Although the Constitution’s First Amendment prohibits government censorship, the public’s reliance on social media platforms has created new opportunities for censorship by proxy. Government officials increasingly bully platforms into suppressing disfavored speakers.

This “jawboning,” or application of informal government pressure, was once used to control prices. Now, it is being used to control speech, circumventing the First Amendment. Two recent revelations have drawn attention to this trend, but they are only the tip of the iceberg.

In the course of his lawsuit against Twitter, vaccine skeptic Alex Berenson discovered White House officials had asked, in Twitter’s telling, “really tough questions about why Alex Berenson hasn’t been kicked off the platform” in the months before his removal.

Weeks later, Mark Zuckerberg revealed on the Joe Rogan podcast that the FBI had warned Facebook about Russian hack-and-leak operations shortly before the New York Post exposed damning information found on Hunter Biden’s laptop. (This was the second time Zuckerberg made this claim; evidently, Joe Rogan’s podcast receives more attention than congressional hearings.)

Both disclosures have prompted concern about government censorship demands. However, as I detail in my recently published paper, “Jawboning Against Speech: How Government Bullying Shapes the Rules of Social Media,” this sort of pressure to censor is a much larger, longer-running problem. While the government’s request that Alex Berenson be removed from Twitter and the FBI’s disinformation warnings were delivered in private, even more forceful pressure has been applied in public.

In 2017, Sen. Dianne Feinstein (D‑CA) told social media companies: “You’ve created these platforms and now they are being misused, and you have to be the ones to do something about it, or we will.” She wasn’t just pushing them to remove Russian misinformation without regard for collateral damage. She was also starting a jawboning arms race in Congress.

Over the past four years, elected officials have become comfortable demanding platforms remove all sorts of disfavored speech. Sen. Ron Johnson (R‑WI) called on Twitter to remove absurd jokes about him. President Donald Trump threatened that unless platforms corrected perceptions of anti-conservative bias, he would “strongly regulate, or close them down.” The Biden administration demanded Facebook remove the accounts of Robert F. Kennedy Jr. and 11 other so-called misinformation “super spreaders.”

All these demands put platforms in a no-win situation. Although government officials can’t order platforms to remove disfavored speech, politicians have many ways of punishing platforms that resist their demands. Antitrust suits, punitive legislation, government contracts, and congressional investigations can all be used to harm platforms that fail to comply.

However, complying with one party’s requests may upset the other party and provoke new demands. Twitter did “something” about Russian hack-and-leak operations by creating new rules for suspected hacked materials. However, when these rules were applied to the New York Post’s story about Hunter Biden’s laptop, Twitter’s representatives were dragged back before Congress.

It is all but impossible for users to determine if their speech is removed because of government pressure or because they violated private platform rules. Social media content moderation is opaque and error-prone, given its massive scale. To the user, an account suspension looks the same, whether it is the result of an overzealous algorithm or a government official’s request.

The creation of the internet and social media unleashed a tidal wave of speech. For the first time in history, anyone could speak to a global audience cheaply and easily. In other countries, discontent with this newfound publishing potential was met with legislation. However, in America, the First Amendment prevents such a response. But, instead of melting away, censorship demands have been routed through informal channels.

This circumvention of constitutional speech rights cannot be easily addressed by the courts. Even when jawboned content moderation decisions can be identified, forcing platforms to refrain from removing controversial content tramples their First Amendment rights.

Additionally, the Constitution’s speech and debate clause grants congressional speech special protections, even speech used to jawbone. Instead, the best check on bullying by elected officials comes via the ballot box. This November, voters have an opportunity to select representatives who respect their speech, even when they disagree with it.