Competition Law Forum Hosts High Level Roundtable on Ecosystem Competition
On 13th May, the Competition Law Forum hosted a roundtable discussion at the British Institute of International and Comparative Law with regulators from the ACCC, CMA, the European Commission, economists, academics and lawyers.
Key messages
The session was held under Chatham House rules so, sadly, if you missed it, you missed out! All is not lost, though, as we have set out below a few of the key takeaways gleaned from our esteemed panel:
• No legal definition or clear consensus about what constitutes an ecosystem for the purposes of competition law. There are various types of ecosystems: business, entrepreneurial or platform ecosystems. Our main discussion focused on platform ecosystems, which tends to be coordinated. Platform ecosystems typically have a hub firm that engages in orchestration activities, although firms can be active in multiple ecosystems, and it is not always clear who the hub firm is. Some argued that the concept of 'ecosystem competition' was largely a matter of old wine in new bottles and the competition system was already well placed to analyse and deal with any potential concerns using existing tools for analysing conglomerate effects. Discussion focused on the possibility that the act of identifying a company as orchestrating an ecosystem may create unwelcome presumptions of this being a problem in itself. There were mixed views about the likelihood of this labelling creating unwarranted action by authorities.
• Potential competition concerns within ecosystems. Despite slightly different ecosystem definitions, the sort of list of potential anti-competitive effects are broadly like those found outside an ecosystem such as self-preferencing and preinstallation. When the hub firm expands its vertical scope by integrating activities that were previously conducted by independent complementor firms, the level of integration density increases as ecosystems become more established. The expansion of digital platform ecosystems has raised concerns relating attribute effects and risks of foreclosure. Self-preferencing, where the hub firm selectively promotes its own complementary goods or services on the expense of independent complementor firms, can be done through algorithmic biases, through manual curation or sorting features, default options where a platform ecosystem comes preloaded with the hub firm's internally developed complementary goods. Those practices in themselves do not always harm competition, they can bring benefits to consumers, but they are harmful to competition when they prevent or inhibit rivals' ability to effectively compete.
• Data can create an informational advantage. Those in favour of the 'newness' of the ecosystem challenge focused on several key challenges thrown up by recent industry and technological developments. These included, the ubiquity and sharing of consumer data, the opacity of data harvesting by firms and the opacity of links between firms with access to that harvested consumer data. Due to hub firms' access to often complete set of data there can be economies of scope with respect to data. They can use data to develop new products and services as well as import it into existing products and services. Hub firms are at an opportune position to track all the data flows and use this to their advantage. They can typically see all the data generated in the ecosystem and use this to decide which product categories to enter, or how to compete. Whereas a complementor firm might only see or observe the data that is generated from its own business activities. However, it was acknowledged that the importance of data and the extent of economies of scale and scope are case specific.
• Years of perceived under-enforcement have led to the risk of over-compensation and was viewed with differing degrees of concern. There was discussion of the problem of authorities over-compensating for their belief that they under-enforced for years at the start of the recent wave of digital firm creation and reading across lessons from past waves of technology that might not read across. The risk of over-compensation, or refighting old battles with new cases was viewed with differing degrees of concern.
• Interdependencies. It is important for merger control not just to look at individual linkages between different products within an ecosystem, but also be aware of interdependencies between these markets. In a merger case authorities take a two-step approach (i) market definition, and (ii) competitive assessment. An ecosystem as a system needs to be assessed whether it is a market in itself that includes all elements of the ecosystem, a primary market with secondary markets around or one overall secondary market in addition to that. In relation to the competitive assessment, an ecosystem in itself does not appear to be a concern, it all boils down into horizontal, vertical or conglomerate effects that can be numerous based on the interdependencies with the ecosystem.
• Remedies. There was some discussion of whether digital markets were better or worse suited to behavioural remedies and how authorities might think about potential enforcement traps and identifying the real locus of harm and not be caught up in a wider fear of 'bigness' or 'ecosystems'.
Top tips to takeaway...
• Ecosystems are not simply multiproduct firms as multiproduct ecosystem firms can be embedded in a multi-actor ecosystem.
• The concept of 'ecosystem competition' is more complex than traditional competition in terms of scale, speed, and network effects, which apply across complementary products rather than product by product and service by service partly because of the data advantages.
• Ecosystems typically ensure interoperability between complementary services, which can be an advantage compared to what is available outside the ecosystem. Whether this is good for consumers is a fact specific question.
• There is more interconnectedness in ecosystem competition and therefore more scope for feedback effects.
• Regulators need to carefully consider whether the competitive harm by a hub firm's activities lead to a lessening of competition within the ecosystem, and balance this against additional value being created for customers.
