I have written this piece for the Washington Post. It begins:

In the Menlo Park, Calif., offices of Meta, discussions probably have already begun to consider what will happen Jan. 7, 2023, when former president Donald Trump’s ban from Facebook for encouraging the violent insurrection at the U.S. Capitol on Jan. 6, 2021, is set to potentially expire. Judging by how large social media companies have responded lately to the aftermath of the 2020 election and the looming 2022 election in which Republicans may take back control of Congress, there’s ample reason to worry Meta will restore the former president’s ability to post on Facebook — allowing him to continue to spread the false and dangerous claim that the 2020 election was stolen from him. Social media networks and other online platforms such as Google’s YouTube and Spotify can, instead, step up their support for reasonable measures to assure both vibrant political debate and protection of American election integrity and legitimacy. That would include keeping Trump off Facebook….

Social media and other new communications technologies are not solely to blame for the metastasizing election lies, but they play a big part. As the 2020 election season geared up and as Trump began spreading his false claims in the midst of the coronavirus pandemic that the election would be stolen or rigged, Facebook and Twitter reacted meekly. Rather than blocking Trump, they slapped labels on his posts saying his claims were disputed or directing voters to more information. Evidence indicates these labels may have backfired, amplifying Trump’s falsehoods and perhaps even suggesting to voters that they were correct. Things were even worse on other platforms: YouTube allowed videos with false accusations about the election to flourish, and its algorithm directed viewers to ever more extreme content. And those who distribute podcasts, such as on Spotify or Apple, appeared to do little policing of incendiary and dangerous election claims.

It took the actual violence of Jan. 6 for Facebook and Twitter to take action. Both chose to remove Trump from their platforms. Twitter made its ban permanent. Facebook initially did, too, but the Oversight Board it created to give it guidance on content told Meta that while deplatforming Trump was justified because he “created an environment where a serious risk of violence was possible,” the company needed criteria for removing politicians and conditions for determining the length of such bans. In response, the company announced that Trump would be booted for two years, followed by an evaluation as to whether he remained a “threat to public safety.” The company explained: “At the end of this period, we will look to experts to assess whether the risk to public safety has receded. We will evaluate external factors, including instances of violence, restrictions on peaceful assembly and other markers of civil unrest. If we determine that there is still a serious risk to public safety, we will extend the restriction for a set period of time and continue to reevaluate until that risk has receded.”…

Meta may soon face great political pressure from the right to show that they are being “fair” to Trump, especially with Republicans likely to take control of one or both houses of Congress after the 2022 elections and consider laws reining in tech platforms the GOP considers unfriendly. It could also have other incentives to let Trump back onto the site: Recent reporting by Judd Legum, for example, suggests that Facebook has not followed its own policies to prevent the viral spread of false political information, allowing fake groups to manipulate its rules to build up millions of followers to further spread election misinformation. Posts containing such misinformation are often among the most shared items on the platform.

It’s not just Facebook. To little fanfare, Twitter confirmed a few weeks ago that it will no longer police false election claims about the 2020 election, apparently because it believes such claims are no longer a threat to election integrity. Twitter told CNN that its civic integrity “policy is designed to be used ‘during the duration’ of an election or other civic event, and ‘the 2020 U.S. election is not only certified, but President Biden has been in office for more than a year.’ The staying power of the “big lie” and the rising threat of election subversion built on that lie shows how wrong that calculation is…

Companies such as Meta, Twitter and Google are private corporations, which have the right to decide what content to include, exclude, promote or demote on their platforms. They already do that with hate speech, pornography and violence. They need to continue to do that with speech threatening the integrity of American elections. Silencing a political leader should be the last resort, given our commitment to free speech and vibrant election contests. But Trump clearly crossed the line well before the Jan. 6 insurrection.