Fair, reasonable and quite unclear

Fair, reasonable and quite unclear

“FRAND” conditions in the EC’s Digital Markets Act

The EC’s Digital Markets Act (DMA) will establish a new set of “dos” and “don’ts” for so-called gatekeeper platforms, as explained in our previous articles here.

Among these obligations, gatekeepers that provide online search services (read – Google) will be required to provide access to their ranking, query, click and view data on fair, reasonable and non-discriminatory (“FRAND”) terms to other online search engine providers. In comparison, gatekeepers that provide access to app stores (read – Apple and Google) will be required to do so on fair and non-discriminatory terms. However, the criterion of “reasonable” is conspicuously absent from these terms for app stores, so these appear to be “FAND” rather than “FRAND” terms.

Is this simply a typo or is there something more deliberate behind not including the “R” for app stores? The debate around access terms for app stores is also very much front and centre in light of the ongoing dispute between Apple and Epic Games, which is currently being played out in US courts.

FRAND/FAND in theory

FRAND terms are typically applied to situations where a product is an important input for certain downstream markets (e.g. standard essential patents relating to a technology that is essential for an industry standard). In these situations, the supplier of that input may hold a degree of market power, so providing access on FRAND terms: (i) limits the extent to which such suppliers are able to take advantage of this market power, whilst (ii) allowing suppliers to (at least) recover the costs of their investment and maintain the incentive to produce such inputs. Further competition concerns may arise if the supplier of that input is also active in the downstream market (and so could, for example, potentially favour its own downstream services). In this situation, FRAND terms may also allow downstream players to compete on a level playing field.

However, what constitutes FRAND is very much in the eye of the beholder. The concepts of “fair” and “reasonable” are subjective and appear closely linked. One would be hard pressed to think of terms that are fair but not reasonable or reasonable but not fair. The concept of “non-discrimination” is perhaps more obvious to define – offering the same terms to equivalent access seekers in equivalent  circumstances – but this is still challenging to apply in practice. For instance, how does one identify whether different access seekers are equivalent?

One interpretation of “fair” is to correct any imbalance of bargaining power between the two sides and to replicate the outcome of negotiations in a competitive market – in other words, a “fair” deal for both sides. “Reasonable” terms of access, to the extent that there is any distinction to “fairness”, might relate more to the price of access. For instance, does the price paid by the access seeker allow it to compete effectively in the downstream market and does the access provider have a sufficient incentive to continue producing this input at this price? If so, such a price might be considered “reasonable”.

FRAND/FAND in practice: app stores and search engines

App stores serve as an important route to market for app developers – without such access, developers would reach a significantly smaller pool of end customers. At the same time, app stores need the content provided by app developers – without such content, the stores would be less attractive to end customers. This is fundamentally a bargaining problem, as both sides would benefit from reaching an agreement with each other. And, as in any bargaining problem, the bargaining power of each side is principally determined by its “walk-away” option, i.e. the payoff from not reaching an agreement (also known as the BATNA, or “Best Alternative to a Negotiated Agreement).

How much can one side live without the other? For app developers, the options may be limited (e.g. Apple’s App Store, Google’s Play Store and/or potentially direct access to customers through non-mobile platforms – although this is a live area of debate in Apple v Epic) and the DMA suggests that app stores are an important gateway for developers to reach end users. For app stores, some apps are likely to be more important than others, but it seems unlikely that one app alone would be singularly important from the perspective of the app store.

This suggests that there is an imbalance in bargaining power between the two sides, which may explain why the DMA’s emphasis is on “fair” access terms. However, it is not clear that it is any less important for access terms to be “reasonable”, even if this is difficult to define. Indeed, this is at the heart of the dispute between Apple and Epic – does a commission rate of 30% reflect a “fair” and/or “reasonable” bargaining outcome?

For search services, access to search data is less obviously a bargaining problem – whilst search engine providers would benefit from having access to the search data of their rivals, those rivals would have little (if anything) to gain from providing it. Indeed, this may be precisely why such search data is not currently offered by search engine providers.

Instead, the DMA’s obligations for search engine providers appear to focus more on providing access to search data on FRAND terms because it is considered to be an essential input for competition – for instance, the DMA refers to access to such data as an “important barrier to entry and expansion, which undermines the contestability of online search engines”. This seems to be more in the territory of conventional access regulation, where the need for both “fair” and “reasonable” terms is well-established (although the challenge still remains of defining what these mean in practice).

Therefore, although the provision of access to app stores and search engines is clearly different, there does not appear to be any obvious reason why the same broad principles of FRAND should not apply. However, the questions around how these principles should apply in practice remain very much open, and regrettably the DMA does little to shed further light on the matter.

Fair, reasonable and quite unclear