Chillin'Competition

The Commission announced this week the formal opening of a case against Amazon (see here). It had also informally done something quite similar almost a year ago (see here), and that first news cycle triggered comments from Pablo that remain current and are even more valuable today. In parallel, the competition authorities from Germany, Austria and Luxembourg closed proceedings against Amazon after the company agreed to modify certain clauses.

The Commission’s case is plagued with interesting legal issues and questions that we look forward to exploring. It is a prime (pun intended) example of the issues raised in contemporary antitrust. I’ve received a few press inquiries about this and already had a chance to discuss the development almost live during a lecture at the College of Europe on “Multi-sided platforms: the lessons from the case law”, so I’ll build on what I explained there (off the top of my head, so this is all likely to evolve).

Bear in mind that we have no information as to whether the Commission’s factual suspicions are well-founded or not, so for the purposes of this post let’s simply assume that they are and focus on the law:

Duality at the core. The case against Amazon is premised on the observation that “Amazon has a dual role as a platform: (i) it sells products on its website as a retailer; and (ii) it provides a marketplace where independent sellers can sell products directly to consumers”. The Commission suspects that Amazon collects competitively sensitive information about marketplace sellers and, according to the press release, it “will focus on whether and how the use of accumulated marketplace seller data by Amazon as a retailer affects competition”.

Challenging vertical integration/a business model. Is the Commission challenging a business model or vertical integration itself? To the extent that one argues that the problem lies in “being umpire and having a team at the same time” (see here) in itself, then this would effectively constitute a challenge to a given (hybrid) business model more than to specific practices [on a different note, the referee/player metaphor was the one traditionally used to challenge the Commission’s own business/enforcement model]. 

If this is the Commission’s thinking (it may well not be), this could have very profound implications, as many companies other than Amazon (including e.g. large offline retailers with private label brands) rely on the same business model. This is particularly true given that the investigation is based not only on Article 102 but also on 101, so its ramifications could extend beyond dominant companies.

As we always say, competition law is business model agnostic. Amazon, for one, has suffered from this when it comes to platform bans (think about it, Coty is also a prime example of the idea that one cannot treat a firm that opts for selective distribution worse than a vertically integrated rival pursuing the same objectives). The use of a given business model, in itself, does not warrant antitrust intervention. It might, if it gives rise to anticompetitive effects in the sense of the case law. As Pablo explained in his post, vertical integration and lack of neutrality is often even procompetitive. Again, that doesn’t mean that specific practices may perhaps be legal/illegal, but one cannot simply presume legality/ illegality just by looking at the business model.

On the threshold of effects. If it’s all about the effects, what effects are we talking about? What legal standard should the Commission apply?

Is the theory then that merchants suffer a competitive disadvantage? The case law tells us that a mere disparity of treatment/competitive disadvantage is not enough to find an infringement (see e.g. Lesson number 7 here ;), MEO, Post Danmark I or Deutsche Telekom (para. 250 where the Court said that the existence of a margin squeeze/forcing rivals to price below cost is in itself, absent anticompetitive effects, insufficient to find an infringement).

Is the theory about unfair trading conditions? If one frames this as an exploitation/unfair conditions/excessive pricing case, one would need to look at whether the price paid by merchants is excessive having regard to the value of the service provided by Amazon’s Marketplace. And that may be pretty hard. Moreover, cases like TeliaSonera also involved unfair conditions and also require a showing of anticompetitive effects. As explained in the previous paragraph, this means something beyond a mere competitive disadvantage.

Isn’t this rather about conditions for access? The idea seems to be not only that Amazon is dominant but that merchants actually depend on Amazon to market their products (if not, it’s clear that there is no foreclosure, right?). In a way, you could say that data allegedly collected by Amazon is part of the price that a merchant pays to be able to sell its products in the marketplace (don’t critics of online platforms actually often repeat that people pay with data?). Amazon may need that data for different purposes, including ensuring that the overall platform/marketplace remains competitive (this is a fundamental point on which I would expect much of the legal discussion to focus). And this is in the nature of the hybrid business model, which has so far worked in this and other sectors. Would we be better off if Amazon re-adopted its previous model and closed the marketplace to third parties? And in that scenario, could a competition authority force Amazon to reopen it again under FRAND terms? The intuitively easy answers to the questions may be telling.

If one thinks about the case law (a big “if” these days), cases like Bronner, Commercial Solvents and Télémarketing all involved similar settings (vertically integrated rivals “favoring” their own services) and all of them require indispensability and the elimination of all competition. Bronner very much emphasizes this point beyond doubt (“self-favour” yourself and read it again).

Are merchants foreclosed/driven out of the market by Amazon’s alleged conduct? At first sight this would not appear to be the case given the continuous growth in merchant sales on the Amazon marketplace and the existence of other channels to market products. In fact, Amazon needs merchants and it is highly unlikely that its marketplace would thrive if merchants didn’t. An automatic assumption of foreclosure would imply presuming that Amazon is somehow an indispensable sales channel, which sounds like quite a stretch (and which, by the way, also appears to be at odds with brands’ appetite for platform bans). It will be interesting to see how the Commission approaches this question.

Increasing competition? If it were true that Amazon really uses merchant data to set its competitive strategy, I guess one could even argue that Amazon would be merely observing where there is scope for greater competition (in terms of price, output or quality) in order to adopt certain decisions, including whether to launch its own product. Under this optic, one could therefore argue that this practice (again, assuming it might exist) would actually enhance competition in every product category.

On similarities with other cases. This investigation is but one more step in a trend/line of thinking that started and peaked with the Google Shopping case (the judicial outcome of which could have a crucial impact on the Amazon case depending on time and on how Amazon plays this). Pablo accurately calls this “common-carrier antitrust”. The irony is that some of the people that propelled and still defend those theories are now faced with their boomerang effect (talk about dual roles…). Expect some creative contortions

Openings and closings. As some commentators have observed, the parallel closing of the German and Austrian cases shows that competition cases can also be quickly and effectively resolved, regardless of their merit, even absent interim measures. But, among other factors, that depends on whether what is at stake is an essential component of the business model or not. This may perhaps be one more reason to focus on cheap exclusion and not second guess business models (which is what some people are now openly advocating for).

To be continued…