Trump’s Social Media Executive Order Is a Mistake

Earlier this week, Twitter crossed the digital Rubicon, albeit tepidly. After President Trump tweeted criticism of mail-in ballot proposals, Twitter attached a note to the bottom of the tweet inviting readers to fact-check the president’s claims and linking to a refutation of them. The White House responded by releasing an executive order that purports to “prevent[] online censorship” by social media platforms. While addressing a hot-button topic, the order abandons decades of conservative doctrine regarding free speech and private property. Thankfully, it appears to be long on bluster and short on action: When stripped of its overheated rhetoric, this executive order does little, if anything, to compromise online platforms’ constitutionally protected rights to editorial control.

What does the executive order say?

The executive order highlights the gatekeeper role that online platforms play when facilitating digital speech, and the resulting power they have to shape the flow of information online. It addresses a long-standing grievance held by many on the right that Silicon Valley companies and their executives largely favor progressive policies, and that their editorial judgments disproportionately censor conservative speakers and views. The executive order notes that Twitter’s officer in charge of “site integrity” has “flaunted his political bias” in his own tweets (though Twitter says the White House targeted the wrong guy).  

It is remarkable how much of this order appropriates rhetoric traditionally deployed by those on the left. For example, the idea that “we cannot allow a limited number of online platforms to hand pick the speech that Americans may access and convey on the internet” would fit comfortably within a brief supporting net neutrality. Indeed, when Professor Tim Wu first coined the phrase “net neutrality,” his primary concern was about ISPs exploiting their own gatekeeper status to suppress speech they disagree with. Similarly, the argument that “these platforms function in many ways as a 21st century equivalent of the public square” has often been deployed by those seeking to abrogate the state action doctrine and place First Amendment restrictions on private entities. And the notion that “we must foster and protect diverse viewpoints in today’s digital communications environment where all Americans can and should have a voice” contains echoes of the Fairness Doctrine, which required broadcasters to cover both sides of controversial issues.

In today’s hyper-politicized environment, one can understand the appeal of this rhetoric and the reality of the grievances they seek to nurse. Social media has emerged as one of our most volatile political battlegrounds, where emotions run high and any platform-based restriction on one’s advocacy can feel like a personal slight. Moreover, any gatekeeper’s perception of particular content will be informed by his or her priors, at least subconsciously. That’s the lesson of “They Saw a Protest,” a 2012 Stanford Law Review article in which people watching video of a political demonstration reacted differently when told they were protesting abortion outside an abortion clinic, or the military’s “don’t ask, don’t tell” policy outside a recruiting center. The article discusses the importance of “cultural cognition,” which it defines as “the unconscious influence of individuals’ group commitments on their perceptions of legally consequential facts.”

While acknowledging this reality, the executive order’s free speech rhetoric makes the same mistake made by net neutrality and Fairness Doctrine proponents: The First Amendment protects only against government limitations on free speech. With very limited exceptions, the Constitution does not require private parties to facilitate speech. I can write a letter to the editor, but The Boston Globe is not required to print it. A student can stand up to speak in my classroom, but I can tell him to sit down. And just because I have something to say, it doesn’t mean Twitter must allow me to say it. Furthermore, if the government were to force Twitter to let me speak against its will, that would violate Twitter’s own free speech rights. Free speech includes the right to prevent your property from being used to communicate a message with which you disagree: This is the First Amendment right of editorial control. As TechFreedom’s Ashkhen Kazaryan pithily put it, “The First Amendment protects Twitter from Trump. It does not protect Trump from Twitter.”

What does the executive order do?

While this high-level policy discussion is interesting and explores the sentiments behind the executive order, it is important to ask what, precisely, the executive order actually does to address the problem it cites. The heart of the order is to discourage politically-motivated censorship by threatening the protections provided by Section 230 of the Communications Decency Act. Section 230 provides two overlapping but distinct protections for internet companies that host third-party content:

In other words, Section (c)(1) allows hosting services like Twitter to post my tweets without risking a lawsuit if my tweet contains something false or misleading. (Of course, as the speaker, I could still be sued.) Section (c)(2) protects Twitter from my wrath if it takes down a post that it decides, in good faith, is objectionable. Importantly, Twitter retains liability for any content that it creates directly — including, incidentally, the “fact-check” that it attached to President Trump’s mail-in vote tweet.

The executive order seeks two changes to the Section 230 framework. First, it seeks to clarify that Section (c)(2)’s “good faith” exception cannot be used as a pretext to stifle viewpoint discrimination. It’s unclear how “viewpoint discrimination” would be determined. But by itself, this change may be insignificant, as few cases involve the (c)(2) safe harbor. Second, the order seeks to import the “good faith” restriction from (c)(2) to (c)(1). This would mean companies engaged in viewpoint discrimination would lose all immunity from liability for third-party content.

This second change is potentially important. Professor Jeff Kosseff has noted, without hyperbole, that Section 230(c)(1) contains “The Twenty-Six Words that Created the Internet.” Platforms allow us to post, and disseminate our views to our networks, precisely because Section 230 protects them from liability for what we say. Stripping Twitter of its Section 230 immunity will lead to more censorship, not less: If Twitter is potentially liable for what I say, it will read it closely and approve every tweet before publishing. It may approve fewer tweets and only after lengthy delays. More likely, the risk of residual liability would pose an existential threat to social media, eliminating Twitter as a tool for discourse. One might respond that Twitter could retain its immunity simply by eliminating viewpoint discrimination. But given the elasticity of this term, and in light of the cultural cognition discussion above, this safe harbor would likely be insufficient. And the fallout would not be limited to Twitter, but to any site that hosts third-party content, including YouTube, any newspapers or blogs with comment threads, online marketplaces such as eBay and Craigslist, and so on.

As is often the case, the law of unintended consequences matters. The executive order would not make social media safer for conservatives. Rather, it could prove the 21st century equivalent of destroying the village in order to save it.

OK, but what does the executive order really do?

Fortunately, the executive order does not actually change Section 230. The statute itself can only be changed by Congress. And although administrative agencies have significant leeway to alter the law by reinterpretation of statutes (which President Obama famously referred to as governing “by pen and phone”), the relevant agencies here are independent of direct presidential oversight. This significantly limits the ultimate effectiveness of the order.

When stripped of its rhetoric, the executive order contains only two legally significant action items:

• It directs the Commerce Department to file a petition for rulemaking requesting that the Federal Communications Commission (FCC) clarify the scope of Section 230 as outlined in the order; and
• It asks the Federal Trade Commission (FTC) to take action under its “unfair or deceptive acts” authority to make sure online platforms do not restrict speech in ways that do not align with their terms of service.

Neither of these requests is likely to result in action. The FTC already monitors online companies’ compliance with their terms of service, and importantly, those terms typically include rules about what can and cannot be posted. In other words, Twitter users agree to editorial control as a condition of using the site.

While the government may file a petition for rulemaking with the FCC, nothing compels the FCC to respond to the request. There is little incentive for the FCC to do so, especially as the request runs contrary to the strong First Amendment protections that the agency has traditionally extended in previous controversies over, for example, net neutrality and the Fairness Doctrine. Moreover, even if it were to do so, it is unlikely to survive judicial review: the whole point of Section 230 was to allow websites to self-censor material without liability. In debate on the bill, co-author Chris Cox stated clearly that his intent was not to create “a Federal Computer Commission with an army of bureaucrats regulating the internet.”

Conclusion

It is worth noting that Section 230 is a quarter-century old. There is an important debate to be had about whether and how it might be modified for the current digital environment. But this executive order does not reflect that nuanced discussion. It is, instead, a proclamation long on rhetoric and short on action. And while it is unlikely to result in any appreciable change to the current doctrine, perhaps that’s acceptable to the White House. The President is reaping the political benefit of addressing a concern that has long bothered his base. And policymakers and academics — including me here — are investing a news cycle in surfacing and showcasing those concerns. For a president facing re-election, that may be sufficient to claim victory.