THE OLA-UBER PRICING ALGORITHM CASE

This article has been written by Harshita Lal, a 4th year B.A. LL.B (Hons.) student at Symbiosis Law School Pune.

Introduction

In the instant case of Samir Agrawal v. ANI Technologies Pvt. Ltd. and Uber India Systems Pvt. Ltd. (Case No. 37 of 2018), the Informant alleged violation of provisions of Section 3 of the Competition Act (hereinafter “the Act”)  by the Opposite Parties, Ola and Uber (hereinafter “OPs”). Mainly aggrieved by the pricing mechanism adopted by the OPs while providing radio taxi services, the Informant alleged that it amounted to price fixing as it hindered the freedom of individual drivers to compete amongst each other. However, the Competition Commission (hereinafter “Commission”) took the view that no case of contravention of provisions of Section 3 had been made out and dismissed the Informant’s allegations against the OPs.

Factual Background

The OPs provide radio taxi services on-demand by the process of matching riders and drivers and providing an estimate of the fare/price beforehand, using an algorithm. The Informant, as an aggrieved consumer of the services provided by the OPs, has raised three allegations:

First, the Cab Aggregators use algorithms that deliver the centralised power of fixing ride prices booked through their respective Apps into their hands, not allowing the drivers who are attached as independent third party service providers, to compete on prices. This was alleged to be a hub and spoke arrangement where the OPs act as ‘hub’ and the competing drivers act as ‘spokes’ colluding on prices;

Second, the acts of price fixing are an imposition of a minimum resale price agreement between the Cab Aggregators and the drivers, as the drivers don’t have the liberty to reject algorithm calculated prices or offer their services for a lower price; and

Third, the Cab Aggregators have considerable personalised information about every rider, owing to the information asymmetry, giving them the ability to discriminate on the basis of price to the disadvantage of the riders.

The Informant supported his aforementioned allegations with the arguments that (i) the Ola/Uber model is comparable to Zomato, Trivago or Airbnb who do not own any restaurants, properties or hotels, respectively, but acts only as platforms that connect buyers and sellers and the price isn’t fixed by the platform and (ii) Ola/Uber had facilitated a cartel of drivers in a digital mode and should be accorded a similar treatment/liability under the Act as the Commission’s judgement of Builders Association v. Cement Manufacturers Association & Ors (Case No. 29 of 2010).

The Commission’s Findings and Decision

The Commission found no substance in the first allegation regarding the existence of a hub and spoke arrangement between the Cab Aggregators and the drivers. In both the Cab Aggregator models, the fare was estimated through their Apps on the basis of ‘big data’ by the algorithm, taking into account personalised information of riders along with other factors, resultantly determining prices for each rider and trip differently. The drivers did accede to the algorithmically determined prices, but there was no agreement between them to set prices through the platform or for the platform to coordinate the prices between them. Therefore, there did not appear to be any such agreement inter-se between the drivers to delegate pricing power to the Cab Aggregators.

The Commission found the second allegation of contravention of Section 3(4)(e) of the Act not tenable, explaining that resale is essential to conduct resale price maintenance, however, there was no sale of any goods/services by the OPs to the drivers, resold by the drivers to the riders. When operating through the OPs platform, the drivers were effectively their extensions or agents, therefore resulting in a single transaction between the rider and the OPs. Further, the algorithmic determination of prices by the OPs was vital to the functioning of the aggregation-based models. These pricing algorithms allowed for modification and optimization of prices based on numerous factors, including available stock and anticipated demand, resulting in fares that were dynamic in nature and updated on the basis of real-time market and traffic conditions. Accordingly, the Commission was of the view that the Informant had arrived at an erroneous conclusion without placing any evidence on record, that an algorithm determined price would eliminate price competition and would necessarily be higher than the prices negotiated by drivers and riders on an individual trip basis.

Price discrimination was alleged by the Informant but not under Section 4 of the Act, where Section 4(2)(a)(ii) deals with prohibition of imposing discriminatory prices by a dominant enterprise. Further, the Act does not recognise collective dominance whereas the market in question had two players, none of which was alleged to be dominant. This position was clarified in previous matters relating to the Cab Aggregators market (Fast Track Call Cab Pvt. Ltd. V. ANI Technologies Pvt. Ltd. and In re Meru Travel Solutions Pvt. Ltd.). The Commission found the third allegation of price discrimination misplaced and unsupported by any evidence on record.

Lastly, the Commission dealt with the Informant’s supporting arguments in the following manner: (i) The Commission found it inappropriate to equate Ola/Uber with Airbnb, Trivago and Zomato as they purely acted as platforms, whereas Ola was held to be a radio taxi operator in Fast Track Call Cab Pvt. Ltd. V. ANI Technologies Pvt. Ltd. and not merely a platform and Uber was held to be a transport service company  in Asociación Profesional Élite Taxi v. Uber Systems Spain SL (C-434/15), which not only intermediates between drivers but also acts as a service provider; and (ii) The Commission found the Informant’s demand for uniform application of the Builders Association judgment in the present case devoid of an understanding of economic literature and practical realities of the digital markets. As Ola and Uber acted as separate entities from their respective drivers where there was no opportunity for the drivers to coordinate their actions with other drivers. Therefore, this cannot be termed as cartel conduct through Ola/Uber’s platform.

Pricing Algorithms and their impact

Pricing algorithms such as the ones used by the OPs are intended to gather and analyse a large amount of relevant market data and price products/services after considering numerous factors. They can enable an enterprise to react instantaneously to the price movements of its competitors, therefore, the speed and complexity are such that cannot be replicated by humans. This creates a perception that the price fixing is a result of machine-driven processes irrespective of the human element creating potential antitrust liability [1].

This gives an advantage to enterprises using such pricing algorithms for exchanging market information and fixing prices thereby colluding tacitly. The difficulty lies in proving the intention or understanding behind such tacit collusion. Therefore, such algorithms may assist with collusion and complicate detection of unlawful agreements. Thus, it is challenging to deal with such anticompetitive behaviour and enforce laws regarding the prohibition of price parallelism.

CONCLUSION

The Competition Act is well equipped to deal with a case of price fixing using pricing algorithms, however, in certain matters where there is no communication between the enterprises and no available circumstantial evidence, just as held in the Eturas case (C- 74/14), the significant option left is to create a presumption that two or more enterprises with similar prices using pricing algorithms would indicate knowledge of the price-fixing carried out by the pricing algorithms and engagement in concerted practice. This rebuttable presumption is acceptable; it does not shift the burden of proof entirely onto the defendants thus not going against the concept of a fair trial.

ENDNOTES

[1]https://www.arnoldporter.com/en/perspectives/publications/2018/04/pricing-algorithms-the-antitrust-implications

Leave a comment

Design a site like this with WordPress.com
Get started