Does Twitter Have an Anti-Conservative Bias?

The following article by Lincoln Network head of policy Zach Graves originally ran at Techdirt:

In an article for Quillette titled, “It Isn’t Your Imagination: Twitter Treats Conservatives More Harshly Than Liberals,” Columbia University research fellow Richard Hanania offers us proof–once and for all–that social media companies are biased against conservatives. Either that, or it’s the latest in a growing list of bogus, exaggerated or otherwise dubiousanti-conservative bias claims (I’ll let you judge for yourself).

“Until now, conservatives have had to rely on anecdotes to make their case,“ Hanania writes. Adding that, “[m]y results make it difficult to take [social media platforms’] claims of political neutrality seriously.” The data he collected (with the help two research assistants, no less) looks at “prominent, politically active” people suspended from Twitter since the company’s launch in 2006.

Accounts included in the data set were selected from individuals and organizations whose suspension was covered in a “mainstream” news outlet, and who expressed a preference for either Donald Trump or Hillary Clinton in the 2016 presidential election.

Out of 22 (!!!) accounts in the data set that met these criteria, 21 (or 95%) were Trump supporters. Despite the small sample size, the author argues this is compelling evidence for Twitter’s anti-conservative bias. Even if conservatives are more likely to break Twitter’s rules, he argues, it “doesn’t seem credible” the disparity would be so wide.

But let’s look a little more closely at this. These are the 22 accounts make up the data set:

1. Rose McGowan (the list’s lone Clinton supporter)
2. Azealia Banks
3. Tila Tequila
4. James O'Keefe
5. Richard Spencer
6. Baked Alaska
7. Roger Stone
8. Gavin McInnes
9. Candace Owens
10. Alex Jones
11. Chuck Johnson
12. Robert Stacy McCain
13. Milo Yiannopoulos
14. Radix Journal
15. National Policy Institute
16. Craig R. Brittain
17. David Duke
18. American Nazi Party
19. James Allsup
20. American Renaissance
21. Jared Taylor
22. Laura Loomer

Scanning the list, you probably noticed the “American Nazi Party.” This is not an anomaly. The bulk of the list is a who’s who of outspoken or accused white nationalists, neo-Confederates, holocaust deniers, conspiracy peddlers, professional trolls, and other alt-right or fringe personalities (go ahead, pick a couple and Google them). It does not include any mainstream conservatives, unless, I suppose, you count recently-indicted Trump campaign advisor and “dirty trickster” Roger Stone.

Reasons listed for banning these individuals in Hanania’s own data sheet include “violent threats,” “harassment,” “inciting violence,” “targeted abuse,” “doxxing,” “pro-Nazi tweets,” and “racist slurs.” Additionally, about a quarter of the accounts listed are still active and no longer suspended.

Kicking off a bunch of Nazis and trolls isn’t very compelling evidence that your average conservative is getting unfair treatment on Twitter. The majority of the “victims” here seem to have been engaged in abuse, and it’s reasonable for a private company like Twitter to kick off people who are undermining the quality of their platform by harassing or threatening other users.

Considering the alt-right’s propensity to scream and yell about getting “deplatformed,” these 22 accounts probably aren’t that representative of Twitter’s 67 million U.S. monthly active users. Nor does their small number (despite the author having two research assistants) indicate a broad, systemic problem.

Of course, social media companies may be not be perfectly neutral when it comes to politics. The Bay Area, where many of these companies are based, is a very liberal place. In 2016, only 9.4% of San Francisco County voted for Donald Trump. It’s entirely plausible that this disposition affects their products and policies in subtle ways. Yet, to date there has not been compelling evidence of systemic bias or a grand conspiracy to silence conservatives (despite this becoming a standard trope in congressional hearings and conservative conferences).

But social media platforms aren’t bastions of free speech, either. Their evolving norms and policies around content moderation raise a host of concerns and issues. At minimum, platforms could do a lot better at being transparent in their enforcement and governance decisions.

For conservatives, as I’ve argued before, crying wolf about censorship is a self-defeating strategy that will only make people not listen when it actually happens. Nazis, while sometimes useful in edge cases around free speech or references to Godwin’s law, are not stand ins for the median conservative American. Targeted abuse or incitements to violence are also not the same thing as free speech. Let’s not get these things mixed up.