long). This is why fandom migrations might be considered a relevant example: They have the group cohesion and coordination to get a substantial number of people to move all at once. But fandom is also a major organizing principle of politics. When Trump supporters, after being kicked off of Reddit, gathered on alternative forums of their own creation—or abandoned Twitter and picked up Parler or Truth Social—those were actions undertaken by a political fandom. If you think of the Millennial liberals associated with Twitter as a subculture that, in many ways, resembles a tightly knit fandom, it’s not impossible to imagine them doing the same.