Why Twitter isn't 'shadow banning' Republican lawmakers

Daniel Howley
Technology Editor

Twitter (TWTR) CEO Jack Dorsey took to the main stage during the House Commerce and Energy Committee’s hearing on the social media network’s content monitoring practices and algorithm use on Wednesday. As expected, it was largely a farce.

The impetus for the meeting was to question Dorsey about whether Twitter has a bias against conservative voices, a theory that’s been gaining steam among Republican politicians as of late. But in truth, it’s all smoke and mirrors.

To hear Republicans tell it, Twitter was blocking conservative accounts en masse, via so-called “shadow banning.” Shadow banning is a way to shut down disruptive users on forum websites by hiding what they post. Since a person can always create a new account when their original is deleted and continue bothering users, the moderator keeps the offending user’s account active, but prevents anyone from being able to see what they wrote. In some cases, the disruptive user might never know they’ve been shadow banned.

Twitter CEO Jack Dorsey testified about ‘shadow banning’ before members of the House Energy and Commerce Committee.

In Twitter’s case, conservatives argue that the social network was keeping conservative politicians’ accounts hidden. But that’s not the case at all. And doing so would only hurt Twitter as a company.

A shadow ban sideshow

Here’s what happened with this supposed shadow ban. When you search for a Twitter user, typing their name into the Twitter search bar will provide an auto-completed search, which makes it easier to find individuals.

In this incident, more than 600,000 Twitter users’ names were not showing up in the autocomplete search. For example, if you typed in the name of Representative Mark Meadows (R-NC), his name wouldn’t appear in the search bar.

If you performed the full search, though, you could still find Meadows, as well as the rest of the impacted accounts. Similarly, if you followed such users, you were still able to see their tweets just as you normally would.

Following a Vice News report on the issue, however, conservative lawmakers latched on to the topic, with Trump even saying that Twitter was shadow banning Republicans and that doing so was “Not good.”

Compounding the matter was Dorsey’s interview with CNN’s Brian Stelter last month during which the CEO acknowledged that Twitter’s employees have their own left-leaning bias, but that their point of view is not reflected in the end product.

“We do not look at content with regards to political viewpoint or ideology,” Dorsey told Stelter. “We look at behavior, and we look at behavior as a signal to add to relevance. We need to constantly show that we are not adding our own bias, which we fully admit is more left-leaning…but we need to remove all bias from how we act, and our policies and our enforcement.”

Despite his explanation, the portion of Dorsey’s statement that gained the most traction was that Twitter employees have a left-leaning bias.

Twitter would be hurting itself

In a blog post written on July 26, Twitter’s legal, policy, trust and safety lead Vijaya Gadde and product lead for Twitter and co-founder of Periscope Kayvon Beykpour explained that the autocomplete search problem likely arose as a result of the way users interacted with the 600,000 impacted accounts.

“There are communities that try to boost each other’s presence on the platform through coordinated engagement,” the post explained. “We believe these types of actors engaged with the representatives’ accounts—the impact of this coordinated behavior, in combination with our implementation of search auto-suggestions, caused the representatives’ accounts to not show up in auto-suggestions.”

Essentially, Twitter was punishing accounts as a result of the way other users follow them. Dorsey confirmed as much on Wednesday and said doing so was not fair to the impacted accounts.

For Twitter, purposely hiding the accounts of conservatives would be detrimental to the company as a business. Twitter relies on advertising to generate revenue. It builds that up by having a large user base. By cutting out conservative voices, Twitter would be hurting itself.

Twitter has done this when it comes to bot accounts and accounts run by foreign actors trying to influence American politics, but it hasn’t done so with normal users unless they have broken the company’s user guidelines.

In this instance, Twitter didn’t seek to silence conservative voices. Instead, it used an algorithm designed to improve the network, which directly impacted how some users showed up in autocomplete searches.

There’s no great conspiracy here. Regardless of how much some want there to be.

More from Dan:

Email Daniel Howley at dhowley@oath.com; follow him on Twitter at @DanielHowley. Follow Yahoo Finance on Facebook, Twitter, Instagram, and LinkedIn