ANTI-COMPETITIVE MERGERS IN TWO-SIDED DIGITAL PLATFORM MARKETS

• Author: Kenneth KHOO1 LLB (National University of Singapore), BSocSci (National University of Singapore), MSc (London School of Economics and Political Science), LLM (Yale Law School); Lecturer, Faculty of Law, National University of Singapore.
• Publication year: 2021
• Citation: (2021) 33 SAcLJ 10202
• Published date: 01 December 2021
• Date: 01 December 2021

I. Introduction

1 The broad range of economic activities related to the use of digitised information and knowledge has been a major driver of economic growth in the 21st century. Known as the “digital economy”,

these activities are set to increase Singapore's gross domestic product by as much as S$13.5bn in 2021.2 Digital technologies are becoming increasingly pervasive in everyday life — today, consumers need only tap their smartphones to order a meal, book a ride or pay a bill. Such advancements have also radically transformed the way consumers engage with written work, music, games and movies. Most content is consumed online these days, with online purchases of digital content replacing brick-and-mortar sales.3

2 The meteoric rise of digitalised content has accompanied the rapid ascent of digital platforms, which harness technology to bring together two or more distinct groups of customers.4 Digital platforms co-ordinate the actions of customers, creating value for them in a way which customers cannot readily obtain without these platforms.5 Thus, an important characteristic of digital platforms is their “two-sidedness” — that is, they act as an intermediary bridging two or more sets of users.6 For example, digital operating systems like “Windows” offered by Microsoft bring together application developers and users by providing a common interface for developer-user interaction, thereby reducing the costs of product development and ameliorating the difficulties associated with hardware selection.7 Many of these platforms have been established by massive conglomerates with ubiquitous names: Google, Facebook, Amazon and Apple are some that come to mind. As of 2019, the four companies have a combined market capitalisation of more than US$4trn.8

3 Despite providing economic benefits, digital platforms raise new regulatory challenges. In recent years, some antitrust scholars have argued that these platforms are decreasing effective competition, with

some describing Amazon and Google as monopolistic giants that rival Standard Oil and US Steel in the Gilded Age.9 In particular, markets where digital platforms operate tend towards monopolies due to strong network effects, large economies of scale and scope, close to zero marginal costs, as well as increasing returns to the use of data.10 On the other hand, other scholars have noted that ostensibly anti-competitive conduct by traditional “one-sided” businesses may be innocuous or even welfare-enhancing when initiated by digital platforms due to their “two-sided” nature.11 As an example, the sale of a product for free may be a profit-maximising strategy for a two-sided platform, rather than an attempt to predate.12

4 In the US, several commentators have suggested that contemporary US antitrust law has failed to protect consumers in this arena, with courts and antitrust authorities still strongly influenced by the non-interventionist bent of “Chicago School” economics.13 Thus, it comes of little surprise that a recent report by the Stigler Center for the Study of the Economy and the State contended that these platforms have “largely avoided any regulation” thus far.14 The report noted that unlike other countries, no US government committee was formed to address the antitrust issues arising from dominant digital platforms.15 Across the Atlantic, perhaps in part due to its ordoliberal roots, competition enforcement in the European Union (“EU”) is far more interventionist.16 However, courts and competition regulators there have not done much better where two-sided platforms are concerned — confusion continues to reign in competition law cases concerning such platforms. In Groupement des Cartes Bancaires,17

for instance, the European Court of Justice held that the General Court erred in law in failing to consider the interactions between the two sides of the market, as well as the indirect network effects present in the two-sided payment systems market.

5 As a relatively late adopter of competition policy, Singapore does not face the same ideological constraints faced by the antitrust authorities in both the US and the EU. As Tan has noted, Singapore's regulatory policies tend to adopt a pragmatic and instrumentalist approach, eschewing ideology in favour of rational, evidence-based reasoning.18 Indeed, competition policy in Singapore resembles what Ayres and Braithwaite have termed “responsive regulation” — the Competition and Consumer Commission of Singapore (“CCCS”) adopts a framework of collaborative capacity building, while retaining the capability to escalate regulation to tough enforcement.19 Reflecting this ethos, the CCCS regularly accepts commitments by firms to address anticompetitive conduct, acknowledging that these firms are not liable under the Competition Act20 where such commitments are accepted.21

6 In 2018, the CCCS had the opportunity to examine several novel issues that arose in a merger between two digital platforms, Grab and Uber.22 Both of these platforms engaged in, inter alia, ride-hailing — the matching of drivers and riders23 for the provision of booked chauffeured point-to-point transport (“CPPT”) services.24 The transaction entailed Grab's acquisition of certain assets, employees, contracts and data from Uber and its subsidiaries in Southeast Asia (“SEA”).25 These assets related to all of Uber's businesses in SEA, including the provision of intra-city transport facilitation and food delivery services excluding its intellectual property other than the data of riders, driver partners, UberEats

merchants, eaters and couriers.26 In consideration for these assets and other benefits, Uber received a 27.5% equity share in Grab.27 Ultimately, the merger led to the exit of Uber from the ride-hailing market.

7 The transaction attracted considerable media attention. Prior to the CCCS's decision, numerous op-eds were written on the matter. Some were in favour of aggressive intervention to prohibit the transaction, while others militated towards a more laissez-faire approach.28 On 24 September 2018, the CCCS finally released its long-awaited infringement decision (“ID”) issuing a S$13m financial penalty against the parties for infringing s 54 of the Competition Act. A series of remedial directions were also issued against the merged entity, Grab. These remedial measures required Grab to, inter alia, remove any exclusivity obligations with regard to its drivers, maintain its pre-transaction pricing on its platform, and to remove any restriction on the acquirers to whom its car rental company, Lion City Rentals, could be sold.29

8 In this article, the CCCS ID regarding Uber's sale of its SEA business to Grab will be critically evaluated. Although the CCCS's decision is laudable in many ways, it is argued that conceptual confusion remains regarding the two-sided nature of the merging platforms. In particular, the CCCS failed to distinguish the transaction price that riders faced on the platforms from the platform fees that both riders and drivers pay to use the platforms.30 The two-sided nature of the ride-hailing platforms involved is most apparent when one considers the fact that riders were paying negative platform fees for some trips on the platforms, in contrast to the corresponding drivers who were paying positive fees for the very same trips. Indeed, much of the transaction price paid by a given rider is in fact a zero-sum transfer to the driver undertaking that trip. Thus, the indirect network externalities in ride-hailing platforms affect not only the potential entry of the platforms' competitors31 but also the pricing

structure (ie, the ratio of rider platform fees to driver platform fees) of these platforms.32

9 Several related issues arise from this central observation. First, if competition law is to pursue welfarist objectives,33 then the ultimate question for the competition regulator is whether the merged platform would have the incentives to raise one or both platform fees relative to a counterfactual where the merger does not take place.34 Notably, this comparison is largely independent of the transaction price involved. In this context, the nature of the indirect network externality in the two-sided platform is critical. As the network externalities here are positive, it is easy to show how the externalities reinforce the merged entity's unilateral incentives to raise one or both platform fees, similar to how a traditional one-sided merged entity has the unilateral post-merger incentives to raise prices.35 However, the same is not true where the network externalities are negative. Here, a platform may have incentives to reduce one or both platform fees post-merger.36 Unfortunately, the CCCS did not consider the two-sided nature of the markets in the ID. Instead, it conflated the notion of an “effective” transaction price with the two distinct platform fees at hand.37

10 Second, in defining markets where two-sided platforms are involved, the competition regulator should consider how the nature of platform competition differs from more traditional forms of competition in one-sided markets. Here, the crucial aspect of the ride-hailing market is that of its nature as an exchange.38 In exchange markets, platforms collect fees from each side simultaneously for a given transaction. As any isolated analysis of the incentives to raise fees on just one side of the platform must be incomplete, the CCCS was correct in defining a single

product market for CPPT services.39 However, defining only one market as opposed to two distinct markets means that a firm would be either on both sides of the market or on none, rendering most of the CCCS's analysis otiose. As will be argued below,40 this “single-market” view may be justified by the nature of the platform's intermediary role. Here, the CCCS's lengthy discussion on driver and rider-side substitutes in the ID also failed to consider the nature of the indirect externalities...