On 28 February 2023, the European Commission (EC) sent Apple a new Statement of Objections (SO) ‘clarifying its concerns over App Store rules for music streaming providers’. Rather than a clarification, or an expansion of the previous SO, the new SO dropped one of the two objections—an unusual move, especially at this stage of the proceedings.
When a startup shuts down, as many of them do, founders are in the habit of writing up a post-mortem providing an analysis of what exactly went wrong so that other founders can avoid making the same mistakes. My idea is to provide a post-mortem for the objection dropped by the EC.
Let us start by tracing the life of the objection we’re mourning. When Apple launched Apple Music in 2015, and thus started competing with Spotify, antitrust concerns immediately arose. Spotify started accusing Apple of using the App Store as a weapon shortly after, and started lobbying for regulation. In March 2019, Spotify launched an antitrust complaint (or rather, campaign), which the EC followed up on with a formal investigation in June 2020.
The EC sent Apple an SO in April 2021, outlining its preliminary view that Apple abused its dominant position in app distribution, first, by imposing its own in-app payment mechanism on music streaming app developers (‘IAP obligation’), and second, by restricting app developers’ ability to inform users of alternative subscription methods (‘anti-steering obligation’). Now, the EC dropped the first objection.
In this piece, I speculate as to why the EC dropped its primary objection, in other words, I try to determine its cause of death (the post-mortem). Before doing so, I discuss what remains of the case (pay respect to ‘those left behind’), and after, I examine the relation with the Digital Markets Act (examine the deceased’s legacy).
1. What remains of the EC’s App Store case?
Remember that, according to App Store rules, music streaming apps qualify as ‘reader apps’ (as do other media apps, e.g., those providing video or e-books). Reader apps are allowed to let their users sign up on the web rather than via the App Store. This is great for developers (they avoid the IAP fee of 30%) and for users (developers pass on some of their savings). It’s not great for Apple (it loses its fee), which is why it prohibited developers from informing users about outside purchase options (the anti-steering obligation).
The Japan Fair Trade Commission investigated the anti-steering obligation for reader apps, but closed its investigation when Apple committed to allow reader app developers to link to their website for account management. In other words, Apple mostly dropped the anti-steering obligation, although restrictions remain. Developers have to request an ‘External Link Account Entitlement‘ and when linking to their website, they cannot include ‘language that includes the price of items available on the website’. This means developers still can’t promote cheaper offers outside of the App Store.
The change was a step forward. Consider Netflix. Like Spotify, it stopped supporting IAP to avoid the App Store fee years ago, which meant users could no longer subscribe via their iOS app. But Netflix also couldn’t direct users to their website. So when a non-subscriber opened the app, Netflix showed them this awkward phrase (doing little to hide its dissatisfaction): ‘You can’t sign up for Netflix in the app. We know it’s a hassle.’ After the JFTC’s intervention, however, the Netflix app is now redirecting potential subscribers to its website to sign up.
Developers of music streaming apps like Spotify can do the same. It’s still not ideal—developers would certainly like to include a message saying: ‘Go sign up on our website, where a subscription is X% cheaper’. But it means that, with its remaining case, the EC can at best hope to improve the already-allowed steering process, in particular by letting developers advertise prices.
In short, given previous developments around anti-steering, dropping the IAP objection means not all that much of the EC case is left.
2. Why did the EC drop the IAP objection?
The big question is why the EC dropped its primary objection. In the press release, the EC limits itself to saying it ‘no longer take[s] a position as to the legality of the IAP obligation’ and the final decision might not elaborate much further. Therefore, I must speculate.
A first option is that the EC simply found that the IAP obligation did not result in any anticompetitive effects. But that would be curious after a first SO that must have detailed such effects. And indeed, it is not too difficult to envisage anticompetitive effects. The EC chose the right case: music streaming apps have tight margins given that they owe most of their revenue (say, 70%) to record labels, so there may simply not be 30% margin left to pay the IAP fee. And when Apple does not subject its own music streaming app to such a fee, competition may be distorted. (For a more in-depth assessment, see the article I co-authored with Daniel Mândrescu).
A second option is that the EC did see a risk of anticompetitive effects but could not build a legal case. This doesn’t sound right either. In its new SO, the EC qualifies the anti-steering obligation as an unfair trading condition; it could have done the same for the IAP obligation—indeed, that’s what the Dutch Authority for Consumers and Markets (ACM) did. Given the aforementioned margin issue, the abuse of margin squeeze might’ve also fitted the conduct (as I’ve argued elsewhere). And those are only some of the options.
A third option is that, in response to the EC’s SO, Apple put forward a compelling justification for its IAP obligation. This option makes most procedural sense, but does Apple have such a justification? A look at the U.S. Epic v Apple judgment is instructive. Apple argued that IAP provides security, including fraud protection. In particular, centralizing payments allows for the detection of fraud patterns. But IAP is not the largest in-app payment service, as third-party payment processors can be used for certain transactions (of physical rather than digital goods). If scale is important, those third parties are actually better placed. Therefore, the judge was not convinced of Apple’s justification.
A fourth option is that remedies would be ineffective. Indeed, the ACM’s decision, which obliged Apple to do away with its IAP obligation for dating apps in the Netherlands, didn’t move the needle. IAP is, after all, just a convenient fee-levying mechanism. After a series of periodic penalty payments, Apple did away with IAP, but still charged a 27% commission (the original 30% minus 3% for payment processing). Competition authorities should consider remedies when starting a case, but some difficulty in implementing remedies shouldn’t deter them either (that would give dominant firms the wrong incentives).
At first sight, there does not seem to be a good (enough) reason to drop the objection. Of course, I do not have access to the EC’s detailed assessment, which likely provides such a reason. We can only hope the EC shares a significant enough part of that assessment.
3. How does the Digital Markets Act (DMA) factor into this?
You could say: ‘All of this doesn’t matter, the DMA prohibits Apple’s IAP obligation anyway.’ While that is true (see Article 5(7)), that approach seems counterproductive. This is a difficult case, but that is why it is all the more important to have a precedent (or, if the conduct is not anticompetitive/justified, a full explanation). The DMA is, after all, a fairly static instrument. Competition law precedent better withstands the test of time, and is thus more useful for future cases.
Is the EC getting lazy then, pausing enforcement until it can apply the DMA? That doesn’t seem to be the case either. After all, the anti-steering obligation is also prohibited by the DMA (see Article 5(4)–(5)) and the EC is still pursuing that objection. When it comes to the IAP obligation, however, the DMA has an added benefit compared to competition law enforcement: it not only prohibits IAP obligations but also mandates FRAND app store fees (Article 6(12)), which can avoid a situation like in the Netherlands.
Finally, what about the idea—promoted also by the EC—that the DMA is inspired by its experience enforcing competition law? There are two readings here. On the one hand, IAP obligations are apparently not a problem under competition law, so why is the DMA prohibiting them? On the other hand, the DMA makes much of the fact that it does not pursue undistorted competition but fairness and contestability. It has been criticized for not clearly distinguishing those goals, but this case might demonstrate the difference: while the IAP obligation apparently does not distort competition, it is considered unfair.
This blog post also appeared as an op-ed on EU Law Live.