The judgment of the CJEU in Budapest Bank (Case C-228/18) is the most recent case that provides guidance with regard to the application of art. 101 TFEU in the context of multi-sided platforms. The CJEU explicitly confirmed the possibility of finding restrictions of competition by object by such players despite the complexities originating from their multi-sided nature. However, the manner in which such findings can be reached in practice, based on the findings of the CJEU, is far from simple and may require building up a track record of platform cases first.
The case of Budapest Bank entails yet another case dealing with Multilateral Interchange Fee (MIF) agreements adopted in the context of the VISA and MasterCard payment card systems. MIFs constitute the fees charged by the cardholder’s bank (also referred to as issuing bank) to the merchant bank (also referred to as the acquiring bank) for each credit card transaction (see here for more information and regulation concerning MIFs).
In the context of this case it was established that in 1996 several banks in Hungary adopted an agreement that introduced a uniform MIF for the VISA and MasterCard credit card systems. The Hungarian competition authority found this practice to be a restriction of competition by object and effect constituting an infringement of art. 101 TFEU (and the national equivalent thereof).
The decision of the Hungarian competition authority was appealed before the national courts where it eventually reached the Supreme Court of Hungary, which was required to assess the findings of the Hungarian competition authority in this matter. In these proceedings, the Hungarian Supreme Court referred four preliminary questions about the application of art. 101(1) TFEU of which the first two were addressed by the CJEU:
- Can the same conduct be classified as an infringement of that provision by object and effect simultaneously (albeit on different foundations)?
- Does the MIF agreement constitute a restriction by object in so far as it fixes a uniform amount for the Visa and MasterCard interchange fee?
The answers provided for these questions are not only useful with regard to the general application of art. 101(1) TFEU as such, but are of particular relevance in the context of multi-sided markets as will be discussed in this post.
Restrictions by object and/or effect
With regard to the first question, the CJEU found the same conduct could have both the object and effect of restricting competition [par. 44]. This is unsurprising, as restrictions by object are considered inherently harmful for competition. A competition authority that identifies an object restriction it is no longer required to look into the effects of such practice, however, this does not mean it is precluded from doing so [par. 40]. Accordingly, when a competition authority or court decides to go the extra mile by showing the effects of the object restriction, they can also indicate that such a restriction by object has the effect of restricting competition.
This finding should, however, not be confused with the possibility that restriction of competition by object can be found based on an effects analysis. If a practice does not display a sufficiently harmful character for it to qualify as an object restriction to begin with, an effects analysis cannot be used as an alternative route to reach such a finding. In such instances the effects analysis would be required to find a restriction by effect instead.
This is confirmed by the CJEU that indicated that courts and competition authorities can qualify the same conduct as a restriction by object and as a restriction by effect provided that such authority or court is able to provide the required proof for making such (separate) finding and to specify to what extent such proof relates to one or the other type of restrictions [par. 43]. Therefore the risk that line between restrictions by object and restrictions by effect may become blurred is somewhat mitigated with this judgment [on this matter see Opinion of AG Kokott in T-Mobile par. 42-49 and Opinion of AG Wathelet in Toshiba par. 40 onwards].
Does the MIF constitute a restriction of competition by object?
In addressing the second question, the CJEU restated that the concept of object restrictions should be interpreted restrictively and relied upon only in cases where the investigated practice is one that is by its very nature harmful to competition [par. 54]. The assessment of the harmful character of a practice must take into account the legal and economic context of each case in light of its specific circumstances. When the case concerns a situation that involves competition on multiple (interrelated) markets such an analysis requires looking into the interactions between the relevant market(s) and the other related market(s) [par. 67-68]. As in any case concerning MIF agreements, in this case the concerned banks as well as VISA and MasterCard maintained that the MIF agreements were needed to ensure a certain balance /ratio between the issuing and acquiring banks in their respective credit card systems [par. 66]. This aspect was considered by the CJEU as highly relevant for assessing whether the MIF agreement could qualify as a restriction by object [par. 71-73]. Furthermore, the refereeing court has also indicated the MIF agreement, despite eliminating competition with respect to MIFs between VISA and MasterCard, may have promoted competition on other parameters. According to the CJEU, if that were indeed true, such circumstances would stand in the way of qualifying the MIF agreement as a restriction of competition by object [par.74-75].
Very interestingly, and particularly important in context of multi-sided platforms), the CJEU (referring to the AG opinion) also noted that finding a restriction by object (and thus refraining from looking into the effects of a concerned practice) requires a significant degree of solid and reliable past experience with such practices [par.76]. Nevertheless, the CJEU explicitly noted that the complex nature of multi-sided markets (e.g. in this case payment cards) is not, as such, sufficient to stand in the way of finding an object restriction [par. 80]. Both these statements made by the CJEU provide valuable guidance with regard to the application of art. 101(1) TFEU to multi-sided platforms.
Finally, in addition to the potential pro-competitive effects of the concerned practice, the finding of an object restriction also requires considering the counterfactual for the investigated practice when assessing its (potentially) harmful ‘nature’ [par. 81-84]. By doing so, the CJEU eliminated the doubt left by the GC in Lundbeck [par. 473].
With regard to the MIF agreement in this specific case the CJEU ultimately left it to the national court to consider whether the evidence available to it was enough to conclude that it displays a sufficiently significant degree of harm that would justify qualifying it as a restriction by object [par. 86].
Implications for multi-sided platforms
The findings of the CJEU in this case entail an evolution of the current practice in the context of platforms. Of course, as in most cases, several uncertainties still remain to be resolved. Nevertheless, the case confirms in a more explicit manner that the multi-sided character of platforms will constitute an important and inseparable part of the legal analysis of their business practices.
Finding a restriction of competition by object in the case of platforms will be difficult due to their reliance on network effects
When looking at the assessment of the CJEU in this case in light of the special character of platforms it becomes quite evident that finding an object restriction may turn out to be a difficult task. The main reasoning put forward by the concerned banks and credit card schemes to justify the MIF agreement was that it was needed to create an optimal balance between the issuing banks and the acquiring banks. Similar arguments were also made in previous MIF case law. This reasoning and its economic rationale were considered in this case (as well as in Cartes Bancaires) an important reason for not labeling the MIF agreement as a restriction of competition by object.
The economic rationale of this practice, i.e. to maintain an optimal ratio between the platform participant groups is, however, inherent in most of the commercial decisions made by platforms. The success of platforms is in fact dependent of their ability to achieve and maintain such an optimal ratio. Failure to do so would lead to a forced market exit due to the dependence of platforms on the direct and indirect network effects that their various customer groups exhibit. Accordingly, finding a restriction by object in circumstances where the concerned practice relates to optimizing the ratio between the platforms’ customer groups would be very difficult, if at all possible. After all, if the practices of the platform are indeed directly related to creating an optional ratio of customers, it is hard to maintain that such practices have their object to restrict competition. To some extent it can even be argued that such actions will also have a pro-competitive potential since they essentially sustain the existence of the platform as such.
Therefore, to find a restriction by object in the case of platforms, a competition authority would ideally be able to provide evidence that the concerned practices do not relate to achieving or maintaining a certain ratio between its various customer groups, nor are such practices related to the multi-sided business model of the platform.
Having such evidence would then allow the competition authority to rely on the expertise of current practice outside of the scope of platforms to qualify a certain practice as a restriction by object. In other words, if the concerned practice is not suitable for pursuing a legitimate and economically rational objective in the context of platforms (or multi-sided markets) there is no reason why the multi-sided nature of platforms as such should stand in the way of finding an object restriction. For example, a price fixing agreement between Expedia and Booking.com with regard to the commission they charge hotels would have no relation to their multi-sided character. Therefore, such practice would not be any different from any other price fixing agreement outside the scope of platforms, meaning that a finding of an object restriction is such a case should not be problematic.
When, however, the concerned practice may also relate to a potentially legitimate and economically rational aim in the context of platforms, finding a restriction by object would not likely be possible at this point in time because of the limited experience that competition authorities have with platforms. In such cases, the investigated practices will have to be, for now, subject to an effects analysis.
Finding a restriction by object in case of platforms requires a ‘learning period’
The CJEU indicated that finding an object restriction requires sufficiently solid and reliable experience with regard to the investigated practices, which would justify such qualification. Given the (relative) novel nature of multi-sided platforms in the context of competition law, it is hard to maintain that such experience exists. This is particularly true when the investigated practices are fully intertwined in the multi-sided business model of such platforms. There are currently only a handful of cases concerning the application of art. 101 TFEU to the business practices of platforms at the EU level and we can observe a similar situation at the Member State level.
So how is such experience to be gained? This is where the CJEU answer to the first preliminary question in Budapest Bank comes in. According to the CJEU the Commission, national courts or national competition authority can label the same practices as being a restriction of competition by object and effect. This combination can be relied upon provided that the required type of evidence for each category is provided. Due to the lack of experience of current practice it is likely that the first cases concerning platforms will often have to be treated as restrictions of competition by effect. However, if certain types of practices that at first were not evidently harmful to competition turn out to always be harmful to competition in the context of platforms (and not be exempted under art. 101(3) TFEU)– then such practices could at some point be qualified as object restrictions.
What remains to be seen after Budapest Banks is how many cases are needed to maintain that such solid and reliable experience has been acquired. In other words, how many incriminating effects analyses are needed to create a restriction by object?
Budapest Bank provides the legal basis for what will be a lengthy road towards building up a solid and reliable experience in the context of platforms. While the requirement to have such experience is justified due to the special character of restrictions by object, fulfilling this requirement in the case of platforms will require a sizeable amount of case law to be built up first. At the same time, business practices that do not interact with the multi-sided business model of such platforms can be found to constitute object restrictions based on the existing experience of current practice. As noted by the CJEU, the multi-sided character of platforms as such does not stand in the way of such a finding.