In May 2020, Google integrated its free video conferencing service Meet into its free client mail service Gmail as a pre-installed service that can only be hidden, not deleted, by the consumers by modifying the settings.

This conduct could be assessed in competition law as an abuse of dominance, namely a tying sensu stricto, consisting of purchasing a product A on the market where the undertaking owns a dominant position, subject to the purchase of a product B. In the Google Meet case, it would mean that Gmail would not be available without Meet. Abusive tying is prohibited under Section 4(2)(e) of Indian law as well as under Article 102(2)(d) of European law. For tying to be abusive, several common elements to the two regimes are required, such as notably the presence of coercion, dominance in the tying product market (product A), two separate products and an anticompetitive outcome. This note will focus on the coercion element being at the centre of the Indian Google Meet decision. While the Competition Commission of India (CCI) ruled that Google’s practise was not anticompetitive, its reasoning may not hold before the European Commission (EC) as the degree and the perspective of the coercion differ. Furthermore, the CCI’s approach of coercion raises some concerns over the risks of type II errors it implies in digital tying cases.

 

India vs EU: Coercion or no coercion?

 While the CCI found that the coercion element of tying was missing, the EC would unlikely reach the same outcome.

The CCI ruled that Google’s conduct did not appear to violate the prohibition of tying as no coercion was exercised upon consumers. It based its decision on two arguments. First, there is no obligation to use Meet for consumers, and the use of competing video conferencing services is authorized without needing to fear adverse consequences. Second, Meet is available without the Gmail ecosystem as the consumer needs a Google account to use it, not necessarily a Gmail ID.

This reasoning could fundamentally differ at the European level following the established case law in tying cases.

As regards the first argument, the EU endorses a lower degree of coercion to demonstrate a tying abuse. Indeed, the EU General Court stated in Microsoft stated that coercion could still exist when the consumer “is not required to use it or is entitled to use the same product supplied by a competitor of the dominant undertaking”. The same approach was reaffirmed later by the EC in its Android decision.

With respect to the second argument, the EU seems to adopt an opposite perspective in the coercion assessment. While the CCI assessed the availability of Meet without Gmail, the European regime of tying looks at this relationship from the other side in examining the availability of the dominant product, Gmail, without the tied product, Meet. This derives from the notions of tying sensu stricto and leveraging, as well as from the requirement of dominance on the tying product market stated in the Article 82-related Commission’s Guidance. All of this refers to using a dominance on the market A to tie this product A with a product on another market B where the undertaking wants to stretch its market power. The availability of Meet without Gmail would only be considered in the European regime in a case of pure bundling; the latter being the tying of two products with none of them being available as standalone. But in the Google Meet case, the tying sensu stricto would have been sufficient to reach the coercion requirements as the EC stated in Microsoft that coercion was present even though the tied product could be easily hidden. Thus, in integrating Meet into Gmail without the possibility to remove it (but with the possibility to hide it), the EC would likely consider that Google’s conduct prevents its consumers from benefiting from Gmail as a standalone product.

 

CCI’s reasoning and its implied risks of type II errors

 The CCI’s reasoning seems flawed on at least four aspects, namely the non-consideration of the consumers’ inertia, the perspective issue, the narrow perception of the ecosystem and the degree of product’s availability. Each of them could lead the competition authority to a type II error as this latter could clear conducts having anticompetitive effects.

First, the CCI did not even consider the role of consumers’ inertia in the pre-installation scheme. Yet, in its assessment of the digital market, the EC showed its considerable magnitude and its impact on competition. In Microsoft, it stated that “while downloading is in itself a technically inexpensive way of distributing media players, vendors must expend resources to overcome end-users inertia and persuade them to ignore the pre-installation”. Consumers’ inertia was also at the core of the Google Shopping decision, where the EC demonstrated that only 1% of consumers were looking at the second page of Google search results. In this case, one click away seems too far away for the vast majority of the consumers. Hence, although there is no formal obligation to use the pre-installed service, the outcome of the conduct will be similar due to consumers’ inertia.

Second, the CCI’s perspective to assess coercion is inconsistent with the strategy at the core of tying abuse, i.e. leveraging. In tying cases, the competition authority should examine the conduct from the dominant market perspective as the undertaking tries to stretch its market power into an adjacent market. The focus should thus be the availability of Gmail without Meet instead of the opposite. Otherwise, it would amount to negate the tying sensu stricto purely.

Third, the CCI’s consideration of the platform ecosystem is too narrow. Instead of thoroughly assessing the Google account ecosystem, the CCI chose to focus its attention on the Gmail ecosystem and looks at the Google account ecosystem only from a linear perspective (see Table 1). But an incomplete answer derives from a limited question. Therefore, in assessing if a Gmail address was needed to use Meet, the CCI could only have an incomplete vision of the entire ecosystem and the relation between the two services. The broader question of the availability of Meet without Gmail would potentially have resulted in another outcome. This narrow vision of the ecosystem can have led the CCI towards a type II error in ruling that Meet was available without Gmail. Although using Meet requires a Google account available without a Gmail address, the creation of a Google account automatically gives the consumer a Gmail address. Hence, the narrow consideration of the ecosystem seems to lead to flawed outcomes as a Gmail address will automatically be created if the consumer wants to use Meet. The availability of Meet without Gmail seems so quite constrained.

Table 1

Fourth, the CCI endorses a low threshold regarding the degree of product’s availability in tying cases. Indeed, the use of Meet is incredibly limited for consumers without a Gmail/Google account as they can use it only if they are invited to participate by a host benefiting from a Google workspace account and that they participate via platforms other than mobile phones. Given the reduced possibility to use Meet without Gmail, the availability of Meet as standalone – and thus, the consumer’s choice – seems rather theoretical.

 

Conclusion

 Although the Indian and the European competition authorities face the same conduct and examine it under the same category of abuse, namely tying, they endorse diverse approaches concerning the coercion aspect. This CCI’s decision highlights some of these differences between the two regimes, such as the perspective as well as the degree of coercion required to reach the threshold of abusive tying. While the EC adopted a strict policy, the CCI set a (too?) high threshold for coercion in tying cases. It did so by failing to consider  consumers’ inertia, by adopting an inconsistent perspective of the conduct, by narrowly circumventing the ecosystem assessment and by lowering the degree of product’s availability required. However, each of these factors could separately lead to a type II error in antitrust enforcement as it would significantly impact the finding of coercion, one of the required elements for tying to be considered anticompetitive.


________________________

To make sure you do not miss out on regular updates from the Kluwer Competition Law Blog, please subscribe here.

image_pdfimage_print

Leave a Reply

Your email address will not be published. Required fields are marked *