All DMA gatekeepers are equal

But are some more equal than others?

In the Digital Markets Act (“DMA”), the EC defines “gatekeepers” as companies that meet certain criteria. In particular, it states that a company is a gatekeeper where it:

  • has a strong economic position, with a significant impact on the internal market and is active in at least three EU countries (the thresholds defined are for the company to have had turnover equal to or greater than €6.5 billion in the last three financial years, or an average market capitalisation of at least €65 billion in the last financial year);
  • has a strong intermediation position, meaning that it links a large user base to a large number of businesses (specifically, a core platform service it provides has more than 45 million monthly active end users in the EU and more than 10,000 yearly active business users in the EU in the last financial year); and
  • has (or is about to have) an entrenched and durable position in the market, meaning that it is stable over time (this is determined by the thresholds in the previous point being met in each of the last three financial years).

These are clearly different criteria to dominance. And the lack of any explicit assessment of market power in the gatekeeper criteria means that there is scope for platforms with quite different market positions – and in particular different levels of criticality as channels for businesses and consumers – to be classified as gatekeepers. In other words, from an economic perspective, are firms that are likely to meet the criteria really “gatekeepers” to the same degree, or is there an implicit hierarchy of gatekeepers that varies across the different core platform services? Or, to go further still, are there some firms that aren’t really gatekeepers at all in an economic sense, but would be classified as such based on the EC’s criteria?

To help illustrate this point consider the example of a greengrocer wanting to branch out into online delivery via a website and/or app. What channels (or “core platform services” to use the EC’s terminology) would the business need to use to reach and advertise to customers? And how many (potential) gatekeepers would it need to reach them?

Our greengrocer would need to set up a website, and this would need to show up in a search engine search. It may also want to set up a social media page, or appear on other third-party platforms who could send traffic its way. Our greengrocer’s app would probably need to also be available on app stores to have any chance of reaching a significant number of customers, and the operating system (OS) used to develop the app is linked to the app store customers download it from (i.e. iOS is needed to reach Apple customers via the App Store and Android to reach Google customers via the Google Play Store). Finally, it may wish to market to customers in some way, possibly through online search and/or display advertising, traditional channels such as print and broadcast media advertising, or physical displays such as billboards or flyers.

These channels are likely to differ in the extent to which they are a “must have” for our hypothetical greengrocer (e.g. using a specific advertising channel may be more discretionary whereas being available in an app store might be more essential). In addition, it is clear that they differ in terms of the number of platforms operating within the space and the strength of individual competitors. Mapping our greengrocer’s (potential) requirements to the core platform services set out in the DMA:

  • Operating systems and app stores: clearly, for this service at least, Apple and Google are the main games in town: most people either have an iPhone or an Android phone, and either use Apple’s App Store or the Google Play Store. This means that there is likely to be no realistic way of our greengrocer avoiding Apple and Google here in order to get its app in front of customers.
  • Online intermediation services: to facilitate its online deliveries, our greengrocer may want to reach customers through an online intermediation service, such as Deliveroo or Amazon Fresh in the UK. While the greengrocer might choose to set up its own delivery infrastructure, these platforms offer a route to customers where the greengrocer does not need to invest in hiring staff or vehicles to make deliveries.
  • Social media services: a social media page could be useful for our greengrocer, but this is perhaps not as important as having an app and website (although this is likely to differ case-by-case). To the extent that it finds a social media page useful, the greengrocer would likely opt for a Facebook, Instagram and/or Twitter page, but it may also/instead opt to have a presence on successful newer social media services, such as TikTok, Snapchat or Clubhouse. On the advertising side, as alluded to above, there are many options outside of social media display advertising, including open display advertising, search advertising, and more traditional routes such as print/broadcast advertising, billboard advertising or even handing out flyers on the local high street. All are ways to capture eyeballs and generate sales.
  • Online search engines: if the greengrocer wants to appear (prominently) in a significant number of people’s search results, it would realistically have to be on Google, given its strong market position (having a share of search globally of over 85%). Although on the advertising side, as noted, there are a number of offline and online alternatives to search advertising for reaching consumers.
  • Cloud computing services: our greengrocer’s app and website would require web hosting. For this, there are various traditional web hosting companies, although more popular and/sophisticated sites (which the greengrocer’s website/app may be) are likely to require cloud-based web hosting. The key players are Amazon (AWS), Google (Google Cloud Services) and Microsoft (MS Azure).

This heterogeneity across core platform services raises the question: are all gatekeepers equal? If by “gatekeeper” we mean the criticality of a firm in its role sitting between business and consumers, then the answer to that question would seem to be no. In some core platform services there appear to be a small number of options, while in others businesses/consumers have many options. There is no direct mapping to firms which are likely to be classified as gatekeepers and others that are not (which may raise other questions around the potential negatives of asymmetric regulation).

A more careful treatment might suggest a hierarchy of gatekeepers that varies across the different core platform services. Extending the gatekeeper analogy, one could think of the “number of gatekeepers” and the “width of the gate” as two important dimensions:

  • The “number of gatekeepers” represents the degree of competition in a given channel. Here we are referring to the total number of credible competitors in the space, rather than only firms meeting the EC’s criteria. The more options a business has for reaching customers via a particular core platform service, particularly where customers can easily switch between platforms or multi-home, the more opportunities a business has for “getting through the gate”. As shown above, a company can offer a core platform service for which there could be many or few alternative providers.
  • The “width of the gate” represents how essential a business might find using a firm providing a particular core platform service – i.e. the extent to which it could avoid using gatekeepers/firms offering the service, dodging the gatekeepers while still reaching enough consumers. If gatekeepers are less essential for businesses/consumers, this could suggest that they are lower in the hierarchy, which may affect the need for regulatory intervention. There is also a potential definitional issue here: if businesses can get by without using a core platform service at all, this might mean that the core platform service is too narrowly defined and there are viable substitutes which fall outside the scope.

The DMA in its current form does not take these factors in to account and would broadly treat all companies deemed to be gatekeepers under the criteria in the same way (albeit with specific obligations that are more targeted to different types of services). However, given the potentially significant differences between the criticality of different core platform services and the operators in them, the EC’s binary approach risks unnecessarily burdening companies which – as the EC also identifies – have created significant value for businesses and consumers. A more nuanced approach could re-visit both the gatekeeper criteria, to bring them closer to identifying true ‘must have’ platforms, and the obligations – to tailor them to achieve a better match to the ‘gatekeeping strength’ of a platform. Moreover, the more the EC sticks to its ‘one size fits all’ gatekeeper criteria, the stronger the case for a more flexible approach to the obligations each gatekeeper is required to meet.

All DMA gatekeepers are equal