Android Remedies: Tearing Down the Wall?

November 2018

CPI Europe Column edited by Anna Tzanaki (Competition Policy International) & Juan Delgado (Global Economics Group) presents:

Android Remedies: Tearing Down the Wall? By Simonetta Vezzoso (Trento University)1

Click here for a PDF version of the article



The control of abusive practices has long been touted as the most arduous area of competition law, fittingly compared to climbing Mount Everest.2 Once the anticompetitive practices have been proven, though, devising appropriate remedies might at times vividly recall the challenges posed by another mythical mountain, namely the K2. While not the highest mountain in the world, climbing the K2 is generally considered harder.

The two recent Google Decisions might well exemplify many of these difficulties. Following long and complex investigations, the European Commission (“Commission”) imposed on Google a whopping total fine of €6.7 billion. While Google has already applied for annulment of both infringement decisions, the antitrust eyeballs are now fixated on the respective remedial phases of the proceedings. Neither in the Google Shopping Case, nor in the Android Case, the Commission dictated a specific remedy but instead left it up to Google to propose measures that can effectively bring the infringements of Article 102 TFEU to an end.

The Google Shopping decision was issued in July 2017.3 The Android Decision has been announced on 18 July 2018, and while the public is still waiting for the pleasure of reading it, Google has already started implementing significant changes to the Android ecosystem as a consequence of its proposed remedies to the alleged anticompetitive practices.

The legal framework for remedies to abuses of dominance under EU law is set by Council Regulation 1/20034 and the relevant case law. The Commission can impose behavioral or structural remedies which are proportionate to the infringement committed. Moreover, the purpose of a remedy in an infringement decision is “to bring the infringement effectively to an end.” When it comes to effectiveness, it is of course a matter of degree. A remedy worth its name should at the very least put an end to the types of conduct considered abusive. Put differently, the dominant undertaking should be prevented from continuing to engage in the future in the conduct that has been found to be abusive (the “sin no more” remedy). While this is obviously crucial, a more comprehensive and effective remedy should also consider stopping the effects of the abusive behavior and therefore, truly eliminating the consequences of the competition law infringement. This would require restoring the competitive process that would have prevailed but for the abusive practice(s). This does not mean trying to go back in time, especially in rapidly changing markets. Efforts should be made, however, to ensure that the market is as competitive as it would have been, if not for the legal violation. Depending on the circumstances of the case, this could mean re-creating the conditions for smaller competitors to compete and innovate. In the most extreme scenarios, structural remedies could be necessary, like divesting parts of the business of the dominant undertaking to substitute for the competitive threat successfully eliminated by the abusive practices of that dominant undertaking. The undeniable challenge here is how to construct a suitable counterfactual. While difficult, this should not be a suitable excuse for tying one’s hands and not engaging the enforcer’s best efforts aimed to the accomplishment of the all-important task of restoring the competitive process in the interest of consumers. This is also to avoid a dangerous externality in the form of the otherwise credible message that infringing competition law pays, especially in winner-takes-all markets.5


The little we know about the Android remedies

Competition Commissioner Vestager made clear that Google was required to bring the conduct effectively to an end within 90 days of the decision.

At a minimum Google has to stop and to not re-engage in any of the three types of practices considered abusive, namely the tying practices, the exclusivity payments and the obstruction to the development of Android forks.6 “In other words,” explains the Commissioner “our decision stops Google from controlling which search and browser apps manufacturers can pre-install on Android devices, or which Android operating system they can adopt.” Moreover, the Decision imposes Google an obligation “to refrain from any measure that has the same or an equivalent object or effect as the practices found abusive”.7  Finally, the Commission specifies that the “decision does not prevent Google from putting in place a reasonable, fair and objective system to ensure the correct functioning of Android devices using Google proprietary apps and services, without however affecting device manufacturers’ freedom to produce devices based on Android forks.”

If the Google Shopping Decision can serve here as a template8, Google in the Android case had until the end of September (60 days) to come up with a remedy proposal. After that, it is realistic to suppose that a number of substantial exchanges between the Commission and Google took place.9 On 28 October 2018 Google started implementing its chosen remedy.

Following Google’s public announcement illustrating the core of the proposed remedies on 16 October 201810, many external commentators have strongly criticized them (but the complainants in the Android Case have remained remarkably quiet).11

Google’s public announcement focuses on a series of changes to the contracts with device manufacturers. First, Google unbundles the Google Search and the Chrome browser apps from the group of other Google mobile applications (G-Suite). This means that, for instance, if the device manufacturer decides to install Google Play (Android’s app store) on her devices, she does not need to install Google Search and Chrome. She is required, however, to install the other apps comprised in the G-suite.

What has also attracted much attention, and early criticism, is another aspect of Google’s proposed remedy package, namely Google’s intention to introduce a new paid licensing agreement for smartphones and tablets wanting to use Google mobile applications. As Google puts it, “(s)ince the pre-installation of Google Search and Chrome together with our other apps helped us fund the development and free distribution of Android, we will introduce a new paid licensing agreement for smartphones and tablets shipped into the EEA.” How much handset manufacturers will have to pay is still unclear. According to documents purportedly seen by some journalists, Google will charge handset makers wishing to use its mobile application package as much as $40 in Europe, depending on the country and the device’s pixel density. Fees would apply to devices activated on or after February 1, 2019.12

On the other hand, the Google Search app and Chrome will be licensed separately and do not foresee a licensing fee. At the same time, Google offers new commercial agreements to device manufacturers for the non-exclusive pre-installation of Google Search and Chrome. In its offering to device manufacturers Google will now compete with app developers in the browser and search markets. This means that a device manufacturer can choose, based on a number of commercial considerations, to pre-install browsers and search apps alternative to Google Search and Chrome, like for instance Bing, DuckDuckGo, Qwant, Firefox Focus, Samsung Internet, Microsoft Edge, Puffin and Opera.

Another modification in the contracts with handset manufacturers relates to the forking restriction. Manufacturers selling handsets that carry the G-suite may also build and distribute forked smartphones. This means, for instance, that Samsung, Motorola, Sony, HTC, LG could start offering Android forks. In this respect, the Commission wrote that it had “found evidence that Google’s conduct prevented a number of large manufacturers from developing and selling devices based on Amazon’s Android fork called ‘Fire OS.13’”

Besides the modifications to the contracts with handset manufacturers succinctly described by Google in its blog post, further bits and pieces of a broader reshuffling of the Android ecosystem are starting to emerge from the ever attentive business press. Thus, for instance, Google will be offering in the European Play Store ‘Google Voice Action Services’. While this app shares the same voice-based capabilities of Google Assistant, it does not provide any web search capability. Users who buy devices from handset manufacturers that do not pre-install Search will still be able to use it.14 It is also interesting that Google is introducing for the first time into its contracts with handset manufacturers the requirement that manufacturers install security patches in a timely matter. It is not clear, however, if the contracts foresee penalties if manufacturers fail to meet Google’s security requirements.15


Of castles and fortresses

For all the current digital hype, the Android case brought by the European Commission is still largely Old School. In particular, the conduct considered anticompetitive by the EU competition enforcer was not embedded in complex code, or black boxes, but was in rather plain sight, namely in the written contracts between Google and device manufacturers.

One of the possible narratives of the theory of harm in the Android case is that Google built a moat around its castle, the search engine, and that some of the tactics and constructions employed were abusive under Article 102 TFEU. Some of those tactics, apparently, comprised extending the fortification to reinforce the castle’s defences.16 Restoring competition in the markets affected by the abusive types of conduct would seem to require reducing some of those defences, like lowering the walls of the fortress,17 including lowering barriers to entry. While promoting competition and innovation within the ecosystem, this could benefit some actual or potential competitors of Google. The fact that these competitors may or may not choose to take advantage of the new opportunities opened up by the remedies is not in the competition enforcer’s control, especially in fast moving digital markets. In some instances, however, the fact that new competition does not emerge and grow is a clear indication that the remedy package could have been better engineered. Ideally, consumers should benefit directly from remedies carefully devised to increase competition and choice, and not only from the overall “distraction” that antitrust enforcement causes to the dominant undertaking.18

The Google Cases, as well as the still to these days passionately debated Microsoft Cases19,make it abundantly clear that, going forward, the remedial phase of a competition law proceeding deserves increased attention.

For once, remedies need to be proportionate, meaning among other things that they must be suitable in addressing the harm at issue.20 In markets where technology changes rapidly, however, addressing the “harm at issue” is rarely straightforward. Thus, for instance, an adequate remedy21 in these markets is necessarily forward-looking, and directly addressing the harm as it materialized in a drastically different technological environment is unlikely to be effective. In this respect, an intrinsic tension may emerge between remedies that are proportionate, or suitable, in the sense that they address the conduct that is found to be illegal, and remedies that are truly effective in restoring competition. This is especially the case in today’s entrenched digital markets, where attacks that do not merely weaken the fortress defenses but that may be successful in breaching the walls of the fortress are notoriously rare.22 This is another significant, albeit often disregarded, way in which competition law is challenged in the era of digitization.

A further remedy-related aspect that needs careful attention is that the two engines driving the current economic and societal changes, namely software and data, are highly malleable and transformative. It follows that competition authorities might encounter serious difficulties in devising mechanisms allowing for their full potential to be preserved in the interest of consumers, while providing effective remedies against anticompetitive practices. Moreover, there might be an ever increasing information asymmetry between high tech-savvy undertakings and competition enforcers, entailing a serious risk of successful strategizing and gaming by the former, especially in the remedial phase. Understandably, competition authorities are reacting to this and other new challenges by increasing their in-house technological competence. This not only might improve the detection of anticompetitive behavior, but also strengthen later phases of competition enforcement, including the restoration of competitive processes where their functioning has been hampered by anticompetitive effects.




1 Trento University, Department of Economics and Management. Professor for competition policy and intellectual property law and Senior Researcher,

2 Andreas Mundt, Closing speech at the 13th International Competition Network Conference in Marrakech, April 25, 2014.

3 European Commission, CASE AT.39740, Google Search (Shopping), 27 July 2017, available at

4 Council Regulation (EC) No 1/2003 [2003] OJ L 1/1.

5 See Cyril Ritter, How Far Can the Commission Go When Imposing Remedies for Antitrust Infringements?, Journal of European Competition Law & Practice, 2016, Vol. 7, No. 9, 587, 589.

6 See EC, Statement by Commissioner Vestager on Commission decision to fine Google €4.34 billion for illegal practices regarding Android mobile devices to strengthen dominance of Google’s search engine, 18 July 2018. See also Simonetta Vezzoso, Android and Forking Restrictions: On the Hidden Closedness of ‘Open’, Market and Competition Law Review, 2018, Forthcoming.

7 See EC, Antitrust: Commission fines Google €4.34 billion for illegal practices regarding Android mobile devices to strengthen dominance of Google’s search engine, Press Release, 18 July 2018.

8 CASE AT.39740 Google Search (Shopping), 27 June 2017.

9 See with regard to the Google Shopping remedies Bo Vesterdorf and Kyriakos Fountoukakos, Appraisal of the Remedy in the Commission’s Google Search (Shopping) Decision and a Guide to its Interpretation in Light of an Analytical Reading of the Case Law, Journal of European Competition Law & Practice, 2017, p. 3 (“Google’s remedy is not subject to a formal approval process (although the authors would note that, by having presumably carefully reviewed, debated, and accepted Google’s 60-day and 90-day notifications, arguably the Commission has implicitly endorsed the remedy as a matter of principle”)).

10 Google, Complying with the EC’s Android decision, 16 October 2018,

11 Bloomberg, Europe’s Attacks on Google Are Backfiring, 18 October 2018.

12 Cfr. The Verge, Google app suite costs as much as $40 per phone under new EU Android deal, 19 October 2018.

13 EC, Press Release, note 7 above.

14 See 9to5Google, Android phones in Europe without Google Search will have a basic voice ‘assistant’, 24 October 2018.

15 See The Verge, Google mandates two years of security updates for popular phones in new Android contract, 24 October 2018.

16 See Nicholas Economides and Ioannis Lianos, The Quest for Appropriate Remedies in the Microsoft Antitrust EU Cases: A Comparative Appraisal, in Luca Rubini (Ed.), Microsoft on Trial: Legal and economic analysis of a Transatlantic Antitrust Case, Edward Elgar Publishing, 2010, 393, 461.

17 Carl Shapiro, Microsoft: A Remedial Failure, Antitrust Law Journal, Vol. 75, No. 3 (2009), 739, 746 ff.

18 Recode, Microsoft executive Brad Smith: ‘If you create tech that changes the world, the world is going to want to govern you.’, 29 May 2018.

19 See for a late example the FTC Panel titled “What Can U.S. v. Microsoft Teach About Antitrust and Multi-Sided Platforms?”, 16 October 2018, available at

20 Ritter 2016, note 5 above, 594.

21 Ibid. (“First, the remedy has to be ‘appropriate’ (or ‘adequate’, or ‘suitable’), in the sense of actually addres-sing the harm at issue—and not some other kind of harm—and addressing it fully, or ‘effectively’”).

22 See Shapiro 2009, note 17 above, 747.


More Articles More Articles More Articles More Articles More Articles