Chillin'Competition

Earlier this month the Bundeskartellamt announced via this press release that it had “imposed on Facebook far-reaching restrictions in the processing of data”. According to its President, Andreas Mundt, this “can be seen as an internal divestiture of Facebook’s data”. The remedies are not the only thing that is far reaching in this case.

I had to read the available materials last weekend in preparation for a training session with a competition authority, but we never actually discussed the case, so here are the preliminary thoughts for discussion that I wrote down then.

In a nutshell, the Bundeskartellamt observed that Facebook was gathering user data also outside of the Facebook website and assigning them to Facebook user accounts. Its intervention seeks to ensure that going forward this will only be possible if users provide voluntary consent. Sounds good, right? Well, let’s see…

Are Facebook users better off? Probably, yes. The decision probably addressed an asymmetry of information issue. The Bundeskartelamt confers great importance to the fact that many users were not aware that Facebook could collect information also from third party websites featuring “Like” or “Share” buttons, even if the user did not scroll over or click on them. Did most Facebook users – myself included, to the extent I may still qualify as one – know this? Probably the very large majority of us didn’t. Although I guess we could have. Now…

Is that a competition law problem? Not really. The fact that an intervention makes consumers better off doesn’t necessarily mean that it’s good, or even legal. I would also be better off if the Belgian competition authority forced restaurants to serve lunch after 2 pm and with a smile, but I understand that’s not their job (I mean the authority’s…). Competition law is not something we can use to fine-tune market according to our preferences.

Market definition. The Decision is premised on the idea that Facebook is dominant in the market for social networks as well as in a market for advertising in social networks. The Bundeskartellamt observes that “services like Snapchat, Youtube or Twitter, but also professional networks like LinkedIn and Xing only offer parts of the services of a social network and are thus not to be included in the relevant market”. The Q&A document explains that other players (e.g Youtube) use business models that are not sufficiently similar to Facebook’s to warrant inclusion in the same market.

Some commentators have welcomed this “flexible approach” based on “functional similarities and differences”. I don’t. First, a formal approach merely based on functionality tells you nothing about competitive constraints. Different functionalities are often actually a way of competing, not a reason to rule out competitive constraints. The functionality approach goes back to United Brands, and I haven’t ever heard anyone saying that was a good approach. Second, it should be evident that different business models can coexist and compete within the same relevant market (think e.g. of vertically integrated companies vs non-integrated rivals or of franchises vs independents. In sum, market definition is not about formal differences in functionality, but about empirical substitutability.

“Users practically cannot switch”(?) Exploitative practices are mostly problematic when users are locked-in to a service and there are barriers to entry and switching. Is this the case here?

The Q&A document states that “because of Facebook’s market power users have no option to avoid the combination of their data”, that there is “lock-in” and that “Facebook is becoming more and more indispensable for advertising customers”. The press release explains that “Facebook users practically cannot switch to other social networks”. The adverb “practically” is not without importance, and I look forward to reading the decision. The staggered 360-degree-turn evolution of precedents on “switching” is fascinating. How can this finding coexist with the GC’s Judgment Microsoft/Skype or with the Commission’s decision in Facebook/Whatsapp? Authorities do not seem to perceive barriers to switching views on this point…

At first sight, it’s not like Facebook is an indispensable service; users can switch, and switch off. The question is: would users leave Facebook if it were to act in a non-competitive way, or in a way that users did not approve of? Press reports published following the Cambridge Analytical case suggest that users can, did and would abandon Facebook in those circumstances.

Data protection provisions as a benchmark for finding an abuse. The Bundeskartellamt has explained that Facebook’s terms of service and its collection and use of data “are in violation of the European data protection rules” and that they have “closely cooperated with leading data protection authorities in clarifying the data protection issues involved”. The obvious question is: which of these leading data protection authorities has declared a violation of data protection rules? To my knowledge, none. And had there been a violation, shouldn’t they have declared it? I don’t know whether a violation existed, but it’s not for me to say, because –like the Bundeskartellamt- I’m at most a competition expert, not a data protection one. There are very competent data protection authorities specifically prepared to deal with these issues.  Even data protection activists have stated that it doesn’t make sense to shift data protection responsibility away from specialized authorities. How can we have specialized authorities but have the call made by a non-specialized one? How can we simply assume a violation that has not been established?

Harm to news publishers from a certain Member State as a standard for amending laws, creating exemptions, and prosecuting creative competition law infringements.

Product refinement and targeted advertising as a problem. The Q&A document explains that “from Facebook’s perspective, the data are of great economic value (…) Facebook can use them to optimise its own service and tie more users to its network (…) In addition, with the help of the user profiles, Facebook can improve its targeted advertising activities”. So practices are problematic because they enable companies to improve their products and offer ads that are more relevant to users. Such allegations are commonplace these days, but sound to me like an efficiency offense. Just as in the case of network effects, one cannot simply assume that something that increases the value of a product/service has negative competitive implications. But, unfortunately, there seems to be little appetite to deal with complexity and ambiguity these days, particularly when it comes to certain “online platforms”. It’s much easier to simply assume they’re just bad.

The convergence problem (and how to avoid it). The Bundeskartellamt states that “such an abuse control proceeding against Facebook would generally also be possible under the relevant norm of Article 102 TFEU. So far, however, only the case law of the highest German court has been established which can take into account constitutional or other legal principles (in this case data protection) in assessing abusive practices of a dominant company” (sic). This is interesting. And the use of the adverb “generally” may be as telling as the use of “practically” above.

If the Bundeskartellamt is confident that its approach was possible under EU Law, why didn’t it apply Article 102 in parallel to the national law? The only legally valid reason would be that trade between Member States was not affected, but that doesn’t seem to fly in this case. That issue, however, was avoided by defining a national market for social networks, based on the observation that German users use social media to network in seemingly isolated silos with other contacts within Germany.

What’s essential to a business model? There is nonetheless an element of consolation in all this. The decision does not target the processing of data generated by Facebook’s own website because “[t]his is an essential component of a social network and its data-based business model”. The idea that one should not lightly challenge the elements underpinning a business model is a sensible one on which we have often insisted. Even if this doesn’t make the rest of the decision right, it is a good way of drawing red lines.