Blog

Written by Dr Steve Unger

Addressing competition concerns in the digital sector

In ‘Regulating the digital sector’, Steve Unger identified four key policy lessons about digital markets from the history of regulating telecommunications. Here he asks: how can we apply these lessons to online platforms?

Circuit board

This post follows on from ‘Regulating the digital sector’.

One of the most striking features of the current debate about online platforms is how much general attitudes have shifted over a very small number of years.

A sector of the economy which used to be admired for its innovation is now seen by many as a source of harm, and an inevitable target of regulation. My view is that this shift in sentiment is dangerous, creating a risk of poorly designed interventions which do more harm than good.

This concern is based in part on recent competition rulings from the European Commission, which have had a high public profile, but which in my view lack effective focus.

For example, one of the Commission’s findings against Google is that it acted illegally by refusing to make its Android operating system available on an open source basis. The Commission argues that this has held back the development of alternative operating systems, and therefore closed off a channel for competing apps and services. I believe this conclusion to be flawed. The use of ex post competition powers to mandate access to the Android source code is neither necessary to enable other companies to develop competing operating systems, nor is it likely to be an effective means of addressing the competition concern which has been identified.   

That said, the power held by online platforms does of course raise potential competition concerns. I am not suggesting these should not be addressed, just that a more measured approach is required.

The competition concerns which have been identified tend to fall into three broad categories:

  • First, the nature of online platforms is such that everyone wants to connect to the same platform. These ‘network effects’ means that at any point in time there are likely to be only a small number of platforms in the market, each with high market shares.
  • Secondly, online platforms have an incentive to cement their position, by turning the transient market power which results from network effects into an enduring position of dominance. They may have the ability to achieve this, by accumulating large amounts of data on consumer preferences, which can be used to optimise the performance of their platforms in a manner that cannot be matched by a challenger.
  • Thirdly, there is a concern that once platforms have become established, they become gatekeepers for a variety of other services. This allows them to extend their dominance into other markets by discriminating against those other services which they regard as competitors.

Interestingly all three of these concerns have parallels in traditional telecoms regulation. However, details really matter when it comes to designing effective responses. And whilst traditional responses are informative, they may not represent the most appropriate approach going forward. Specifically:

  • Network effects are a key characteristic of traditional telephone networks, to which the established regulatory response has been a requirement for different networks to interconnect with each other. Given this, it is not surprising that interoperability between different online services has emerged as a potential mechanism for mitigating the impact of network effects on these services. However, the parallel with traditional telephone networks is imperfect, given that a modern smartphone provides consumers with simultaneous access to multiple services. And there is a risk that the common technical standards which are required to deliver full interoperability between those services will hold back innovation.
  • Enduring dominance is a key characteristic of the ‘last mile’ of telecommunications networks, due to the economies of scale associated with these networks. The traditional response has been various forms of network access obligation. Given this, it is again not surprising that access to large datasets has emerged as a potential mechanism for making online platform markets more contestable. However, more work is required to understand which datasets are most important as an enabler of competition, and how to address any privacy concerns associated with making these more widely available.
  • The use of discriminatory behaviour to try to extend dominance to adjacent markets is again a common characteristic of telecoms markets. The traditional response has been the imposition of non-discrimination obligations, such as the Net Neutrality rules which apply to internet service providers. An approach of this kind might be appropriate for online gatekeepers, but there are practical challenges designing rules which allow for those forms of discrimination which are beneficial (such as the prioritisation of emergency calls by communications networks, or the curation of content by online publishers) whilst preventing discrimination whose primary purpose is anti-competitive.   

I am encouraged by the recent publication of the Furman review in the UK (‘Unlocking digital competition’), and the report of the special advisors to the European Commission (‘Competition policy for the digital era’). Both reports acknowledge the important competition issues raised by online platforms, indicate potential solutions, but recognise that further work is required to get the detail of that right. 

It’s that detail which will be the subject of further work by myself and many others, which I will be writing about in future posts.

  • About the author

    Dr Steve Unger, Affiliated Researcher

    Steve Unger was until recently a Board member of Ofcom, the UK regulator responsible for digital communications. He had various responsibilities, including setting regulatory strategy for the UK, representing the UK in international negotiations, and leading Ofcom's technology programme. He spent a brief period as ...   Learn more

    Steve Unger