Here we go again. Another Android antitrust drumbeat. In April, the European Commission announced a formal investigation into Google’s Android operating system. This is not the first time antitrust allegations have been leveled at Android.
Let’s take a short trip back in time. In early 2014, a flurry of media reports emerged accusing Google of anticompetitive conduct surrounding its Android licensing arrangements. The reports cited industry insiders and experts who had “examined” contractual arrangements that surfaced through an unrelated court proceeding between Google and Oracle on intellectual property claims. Much of that early reporting was fueled by a detailed blog post and paper by longtime anti-Google “consultant” Ben Edelman which accused Google of leveraging its monopoly power through secret nefarious contractual arrangements with device manufacturers (most of which actually had nothing to do with the Android operating system itself). Unfortunately, much of the reporting (and much of Edelman’s analysis) turned out to be misleading or just plain wrong. Industry followers, open source experts and academics followed on to poke skyscraper-sized holes in the initial reporting. (I had my own take.)
FairSearch, an organization funded by Google’s competitors aimed at bringing regulatory scrutiny on the Mountain View company, used the Edelman claims as the basis for a complaint it filed in April 2013 with the European Commission — shortly after the original round of media coverage — claiming that Google’s below-cost distribution of Android (read: free, open source) was predatory pricing that made it difficult for Google’s competitors to compete. It also claimed that Google’s practice of offering its suite of mobile applications in a package instead of a la carte (via so called MADA agreements) foreclosed competition in mobile platforms and applications.
In short, the claims were laughable. The open source community cried foul, pointing out the dangerous implications such a precedent would set given that all open source software is available for free. Furthermore, commenters noted the irony of a group funded by proprietary software companies attacking the free distribution of open source software as “predatory.” Others pointed out that the MADA agreements are standard operating procedure for companies trying to build user friendly products and ensure that customers have an expected suite of services available to them “out of the box.” At best, I thought, these claims were a sideshow. A PR stunt orchestrated to keep Google’s PR and legal teams fighting on multiple fronts and that they would fade quickly.
Well, FairSearch is back on the scene with a new “report” (that recycles its tired old claims). Days after the report was launched, another company conveniently filed a complaint against Android in Europe after being banned from the Google Play app store because it failed to comply with the terms of service. All this is seemingly aimed at prodding the European Commission into action, after the U.S. Federal Trade Commission and the Korea Fair Trade Commission found no wrongdoing with Google’s mobile practices.
The new FairSearch report is littered with too many factual inaccuracies and logical fallacies to dispense with in one blog post. I will address the two overriding takeaways from the report, one in this post and one in a follow up post. In this post, I will address the claims that Google is a dominant player in the mobile space who possesses monopoly power. For competition authorities to act on an abuse of dominance charge, it must first be established that monopoly power exists.
Second, which I will address in a follow up blog post, is the fact that the majority of FairSearch’s complaints are centered on arrangements and contracts designed to overcome potential pitfalls of open source projects — fragmentation and compatibility issues. In short, it appears that FairSearch, and its funders, are encouraging regulatory intervention aimed at hurting the competitive viability of Android by increasing Android’s fragmentation. Clearly this would benefit Google’s competitors, but it certainly wouldn’t benefit users who want a consistent, bug free experience. Nor would it help application developers that depend on a well functioning, compatible Android ecosystem that allows their applications to work across multiple devices.
Google, Android and Monopoly Power
In monopoly cases, it is necessary to establish that a firm possesses monopoly power. In short, this means that a firm has the ability to act independently of competitive pressure. Under such conditions, a “dominant” firm can raise prices and act anticompetitively without the fear that users will choose another product or service. If there is no finding of monopoly power, then there is no antitrust case. Period. End of story.
With this in mind, it is blindingly apparent that Google does not possess anything close to monopoly power in the mobile space. For those who want a quick overview of my argument, here are the bullet points.
- Google can’t pursue a price predation strategy because it can’t raise prices on open source software.
- Android users are flocking to iOS, so Android is not shielded from competitive pressure.
- Apple, not Android, takes home the vast majority of mobile phone profits.
- Even Google reportedly makes more money off of iOS users than it does off of Android users.
- Global Android shipments actually declined in Q4 2014 (while iOS shipments increased 90% and Windows Phone shipments increased 19%).
- Competitors and startups can use Android’s open source code to create their own mobile operating systems. High profile companies already doing this include Amazon and venture capital flush Cyanogen. In fact, in Q4 2014 85 million units of “forked” Android smartphones shipped. Forked Android OSes are completely independent of Google. Ironically, the biggest competitive threat to Android is the Android Open Source Project.
- Because Android is open source, Google has limited control over what device manufacturers do with it. They are free to customize the UI and give favorable placement to their own products and services.
Alright, now for the deep dive.
Let’s dispense with the pricing arguments first as they are the most absurd. Since Android is freely distributed under an open source agreement, the price of Android is $0 and Google is effectively barred from raising that price. In fact, FairSearch is just complaining that price is too low:
Google’s distribution of Android at below-cost makes it difficult for other providers of operating systems to recoup investments in competing with Google’s dominant mobile platform.
In competition policy, a predatory pricing strategy would involve a firm distributing a product below cost to drive rivals out of the market so it could raise prices in the future. Since Google can’t raise prices on Android, this claim lacks any credibility. Second, Google would need to be driving competitors out of the market. How’s that going? In Q4 2014, global Android shipments actually declined! So Android is losing market share and has no effective way to raise prices. If there is an evil genius behind this strategy somewhere in Mountain View, they should probably be fired.
So, that part of the complaint is basically just competitors whining that it’s hard to compete with free. Obviously, this is not a competition policy problem. It’s a business model problem. For one, how would consumers be helped if the complainants had their way and Google was forced by regulators to charge for Android?
Moving on, I turn to the claim that Google’s possession of monopoly power allows it to behave anticompetitively. FairSearch’s only datapoint to undergird this argument is that Android has a high market share. They highlight a graph showing Google has had 70%+ mobile OS marketshare for the last couple of years. There is little more evidence to their charge than that. They conveniently gloss over the fact that 30% of the Android phones shipped in the recent past are actually forked versions of Android that are entirely independent of Google’s Open Handset Alliance. Furthermore, as illustrated by the most recent sales figures, millions of Android users are switching to iOS. Where forked Android distributions represent a serious competitive threat on the low-end of the market, iOS — as demonstrated by the millions of Android users switching to iOS — represents an existential threat to Android for the most lucrative mobile phone users. And the rate of users switching from Android to iOS is actually increasing. As one NYT columnist pointed out, Android is in a precarious position. For clarity’s sake, I will state the obvious: a product that is in a precarious market position cannot also have a dominant market position.
Almost all of these forked versions remove Google products and services and replace them with proprietary offerings. The remaining 70% are still distributed under an open source licensing agreement. Google exerts far less control over Android then a proprietary closed source OS owner is able to exert. As a result, Google has few options to “abuse” its alleged dominance. Professor Torsten Koerber addresses this in a comprehensive paper on the subject:
Google offers Android to OEMs on a royalty-free basis. The licensees are free to download, distribute and even modify the Android code as they like. OEMs can create mobile devices that run “pure” Android…or they can apply their own user interfaces (IO) and thereby hide most of the underlying Android system (e.g. Samsung’s “TouchWiz” or HTC’s “Sense”). OEMs make ample use of this option.
This explains why the other telltale signs of monopoly power are non-existent in the mobile OS market. Despite Android’s share, it is actually Apple that makes 89% of the profits in the mobile platform space, with Android OEMs collectively coming in at a paltry 11%. Google, for its part, makes nearly $12 billion dollars in mobile advertising revenue — the main way it monetizes its work in the mobile marketplace. But, there is a catch. Google makes nearly three fourths of that money from iPad and iPhone users. For all the charges of its dominance, Google makes significantly more revenue off of users of Apple’s iOS for which Google pays Apple a premium to be included as a default search engine on iOS (and Google apps are not even pre-installed, users have to download those Google Apps and Services, which flies in the face of FairSearch’s claims that not being pre-installed on a platform forecloses competition… but I digress). Dominant firms do not pay a premium for placement on their biggest rival’s platform.
The reason for this non-intuitive outcome is that Google does not possess monopoly power. The so called anticompetitive actions Google is accused of are contractual agreements designed to guard against fragmentation of Android and to present Android users with a good user experience (more on that in a later post). Other companies are free to customize Android and include their own proprietary applications and services on the phones. Users are also free to download whatever they want. Unlike Android’s proprietary competitors, Google has few options available to it to act anticompetitively, if this were even possible.
Also, ironically, the competition within the Android ecosystem that Google and the other members of Android’s Open Handset Alliance have fostered has actually started to work against Google. Because mobile handset manufacturers can’t extract much of a premium for consumers due to the large number of Android phone offerings, manufacturers are looking to make money through offering their own Android-based products and services that compete directly with Google’s products. Google cannot stop this. In a recent NYT column discussing Android’s vulnerability, Farhad Manjoo makes this exact point:
Because Xiaomi and others don’t make much of a profit by selling phones, they’re all looking for other ways to make money — and for many, the obvious business is in apps offering mail, messaging and other services that compete with Google’s own moneymaking apps.
Even though it offers a compatible version of Android, Xiaomi has stripped out all of Google’s products and services from its out of the box offering. Here is an excerpt of a review of Xiaomi’s flagship Mi Note smartphone:
Instead of the Google Play store, it’s got the Mi Market. Instead of Google search, it’s got Baidu. Instead of Twitter, it’s got Weibo. Instead of WhatsApp, it’s got WeChat. China’s an alternate universe where Google’s apps don’t need to exist because there are plenty of homegrown alternatives. The Mi Market is exclusive only to China, though; Mi Notes in other Asian regions do have Google Play services pre-installed.
Xiaomi has more than 15% of the Chinese market and aggressive plans for international expansion. (In fact, 70% of Android phones distributed in China do not offer Google Play services.) Couple that with the fact that other new entrants and non-OHA members can “fork” the Android source code and build their own operating systems completely independent of any Google influence. In fact, many other companies are. Amazon is one notable company who has done so successfully. Another, Cyanogen, just received a huge influx of venture capital — and a strategic partnership with Microsoft — to develop an alternative mobile OS that uses Android’s open source code. The irony of this being of course, one of Google’s greatest competitive threats comes from Android itself, as potential competitors don’t have to build an operating system from scratch. They can simple take the Android Open Source Project (AOSP) and modify it to their heart’s content. This is not a theoretical threat either: 85 million forked versions of Android shipped in Q4 2014, compared with 205.86 million units of standard Android phones.
So, to recap, claims that Google is dominant in the mobile OS market rest on misleading market share tallies, willful sleight of hand, and a lack of understanding of the dynamics of open source software. The successful growth of an open source competitor to proprietary mobile OSes has fostered intense competition in the smartphone market, and by decoupling the OS from the hardware, Android has allowed device manufacturers to aggressively enter the market without the worry of having to build their own OSes and without having to foster their own ecosystem of compatible apps from scratch. This competition has proved a boon to consumers. So much so that we are even seeing Android smartphones below the $100 price point. That being said, open source software has some competitive disadvantages as well. Not least of which being fragmentation and compatibility, seeing that no one company is in control of what direction the source code moves in. (This was a problem for Linux.) Most of FairSearch’s complaints attack agreements designed to keep Android from fragmenting, and ensuring it remains compatible across a wide range of phones and tablets that use different hardware and have different screen sizes. Furthermore, the MADA agreements are aimed at ensuring that mobile users have a full suite of services available to them when they first take the smartphone out of the box, much like their proprietary OS competitors have. I will stop there, as that topic is the subject of my next blog post in this series.