-Analysis-
PARIS — Elon Musk is the world’s leading reckless driver. The ever unpredictable CEO of Tesla and SpaceX is now behind a very different wheel as the new head of Twitter.
He began by banning remote work before slightly backtracking and authorizing it for the company’s “significant contributors.” Now he’s opened the door to Donald Trump to return to Twitter, while at the same time vaunting a decrease in the number of hate-messages that appear on the social network…all while firing Twitter’s content moderation teams.
But this time, the world’s richest man will have to make choices. He’ll have to limit his otherwise unconditional love of free speech. “Freedom consists of being able to do everything that does not harm others,” proclaimed the French-born Declaration of the Rights of Man in 1789.
Yet freedom on social networks results not only in insults and defamation, but sometimes also in physical aggression.
Indeed, researchers have accumulated an abundance of evidence to prove the effects that’s becoming impossible to deny.
Trump the inciter
Donald Trump himself is a magnificent case study. The tweetomaniac and former U.S. President sent more than 50,000 tweets to his over 90 million followers, with sometimes undesirable effects. Not just his Dec. 19, 2021 messages calling for a large protest in Washington on January 6 to protest against alleged (but systematically disproven) electoral fraud: “Be there, will be wild!”
After the 2016 primary, the number of hate crimes against Muslims doubled in the United States
The ex-casino owner often took aim at minorities. Karstel Müller, of the University of Singapore, and Carlo Schwarz of Bocconi University in Italy, scrutinized Trump’s anti-Muslim tweets. After the 2016 primary, the number of hate crimes against Muslims doubled in the United States. And the increase was strongest in the counties where residents were most active on Twitter.
The researchers used statistical tests to verify that this increase in hate crime didn’t have other causes. Then, they examined the link between presidential tweets and assaults. Trump was more likely to tweet anti-Muslim statements when he played golf (which he did one out of every four days in 2017), undoubtedly because he was away from the White House, or while in the company of Dan Scavino, his former caddie turned social media advisor.
Virus of violence
The presidential tweets were followed by an increased attention placed on Muslims in television news media, most notably on Fox News. And ultimately, it led to an increase in assaults against Muslims.
Meanwhile, a team of economists at Texas A&M University worked on another strand of Trump messages: those targeting the “China virus” in March 2020, when the COVID-19 pandemic first made inroads in the United States.
From March 16-18, Trump published six tweets focused on the “China virus,” an expression that had not appeared in public discourse until then. On March 19, the United States registered a spike in Google searches for “China virus,” often followed by Trump’s name. And March 20, a spike in assaults against Asians, twice as many as the day before.
Russian racism on VKontakte
This isn’t specific to Trump, or even to the United States, Ruben Enikolopov and Maria Petrova, fro the Institute for Political Economy and Governance (IPEG) in Barcelona, showed in work done with Leonardo Bursztyn (University of Chicago) and George Egorov (Northwestern University) that there was a link between social networks and racist assaults in Russia.
Greater penetration of social media led to increased hate-crimes with a racist nature.
The researchers studied VKontakte, the Russian alternative to Facebook, which was unevenly developed in different Russian cities. “Greater penetration of social media led to increased hate-crimes with a racist nature, and this effect was stronger in the cities that had greater nationalist sentiment before the arrival of the social network,” the researchers concluded.
Like in the United States, social media messages reinforce existing beliefs and incite people to act on them, more than they actually create the beliefs in the first place. Other research has shown the same mechanisms at work in Germany.
How did Germany regulate speech on social media?
Even if banning social networks isn’t in the realm of possibility, moderating them is. That’s what Germany mandated in 2017 after the arrival of more than one million refugees and a torrent of racist messages on social networks. The “Netzwerkdurchsetzungsgesetz” (NetzDG) law obliged social networks to remove hate messages within 24 hours of their being reported.
The law was a game changer in Germany. The two researchers who studied Trump’s anti-Muslim tweets, Müller and Schwartz, examined what happened after the law’s entry into force, this time with Rafael Jiminez-Duran (from the Social Science Research Council). According to their recently published paper, the law has been associated with a decrease in the number of toxic messages, and a decrease in anti-refugee crimes in cities with a large number of far-right Facebook users.
Elon Musk is going to have to do something he’s not very used to: find a balance, between freedom of speech and responsibility. The good news for him and other social media titans is that, while it might be complicated, it can be done.