Some cautionary notes on platform regulation
Diego Zuluaga // 03.11.2015
In a post last week, I welcomed Competition and Markets Authority chief Alex Chisholm’s plea for regulatory prudence on online platforms. As Chisholm indicated, the unknowns that competition authorities face when analysing any sector are magnified in the online sphere, where markets are highly dynamic, market shares are large but vulnerable and ephemeral, and where incumbents commonly face competitive pressures from both within and outside their relevant industry. In such an environment, regulatory error is both more likely and more costly, and regulatory action rarely takes place before the problem it sought to address is resolved by market forces. The CMA boss is therefore right in cautioning against ex ante regulation in this sphere.
However, the pressures to come up with a statutory framework to regulate online platforms are significant in Europe at the moment – both in the Union as a whole and in specific Member States like France. It is therefore essential to explain specifically why it is a bad idea for governments to legislate on platforms as a separate industry.
The first issue is definitional. Namely, we need to establish what online platforms have in common and what makes them different from other players on the Internet and from their offline competitors. Otherwise, there could be no rationale for regulating platforms just because they are platforms. We know a platform when we see one – Google, Facebook, Airbnb, Uber are all examples of online platforms. But what do a search engine, a social network, a room-sharing and a ride-sharing app have in common – and what separates them from the Yellow Pages, WhatsApp, a hotel chain and London black cabs, respectively? A feature that many economists argue is common to platforms is the fact that they operate in two-sided markets, i.e. Google and Facebook offer information to users – albeit at zero monetary cost – and they sell online space to advertisers. Similarly, Airbnb and Uber cater both to those with a room/ ride to offer, and to those looking for lodging/ transport.
The latter play a role more akin to traditional intermediaries, breaking barriers between suppliers and consumers. But even search engines and social networks serve to match willing sellers with online users, linking together (as my IEA colleague Philip Booth has pointed out) a public good – online information, which is by design non-excludable and non-rival – with a public ‘bad’ – advertising, although the latter need not always be a ‘bad’ to consumers. Indeed, given how effective algorithms are getting at offering users what they might wish to buy, online ads are presumably becoming less and less of a public ‘bad’: less obnoxious and more relevant to the given user. Another way to look at online platforms is as mechanisms designed to reduce informational asymmetries and other transaction costs between sellers and buyers – thus increasing opportunities for mutually advantageous trade.
But that is where the similarities end. Platforms operate in wildly different markets, rely on a variety of infrastructures – online browsers, smartphones, apps – and ecosystems – Android, iOS – and compete with a range of online and offline rivals. Given that competition policy is concerned with consumer welfare in the relevant market where an undertaking operates, and that regulation in this sphere is aimed at preventing anti-competitive practices, it would not appear reasonable to regulate very disparate businesses – and, importantly, not some of their competitors in each case – just because they share a very broadly defined business model – match-making – or because they have similar market effects, i.e. reduced transaction costs – which, in any event, are a common feature of process innovations. Platforms thus do not seem to share enough characteristics to warrant a common regulatory approach.
The second problem in proposals for platform regulation is that their economic rationale is shaky. For instance, the French National Digital Council (Conseil National Numérique) has proposed to apply the so-called ‘essential facilities doctrine’ to “dominant” online platforms. This doctrine, traditionally applied to utilities, mandates that certain monopolistic suppliers have a duty to deal with competitors when the latter require a good or service owned by the former in order to be able to compete with it. It already becomes apparent that the essential facilities doctrine requires enormous stretching to be applicable to online platforms. In a world where there are dozens of globally competitive search engines – horizontal and vertical – with various ways to reach users – direct browsing, search on other platforms, apps, and so on – can one really speak of a monopolistic supplier? What’s more, can one consider that the facilities owner controls access to a resource that is essential for competitors to challenge it? Any casual analysis of the platform environment – constantly evolving and giving rise to new players – suggests otherwise.
More sophisticated critiques of online platforms in favour of regulation are nonetheless dubious. For instance, some proponents of regulation argue that data – of the sort obtained from users as they make use of a platform – can be a barrier to entry for new players. But the theory does not stand up to scrutiny, since data is non-rival – one firm’s use of it does not preclude another from doing the same – and because data is useless unless the relevant undertaking can make profitable use of it – for instance, through a sophisticated algorithm that makes accurate predictions of what a user is looking for on the basis of her search. Neither do the empirics support the assertion, as we have experienced an ongoing stream of new players in all sorts of platform markets in recent years – and no slowdown is in sight. As so often in competition discussions – particularly of the innovative industries – proponents of regulatory action put the cart before the horse, by coming up with theoretical constructs in which a rationale for intervention might be justifiable, and failing to consider drawbacks to the theory – let alone whether it comes close to reality.
The final weakness that calls for the regulation of online platforms suffer from is a reliance on outdated models that are being gradually phased out by the regulators themselves. As Maxwell and Pénard (2015) point out in a recent study for Hogan Lovells, ex ante regulation of the sort we have known in Europe in the last few decades was designed in the context of monopolistic or quasi-monopolistic players in the telecommunications and utility sectors, which owed their dominance to government privilege. Indeed, they tended to be state-owned or formerly state-owned, and competitors were legally barred from challenging them. It was in an attempt to introduce greater competition in a deliberately uncompetitive market that European regulators developed their models.
Yet, as state-sponsored dominance fades with time, those models are becoming less and less useful – and they are even less relevant for sectors which have never experienced anti-competitive state intervention. This, by the way, is recognised by the European Commission, which in 2009 stated that “the aim is progressively to reduce ex ante sector-specific rules as competition in the markets develops and, ultimately, [for them to be replaced] by competition law.” The fact about ex ante regulation that its 21st-century advocates seem to forget is that it was developed in very specific circumstances, for very specific sectors. And disruptive, market-driven online innovations are not one of them.
The need to avoid ex ante rules for online platforms does not just stem from a high likelihood of costly regulatory errors – although this is very important, and the CMA’s Chisholm was right to point it out last week. Regulation is fundamentally inappropriate because key questions of definition, economic theory and empirics, and competition policy precedents have not yet been answered in the affirmative – not by a mile.
Diego Zuluaga is Deputy Director of EPICENTER, and a Research Fellow at the Institute of Economic Affairs.
EPICENTER publications and contributions from our member think tanks are designed to promote the discussion of economic issues and the role of markets in solving economic and social problems. As with all EPICENTER publications, the views expressed here are those of the author and not EPICENTER or its member think tanks (which have no corporate view).