Today’s decision imposes a record fine of €4.34 billion on Google. In such an innovative and competitive industry, a decision and fine on this scale sends the wrong message. This post argues there is no obvious foreclosure, explains why Microsoft/WMP is an inadequate precedent (though Microsoft/Skype, in contrast, is directly on point), and asks whether intervention risks reducing innovation, raising prices and harming consumers.
The Commission’s findings at a glance
The Commission concluded that Google is dominant in the markets for general internet search services, licensable smart mobile operating systems and app stores for the Android mobile operating system.
The Commission identified the following abuses:
- requiring manufacturers to pre-install the Google Search app and browser app (Chrome), as a condition for licensing Google’s app store (the Play Store)
- making payments to certain large manufacturers and mobile network operators on condition that they exclusively pre-installed the Google Search app on their devices
- preventing manufacturers wishing to pre-install Google apps from selling a single smart mobile device running on alternative versions of Android that were not approved by Google (Android forks)
In Search of A Theory of Harm
Is a “licensable” operating system market too narrowly drawn, and does it mask the true nature of competition in this sector?
The Commission defines a market for “licensable” mobile operating systems. Its press release shows that it does not see other operating systems used by Apple and Blackberry as in the same market because the latter are not available for licence by third party device manufacturers.
Even if few other manufacturers license their OS on the scale that Android does, it is counterintuitive to argue that there is not fierce competition between operating systems.
The very existence of Android is evidence of the competitive pressure exerted by operating systems on each other. The sea change was the advent of the touch screen smartphone in 2007. Competition from Android – an open sourced operating system delivered for free to handset makers – has been vigorous ever since to keep pace with smartphone innovation. Since 2008, Google has delivered 14 major releases of Android. There have been year on year innovations: swiping, security, cameras, fitness functions, facial recognition, multitasking gestures, payment systems…
If Android were to provide a lower quality experience, or relax its innovative effort, compared to its rivals, smart phone users, always demanding the latest technology, would surely switch in droves at the next, typically annual, round of handset launches. Operating systems are subject to intense competitive pressure and live or die by their latest release.
This intense competition produces compelling consumer benefits. Android makes cutting edge handset technology available to a mass market, and at a reduced price. One statistic suggests the average price of an Android smartphone halved over the period 2010-2016, from $441 to $215.
Is the foreclosure story plausible?
The Commission alleges that Google’s Android licensing practices foreclosed competitors through pre-installing apps or pre-setting default search services.
Publicly available information points the other way. It seems that handset makers can choose not to take Google’s suite of apps and offer their own instead. An example is Amazon’s tablet operating system, FireOS. In China, handset makers offer a local suite of apps rather than Google’s.
Where handset makers do distribute Google’s apps, then the question arises whether competing apps or services are actually foreclosed. Mobile phones are intensely personal, and personalised, devices. A quick fact check suggests there were 82 billion downloads from Google Play in 2016. 80% of consumers customise their home screen. So intuitively it seems implausible to say that consumers would put up with an undesired (or lower quality) app, just because it comes pre-installed. Some of the most popular apps – Facebook, Dropbox, WhatsApp – are more popular with consumers than their Google equivalents, whether or not the Google apps are pre-installed. Among non-Android device users, you would be hard pressed to find any who have not downloaded the popular Google Maps or Waze. And that is regardless of whether the non-Android device has another map function pre-installed. The same is true of Chrome and Search. These are popular apps downloaded by non-Android device users regardless of whether the non-Android device has other search engine or browser apps preloaded.
The Commission press release notes that, on Android devices (with Google Search and Chrome pre-installed), more than 95% of all search queries were made via Google Search, whereas on Windows Mobile devices (where Google Search and Chrome are not installed) less than 25% of all search queries were made via Google Search (meaning that more than 75% of search queries happened on MS Bing).
However, that does also mean that users are five times more likely to switch from a Bing default to Google (25% Google vs 75% MS) than a Google default to Bing (95% Google, 5% (mainly) Bing). This suggests quality and preferences overcome pre-installation (which of course is precisely what drives people to download other popular apps, e.g. DropBox, Facebook, WhatsApp).
EU precedent militates against any finding of pre-installation foreclosure in relation to the apps in question. In Microsoft/Skype the Commission and, on appeal, the Court, examined whether, if Microsoft chose to install Skype on the desktop post acquisition, that would lead to foreclosure of rival video telephony services. The conclusion was there was no technical or economic constraints which prevented users from downloading several communications applications on their operating device, especially as the software concerned was free, easy to download and took up little space on hard drives. Even more recently in Facebook/WhatsApp (2014), the Commission found that, in relation to the apps in question, multi-homing was facilitated by the ease of downloading – which is generally free, easy and does not take up much smartphone capacity. The Commission also found that using multiple consumer communications apps was easy because a user does not have to log in each time, when switching an app.
Facebook/WhatsApp and Microsoft/Skype (upheld on appeal in 2013) are clearly better precedents than the older case of Microsoft/WMP. The alleged foreclosure in Microsoft/WMP occurred in a world without apps, or appstores, or user-driven device customisation available with a few swipes and clicks.
The Canadian Competition Bureau did not find evidence of foreclosure when it discontinued its investigation into Google in 2016. It found that search engines “can and do” compete to appear as the default search engine on smartphones etc. Switching activity over time was significant and manufacturers were able, for example, to ship multiple devices with different pre‑loaded search engine defaults. The Bureau noted specifically that consumers “can and do change the default search engine on their desktop and mobile devices if they prefer a different one to the pre‑loaded default”.1)http://www.competitionbureau.gc.ca/eic/site/cb-bc.nsf/eng/04066.html.
There are offline precedents too. Think back to Van den Bergh Foods (ice cream cabinets), Heineken (beer dispensing equipment) and Coca-Cola Undertaking (drinks fridges). The conclusion from those cases is that full equipment exclusivity (for example, the ice cream cabinets in Van den Bergh Foods) might be capable of foreclosing. But if the supplier allows rivals sufficient space (Coca-Cola Undertaking) or makes the restraint easy to change (Heineken) then foreclosure risks do not arise. In Google’s case, the application of these offline precedents suggests foreclosure is unlikely. If Google has only a few of the approximate 70 or so spots on a device and these can easily be deleted, replaced or added to by rivals, or pre-sets changed, there would seem to be no risk of rival exclusion.
How should efficiencies be evaluated?
Android has created a mobile ecosystem generating significant benefits. This balance between distribution of Google’s revenue generating apps as the quid pro quo for Android’s ongoing R&D would seem to be the clearest example of pro-competitive benefits. The benefits of the “anti-fragmentation” agreements should also be spelled out. It is not the same as ‘non-differentiation’: the approach is necessary to preserve the attractiveness of the ecosystem for app developers. It seems likely that developers would not code, or code far fewer, apps for multiple versions of forked mobile OS. And this of course drives efficiencies: lower prices, greater output and innovation in the Android ecosystem. It also ensures that it is easier for consumers to switch between devices, increasing competition and innovation between OEMs.
Many would say pre-installation (or preset services) are at most a transitory – and above all non-exclusionary – marketing benefit. Consumers can and do swiftly switch away from apps, or default services, they do not want. And the quid pro quo in terms of efficiencies would appear overwhelming. Device makers receive a free operating system, consumers have the latest innovations at their fingertips and the prices of handsets fall.
It may be questioned what impact the decision will have on these prima facie highly desirable outcomes. We do not know whether Google will change its licensing practices for Android, or reduce its R&D efforts, if it cannot secure returns from revenue-generating apps or services. But either outcome would seem to be damaging to competition and consumers. Certainly not the message that should be associated with the biggest fine in antitrust history.
Given the ease with which consumers can switch, the decision is an attempt to overcome consumer reluctance to move away from a product they are happy with, in order to drive broader policy objectives in the digital sphere.
References [ + ]