[Harshit Upadhyay and Sanigdh Budhia are third-year B.A. LL.B. (Hons.) students at Gujarat National Law University, Gandhinagar]
The underlying idea of a majority of platforms in the digital market has been to connect two or more completely different sets of users with differing motivations together, where the platform only acts as an intermediary between them. For instance, Ola and Uber connect users seeking transportation with drivers through their online platforms. This gives rise to the multisided nature of these platforms, where they simultaneously participate in multiple connected markets. The multisided nature of these platforms makes it difficult to delineate relevant markets for these platforms under Section 2 (r) of the Competition Act, 2002. However, while delineating relevant markets, there cannot be a uniform approach since multisided platforms can further be classified into ‘transaction’ and ‘non-transaction’ platforms. While the Competition Commission of India (“CCI”) recently, in the BookMyShow case, recognised the multisided nature of the platform, it is yet to acknowledge that different approaches need to be adopted for the delineation of different types of multisided platforms. Against this backdrop, this article argues that the Indian competition regulator should consider the nature of different multisided platforms while delineating relevant markets. Further, the authors suggest that the approach taken by the US Supreme Court in Ohio v. American Express can be incorporated into the Indian competition regime.
Understanding Multisided Platforms
Multisided Platforms
There are multiple overlapping definitions of multisided platforms in contemporary literature. The debate regarding the same remains unresolved. However, there is a wider agreement on certain components in all the definitions. The four primary elements in all of these are the existence of (i) the platform; (ii) users brought together by the platform; (iii) the pricing structure applied by the platform; and (iv) feedback effect, that is, decisions of the users on one side of the market impact the other side. Among the oldest examples of the multisided platform could be newspapers with readers and advertisers.
Distinguishing Transaction and Non-Transaction Platforms
The impact of the feedback effect can vary based on the nature of the platforms and needs to be assessed individually, taking into consideration the nature of the platform. The multisided platforms can be classified into two groups: (i) transaction platforms and (ii) non-transaction platforms. In transaction platforms, there is a single simultaneous transaction between the participants, for example, on online ticket booking platforms. While in non-transaction platforms, a single simultaneous transaction is not observable, for instance, in social media applications. The existence and extent of the feedback effect make a crucial impact on the behaviour of these platforms in the market and is a key determinant of their market position.
Delineation of Relevant Market in Ohio v. American Express
In Ohio v. American Express, it was contested that the anti-steering clause in agreements between American Express (‘Amex’) and businesses that accept Amex credit cards is anti-competitive. Amex facilitated credit-card transactions between consumers and merchants and thus was running a transaction platform.
The court, in this case, while delineating the relevant market, took an unconventional approach. Instead of completely relying on the substitutability test, i.e., finding the complete substitute for Amex, it distinguished between the nature of the platforms. It distinguished transaction platforms from non-transaction platforms and found that Amex is a transaction-based platform. Therefore, one integrated market shall be defined for the same. The court further propounded that in the case of non-transaction platforms, two distinct connected markets should be delineated because of their different typology.
The Amex Court based its reasoning on the article by Lapo Filistrucchi, Damien Geradin, Eric van Damme and Pauline Affeldt. Filistrucchi and others in their article have explained how in a two-sided transaction platform, say credit-card, a transaction can occur only if both the merchant and the cardholder simultaneously agree to use the same network. This results in network effects on both sides of the market in the completion of the transaction. A market could be treated as one-sided if the impact of indirect network effects and relative pricing on one side of the market is minor, like in newspapers that sell advertisements. In newspapers selling advertisements, newspaper readers are largely indifferent to the amount of advertising that a newspaper contains, and hence, the market for newspaper advertising behaves like a one-sided market.
The Indian Approach
In Transaction Platforms
Before the BookMyShow case, the CCI did not completely factor in the interdependence of different sides of the market on each other. The CCI used to delineate the market based on the context in which the case arose, as seen in FHRAI vs MakeMyTrip, where the market was delineated from the perspective of the hoteliers.
However, in the BookMyShow case, the feedback effect was implicitly recognised in transaction platforms by giving due consideration to both sides of the market, i.e., both consumers and multiplexes. Further, while assessing dominance, the CCI also considered the dependence of one side of the market on the other, where dominance on the side of consumers allows BookMyShow to enter into exclusive agreements on the multiplex side. Thus, the CCI delineated one integrated relevant market of the market for online intermediation services for booking of movie tickets in India.
In Non-Transaction Platforms
The approach adopted by CCI till now in non-transaction platform cases like Matrimony.com v. Google (‘Matrimony Case’) and Umar Javeed v. Google (‘Google AppStore Case’) is analytically incomplete. The CCI in the Matrimony case delineated two separate relevant markets for both sides of the platform. CCI based its reasoning on substitutability test, opining that the online general web search services (consumer side market) were not substitutable with search advertising services (advertiser side market). While in the Google AppStore case, the CCI delineated the market as the market for licensable mobile operating systems for smart mobile devices and the market for app store for android OS. Here, the CCI again did not consider the multisided nature of the app store, which connects smartphone users with app developers. Similarly, in XYZ v. Alphabets Inc., where the CCI again delineated the market for Google AppStore, it did not factor in its multisided nature.
The CCI, in the Matrimony case, should have considered the nature of Google as it provides different products and services to consumers on different sides of its platform. It failed to analyse this complex inter-connectedness between different sides of Google, even though it recognised the two sides, i.e., consumer and advertiser. However, later in cases concerning Google AppStore, the CCI completely failed to take into account the multisided nature of the platform.
Distinguishing the Market Delineation Based on Typology of Platform
Delineation of the relevant market should be based on the typology of the platform. Transaction and non-transaction platforms face differing nature of competitive constraints and feedback effects and hence should be dealt with differently.
Differing Competitive Constraints
Transaction platforms only compete with other similar transaction platforms, unlike non-transaction platforms, which face different competitive constraints in connected markets. For example, in newspapers, from the perspective of advertisers, the relevant competitors include players in the television market, while the same is not true from the perspective of readers. But in transaction platforms like mobile payments apps, the primary competitive constraint comes from other similar apps.
Feedback Effect
The importance of the feedback effect of transactions in different types of platforms is also different. The feedback effect in the non-transaction platform is generally uni-directional, while the same is not the case with transaction platforms. For instance, in newspapers, feedback effects are stronger from the perspective of advertisers, while from the perspective of readers, it is close to zero since the newspaper readers are indifferent to advertising which a newspaper contains.
The choice between delineating two interconnected markets or one integrated market will play a crucial role in competition analysis. In the case of transaction platforms, one integrated market should be defined as considering the connected sides of the market separately would lead to mistaken inferences. While for non-transaction platforms, different interconnected markets need to be delineated.
In an integrated market, the conduct needs to be holistically analysed from all sides of the market, while in different interconnected markets, the net competitive harm in any side of the market is enough to make the conduct anti-competitive because of the difference in feedback effect and competitive constraints.
The Way Ahead
In the BookMyShow case, the platform was a transaction platform since a single simultaneous transaction between the participants, which are the consumers and multiplexes, is observable. Therefore, the approach of CCI while delineating one integrated market incorporating both sides of the platform is a step in the right direction. However, in non-transaction platforms, the approach of CCI is inconsistent. The CCI, in some cases, has overlooked the multisided nature of the platform. Whereas, in others, it has failed to take into account their interconnectedness.
As ruled by the US Supreme Court in the seminal ruling of Ohio v. Amex, the CCI should take into account the difference in the nature of feedback effects and competitive constraints in transaction and non-transaction platforms. To contextualize the same – In a transaction platform, one integrated market shall be delineated, incorporating all the sides of the platform, while in a non-transaction platform, multiple distinct inter-related markets have to be delineated. Further, in the transaction platform, the net harm on all sides market of the market needs to be seen holistically to determine whether the conduct is anti-competitive or not. While for non-transaction platforms, the markets have to be seen separately even though due regard has to be given to their interdependence for the determination of anti-competitive conduct.
– Harshit Upadhyay and Sanigdh Budhia