After almost three years since the Commission sent Apple its statement of objections, which was significantly trimmed down, the Commission reached a finding of abuse for which it imposed a whopping fine of 1.8 billion euros. Alongside this case, Apple was also involved in an almost identical case running parallel in the Netherlands, with similar findings. Meanwhile, during these procedures, the Digital Markets Act, which covers the main constraints in both cases, entered into force and is expected to change the AppStore terms significantly. The interplay between these three developments delivers an outcome that displays the great added value that the DMA can bring to complex cases, the limits of strategic enforcement, and the potential consequences of overstepping them. So, while there is no doubt that the Apple App Store case will become another hallmark case, it will be one, or better yet, the first, in a category of its own.
Evolution of the case
The investigation of the Commission in the case of the Apple App Store started with a blog post by Spotify’s CEO on Apple’s unfair treatment, which has now turned into a dedicated website. Eventually, the post became a complaint to the Commission, which launched its investigation in 2020. The initial investigation was intended to look into Apple’s App Store terms and conditions for music streaming apps and ebook distribution apps covering (i) the mandatory use of the In-App Purchase (IAP) for several categories of apps offered in the App Store for which they pay a 30% transaction fee; (ii) the prohibition imposed on app developers to inform consumers of the possibility to register and /or make purchases of the app services outside the of the respective app that were often cheaper since these were not subject to the App Store transaction fee (‘the anti-steering prohibition’).
The concerns raised by the App Store terms was that it essentially forced Apple’s competitors in vertically related markets to choose between raising their prices or removing the payment processing option from their app altogether. In the case of the IAP, it was also noted that it de facto dis-intermediated app developers from essential customer data, which was only accessible to Apple as the system owner. Over time, however, the scope of the investigation became much narrower. When the first statement of objections was sent out, almost a year after the start of the investigation, it was focused solely on music streaming apps. There was no mention of ebook distribution apps anymore. One may wonder why, however, given that the identified terms of the App Store are identical for both categories of apps (and others). Fast forward almost two years, the Commission clarified the initial statement of objections, further narrowing down the case’s scope. Based on this clarification by the Commission, it was no longer concerned with the mandatory use of the IAP. Instead, the Commission focused only on the anti-steering prohibition, which it considered to be a form of unfair trading conditions as it was neither necessary nor proportionate. It was found to be detrimental to consumers and negatively affected music streaming app developers. Approximately a year after this trim-down, its final decision focused solely on the anti-steering prohibition that was found to constitute an abuse of dominance under Article 102(a) TFEU, resulting in a fine of 1.8 billion euro. While the case of the Apple App Store is certainly one of great importance in the context of digital markets in general and mobile ecosystems in particular, the overall approach and outcome raise some questions concerning its coherence and effectiveness.
Market Definition
As with any abuse of dominance case, the starting point of applying Article 102 TFEU is the market definition that is required for establishing dominance. In this case, the market definition entailed a more challenging exercise due to the AppStore’s two or multi-sided nature, which required simultaneously assessing demand substitution from multiple customer groups. Academic literature and practice would indicate that the way to approach this matter comes down to determining whether to define a single relevant market for the platform (in the case of the App Store) or defining separate (yet interrelated) markets for each of the customer groups (or sides) of the App Store. This approach, also known as the single and multi-market approach, has recently been officially acknowledged by the Commission in its new Notice on the definition of the relevant market. On the face of it, it would appear that the Commission did exactly what the Notice indicates. The finding that Apple abused its dominance in the market for the distribution of music streaming apps through the App Store indicates that the Commission most likely opted for a single relevant market within the realm of iOS. As the App Store is the only distribution channel in this context, it is unsurprising that dominance was established.
Nevertheless, finding a dedicated distribution market for a specific kind of app (music streaming) is not entirely intuitive. As iOS is a closed system for app distribution, why should separate markets be defined for separate app categories? After all, all apps offering paid content are in the same boat when it comes to exclusive distribution via the App Store and the anti-steering prohibition. Accordingly, the position of dominance should be the same across all such apps. Creating a dedicated sub-segment for music streaming apps gives the impression that Apple’s dominant position on app distribution may not extend over all apps offered in the App Store, which is quite unlikely.
This could only happen if (a) different app categories have different substitution possibilities for iOS apps and/or (b) the app distribution services of the AppStore and Play Store are considered substitutes for some categories of apps. The first condition could not have been met at the time of investigation, and the second condition is also unlikely to have been met as iOS users (i.e., consumers) rarely switch to Android devices rather than substitutes for developers. If these two criteria are not met, one may wonder how this approach fits in with the single and multi-market distinction indicated in the Commission Notice. Answering this question is relevant for the App Store and all cases concerning platforms that bring together consumers and a large heterogeneous group of business users, such as sellers on a marketplace or accommodation owners on a booking platform. Hopefully, the official decision document of the Commission will shed more light on this matter, mainly since it is not only the Commission that chose to approach the market definition but also the Dutch competition authority (ACM), as will be discussed.
Theory of harm and remedies
According to the Commission’s press release, the theory of harm in this case is relatively straightforward. Since Apple does not allow app developers to inform users (i.e., consumers) of alternative payment routes outside their respective apps, such users end up making use of the payment options within the app, which are more expensive due to the commission fee imposed by Apple that is often passed on (in whole or in part) to them. Where app developers chose not to include a payment or subscription possibility, consumers were harmed, according to the Commission, because they had to engage in a more cumbersome search for subscription options or they never subscribed to the respective services at all because they could not find such options. Overall, the theory of harm is not complicated to follow and is quite credible. However, its corresponding remedy, if not further specified, is quite unlikely to be effective when looking at the ongoing parallel case in the Netherlands.
According to the Commission, the harm caused to consumers in this case is directly related to app developers being prevented from informing them of more affordable avenues for their subscriptions. Consequently, it is quite sensible that the remedy imposed requires Apple to remove the anti-steering prohibition from its terms and conditions. Requiring Apple to remove this obligation does not mean the solution will lead to more affordable options for consumers. This is because the remedy does not introduce any constraints for the price setting of the App Store that inherently impacts what consumers end up paying. Accordingly, if Apple is only required to abandon the anti-steering prohibition and nothing more, this results in a rather superficial solution that does not change much in practice. This is not a purely speculative remark, but a realistic and likely outcome displayed by the parallel case running in the Netherlands.
Unlike the Commission, the ACM’s case concerned the obligatory use of the IAP and the anti-steering provision, which Apple was required to abandon. Nevertheless, in this case, nothing was (formally) required with respect to the App Store price setting rules. The outcome was, unsurprisingly, not very effective. Apple now allows developers to choose one of the following options: i) continue using Apple’s IAP; ii) use a third-party payment system within the app; iii) include an in-app link directing users to the developer’s website to complete a purchase; or iv) use a third-party payment system within the app and include a link directing users to the developer’s website to complete a purchase. For app developers using options ii-iv, Apple will reduce its commission fee by 3%.
This outcome, which arguably complies with the requirement to abandon the harmful restrictions of the App Store, at least formally speaking, can hardly be said to be effective. In practice, it corners app developers and forces them to make a difficult switching choice for a negligible reduction in fees, which would become even lower once the transaction fees of third-party solutions are calculated into it. If anything, it may make things even more expensive. Until now, the lucky few consumers who found their way to alternative payment /subscription routes could (at least theoretically) have escaped the fee altogether, thereby requiring Apple to introduce the anti-steering obligation. With this new options menu, it would appear that this prospect is replaced with a choice between a fee of 30% and 27% (excluding additional third-party costs) being passed on in whole or in part to consumers. While there were some indications that the ACM may not let the case end in this manner and intervene in Apple’s App Store pricing, it is clear that not tackling such matters within the ambit of the theory of harm can have significant consequences.
Coming back to the case of the App Store before the Commission, a similar outcome can also be expected because the obligation to remove the anti-steering prohibition does not mean that Apple is also required to continue allowing the out-of-app purchases to be commission-free. Suppose the formal decision of the Commission does not include a more detailed analysis of harm and remedy prescription. In that situation, the case will not be able to address the identified harm to consumers in this case adequately. Ironically, it may even increase it, as by now, due to the media coverage of this case, many consumers have learned that they are better off signing up for (music) subscription services outside of their iOS app (when possible).
In this sense, the case offers an excellent example of how narrowing down and simplifying a theory of harm for the purpose of ‘winning’ a case can have detrimental consequences for the remedy stage of the case. An (over)simplified theory of harm leads to a significantly compromised corresponding remedy. In this specific case, the simplified theory is also, to some extent, regrettable due to the similarities it creates with the case against Apple at the national level. Fortunately, in this case, the DMA can serve as a safety net, which makes this shortcoming, or perhaps deliberate strategic choice, worthwhile, displaying very well the added value of having both ex-post and ex-ante frameworks working together.
The DMA and its relevance for the Apple AppStore case
On the 5th of September 2023, Apple was designated as a gatekeeper under the DMA with respect to its AppStore. Although the scope of the gatekeeper status is currently challenged in appeal before the General Court by Apple, which contends that it should not extend to iPads, the designation with regard to iPhones will remain unchanged. Accordingly, as of 7 March, multiple obligations must be accounted for in the AppStore’s terms and conditions. Several of these obligations tackle precisely what the Commission’s case under Article 102 TFEU left out. Some of these may even go further than what would have been attainable within the ambit of this case.
The obligation to use IAP for app developers, which was dropped at an earlier stage of the case, is now covered by Article 5(7) of the DMA, which seems to prohibit it. Similarly, the anti-steering prohibition is covered by Article 5(4) of the DMA. Interestingly, this latter provision states that such a possibility should be provided for business users (i.e., app developers) free of charge. The question is, what does that entail in the context of the AppStore? Would this allow app developers to steer consumers away from their app to complete the respective transaction free of commission? If that is the case, one can expect that Apple would have to re-adjust its AppStore fee for all developers as otherwise it would be facing a non-negligible free-rider risk.
Beyond these two provisions, there are a few additional ones that would contribute to opening up Apple’s closed system for more freedom of choice for both consumers and app developers. To start with art. 5(3) of the DMA, which prohibits price and terms parity clauses, will ensure that app developers can offer more attractive offers outside their respective apps offered through the AppStore. Furthermore, art. 5(5) of the DMA will ensure that subscriptions and services acquired outside of apps distributed in the AppStore can nevertheless be used with such apps and on iOS in general. Beyond these obligations that directly address the concerns initially identified in the App Store case at the national and European level, Articles 6(4) and 6(12) of the DMA will significantly change the current circumstances concerning app distribution, by obligating Apple to allow third-party app stores to be developed for iOS. Finally, Article 6 (7) of the DMA requires that Apple will provide full interoperability to third-party (services and hardware) providers, thereby ensuring the app developers that do make use of the new possibilities created by the DMA are not ‘punished’ for it. Therefore, these provisions will help create new options for app developers if the internal changes within the AppStore do not lead to a desirable outcome involving lower fees and more access to customers’ financial data.
The cumulative effect of these provisions will bring more change in practice than would ever have been possible through a single competition law remedy, regardless of the theory of harm brought forward. Furthermore, addressing these matters through the DMA route from the Commission’s perspective is that no anti-competitive harm must be proved for enforcement to occur. When looking at the public (non-confidential) version of Apple’s DMA compliance report, it appears that Apple is on its way to implementing all these obligations, at least formally. Its approach to the communication of the various changes it will introduce with the DMA is, however, rather odd. The published non-confidential summary report is extremely limited in detail, even by Apple’s standards. By comparison, Apple’s website for developers strangely includes far more detailed information. This mode of communication defeats the point of such reports, which is to allow third parties to evaluate Apple’s compliance with the DMA. In the case of Apple, it is clear that its website provides better information for this purpose than the dedicated DMA report, which should not become the standard for the future. Moving beyond this shortcoming and zooming into the changes adopted by Apple, it is clear that some of these are pretty significant if applied in good faith. Apple will accept third-party app stores and direct downloads from app developers’ websites. Both options are, however, subject to relatively stringent criteria, especially for direct downloads where developers must first be ‘worthy’ of this option. This requires, among other things, having a large customer base and a good contractual relationship with Apple for the past two years.
Regarding the payment processing option, the sub-optimal option implemented in the Netherlands has become the default. This means that app developers who wish to distribute their apps through the AppStore can choose between i) continuing using Apple’s IAP, ii) using a third-party payment system within the app, iii) include an in-app link directing users to the developer’s website to complete a purchase; or iv) use a third-party payment system within the app and include a link directing users to the developer’s website to complete a purchase. The possibility of bringing IAP into the mix is NOT allowed. Similarly, the fees for app developers using options ii-iv will be reduced by 3%. Therefore, in the context of the App Store, the same concerns as in the Netherlands will continue to exist under the scope of the DMA in principle. Nevertheless, the gravity of such concerns may be somewhat reduced because Apple appears to have introduced lower fees across the board. Although this will not make third-party payment solutions more attractive for developers, it may make doing business with Apple less expensive for app developers.
Outside of the scope of the App Store, these new adjustments also bring along new types of fees. Third-party app stores and directly downloaded apps will be subject to a yearly download fee of 0.50 euros, labeled as the Core Technology Fee (CTF). Considering the potential number of iOS users, this modest fee can translate into (hundreds of) millions of euros for Apple. Arguably, this is quite a fair compromise as it would be unreasonable to request Apple to manage a large part of the technical side of such apps and app stores at no cost. Beyond this core fee, it seems that it is up to the respective apps and app stores to set up their own fee structure to be as competitive as possible.
Once the Commission and third parties offer their views, it will become clear how much the DMA can help bring this case to a desirable end. This is not only for the new fees but, more importantly, for the technical process put in place to make these options feasible and viable. For the time being, it can be said that the DMA will likely fulfill its role of complementing competition law enforcement quite well. Nevertheless, the DMA cannot help getting rid of the (two) elephant(s) in the room, namely having almost identical cases running at the Commission and national level and, of course, the calculation of the fine in the case before the Commission.
Overlaps with the case against the App Store in the Netherlands
While the synergy between the DMA and the AppStore case before the Commission can be said to be well-calibrated, the same cannot be said with respect to the case in the Netherlands. This is partly due to the Commission narrowing down and changing the theory of harm. When both cases were initiated, it appeared as though the idea behind the two cases was that the case before the Commission would deal with the potential exclusionary effects of the App Store rules. In contrast, the case in the Netherlands would concern the potentially exploitative effects of these rules, on which the Commission traditionally does not focus. This division of concerns and the fact that both cases were focused on different relevant markets helped prevent any procedural concerns about jurisdiction and allocation of cases under Regulation 1/2003 from arising.
The case in the Netherlands proceeded as planned in this respect. Although the scope of the case was also narrowed in its initial phase with respect to the relevant markets covered by it, the core claim remained the same, namely that the combination of the IAP obligation and anti-steering prohibition constituted a form of unfair trading conditions to the detriment of (dating) app providers. In the case before the Commission, the core of the case appears to have changed significantly. Where the case appeared to tackle the potential foreclosure effects of the IAP obligation and the anti-steering prohibition for (music streaming) app providers, it turned out to revolve around exploitative effects that the anti-steering prohibition may have on consumers.
The foreclosure angle to the case seems to have taken the back seat and is only marginally mentioned in the press release announcing the finding of abuse. If that is indeed the focus of the case, one may wonder what the logic of these two cases is, having run in parallel. Tackling two perspectives (consumers and app developers) of the same problem is hardly a good reason, particularly in a case that concerns a tightly closed platform like the AppStore, where any harm arising on one side of the platform is visible and passed on to the other. The same is also true regarding the different relevant markets tackled in these cases, which, as mentioned above, is a bit questionable given that all kinds of (paid) apps appear to experience the same constraints when it comes to the distribution and capitalization of iOS apps.
Although the above is no consequence for the case before the Commission to the primacy that the Commission has when opening investigations, the outcome of almost identical parallel procedures is, at the very least, not optimal when it comes to using the limited resources of the Commission and NCAs. This is even more so in this case, where the DMA will complement such efforts to a great extent. In such instances, the principal added value of a competition law procedure is to ensure that anti-competitive behaviour that took place in the past is penalized and that the door for follow-on claims for recuperating the damages caused by such actions is opened. However, for that purpose, one well-constructed case should generally be sufficient. Therefore, in future cases, where the facts presented before the NCA and Commission do not significantly differ, one could argue that the better route would be for the Commission to take on the entire issue, also covering the various national angles that may come into play.
The fine and its (unprecedented) calculation
Like in previous big tech cases, the fine in the case of the App Store did not disappoint when it came to making the news. While for most of the public, it would commonly be the amount that draws one’s attention, for most competition law practitioners (and academics), it is its composition that is most remarkable. Although no concrete figures have been published, the press release and additional comments by Vestager emphasize that the fine was set at that level for deterrence purposes. According to the Commission guidelines on fines from 2006, the Commission can increase the fine by an undefined amount to ensure sufficient deterrence in cases where the undertaking concerned has a significant turnover beyond the sales to which the infringement relates. This is undoubtedly the case with Apple, as its overall turnover from the App Store significantly exceeds its sales from the music streaming app subscriptions. Given that the basic amount of fines relates to a proportion of the value of sales affected by the infringement (generally being 30%), it would appear that the jump to 1.8 billion euros is quite steep.
When looking at the entire context of the App Store, the amount of the fine can arguably be justified since the harm identified in this case is most likely to be similar for almost all types of paid apps subject to these same conditions. However, the narrowing down of the case has the unfortunate effect of making the fine appear disproportionate because the deterrence hike is likely to be significant. If, by contrast, the case had looked more broadly at the anti-steering obligation for all paid apps subject to it, this impression would have had no basis. Gaming apps alone, for example, would bring sufficient revenue to justify such a fine. Accordingly, while framing the case may have given the Commission the upper hand regarding the chances of winning the case, it does present some difficulties in setting the fine. In this respect, one may even consider the synergy with the DMA slightly problematic. If the DMA essentially regulates the problematic behavior in this case, what elements of deterrence remain to be covered by the case? Case-specific deterrence can arguably be said to have been handled by the regulatory framework of the DMA, as re-introducing the anti-steering provision would constitute a direct infringement of it. What remains then is a more general notion of deterrence intended to prevent anti-competitive practices, and perhaps more specifically abuses of dominance, from being put into practice by Apple and other undertakings. Both readings can be done based on the wording of the fining guidelines, and it is not entirely evident that deterrence-motivated increases in fines, as significant as in this case, can be justified under such circumstances. It remains to be seen what the EU courts will decide on this matter, as Apple already made it public that it will appeal the decision. Given the probable (significant) discrepancy between the value of sales in the relevant market covered by the case of the final amount of the fine, it would not be a surprise if the fine would be reduced. This is not because the Commission got the facts wrong but due to the substantive (strategic) choices made during the case that bring along corresponding procedural constraints. In other words, the Commission can’t have the cake and eat it, too, or can it?
Final thoughts
The Apple App Store case will undoubtedly become one of the hallmark cases of competition law in digital markets. Unlike the previous cases of Microsoft, Google, Amazon, and Facebook, the case does not present us with much substantive novelty other than perhaps an unusual approach to the market definition. Instead, this case is unique because it clearly displays the potential added value of the synergies between competition law enforcement and the DMA. Competition law enforcement is used to penalize past anti-competitive behaviour while also ensuring deterrence. The DMA complements this process by regulating the respective practices in a manner that goes beyond what EU competition law can offer. At the same time, it may also show the implications of constructing competition law cases around narrow theories of harm. Building up cases that are easier to ‘win’ may mean compromising on the fine that can be imposed, as such, on deterrence.