The European Commission’s competition investigation into Google’s Android mobile operating system (OS) has unsurprisingly raised a lot of attention and commentary. It’s fair to say that so far most comments focused on the ‘abuse part’ of that investigation. That is the part which deals with the question whether certain provisions in the Mobile Application Distribution Agreement (MADA) relating to the bundling of Google apps and the anti-fragmentation agreement (AFA) constitute an abuse under Article 102 TFEU. In fact, the trade association I work for has addressed a letter to Commissioner Vestager explaining why the MADA and the AFA are key to the functioning of the Android ecosystem, spurring innovation and competition.
It goes without saying that readers who are more familiar with the details of competition law will know that the abuse part of any Article 102 investigation formally is the last step competition enforcers take (while arguably being the most interesting). Before one gets to the question of whether a company abused a dominant position, one has to find whether the company actually has a ‘dominant position’. After all, a company can only abuse a dominant position if it has one. This post will discuss this issue in the context of the Android investigation.
But first, let’s continue with competition law basics: how does one find ‘dominance’? Conceptually, a finding of dominance involves a two-stage assessment. First, one has to define the relevant market. While that definition can involve complex economic assessments, it is essentially a matter of substitutability. Where goods or services can be regarded as substitutes or interchangeable by the consumer, they are within the same product market. Second, competition authorities look at whether a given company has a ‘dominant position’ on the relevant market. For economists, companies with a dominant position are companies that have substantial market power. In the often recurring words of the Court of Justice of the EU (CJEU), a dominant position:
“relates to a position of economic strength enjoyed by an undertaking which enables it to prevent effective competition being maintained on the relevant market by affording it the power to behave to an appreciable extent independently of its competitors, customers and ultimately of its consumers”. (Case 27/76 United Brands v Commission, para. 65)
Importantly, the Commission explains in its Guidance on Article 102 Enforcement Priorities that the notion of ‘independence’ “is related to the degree of competitive constraint exerted on the undertaking in question” and where competitive constraints are ineffective, the “undertaking’s decisions are largely insensitive to the actions and reactions of competitors, customers and, ultimately, consumers” (para. 10).
With this in mind, let’s turn to the ongoing Android investigation. To most people, competition experts and non-experts alike, one aspect of the Commission’s investigation stands out: the market definition. When the Commission announced the Statement of Objections (SO) it sent to Google, it held that the company has a market share of more than 90% in the market for licensable smart mobile operating systems. The word ‘licensable’ is key because it means that Apple’s iOS which powers all iPhones and iPads is outside the scope of the relevant market. So are iPhones really not competing with Android-powered smartphones? (Leaving aside the question of whether tablets should be included in the market definition for now).
One can argue the answer to this question depends on perspective. If you’re an original equipment manufacturer (OEM), like LG or Samsung, iOS is not an option for you, indeed. Apple does not license iOS, leaving OEMs with the possibility of either licensing from a mobile OS provider or developing their own OS. OEMs are also free to develop a ‘forked’ OS based on the Android open source code, of course. From the perspective of app developers and consumers, however, the answer to the question above would be dramatically different.
Without the need to dive into a complex econometric analysis, it is fair to say that in the eye of consumers iPhones and Android-based smartphones are pretty interchangeable, particularly as regards the higher-end Android phones that have started to compete in the price range of the top iPhone versions. Today consumers can seamlessly switch from OS to OS — Apple even developed an app allowing consumers to easily move from Android to iOS. That may be one of the reasons why Apple has repeatedly stressed the increased number of ‘movers’ coming from Android in its earnings calls. At the end of 2015 Tim Cook said that 30% of customers who recently bought an iPhone switched from an Android mobile device. Furthermore, competition between Android and iOS is set to increase among mid-range smartphones as Apple continues to sell its entry-level iPhone SE as well as iPhone versions right below the newest one. Consumers will continue to look for the best mobile experience even if that entails a change of the OS.
It is also fair to say that in the eye of developers, iOS and Android compete for their app development efforts. While Android powers the biggest volume of smartphones, iOS has proved to be by far the most lucrative OS. Data recently released by App Annie shows that in Q2 2016 the worldwide app revenue generated by Apple’s App Store was twice as high as the revenue generated by Google Play — even though Google Play had twice as many worldwide app downloads as the App Store. Interestingly, the data also showed that when compared to Q1, the revenue gap in favor of the App Store has increased (and in fact by 10 percentage points), indicating that iOS is becoming even more important for app developers. The recent dispute between Spotify and Apple confirms this and shows who the real power player in the mobile sphere is. Without going into the details of this fight, suffice to say that iOS Spotify subscribers are extremely important because they generate most of Spotify’s mobile revenue. Hence, it’s understandable that Spotify has an allergic reaction to any App Store terms and conditions they consider as competition-distorting. But the key point is that iOS users are not only important for Spotify but for many other apps.
Going back to the Android investigation, it seems illogical to leave out what I will call the ‘Apple factor’, i.e. the presence of a clearly weighty company in the mobile space despite its lower share in mobile OSs. Competition in the mobile sphere is better described as competition between competing ecosystems — of course, today the two main ecosystems are iOS and Android but in fast-moving technology markets one should never neglect competition, particularly when exerted by a company as sophisticated and well-resourced as Microsoft. If you like, you can replace the word ‘ecosystem’ with the trendier word ‘platform’ or even ‘multi-sided platform’. In fact, the multi-sidedness of a platform like a mobile OS is key to understanding the dynamics of competition in this market. For any mobile OS platform, ranging from Apple’s closed platform to Android’s open business model, the ability to keep the various sides of the platform happy is key. Arguably the three most important sides in the mobile OS market are consumers, app developers and content providers. Keeping all of them happy in a balanced way is no easy task. In light of the numbers above, Apple clearly excels at keeping consumers and app developers happy. Some commentators even refer to an “iOS bias” to explain developers’ preference for trying new features with their iOS apps before launching them later on Android.
To define a market as narrow as ‘licensable mobile OS’ essentially means that there must be a separate market for ‘non-licensable mobile OS’ and that companies active in these two markets do not compete. In light of the ‘Apple factor,’ that is more than just counterintuitive. But of course in the context of the Android investigation people can say that it does not matter much because even if one includes iOS into the equation, Android’s share in mobile device shipment will still be around 80% which would not dramatically change a finding of dominance.
Well, that would be true if we stubbornly stick to market share as the most important indicator of dominance. However, let’s look back at the key ingredient of a dominance finding: independence. Dominant companies can act independently of competitors, customers and consumers because they are not constrained by the market. It’s essentially an economic situation which allows you to be completely numb and unresponsive to market forces. It allows you to profitably raise price, limit output, suppress innovation, reduce the variety or quality of goods or services and deprive consumers of choice. In light of Apple taking the lion’s share of profits in the mobile economy, how likely is Google’s ability to totally neglect the Apple ecosystem, the demands of app developers and the wishes of Android users? In a multi-sided market setting, one unhappy side will very quickly translate into other sides becoming unhappy, straining the whole ecosystem and pushing consumers towards the competing ecosystem. And that is not to mention that the overall flexibility of the Android ecosystem diminishes Google’s ability to exert control over the final Android experience users get on their mobile devices.
While I believe that the true degree of Google’s economic and competitive independence is key to the Android case, let me just mention a few words on market shares in dynamic online platform markets. In fact, the European Commission summed up that matter in its Microsoft/Skype merger decision quite nicely, even if primarily related to consumer communications services:
“Market shares only provide a limited indication of competitive strength in the consumer communications services markets. [C]onsumer communications services are a nascent and dynamic sector and market shares can change quickly within a short period of time. Furthermore, almost all communications services are offered free of charge.” (para. 78).
In reference to this paragraph, the Commission went on to stress that “market shares are not the best proxy to evaluate the market power of providers of consumer communications services and they only give a preliminary indication of the competitive situation in these dynamic markets.”
If one were to replace the words “consumer communications services” with the words “mobile operating systems” in that sentence, it would still make a lot of sense.