There is an intense academic discussion regarding whether consumers and business users are exposed to conduct that may amount to competition law abuses when using Internet services. The discussion is connected to the Internet phenomenon of ‘platforms’ or intermediaries. The multitude of direct customer–supplier transactions making up everyday business conduct are, to an increasing degree, replaced on the Internet by an intermediary, the platform, matching the customer with the supplier. Platforms are able to perform role because they provide efficient and easy matching. Further, internet platforms may, due to certain special and somewhat unique characteristics – like network effects, tipping and path dependency – become central ‘hubs’ between purchasers and suppliers.
This chapter, available here, focuses on the application of competition law vis-à-vis the platforms collecting personal and non-personal data. It considers questions such as: may competition law be used to gain access to intermediaries’ data, and the infrastructure around that data? May competition law be used to limit the impact of strong platforms and ‘hubs’ in the digital economy? The chapter will also consider sector-specific regulation on data.
Chapter 2 looks at platforms and their ecosystems.
Platform business activities take place at three different ‘levels’ of competition. At the first level, there is competition between platforms or ecosystems, i.e. ‘inter-platform’ or inter-ecosystem competition. While competition at this level is typically fierce, certain behaviours – e.g. platform providers regularly exclude or restrict access to competing platform providers on their own platforms – might infringe competition law. At the second level, one finds intra-platform competition. The question here typically concerns the type of restraints that platform or ecosystem leaders can lawfully impose on business users within their ecosystems. A third dimension of competition might consist of the creation, collection, storing, commercial use, and dissemination of data. System leaders may act as data gatekeepers, controlling and locking in their customers while simultaneously excluding or limiting data interoperability with other systems, specific devices or machine-producing firms and customers.
While it may be difficult to identify abusive practices in the context of inter-ecosystem competition – since no platform or ecosystem leader may have sufficient market power –, abuses are more likely as regards intra-system competition. Within an ecosystem, firms compete with system leaders and with each other on downstream, upstream or aftermarkets. Since a system leader will often be able to ‘govern’ the ecosystem, it is more easily qualified as being dominant and, hence, as being able to abuse its market power.
Chapter 3 considers the relevance of data for dominance in the digital economy.
It is increasingly acknowledged that controlling data is key for firms to be able to gain and hold market power. In digital ecosystems, platforms are the main collectors of data, and they are often network managers as well. Platforms can thus create bottlenecks and replace direct customer–supplier for their own interfaces. This enables platforms not only to act as retail service providers, but also to collect and transfer commercial data regarding purchasers, products and suppliers.
Therefore, an important question is whether control of data flows amounts to market power in the ecosystem and connected markets. In practice, the identification of dominance should also take into account direct and indirect network effects, network externalities and market tipping dynamics. Nonetheless, high-quality data and better predictive modelling tools, combined with a data-driven business model, can contribute to the creation and maintenance of dominance by platforms in digital ecosystems. It is common for data to be collected and processed by the ecosystem leader, enabling it to have an exclusive on that data or to give access to this data to third parties arbitrarily or in a discriminatory fashion. Taken together, these data practices may make it easier to establish market power as regards intra-ecosystem markets.
Inasmuch as markets have tipped and system leaders are perceived to be regulators within their ecosystems, one may even reach the conclusion that a company is dominant without needing to engage with any in-depth analysis of fluid relevant markets. In such circumstances, there will be little – if no – competition for that network, and the system leader therefore has a special responsibility to create a level playing field for downstream, upstream and connected markets.
Chapter 4 deals with anticompetitive conduct.
Can a platform infringe competition law by having a data advantage and leveraging that advantage to further its market dominance? To implement a data-driven business model or to make use of network effects is not anticompetitive in itself. In the digital economy, it seems that most anticompetitive behaviours by platform providers are discriminatory exclusionary practices – with intra-ecosystem restraints on competition seemingly more relevant than inter-platform competition. For example, the system leader could collect and provide access to data or the result of predictive modelling to a specific firm (or keep it to itself), while refusing access to others. This would enable the dominant firm to leverage the data advantage against competitors.
A related possibility is to base a finding of abuse on a monopoly leveraging theory of harm. This would require two separate markets (data market and device market); unlawful business conduct by the dominant Internet intermediary on the primary data market (e.g. lock-in, non-access to data, non-assert requirement, or discrimination in access to data); subsequent entry onto a (competitive) secondary market, either by the Internet intermediary or a proxy; exclusionary effects on the secondary market, in particular the foreclosing of equally efficient competitors; and the absence of an objective justification for not providing access to the data.
These theories of harm are the outer edges of existing competition law. Light might be shed on their relevance by ongoing investigations into the digital sector – particularly the EU and German investigations into Amazon’s practices as regards third-party vendors in its ecosystem.
Section 5 looks at sector-specific regulations and data protection rules.
Competition law is a heavy-handed instrument. In reality, as EU antitrust law stands today gatekeepers are for now rather ‘safe’. Instead, public authorities seem to be relying on sector-specific regulations to create competition in the digital economy.
The EU has introduced a platform-to-business (P2B) Regulation, which targets the platform–business interface and mostly focuses on transparency. However, this instrument only targets data practices marginally – Article 7 obliges providers of online intermediation services to provide business users with a clear description of the scope, nature and conditions of their access to use certain categories of data; while Article 6 requires providers to be transparent if they intend to discriminate by giving better access to data to affiliated firms than to other business users.
There are also several instruments in various sectors which either have been recently implemented or are being developed in the EU. The PSI Directive governs access to government data with the goal of creating a level playing field when government data is an input to a commercial activity. The eCall Regulation for in-vehicle systems lays the groundwork for a future where competitors are able to access data originating from cars used by individuals, by requiring the vehicle to use or be connected to a standardised, secure and open-access platform. Finally, the recently updated Directive on Payment Services (DPS II) creates a right for third parties to access the banking consumer data under certain circumstances.
This is a thoughtful piece. I particularly enjoyed that it focuses squarely on what is special about data, and how it influences competition analysis. It is interesting to me that this focus leads the author to the conclusion that regulation seems to be a more effective tool than competition – despite the earlier sections of the paper seemingly endorsing very expansive approaches to competition enforcement. However, as the author acknowledges that such an expansive reading faces significant headwinds in practice, I think these sections are better read as providing a description of ‘proposals’ to use competition law to tame digital platforms.
Nonetheless, there is a section in the paper that struck me as being a bit too ambitious – the one about establishing dominance. I have some sympathy for arguments that market definition is unnecessarily cumbersome in digital contexts, but that is tempered by a belief that we can look directly at anticompetitive effects. This, however, is different from generically considering that platforms are dominant and subject to level-playing field duties merely because we call them ‘regulators’ within their ecosystem. Regulatory instruments, of course, can impose such a duty, but it is hard to connect such an approach to competition law as we usually perceive it – i.e. as requiring the identification of specific anticompetitive conduct by companies with demonstrable market power.