The Google Commitments – Testing Substantive Theories Through Remedy Discussion
Kluwer Competition Law Blog
July 4, 2013
Please refer to this post as:, ‘The Google Commitments – Testing Substantive Theories Through Remedy Discussion’, Kluwer Competition Law Blog, July 4 2013, http://competitionlawblog.kluwercompetitionlaw.com/2013/07/04/the-google-commitments-testing-substantive-theories-through-remedy-discussion/
Over the last two months, the Commission has been market testing commitments that Google has offered to resolve the Commission’s pending investigation. Many have made thoughtful comments (see for example here, here, and here). But there has also been criticism from the corner of some of Google’s competitors. I represent Google in the Commission’s proceedings and offer here some personal thoughts from a practitioner’s perspective.
The debate over Google’s remedies provides important insights for antitrust analysis. This debate is not only relevant for remedy design, but also exposes weaknesses in the complainants’ underlying substantive theories. Remedy design is not an after-thought at the end of a process that is abstracted from substance. It is intimately connected to substantive analysis. Reflection on remedies can therefore serve as a useful test for substantive complaints, especially in cases (such as the present one) that involve novel abuse theories.
Most of the public debate has focused on the design of Google’s search results. The complainants’ abuse theories consist of allegations that Google improperly favors what they call Google’s vertical search results. These complaints are directed against elements of Google’s search results design that group search results for a particular category and present them in dedicated formats.
For example, if a user asks Google for “restaurants in Brussels”, Google may display links to restaurants located in Brussels as a group together with a map that identifies the location of each restaurant. If a user asks Google for “William and Kate”, Google may display grouped links to news articles about the royal couple and a group of links to pictures in the form of image thumbnails. Similarly, Google has developed specialized forms of ads. For example, if a user asks Google for “digital cameras”, Google may display a group of specialized product ads that link directly to merchants selling digital cameras together with thumbnail images of cameras and product prices.
These specialized designs make it easier for users to select results that are most relevant for them and in this way enhance the quality of Google’s organic search results and ads. Other search services, such as Bing or Yahoo!, have adopted similar designs, which suggests that they also consider these designs to be quality enhancing and beneficial to users.
The European Commission, accordingly, does not question Google’s display of specialized search result groups or the use of different search result designs. The preliminary concerns that the Commission has identified relate to organizational links that Google displays as part of specialized result groups (e.g., as headers or footers) and that lead to additional results for a specific category (e.g., more restaurant results, more news results, more image results, or more merchant results). As explained in its memorandum, the Commission’s preliminary concern is that users may not be sufficiently “aware” of the nature of these links and that links to third-party vertical search services may not be sufficiently “visible” relative to these links.
Google’s proposed commitments squarely address the Commission’s preliminary concerns through a combination of several measures:
– Google will label the organizational links in question and will provide additional information to make users aware of their particular nature.
– For specialized search results that Google monetizes, Google will distinguish such links further through additional design elements.
– Also, for specialized search results that Google monetizes, Google will display three links to third-party vertical search sites alongside the organizational links in question. These remedy links will lead, where possible, to the results page of the third-party vertical search site for the same query that the user entered on Google.
These measures provide users with additional means to exercise an informed choice. Users are made specifically aware of the character of the links in question and users who want to access search results of third-party vertical search sites for queries that they enter on Google receive another prominent option to do so through the remedy links.
Complainants that have publicly criticized Google’s proposal do not seek a constructive solution for the Commission’s preliminary concerns. Instead, unlike the Commission, they challenge Google’s display of specialized search results more fundamentally. Some demand, for example, a restrictive “one size fits all” approach to search that would prevent adapting search result design intelligently for different result categories. Such demands want to turn the clock of search result design back to a basic list of blue links. Others demand that Google’s specialized search results be replaced by search results from third-party vertical search sites: If a user asks Google for restaurants in Brussels, Google should not show what it considers to be the most relevant restaurants, but a list of restaurants compiled by another site, thereby eviscerating the core value proposition that Google offers its users.
These demands do not fit the Commission’s preliminary concerns and as such disqualify themselves. But the complainants’ demands are not just defective in terms of remedy design. Their remedy demands also expose underlying flaws in their theories of abuse. The complainants’ various demands all come down to one proposition: Google should not be allowed to provide directly responsive results for queries that users put to Google or should not do so in a manner that makes it easy for users to find them.
But if Google shows its own directly responsive results for queries that users enter on Google, that is not “favoring” or anti-competitive conduct. It represents the very essence of what a search service does. The results identified by one search service are not the same as the results identified by another – they are not interchangeable commodities. Each search service differentiates itself and competes on the merits of its search results. Preventing Google from showing what it considers to be the most relevant, directly responsive results for a query would not promote competition. It would restrict competition and obstruct innovation.
What the complainants ultimately seek is free promotion. Free promotion that would, nota bene, come at the expense of sites that Google identifies as directly relevant to the query, such as restaurant sites that Google references in response to a restaurant query. Ultimately, this would be to the detriment of users. Users who enter restaurant queries on Google ask for Google’s assessment of the most relevant restaurants, not for the assessment of another site. Denying Google the ability to respond to these queries in the most direct and effective way would leave users with a lower quality service and an inferior user experience. There is no conceivable legal basis under EU competition law for theories that postulate such outcomes. The complainants do not identify an anti-competitive restriction of competition. They seek to instrumentalise competition law for their own ends, in ways that are neither in the interest of competition nor consumers.
Indeed, a recent judgment of the Landgericht Hamburg soundly rejected similar favoring theories in connection with an interim measures request against Google’s weather search results. The court observed that the complainants “have been participating and will continue to participate in Google Search as ‘free riders’. They demand favorable positioning without offering compensation”. The court concluded that the complainants have no right to raise such demands under competition law. The court stressed that “the prohibition of an abuse of a dominant position does not have as its objective to preserve outdated business models that cannot withstand change”. Google therefore cannot “be prevented from implementing innovation that it deems beneficial so that other companies […] can appear in the usual way in organic search results”.
The FTC in the US Google investigation reached similar conclusions, noting that “Google adopted the design changes that the Commission investigated to improve the quality of its search results” and emphasizing that “product design is an important dimension of competition and condemning legitimate product improvements risks harming consumers”.
In sum, the complainants rely on theories that are misconceived as a matter of legal substance and raise demands that do not seek to resolve the Commission’s preliminary concerns. Google’s proposed commitments, on the other hand, offer a constructive and effective solution for these concerns. That solution comprehensively addresses the Commission’s preliminary concerns while preserving the efficiencies of Google’s search designs and Google’s ability to innovate. It is a testament to the Commission’s and Google’s sophistication and professionalism that, in spite of differences, both have worked to find a practical and efficient outcome.