But the court shifted again, Lakier says, toward interpreting the First Amendment “as a grant of almost total freedom” for private owners to decide who could speak through their outlets. In 1974, it struck down a Florida law requiring newspapers that criticized the character of political candidates to offer them space to reply. Chief Justice Warren Burger, in his opinion for the majority, recognized that barriers to entry in the newspaper market meant this placed the power to shape public opinion “in few hands.” But in his view, there was little the government could do about it.
Traditionally, conservatives have favored that libertarian approach: Let owners decide how their property is used. That’s changing now that they find their speech running afoul of tech-company rules. “Listen to me, America, we were wiped out,” the right-wing podcaster Dan Bongino, an investor in Parler, said in a Fox News interview after Amazon pulled its services. “And to all the geniuses out there, too, saying this is a private company, it’s not a First Amendment fight — really, it’s not?” The law that prevents the government from censoring speech should still apply, he said, because “these companies are more powerful than a de facto government.” You needn’t sympathize with him to see the hit Parler took as the modern equivalent of, in Burger’s terms, disliking one newspaper and taking the trouble to start your own, only to find no one will sell you ink to print it.
One problem with private companies’ holding the ability to deplatform any speaker is that they’re in no way insulated from politics — from accusations of bias to advertiser boycotts to employee walkouts. Facebook is a business, driven by profit and with no legal obligation to explain its decisions the way a court or regulatory body would. Why, for example, hasn’t Facebook suspended the accounts of other leaders who have used the platform to spread lies and bolster their power, like the president of the Philippines, Rodrigo Duterte? A spokesman said suspending Trump was “a response to a specific situation based on risk” — but so is every decision, and the risks can be just as high overseas.
“It’s really media and public pressure that is the difference between Trump coming down and Duterte staying up,” says Evelyn Douek, a lecturer at Harvard Law School. “But the winds of public opinion are a terrible basis for free-speech decisions! Maybe it seems like it’s working right now. But in the longer run, how do you think unpopular dissidents and minorities will fare?”
Deplatforming works, at least in the short term. There are indications that in the weeks after the platforms cleaned house — with Twitter suspending not just Trump but some 70,000 accounts, including many QAnon influencers — conversations about election fraud decreased significantly across several sites. After Facebook reintroduced a scoring system to promote news sources based on its judgment of their quality, the list of top performers, usually filled by hyperpartisan sources, featured CNN, NPR and local news outlets.
But there’s no reason to think the healthier information climate will last. The very features that make social media so potent work both to the benefit and the detriment of democracy. YouTube, for instance, changed its recommendation algorithm in 2019, after researchers and reporters (including Kevin Roose at The New York Times) showed how it pushed some users toward radicalizing content. It’s also telling that, since the election, Facebook has stopped recommending civic groups for people to join. After Jan. 6, the researcher Aric Toler at Bellingcat surfaced a cheery video, automatically created by Facebook to promote its groups, which imposed the tagline “community means a lot” over images of a militia brandishing weapons and a photo of Robert Gieswein, who has since been charged in the assault on the Capitol. “I’m afraid that the technology has upended the possibility of a well-functioning, responsible speech environment,” the Harvard law professor Jack Goldsmith says. “It used to be we had masses of speech in a reasonable range, and some extreme speech we could tolerate. Now we have a lot more extreme speech coming from lots of outlets and mouthpieces, and it’s more injurious and harder to regulate.”
For decades, tech companies mostly responded to such criticism with proud free-speech absolutism. But external pressures, and the absence of any other force to contain users, gradually dragged them into the expensive and burdensome role of policing their domains. Facebook, for one, now has legions of low-paid workers reviewing posts flagged as harmful, a task gruesome enough that the company has agreed to pay $52 million in mental-health compensation to settle a lawsuit by more than 10,000 moderators.