On August 6, “X Corp.,” the company formerly known as Twitter (and controlled by Elon Musk), filed a wide-ranging federal-court antitrust lawsuit against the World Federation of Advertisers (“WFA”) as well as a number of large, international corporations. The basic claim made in X’s complaint is that the WFA and corporations who belong to it, acting through a WFA committee called the Global Alliance for Responsible Media (“GARM”), organized an economic boycott of X in violation of U.S. antitrust laws. In the space below we look at the serious barriers to X’s lawsuit both posed by current Supreme Court doctrine under antitrust law, and under the First Amendment.
The WFA is a nonprofit trade association representing most of the largest purchasers of advertising in the world. In June of 2019, the WFA, in coordination with a group of large advertisers and nearly all major advertising agencies, established GARM as a committee of the WFA. GARM’s purpose was “to improve the safety of online environments.” In practice, that meant that GARM developed “brand safety” standards identifying harmful online content which advertisers would not want to be associated with (think pornography or Nazi propaganda), and monitored platforms to ensure that GARM members’ advertising did not appear in connection with such harmful content. GARM also worked with digital media and social media platforms to confirm that those platforms’ advertising practices complied with GARM standards.
The events that led to this lawsuit arose when Elon Musk purchased Twitter (and eventually changed its name to X), a transaction that was announced in April of 2022 and closed on October 27, 2022. In light of certain public comments made by Musk, concern arose among WFA members that, under Musk’s control, Twitter would no longer abide by GARM standards regarding harmful content. In response, on October 31, 2022 the WFA posted a public letter to Musk calling on Twitter to abide by its commitments to brand safety, and the head of GARM posted a similar public statement (ironically, on Twitter). Musk responded that “Twitter’s commitment to brand safety remains unchanged.”
Nonetheless, in the weeks following the Musk takeover concerns about harmful content on Twitter grew (X’s complaint does not mention this but press reports at the time suggested that Twitter’s content moderation policies had relaxed markedly under Musk’s ownership, especially after Musk fired a huge percentage of Twitter’s workforce, including its content moderation staff, in November of 2022). As a result, GARM surveyed its members to determine their experiences with the new Twitter, and several GARM members contacted GARM to solicit its views on Twitter’s compliance with brand safety standards. Ultimately, a number of major advertising agency holding companies recommended suspending advertising on Twitter because of brand safety concerns, and most major advertisers, some after consulting with GARM, suspended advertising on Twitter.
Before considering the First Amendment barriers to X’s lawsuit, it should be noted it is at least doubtful if these alleged facts—which are largely pulled from X’s own complaint—establish an antitrust violation at all. Under current doctrine, the key to establishing a violation of Section 1 of the Sherman Act, the antitrust law invoked by X, is to prove the existence of an agreement (typically among competitors) that has an anticompetitive purpose and effect—the language of Section 1 prohibits “every contract, combination . . . or conspiracy, in restraint of trade or commerce.” The archetypal example of such an agreement is for competitors within an industry to agree (effectively as a cartel) on a price they will charge for a particular good or service. X’s complaint, in essence, argues that the defendants in this case, working through GARM, engaged in a “group boycott” of Twitter/X with unlawful anticompetitive economic goals.
But is that really a plausible reading of this scenario? One obvious problem is that the major advertisers who make up the WFA and GARM, and are among the defendants in X’s lawsuit, are not competitors with respect to the goods and services they provide. They include a Danish renewable energy company, Mars (famous for its candy), and CVS Health (famous for its pharmacies). While all these firms of course buy advertising and thus participate in the market for ad space, they do not serve the same customers. As such, it is not at all clear that the coordination among them alleged by X (assuming for now that there was coordination) constitutes the sort of agreement to advance collective economic self-interest that is at the heart of Section 1’s prohibitions.
Indeed, there is one strong indicator that participants in GARM did not see themselves as participating in such an agreement. When competitors establish the type of price-fixing cartels at the core of Section 1’s prohibitions, they inevitably do so in secret (at least after 1890, when the Sherman Act was enacted). And X’s complaint suggests that the entities it sued were engaged in a similarly nefarious undertaking by regularly describing their actions as a “conspiracy.” But in fact, GARM’s actions were all public and highly visible—remember, GARM’s head tweeted about his concerns following Musk’s takeover. It is true that within a week of X’s lawsuit being filed GARM shut down; but at least according to GARM that was due the financial burden of defending the lawsuit (GARM, remember, is nonprofit), not any admission of guilt.
Furthermore, it is at least questionable whether the facts as alleged by X do make out the existence of the requisite “agreement” or “conspiracy” needed for a Section 1 lawsuit to succeed. Although GARM established standards, and its members generally did not advertise with platforms that failed to meet those standards, there is no indication that the advertisers agreed with one another—or even discussed joint actions—to boycott noncompliant platforms. Nor is it obvious why it would have been in their economic interest to conspire; X’s complaint itself concedes that X’s advertising prices were lower than their competitors. As such, it would seem more accurate to characterize GARM as a sort of standard-setting body for platform quality, compliance with which was entirely voluntary among members.
There is, however, one caveat which makes the agreement question harder. In a 1939 decision called Interstate Circuit v. United States, the Supreme Court recognized the possibility that Section 1 of the Sherman Act can be violated when competitors all agree with a central actor to act in an anticompetitive way—what is today called a “hub-and-spoke conspiracy.” And that seems to be the kind of conspiracy X is alleging, with GARM acting as the hub to try and force Twitter/X to raise the quality of the advertising services it was selling.
Even in the hub-and-spoke situation, however, the cases are clear that there must exist some agreement among competitors—the so-called “rim”—to abide by the anticompetitive agreement. Admittedly, the cases do not require evidence that competitors actually communicate with each other to establish an agreement. But the cases do typically involve proof that the “hub” coordinated agreement, often by assuring each competitor that the others had agreed to stick to the agreement, as long as everyone else did. And it is just not clear that there are enough facts in the X complaint to make out such a horizontal agreement among the advertisers who made up the rim of the alleged hub-and-spoke conspiracy.
Even assuming that the decisions by GARM and the corporate defendants constitute some kind of mutual agreement not to advertise on X—rather than (as seems quite plausibly the case) separate and independent, albeit similar, decisions by each of the entities to abide by GARM’s principles in the context of X—the antitrust lawsuit faces hurdles arising from the intersection of the regulation of business activity and the First Amendment.
In at least two separate lines of cases, the Supreme Court has recognized that coordinated activity by business firms or individuals that might impact marketplace competition is nonetheless insulated from government regulation under First Amendment principles. In the so-called Noerr-Pennington line of cases, the Court has held that coordinated lobbying activities by businesses lie outside the scope of the federal antitrust laws, even if the object of the lobbying is the enactment of policies and regulations that would have anti-competitive consequences. So, for example, in the 1961 Noerr case (Eastern R. Conference v. Noerr Motors), the Court held that a public relations campaign orchestrated by railroad companies, who were in competition with each other and also with the trucking industry, could not give rise to antitrust liability even if the goal of the PR campaign was the enactment of regulations that would burden the trucking industry so as to hinder competition between it and railway transportation. As the Court has described the essential teaching of Noerr, “at least insofar as the railroads’ campaign was directed toward obtaining governmental action, its legality was not at all affected by any anticompetitive purpose it may have had.”
The Noerr doctrine relating to lobbying and petitioning government for redress does not have direct application to the X lawsuit, insofar as the defendants’ actions pulling advertising business from X is not, as far as can be discerned, connected to any campaign to induce the government to adopt any policies or regulations whatsoever.
But another line of cases, represented by the 1982 Supreme Court decision in NAACP v. Claiborne Hardware, does seem much more directly applicable to the lawsuit filed by X. The Claiborne Hardware case “arose after black citizens boycotted white merchants in Claiborne County, Miss [and some] white merchants sued under state law to recover losses from the boycott.” The Court concluded that the “right of the States to regulate economic activity could not justify a . . . prohibition against a nonviolent, politically motivated boycott designed to force governmental and economic change and to effectuate rights guaranteed by the Constitution itself.” Thus, even though the boycotting customers in Claiborne Hardware were engaged in concerted economic activity that could affect the competitive conditions in the Claiborne County business community—as opposed to simply lobbying, as in Noerr, for governmental policies that might change those conditions—the Court insulated the boycotters from liability under state laws seeking to protect fair economic competition and held that “the nonviolent elements of [the boycotters’] activities [were] entitled to the protection of the First Amendment.”
Not all boycotts, however, are afforded First Amendment protection. In 1990 in FTC v. Superior Court Trial Lawyers Ass’n, the Court permitted the Federal Trade Commission to end a boycott undertaken by a “group of lawyers in private practice who regularly acted as court-appointed counsel for indigent defendants in District of Columbia criminal cases [and who] agreed at a meeting of the Superior Court Trial Lawyers Association to stop providing such representation until the District increased group members’ compensation.” Although the lawyers argued that, like the boycotters in Claiborne Hardware, they too were interested in advancing equality and social justice rights—those of indigent clients—the Court declined to confer protection from liability, and in so doing distinguished the Claiborne Hardware case:
It is true that the Claiborne Hardware case also involved a boycott. That boycott, however, differs in a decisive respect. Those who joined the Claiborne Hardware boycott sought no special advantage for themselves. They were black citizens in Port Gibson, Mississippi, who had been the victims of political, social, and economic discrimination for many years. They sought only the equal respect and equal treatment to which they were constitutionally entitled. They struggled “to change a social order that had consistently treated them as second class citizens.” As we observed, the campaign was not intended “to destroy legitimate competition.” Equality and freedom are preconditions of the free market, and not commodities to be haggled over within it.
The same cannot be said of attorney’s fees. . . . [O]ur reasoning in Claiborne Hardware is not applicable to a boycott conducted by business competitors who “stand to profit financially from a lessening of competition in the boycotted market.” No matter how altruistic the motives of [the lawyers] may have been, it is undisputed that their immediate objective was to increase the price that they would be paid for their services. . . . [T]he undenied objective of their boycott was an economic advantage for those who agreed to participate. . . . Such an economic boycott is well within the category that was expressly distinguished in the Claiborne Hardware opinion itself.
For purposes of X’s lawsuit, the $64,000 doctrinal question is whether the boycott of X (assuming it can even be seen as a coordinated boycott) falls on the Claiborne Hardware or the Superior Court Trial Lawyers Ass’n side of the line. To us, the answer is pretty clear that Claiborne Hardware is the controlling precedent, for several related and mutually reinforcing reasons.
First, unlike the trial lawyers in DC, the defendants in the X dispute are not even alleged to be trying to fix prices in a way that would work to the X boycotters’ financial advantage and that triggers application of a per se rule of antitrust violation. As noted earlier, even the complaint by X concedes that defendants who eschewed advertising on X would be paying more, not less, to advertise on other platforms that comply with GARM norms insofar as (because X was not a viable alternative) these other platforms would have more leverage to charge the defendants higher prices.
To the extent that the complaint alleges that defendants sought from X not lower prices but better service, that is, service that better suited defendants’ values, that is true in every boycott case; in the real world, every decision by a set of individuals or firms is likely to have potential long-term economic as well as political consequences. That may be one reason that not all group boycotts are considered per se illegal under antitrust law. In Claiborne Hardware, for example, the Court described the boycotters as seeking to accomplish “government and economic change” (emphasis added), and yet the fact that the boycotters might have received some economic benefits from the elimination of racial discrimination didn’t deprive their boycott of constitutional protection. So too, the fact that the defendants in the X case might believe they are better off in business terms if they don’t associate with X doesn’t in any way diminish the fact that their primary motive is to vindicate their expressive freedom not to associate, just as the boycotters in Claiborne Hardware were seeking to vindicate their rights to equality. As the Court pointed out in block quote above discussing Claiborne Hardware, both freedom and equality are preconditions of a properly functioning market, not commodities to be haggled over.
Second, like the boycott in Claiborne Hardware (but unlike the one by the DC trial lawyers), the alleged advertising boycott in the X case is not being pursued to hinder any “legitimate competition” or obtain “special advantage for [the boycotters].” The defendants in the X lawsuit—who come from a variety of industries and who don’t compete with each other in the same consumer markets—aren’t trying to extract from X any concessions that wouldn’t also be available to any other would-be advertiser. That is, they are not seeking to extract any benefits that are particular to them. As noted above, they are not seeking lower advertising prices for themselves but instead seem willing to pay higher prices in order to vindicate their expressive values. And if X, in response to the alleged boycott, changes its policies and moves in the direction of complying with GARM best practices, such changes would redound not to the specific benefit of the defendants, but instead to all would-be advertisers who might support GARM principles.
This brings us to a few ways in which the case for First Amendment protection of the X defendants is in some respects even stronger than that of the Claiborne County boycotters. One is that the X defendants are focusing their efforts on the very entity whose expression they find unattractive, rather than targeting a third party to indirectly exert pressure on others to change. In Claiborne Hardware, the boycott was intended not simply to get the targeted businesses to alter their practices, but to put pressure on government officials (who presumably would be worried about bad publicity and a possible reduction in tax revenues) to change public policies. In some respects, then, as the Court itself noted, the Claiborne Hardware boycott was a “secondary” boycott, whereas the X defendants are through their actions registering their disagreement with X directly. That matters, because even though in both settings the boycotters are trying to influence debate on matters of public interest, the directness of the X customers’ boycott makes their expressive motives particularly clear, and it also means that there is the less potential for collateral damage to marketplace operations; only those entities who are alleged by the X defendants to be engaged in undesirable business practices—X and other platforms that eschew GARM principles—are subject to any economic consequences. In Claiborne Hardware, by contrast, not all the boycotted white merchants were equally guilty of whatever wrongs the boycotters were protesting, and some indeed were probably innocent pawns in the larger picture.
All of that raises a final, and quite powerful, reason the complaint by X is particularly vulnerable under the First Amendment: the defendants in the X case are themselves in engaged in protected First Amendment activity when they choose whether and where to advertise, much more so than were the boycotters in Claiborne Hardware when they chose where to shop. In Claiborne Hardware, the First Amendment was invoked by the Court to protect the equality rights of the boycotters; in the present case, the First Amendment is relevant and needed to protect the First Amendment rights of the alleged boycotters.
The Supreme Court has made clear (whether or not it has been completely correct in doing so) that commercial speech, including advertising undertaken by large corporations, is rigorously protected by the First Amendment. Although advertising may not occupy quite the same doctrinal plane as political speech (insofar as advertising can be made subject to disclosure mandates to safeguard against consumer misunderstandings or fraud and to other regulations not relevant here), government cannot regulate commercial speech in any way it wants. So even if the X defendants are engaged in commercial speech (and it warrants mention that their disagreement with X’s values might easily implicate political speech norms, including the so-called marketplace of political ideas, and implicate the most robust First Amendment protections, as advertisements concerning abortion did in the Bigelow v. Virginia case ), their decisions about how, where, and what to advertise are protected by the First Amendment.
And just as (in the context of core political speech) an abortion protest in one venue (outside a clinic) is different from a protest in another venue (in a park), so too the “where” of advertising is intimately connected with the message the listener hears and the instrumental effect the speech has. So requiring defendants to advertise on X (as the lawsuit seeks to do) more directly impairs the defendants’ rights of expression than would forcing the Claiborne County boycotters to shop in white-owned Claiborne stores. Perhaps the Claiborne Hardware boycotters would have felt they were “sending the wrong message” by shopping at white-owned stores, but buying a hammer at a hardware store is not inherently expressive the way advertising on a particular platform is: any issue of compelled expression in Claiborne Hardware would have been much more attenuated than the obvious compelled expression that would result from requiring unwilling advertisers to run ads on X.
This final point makes clear that Claiborne Hardware isn’t the only First Amendment precedent that bears heavily on X’s claims. Just over a year ago, the Court in 303 Creative v. Elenis said that a web designer whose websites express and communicate her ideas could not, consistent with the First Amendment, be forced to do business with clients whose own messages (in that case the celebration of same-sex marriages) she did not want to be associated with. If a seller of inherently expressive services (like a web designer) can’t be compelled to provide speech, then, we think the Court would conclude, neither can a consumer (advertiser) whose consumer activity is also inherently expressive. (And thus far the Court has not embraced any distinction between individuals and large corporations in this regard—recent cases like Citizens United and Hobby Lobby—and an older one involving Pacific Gas & Electric—reflect the modern Court’s full willingness to protect free speech rights of large and even for-profit business entities.) In any of these settings, the would-be speaker (the designer or the advertiser) cannot be required to engage in expression in tandem with others whose messages the speaker finds objectionable. And just as the desire to protect the consuming public did not permit Colorado to compel speech in 303 Creative, neither can the desire to protect marketplace competition in the present case justify compulsion of speech. Nor would, we think, it have mattered in 303 Creative if the web designer there were part of a group of web designers all of whom were committed to avoiding association of their own messages with those of same-sex-wedding celebrants. The government simply can’t compel unwilling expression under these circumstances.