Home >

On the Morality of Apple's 30% Cut

On the Morality of Apple's 30% Cut

How things could be different for app developers and users alike.

Posted on August 24, 2020

The free WordPress iOS app, which helps manage sites created with the open source content management framework of the same name, is the most recent victim of the Apple App Store’s police, following Fornite’s—still ongoing—saga.

The WordPress case is a source of controversy because Matt Mullenweg, the founder and lead developer, was forced to add payment functionality which is unrelated to the app’s purpose. All of this just because wordpress.com happens to resell domain names, which helps fund the open source project. The publishing of any “free” apps, that allows payments or subscriptions to take place outside of the App Store—bypassing Apple’s 30% commission—is in violation of Apple’s terms and conditions.

Thom Holwerda, chief editor of OSNews, calls it an extortion fee; others, the Apple Tax. Moreover, the fact that large enough bullies can get away with a discount—in contradiction to Tim Cook’s claim that fees are applied to all developers evenly—only helps to add insult to injury. It is no wonder that people are starting to question both the legality and morality of Apple’s cut, which is larger than the world’s average corporation tax, at roughly 25%.

In the analysis of this dilemma, two considerations are at odds; the need to reward innovators economically (by means of patents, intellectual property rights, etc.) and the need to foster competition by dismantling monopolies, like in the historic cases of AT&T and Standard Oil.

Apple’s natural position, in a supposed free market world, is simple: “If you don’t like my lemonade, go to someone else’s lemonade stand”. Except, of course, that the other lemonade stand is Google’s, which suffers from similar draconian terms and conditions. With BlackBerry and Microsoft out of the game, the effective market condition is that of a duopoly.

If we look at the game console’s business, hardware is often subsidised (especially in the early years when technology is new and the investment in state of the art manufacturing facilities needs to be amortized), and profits can only be obtained from the royalties and commissions charged to game developers. Apple and Google could make a similar argument, but this is not factually the case, given that mobile devices are rarely sold at a loss. On the contrary, mobile device inflation has been several times above wage growth in the last five years. An app store is an industry in and of itself, not a compensating revenue mechanism for some other, non-profitable line of business.

There is also the argument of curation costs; the fee is not a charge on the developer, but the cost of weeding out the other bad apps that would devalue the app store as a whole—including the app subjected to the fee itself. This is similar to municipality (borough) property taxes. Good neighborhoods pay higher taxes because it is the only way to keep the streets clean and well policed.

By definition, most economic incentives for innovators are in the form of monopoly rights—albeit temporary—so, I don’t believe that the Apple App Store’s fee existence (and concrete amount) is the problem per se. I believe that the issue is more fundamental: it is the failure of government, and the triumph of markets, in providing what has become one of most important resources in the world—computing—to its citizens.

Take the case of water. Even countries where private corporations run water supplies, there are price controls, and strict enforcement of healthy & safety regulations. However, there are also “premium” products, provided by strictly for-profit corporations, like Evian. Both co-exist; affordable, ubiquitous, drinkable, tap water, and premium bottled water.

But let’s get back to the issue of universality of computing. Why computing; isn’t this discussion about mobile devices? Yes, the most universal computing device is not something like the OLPC laptop anymore, but a pocket, battery-powered computer with radio capabilities (e.g. 2G, 4G, etc.): in other words, a smartphone. Despite appearances, for the vast majority of the population, this is not a luxury item—to be bought at an all-glass shrine—but an essential survival necessity.

We have to remember that the Internet is not the product of free markets, but of government. Yes, tax-funded government. Commodity-wise, universal computing (which is by definition mobile computing) is way too important to be left to markets. But doesn’t the web already give us the ability to run applications ubiquitously? Well, implementing a modern web browser is so much more difficult today than even Microsoft has given up. Browsing the web is today pretty much tantamount to running Chrome. But let us pretend that Mozilla is still well funded for a second.

The question is, how do we create a commodity-like, universal compute “platform” that protects citizens from the vagaries of private market interests. Take the case of banking mobile applications. Mobile banking is more essential in remote, deprived villages, than it is in New York or London, where the nearest ATM is a few meters away.

My bank, for instance, has long made me dispose of my hard RSA token in exchange for a “digital” app equivalent. Naturally, the app only runs on either iOS or Android. Not a problem, there are $10 Android devices in developing markets, but this is not the point.

What would happen, for instance, if a loan is 30% more expensive if I apply for it through the app? In a developing market, that 30% could be added to an app that reports commodity market prices used by farmers. In essence, things that were free become, all of a sudden, for profit. Companies that may not even have a merchant account, would need to add one so that they can charge a dollar, so that 30 cents can be conferred to Apple.

My preferred—pie in the sky solution—would be a fully open source platform (maybe based on the likes RISC-V) but this is too impractical. Mobile components are already obsolete by the time that patents have expired or by the time that they are available as off-the-shelf parts rather than as SoCs. When the GNU/Linux revolution emerged in the x86 space, PC parts were already part of a highly commoditized market of interchangeable parts. Even if the hardware problem could be cracked, there is the chicken and egg situation when it comes to applications; developers only write applications when there is demand for the relevant mobile platform, but there is only demand for such a platform if there are apps in the first place.

Given that the practical measures live in the legal space, I provide three ideas that try to honour, to a certain extent, the need for app store’s owners to profit from their investments—so that they can keep innovating, as per neoliberalism’s ethos (or lack thereof).

The first is making side-loading a universal right for all users of any compute platform: this is somewhat the case in the Android world and the de facto way in which things still operate in the Wintel world, notwithstanding Microsoft’s attempts to push their “me too” app store.

In this model, app fees associated with a given app store are simply a convenience fee, with which the vast majority of the population is OK in exchange for avoiding the hassle of searching applications, avoiding spyware and so on.

The second option is mandating the need to support a universal, non-proprietary app execution model: there is no need to bring back J2ME from the grave here. The Progressive Web App spec pretty much achieves this, but this would require such applications to be ring fenced from each applicable app store’s current terms and conditions. The benefit here is that the maturity in terms of sandboxing is high and getting even higher.

In this model, the app store owners can still demand a premium for the privilege of running fast, native code. The drawback here is that app store owners could have the perverse incentive of being negligent about optimising their web-based execution environment, in favour of the native one.

The third, and last option, is making fees timeboxed and/or relative to the quantity and size of downloads. Patents don’t last forever; it is reasonable to expect that the fees related to the privilege of having access to an app store would decrease over time. In book publishing, for instance, the converse is true; royalties go up the more copies a book sells because the publishers’ costs are recouped when selling the first copies.

As sensible as this sounds, the track record in models like this is poor when considering similar closed platforms such as game consoles. The reason why is that this type of policy can always be dressed as an “anti-market price control regulation that stiffens innovation”. Even if a patent-like timebound mechanism could be enforced, vendors would just create a “app store 2.0” where the patent-like clock starts from zero again. Corporations don’t like governments setting prices; they react better to black and white DOs and DO NOTs. So the solution that most developers want (simply sensibly lower fees, say 5-10%) is, in my view, the most unlikely.

Conclusion

We live in a world where current policy is geared towards the formation and protection of monopolies (when it applies to concrete commodities, like Apple’s App Store), in the industries where it matters (technology, pharmaceuticals, food, etc.) under the hypothesis that the inability to profit under said monopolies would prevent the innovators from bringing the service or product into existence in the first place.

As we had discussed, products and services, like the Internet and the Web themselves, cannot only be brought into existence by the market, but by the government too. Big government is not necessarily evil; in the US, it put a man on the moon, and in the UK, it created the NHS.

Finally, smartphones are not what their name suggests. They are neither smart nor phones (for the most part) anymore. They are ordinary rather smart—otherwise they wouldn’t be carried by nearly half the world’s population—and they can barely be called phones since the implementation of the phone capability (the radio component and antenna) constitutes a tiny fraction of the total device’s real estate and cost profile. A smartphone has become effectively the computer of every man in the street.