Regulating platforms? A competition law perspective

A few days ago I spoke at CEPS about the debate concerning online platform regulation that is attracting some interest these days as DG Connect’s and the House of Lords’ consultations are ongoing (my presentation and a video interview are available here). This is a most interesting issue although, admittedly, one that not so long ago I would not have expected to be an issue at all.

In the wake of this event we thought that perhaps it would be useful to contribute a bit to the debate, so here you will find a summary of what I said at that conference. In order to complement it, we have decided to engage in a dialogue with our friends at CCIA and their DisCo (Disruptive Competition) Project blog (here or here). This post will also be published here, and we will soon be posting a guest contribution from them. Any comments you might have will certainly enrich the debate.

Not being an expert in regulation, my views on the subject are eminently related to competition law and to its application to multi-sided markets which, as you know, is one of the fields in which I have recently done some work, advising platforms, non-platforms competing with platforms or simply reflecting on wider policy issues (e.g. here or here). The competition law perspective is a particularly useful one because competition law seems to be the elephant in the room, much at the root of these discussions.

Indeed, many of you will recall that ex ante “platform” regulation went from being a non-issue to being very much an issue when a number of Member States (notably Germany and France) expressed frustration at how competition law would not be enough to tackle some problems (not clear which) caused by “some” (apparently not all) platforms (some examples of such statements are available here, here and here). [This concern about the possible shortcomings of competition law coincidentally emerged at a time when some thought that the ongoing Google investigation would fail to establish a “neutrality” obligation incumbent upon Google’s search activities] And since competition law was seen as insufficient (read: did not lead to the outcome that some expected), some thought that it would be idea to either change competition law or to bypass it by adopting specific regulation.

What kind of animal is a platform?

In this context, the word “platform” seems to have been chosen to encompass those “some platforms” that people had in mind. However, as explained in previous posts written on the DisCo Project blog (here or here), it may not be the best term to identify a category of companies subject to specific regulation.

Some of you may have heard of the expression “The Law of the Horse”. This is a term coined by American judge and antitrust expert Frank Easterbrook in a now famous conference in the US. In this conference he explained that there is no “law of the internet” more than there is a “law of the horse”; that there are laws of contracts that apply when horses are sold, of animal husbandry that apply when they need care, of laws of gambling that regulate when they race, but there is no law of the horse. Nowadays some partisans of regulation are trying to create some sort of “Law of the horse platform”. But then of course we come back to the somehow relevant question of what a platform is…

I don’t know if you are familiar with the Indian story of “The Blind Men and the Elephant”. This is a story in which several blind men are asked to describe what an elephant looks like by touching different parts of its body. The one touching the leg says the elephant is like a pillar; the one touching the belly says it must be like a wall; the one feeling the trunk believes it must look like a tree branch; the one touching the ear is convinced an elephant resembles a hand fan, and so on.. This story illustrates the fallacy that one’s subjective experience can be true but at the same time it is inherently limited and cannot account for the totality of the truth.

We see something similar regarding “platforms”: some appear to extrapolate certain features or problems from a limited number of companies to a whole business model, but those problems are neither exclusive nor common to “platforms” (as defined in the Commission’s consultation). Before doing such a thing, one should perhaps understand how markets work and why companies do what they do.

Having this complete view would certainly be necessary, for, as stated by Easterbrook in his talk The Law of the Horse, “the blind do not make good trailblazers”.

The wrong question

As already noted, the question many are asking is whether competition law is sufficient to address the challenges raised by platforms or whether we need a new framework; does competition law need to adapt?

In my view, and whereas enforcement may need some refinements (e.g. merger notification thresholds may not be well suited for some mergers—see here—and we do not have good economic tools to assess demand-side efficiencies–see here), there is no other branch of law that, over more than a hundred years, has proved similarly flexible, adaptable and accommodating of the evolution of markets and economic thinking than competition law.

In my view, the questions that are being posed now are the wrong ones, so I would suggest that instead of looking at supposed flaws in competition law, perhaps we should look to competition law to extract some lessons.

The above includes understanding why competition authorities are sometimes reluctant to intervene, or why issues that are perceived to be problematic by the lay public are not understood as such by experts in the field. Also, it would be worth reflecting on whether there may be a possibility that if competition law has not done more regarding “platforms” it might be due to the fact that there may indeed be very good reasons for it not to do more.

Why competition law can teach us?

Some may wonder whether competition law can really teach us something about platforms and about how to deal with them. If you ask me, it sure can.

You see, the cornerstone provisions in competition law are merely a couple of articles in the Treaty (or sections in the Sherman Act) and have over the years evolved as interpreted and applied by Courts and specialized agencies to changing markets and in light of mainstream economics free from changing political priorities. Some aura of apparent complexity has also enabled it to evolve soundly, free from the inference of small politics (by which I mean that that its underlying principles are not affected by movable or fashionable political considerations, which is a good thing; by the way, I developed those ideas here). All this has turned competition law into an ever-evolving distillation of common-sense principles infused with mainstream economics.

In addition to that, competition law has a privileged view of the main issues often linked to “platforms”. Indeed, many have resorted to our discipline to assess issues such as, for instance, portability and self-favouring (Google case), price restriction in the form of MFNs (several EU and national cases), excessive pricing (see the MIFs cases regarding Visa and Mastercard), interoperability denials (Microsoft) and even privacy considerations in many recent debates and to some extent in Facebook/Whatsapp. I can’t think of any other discipline having a better view or understanding of what goes on in those settings.

Finally, competition law does not assess all these matters superficially or in the abstract, it analyses them technically and objectively, in specific cases and against the background of factual and economic evidence. Compare that, for instance, with this example of a thoroughly thought out proposal for public intervention… Against this background, ignoring the lessons that competition law experience can offer to the debate would probably not be wise.

What competition law can teach us

• On whether and how to intervene. Over the years, experience has taught competition law to be humble, that with great power comes great responsibility, that authorities should also intervene when a clear problem is identified, and that when they do so they should only act with proportional remedies. We have come to accept that markets and competition cannot be perfect, and that is why we aim at “workable” competition; if we strive for perfection, we may be messing with complex realities with unforeseen consequences (e.g. what is the cost of imposing non-discrimination obligations across the board? Does it make sense? Do we have the knowledge to be sure this is an appropriate remedy?)

• On the identification of “platforms”. According to the Commission’s consultation a platform is essentially an intermediary operating in a two-sided market bringing together different but interdependent groups of users. In competition law there have been plenty of discussions in recent years about how to adapt some of our tools to the peculiarities of two-sided markets. In any given case, and given that many markets present multi-sided features, the first question that we need to address is whether two-sidedness is enough to matter. And we have come to find that this question is not one that can be answered ex ante in the abstract, but on a case-by-case basis and in the light of empirical evidence. Our experience shows that picking multi-sidedness as the decisive element for specific regulation to apply may be tremendously problematic.

• On agnosticism towards business models and public distortions. Over time we have also learnt to move away from the analysis of business models to the assessment of competitive constraints. When we assess a conduct or merger we do not look in isolation at how the business is organized, but at its competitive impact in a relevant market where different business models may compete. In many markets “online platforms” compete with offline platforms or with non-platforms. Regulating only some of the players active in a market may skew and distort competitive conditions in an undesirable way (one only needs to look at the “regulatory asymmetry” concerns raised in regard to the sharing economy which is, by the way, subject to the same consultation that also extends to “online platforms”). Competition lawyers are very aware of the fact that most, and the most serious, restrictions of competition often have a public origin.

• Ceci n’est pas market power. Competition authorities, Courts and lawyers have also learnt to abandon some of our traditional reflexes when it comes to platforms. We now know that appearances of market power in online settings often cannot be trusted, because, among others, the markets are extremely dynamic and there are strong actual or potential competitors, services may be provided for free, there may differentiation, multi-homing, easy switching, interoperability or low barriers to entry. This is not just theory, but also a competitive reality that has been assessed and confirmed in several cases regarding “online platforms”, such as Google/DoubleClick, Microsoft/Skype (you know my take on that one) or Facebook/Whatsapp.

• Size does not necessarily matter (this is something that competition lawyers know all too well…) or that it may not necessarily be a bad thing. Multi-sidedness is about network effects and network effects are a positive externality; the main defining characteristic of these settings is that scale generates benefits. This also means that:

• Business practices carried out in multi-sided markets will often be competitively ambiguous, because the same features that yield market power might help achieve optimal scale/demand side efficiencies (hence the title of my piece on the subject: the double duality of two-sided markets). Certainly some practices will be anticompetitive also in these settings, but this is something that will have to be looked at carefully and on a case-by-case basis, but is not something that can be decided in the abstract. This was actually the main message of the ECJ in Cartes Bancaires, and once again seems to reveal that ex ante regulation may not be ideal.

• Competition law is always there, ready to kick in in those specific cases where there things go wrong and there may be problems felt in markets. As noted earlier, we have seen this with portability restrictions, interoperability denials, pricing restrictions, excessive pricing, etc.

• But competition law isn’t the answer to every problem. Despite the above, over the years we have also come to realize that competition law is not the answer to every problem. Expecting the contrary indeed might lead to frustration or to a feeling of insufficiency. This is a message in which I have insisted repeatedly in the context of the debate on the role of competition policy in addressing data protection/privacy concerns (see here among others).

• On consumer choice as the bottom line. Competition law is about consumer choice, much like, at least in my view, public policy should be about enabling informed choices. If you ask me, the best way of protecting consumers from the perceived problems in these markets would mainly require informing/educating them, and then making sure that they are not artificially locked-in to a given platform (which is what competition law is here to do in these and other markets). It is often said that “platforms” are powerful because of the information they control; but if information is power, we should perhaps also give it to consumers…