OLYMPUS DIGITAL CAMERAKonstantina Bania from the European University Institute questions whether the commitments agreed to in the slow and inefficient Google antitrust investigation will in fact restore competition.

After dragging on the Google investigation for more than three years, the European Commission announced on February 5, 2014 that it was satisfied with the most recent package of proposed solutions and that it would therefore be moving towards a commitments decision (a commitments decision accepts the undertakings offered by the company under investigation and, as opposed to a prohibition decision, does not conclude on whether there has been or still is an infringement).

Failure to restore competition in a swift and effective manner

The Commission has opted for pursuing the commitment route because, as Commissioner for Competition Joaquín Almunia put it, fast moving markets particularly benefit from a “quick resolution” of the issues identified. The Microsoft ten-year saga is a good example illustrating why restoring competition at an early stage may be preferable to lengthy proceedings. Yet, the Commission opened its probe as early as 2010 and we are still waiting for the commitments made by Google to become legally binding. The Commission may have had to deal with complex issues (e.g. the practice of algorithm manipulation to favor one’s own services whilst excluding competing websites from web search is new to antitrust law), but moving at such a slow pace renders this procedure devoid of any benefit and violates the involved company’s right to have its affairs handled within a reasonable time (enshrined in Article 41 CFREU).

What about the proposed commitments?  

The issue of whether Google manipulated its algorithms to favor its own services and downgrade competitors has been the most debated allegation against the Internet giant. To alleviate the Commission’s concern that this practice may result in the exclusion of other websites from online search, Google offered (for a five-year period) to label promoted links to its own specialized search services so that users can distinguish them from natural search results, separate these links from other search results by graphical features (e.g. a frame), and display links to three competing search engines close to its own services, in a place that is visible to users.

Those opposing the acceptance of the above proposals raise some convincing arguments over why Google’s suggestions do not seem appropriate to address relevant anti-competitive concerns. For example, the proposal to display links to three competing services justifiably raises the question as to who will determine the promotional parameters. If it is Google’s responsibility, then that commitment would give it a wide margin of discretion thus, possibly falling short of resolving the issue of discriminatory choices for users. In addition, the undertaking to visually separate links to Google services from links to competing services may have an effect opposite to the one the Commission desires as it may induce the consumers to click on the results that are highlighted in a frame (i.e., Google services).

Finally, as the commitments currently stand, Google will not be required to abide by its own “host-crowding rule”. This rule forms part of a set of best practices that Google has designed over the years and is aimed at eliminating the duplication of results in order to safeguard diversity in the search. On the basis of the host-crowding rule, if, following a given query, there are many results in a single web directory, only the two most relevant results for that directory appear in the list of organic results. However, Google has not been applying this rule to its own services. Since May 2007, Google has been implementing a universal search model whereby, following a given query, the algorithm searches all its content sources, for instance Google News, Google Maps, YouTube, etc., ranks all the information in its possession and delivers an integrated set of results, which may be more than two per page.

In view of the above, there are strong doubts as to whether the proposed changes will manage to eliminate or neutralize the effects that the practice of search manipulation has had on competition.


It goes without saying that the ramifications of a search engine engaging in algorithm manipulation go well beyond the improper functioning of the market. This type of behavior is particularly harmful not only because it may lead to anti-competitive prices in the advertising markets, but also because in an environment where content is abundant, but attention is scarce, the process of selecting contents of interest to the online user is becoming increasingly important. 73% of European households are now connected to the Internet and search engines constitute the main entry gate. In other words, search engines are increasingly evolving into the new bottleneck for content access. The Commission, whose investigation into Google’s anticompetitive behavior has been neither diligent nor time efficient, seems to have largely undermined the (economic and non-economic) effects created by the new forms of artificial scarcity that digital intermediaries have been introducing in the media landscape.

This post gives the views of the author, and does not represent the position of the LSE Media Policy Project blog, nor of the London School of Economics.