Brussels à jour

Google/Fitbit – A New Horizon for Behavioral Remedies in EU Merger Control?

In the wake of the European Commission’s 2014 decision in Facebook/Whatsapp, [1] data has taken an increasingly prominent role in merger analysis. As data-centric theories of harm take the center stage in “big data” deals, practitioners and academics have raised the issue whether these novel theories of harm require the European Commission to re-think its approach to remedies, which has traditionally strongly favored structural over behavioral remedies. [2] While there is some precedent such as, e.g., Intel/McAffee [3] and DriveNow/Car2Go [4] in which the European Commission has agreed to accept behavioral remedies, data-driven companies will closely follow the ongoing investigation of Google’s proposed acquisition of Fitbit, anxious to learn whether the European Commission will further open the door for behavioral commitments to resolve concerns around the collection, merging and use of data.

In Google/Fitbit, [5] the European Commission is concerned about Google’s ability to use Fitbit’s data, in particular health and location data of Fitbit users, for advertising purposes and thus to strengthen its market position in the online advertising space. An additional concern is related to Google potentially having an incentive to prefer Fitbit over competing healthcare wearable producers when it comes to access to Google’s Android ecosystem. While the latter concern resembles a more traditional foreclosure theory of harms, the former appears to be more innovative in the sense that it considers the competitive effects that may be associated with the acquisition of control over potentially valuable data sets. In order to resolve these concerns, Google has offered to commit not to use Fitbit’s data in targeting advertisements and to separate Fitbit’s dataset from its own by way of ring-fencing the data, with Fitbit being the sole legal controller and no access to the data by Google, for a time period of five years. According to press reports, the European Commission is not satisfied with the commitments offered so far and is considering to request from Google that it will not use Fitbit’s data to enhance its search and to grant third parties equal access. The European Commission and Google have until 4 August 2020 to agree on a data remedy in phase 1. Otherwise the European Commission will open an in-depth phase 2 investigation.

The European Commission’s Remedies Notice expresses a clear preference for divestiture remedies. [6] This is because, like a merger, they affect the very structure of the market and thus are capable of directly addressing the cause of harm. They also solve competition concerns in the long term, and they provide administrative efficiency without need for monitoring. However, technology mergers often confront regulators with novel theories of harm for which divestiture remedies may either not offer a suitable resolution or be disproportional, particularly when the company value is mainly driven by its possession of or access to data.

The European Commission’s Remedies Notice (still) offers a useful starting point for data-driven companies contemplating potential remedies for their deals. The Notice, first and foremost, reminds companies that, as divestitures, behavioral remedies must eliminate competition concerns entirely, must be comprehensive and effective from all points of view and must be capable of being implemented effectively within a short period of time. This is true irrespective of whether the commitments aim at granting access to goods or services (or data), removing of links within competitors, or resolving competition concerns in other ways. A re-read of the European Commission’s 2005 Merger Remedy Study [7] may also still prove useful, particularly in data-driven markets. According to the Study, parties should be transparent about the information needed by the European Commission to assess the viability of a remedy, particularly in dynamic markets where future developments are difficult to predict and the parties will have a better understanding of the effects a given remedy might have. Specifically with respect to behavioral remedies, the wording of the commitments should be clear, precise and comprehensive, and there should be an effective monitoring and dispute resolution mechanism. Although the Study dates back to 2005, it remains relevant today. In big tech and data mergers, market dynamics can be very hard to predict. And given the flexibility of some tech business models, enforcement may even be more difficult than it was in 2005.

Recent precedents offer further guidance. These suggest that, in particular where the concerns relate to data access or a platform which should remain accessible to the target’s competitors, commitments to grant access stand a good chance to win the European Commission’s support.

  • In Intel/McAffee, [8] the Commission had doubts whether competing IT-security services might at least partly be banned from the market if Intel, being a chipset producer with a high market share, were to acquire McAffee, which provided such IT-security services. The European Commission considered it likely that after the merger Intel, by way of technical tying, interoperability hampering and commercial bundling strategy, could exclusively tie McAffee’s security solutions to its computer chipsets. [9] To resolve the European Commission’s concerns, Intel committed to providing competing IT-security companies all necessary technical information to use Intel’s chipsets the same way as McAffee [10] did. This commitment was made for a duration of five years and overseen by a monitoring trustee. [11]
  • In DriveNow/Car2Go [12] the parties, active in the market for car sharing services and combined multimodal mobility apps, committed to granting (i) application programming interface (API) access to competing mobility aggregator platforms and (ii) access for all interested car sharing providers to the parties’ combined multimodal app “moovel”. [13] Both commitments were valid for two years after project rollout but could be extended for further two years if no competitor stepped into the market. [14] These commitments were likewise overseen by a monitoring trustee. [15]

Depending on the specific theory of harm, other behavioral commitments may also be suitable in the context of access to data. Instead of committing to allowing competitors to access a certain platform, companies could commit to allowing competitors access to their own proprietary data. When Google acquired airfare pricing and shopping software developer ITA software, the DOJ requested a behavioral commitment by Google to allow other search engines access to ITA software’s data for their respective flight searches and shopping software, which Google entered into for a period of five years. [16]

Google’s proposed commitments in Google/Fitbit differ from these precedents in that they are not aiming at allowing competitors to use a specific platform or data, but would have Google commit not to use Fitbit’s data. This seems comparable to a commitment not to raise prices post-merger or to a “hold separate” commitment. If the European Commission were to accept Google’s commitments this might open a much welcomed door for deals for which structural – or even more traditional behavioral – remedies may be unfeasible.

In a market test sent out to other makers of wearable devices, app developers and other online service providers as well as healthcare providers by the European Commission, besides covering questions on the type of data Fitbit has in its possession and Google’s potential use of it for advertising purposes, the European Commission also asked recipients whether the foreseen monitoring mechanism through a trustee would be sufficient and whether the proposed time frame of five years is sufficient. Enforcing behavioral remedies is generally more complicated than structural remedies. With the business environment constantly changing over the term of the remedy and potentially also the business model being adapted over time, monitoring the adherence to the terms of the remedy can be challenging. Apparently, this issue has been raised by recipients of the market test, who questioned the monitoring trustee’s ability to effectively ensure Google’s compliance with the commitments. Private enforcement by way of an expedited arbitration procedure may be one possible route to be explored.

Irrespective of how the European Commission will come out on Google/Fitbit, going forward the focus may shift from whether a data remedy is suitable at all to how such a remedy can be designed to be effective. Devising an effective enforcement mechanism, which does not require constant monitoring by the European Commission and/or a monitoring trustee may be one of the keys to success. Such mechanism should ideally be self-enforcing and/or could provide for a fast-track dispute resolution procedure for any issues that may come up over the time of the commitment.

  1. Commission Decision of 3 October 2014, Case No COMP/M.7217 – Facebook/WhatsApp.
  2. Commission notice on remedies acceptable under Council Regulation (ES) No 139/2004 and under Commission Regulation (EC) No 802/2004 (2008/C 267/01), para. 17, 61.
  3. Commission Decision of 26 January 2011, Case No COMP/M.5984 – Intel/McAffee.
  4. Commission Decision of 7 November 2018, Case No COMP/M.8744 – Daimler/BMW/Car Sharing JV.
  5. Commission, Case No. COMP/M.9660 – Google/Fitbit.
  6. Commission notice on remedies acceptable under Council Regulation (ES) No 139/2004 and under Commission Regulation (EC) No 802/2004 (2008/C 267/01), para. 17, 61.
  7. European Commission, Merger Remedies Study 2005, retrievable at: https://ec.europa.eu/competition/mergers/legislation/remedies_study.pdf.
  8. Commission Decision of 26 January 2011, Case No COMP/M.5984 – Intel/McAffee.
  9. Commission Decision of 26 January 2011, Case No COMP/M.5984 – Intel/McAffee, paras. 291 et seq.
  10. Commission Decision of 26 January 2011, Case No COMP/M.5984 – Intel/McAffee, paras. 337 et seq.
  11. Commission Decision of 26 January 2011, Case No COMP/M.5984 – Intel/McAffee, paras. 344.
  12. Commission Decision of 7 November 2018, Case No COMP/M.8744 – Daimler/BMW/Car Sharing JV.
  13. Commission Decision of 7 November 2018, Case No COMP/M.8744 – Daimler/BMW/Car Sharing JV, para. 338.
  14. Commission Decision of 7 November 2018, Case No COMP/M.8744 – Daimler/BMW/Car Sharing JV, para. 339.
  15. Commission Decision of 7 November 2018, Case No COMP/M.8744 – Daimler/BMW/Car Sharing JV, para. 339.
  16. United States District Court for the District of Columbia, Judgement of 5 October 2011, Case 1:11-cv-00688-RLW – U.S. v. Google Inc. and ITA Software, Inc.
 
 
» DOWNLOAD Newsletter (PDF)

FOLLOW US

Don’t miss any edition of our Brussels à Jour Newsletter.
You can simply follow the hashtag #Brusselsajour on LinkedIn to make sure you receive our updates in your feed.