Apple’s EU app rules comply with Digital Markets Act unless judges take purposive interpretation to higher level

Shortly after Apple announced its new EU app rules in preparation for the entry into full force of the DMA, games fray explained various issues:

• the direct installation of apps (other than alternative app stores) will not be possible (contrary to widespread misreporting and misleading headlines) (January 26, 2024 games fray article);
• Apple knows that its last line of defense will be the claim that its terms are merely a fair, reasonable and non-discriminatory (FRAND) royalty for intellectual property and other valuable assets, a question that can keep courts busy for many years with an uncertain outcome (another January 26, 2024 games fray article); and
• the seemingly low 50-eurocent per-install per-year fee means that no one who makes a free app (such as an e-commerce, customer service or advertising-sponsored app) will want to use an alternative app store (given that Apple will discriminatorily carry those apps for free) and even most apps with in-app purchasing, apart from maybe adult content (which Apple doesn’t allow on its App Store), stand nothing to gain from using alternative app stores as there would be hardly any saving of fees but users would have to uninstall and reinstall an app to get it from another store (January 28, 2024 games fray article).

Now let’s look at it from the angle of what could go right (in terms of creating effective competition in app distribution), and how likely that positive scenario is to play out. It’s not impossible that Apple’s gamble will turn out to have been a mistake, but the odds are fairly long against those challenging Apple’s EU app rules on the basis of the DMA in its current form. With a few limited changes such as a Most Favored Nation clause (as discussed in a previously-mentioned article), an anti-disruption clause (that would prohibit Apple from requiring app makers to offer different apps in the EU than elsewhere only to benefit from the freedoms afforded to them by the DMA) and a different mechanism for FRAND determinations, the law could achieve its objectives with a reasonably high degree of certainty. There will be a follow-up article in the not too distant future in which games fray will make and explain such proposals.

For now, there definitely is a lot of blowback in Apple’s face, such as from large players like Meta (February 2, 2024 games fray article) and Microsoft (January 30, 2024 games fray article). The Coalition for App Fairness (CAF) is building a library of short TikTok-format videos in which politicians, advocacy groups and some smaller companies demand that the DMA be vigorously enforced now. The politicians speaking out there were in favor of the DMA during the legislative process and still claim that it is a game-changer law that safeguards competition. But they won’t decide on how the law gets applied and interpreted now. What they say will be taken note of by the European Commission, and will be ignored by the EU judiciary (separation of powers).

Not a lot of legal talent available in Brussels to go against Apple when Google has largely similar interests

There are signs, even in LinkedIn comments by antitrust lawyers, of Apple potentially facing a mix of public (EC) and private (companies and trade groups) enforcement. Even with “only” 40 people in the digital markets branch of the EC’s Directorate-General for Competition (DG COMP), the EC is in a position to tackle at least the number one priority. If there is one company they must view as the most urgent target, it’s Apple because of its arrogant and belligerent attitude. The EC simply can’t condone all that Apple is doing here because it makes mockery of the EU’s regulatory powers.

Those 40 EC officials won’t be alone. There are major private interests involved here, and companies who want to support DG COMP can do a lot of the grunt work, feeding the EC with studies and legal analysis. There may soon be multiple external experts (lawyers, economists, technical experts) working on this per each EC official in charge of DMA enforcement. Of course, Apple will also use an army of lawyers (or an army of armies of lawyers). And there’s a major problem: it’s hard to find good competition law firms in Brussels that haven’t done work for either Apple or Google, which in some cases is indirect (such as representing astroturfing organizations like ACT | The App(le) Association or Google’s Developer Alliance).

Even though Apple and Google compete, and some of Apple’s actions (such as the destruction of PWAs) are not in Google’s interest, the two mobile duopolists are more aligned than divided when it comes to app distribution monopolies. If Google surprisingly altered course (which would make sense in light of the outcome of the Epic Games v. Google trial in the U.S., and Google did something similar on standard-essential patents after it settled all cases with Apple and Microsoft), then there would be access to additional legal talent such as Baker Botts’ antitrust lawyers (pro-Google, not pro-Apple).

The relatively strongest European antitrust law firm (though it is “only” a boutique, albeit one that has successfully expanded already to the UK (with former Competition & Markets Authority officials as partners) and set up shop in the Netherlands) that can go against Apple, Google and Amazon is Geradin Partners.

No conflict concerning Apple, Google and Amazon also applies to Hausfeld. In fact, they can go against any gatekeeper. In the U.S. they’re primarily a class-action firm, and Apple would try to hold against them the terms on which they settled a Google Play Store case there, though European judges don’t care about that (formally, U.S. judges don’t either, but in the U.S. the credibility of counsel is psychologically more important). In the EU, Hausfeld is more than an antitrust damages and class action firm. They’ve done great work, such as in a German case against Google that forced the German government and Google to abandon a health information-related partnership.

Then there’s an even larger firm, Clifford Chance, which routinely goes against Apple and Google, but then uses the experience from those cases to defend Amazon. Also, Clifford brought EU antitrust complaints on Spotify’s behalf (where Apple fended off 95% of the complaint in terms of economic and strategic importance and Clifford may now get a 5% win) as well as Epic’s (a complaint of which one never heard again after it was filed three years ago). There’s also the European Super League soccer breakaway league case, where Clifford can say that the press department of the ECJ declared its client the winner, though it was simply an incorrect wording as the underlying fact pattern was outdated (December 21, 2023 LinkedIn article; see analysis by Norton Rose Fulbright: “Owing to UEFA having amended its rules in 2022, [the ECJ case was] only [about] whether the old UEFA system complied with EU competition law.”).

So, Clifford Chance is large, but suboptimal, Hausfeld in the EU is in a stronger position for this kind of work than the firm would be in the U.S., and Geradin Partners is free from conflicts of interest and has great legal talent on board, but Apple’s lawyers will outnumber everyone else’s team.

Parties seeking to support the EC’s enforcement efforts and/or contemplating their own action should look beyond Brussels. In fact, Paris is a short train ride away. A major business hub and regulatory venue. Dazi Law is also free from conflicts concerning the gatekeepers, and the firm’s founder, Fayrouze Masmi-Dazi, has done some interesting work particularly against Apple that games fray is aware of.

Literal vs. teleological/purposive interpretation

What Apple’s lawyers did when they developed the company’s response to the DMA with executives and engineers was to interpret the statutes literally and narrowly. Two simple examples:

• The DMA says they must allow rival browsers to come with their own browser engines, which means that browsers like Chrome (Google) and Edge (Microsoft) could become much more functional on iOS and enable web apps to deliver greater performance and a far better user experience than before. And users can choose a default browser. But the DMA does not say that Progressive Web Apps (pseudo-apps that actually just open a website, but can then run in full-screen mode and give users an app-like experience if the browser is better than Safari) must continue to run, if the user made that choice, in a rival browser. So Apple just doesn’t let PWAs run in alternative browsers anymore, and even degrades their user experience in Safari (as Apple knows that with rival browsers it may have to improve Safari, for general web surfing, at a far faster pace than in the past, where Apple wanted to ensure web apps wouldn’t really be able to compete with native apps).
• The DMA says they must allow the installation of apps outside their App Store (a core platform service in DMA lingo) and they must allow alternative app stores. But the DMA does not say explicitly that any app must be installable straight from the web (as opposed to some app store, be it Apple’s or a third-party store). Instead of allowing both alternative app stores and “sideloading” (a pejorative term for direct installation from the web), Apple interprets the statute as being complied with by simply allowing the direct installation of the frontend apps of alternative app stores (but not of other apps).

What Apple interprets very expansively, however, are references in the DMA to its right to be compensated on FRAND terms and also its rights to safeguard privacy and security.

The EC now has to develop its own interpretation, which will be court-reviewable. Private parties can go to court directly. They will probably try to find out first whether the EC is going to take action that they’d rather throw their weight behind than turn this into a Meta v. Apple or Microsoft v. Apple dispute. But things may take quite long with the EC, and it’s far from certain that the EC will take meaningful action. Also, if the EC is the actual litigant and others are merely amici curiae, there’s always the risk of the EC settling for a suboptimal result.

Let’s assume now that someone, be it the EC, private parties or a combination of both, has to litigate against Apple. If the EC makes a decision Apple doesn’t like, Apple will appeal, while private parties have to go to court and seek an injunction against Apple to change market dynamics. Damages claims could also help, but the key battlefield would be injunctive relief.

Under a literal interpretation of the DMA, and with an in dubio pro reo (in case of doubt, decide for defendant) and nulla poena sine lege (no penalty without a law) approach, Apple would come away either unscathed or, more likely, would be forced to make some minor adjustments and the fees it seeks to collect from alternative app stores as well as from app makers who wish to offer their apps through alternative stores would ultimately be determined by the courts of law (FRAND determination). If it all comes down to literal interpretation, nothing will change in the marketplace for years, and even thereafter, Apple will have defended a sufficient part of its new EU app distribution rules that the App Store is going to remain, more or less, the monopoly that it is today.

It’s important to consider that the DMA is a sui generis type of regulation and not an amendment to antitrust law. Simply put, the DMA doesn’t say that if someone has gatekeeper power, there’s a reversal of the burden of proof (presumed guilty until proven innocent), or that any abuse of a major imbalance in bargaining power (a concept that is already and increasingly popular in some Asian jurisdictions) constitutes an abuse of market power even if it didn’t under traditional antitrust law. The DMA is what some call an ex ante (before market power and a conventional abuse thereof are proven) regulation. If one interprets a regulation literally, then it’s like, say, traffic rules. If a car crosses a street despite the traffic light being red, that’s a clear violation; but if the rules don’t say you’re not allowed to use a smartphone while driving, you won’t be fined unless and until they adapt the rules.

What is the EU’s judicial philosophy going to be? Given that the DMA is a novel kind of regulation, there is considerable risk for both sides, but for Apple it’s worth taking that risk and that’s why they’ve unmistakably decided to duke it out in court.

The key to any impactful ruling against Apple in the DMA context will be a purposive interpretation, or a teleological one. To some, the two are synonyms, while others make the distinction that a purposive interpretation identifies the purpose of a given statute while a teleological one looks beyond the statute at hand and takes constitutional values and other laws into consideration.

Here, the DMA itself actually makes its purpose very clear. The recitals are extensive. For the purposes of this DMA discussion, we can assume (at least for the time being) that the result won’t differ (there’s no contradiction between the spirit of the DMA and the EU’s de facto constitution and other laws), and a teleological interpretation is ultimately also a purposive one.

Here’s an explanation of what effects a teleological interpretation would have. It’s from an Irish decision (HMIL Ltd v Minister for Agriculture and Food, 8 February 1996):

“Teleological interpretation is used for three purposes:
(1) to promote the objective for which the rule of law was made;
(2) to prevent unacceptable consequences to which a literal interpretation might lead, and
(3) to fill gaps which may otherwise exist in the legal order.”

The objective of the DMA is clearly to open markets, while Apple seeks to uphold its walled garden to almost the full extent regardless of the DMA. The consequences of a literal interpretation are obvious in this case: it’s a “one step up and two steps back” game where Apple will literally comply with one rule but then come up with one or more anticompetitive measures to minimize the effect. And there arguably are gaps, though the DMA is fairly detailed. And that is the problem: a teleological or purposive interpretation is not the starting point of analysis. The very first question is whether the statute is clear enough in its own right and so specific that it must be interpreted literally. And the more detailed a law or regulation is, the harder it is to argue that there are gaps or loopholes.

The ECJ sometimes does interpret statutes purposively. Its current president, Professor Koen Lenaerts, whose term as a judge has almost four years left (press release), authored a paper in 2007 (PDF) in which he said “[t]he teleological approach commands a prominent place in the Court of Justice’s interpretative methods.” However, the focus there was on how to interpret the EU Treaty, which is rather abstract and general, while the DMA is a highly specific regulation for a few types of technology markets.

The following passage from his paper shows what place teleological interpretation has in EU jurisprudence, and indicates that President Lenaerts considers “teleological” to be synonymous with “functional” and “purposive”:

“The Court of Justice utilizes three primary methods of interpretation. First, literal interpretation looks to the text and wording of the law. Second, systematic interpretation denotes reference to the context of the law, for example, to its historical background and to its place in the system of the Treaties. Third, teleological interpretation, also called functional or purposive interpretation, constitutes the method by which the Court engages in the process of interpreting provisions of Community law by choosing ‘the interpretation which best serves the purpose for which the provision was made.’”

In a 2016 ruling, the ECJ held those placing hyperlinks to copyright-infringing third-party websites potentially liable. The EU Copyright Directive didn’t say so, and the Advocate General thought the language of that EU law was clear enough that a copyright holder would have to sue the actual infringer, not someone who just places a link. Paragraphs 30 and 31 reflect a strongly teleological reasoning:

With that kind of approach, Apple could be in trouble, but it was just about the interpretation of one term. The question in that copyright case was whether a hyperlink is a “communication to the public” like an advertisement.

It’s possible that “sideloading” could be required and Apple’s narrow interpretation (that alternative app stores are sufficient to comply with the statute) could be rejected. That would be somewhat similar to the question in the copyright case: a purposive interpretation would mean “in case of doubt, rule in favor of freedom and choice.” But even then, Apple would point to the fact that the DMA gives it the right to protect privacy and security.

The problem with Apple’s EU app rules is not so much that a single one of those rules alone defeats the purpose of the DMA. Some rules are more impactful than others, but in the end it’s about the overall set of rules. It’s about the aggregate effect of a complex web of rules. It’s a thicket of rules that Apple had promulgated before and the new EU-specific ones, and it involves some technical measures such as preventing PWAs from running in alternative browsers or, even if in Safari, in full-screen mode.

From a rule-of-law point of view, that poses a major challenge. If a public enforcer or private party claims that the combination of certain Apple rules thwarts the DMA, the courts of law will want to know what particular Apple rule(s) they should declare illegal (and, as a direct result, unenforceable). If it’s the combination of three or four issues that prevents competition in iOS app distribution from emerging, does that mean that every one of those rules is illegal? Maybe a particular rule would be acceptable in isolation, or maybe in combination with other (softer) rules.

That’s why it was mentioned further above that the DMA is not a modification of antitrust law. It’s a new regulation. If it was about lower hurdles and fast-track proceedings under EU antitrust law, the EC and the courts would find it easier to deem a thicket of rules anticompetitive than in connection with an industry-specific, detailed regulation. This is not to say that it was a mistake: lowering the hurdles under general antitrust law would have been problematic and might never have come to fruition. But one has to be aware of the limitations. If one decides to craft a detailed regulation that specifically discusses app stores, then one has to get it right. While there’s a lot of good stuff in the DMA, it’s far from waterproof, and that’s what Apple is capitalizing on.

Another problem: technical smokescreens (including, but not limited to, security and privacy)

Finally, even if the dogmatic problem outlined above was overcome and the courts were prepared to rule on the basis of the net effect of all of Apple’s EU app rules (as opposed to focusing on literal violations of single DMA passages), there would still be another major challenge:

The judges would have to understand that Apple’s privacy and security pretexts are weak. Yes, if there’s some manual app review, potentially one might find privacy risks or security violations; but no, Apple’s app review is primarily focused on enforcing Apple’s tax collection and the enforcement of other rules that are all about Apple’s commercial interests.

There will also be questions of consumer behavior involved. What does it mean to uninstall an app and then reinstall it from another app store? How many customers does an app maker lose? What are the practical implications of having to offer one app version (under one app ID) in the EU and one in the rest of the world? It’s not that judges can’t understand this. Some of it shouldn’t be hard to understand as they all have smartphones by now. But there’s no guarantee that they will understand.

There must be some enforcement attempts by the EC and by private parties. What Apple is doing can’t be accepted without a fight. But if the objective is to bring about significant change, and in the foreseeable future, then the EU has to amend the DMA after the elections. Maybe some of the naysayers who opposed certain measures in the past will now agree that a law that may work for parties that are reasonably respectful of the legislative intent behind a law won’t suffice when you have to deal with Apple.