ALGORITHMIC COLLUSION AND HUB & SPOKE CARTEL: STUDYING ANTI-COMPETITIVE BEHAVIOUR IN ONLINE PLATFORM
[Avanti Deshpande is a fourth-year student at ILS, Pune]
Digital economies have upended traditional economic models and the conventional understanding of competition law. Technology and digitization promise pro-competitive benefits such as increased efficiency and better consumer welfare, however, it also has insidious ways of fostering anti-competitive behaviour. The advent of disruptive innovators alongside the development of pricing algorithms has given rise to several concerns regarding their use to facilitate anti-competitive behaviour such as price fixing on various digital platforms. This concern has been legitimized by numerous judgements around the world where the technology of pricing algorithms has been effectively utilized to enable anti-competitive agreements.
Indian Legal Framework
A cartel is a form of an anti-competitive agreement, and described in Section 2(c) of The Competition Act, 2002 (“Act”) which states that,
“Cartel” includes an association of producers, sellers, distributors, traders or service providers who, by agreement amongst themselves, limit, control or attempt to control the production, distribution, sale or price of, or, trade in goods or provision of services;
Simply put, cartels are agreements between enterprises to not compete on product, price or customers with the goal of raising prices above competitive levels to the detriment of the consumers and the economy. Section 3 of the Act which discusses anti-competitive agreements prohibits any enterprise or association from entering into agreements which cause or are likely to cause an appreciable adverse effect on competition (AAEC) in India. This has a wide scope for the term ‘agreement’ and includes informal and unwritten agreements as well. Section 3(3) of the Act prohibits collusion and is important in the context of cartelization. Section 3(3)(a) can be broken down as follows:
An agreement was entered into; a decision was taken, or, a practice was carried on;
by persons, an association of persons, enterprises or association of enterprises;
which has directly or indirectly determined the purchase or sale prices;
and, shall be presumed to have an AAEC.
What is a Hub and Spoke Cartel?
The rise of digital technologies and the internet has led to a change in the anti-competitive behaviour of market players, and thus, traditional cartels as we know them, seem to be nearing their end. In competition law, agreements between market players are classified as horizontal or vertical. A horizontal agreement is an agreement between competitors, and are thus operating at the same level in the market or the supply chain, while as a vertical agreement denotes firms or enterprises operating at different levels of the production or distribution chain.
In the case of a hub and spoke cartel, the communication between competitors is not done directly, it is done via a third party, which takes the form of a hub and acts as the facilitator of the cartel. Thus, the hub can be a common manufacturer, retailer, service provider or even a trade association. The spokes are the competitors who form the colluding parties. They are connected by the horizontal agreement between them which forms the rim and thus completes the cartel.
The hub acts as a common contractual partner, which is active at a different level of the production or distribution chain and often plays the role of stabilizing the cartel. The hub makes it especially difficult for authorities to detect anti-competitive behavior, as by acting as a conduit, makes the cartel even more efficient as compared to a regular cartel, which operates without a hub.
Pricing algorithms are sensitive in nature and make anti-competitive behaviour cost effective and quick, and safeguard it from detection . Principally, there exists four methods through which algorithms can be used to facilitate collusion amongst competitors, out of which the hub and spoke arrangement is one.
Plainly put, an algorithm can be explained as, “a sequence of rules to be applied in precise order to carry out a task.” Further, a pricing algorithm can be described as, “an algorithm that uses price as an input, and/or uses a computational procedure to determine price as an output.” Thus, pricing algorithms are essentially designed to collect and analyze immense amounts of relevant market data in order to price products and services while taking into account an extensive set of factors.
Here, it is essential to understand that, algorithmic collusion is mainly of two types, algorithmic tacit collusion and algorithmic express collusion. Tacit collusion or tacit coordination refers to an anti-competitive market outcome which is achieved without the need for explicit communication between competitors. Algorithmic tacit collusion or conscious parallelism is a situation wherein higher prices, which is the anti-competitive outcome, can be achieved without rivals having agreed to tamper with prices. Furthermore, there are two main methods via which algorithms are used to reach collusive outcomes. The first being, that an existing price fixing agreement between competitors is facilitated by an algorithm, and the second, the algorithms are devised in a manner so as to reach a tacitly collusive outcome.
In their work, Anderson and Huffman assert that, firstly, Uber drivers did not form and do not own Uber, secondly, the price that is charged will be determined by Uber’s pricing algorithm and lastly, that each driver is aware of the fact that all of the other drivers are simultaneously in similar agreements with Uber and that all the drivers have agreed to this arrangement.
With regards to this issue, the Competition Commission of India (CCI) unfortunately took a narrow approach in the case of Samir Agrawal v. ANI Technologies Pvt. Ltd. (which was recently upheld by the NCLAT) and dismissed a case of facilitating a hub and spoke cartel via tacit collusion against cab aggregators Ola and Uber. Uber in its Terms and Conditions states that it is a ‘technology platform’ and asks the user to recognize that it does not provide any transportation or logistics services and that all such services are provided by an independent third party contractor, who is not employed by Uber. Similarly, Ola too in its User Terms does not identify itself as a transportation service, but merely as an electronic platform to facilitate aggregation of vehicles. This means that, the drivers are not employees or workers of Ola/Uber, but are independent contractors. However, despite being independent contractors, the drivers have no bargaining power and are not at liberty to set their own prices; but the prices are set by Ola/Uber via a pricing algorithm. Here comes the relevance of the fact that, in a hub and spoke cartel, cartel members do not communicate directly in order to align their behaviour, but, rely on a third party or intermediary to do so. Additionally, just because the third part may not “active”, does not absolve it of its liability, as cartel facilitators are certainly liable in competition law as well.
The situation in the Samir Agarwal case, is in a sense, comparable to the Eturas case, where hub and spoke cartels were first detected in the online sphere. In Eturas case, an online travel booking system provided by a Lithuanian administrator, allowed travel agencies to put up travel bookings for sale on their websites through a uniform presentation method determined by Eturas. A cap on the online booking discounts of 3% was subsequently introduced by Eturas. A message regarding the information of this restriction was sent to all the travel agencies who were participating in the online platform of Eturas via an internal electronic message. The Court noted that the travel agents knew of the message and did not publicly distance themselves from it or report the practice to the competition/administrative authorities, were presumed to have participated in the cartel. Even if they were not aware of it, as all the agencies had received the electronic message from Eturas, they tacitly assented to it and hence, the agencies had indulged in a concerted practice and their behaviour was anti-competitive.
One of the main objections of the CCI in Samir Agarwal was that, for a cartel to exist, there needs to be a conspiracy to fix prices, and for a conspiracy there needs to be collusion and mere assent of the drivers to the algorithmically determined prices could not amount to collusion between the drivers. However, in the opinion of the author, the CCI here took a very narrow understanding of the term “collusion” and failed entirely to account for tacit collusion or even understand the nuances of tacit collusion in the online sphere. In Eturas too, there was no communication between the online travel agencies, just as there was no communication between the drivers, but the presumption of participation still existed. Thus, in the cab aggregators case too, there was tacit collusion with the cab aggregator acting as the hub, the drivers who are the competitors acting as the spokes, and the prices co-ordinated by the hub (Ola/Uber) and followed by the drivers. Furthermore, the drivers entered into the agreements with the active knowledge that other drivers (their competitors) are doing the same, thus culminating in a hub and spoke cartel. Following the reasoning adopted in the Eturas case, usage of a common third-party algorithm.
The pace of growth in business with context to the digital sphere has been rapid and has in turn resulted in several competition concerns, which the existing legislation is ill equipped to handle. Regrettably, the CCI missed the mark in the Samir Agarwal case, and adopted a very traditional approach, while failing to embrace a more nuanced approach, more in line with the technological realities of the 21st century. The COVID 19 crisis, has further highlighted the importance of the internet and online market places. As such, competition law concerns with regards to the digital sphere can only be expected to grow in the future. The Draft Competition (Amendment) Bill, 2020 has suggested a number of amendments to the Competition Act, 2002, one of them being, an expansion of the scope of the term ‘cartel’ to include hub and spoke cartels that will hopefully address tacit collusion, which is a step in the right direction.