Competition Law issues arising from Uber’s Model

Competition Law  issues arising from Uber’s Model-Does an increase of prices, regulated by Uber algorithm, constitutes – during Strikes Period in Paris – an infringement of EU rules under competition law?

by Dr. Nikolaos Pitsos LL.B Athens, Dr.iur. , LL.M FU-Berlin

Particularities of Uber services

New digital products and business models as well, as the special characteristics of digital markets, have created new challenges for competition policy and enforcement. Often  multi-sided platforms markets should be taken  into consideration which , when assessing individual cases , creates further complexity regarding important factors for a competition related assessment, such as  the market definition and market shares of trading platforms and online dealers,  the buyers or market  power and third party restrictions. Despite the growing expansion of these services, in most countries there is still no regulatory framework addressing these problems.

Uber model belongs to the so-called forms of collaborative economy, which, as an “umbrella term”, encloses different types of economic activities, somewhat informal, though all of them dependent on online platforms that bring together providers of different goods and services and users. Enterprises of collaborative economy are created by many of contracts between suppliers and the platform, which controls the terms of service in suppliers’ contracts with consumers. However, one main  difference (of collaborative economy) is that users on the supply side of the platforms are often not service providers acting in their professional capacity, but rather private individuals offering services on an occasional basis (so called “peers”).

Based on the examined practice and of Uber’s Business Model it can be argued that main purpose of Uber is to provide the carriage of passengers by road services. For that purpose, the passengers use the Uber’s application, a digital platform which enables the establishment of contractual relationships between the involved parties. However, regarding the price determination Uber’s charge differs from the price mechanism routed in traditional taxi services. Unlike the taxi services Uber models allows for drastic changes in prices for the same routes, depending on several different factors just like i) traffic congestion, ii) the level of supply and demand on the passenger location, iii) the distance between individual Uber drivers and the passenger, and iv) the general level of free carriage capacity and individual passengers orders in the broader area of passenger location, etc. According to Uber’s webpage, prices may briefly increase for a number of reasons such as bad weather conditions, public holidays, etc. The so-called Uber algorithm for every individual ride calculates all this relevant factors. In addition, Uber receives payments directly by the application users, collects its fee from the accepted amount and then pays the remaining amount to Uber drivers. As a result of the above described “trilateral- relation”, Uber driver and application user are not capable of directly negotiating the terms of the carriage contract, like f.i price, as all relevant aspects of contract negotiation are rendered through the digital platform.   


Online Platforms, Algorithms and Competition Law

The originality of concerns that Uber’s Model could raise, was well developed in Meyer vs Kalanick case in US. The legal question raised, was if the central pricing algorithm reduces price competition and enables a form of collusion and price fixing orchestrated by the platform and the drivers. Uber has faced both a claim for horizontal price-fixing and conspiracy between the drivers themselves and a possible vertical conspiracy between the platform and the drivers. Nevertheless, the legal classification of workers in the sharing-platform economy, like Uber, is important for several competition issues, as f.i the legal issue of coordination and collusion. As Uber’s activity has been expanding worldwide, also the efforts to impose regulation have grown with numerous jurisdictions in USA employing a wide range of measures (see Derek McKee, Finn Makela, Teresa Scassa, Law and the “Sharing Economy”: Regulating Online Market Platforms). Regarding Uber, some court’s decision (f.i Canada) have concluded that drivers should be considered as employees when in some American states drivers have been classified by the courts as independent contractors. This issue has been also a long controversy debate in decisions of the ECJ (see f.i also relevant Case of the European Court Uber France Case SAS C‑320/16). In these decisions among others it has been noticed that “Uber exercises decisive influence over the conditions under which that service is provided by those drivers” (Case C-434/15 Asociación_ Profesional Elite Taxi v. Uber_Systems Spain EU:C:2017:981, para 39) as an indication for the legal classification. 

Furthermore, the extend use of algorithms from online platforms and the way that their use may affect competition law has been an issue of intense research during recent years. The use of pricing algorithms from dominant and  non-dominant firms, even if such use had resulted in higher prices for consumers  didn’t always fall within the existent  competition law framework for many countries. One of the concerns is that algorithms have the capacity to easily determine the means of profit optimization for the companies and that also in these cases there is a very small human intervention, which further is making difficult for antitrust authorities to establish any anti-competitive practices. However, recently the legal scholars are starting to reconsider the relationship between machines and humans examining if during the use of relevant algorithm any illegal action could have anticipated or was predetermined. In that sense, it has been noticed that pricing algorithms should be built in a way that does not allow them to collude. After the Commission decision imposing a record fine on Google for an abusive practice perpetrated through its search algorithm, it was clear that companies could not always escape responsibility for collusion by hiding behind a computer program. In other words, only the use of an algorithm does not necessarily prove the “neutrality” or the absence of any human intervention in a possible infringement. As also the EU Competition Commissioner Vestager recently stated  “what matters is how these algorithms are actually used” which according to one approach” sensibly suggests that for now the Commission’s focus will remain on the more clear-cut cases of collusion”. (See S. Lawrance and M. HuntWill pricing algorithms be the European Commission’s next antitrust target? “2017. However, also “under the algorithms approach” it should not be forgotten that also a “normal” marketplace operates by matching the excess capacity of an individual supplier with the temporary need of an individual consumer.

Assessing dominance and “success” for online platforms

The definition of the relevant  market, the measurement of constraining market power and the definition of a dominant position in the so called “collaborative platforms”, like Uber’s  can present different complexity. More precisely and given the fact that digital platforms operate in two-sided markets, the first step of a market definition is to evaluate if it should be the definition of one or two markets, and, more precisely, to determine the market in which sharing economy firms compete. (See more See, e.g., L. Filistrucchi, D. Geradin, E. van Damme & P. Affeldt, “Market Definition in Two-Sided Markets: Theory and Practice, 10(2) J. Comp. L. & Econ. 293 (2014) )

In some cases, the effort to define the relevant market at the same time could easily justify the success of the platform. Taking the example of Uber under consideration, Uber belongs to a two-sided business model, which makes full use of and derives benefits from the indirect network effects that exist between drivers and users. In practice Uber’s platform develops successfully and manage to be efficient not only because Uber  through the platform controls the proper use  of cars and drivers,avoiding unnecessary offer of services,  but also  because it creates a so called “ self-reinforcing effect” arising from the fact that “a greater number of drivers attracts more users and vice-versa” (See. M.Colangelo/M. Maggiolino Uber: A new Challenge for Regulation and Competition Law, in Market and Competition Law Review Vol.1 Oktober 2017 p.47(60) and also D. Evans and M. Noel, Defining AntitrusMarkets When Firms Operate Two-Sided PlatformsColum. Bus. L. Rev.3 (2005): 667. In other words online platforms offer a significant advantage in order to control the demand of a market in real time and consequently to be in a position of making the proper adjustments from the supply-side. In addition, the fact that the user usually disposes a choice between different prices, provided by the drivers, creates a competitive advantage for Uber in comparison with other relevant markets such as taxi drivers, where minimum prices in many countries are defined by the State. Uber’s success is also connected with the fact that, by matching users and drivers via its platform, it exploits the interconnection between at least two different markets and consequently the strict relation  between  two different demands: the demands (and market) of transportation and the demand of driving(See.M.Colangelo/M. Maggiolino Uber: A new Challenge for Regulation and Competition Law, in Market and Competition Law Review Vol.1 Oktober 2017 p.47(60).Regarding the establishing of market power traditional indicators such as network effects, product differentiation, innovation and potential competition need also to be reassessed in a platform context, like Uber.

However, competition authorities seem to have most difficulty with the finding and applying a concrete theory of abuse, despite the fact that market power is well established. In a relevant Uber Case in NY, the Case Meyer vs Kalanick, a private consumer class action in the US against Uber, the dominant position of Uber was examined according the specific effects to the competitors. More precisely the Court found that Uber’s market position has already helped force Sidecar out of the marketplace, while dominant position and considerable name recognition made it difficult for potential competitors to enter the marketplace. The claim of Uber that company’s terms allows drivers to charge less, which mean to deviate from the price that the algorithm suggest, has been rejected by the Court since the application lacks the mechanism for such a possibility. In addition, in Meyer Case and before the case was submitted to arbitration, the Court held that individual driver agreements with Uber platform could be aggregated to produce a horizontal agreement among individual drivers. In the theory of Competition law Uber’s Model brings in mind the construction of the so called, “Hub-and-spoke arrangements”, which, in the case United States v. Apple Inc has  been held to be a per se violation of the Sherman Act.( See United States v. Apple Inc., 791 F.3d 290 (2d Cir. 2015), See also Belleflamme, P. / Peitz, M.  The Competitive Impacts of Exclusivity and Price Transparency in Markets with Digital Platforms, published in, November 2019). Hub-and-spoke arrangements are considered the horizontal restrictions on the supplier or retailer level (the “spokes”), which are implemented through vertically related players that serve as a common “hub” (e.g., a common manufacturer, retailer or service provider).Under these conditions  the “hub” facilitates the co-ordination of competition between the “spokes” without direct contacts between the spokes.

In another relevant Case in India, the company Meru Travels Solutions filed an information with the Competition Commission of India (CCI) alleging that Uber was engaging in below-cost pricing for its services in order to eliminate competition“. Τhe information was filed by alleging that Uber was engaging in predatory pricing by offering huge discounts, in addition to the already reduced tariffs to customers and unreasonable high incentives to drivers to keep them attached to its network. More precisely the behavior of Uber presented typical indications of predatory pricing given the fact that these incentive schemes had as a result voluntary losses on a per trip basis for Uber, which it could be absorbed from Uber because of the (alleged) market power it enjoys, by way of PE funding, network of cars and drivers, etc. The result of the incentive schemes is that Uber is engaging in pricing that is below the cost of providing the services. 

The Commission delineated the relevant market to be the ‘market for radio taxi services in Delhi’ based on its previously decided cases on radio taxi services. The informant had relied on a TechSci report to establish dominance of Uber in the market. However, the Commission observed that the TechSci report was not reliable. First, the Commission further observed that the fluctuating market shares of different competitors showed that the radio taxi service market in Delhi was competitive in nature; the second Ola’s presence in the relevant market exerted a significant competitive influence; and therefore the Commission had concluded that Uber was not holding a dominant position in the relevant market. The Commission therefore closed the case under section 26(2) of the Competition Act, 2002.However, Meru filed an appeal under Section 53B of the (Indian) Act against the order passed by the Commission and the COMPAT (Competition Appellate Tribunal’s) observed that the Commission had erred while delineating the geographical market.

Recently, the Supreme Court of India, held that (i) if Uber is incurring losses on its trips, by way of discounts to customers and incentives to cab owners, such conduct is bound to affect its competitors and would, therefore, fall under Explanation (a); and (ii) therefore, according to the Supreme Court, Uber enjoys a dominant position and consequently dismisses an appeal against an investigation into allegations of abuse of dominance.

In another relevant case – Agrawal v. ANI Techs./Uber, the Competition Commission of India (during an investigation  of a hub-and-spoke conspiracy in the context of a platform enterprise) appears to have required evidence of communication among individual suppliers rather than simply  referring  to the communication-channels “as a function of mutual participation on the platform”. (See See also Belleflamme, P.
/ Peitz, M.  The Competitive Impacts of Exclusivity and Price Transparency in Markets with Digital Platforms)

However, and with regard to online platforms, like Uber, it can be argued that one form of abuse could be the exploitation of market power and more precisely exploitation against consumers’ choices. Digital platforms maintain “the advantage”, through the gathering of data, to monitor the customers’ activities and react to market changes in real time. This adjustment is not necessarily related always to a “price decision”. The form of this exploitation depends on the nature of the services, provided by the online platform (f.i in Facebook Case the German competition authority (Bundeskartellamt) accused Facebook for imposing unfair business terms on its users, which allow it to amass data limitlessly generated when using third-party websites).

One of these changes could be an increase regarding demand during and because of a strike period on others private or public means of transport, like recently was the case in France.  Here exactly could be raised the main legal question: given the fact that computer algorithms can easily adapt themselves to market conditions or raise the prices to reflect the demand and supply conditions of the relevant market at a specific period, could such an “adjustment” be qualified as an “abuse of dominance”? Should this price increase be justified by the fact that it is a result of an algorithm-adjustment and not a result of a (human) strategic enterprise decision? Supposing that Uber’s algorithm (also referred as an “algorithmic monopoly”) “may mimic a perceived competitive price rather than the true market price therefore raising suspicions of manipulation of the perceived market price” (see Hatzopoulos, V. The Collaborative Economy and EU Law, p.130) should such a behavior be considered as a violation of competition rules? Uber defense is that the use of algorithm is only for determining what the market (price) is (this was one of the main statement made by Uber’s former CEO in the case Meyer vs Kalanick). However, any proof that an algorithm has been used in an “abusive” or “facilitating cartel” way remains in our opinion a difficult task for antitrust authorities and plaintiffs.

In any case it should exist an upper limit, a “filtering”-point for price increase even if it is accepted that algorithm is functioning only to reflect the (real) market price. As through, our analysis has been shown (USA and India) and regarding the theory of abuse of dominance, which can be used, this can differ from one jurisdiction to another. An interesting idea for examining the setting prices by Uber, at least in EU Competition law, would be the theory of excessive prices. This form could reflect an upper limit both to a price increase and to the use of market power by Uber, especially in periods with an increased demand. According to the traditional approach of EU Competition law, a dominant firm can abuse its position by charging unfair prices. Among other prohibitions, this prohibition prohibits excessive prices that are ‘too high’. However, this form of abuse has remained underdeveloped conceptually and in practice at the EU level, so there is ambiguity regarding what constitutes an excessive and therefore unfair price. The «classic» European case on excessive prices –United Brands, had established the standard test used by the Commission and European competition authorities when investigating excessive prices. The United Brands test has two limbs, consisting in determining i) “whether the difference between the costs actually incurred and the price actually charged is excessive” and ii) “if the answer to this question is in the afirmative, whether a price has been imposed which is either unfair in itself or when compared to competing products”. Another main obstacle to effective implementation of the prohibition of excessive pricing is the lack of a commonly agreed approach upon how to measure it. Traditionally, antitrust authorities worldwide in excessive pricing cases often base their claims on a comparison of the dominant firm’s price with some competitive benchmark, such as the firm’s cost of production, or the firm’s own price in other time periods, in different geographical markets, or in different market segments. As our analysis, based on Competition law, has proved, the step-by-step procedure it should be followed regarding the assessment of a dominant position for online platforms, like Uber, doesn’t necessarily present many originalities (f.i market definition, the level of market shares etc.). However, and regarding the proof and implementation of more concrete theories of harm for online platforms, like predatory pricing- practice or collusion during the use of algorithms, could raise many complex and indigenous legal questions, where traditional tools like, the theory of “excessive prices” could be used, in cases where f.i an increase of prices, regulated by Uber algorithm, could reach a level, which it could be not simply justified from an adaptation of an algorithm to the unusual  (like the strike in Paris) market – from the demand and supply-side-conditions. The broaden use of artificial intelligence in many aspects of modern times has already influence the  nature of competition restraints, which enforcement agencies will need to tackle.