Social Media Power and the Newspaper/Broadcaster Analogy
Another excerpt from my Social Media as Common Carriers? article (see also this thread). We're still on Part I, which asks whether it might make sense to mandate viewpoint neutrality of social media platforms, in at least some contexts—but if you're waiting for the First Amendment analysis of whether such mandates are constitutional, we'll be turning to that shortly.
[* * *]
Now of course there have also long been rich, powerful organizations with the power to influence public debate: newspapers and broadcasters. Everything Justice Stevens said about business corporations generally in Citizens United also applies to newspaper corporations.
To be sure, some media outlets, such as magazines of opinion, acquire power because their audience agrees with their views; that might distinguish them from other businesses whose power has "little or no correlation to the public's support for the corporation's political ideas." But that's not so for many of the most powerful media outlets, such as newspapers: Especially when there was only one newspaper in town, people often subscribed to it because of its classifieds, coupons, TV listings, or nonpolitical local affairs coverage, not because they agreed with its ideology.
Yet newspapers and broadcasters have long been seen as entitled to pick and choose which opinions to publish. And I support the continued freedom of newspapers and broadcasters to make such choices, both as a policy matter and as a First Amendment right (more on that in the next Part). Newspapers and broadcasters shouldn't be seen as common carriers, because of three related features of such media.
[1.] The limited space in a newspaper and limited time on a broadcast channel make editorial judgment necessary. A newspaper can't publish all the items submitted to it (especially given how many submissions it would get if it had such an obligation). Nor can a newspaper even adhere to a viewpoint neutrality rule, given the number of viewpoints, thoughtful or crank, that could be submitted on any subject. Twitter and Facebook, on the other hand, can host all viewpoints. When they exclude certain viewpoints—a tiny fraction of all the items that are posted on their services—they are certainly not doing it to save disk space.
[2.] Readers and viewers rely on newspapers and broadcasters to help avoid information overload, as well as to exclude material that readers and viewers might find offensive or useless. For these publishers to be useful to the public, they need to publish 1% (or perhaps much less) of all the viewpoints available for them to publish.
Twitter and Facebook, on the other hand, even when they delete some views, likely distribute 99% of all the viewpoints submitted to their services. Readers don't count on social media platforms to fight information overload using their hosting decisions (though readers may count on social media platforms to select the most interesting material in their recommendation decisions).
[3.] Both for newspapers and particular broadcasts, readers and viewers tend to consume them as a coherent product—they may read a newspaper (or at least a section) cover to cover, or watch a whole half-hour newscast, or even keep a channel on for hours on end. They do this to get an aggregate speech product, "today's news" (or perhaps "this weekly magazine's viewpoint on the past week's news"), again trusting the publisher's editorial judgment. Requiring publishers to include certain material in the product denies viewers the coherent speech product that they seek.
The major platforms, on the other hand, are not generally in the business of providing "coherent and consistent messaging" the way that, say, an ideologically minded magazine or cable news channel might be. Even if Facebook and Twitter deliberately exclude some viewpoints, the aggregate of all the material they host is very far from coherent and consistent.
Those, then, are the reasons the platforms' hosting function differs from the media's function. But our legal system has also long been concerned about concentration of economic power, and its influence on democratic debate, even among the media. We've seen that in the FCC's media cross-ownership rules, which aimed to limit consolidation of newspaper and broadcaster power within the same city. We've seen it in attempts to use antitrust law to preserve two-newspaper towns, for instance by insisting on joint operating agreements that maintain each newspaper's editorial independence even when the newspapers must merge for financial reasons. We've seen it in the Fairness Doctrine, the Personal Attack Rule, the requirement that broadcasters sell ads to political candidates, and the requirement that they not censor ads bought by political candidates.
Not all these policies, to be sure, are good or constitutionally permissible. I think the Fairness Doctrine, for instance, was a mistake, partly because it required the government to decide (1) when some speech by broadcasters triggered a duty to present other speech, and (2) which other speech broadcasters needed to air in response. Broadcasters had a duty to devote time to "controversial issues of public importance," and then had a "duty to present responsible conflicting views." Policing those vague lines would itself inevitably involve viewpoint discrimination, unless the FCC were prepared to insist, for instance, that any pro-tolerance views triggered a duty to present racist views.
The Fairness Doctrine also limited broadcasters' own speech, both by diminishing the time available for that speech and by creating a disincentive to air controversial speech that would yield Fairness Doctrine obligations. I'm glad that Reno v. ACLU essentially foreclosed the implementation of a Fairness Doctrine for Internet sites. But a rule of viewpoint neutrality in platform hosting, where platforms have essentially unlimited storage space—unlike broadcasters, who face a sharply limited broadcasting day—wouldn't pose the same problems.
Yet in any event, the history of these policies, with all its missteps, shows an enduring concern about undue domination of political debate by powerful economic entities. Congress and the courts have not treated broadcasters as "common carriers" because of the countervailing reasons for publisher editorial discretion that I laid out above. But while that decision makes sense for such publishers, it may not make sense for other platforms.
And indeed it seems to me that social media platforms, in their hosting function (rather than their recommendation function), are more like phone companies (whether often-monopoly landlines or competitive cellular companies) than like newspapers or broadcasters. They can indeed host all viewpoints. Their decisions to stop hosting certain feeds doesn't help readers deal with information overload, since it reduces the number of hosted feeds by only a small fraction. And people don't go to Facebook and Twitter to see the aggregate of all the pages they host, the way some people read a newspaper section or a magazine cover to cover.
Moreover, there are hundreds of newspapers throughout the nation and several major TV networks. Facebook and Twitter have no major rivals in their media niches. Indeed, actions such as Amazon Web Services', the Google Store's, and the Apple Store's blockade of Parler help make such near-monopoly status likely to endure.
 Citizens United, 558 U.S. at 883 (quoting Austin v. Michigan Chamber of Commerce, 494 U.S. 652, 660 (1990)).
 In Benjamin Franklin's words, "having contracted with my subscribers to furnish them with what might be either useful or entertaining, I could not fill their papers with private altercation, in which they had no concern, without doing them manifest injustice." Benjamin Franklin, The Autobiography of Benjamin Franklin 88 (1850).
 "Social media providers, in contrast, routinely use algorithms to screen all content for unacceptable material but usually not for viewpoint, and the overwhelming majority of the material never gets reviewed except by algorithms. Something well north of 99% of the content that makes it onto a social media site never gets reviewed further." NetChoice, LLC v. Moody, No. 4:21CV220-RH-MAF, 2021 WL 2690876, *8 (N.D. Fla. June 30, 2021).
 Ashutosh Bhagwat, Do Platforms Have Editorial Rights?, 1 J. Free Speech L. __, 5-6 (2021).
 See, e.g., FCC v. Nat'l Citizens Comm. for Broadcasting, 436 U.S. 775, 780 (1978) (discussing FCC's interest in "preventing undue concentration of economic power" in "mass media ownership"); Multiple Ownership of Standard, FM and Television Broadcast Stations, 45 FCC 1476, 1476–77 (1964).
 Newspaper Preservation Act, 15 U.S.C. §§ 1801 et seq.
 Red Lion Broad. Co. v. FCC, 395 U.S.367, 370 (1969).
 Id. at 378.
 47 U.S.C. §§ 312(a)(7), 315(a).
 Id. § 315(a).
 See generally Thomas G. Krattenmaker & Lucas A. Powe, Jr., Regulating Broadcast Programming 240–74 (1994); Samples, supra note 38, text accompanying nn.47–48.
 NBC v. FCC, 516 F.2d 1101, 1110 (D.C. Cir. 1974); In re Responsibility Under the Fairness Doctrine, 40 FCC 576, 577 (1963).
 521 U.S. 844, 868–69 (1997).
 Krishnamurthy & Chemerinsky, supra note 29. To be precise, server space does cost money; but there's little reason to think that the relatively few items that platforms choose to block based on viewpoint occupy a material fraction of the space that Facebook uses.
 See, e.g., FCC v. Midwest Video Corp., 440 U.S. 689 (1979); 47 U.S.C. § 153(11).
 "[T]he power that platforms such as Twitter and Facebook possess is far greater than that of individual broadcasters who compete with one another as well as with satellite and cable networks." Krishnamurthy & Chemerinsky, supra note 29.
 See supra note 52.
 For an argument that such monopoly status is overstated, and in any case is unlikely to endure, see Samples, supra note 38, at text accompanying nn.38–42. Epstein, supra note 9, at 5, responds briefly, expressing some sympathy for Samples' position but also some doubt.