What sets platforms apart is their possibility to effectively cross-subsidise between the different user groups that are party to a transaction. Platforms often treat one side as a profit centre and the other as a loss leader, or, at best, as financially neutral. As a result, platforms must choose not only a price level, but also a price structure for their service.
Given this, the present article, available here, explores how potentially abusive behaviour involving free products (both goods and services) can be assessed under competition law.
Section II looks at different dimensions of offering free goods and services.
Free online offerings have become ubiquitous. This reflects lower costs brought about by the existing digital infrastructure (e.g. processing power, bandwidth, storage). However, companies still want to make a profit. In practice, offering services for free has the potential to attract the critical mass of customers that will allow a company to maximise its profits across its various products. There are three main business models that support the supply of free products: (i) direct cross-subsidisation; (ii) providing a different service for a price on one side of market, which monetises the users to whom free services are provided on the other side of market; (iii) freemium, whereby any customer can obtain a free service, while some to pay a premium for extra or enhanced services.
Section III considers how the provision of free services relates to predatory pricing.
A first question that arises as regards free products is whether they amount to predatory pricing. Under EU law, prices below average variable costs are assumed to be anticompetitive, while prices between total and average variable cost are abusive if they are part of a plan for eliminating a competitor. The reason to sanction such practices is because such prices can drive from the market undertakings as efficient as the dominant undertaking but which, because of their smaller financial resources, are incapable of withstanding the competition waged against them. Unlike in the US, no prospect of recoupment has to be established in the EU. On the other hand, recoupment may still be relevant to exclude economic justifications other than the elimination of a competitor, or to assist in establishing that a plan to eliminate a competitor exists.
A price of zero is necessarily below cost. Therefore, the provision of any free product could be prima facie abusive under the predatory pricing doctrine. However, digital products are not actually free: somewhere, somehow, someone is paying. The business decision of who is going to pay what depends on a complex combination of the price elasticity of users on each side of a platform, the relative strength of indirect network effects and the users’ ability to multi-home (i.e. connect with several platforms). Offering consumers a free service can thus make perfect sense from a business point of view, as someone (else) is footing the bill. As such, standard cost-based tests for detecting predatory pricing generally make no economic sense when applied to multi-sided businesses. In two-sided markets, both market sides should be taken into account when examining whether prices are predatory. This is not to say, however, that online platforms are inherently unlikely to engage in predatory pricing. In fact, the economics guiding platforms may just give them a reason to do so. The value of the platform is determined by the quantity (and quality) of the users on either side. Platforms are thus engaged in a race to attain a workable number of users (‘critical mass’), especially as—in the end—no more than a few platforms (and often even one) dominate the market. In doing so, they often pursue a growth-over-profits strategy, opting to reinvest rather than to turn a profit.
Section IV focuses on excessive pricing.
Consideration for free services often consists of the consumers’ information and attention. One may wonder, then, whether a business could demand too much information or attention from its users, so that its conduct amounts to excessive pricing. Here, it is useful to distinguish between information and attention costs. The information cost for users consists in handing over their personal data, either consciously or involuntarily. Attention costs include time spent on undesired activities such as looking at advertising.
Despite this, competition law focuses mainly on price. Until effective methods to account for the ‘cost’ of information and attention are found, competition authorities will rely on non-price-based theories, as happened in the Bundeskartellamt’s case against Facebook. The abuse at issue in this case concerns the way in which Facebook processes user data – in particular how Facebook makes use of its social network conditional on it being allowed to limitlessly amass data generated by third-party websites, and merge it with data from the user’s Facebook account. In its case that this amounts to an abuse of dominance, the Bundeskartellamt relied heavily on the Data Protection Regulation (‘GDPR’). The Bundeskartellamt also referred to the case law of the German Federal Court of Justice, under which contract terms may be deemed abusive if they violate the German Civil Code (in particular when such terms are imposed by a party with superior economic power). More specifically, Facebook’s terms and conditions were found to violate the GDPR, as there was no effective consent for such extensive processing of data – particularly because Facebook users cannot be said to have consented “voluntarily” to such processing, since consent is a prerequisite for them to use Facebook in the first place.
As users ‘pay’ Facebook with their data, one might expect this abuse to amount to a form of excessive pricing. Instead, the authority qualified Facebook’s data processing as an imposition of unfair business terms, which may have something to do with the high bar to establish excessive pricing. Coming up with an effective theory of harm related to this conduct is more difficult, given that users do not pay an obvious price for Facebook. However, the Bundeskartellamt argued that the damage for the users lies in a loss of control: they are no longer able to control how their personal data are used.
The author thinks it is fair to question whether the Bundeskartellamt’s decision unnecessarily blurs the lines between competition and data protection law. The connection between the two, which the Bundeskartellamt finds in Facebook’s market power, is somewhat tenuous. Therefore, it might be better to defer to the GDPR when it comes to data protection. It would follow that any violation of data protections rules would be the exclusive responsibility of the designated data protection supervisor, and competition authorities could expend their time and resources on other matters.
Further, there are cases where neither competition nor data protection law offer the right tools to assess and remedy exploitative practices related to data. In such cases, consumer law can apply – as has happened in Italy. Not unlike its German counterpart, the Italian Authority investigated Facebook for collecting user data, both on Facebook and on third-party websites, without conscious user consent. Rather than a competition law offense, however, the Autorita considered such conduct an unfair commercial practice in violation of the Consumer Code. It is also possible that competition law is the right lens to approach these practices, but ‘data as price’ is the wrong way to think about it. Instead, one should look at data processing, and therefore privacy, as affecting the quality of services—another parameter of competition.
This paper provides a very good overview of the literature on the application of predatory pricing to multisided markets. It also provides many clear descriptions of predatory pricing cases in multisided markets in Europe, as well as of cases that could be said to apply excessive pricing by analogy. The paper does not contain any groundbreaking insight, but it provides a very nice depiction of the lay of the land.