In the past, a Twitter employee suggested the company couldn’t ban white nationalist, alt-right, and other right-wing hate speech and terrorist groups on Twitter because their algorithms would also sweep up Republican politicians. The party has become so aligned with the dangerous alt-right that the two have become indistinguishable. However, Twitter’s situation is worse than unequal application of their own rules. Twitter’s own research shows they boost conservative views on the platform.
In the United States, far-right ideas and conspiracy theories have worsened the COVID-19 pandemic, lead to an attempted coup to install Donald Trump as president, and has increased the political divide between conservatives and both moderates and liberals. Furthermore, among marginalized communities, Twitter is infamous for its reluctance to do anything about harassment, especially if targets of hate are smaller, personal accounts. Both Twitter and Facebook have a problem with enforcing their own rules, and, as it turns out, it may have lead to boosting right-wing voices over others.
Twitter’s Own Troubling Findings
Twitter conducted a study to examine the bias in their algorithmic timeline. Since 2016, there have been two ways to view Twitter. The first is what you’d expect: reverse chronological order. You see the most recent tweets first, and older tweets as you scroll. However, this isn’t the default. Instead, you get an algorithmically-generated feed. This finds content from people you follow, their interactions, and content Twitter thinks you’ll likely interact with. To ensure it doesn’t have bias, and could potentially push users to one political ideology over another, Twitter studied their algorithmic feed.
Instead, they found that it could introduce a pro-conservative bias.
Twitter’s study looked into politicians and news organizations from seven countries: Canada, France, Germany, Spain, the U.K., and the U.S. They used AllSides and Ad Fontes Media, third party organizations, to sort news organizations along bias lines. Meanwhile, they used the political affiliation of politicians to measure bias. Twitter found that, in six out of seven countries, they increased the amount of right-leaning content users would see. Instead of a fair, unbiased view, Twitter was presenting conservative voices over liberal ones in all countries in their study with the exception of Germany.
It’s worth noting that, because of Germany’s anti-Nazi and anti-hate laws, Twitter monitors content more closely in Germany. They can’t allow pro-Nazi speech or hate speech, as they can in other countries. This is potentially why Germany was the only country that did not show this bias.
Turns out when you filter out violence, Twitter won’t boost right-wing ideals as much. What does that say about modern conservatism?
What Twitter’s Doing About Bias
In short? Nothing. That is, nothing to the service itself will change anytime soon. Twitter did call the unfair amplification of only one political side’s opinions as “problematic” and stated they need to do “root cause analysis” next. They know what is happening, that conservative voices have an advantage on Twitter, but not precisely why. All Twitter really knows is that the home page, as it is now, amplifies right-leaning voices. Further research could figure out what caused this, and how to fix it within the current system. That could be promising in the future, but currently, it’s allowing a problem to exist and continue to grow. It’s also something harder to study, as why a person’s content is more engaging or spreads more than another person’s view is difficult to quantify and predict. You might be able to say that a republican’s tweets get more interactions, but that doesn’t necessarily say that their content is more engaging, especially if Twitter’s feed algorithm is already providing a boost.
Twitter’s own research here suggests they could fix the problem overnight by looking at what makes Germany different. Germany was the only country in the study that didn’t have clear bias. It’s also the only nation that has specific anti-hate rules on Twitter.
Its possible that Germany’s specific anti-hate enforcement changes the way right-wing politicians and pundits reach their audience. They can’t do it with the same violence-inspiring vitriol that other politicians and pundits use. Even if Twitter’s algorithm is not to blame, the hate speech permitted on the platform allows the right to spread a potent message to their followers, who are more likely to amplify it.
Twitter’s standard, non-chronological feed promotes material it believes you’re most likely to interact with. This includes things you agree with as well as things you do not. The latter, which can range from harassment to fake news, is most frequently aligned with the far-right. These are opinions that pull from an aggressively supportive base and draw ire from everyone left of the center-right, inclusively. Extremist, far-right believes, are simply more profitable, from an engagement perspective, and neither Twitter nor Facebook have been willing to give that up. If they were, policies that align with Germany’s strict anti-hate laws would be worldwide. Until dangerous content is no longer profitable, social networks aren’t going to ban it.
A Conservative Takeover on Social Media
Twitter and Facebook are just two of the sites that have found right-wing content to be more lucrative than banning hate speech or COVID-19 misinformation and conspiracies. These extreme far-right opinions have lead to harm in the real world, including violence and deaths. In fact, misinformation spread on Facebook, which outnumbered real stories and skewed heavily conservative, likely contributed to Donald Trump winning an electoral victory in 2016, becoming the U.S. president for one term. Meanwhile, groups on Facebook planned out the January 6th attack on the U.S. Capitol, in an effort to reinstate Donald Trump, despite his election loss in 2020.
Twitter’s transparency is to be commended. The fact that they’re doing little to fix the problems they’ve created, however, is just as damning as hiding that knowledge. Twitter has always known the solution to their problems. They harbor hate. The only country that, through legislation, has banned hate, is also the only country without right-wing bias. Twitter knows what to do: ban the hate. Too bad hate is profitable, and in nations where politics and business share an uncomfortably close relationship, that’s all that matters.