You are here
How Oracle v. Google Magnifies Google’s Android Antitrust Vulnerabilities
Submitted by Scott Cleland on Wed, 2016-05-04 19:07
1 Oracle v. Google case + 1 EU Android Tying Case = 3
While the U.S. Oracle v. Google Java API copyright case that will recommence in public court this month has been completely independent of the EU Google-Android antitrust case, in sovereign jurisdiction, type of law, legal process, timetable and alleged offense, these two cases ultimately could have huge, much underappreciated implications for each other, because they are both about the same thing -- purposeful illegal actions that Google chose to do to extend its search-related dominance into mobile via Android.
In a nutshell, this piece explains why the sum of the Oracle v. Google and EU-Android antitrust cases is greater than the parts. Together they expose:
The software at issue here, Oracle’s Java, is the leading software language for developing cross-platform, software applications – e.g. smartphones and other devices and apps. Google decided to use Java Application Program Interfaces (APIs) as the core APIs in its new mobile operating system, which it offered to the market as open source software for free. (APIs are the foundational building blocks -- the routines, protocols, and tools -- upon which programmers and app developers build new software and apps.)
After months of negotiations between Oracle and Google in 2010, Google became the first and only company at that time to refuse to license Java, publicly deeming it could use it under “open source” and copyright “fair use” without permission from, or payment to, Oracle. Oracle shortly thereafter sued for copyright infringement.
Google won in District Court, but ultimately lost in the appeals process that went up to the Supreme Court, and which ultimately and effectively affirmed the validity and enforceability of Oracle’s Java API copyrights of the “structure, sequence and organization” of its foundational Java APIs.
The current remaining issue for the court to resolve is effectively the amount of damages due Oracle. That’s because Oracle is rebutting Google’s “fair use” claim with evidence that Android has generated $31b in revenues and $22b in profits for Google, belying an “extraordinary magnitude of commerciality” that undercuts Google’s defense that its Android use of Oracle’s Java APIs is non-commercial, philanthropic, fair use. Concerning the amount of damages, Google suggests millions of dollars, while Oracle is asking the court for around $9b in damages.
Meanwhile the EU’s Commissioner for Competition, Magrethe Vestager, issued a Statement of Objections against Android, that tentatively concluded that Google: “abused its dominant position by imposing restrictions on Android device manufacturers and mobile network operators;” … “implemented a strategy on mobile devices to preserve and strengthen its dominance in general internet search…;” and is now “dominant in the markets for general internet search services, licensable smart mobile operating systems and app stores for the Android mobile operating system.”
Relevant to the Oracle v. Google case, the EU has concluded Google has extended its search dominance to Android OS and Google Play in part by contractually making “the licensing of the Play Store on Android devices conditional on Google Search being pre-installed and set as default search service.”
Why the Sum of These Cases Is Greater Than the Parts
1. Together these cases expose a more complete anticompetitive Android story.
In a sense each case fills out and strengthens the other because they are not isolated; they are sequentially the first and second halves of the same story.
Evidence presented in the original Oracle v Google decision indicated that Google considered other alternatives to using Java APIs, including creating their own, but concluded none were satisfactory or offered the benefit of the millions of developers that already programmed with Java. So Google decided to build Android upon the foundation of Java’s APIs because it was necessary to successfully jumpstarting a 2008 first-mover advantage for both Android as an open source smartphone operating system and Android Market as the app store predecessor of Google Play.
To gain first mover advantage, Google had to use Java APIs without any Java license or payment, in order to: credibly call Android “open source” via the Open Handset Alliance; give it away for free to incent device manufacturers and mobile operators and drive fastest unit growth; and create an Android app-development community that would need an Android app sales channel like Android Market/Google Play to create an Android ecosystem.
Simply, Google’s infringement of Java’s IP was the linchpin anticompetitive catalyst that ignited the viral growth that would lead Google from having no OS and no app store until late 2008, to commanding the dominant Android OS and dominant Google Play app store in the world today.
Then to keep that original ill-gotten first-mover-advantage and momentum going, Google anticompetitively leveraged its Google search dominance network effects into mobile search, OS and app distribution, by contractually requiring device manufacturers and mobile network operators that used Android to effectively preinstall Google Search as the default search engine, Chrome as the default browser, and Play as the default app store.
Connecting these two cases, Google’s purposeful mass copyright infringement of Java APIs was an ill-gotten and anticompetitive Google first-mover advantage for Android, and Google’s contractual Android restrictions on device manufacturers and mobile operators were ill-gotten anticompetitive network effects that synthetically catapulted: Google Search to dominate smartphones; Android to dominate mobile OS; and Play to dominate Android app distribution -- in an amazing ~seven years time!
2. Together these cases expose an Android OS that isn’t really free or open source software, but practically a cross-subsidized, hybrid-open-proprietary, Android-Java OS that effectively fixes a predatory price of zero for a dominant mobile OS that now is known to have a material incremental cost – i.e. needed Java license fees.
Think about how the Oracle v. Google case has for the first time exposed the real economics of Android, which in turn undermine Google’s “free” and “open” claims, and which in turn again strengthen the EU antitrust case.
What has changed as a result of Oracle v. Google? The apparent economics of Android have changed.
The courts have already effectively determined that Google infringed (stole or “scraped”) 37 Java API packages owned by Oracle. The effective question is the actual amount of the financial reparations due to Oracle from Google for its seven years of mass infringement (likely in the billions of dollars given Google’s ill-gotten profits), and potentially any going-forward license fee of some type and amount, likely on a per-device, and/or a per-app basis.
In other words, Google now has a debt due to Oracle for Google’s use of its Java APIs without permission or payment. Put yet another way, Oracle has exposed the real economics of Android.
If the real economics of Android have changed, what does that mean for the EU’s Android-tying case?
Android’s Not Really Open Source: It means that Google-licensed Android, is not the true open source software model that the EU envisioned would be pro-competitive in its antitrust case against Microsoft that stopped Microsoft’s abuse of its dominant operating system by denying interoperability with other software via APIs. It also means Android is not open source software, but actually a hybrid-open-proprietary, Android-Java operating system, because without Oracle’s Java APIs, Android literally can’t function as either an operating system or an app-developing ecosystem.
Even if Google could somehow replicate the functionality of all 37 infringed Java’s APIs without further infringing Java again, it would take years: for the ~2 billion functioning Android devices in the field to switch to the new OS; millions of developers to learn Google’s new APIs; and for the 1.4 million apps on Google Play to all be redeveloped.
Sadly and ironically, Google has proven that open source software, just like proprietary software, can be abused, anticompetitive and prone to monopolization.
Android Is Really Not Free: What this means is that there is a real and materially large fixed COST attributed to Android, in addition to a potential Java license fee for every Android device or app sold that Google will owe Oracle. Before, Google claimed internet distribution of Android was incrementally so near cost-free that it naturally should be priced as free.
Now consider this new Android cost/price fact predicate in an EU-Android-tying antitrust context. If a firm is extending its dominance in one market (say search) to additional markets (say OS and app store) in part via pricing one of the goods or services at zero to rapidly grow users when the service has known major fixed costs and significant incremental costs, an antitrust authority can conclude there is anticompetitive cross-subsidization and predatory pricing or dumping to kill off any actual or potential competition. It apparently has played out that way in the mobile OS market over the last seven years.
Simply, the Oracle v. Google case provides economic facts of cost and price that implicate a broader monopolization case than just the current more narrow and easily remedied EU-Android-tying case.
Requires a More Comprehensive Android Remedy. Bringing these cases together also means the potential discussed remedy of just prohibiting Google from anticompetitively restricting contracts of device manufacturers and mobile network operators would not be enough to restore competition, because Google could still anticompetitively maintain its dominance over multiple markets via predatory pricing and cross subsidization of a free Android OS.
3. Together these cases expose a clear Google anticompetitive pattern of behavior of abusing different sorts of APIs that could help expand the EU-Android-tying case into potentially a broader Google monopolization case.
Oracle v. Google spotlights how Google stole/scraped Java APIs that it easily could afford to license, for the purposes of creating a dominant free open-source mobile operating system and app ecosystem.
Many do not know that three different antitrust jurisdictions already have opposed Google’s abuse of advertising APIs as anticompetitive.
Consider what the the Canadian Competition Bureau did to stop Google’s abuse of APIs just last month: “As a result of an in‑depth investigation, the Bureau concluded that Google used anti‑competitive clauses in its AdWords Application Programming Interface (API) Terms and Conditions. The Bureau concluded that these clauses were intended to exclude rivals and negatively affected advertisers. Google has removed these clauses and has provided a commitment to the Commissioner not to reintroduce them (or others which have the same effect) for a period of five years.”
Consider that the U.S. FTC similarly found Google’s advertising API restrictions anticompetitive and Google committed in a settlement to stop the practice in 2013: “Google will remove restrictions hampering advertisers’ management of their ad campaigns across competing ad platforms.” …“Some FTC Commissioners were concerned that Google’s contractual conditions governing the use of its API made it more difficult for an advertiser to simultaneously manage a campaign on AdWords and on competing ad platforms, and that these restrictions might impair competition in search advertising.”
Google’s abuse of Adwords APIs also has been the subject of EU antitrust investigation since 2012.
The anticompetitive abuse of APIs in search advertising in the search antitrust case, combined with Google-Android’s de facto “scraping” abuse of Oracle’s APIs to dominate mobile search, mobile operating system, and Android App store, is additional evidence that the EU Android case may start as a straightforward tying case, but could expand into a broader monopolization case given how integral and pervasive Google’s anticompetitive behavior appears to be.
Put the Oracle v. Google and EU-Android cases together and they are more than the sum of the parts.
At a high level, together they expose Google effectively engaging in a classic “bait and switch” fraud on a global scale, in selling a “free and open” Android OS to expand Android’s dominance that was neither really open or really free.
Android really is a monopoly cross-subsidized, hybrid-open-proprietary, mobile operating system that effectively fixed a predatory price of zero over several years to ensure extension of their desktop search dominance into dominance of mobile search, Android OS, and Google Play.
From a holistic antitrust level, these cases together illuminate a predatory Google strategy to dominate mobile from the beginning to the ongoing, and to dominate most all aspects of the mobile space that rely on Google search, search advertising, mobile OS and Google Play.
These cases together are one of the rare instances where 1 +1 = 3.
Scott Cleland served as Deputy U.S. Coordinator for International Communications & Information Policy in the George H. W. Bush Administration. He is President of Precursor LLC, an emergent enterprise risk consultancy for Fortune 500 companies, some of which are Google competitors, and Chairman of NetCompetition, a pro-competition e-forum supported by broadband interests. He is also author of “Search & Destroy: Why You Can’t Trust Google Inc.” Cleland has testified before both the Senate and House antitrust subcommittees on Google and also before the relevant House oversight subcommittee on Google’s privacy problems.