The topic of the essential facility doctrine has made unexpectedly frequent appearances in EU case law over the past few years, which continues to refine the application scope of Bronner. The recent preliminary procedure in the case of Google Android Auto will inevitably become part of this process, which may have some significant consequences in the context of digital platforms. The Opinion of AG Medina adds a new perspective for assessing the applicability of Bronner, namely by looking at the functional and commercial purpose of the facility at hand. Furthermore, it depicts the limits of the objective justification under art. 102 TFEU in cases concerning the restriction or denial of market access. Both matters are of great importance in the context of the digital economy and, specifically, digital platforms where conflicts concerning access and the quality thereof are expected to arise whenever there is a misalignment between the interest of the respective platform and its commercial customers. Accordingly, clarification by the CJEU would be valuable for future case practice and have significant consequences for big tech platforms. Therefore, it requires a great deal of nuance to ensure both coherence with previous practice and effectiveness for future practice, which will be challenging when looking at the AG Opinion that offers a mixture of valuable and concerning considerations to be addressed.
Background to the case
The factual background to the case concerns a dispute between ENEL, an electric car charging services provider, and Google. The conflict between the two parties stems from the lack of interoperability between Google Android Auto and the app developed by ENEL that was designed to allow users to (i) search for charging stations on a map and book them, (ii) transfer the search to Google Maps to navigate to the selected charging station; and (iii) start, stop and monitor the charging session and the payment relating to it. The app was already available on Google Android OS for smartphones but not for Auto (i.e., the dedicated car version installed on the cars’ entertainment system).
Given the evident relevance of ENELs app by the name of Juicepass for owners of electric cars that run Android Auto, it requested access to Android Auto. Upon request, Google informed ENEL that it could not grant it access as the technical template needed for apps like Juicepass had not yet been developed. At the moment of the request, there were only templates for media and messaging apps and for Google’s navigation apps (Maps and Waze). Following this reply, ENEL requested that Google make its templates compatible with its Juicepass app, which Google refused to use based on security concerns and the need for it to allocate resources sensibly for developing new templates. Following this refusal, ENEL filed a complaint to the Italian competition authority (AGCM), claiming that such actions constituted an abuse of dominance.
The AGCM confirmed the position of ENEL and found that Google’s refusal constituted an abuse of dominance under art. 102 TFEU, resulting in a fine of approx. 102 million euro. According to the AGCM, Android Auto was an indispensable product for developers of apps that are aimed at drivers. Furthermore, the AGCM also found that Google had intended to favor its proprietary Google Maps app to the detriment of other apps like Juicepass that could compete with it.
Google first appealed the decision of the AGCM before the Regional Administrative Court of Lazio, which dismissed its action in its entirety, and then later before the Italian Council of State, which stayed the proceeding to ask for guidance from the CJEU in the matter of the application of the essential facility doctrine according to Bronner. The questions referred to by the Council of State allow the Court to clarify the applicability of Bronner in general and the challenging context of digital platforms that concern this case.
The Opinion of AG Medina in this case offers a preview of the challenging nature of this matter. The approach to Bronner offered in the Opinion, if followed by the CJEU, can offer a valuable development for current practice despite the difficult compromises it entails with respect to objective justification arguments. At the same time, overstretching such an approach could lead to unwarranted detrimental consequences for digital platforms.
The Good: a new perspective on the essential facility doctrine
Over the past few years, multiple cases have refined and clarified the boundaries of the Bronner case law. In Slovak Telekom, it was indicated that Bronner would not apply where the restriction of access is not the result of an outright refusal but rather the result of a so-called constructive refusal consisting of unfair trading conditions that make access to the respective facility economically non-viable. This is quite reasonable given that the choice of dominant undertakings to deal with third parties is protected by fundamental rights, whereas the possibility to impose unfair or exploitative trading conditions is not. Accordingly, in this latter scenario, the balancing act incorporated in Bronner does not arise, thereby removing the need for its application. Furthermore, it was added that Bronner would also not apply when the matter of access to the respective facility has already been covered by (secondary or national) legislation. This is because the balancing act incorporated in Bronner is presumed to have been incorporated in the legislative process of such a framework with the result of access to the facility being mandated. Under such circumstances, there would be no room for redoing such exercise in the context of a competition law case.
This second point was also confirmed in Lithuanian Railways, which added that the application of Bronner may also depend on whether the respective facility resulted from an entrepreneurial risk. This is because, next to the protection of fundamental rights, the Bronner criteria have been developed to ensure that undertakings are incentivized to invest in creating such facilities. In situations where the respective facility came into being through other means, such as public funding, this aspect of entrepreneurial risk and the incentive would be absent, meaning that the application of Bronner would be unwarranted. Finally, the Court also noted that destroying one’s facility does not equate to a refusal in the sense of Bronner, which is intended to do quite the opposite: reserve it for the dominant undertaking itself to use.
More recently, in Bulgarian Energy, the General Court restated the importance of additional regulatory frameworks that regulate access to the facility and evidence of an outright refusal to supply on the part of the dominant undertaking. Next to this, the GC also addressed the legal status of the dominant undertaking with respect to the respective facility. Here, the GC noted that ownership, as such, is not per se required; an exclusive right to make use of the facility was sufficient. What was important in this respect, however, was that the dominant undertaking paid a fixed yearly fee to obtain this exclusive right, which constituted an investment and, thus, an entrepreneurial risk in the sense of Bronner, according to the GC [paras. 263-269].
Now, with the recent case of Google Android Auto, the CJEU will have the chance to make another refinement to the application of Bronner, focusing this time on the nature of the facility. According to the AG opinion, Bronner should not apply to the refusal in this case due to the nature of the facility itself. Google Android is, in principle, an open-source software platform designed to be developed with the help of third parties and, most importantly, was designed to facilitate the participation of third parties (i.e., app developers and providers) [paras. 36-40]. In this respect, the creation of this platform and its viable existence depends on its popularity among such third parties and, thus, inherently also, their access to it. In other words, the nature of the Android Auto platform and the rationale of its business model rely significantly on third parties developing apps for it. Consequently, according to the AG, the choice of whether or not to open the facility up to third parties was already made when Google decided to create such a platform.
The logic here is that if the platform is designed to cater to third parties and, in fact, may not viably survive without them, then it can be presumed that access to it is by default granted as a matter of principle. When going back to the balancing act of Bronner, this would entail that the dominant undertaking, in this case, Google, utilized its fundamental right by opening up to third parties rather than refusing them access. Therefore, there would be no need to apply Bronner if access is later refused. Furthermore, in such circumstances, the refusal cannot have the purpose of reserving the facility (the Android Auto platform) for the dominant undertaking.
Overall, as a matter of principle, this approach is sensible and could represent a valuable development of current practice. If a facility is created with the core purpose of facilitating access and harnessing it as a core part of its business model, then there would seem to be no reason to re-evaluate this aspect under the strict criteria of Bronner every time a third party is denied access. This is similar to situations where access is mandated by a regulatory framework, as seen in Slovak Telekom and Lithuanian Railways. Furthermore, lifting the protection of the Bronner case would, in principle, not impact the entrepreneurial incentive to invest in creating the respective facility, as the facility’s purpose was always to be accessed by third parties. Accordingly, an access remedy would not significantly limit the commercial freedom of the dominant undertaking since it was never designed to be solely used by it.
Applying this logic to the specific case of Android Auto would, therefore, seem sensible. Android Auto, like Android OS on smartphones, has indeed been developed to allow app developers to participate. Without such participation, it could never have survived, as showcased by Windows Mobile, Symbian, and BlackBerry. In the case of Android Auto, the scope of the target of third parties is admittedly perhaps narrower as compatible apps are expected to be, in a way, car-related, such as in this case with the ENEL charging app.
In such a context, it would be reasonable to presume that with the creation of this OS version, Google intended to open it up to such app developers; the only thing that would remain to be decided is when and how that should happen. These aspects relate, however, to the conditions and quality of access and not to the question of access as such and, therefore, do not require Bronner to be applied.
When applying this same logic to platforms outside the specific context of this case, caution is required to ensure that the added value of this approach does not implicitly make it impossible for platform entities to rely on Bronner.
The (Potentially) Bad: defining platforms as inherently open facilities in the context of Bronner
The potential concern associated with the approach of AG Medina stems from the lack of nuance that the opinion has when it comes to the discussion of platforms and the implications of their structure for the application of Bronner. The discussion of platforms’ open nature, showcased in the case of Android Auto, is formulated in general terms almost throughout.
This seemingly minor issue can have, however, significant consequences if adopted similarly by the CJEU. The reason for that is the platforms are inherently designed to function as intermediaries that bring together various groups of customers and capitalize on their interactions. Accordingly, all platforms are inherently designed to offer access to third parties. Without participation from these parties, the existence of such structures is simply not possible.
Consequently, if the matter of openness in the abstract is sufficient to determine that Bronner does not apply to a refusal to grant access to a respective platform, adopting the AG’s approach will eliminate the protection offered by this case law by default. Evidently, that cannot be the aim or the desired outcome of this case for developing the case law concerning the essential facility doctrine. Excluding
Therefore, preserving the added value of this new approach would require looking further into the matter of openness in each case based on its specific circumstances and applying it in a that is suitable for the character of the respective platform. Going back to the case of Android Auto and considering its open character and intended use, it would be reasonable to distinguish between car-related and non-car-related apps when it comes to mandating access.
Since Android Auto was arguably designed to support car-related apps, it would not be sensible to apply Bronner to refusals concerning such apps for the above reasons. When, however, the refusal concerns non-car related apps, such as a hotel room booking app, a text or photo editing app, or a retail marketplace app that have little to do with operating a car, then Bronner should arguably apply. In such cases, it cannot be argued the aim of the platform was always to attract such third parties and, therefore, that the acceptance of access can be presumed to have been granted by design. Consequently, an access remedy would interfere with the fundamental rights protected by Bronner as well as the entrepreneurial incentive to invest in the creation of the platform.
Therefore, making use of the approach offered by the AG Opinion in an optimal manner in the case of digital platforms would dictate accounting for their inherent open nature in tandem with their respective business model and functional purpose. Here, it is essential to note that such an approach should also be dynamic, as such aspects of the platform may change over time. When that occurs, such changes should also be used to re-evaluate the possibility of relying on Bronner when denying access on a case-by-case basis.
The Ugly: the objective justification for restricting access
The topic of objective justification has long been considered a theoretical rather than practical one, as the number of cases where such arguments were successful is extremely limited. Nevertheless, it remains, formally speaking, an inseparable aspect of the art. 102 TFEU assessment. Accordingly, as with many art. 102 TFEU cases in the past, this topic was also addressed in the AG Opinion of Android Auto. Although the AG considered that Bronner should not be applied in this case, the discussion on the objective justification constitutes an integral part of the refusal to supply the test of Bronner and, thus, is also relevant for such cases in the future.
The discussion on the objective justification, in this case, shows the challenge of ensuring that neither the dominant undertaking nor the party requesting access misuses their respective position. This results in a somewhat logical but, at the same time, problematic compromise. With respect to the party making the access request, it is required, according to the AG, that it makes a genuine request for access, thereby taking a proactive role in helping to facilitate this in the event it is granted. This is rightfully a sensible requirement to be introduced to prevent parties from making requests, which they can expect will be refused, solely to get the dominant undertaking involved in an abuse case under art. 102 TFEU. This is particularly so in cases where Bronner would not apply, thereby lowering the difficulty of finding an infringement, which could be problematic where the abuse is established based on potential rather than actual effects on competition.
Regarding the dominant undertaking, the AG accepts that the refusals due to technical and financial considerations are, in principle, legitimate. However, such arguments cannot serve the dominant undertaking in perpetuity. Restating this motive to justify a refusal to grant access within the ambit of Bronner or out of it will, at some point, fail to meet the proportionality requirement that forms part of the objective justification. Therefore, for refusals stemming from technical or financial reasons to be legitimate, the dominant undertaking must discuss these reasons with the party requesting access to seek a mutually acceptable solution.
Overall, until this point, the approach offered by the AG seems once more logical. However, matters become more complicated when indicating how such discussion and envisaged solution should materialize. According to the AG, when the refusal stems from technical or financial reasons, the dominant undertaking should provide the requesting party with an overview of the costs of resolving such matter and allow the latter to finance this. To ensure this is a feasible solution, the dominant undertaking should ensure that its calculations are reasonable and proportionate [para. 75].
On the face of it, these additional requirements seem intuitive, and if all the parties involved act in good faith, it may work quite well. In practice, however, this suggestion only shifts the potential of misconduct to a newly introduced factor down the line in the discussion on facilitating access. At this stage, the requesting party could persistently sustain that the costs provided by the dominant undertaking are unreasonable to keep its costs as low as possible. Conversely, the dominant undertaking could use this possibility to allow third parties to co-finance some of its developments. Admittedly, despite its flaws, the approach offered by the AG nevertheless offers a better way forward for current practice. It is hard to see what a more optimal solution would look like.
Conclusion
The case of Android Auto offers yet another opportunity for the CJEU to refine the applicability of the essential facility doctrine. If followed, the suggestions included in the AG Opinion would introduce a new, valuable way for assessing when Bronner should apply based on the specific characteristics of the respective facility. When applied, particularly in the context of platforms, such an approach should account for the (open) nature of the facility in tandem with its functional purpose and business model. In such cases, when the objective justification for refusal (within the ambit of Bronner or out of it) is assessed, the potentially disingenuous intentions of all the parties involved should be filtered out as much as possible to allow (financial or technical) barriers to access to be removed in good faith. It now remains to be seen whether and, if so, to what extent the CJEU will decide to follow the suggestions of AG Medina.
This blog post was first published on the author’s blog here.
________________________
To make sure you do not miss out on regular updates from the Kluwer Competition Law Blog, please subscribe here.
