Apple critics, EU Commission will get value out of March 18 workshop only if they recognize DMA limitations

As games fray has previously explained, Apple’s EU app rules, at least for the largest part, comply with the DMA unless the regulation is interpreted more purposively than the EU judiciary has so far interpreted comparable statutes (February 19, 2024 games fray article). In light of that as well as the fact that a court determination of fair, reasonable and non-discriminatory (FRAND) terms would be time-consuming and fraught with uncertainty (January 26, 2024 games fray article), games fray is convinced that only a near-term legislative amendment realistically has the potential to open up the market for iOS app distribution in the EU before the end of this decade (January 20, 2024 games fray article).

Stakeholders must consider four harsh realities:

• The DMA was an interesting initiative, but in its current form the part concerning app stores is deficient in some critical respects.
• A chain is as strong as its weakest link, and it takes a multiplicity of elements to make alternative app stores work for the masses (as opposed to merely “adult content”). Even if some of Apple’s rules were modified, there probably still wouldn’t be a mass market opportunity for alternative app stores.
• The European Commission will likely identify Apple’s conduct as uncooperative, but has not taken swift and decisive action against Apple in any competition context so far (with maybe the exception of tap-to-pay services, where Apple has offered commitments). And the EC has never picked a fight with anyone over FRAND rates.
• Apple is not susceptible to public pressure. Even bad press such as a recent Financial Times article (February 22, 2024 X post by games fray) won’t dissuade Apple from its charted course. The sole exception where Apple ultimately acted flexibly involved a pandemic-related small business promotion with Meta/Facebook.

Having said that, here is what games fray believes everyone should strive to accomplish at and around the March 18 workshop.

Shut out Apploturfers

At a workshop during the preparatory phase of the DMA, the EC incredibly allowed ACT | The App Association to speak out on behalf of small app developers. But ACT is, like Apple, an enemy of app developers (regardless of size). ACT takes anti-developer positions with absurd arguments according to which developers benefit from Apple’s heavyhandedness.

If the EC does not want to make itself and the DMA enforcement process ridiculous, it has to decline to let ACT participate. Apple critics should also make it clear that ACT’s deceptive lobbying on Apple’s behalf cannot be tolerated. In the end, ACT’s participation is just going to cost time that actual app developer representatives need to explain the issues.

At this stage, there is no more excuse if the EC allows ACT to speak at the event (unless ACT representatives are considered part of Apple’s delegation and use up some of the time allotted to Apple). In an email that ACT sent to many (if not all) Members of the European Parliament, the company itself (yes, it is a company, not an association) stated that small app developers pay no fees (February 22, 2024 ip fray article). ACT told MEPs that it secures funding from Apple and other corporations to offer small companies free membership, which is an insult to human intelligence as it means that those small companies have no say. In fact, those small companies do not get to vote on who runs ACT: there is no general assembly where such a vote would take place, as ACT is an Apple-controlled company.

If the EC held an event about the situation in Ukraine, it couldn’t allow an organization to speak on behalf of Ukrainians that says it secured funding from Russia to offer Ukranians free membership.

In 2022, a Bloomberg article revealed that even by its own admission, ACT receives more than half of its funding from Apple. Bloomberg also interviewed four former ACT employees and learned that Apple effectively gives ACT directions and that it is a gross understatement to say Apple foots more than half of the bill.

Get help from law firms and have them help the EC, but don’t believe them if they overpromise

For law firms, the DMA is a full employment act if they are either retained by Apple or by one or more Apple critics. Trying to enforce the current DMA against Apple will take many years as Apple routinely exhausts all appeals.

The EC will need help from lawyers paid by Apple’s critics, but no EU law firm will be in a position to achieve in the coming years that the market for iOS app distribution in the EU becomes seriously contestable. And Apple won’t back down. The theoretical maximum for a DMA non-compliance fine is 10% of worldwide turnover, but whatever fine is imposed is court-reviewable and must be proportionate to the seriousness of an infringement. In the Spotify antitrust case, the theoretical limit is the same, yet the EC is now expected to impose merely a $500 million fine.

No law firm will be able to truthfully promise for a number of years (in games fray‘s estimate, 5+ years) an enforceable court ruling on the fees Apple may charge those operating alternative app stores or distributing their apps through them.

Categorize the issues by how they can be realistically tackled

It does not make sense to expect miracles from the EC’s DMA enforcement team. At the end of the day, the enforcers can only do what the law enables them to do, and that is nowhere near as much as some Apple critics may have hoped.

The most effective and efficient approach will be to assign the issues to the following categories:

• Apple rules and conduct that can be realistically challenged through enforcement of the current DMA. For instance, there may be an interesting discrimination case over the fact that alternative app stores, because of the 50-eurocent per-install per-year fee for apps distributed through them, are going to be unattractive for the purpose of distributing free apps, such as e-commerce storefronts, while Apple’s own App Store carries them without such fees involved.
• Apple rules and conduct that the courts of law may deem unlawful if they go reasonably far in the direction of purposive interpretation, but which will be deemed lawful if the interpretation is literal or only moderately purposive. An example for this is what Apple has done to Progressive Web Apps (PWA). The DMA wants to make the market for mobile browsers, including browser engines, competitive. Apple will now allow alternative browser engines, but doesn’t want PWAs to become a competitive threat to native apps. A court may feel that the DMA fails to address PWAs, but it may also deem this a violation of the DMA’s general anti-circumvention clause.
• Apple rules and conduct that cannot be realistically challenged through regulatory action and litigation based on the current DMA, but requires an amended DMA. The most important item in this category is FRAND: the question of what fees Apple can impose without violating the DMA requirement to provide access on fair, reasonable and non-discriminatory terms. That is objectively the most important question involved and will be discussed in more detail below. Another item that probably falls into this category is the regional fragmentation that Apple imposes by requiring those who seek to benefit from certain DMA-related liberties to make separate apps (with unique IDs) only for the EU market. It’s probably not a case for the anti-circumvention clause as Apple can argue that the DMA regulates only what happens in the EU market, not the interplay between apps that are distributed in the EU with app distributed abroad. Apple can also point to an existing system architecture where certain entitlements are connected to an app ID. It would be trivial for Apple to provide entitlements on a per-app per-region basis, but Apple is not required to do development work unless the DMA establishes a clear requirement. Not to go beyond the call of duty is not the same as circumvention.

App Store critics should make themselves familiar with what FRAND means in practice. The term is used in antitrust contexts all the time, but there is hardly ever a court decision on what FRAND actually means, numerically speaking, in a specific case. The only field in which FRAND determinations are made repeatedly is the licensing of standard-essential patents (SEPs), a topic covered by games fray‘s sibling site ip fray.

If FRAND wasn’t complicated and resource-intensive, regulators like the EC would have enforced the antitrust laws to bring down abusive charges. In practice, regulators typically focus on structural questions, but keep clear of numerical questions. Lawmakers prefer not to engage in governmental rate-setting either, with a very few exceptions such as mobile roaming in the EU.

FRAND doesn’t mean that anything happens quickly. FRAND takes time. In each of the two most recent UK FRAND cases, the court needed approximately one year after a trial to come up with a FRAND ruling. And there was a lot of work before those cases went to trial.

FRAND also doesn’t mean that courts can just use their judgment and set a rate that seems fair. In practice, FRAND means that courts are looking for points of reference from the marketplace. In the App Store context, that favors Apple:

• If app makers complain, Apple can always argue that even its standard App Store fee of 30% is FRAND. Courts love comparables: they focus on other commercial agreements as real-world facts that enable them to tell whether a certain rate is abusive or acceptable. The problem with the 30% rate is that Apple can point to the Google Play Store, even to some other Android app stores, to video game consoles and to Steam. There are reasons for which video game consoles can be distinguished from mobile app stores, but Apple is going to be able to point to a lot of data point that suggest 30% is a market rate. Courts are going to be extremely reluctant to brush that “evidence” aside.
• If it’s about alternative app stores, Apple charges a per-year per-install fee of 50 eurocents. There is no point of comparison for that one. Psychologically, it appears low. Economically, it makes a far greater impact than the number suggests at first sight (January 28, 2024 games fray article). The problem for the courts is that it all depends on only one comparable: Google doesn’t impose a fee like that on alternative app stores and the apps carried by them. But the DMA doesn’t say that zero is the right number. It says Apple may ask for something, and there’s no alternative non-zero number to the 50 eurocents that anyone can point to.

The structural flaw of the DMA is that its FRAND requirement is non-qualified.

FRAND experts know that it’s all about the parameters. That’s why FRAND arbitration agreements in the SEP licensing context tell the arbitrators very specifically what criteria to apply and what arguments to consider.

Parameterless FRAND as under the current DMA makes it easy for Apple to justify its demands with the following arguments:

• intellectual property in application programming interfaces (APIs): such IP is rather weak if one actually has to enforce it, but the DMA enforcers at the EC and the courts of law will have to consider it for valuation purposes
• intellectual property in development tools; that kind of IP is definitely stronger than in APIs, and the argument that Apple makes those tools available for a seemingly low annual fee doesn’t count because Apple did so only on the basis that app makers would ultimately have to distribute their apps through the App store
• app review costs: Apple will still perform a basic review to notarize apps; that review will be 99%+ automatic, but Apple can argue that it also costs something
• access to the customer base: Apple can then, again, point to all the platforms that charge 30%, which means that 3% is attributable to payment processing while 27% is for access to a customer base

It is very unlikely that any FRAND determination under the current DMA will bring down Apple’s charges at a game-changing level. The EC may not even want to touch a subject it has historically avoided, and the courts of law will be reluctant to make major adjustments. Unlike in the U.S., a “margin squeeze” argument can be made under antitrust law in the EU, so App Store critics can also try to argue that under the DMA there must be a sufficient margin for alternative app stores to operate. But the DMA won’t help unless and until it limits Apple’s compensation to just API-related IP and ensures that Apple can’t use its gatekeeper powers to extract supra-FRAND rates in accordance with the seventh item of games fray’s proposed DMA changes (February 20, 2024 games fray article).

Apple is very familiar with how FRAND determinations work, and in the SEP context Apple fights hard for certain parameters such as “smallest saleable patent-practicing unit” and the value of patent-protected innovation prior to inclusion in an industry standard. At an abstract level, very similar questions will come up in the DMA FRAND context. Most of Apple’s critics have never dealt with the subject. Of the companies that have spoken out against, or are otherwise known to oppose, Apple’s new EU app rules, only Microsoft (January 30, 2024 games fray article) has ever been a party to a FRAND rate dispute, and that was 12 years ago.

Apple’s critics will have to be realistic about what the EC is equipped and prepared to do. They can fill hundreds or even thousands of pages with legal and economic argument to urge the EC to challenge Apple’s rates, but the EC has not done anything like this before and, sadly, is unlikely to have an appetite for it this time around. The EC should make it clear to stakeholders what it is realistically willing to do, and what the timeframe would be.

In all likelihood, the EC’s scarce enforcement resources (compared to Apple’s vast resources) will be used more efficiently if the EC and stakeholders openly discuss the limitations of the DMA in its current form, and then set the right priorities. If the EC acknowledged that certain aspects of Apple’s conduct can’t be challenged under the current DMA (at least not in a promising way), the debate can move on to what EU lawmakers should do, even well ahead of the review clause in the DMA (2026). Some enforcement of the current DMA will make sense, but the two tricky F items (FRAND and fragmentation) can most likely not be solved, and especially not in the foreseeable future, unless the EU’s legislative institutions realize that Apple has just provided a reality check that the naysayers in those institutions can’t ignore.

The current DMA is like a car that an automaker releases into the market, only to make a major recall because some serious issues need to be fixed as soon as possible. That’s awkward, but that’s the way it is. Everything else is a waste of time and money.