Streetmap v Google: lessons for pending Article 102 TFEU cases (including Google itself)
On Friday last week, the High Court of England and Wales ruled on the dispute between Streetmap and Google (see here). It is a really interesting read, and one that shows that – whether or not one agrees with the outcome – courts can deal effectively with complex competition law matters (I am told that I should not jump to conclusions too readily: apparently Mr Justice Roth, author of the decision, is an active member of the Association of European Competition Law Judges and attended in that capacity the seminal lecture on two-sided markets that Alfonso gave in Uppsala 😉 ).
The obvious appeal of Streetmap v Google is that it raises pretty much the same issues that the Commission will have to address in its own – and pending – Google case (see here and here). In essence, Streetmap argued that Google’s prominent presentation of its mapping services amounted to an abuse of a dominant position. As many people have argued, the combination of Google Maps and the search results is a form of abusive ‘bundling’.
Mr Justice Roth did not seem very impressed with the arguments brought by Streetmap. He concluded that Google’s practice was not reasonably likely to have an appreciable anticompetitive effect and that, in any event, it was objectively justified. I, on the other hand, am very impressed with the decision. Mr Justice Roth’s analysis is penetrating and creative. Here and there, I have found arguments about which I had not thought before.
The decision provides particularly interesting lessons for Google and other ongoing Article 102 TFEU cases. I can think of the following:
- Effects need to be showed for some practices, not simply assumed. Post Danmark II made it clear that, as far as some practices are concerned, a likely anticompetitive effect must be shown. In such cases, assuming that the practice is capable of having such an effect is not enough. It would be necessary to examine the features of the relevant market and how, in that context, the practice would lead to foreclosure. Mr Justice Roth applies this principle and concludes – rightly in my view – that Google’s conduct is a ‘by effect’ practice.
- There must be a causal link between the abuse and the anticompetitive effect. This is one of the fundamental aspects of the ongoing Google case, and one that is often ignored. In fast-moving markets, the exclusion of rivals is not necessarily the consequence of an abusive practice. Rivals may not be able to adapt to changes in consumer demand. Their business model may be the relic of a past era. As a result, they would have been driven out of the market irrespective of the behaviour of the dominant firm (the counterfactual, again!).
Post Danmark II was clear in stating that a ‘by effect’ practice is only contrary to Article 102 TFEU where the ‘anti-competitive effects’ are ‘attributable’ to the dominant firm (para 47, emphasis added). Mr Justice Roth applies this principle and appears to take the view that Streetmap’s decline is not attributable to Google’s practice. He seems to suggest that it would have happened anyway.
The Commission faces a major challenge in this regard in the pending Google case. Assuming there has been a decline in the traffic towards some price comparison websites, the Commission would have to show, to the requisite legal standard, that this is the consequence of Google’s practices – as opposed to the consequence of the evolution of markets and, in particular, the rise of Amazon, eBay and others.
- There is confusion about the legal test that should apply to Google. Unsurprisingly, Streetmap argued that Google’s practice was a form of bundling. This position is controversial (see here). These days, consumers expect more than the proverbial ten blue links when they perform a search on Google. Consumers’ assumption is that, where relevant, other affiliated services (including maps, images and youtube videos) will also be displayed.
Is it possible to argue, against this background, that Google’s practice is a form of abusive bundling and/or that the practice is not objectively justified? The Commission conceded in its Guidance that two products are distinct only where ‘a substantial number of customers would purchase or would have purchased the tying product without also buying the tied product from the same supplier’.
Where, conversely, consumers would only obtain the tying product with the tied product, the practice is most probably objectively justified and as such a source of efficiency gains that benefits consumers.
Will the Commission depart from the Guidance in Google? What are the implications? Nicolas Petit has recently written an interesting paper on the impact of the Guidance on administrative discretion. The conclusion that would follow logically from Nico’s paper is that the Commission cannot depart from the approach to tying and bundling sketched in the document. I have written elsewhere that the Guidance is a pre-commitment device – a promise to act in a certain way – that cannot simply be disregarded.
- The rejection of the de minimis doctrine does not mean that it is not necessary to show an effect. In Post Danmark II, the Court of Justice refused to set a de minimis threshold. Some people interpreted this passage of the ruling as meaning that there is no need to show an anticompetitive effect in the context of Article 102 TFEU. This position is not correct, as I explained elsewhere. What the Court held in Post Danmark II is that, where an anticompetitive effect is shown, this effect will be appreciable. In any event, the anti-competitive effect – and the causal link between the practice and the effect – will have to be established, and not simply assumed.
Mr Justice Roth adds an interesting twist to this question. He claims that the conclusion of the Court in Post Danmark II is only valid where the abuse and the exclusionary effects take place on the same market. In the case of leveraging, it would be necessary to show that the anticompetitive effects on the non-dominated markets are appreciable. Not everybody will agree with this position, but the underlying reasoning and approach are, in my view, correct.