Elon Musk's Lawsuit against the Companies That Have Boycotted Twitter by Pulling Out Their Advertising.
X, the former Twitter, has sued an advertising trade assocation, GARM, for antitrust (anti-monopoly) violations along with several corporations that participated in the boycott. (I dislike the name “X”, so I’ll use “Twitter” for this essay.) Shortly after Musk bought Twitter, GARM advised its members to boycott it by ceasing to advertise there. Musk says that this is “a conspiracy in restraint of trade”, as with companies that coordinate with each other to have high prices. The restraint of trade here is even more obvious. Rather than restraining trade by manipulating prices and diminishing sales volume, GARM is directly restraining it by organizing a boycott.
GARM dissolved itself a few days after the lawsuit was filed, but presumably the lawsuit will continue against the other defendants. The defense would be that the boycott was organized for good reasons, a valid defense, if true.
The typical anti-trust case has sellers organizing to hurt the buyers by raising prices. This is a bit different. Here we have buyers of ads who organized to hurt the seller by boycotting it. If the buyers did this to pressure Twitter to reduce its ad rates, they would clearly be liable and would have to pay treble damages to Twitter (three times the actual damage). They did not do this, though. Their boycott had a different motivation. GARM says that since Twitter has so many disreputable posts now that Musk has eased censorship, GARM recommends that advertisers avoid Twitter so their ads don’t appear on a platform so open to such ideas, and the members simply followed the advice. Since the members of GARM include the big advertising companies and big advertisers, the boycott has drastically reduced Twitter’s ad revenue.
Being a group boycott is what makes this illegal. If Unilever decides by itself for whatever reason to boycott X, that’s legal. If Unilever boycotts as part of a group, that’s illegal. It works out the same way as for seller monopoly. If one company raises prices, that’s legal. If two companies get together and agree to raise prices jointly, that’s illegal.
The same idea of group versus individual boycott comes up in the early-20th-century Japanese law of ostracism. I mention this because Mark Ramseyer and I have an article forthcoming on this.1 If one member of a Japanese village was ostracized by the rest, he ccould sue them and win. Any one villager was free to ostracize someone, but he couldn’t do it as part of a group effort. Our paper focuses on the question of why anyone would go to court in such a situation when it really was impossible for the court to require the other villagers to interact with the villager who sued. Our explanation is that what the victim of ostracism wanted was public vindication by the court, to have a neutral authority declare that after hearing the evidence, the act of ostracism was unjustified. In the case of X something similar is going on. What Elon Musk wants is not just money damages, which will not be very high even if he wins, but public vindication of his claim that the Establishment is out to ruin him.2
GARM says that all it is doing is giving advice to advertisers. Twitter says that the advertisers banded together when GARM was formed in 2019 to impose the censorship standards they wanted onto all the Internet platforms— not just Twitter, but Facebook, LinkedIn, and so forth. The counterargument is that GARM’s members want restrictions on conservative posts, for some to be banned, and others slowed down. The advertisers didn’t think they could do this independently, so they created GARM to increase their market power.
It’s odd that buyers would have to organize to get the seller to produce the kind of product they want. After all, the seller can charge a higher price if his product is more attractive. If computer companies all want a certain design in the next generation of memory chips, they wouldn’t have to band together to force the chip makers to provide that design. The chip makers would do it voluntarily, in order to sell more chips. Why, then, did the buyers organize to get the particular kind of censorship GARM offered?
I will tentatively suggest that a certain group of advertisers wanted that level of censorship, and they formed GARM to impose it on other advertisers who did not. These other advertisers then joined GARM, even though they didn’t like that level, because they at least wanted to be involved in the process of setting group policy, rather than staying outside without any influence at all. The companies that did the initial organizing wanted to induce every platform to do the kind of censorship they wanted, rather than to allow individual negotiations between advertisers and platforms that would result in varying levels of censorship. With individual negotiation, most companies would settle for their own ads not being placed next to undesirable posts that might turn off consumers, rather than hold out for the entire platform being censored, even the parts where they don’t advertise.
Let me make up some numbers for an example. Suppose that without GARM, Google will set the censorship level for various companies at 2 and 8. Level-2 advertisters don’t want much censorship; they just want to have their ads placed next to the most likely consumers. The advertisers that prefer level 8, on the other hand, have political objectives and are willing to sacrifice profits from selling to people who like conservative posts. They don’t want to let the level-2 companies place ads at lower cost and higher value— especially if they are a level-8 advertising agency that is competing with other ad agencies for the marketing accounts of big corporations. Thus a level-8 company will want to somehow stop the level-2 companies from getting access to low-censorhip internet platforms. The level-8 companies set up GARM, creating market power such that the platforms would fall into line with level 8. The level-2 companies would either stay out of GARM, or join in the hope that by at least being in the organization they could move the level to 7 instead of 8. The end effect is to hurt the platforms and hurt the level-2 advertisers. The level-2 advertisers could try to do the same by setting up their own version of GARM, but that’s hard to do once such a trade association has already been set up.3
There is another consideration, an even simpler one. If GARM’s members were internet platforms— Facebook, You-Tube, and so forth— then it would be an obvious seller anti-trust violation to organize an advertiser boycott of Twitter, their competitor. Facebook and You-Tube are indeed GARM members. Members of GARM include Meta (which owns Facebook and Instagram), You-Tube (owned by Google), TikTok, and LinkedIn, along with Microsoft, the owner of LinkedIn. The platforms don’t hold the power in GARM, but they are members, and it is in their interest to see Twitter suffer.
It will be interesting to see what happens to this case. As I said earlier, I think the motivation is to publicize the coordinate attack on Twitter, and find out more details— secret memos, interviews with GARM executives, and such— in the course of the lawsuit. Simply publicizing what was going on might break the boycott by embarassing the companies doing it in front of their shareholders. Or, it may be that the court will grant Twitter the injunction it wants, a court order to stop boycotting Twitter, though I can’t see how a court could enforce such an injunction.
J. Mark Ramseyer & Eric Rasmusen, "Ostracism in Japan," forthcoming, Asian Journal of Comparative Law. There’s a good story about why we two ended up publishing this paper on ostracism in a Singapore-based journal, but I will tell that story another day. Suffice it to say that many people dislike other things we have written. Or, of course, it might just not be a very good paper.
IThe main defendant, GARM, has few assets of its own, and it is dissolving rather than fight the lawsuit. That won’t stop the boycott, which now that it has been organized does not need a leader. I suppose Twitter will continue its suit, but just against the individual corporate defendants such as Unilever.
In the field of “game theory”, we call such a situation a “coordination game”. If someone starts a level-8 trade association and companies join it, the fact that it has so many members means nobody would want to join a little new level-2 trade association. They’d only move if most of the other members of the initial trade association moved. If, however, the level-2 companies had organized a trade association first, to impose level-2 on platforms, every other firm, even the level-8 ones, would join it because so many other companies had joined it, a self-fulfilling prophecy. It’s probably is no accident that the level-8 companies organized first, though. The level-2 companies don’t care if the level-8 companies negotiate special agreements that raise their costs. The only reason to organize is to defend against a level-8 trade association.
Seems like Elon's most obvious countermove is to publish a list of GARM's constituency. Then it's up to us to decide whether to continue patronizing them.
Also, if GARM's organizer(s) are known, they would be vicariously liable for GARM's torts. So while GARM may no longer exist, it's creator(s) do(es).