Competition law and digital ecosystems: Learning to walk before we run

Abstract

The paper analyzes the challenges faced by competition authorities with respect to the digital sector. Borrowing insights from the business policy literature and from the economic literature, the paper first analyzes the specificities of digital firms (multi-sided platforms and ecosystems) with respect to their development and competitive strategies. Building on this foundation the paper explores some of the challenges of applying traditional competition analysis to competition in the business sector. We then discuss a number of issues relevant to competition law enforcement in the digital sector starting with the role of data, competition within ecosystems and between ecosystems, consumer biases, and the role of gatekeepers. We conclude with a research agenda for economists and competition authorities.

1. Introduction

Competition authorities are under severe political pressure to intervene quickly against the digital behemoth for a variety of reasons. The extraordinary financial success of a few companies, some of which have benefited from the coronavirus disease 2019 (Covid-19) crisis, combined with the idea that they have been highly successful at optimizing their tax status and avoiding paying taxes, have cemented a political consensus that they should be scrutinized and that their power should be curtailed. The inability of antitrust authorities to control the merger wave initiated by the Google Apple Facebook Amazon Microsoft (GAFAM) in the USA (Parker et al., 2020) and the fact that in Europe a limited number of cases brought by the European Commission against Microsoft, Google, or Amazon have not had a tangible effect on their behavior (Kwoka and Valletti, 2021) has meant that a number of critics argue that direct regulation of the GAFAM would provide a better way to control them than ineffective competition law enforcement (Bourreau and Perrot, 2020). The rapidity of the changes in the digital sector and the idea that once entrenched market power may be exceedingly difficult to curb have meant that competition authorities have felt that they have to react fast, possibly even before they fully grasp the intricacies of the digital sector.

A number of recent reports, some commissioned by competition authorities, have analyzed the specificities of the digital sector, making suggestions on how to tackle such a frontier in competition law enforcement. For example, ‘Competition policy for the digital era’ report (Crémer et al., 2019) commissioned by Ms Vestager, EC Competition Commissioner, argued that the specific characteristics of platforms, digital ecosystems, and the data economy required established concepts, doctrines, and methodologies, as well as competition enforcement more generally, to be adapted and refined. In particular, the report suggested that the timeframe and standard of proof of the consumer welfare approach should be rethought to avoid under-enforcement. It also suggested that competition enforcement should protect competition for the market as well as competition in the market and that in some cases data access—and possibly data interoperability—may need to be imposed through regulation. It added that the acquisition of targets with a low turnover but a large and/or fast growing user base and a high future market potential needed to be controlled and that injecting some ‘horizontal’ elements into the analysis of ‘conglomerate’ mergers may be necessary to control the creation of dominant ‘ecosystems’ through mergers and acquisitions.

In May 2019, the Furman report (Report of the Digital Competition Expert Panel) (Furman et al., 2019) was published in the UK. This report suggested that digital markets are subject to ‘tipping’, in which a winner will take most of the market, and that if concentration in digital markets can have benefits, it can also give rise to substantial costs, for example from lower innovation. It concluded that neither merger control nor antitrust enforcement are well designed for the intensive and ongoing work that needs to be done to facilitate competition and entry by making it easier for consumers to move and control their data, and for new digital businesses to interoperate with established platforms. It suggested that merger control should be more forward-looking and take better account of technological developments and that, with respect to antitrust, a greater use of interim measures to prevent damage to competition while a case is ongoing as well as an adjustment of the appellate standard were necessary for the digital sector.

In the USA, in October 2020 the Antitrust Subcommittee of the US House of Representative’s Judiciary Committee (REF) released a staff report analyzing the challenges presented by the dominance of Apple, Amazon, Google, and Facebook and their business practices. The report proposed a series of radical remedies to restore competition in the digital economy, to strengthen antitrust laws, and to reinvigorate antitrust enforcement. The report recommended structural remedies such as the separation of digital platforms from commerce and discouraging mergers, resulting in 30% or more market share in order to prevent the increase in concentration. The report also discourages acquisitions by dominant companies of direct competitor start-ups or start-ups operating in adjacent markets. Its regulatory recommendations suggest rules requiring non-discrimination, data portability, and interoperability. Setting the threshold for market dominance at 30% share, the recommendations further call for a European-style offense of abuse of dominance. Additionally, the report called for special rules for dominant companies including non-discriminatory access to their essential facilities and a prohibition on tying products and services together.

Although they have very different perspectives these European Union (EU), UK, and US reports have one point on which they agree: all three consider that, as they are practiced today by competition authorities, antitrust or competition law enforcement and merger control are inadequate or insufficient to deal with competition issues in the digital sector. If the US report proposes to abandon the consumer welfare standard altogether, a proposal which has little chance of success of being adopted, the other two reports suggest that the difficulty is not with the standard applied by competition authorities but with the tools that they use to carry on their analysis. Those tools need to be adapted or new tools need to be adopted. The remainder of this paper considers some of the key issues that emerge in this digital context, starting with the nature and issues that firms and multi-sided platforms as well as ecosystems entail, and what this means for antitrust law and its enforcement. I then consider particular issues of interest, such as the role of data, and then competition within and between ecosystems, behavioral issues (and in particular, consumers status quo bias) before turning to a pragmatic, rather than analytical approach around gatekeepers which has emerged to tackle some of the issues of dominance. I conclude with synthetic observations and priorities for research and practice alike.

2. Firms and multi-sided platforms

Why do we need to adapt competition analysis tools to the digital sector? The answer is that many standard tools of analysis used in traditional competition law enforcement rest on strong hypotheses about the organization of economic activities. First, hierarchical firms are supposed to operate on predefined markets for goods or services where they meet consumers. Second, those firms are considered to operate on markets where similar firms are competing with them by offering substitutable products or services to those that they offer. Thus, a market is the locus of competition. Third, those firms sell their products or services for a price to consumers. Their goal is to maximize their profits on the markets on which they sell the product or services they supply by producing up to the point where their marginal revenue and their marginal cost are equal. Also, because they seek to maximize profits, firms will never choose to charge a price which is below their average variable cost. Their marginal revenue and ultimately the price they charge is therefore a function of the intensity of competition they have on the market and the goal of antitrust is to ensure that the price of those products or services remains competitive. An individual firm may have market power if it has a large share of the market and is protected by barriers to entry. The concern with a cartel or a merger is that it may exacerbate these issues—all of which can be assessed in straightforward ways. If we now turn to the digital sector, two major characteristics differ from traditional markets.

First, the internet abolishes distances and reduces the cost of interaction and transactions between people and groups of people. Multi-sided platforms exploit the interdependent interest between several groups of users of a platform (e.g. between advertisers and users). As a result, multi-sided platforms are prevalent. This means that the delivery of services is organized differently in the digital sector than in one-sided firms operating on traditional markets.

Second, digital technologies allow firms and platforms, via digital interfaces, to gather real-time data on users and/or consumer choices, on the environment in which they operate, and on the alternatives they consider, as well as data on the products or services consumed and the reaction of consumers or users to these products (through sensors), to store this data (on the cloud), and to analyze it (through artificial intelligence algorithms). Digital technologies thus offer unparalleled opportunities for firms to develop insights on individual customers and help firms develop products or services that are customized to customer preferences (Subramaniam, 2020). Access to data and use of data thus become strategic assets for digital firms to increase their value proposition.

There are several types of platforms: transaction platforms where potential sellers and buyers can meet and exchange via the platform (e.g. eBay); services platforms (where the platform provides some services which attract users and allow suppliers or advertisers to reach these users, e.g. Google) and social media (where people interact, e.g. Facebook). A platform seeks to maximize the value for its users of interaction between them. This means, among other things, that there is no longer a direct relationship between the price charged on one side and the cost of servicing consumers on this side. Indeed, acquiring additional users on one side (e.g. users of Google Search), even non-paying users, may be very profitable to the platform if this increases sufficiently the value of the platform to another (paying) side (e.g. advertisers).

This type of platform may raise complex questions for competition authorities with respect to the definition of the market on which these platforms operate, with respect to the assessment of market power as well as with respect to the assessment of their strategy from a competition law standpoint and with respect to the qualification of abusive practices. Each individual side of a platform may not be a relevant market as the platform does not try to maximize its profits on one side independently of the other side but takes into consideration the interaction between the users on the various sides.¹ If each side is not an independent market, the assessment of the market power of a platform is likely to be more complex than the assessment of the market power of a firm operating on a traditional market. Indeed, for instance, a platform may be the only provider of a category of service to one side (to the final consumers benefiting from the services of the platform) but at the same time on the other side of the platform (with respect to business users of the platform, such as the advertisers or providers of services using the platform) be in competition with other platforms offering different services but also attracting a large number of users. In such a case the platform may not have market power in spite of the fact that no other platform offers the same service as the one offered free to its final consumers.

In addition, price competition is often meaningless on the side(s) of the platform where a service is delivered free of charge in order to attract a large number of other (paying) users on other side(s) of the platform. On the side or sides on which they do not charge a price, platforms compete on quality and innovation. As the Organisation for Economic Co-operation and Development (OECD) secretariat observed (OECD Competition Committee, 2018): ‘When the prevailing price in a market is zero, the offering of improved functionality or new features may be one of the only ways for firms to capture market share’. Indeed, in Google Shopping, the EC noted (European Commission, 2017) (para. 268) that ‘In so far as users expect to receive a service free of charge, an undertaking that decides to stop innovating may run the risk of reducing its attractiveness, depending on the level of innovation on the market in question’.

These characteristics mean that the competition law instruments, such as market shares or concentration ratios, used for traditional markets in order to assess dominance cannot be easily used when it comes to platforms because of the multiplicity of services they offer simultaneously to different groups of consumers. It also means that market power indicators based on a comparison of price and cost (such as price–cost margins or the Lerner index) cannot be used on each side of the platform to assess its market power. It finally means that the characterization of an abusive practice of a platform may be either different or more complex than in the case of traditional markets (Jacobides and Lianos, 2021). For example, pricing below cost (on one side of the platform) may be a profit-maximizing strategy and therefore not qualifiable as a predatory practice unlike the case on traditional markets. But a degradation of quality may be an abuse of dominance by the platform with the added difficulties that quality is often not easily observable and that quality is multidimensional, which means that the degradation of some elements of quality may in fact allow an improvement in other dimensions of quality (e.g. the use by a platform of data to which it should not have had access may be a degradation of the privacy of the users but also allow the platform to give a more relevant service to those users).

Platforms have been studied extensively over the last two decades thanks to the fact that non-digital platforms such as payment platforms had attracted the attention of competition authorities. The fact that they have been studied extensively in the economic literature does not mean that agreed-upon solutions have been found on how to adapt competition enforcement tools to platforms. The payment platform cases are an area of antitrust law where we have seen contradictory decisions by competition authorities on how they should be handled as well as much criticism of academics about the approach taken by courts to these new problems, for example with respect to the Amex case in the US Supreme Court (US Supreme Court, 2018) and the Groupement Carte Bancaire case in the EU (Court of Justice of the European Union, 2014).

3. Multi-sided platforms and ecosystems

The rapid evolution of digital platforms forces us to consider new layers of complexities. As platforms have developed, a number of them have evolved into multi-product or multi-services ecosystems. These forms create new challenges for competition law enforcement. The concept of ‘ecosystem’ comes from the strategic management literature (Jacobides et al., 2018). The term digital ecosystem typically designates a complex set including a generic modular technology which can be used to supply different types of products or services and a platform which uses the modular technology to produce a set of services which it combines with other services offered by third parties (complementors) in such a way that the attractiveness of each service is enhanced by the offerings of the other services. The business strategy literature distinguishes between two types of digital ecosystems (Subramanian et al, 2019).

Production ecosystems arise from the use of digital technology to better connect interdependent activities and enhance a firm’s value proposition through its value chain. An example of how a production ecosystem can enhance the value proposition of a firm is insurance companies’ use of digital technologies to gather data allowing them to offer new services, such as insurance contracts where premiums vary with the drivers’ recorded driving performances, in addition to selling their traditional products.

The second type of digital ecosystem, known as consumption ecosystems, which are the focus of this article, also use product or service generated data to identify demand-related complementary or interdependent services or products and to co-offer them either directly or through complementors. In other words, in consumption ecosystems, the platform makes the value proposition of each product or service offered by the digital ecosystem more valuable than if it were offered on a stand-alone basis. Such ecosystems are often based on modular technological algorithms which can be used to provide a wide variety of services.

With respect to consumption ecosystems, we can think of Apple as a multiple product based ecosystem which offers access to a multiplicity of devices and of services which can be accessed on the various devices to the benefit of consumers. Another type of digital consumption ecosystem would be Google, which is a service-based ecosystem built around the distribution of a free Android operating system to independent Original Equipment Manufacturers (OEM) and that provides multiple services (such as mail, office software, maps, cloud services, etc…) supplemented by third-party complementors through applications which can be accessed on the Google Play Store. Google services are related to one another in the sense that Google pools the consumer data it gets from its users to permit a high degree of service personalization across all services. Amazon, which was originally a digital book retailer, complements its offering with the distribution of numerous other products, as well as entertainment, video services, gaming services, payment services, business solutions, a cloud storage service (Amazon Web Services), and the offerings of a vast array of independent suppliers selling through its marketplace. Each Amazon business segment plays a role in the consumption of other types of products and services belonging to its ecosystem. For example, third-party resellers bring in the volume the tech giant needs to leverage its investments in fulfillment and allow it to experiment with innovations in retailing.

The participants in a digital consumption ecosystem may be competitors offering related services or competitors with the services provided by the platform who nevertheless have an interest in being part of the ecosystem due to the fact their service is more valuable to consumers when it is offered as part of a set of complementary services than if they offered it separately on their own. This means that within an ecosystem there can be a complex combination of competition and cooperation between the platform and the complementors (Dietz et al., 2020).²

The role of the core platform in a digital ecosystem is to provide, through the use of the modular technology, the distinctive core services or products which will be the basis for an attractive offer of complementary services or products, and to choose the additional products and services from independent providers in a way that maximizes the value of the whole ecosystem, as well as to establish the governance of the ecosystem and to resolve conflicts between its members (Iansiti and Levien, 2004; Iansiti and Lakhani, 2020; Jacobides, 2021).

To better comprehend what makes an ecosystem successful and what are the critical elements of competition between platforms and ecosystems, it is interesting to compare the insights one gets from the economic literature with those provided by business policy literature. The economic literature has largely focused on the role of network effects, scale economies, learning economies, and data accumulation as constituent elements of the path to success of the first movers with a natural tendency of markets to tip. These same factors are considered to be instrumental in protecting the largest ecosystems against entrants or smaller players. This raises the concern that the largest ecosystems insulated from competition may have less of an incentive to innovate than if they were subject to a more intense competitive pressure. However, the business literature offers a more nuanced vision.

If we now consider the business policy literature, Michael G. Jacobides, Nikolaus Lang, and Konrad von Szczepanski (Jacobides et al., 2019), for example, focus their research on the factors of success for a digital ecosystem. Their data show that to explain the success of a platform it is necessary to go beyond the widely held idea that ‘scale begets interest, which begets scale, leading to winner-takes-all markets’. ‘Being an early mover, is not sufficient to ensure long-term success or dominance, for several reasons. First, the ecosystem’s product or service may simply not be suited to some customers’ needs. Second, as technologies and customer needs evolve, later entrants can jump in and take advantage. And third, in their haste to be first, early movers sometimes build the wrong ecosystem or choose the wrong path’. As counterexamples to the idea that late comers are necessarily at a disadvantage, they point to the bankruptcy of Symbian, an early mobile operating system which had a 70% market share and to IBM Watson a latecomer smart-health space which in 3 years between 2014 and 2017 grew explosively, thanks in part to an automated and streamlined onboarding process and developed an ecosystem of 53 partners. The factors of success they identify are a strong user base (a leading share in a market), a large number of partners bringing expertise from other industries (e.g. Amazon has 67 core partners, roughly double the number of its e-commerce peers, thereby covering logistics, finance, media, and telecommunications), a big geographical footprint, and a robust capacity for cooperation because the core platform must manage dozens of partners, across multiple industries and countries, and different types of relationships. Reeves et al. (2019) provide further details on why ecosystems more often than not fail.

Along the same line, some economists have called into question the importance of network effects, winner take all (Casanova, 2020). For example, Jordi Casanova argues that the advantage of large platforms is mostly due to their ability to benefit from inter-market network externalities which allows them to have more users and operate on more markets. Consequently they are able to better target advertising and this allows them to get comparatively higher advertising revenues and to engage in more research and development (R&D). For the rest, the evidence Jordi Casanova presents supports the idea that second movers can copy what first movers have done and gain market share with lower investments than dominant players and that market leaders need to invest and innovate more if they want to keep their technological leadership over time.

4. Platforms and ecosystems: insights for competition law enforcement

There are a number of insights that one can derive for competition law enforcement from this brief consideration of ecosystems.

First, as mentioned earlier, technical barriers to entry associated with the size of the ecosystem may not be the most crucial element explaining the dominance of a few highly successful platforms. The design of these platforms and their organizational capacities may be equally important factors of success. This means that potential competition among platforms may be more significant than what the market shares or size of these platforms would suggest.

Second, understanding the factors which may determine the success of a platform may be useful when analyzing the acquisition of a start-up by a major platform to establish the counterfactual and get a sense of whether an alternative acquirer might have been a possibility or whether the start-up would have been likely to develop into a successful platform had it not been acquired.

Third, the description of these ecosystems reveals that while production ecosystems tend to remain within in the confine of their value chain (and therefore on the same relevant market as they were before the emergence of the digital ecosystem), the situation is different for consumption ecosystems where the digital platform seeks to develop the combination of complementary services or products which would maximize its value for its consumers, given its technology.

This has several implications for antitrust or competition law enforcement. First, the boundaries of a consumption ecosystem encompass a number of demand-side complementary product or services. In addition, the boundaries of such complements are open-ended. Additional or different combinations of complementary services or products can displace existing ones or non-digital firms on specific markets. Ecosystems can move fluidly from one market to a complementary market thanks to their modular technologies and their analysis of consumption generated data. For example, an e-commerce-based ecosystem can acquire extensive information about the consumption habits of consumers, their credit history, their location, their planned acquisition of a car, the kind of cars they may be interested in, etc. This information allows those ecosystems to have an edge, allowing them to add a loan function to the combination of services they already offer, thus competing with banks on consumer loans.

The notion of relevant market does not define either the locus of activity of digital consumption ecosystems or the locus of competition between ecosystems. An ecosystem providing a different set of more attractive complementary products or services may displace an existing digital ecosystem. This means that the market share that an ecosystem has on any particular market (and the existence of barriers to entry on this particular market) may not be relevant indicators of its market power³. This also means that when one assesses the competitive impact of a merger of consumer complementary services ecosystems in the digital sector, one should consider potential competition to be the set of alternative combinations of complementary products or services which could be offered in competition with those offered by the merging ecosystems. This set is, by the way, not restricted to ecosystems which have the same business model as the business model of the merging ecosystems. For example, the Google ecosystem was developed to compete with the then dominant Apple ecosystem even though the Google ecosystem is centered on a set of service which can be delivered on terminals, whereas the Apple ecosystem is based on a set of terminals which can deliver various services.

Thus, to understand who could be competing with whom in the digital sector, one has to look at a much wider set of possible candidates than in traditional antitrust cases where one can limit oneself to the potential producers of a specialized but clearly defined product. Understanding the trends of demand by platform users (so as to understand, e.g., that a photo app on mobile phones could compete with a social media on personal computers because social media users are migrating toward mobile phone services and that they increasingly feel that exchanging photos is a crucial part of socializing) is crucial to the definition of potential competition. Clearly, competition authorities need to invest in understanding the consumption trends of a fast-moving digital world.

Second, among digital platforms and ecosystems competition on innovation and on quality is the rule rather than competition on prices, at least when it comes to consumption ecosystems. Also innovations tend to be disruptive, that is they fundamentally change the value proposition of established dominant platforms or ecosystems and replace them with more attractive new service combinations. Quality is much more difficult to assess objectively than price and it is multidimensional. This raises two difficult questions for competition authorities, respectively, related to the relationship between competition and innovation and to the relationship between the size of ecosystems and the quality of their services.

The relationship between competition and innovation is complex and not always well understood and the relationship between competition and disruptive innovation is even less well understood (see Aghion et al., 2005; Baker, 2007). For example, competition economists debate the question of whether the acquisition of small innovative start-ups by large platforms (which under certain circumstances can reduce competition) accelerates the rate of innovation in the digital sector by giving an incentive to start-ups to innovate in order to be bought for a hefty amount of money or alternatively can reduce innovation (by allowing a large platform to prevent the innovations from developing independently). It has been suggested that large platforms engage in killer acquisitions of app developing firms when they feel that those firms could either compete with one of their products or services or become the basis of a new offer of complementary services competing with their ecosystems (see Parker et al., 2020). The allegation is therefore that they suppress potential competition through those mergers. But in such cases, the description of the counterfactual necessary to establish the anticompetitive nature of the transaction is particularly difficult. What would have happened to the start-up had it not been bought by the merging ecosystem? Would it have developed into a successful ecosystem or would it have remained an isolated specialized platform? If the service offered by the start-up is going to be integrated into the offering of the large ecosystem, is the merger good for consumer because it allows an innovation to reach the market, or is it detrimental to consumers because it eliminates a potential competitor? Unless competition authorities have a good theory of what makes a digital start-up grow and become successful, their assessment of the effect of such mergers will be controversial as Petit and Teece (2021) note, and they may have to turn to the business economic literature to find clues.

Third, the quality of the services provided by an ecosystem can result from the network effects and/or from the interaction between the complementary services it offers. The quality of service can also benefit from the increased access of the platform to data necessary to feed the algorithms it uses to produce the service offered. Thus, the increased size of an ecosystem may simultaneously allow it to degrade the quality of its services (e.g. with respect to the quality of data protection it offers) because of its increased market power while also allowing it to deliver higher-value services to its consumers. A challenge for competition authorities is to prevent quality degradation due to the increased market power of platforms while keeping the quality benefits due to the size of the platform. Deconcentration measures may simultaneously affect both dimensions, thus leading to uncertain results for users’ welfare.

5. Data as an input and privacy issues and competition policy

Data play several simultaneous roles for platforms and consumption ecosystems. They can be an input exhibiting increasing returns if used in an artificial intelligence context; they can be a strategic asset allowing the platform to enhance its economic power and prevent entry; or they can be a valuable by-product of the activity of the platform which can be commercialized and the respect of privacy can be a non-price parameter of competition (Ocello et al., 2015).⁴

Competition authorities have had only limited experience with the economics of data in the non-digital world since in the non-digital world the link between the accumulation of data and the quality of a service provided using this data is less pronounced than it is in the digital world.⁵ The broader problem is that competition and data issues interact in novel ways in this context (Kira et al., 2021).

One possibility which is of concern to competition authorities is that large ecosystems or platforms strategically deny access to their data to potential competitors (whether a competitor of the platform within the ecosystem or a competing ecosystem) when access to these data is necessary for the competitors to match the level of quality of the offers of the large platform ecosystem.

For example, in the EU a dominant firm must grant third parties access to data it has collected if (i) the data constitute an ‘essential facility’ to the activity of the third party asking for access (i.e. the data are indispensable for the business activities of the third party), (ii) there are no objective considerations that would justify a refusal to grant access, and (iii) it is likely that a refusal would exclude all competition in the other market and refusal to grant access to information has been considered abusive in the Microsoft case (European Commission, 2004).

However, in practice such an approach requires having the answer to several complex questions. The first question is whether the data gathered by the ‘dominant’ ecosystem could be duplicated (gathered through other means) by the competitors and, if so, at what cost disadvantage, if any. The second question concerns the conditions [organization, format (processed data or raw data), and timing] under which the data must be made available to competitors. The third question is whether such an obligation could backfire by reducing the incentive of large ecosystems to gather data which they could use to increase the quality of their services to the benefit of users or consumers. All these issues are being heatedly discussed in the EU in the context of the debate on connected cars. Automobile manufacturers of these cars favor the ‘extended car’ concept whereby the data generated by connected cars will be gathered on the external server of the car manufacturer and then selectively passed on to interested business users rather than being directly accessible in real time on the connected cars by competing providers of after-sales or complementary services (such as insurance companies or independent car repair shops).

In merger control cases among digital platforms, competition authorities may have different types of concerns. First, they may fear that the merger will allow the acquirer to get access to the data gathered and accumulated by the target, thus giving the merged firms an advantage which could not be duplicated by competitors. Second, they may fear that the merged firms may degrade the privacy protection of their offerings, thus degrading the quality of their service to the detriment of consumers or by implicitly increasing the price of their service.

In its examination of the Google/DoubleClick (European Commission, 2008) and Facebook/WhatsApp (European Commission, 2014) mergers, the EC has analyzed in detail for each case the effects of the combination of the two sets of the merging companies’ data and whether these mergers could have led to the foreclosure of competitors. The Commission, however, decided that these mergers would not create a competitive advantage which could not be replicated by competitors of the merging firms.⁶

An interesting aspect of the EC decision is its refusal to consider the potential effect of the merger on privacy protection issues of the data collection.⁷ However, there is no consensus within the EU on whether privacy-related concerns should be considered to be outside the scope of competition law.⁸ In the USA, the Federal Trade Commission (FTC) also cleared the Facebook/WhatsApp merger in 2014 (Federal Trade Commission, 2014) without an in-depth investigation.

The Facebook/WhatsApp merger decisions, however, turned out to be controversial on both sides of the Atlantic. In August 2016, after WhatsApp announced updates to its terms of service and privacy policy, including the possibility of linking WhatsApp users’ phone numbers with Facebook users’ identities contrary to what the parties had said, the Commission fined Facebook (but did not rescind its decision to allow the merger based on the fact that it had established through the ‘even if’ analysis that this link would not lead a competition problem on the online advertising market). This led to a controversy over whether, and if so why, the EC had missed a crucial dimension of the case by refusing to look at the implications of the merger for privacy protection.⁹

This view was echoed by Bureau Européen des Unions de Consommateurs (BEUC) the European consumer organization which stated (BEUC, 2017) ‘It is very disappointing that the Commission decided not to revise its original decision on the Facebook merger with WhatsApp. It is crucial in our data economy that competition bodies more closely scrutinize the potential consumer harm of a merger between data-heavy companies. The Commission failed to do so when it gave the go-ahead to the Facebook/WhatsApp takeover’. Consistent with Kira et al. (2021), Deutscher (2018) contended that: ‘commentators and antitrust authorities have not assessed to what extent the accumulation of personal data not only harms competitors and advertising customers, but might actually also have immediate negative effects on consumers on the “free” user side of online platforms’.¹⁰

In September 2020, 48 U.S. state attorney generals brought an action against Facebook to redress violations of Section 2 of the Sherman Act, 1SU.S.C.§ 2, and Section 7 of the Clayton Act, 15U.S.C. § 18, arguing, notably, that following its acquistion of WhatsApp, Facebook degraded Instagram users’ privacy by matching Instagram and Facebook Blue accounts so that Facebook could use information users had shared with Facebook Blue to serve adds to those users via Instagram.

¹¹

Time will tell whether the attorney generals are eventually successful in pushing a third approach to allow consideration of privacy within the context of antitrust or competition law (degradation of privacy as a reduction of consumer choice).

What is clear at this point is that there is considerable controversy on how data and privacy issues could be integrated into the competition assessment of ecosystems and whether they should be (Economides and Lianos, 2021). There is strong resistance in many jurisdictions to using competition law to deal with privacy issues but as there are cases pending in several jurisdictions, time will tell in which direction the courts will go.

6. Competition issues within ecosystems

In transaction ecosystems and in multi-service or multi-product ecosystems, within-ecosystem (i.e. competition issues arising among the participants of an ecosystem) competition issues may also arise particularly when the platform is competing with some of the business users of the platform or some the providers of complementary services (Jacobides and Lianos, 2021). The core platform of the ecosystem may refuse access to application that competes with its own applications or, if access to third-party applications is allowed, the platform may discriminate in favor of its own services to the detriment of the services offered by independent providers of complementary services or by business users of the platform. Furthermore, the platform may use some of the information it receives through the activity of independent businesses using the platform to piggyback on their successful offers.

In those cases, competition authorities tend to focus on the fact that the central platform of the investigated ecosystem has a dominant position on a service and on the fact that the platform abused its market power either by restricting access to the market for this application or a complementary application or by restricting the ability of its competitors to compete on these markets.

In the 2017 Google Shopping case (European Commission, 2017), where Google was fined €2.42 billion, the EC found that Google dominated the market for general algorithmic online search in national markets throughout the European Economic Area (EEA) and that it leveraged this dominance into the market for comparison shopping services by in general favoring its own shopping comparison service, Google Shopping, in general search results over third-party comparison shopping services. The Commission considered that Google systematically gave prominent placement to its comparison shopping service at or near the top of its general search results, while demoting rival comparison shopping services. As a result, the comparison shopping service of Google was much more visible to consumers in Google’s search results than those of its competitors, which led consumers to click more often on Google Shopping. Given the importance of Google’s search engine as a source of traffic, Google’s preferential treatment of its own comparison shopping service resulted in a substantial and durable loss of traffic for its competitors’ competing shopping services. Google’s practice was thus found to be an infringement of its special responsibility as a dominant firm, is not justified, and is capable of foreclosing competition in the comparison shopping market.¹²

There are numerous other examples of this nature in many jurisdictions.

As an example, Russia has dealt with three cases of this nature. In the Russian Yandex-Google case (FAS, 2016), the Federal Antimonopoly Service (FAS) concluded that Google had a dominant position in the market of pre-installed app stores for smartphones running on Android operating system (OS), based on the facts that on the one hand smartphone manufacturers using the Android OS represented more than 50% of the market for mobile phones in Russia since 2012 and, on the other hand, the Google Play Store which allowed the downloading of application was hugely popular and therefore pre-installed on nearly all mobile phones using the Android operating system.¹³ Similarly, in its case against Apple, it concluded that Apple had a dominant position in the distribution market for iOS applications.¹⁴

In 2017, following a complaint by Kaspersky Lab, the FAS investigated whether Microsoft had favored its own antivirus application and encouraged users to abandon third-party antivirus applications.¹⁵ As a result of this case Microsoft significantly changed its company policy. Microsoft modified the ‘Antimalware Platform Requirements’ that govern the relationship between the corporation and independent antivirus software vendors and eliminated all calls for the abandonment of third-party software. Similar cases have been opened elsewhere.¹⁶

Another type of case aims at practices designed to protect the position of a key service of a platform. The platforms may want to limit the strategic freedom of business users or of independent providers of complementary services active on their platform in order to protect themselves against the potential competition of other ecosystems or platforms¹⁷

Those cases raise several perplexing issues. First, in most of these cases, each application is considered to belong to a specific market and the question raised is whether competition on this specific market has been restricted.¹⁸

Yet we know from the business policy literature that the choice of the complementary services offered on a platform is a crucial choice for the ecosystem both in order to establish its identity and to increase its value for users in the competition among ecosystems. Therefore, the value of one particular service to an ecosystem may be much more important than the value of the service if it were offered on a stand-alone basis.

If we assume that the platform has a choice between several applications which provide the same kind of service whose value would, thanks to the complementarity with the other services offered on the platform, increase the value of the platform over and beyond the value of the service offered on a stand-alone basis, it is difficult to understand why a platform would not choose the application which best matches the other applications and services offered on the platform thus maximizing the overall value of the ecosystem. For example, if we go back to the FAS Russia Microsoft case, it would seem that in order to maximize the value of its ecosystem, Microsoft would have an incentive to choose the best antivirus available irrespective of whether or not it is its own antivirus. Similarly, Google has made the point that Google Search is better than other search engines not because it is pre-installed but that it is pre-installed because it is a superior search engine to those of its competitors.

Thus, first, the choice of which app will be chosen for pre-installation on a platform will have consequences not only on competition between the applications considered but also on competition between the ecosystem it participates in and other ecosystems Therefore, the consequences of the choice of an application made should be considered, by competition authorities, from the standpoint of competition on the application market and competition among ecosystems. Separating the two analyses does not adequately reflect all of the competition implications of the choice of a particular application.

Second, when an application that provides a service for users is denied access to a platform (or is a victim of the self-preferencing of the platform for its own application), it is only if we assume that the possibility to pre-install the competing application (and to have the platform treat this competing application in a non-discriminatory way) would either increase the quality of the platform application or the quality of the competing application over and beyond the initial quality of the platform application that there would be the possibility of a consumer welfare gain from increasing the competition for particular applications in the ecosystem. This possibility, however, cannot be assumed to exist in all cases. Indeed when the quality of an app is related to the intensity of its use (e.g. through the better training of artificial intelligence algorithms) there is a possibility that the quality of both the platform application and the competing application(s) on the same platform may be decreased by the fact that they compete for users on an equal footing (and that each one has fewer users than would be the case if there was only one application pre-installed). Thus, a case-by-case analysis of the quality dimensions of the applications considered and their relationship with the quantity of data collected by the platform is warranted.

Third, the usefulness of competition between applications on a platform (assuming they have equal access to the platform and are treated in a non-discriminatory way) will very much depend on whether or not users make competition work between these platforms. In other words it will depend on the ability of consumers to reward the providers of best quality applications. If, as in the Russian Yandex-Google case consumers are considered to be passive or inert, increasing their choices may not result in an increase in their welfare. We come back to this issue later on.

7. Competition between ecosystems

When we consider competition among multi-product ecosystems such as Apple or Google the standard analyses used in antitrust become even more difficult to use to assess competition issues among ecosystems, as there is a lack of criteria on how effective intra-ecosystem competition is (see Jacobides, 2021). A number of merger cases are relevant to this issue.

In the digital markets, large platforms have engaged in a buying spree often acquiring digital start-up firms with very low turnover at very high prices.¹⁹ A number of those firms were platforms with one or a few innovative functionalities using a specific algorithm which after the merger was integrated into the ecosystem of the acquiring platform. A few turned out to be enormously successful for their acquirers such as the acquisition by Facebook of Instagram (a photo application with 30 million followers) in 2012 for 1 billion US$, followed by the acquisition of WhatsApp (with 450 million followers) in 2014 for 19 billon US$. Today Instagram has more than 1 billion followers and Facebook has more than 2 billion followers.

From the competition standpoint, the concern about mergers in the digital sector is based on the idea that the acquired start-up firms can provide competitive pressure because they are able to siphon off or ‘cream skim’ customers and collect valuable data. Thus, the acquisition of start-up companies by large firms may benefit the incumbent by reducing competitive pressure of potential entrants on the periphery of the market or by preventing future entry and expansion by such start-up firms that could undermine the incumbent’s dominance. Nevertheless, an important question for competition authorities is how to analyze such mergers. It should be recalled, for example, that the Facebook Instagram merger was investigated by the Office of Fair Trading (OFT) in 2012 (OFT, 2012)²⁰ and it concluded that upon further review of this matter, it now appears that no further action is warranted by the Commission at this time. Accordingly, the investigation has been closed. However, less than 5 years after the merger, Instagram had more than 500 million active users and had become the second most popular social media network in the world, behind only Facebook (You Tube excluded), and the number of advertisers on Instagram had grown to more than 1 million, in part due to the fact that Instagram was then considered to be the best social media platform for customer engagement.

It is thus interesting to mention the fact that the FTC has decided in 2020 to revisit the issue of the acquisition of Instagram by Facebook (FTC, 2020).

In its 2020 complaint, the FTC writes: 13. (…) In February 2012, Mr Zuckerberg acknowledged that if left independent—or in the hands of another acquirer like Google or Apple—Instagram threatened to leave Facebook Blue ‘very behind in both functionality and brand on how one of the core use cases of Facebook will evolve in the mobile world.’ Emphasizing that this was a ‘really scary’ outcome for Facebook, Mr Zuckerberg suggested ‘we might want to consider paying a lot of money’ for Instagram.

14. Mr Zuckerberg recognized that by acquiring and controlling Instagram, Facebook would not only squelch the direct threat that Instagram posed, but also significantly hinder another firm from using photo-sharing on mobile phones to gain popularity as a provider of personal social networking. As Mr Zuckerberg explained to Chief Financial Officer David Ebersman in an email, controlling Instagram would secure Facebook’s enduring dominance around one of the few social mechanics that could provide a footing for a competing personal social networking provider:

[T]here are network effects around social products and a finite number of different social mechanics to invent. Once someone wins at a specific mechanic, it’s difficult for others to supplant them without doing something different. It’s possible someone beats Instagram by building something that is better to the point that they get network migration, but this is harder as long as Instagram keeps running as a product. [Integrating Instagram’s functions into Facebook] is also a factor but in reality we already know these companies’ social dynamics and will integrate them over the next 12-24 months anyway. The integration plan involves building their mechanics into our products rather than directly integrating their products if that makes sense… . [O]ne way of looking at this is that what we’re really buying is time. Even if some new competitors spring[] up, buying Instagram, Path, Foursquare, etc now will give us a year or more to integrate their dynamics before anyone can get close to their scale again. Within that time, if we incorporate the social mechanics they were using, those new products won’t get much traction since we’ll already have their mechanics deployed at scale.

 15. On 9 April 2012—the day Facebook announced it was acquiring Instagram—Mr. Zuckerberg wrote privately to a colleague to celebrate suppressing the threat: ‘I remember your internal post about how Instagram was our threat and not Google+. You were basically right. One thing about startups though is you can often acquire them.’

Three things are striking from the FTC complaint.

The first one is the acknowledgment by Mr Zuckerberg that there are a ‘number of different social mechanics to invent’. This confirms that ecosystems compete by offering differentiated combinations of services which are complementary in an effort to maximize the value to users of the interaction between the various sides of their core platform rather than by competition head to head offering the same combination of services. Mr Zuckerberg saw early on (as early as 2011) that photo sharing apps could become an essential element of a combination of services which would make a social media platform successful for mobile users. Having tried to develop such an app but recognizing that Instagram was superior to its own photo app, the Facebook founder decided to acquire Instagram to integrate its dynamic by offering a photo-application-based social media for mobile users. The acquisition made sense in the context of an evolving dynamic competition among ecosystems in which a new offering integrating an element which hitherto had not been a crucial part of social media offerings could become highly successful given the migration of a large part of social media users from personal computers to mobile phones. Furthermore, being the first to build a social media offering for mobile users offering an already developed but not yet widely distributed photo sharing application would allow Facebook to benefit first (and more effectively) from the network effects and give it an advantage over potential competitors.

Mr Zuckerberg suggested that, without the acquisition, Facebook could be displaced and Instagram including the functionalities of a photo sharing application and of communication among its users could become the dominant social media. It interesting to note that this comment implies that, as previously mentioned, competition among ecosystems is for the market rather than for a market share of the market. Thus, in this digital world of ecosystems dominance is not an indicator of lack of competition but rather the result of it but dominant forms are threatened by other ecosystems able to offer a more attractive combination of communication services to users.

The second striking element of this excerpt is that it helps us understand why the traditional tools of merger control used by competition authorities are inadapted to grasping the competitive implications of such a merger. Typically, competition authorities think in terms of existing markets narrowly defined at the time of the transaction and ask themselves whether competition on these markets might be diminished through a unilateral or coordinated effect. Thus, the OFT looked at the social media market as it was (and as it was assumed to remain) at the time of the merger and at the photo app market as it was at the time of the merger and concluded that there were competitors to Instagram on the market for photo applications and that there were competitors to Facebook on the market for social media. It also concluded that a digital photo application in itself was unlikely to become a social media in the short run. What the OFT missed (because it was thinking in terms of restriction of competition on existing markets) was the fact that the acquisition was the sign of possible transformation of the business model of social media with the emergence of an innovative social media product based on a combination of a photo application functionality and other communication functionalities among users.

The third striking element is the fact that there was a debate within the Facebook management about who was the most important competitor to Facebook: Google+ or Instagram. Thus within the company, understanding where the competitive threat was coming form was open to debate.

The US FTC, after conducting a nonpublic investigation of the US$1 billion Facebook Instagram merger in August 2012, cleared it at the time (while it decided to challenge it in 2020) possibly because, at the time, it feared that it would not prevail in court. Indeed, the Supreme Court of the United States in the Marine Bancorporation case of 1974 (US Supreme Court, 1974) established that ‘two preconditions must exist’ before an actual potential competition theory ‘establishes a violation of § 7 of the Clayton Act. The first condition is that the potential competitor could enter the market at issue absent the merger; the second condition is that such entry would produce a likelihood of deconcentration or other significant procompetitive effects’. Moreover, with respect to the first condition, the Court implied that ‘unequivocal proof’ of actual future de novo entry is required. Arguing on the one hand that a photo sharing application could become a social media and that therefore the merger was restricting potential competition (at a time when no such move from the photo sharing application market to the social media market had ever taken place) might have been difficult given the high standard of proof required by US jurisprudence. Alternatively, arguing that Facebook and Instagram were actual competitors on the market for the attention of users of social communication services and that their merger would significantly alter the structure of this market would have been equally difficult. Finally, convincing a judge that the merger was anticompetitive even though Instagram and Facebook would still be free for users after the merger, as they were before the merger, might have been equally difficult as US judges are used to thinking that transactions should be opposed if and only if they are likely to raise prices and thus reduce consumer surplus.

Similarly in 2014, the US FTC cleared Facebook’s $19 billion acquisition of the messaging application WhatsApp, although the FTC did send both companies a letter reminding them of their obligation to maintain privacy practices in accordance with the WhatsApp user agreement in place at the time that user data were collected (Federal Trade Commission, 2014).

There is something disturbing about the fact that the US FTC is now trying to challenge a merger it cleared by arguing, inter alia, that at the time it took place the goal and the effect of the merger were to eliminate the main source of competition that Facebook faced and to deprive its competitors from the ability to challenge its dominance.²¹ If the FTC is right today, it must be that its analysis was wrong at the time of the merger.

Overall, the static analysis of relevant markets, the narrowness of the interpretation of the consumer welfare standard, the high burden of proof when it comes to the risk that potential competition may be lessened, and the difficulty to integrate into competition analysis the inherent instability of business models of digital ecosystems make merger control as it is traditionally implemented an ineffective tool to identify problematic mergers in the digital world.

If our traditional tools of antitrust and merger control need to be modified when it comes to assessing competition among digital ecosystems, this is not to say that the underlying goal of competition, i.e. the protection of consumer welfare, is inadequate. There is no valid argument that consumers should not be protected against possible abuses of market power by large digital ecosystems. But we need to be able to reliably assess how consumer welfare can be defined for the users of the platforms who are offered services at a zero price and we must be able to distinguish between pro- and anticompetitive behaviors or transactions. Unfortunately, the competition analysis tools which work satisfactorily for the rest of the economy do not perform well in the digital world.

8. Behavioral issues: consumers’ status quo bias

Two European cases against Google, the 2017 Google Shopping case where Google was fined €2.42 billion (European Commission, 2017) and the 2018 Google Android case where Google was fined €4.34 billion (European Commission, 2018), share a common feature.

According to the Commission, by giving prominent placement to its own comparison shopping service and by demoting competitors, Google has given its own comparison shopping service a significant advantage compared to rivals because: The evidence shows that consumers click far more often on results that are more visible, i.e. the results appearing higher up in Google’s search results. Thus, a behavioral bias of consumers was alleged to explain why the behavior of Google was anticompetitive. Similarly, in the EU Google Android case, the Commission considered that tying the Google Search App and Google Chrome with Play Store resulted in the fact that Google Search and Google Chrome were pre-installed and impossible to uninstall on mobile phones on which Play Store was installed. This constituted a restriction of competition because it provided Google with a significant competitive advantage that competing general search service providers and competing non-OS-specific mobile web browsers could not offset. To justify this the Commission argues that: (781) ‘The reason why pre-installation, like default setting or premium placement, can increase significantly on a lasting basis the usage of the service provided by an app is that users that find apps pre-installed and presented to them on their smart mobile devices are likely to  “stick” to those apps.…. Users are more likely to try these applications/services based on their prominent visibility and once they are using them, they usually continue to do so. It is an easy way to obtain new users and deliver almost automatic stickiness for an application or service.’  (782) Users are unlikely to look for, download, and use alternative apps, at least when the app that is pre-installed, premium placed and/or set as default already delivers the required functionality to a satisfactory level. Once again, a behavioral bias of consumers is alleged to explain why the practice is anticompetitive.

These two cases follow in the footsteps of the EU Microsoft decision of 2007 where Microsoft was fined €492 million by the EU Commission for two abuses of dominant position, one of which was Microsoft’s tying of Windows Media Player with Windows, its OS.²² This significant competitive advantage for Microsoft exists, in spite of the alleged superiority of other media players, if and only if Windows consumers just use the media player which is installed on their computer and do not bother to look for a ‘better’ media player. Likewise, in the USA, the US Justice Department (DOJ) filed an antitrust lawsuit against Google, alleging it uses anticompetitive tactics to preserve a monopoly for its search engine and related advertising business and on Thursday December 17, 2020 a group of 38 US states and territories sued Google claiming that it illegally maintained a monopoly in general search and search advertising through anticompetitive conduct and contracts. The DOJ lawsuit argues that Google has entered into a series of exclusionary agreements that require that Google Search be set as the preset default general search engine on billions of mobile devices and computers worldwide and, in many cases, prohibit pre-installation of a competitor search engine. Interestingly, the DOJ complaint does not explain why it believes that the pre-installation agreements have foreclosed distribution opportunities to rival search engines given the fact that these rival search engines can be easily installed on mobile devices through Google Play.

These cases raise a number of troubling issues.

In order to find an abuse on the part of the investigated firms, the EC has to rely on the behavioral economic literature and argue (i) that when it comes to the digital products considered there is a failure of consumers to rationally choose what is in their best interest and (ii) that the practice of the firm investigated was a deliberate attempt by the dominant firm to exploit this demand-side market failure to the detriment of its competitors and ultimately of consumers. But there is no explanation of what causes this demand failure or whether the consumer biases are exogenous (observable whatever the characteristics of the competing products) or endogenous to the variety of products available (observable only when the competing products are so similar as to be indistinguishable).

It could be that consumers do not have the necessary information about what their options are. Consumer protection laws or privacy laws could help remedy this type of situation. But alternative explanations are also possible. For example, it could also be that consumers do not think that the minor qualitative differences between the products offered are sufficient for them to bear the cost of the search of an alternative.

If that is the case, it is difficult to see how the alleged violations decreased the welfare of consumers who chose not to look for an alternative to the pre-installed product in spite of the fact that this alternative was readily available or to understand how forcing them to exercise a choice that they do not feel they need to make would increase their welfare. It is equally difficult to see why suppliers of these alternative products would invest more in minor innovations if they had better possibilities of access to consumers if consumers did not value minor innovations in these products (and therefore did not base their choice on the technical differences between these products). Finally, it is not established that competitors of the dominant firms would be deterred by the pre-installation from investing in R&D for major innovations if such major innovations changed the attitude of consumers with respect to the benefit they could derive from seeking an alternative to the pre-installed product.

In addition, the lack of analysis of the importance of the assumed biases of consumers is an obstacle to determining whether such biases could result in a foreclosure of applications which are not pre-installed or are given preferential treatment. Indeed, invoking a status quo bias of consumers is insufficient if we do not know how important this bias is. Does it result in the fact that consumers never compare the alternatives they have? Does it result in the fact that consumers do not compare alternatives as often as they would if they were perfectly rational but still compare sufficiently frequently for the pressure of their choices to be felt by the suppliers of digital services? Those empirical questions need to be answered to get a sense both of the reality of the possible anticompetitive effect of the practices and of their severity.

Thus, understanding the reasons for the demand-side failure as well as the nature and the severity of such failures are crucial both for the qualification of the practice if one believes that the goal of competition is to protect consumer welfare and for the design of remedies, as the Commission found out in the Microsoft case by forcing Microsoft to offer consumers the option of a choice of media players which was a resounding failure.

9. Realpolitik of the role of gatekeepers

To tackle some of the challenges of competition law enforcement in the digital sector on December 15, 2020, the EC unveiled a proposal for a regulation on contestable and fair markets in the digital sector (European Commission, 2020). This proposal is an attempt to complement its existing enforcement powers which it considers are neither sufficient to address some of the competition concerns raised in digital markets nor have sufficient deterrent effect on the largest digital players.

One important part of the proposed regulation deals with the within-ecosystem relationship between some platforms (integrated or not) and third-party suppliers of complementary services. The obligations are designed to protect the businesses using the platform either to provide complementary services to the core services of the platform or as a marketplace against the possibility that the gatekeeper will use their data to outcompete them or will self-preference its own services to the detriment of the services these businesses offer by prohibiting the gatekeeper from discriminating, or imposing business restriction on the users of the platform and by facilitating the transferability of data and the interoperability of their services with the services offered by independent businesses.

The obligations would prevent the gatekeepers from combining personal data from their core platform services with data from other sources (including other services offered by the gatekeeper) or from restricting business users from contracting with end users outside of the gatekeepers’ ecosystems or from requiring business users of the platform to use, offer, or interoperate with any identification service of the gatekeeper in the context of providing its services via the relevant gatekeeper’s core platform services. Other self-executing obligations would protect the advertisers and publishers to which gatekeepers provide advertising services by imposing transparency obligation on the gatekeeper with regard to the price paid by the advertiser and publisher, as well as the amount of remuneration paid to the publisher, for the publishing of a given ad and for each of the relevant advertising services provided by the gatekeeper.

The second category of obligations ‘susceptible to be further specified’ would include: not using data acquired by the platform concerning business users to compete with those users unless the data is publicly available; allowing end users to uninstall any pre-installed software applications on its core plaform service; allowing installation and use of third-party software applications or software application stores (subject to certain carve-outs); not technically restricting end users from switching between and subscribing to software applications and services accessed under a gatekeeper’s operating system; giving business users and providers of complementary services access to and interoperability with the same operating system, hardware, or software features that are available or used by the gatekeeper in the provision of any ancillary services; refrain from using aggregated or non-aggregated data (which may include anonymized and personal data if this data is not publically available) to offer services similar to those of their business users; provide, on request, to third-party providers online search engines with access (on fair, reasonable and non-discriminatory terms) to ranking, query, click and view data related to free and paid search generated by end users via the gatekeeper’s online search engines, subject to anonymization of the quick, query and view data that constitutes personal data; provide business users with fair and non-discriminatory general conditions of access to the gatekeeper’s software application store.

In short, in proposing this kind of ex ante regulation the EC wants to be able to prohibit some of the practices it has examined in the past without having to define relevant markets, to assess market dominance or to bear the burden of establishing that these practices are capable of restricting competition. This is definitively a move away from an economics-based approach to competition law enforcement. It is noteworthy that Ms Vestager has publicly justified this proposal on the grounds that many of those practices were unfair to the business users of platforms.

It is somewhat difficult to follow the logic of this proposition from a competition standpoint. The definition of ‘Gatekeepers’ is, firstly, limited to providers of ‘core platform services’ [Digital Markets Act (DMA), Art. 2(2)]. According to the Commission, core platform services include (i) online intermediation services (e.g. marketplaces, app stores, and online intermediation services in other sectors like mobility, transport, or energy), (ii) online search engines, (iii) social networking, (iv) video sharing platform services, (v) number-independent interpersonal electronic communication services, (vi) operating systems, (vii) cloud services, and (viii) advertising services.

Furthermore, not all providers of those services will be considered to be ‘Gatekeepers’. The Commission states in its proposal (DMA, p. 2) that ‘(s)uch gatekeeper status can be determined either with reference to clearly circumscribed and appropriate quantitative metrics, which can serve as rebuttable presumptions to determine the status of specific providers as a gatekeeper, or based on a case-by-case qualitative assessment by means of a market investigation’.²³

This definition of ‘Gatekeepers’ is well adapted to activities which are dominated by very large platforms such as digital retail (with Amazon) or social networking (with Facebook).

However, its applicability raises some questions for sectors in which a number of competitors each operate a platform which serves as an important gateway for business users to reach the users of the platform.

For example, it is not obvious that the definition applies to automobile manufacturers of connected cars. Yet, in the extended vehicle model, which is the reference model that automobile manufacturers have pushed, each car manufacturer has a privileged and exclusive in-time access to the data provided by the connected cars of its make since these data are automatically transmitted by automobiles and stored on the manufacturer’s server. Because the car manufacturer is in potential competition with the independent providers of a number of after-sales services (such as car repair or maintenance shops or insurance companies) or complementary services (such as mapping services or audio visual services), the competitors are entirely dependent on the transmission by the manufacturer’s server of the information generated by the connected car in real time to be able to offer those services competitively with the car manufacturer.

Thus, although there is reason to believe that non-discriminatory access to the data gathered by manufacturers of connected cars is crucial for the independent after-sales service providers competing with them, it is not clear that each one will meet the quantitative thresholds set by the Commission for the rebuttable presumption of ‘Gatekeeping’ or that these car manufacturers will be (qualitatively) considered as having a significant impact on the EU internal market in the assessment of a market investigation. Much will depend on the interpretation given by the Commission of the relevant EU ‘internal market’.

Secondly, the applicability of the regulation is limited to platforms providing core services which have an ‘entrenched and durable position’ either at present or foreseeably ‘in the near future’.

But the quantitative elements on which the presumption of ‘an entrenched and durable position’ is founded (i.e. their turnover in the EU or the number of active end users or business users in each of the last 3 years) tell us nothing about whether they will have an entrenched or durable position in the future. Furthermore, as we saw, competition authorities have a difficult time understanding the logic of competition between ecosystems, and it is not clear that the EU Commission will be able, in a market study, to assess qualitatively which platforms have an entrenched and durable position and which do not (as mentioned earlier, the rapidity with which Instagram became a major social network was not recognized by competition authorities at the time of the merger between Facebook and Instagram).

Similarly, back in 2005 when MySpace was acquired by NewsCorp (Graham, 2009), ‘millions of teenagers across the world used MySpace, spending hours every day connecting with each other online and fine-tuning personal profile pages that reflected their tastes and personalities’. Yet within 3 years Facebook, which had been created in December 2004, had become a serious competitor and within 5 years (i.e. in December 2009) MySpace’s share of the social networking market has tumbled from 66% to 30%, with MySpace losing US $100 million in 2009. Facebook had expanded its user base, articulating a qualitative product differentiation between itself and its competitors and integrating new ways of engaging users into its suite of social networking functions by offering new features and functionalities (Glick et al., 2020). Facebook surpassed its main rival, MySpace, to become the most popular website in the United States in 2009, just 5 years after its founding. There is little doubt that had the proposed EU regulation been in force in 2005 MySpace would have been considered to have an entrenched and durable position. Yet it was all but gone within 4–5 years of its acquisition by NewsCorp.

As NewsCorp found out and, as the Competition and Markets Authority (CMA) demonstrated, in the previously discussed Facebook Instagram merger, it is particularly difficult to assess whether a platform has an entrenched position on a market and what seems to be a ‘durable’ entrenched position can, in fact, evaporate very rapidly.

What the EU Commission’s approach seems to ignore is that competition among ecosystems comes about through disruptive innovation and a fight among ecosystems for the market rather than in the market. In such circumstances, the past is a poor predictor of the future (Petit and Teece, 2021).

Third, some of the obligations to be imposed on the ‘gatekeepers’ platforms contemplated by the EU proposal seem to misunderstand the function of an ecosystem and could actually restrict competition or innovation of such ecosystems in the name of fairness or of protecting competition within an ecosystem.

For example, preventing gatekeepers from combining personal data from their core platform services with data from other sources (including other services offered by the gatekeeper) is likely to restrict the information a platform can use to improve its value proposition to consumers. This proposal seems to ignore, on the one hand, the fact that in an ecosystem it is the provision of complementary services which increases the value of the ecosystem for its users and, on the other hand, the fact that for a number services offered, the quality of the application is, ceteris paribus, increased by the amount of data used to train the artificial intelligence algorithms which allow the provision of these complementary services.

If it seems sensible to argue that the consumers should have the right to oppose the use of their data for other services than the ones for which they provided their data, a prohibition per se of the combination of data from gatekeeper core services with the data from other services is very likely to limit the quality of the services they render and hence the welfare of the users.

Similarly, the obligation for a platform which is also a marketplace to not use data acquired by the platform in relation with business users to then compete with those business users, unless the data are publicly available is clearly likely, in the name of fairness, to decrease the intensity of competition among the participants to an ecosystem to the detriment of the welfare of the users.

Asymmetric data sharing regulatory obligations or interoperability obligations imposed on platforms seen as ‘gatekeepers’ may allow smaller competitors to upgrade the quality of their service offering and overcome their disadvantage of size or allow independent service providers to offer services to users of the platforms in competition with the platform itself. But such obligations may also decrease the incentive of platforms to accumulate data and have a deleterious effect on the quality of services offered to consumers and on innovation.

As Kerber and Schweitzer (2017) argue, interoperability obligations can lead to a higher level of homogeneity among platforms To the extent that uniform standards and interfaces are used, the possibilities of firms to develop their own specific products and services are limited, because they have to comply with these standards and interoperability requirements. This will limit the scope for innovation and therefore the extent to which specific consumer preferences can be fulfilled by way of product differentiation. Although greater interoperability may lead to more innovation and competition with regard to complementary products, it also can lead to less innovation and competition with regard to the standards and interfaces themselves, which may have the characteristics of natural monopolies (with all their negative consequences). Furthermore, the openness of products and platforms for complementary  products can lead to higher risks for consumers, if the complementary products offered by other firms are not monitored closely with a view to their interoperability, quality, and safety. …The complex trade-offs between benefits and costs of a higher degree of interoperability suggest the need for a careful and separate analysis of each specific interoperability issue, caution regarding a (top down) imposition of mandatory standards and interoperability obligations, and a greater focus on unilateral solutions of interoperability problems, such as adapters or converters. Within the framework of Art. 102 TFEU, EU competition law may be better advised to develop a workable test to address hurdles for unilateral interoperability solutions created by dominant firms, than to continue focusing on the essential facilities doctrine to mandate interoperability.

If data portability is ineffective to counter the network effects that dominant platforms may enjoy, it is sometimes considered useful to alleviate any data-induced lock-in effect. However, even this solution may not be effective in all cases. For example, Gabriel and Weinberg (2019) note However compelling in theory, few have investigated whether competitors can actually use ported data to create or grow competing platforms This gap is particularly troublesome because we found no competitive products built on ported data, despite the fact that many large platforms have enabled users to export their own data for years.  … An overreliance on data portability may distract from more effective tools for addressing concerns with large platforms.

The regulatory framework for gatekeepers proposed by the EC as a complement to competition law enforcement raises three questions which deserve further elaboration respectively concerning its scope, the underlying analysis of platform competition, and the remedies proposed. These questions no doubt will be (and should be) further discussed in the next few months during the consultation phase on this proposal.

First, the proposal does not seem to aim at solving the competition issues raised by gatekeepers in the digital sector in general but limits itself to the sub-set of these problems raised by a small number of very large platforms without providing a clear rationale for this choice. Thus, it is difficult to avoid the impression that this proposal is driven more by the political desire to act against these large platforms than by the desire to promote competition and innovation in the digital sector in general.

Second, the proposed regulation seems to have partially overlooked a number of specific features of competition within and between ecosystems. In particular, it overlooks the essential role of core platforms in monitoring the complementarity beween the services they offer and those offered by their third-party complementors within the ecosystem in order to promote competition between ecosystems. It also overlooks the fact that restricting access to the data core ecosystem platforms use to train their algorithms may have the undesirable consequence of lowering the quality of services offered to their users. Finally, competition among ecosystems is not competition for market shares among firms in preexisting markets, but competition for the market through disruptive innovations.

Third, with respect to remedies, the proposal does not seem to sufficiently take into consideration both the trade-offs involved in the interoperability of platforms and the limits of data portability.

With respect to merger control, article 12 of the proposed regulation (European Commission, 2020) creates an obligation for ‘Gatekeepers’ to notify any intended merger or acquisition that involves, within the meaning of the EU Merger Regulation, another provider of core platform services or of any other services provided in the digital sector. The obligation, which applies to all transactions whether or not they would trigger a notification under the EU or the national merger law, is designed to inform the Commission on whether there is any need to revise the individual gatekeeper designation and to help the Commission monitor the general trend in the digital sector. Thus, the EC proposal is different from proposals in other jurisdictions which envisage a derogatory merger control review process applying to certain digital firms.

However, this does not mean that the EU Commission is not intent on controlling mergers by large firms with start-ups having a very low turnover in order to prohibit killer acquisitions.

On September 11, 2020, Ms Vestager, the EC Commissioner, in a speech to the International Bar Association, rejected the idea of an amendment to the EU Merger regulation in order to be able to control possible killer acquisitions (Vestaeger, 2020). However, Ms Vestager stated that from mid-2021, the EC will accept referrals, made under Article 22 of Regulation (EC) No. 139/2004 (known as the ‘Dutch clause’), by Member States of mergers affecting trade between Member States but which do not have Community dimension and which the Member States Competition Authorities are not competent to review under their national merger control rules.

It is thus likely that the EU Commission will be able to review most mergers which could be killer acquisitions in the digital sector (or in other sectors such as the pharmaceutical sector). However, in light of our discussion of the substantive issues raised in the digital sector, it remains to be seen if the Commission will be able to distinguish between pro- and anticompetitive mergers in this sector.

10. Conclusions

In a pragmatic sense, we are in a period of flux, but we can expect tightening of competition law. The US under-enforcement of competition law in the digital sector cannot be sustained much longer, both for political and for technical reasons. But the assertive enforcement of competition laws in the digital sector by other jurisdictions, on the other hand, has so far been based on shaky analytical grounds, a poor understanding of competition within and between ecosystems and of their interaction and confusion about the goals of competition law. If the EC proposal to regulate the digital sector purports to give the EU Commission new means to intervene in the digital sector, it needs to be seriously refined to avoid the risk that it may backfire and ultimately restrict both competition and innovation in the digital sector.

In order for competition law to become more relevant in the digital sector, competition law methodology and instruments need to be adapted to the complex specificities of this sector, whatever the jurisdiction, as the challenges with platforms and ecosystems are real. This adaptation has not taken place for a number of reasons.

First, competition authorities are generally reticent to say that the instruments they have relied on in the past and the methodologies they have used have limited applicability in the digital sector for fear that the use of a new conceptual apparatus and new instruments for this sector would meet the skepticism of judges who value the stability of jurisprudence more than the reliance on new economic thinking or tools.

Second, there is a difficulty for non-specialist regulators to accept the explanations of the business leaders of the digital sector on why and how competition in the digital world is different from what it is in non-digital sectors. Competition authorities having a long experience of listening to business leaders in all sectors of activities claim that competition law provisions should not be applicable to them the way they are applicable to other businesses because of the specificity of their activity. These authorities are, in general, rightfully skeptical of such claims and so they are biased toward relying on their past successful theories of harm rather than trying to adapt to a new economic logic.

Third, the fast growing academic literature on the digital sector started with business policy articles discussing the business models of digital firms and has only relatively recently attracted the attention of market competition specialists. There is thus a genuine difficulty in understanding what this literature implies for competition law enforcement.

The review of past decisions and the consideration of the business policy literature on the digital sector, however, do suggest a few urgently needed improvements to competition law enforcement in the digital sector.

First, there is a need to deepen our understanding of barriers to entry in the digital sector. The economic literature has largely focused on the role of network effects, scale economies, learning economies, and data accumulation as constituent elements of the protection that the largest platforms enjoy against entrants or smaller players. However, as we saw, some economists have called into question the importance of these factors (Casanova, 2020) and found that second or third movers actually benefit from an advantage over the first movers.

Second, there is a need to find new indicators of market power for platforms and ecosystems than those used for single-market firms (indicators which are largely based on cost price comparisons), taking into account the fact that platforms are multi-sided, that users on various sides interact, that, on some sides, the services are offered for free, and that, on these sides, competition on quality is crucial to attract the maximum number of users which is the objective of the platforms.

Third, there is a need to recognize that competition between ecosystems is, to a large extent, the result of differentiation of the combination of services that the ecosystems offer to their users and that this competition has a winner take all dimension. This has a number of implications. It means first that market shares (however they are calculated) at any one time may not be an indicator of dominance in the future since a completely different combination of services may replace the currently existing dominant ecosystem. This also means that to assess potential competition among ecosystems one has to look, first, at the extent to which the technology of various specialized platforms (even if they offer very different services than the ecosystem considered) could be used for the distribution of a number of other complementary services and, second, at the evolving demands of platforms users for a combination of services. Finally, this means that there is a need to use the teachings of the business policy literature to understand the factors of success of a digital start-up in order to assess whether the acquisition by a platform at the center of a large ecosystem of a small application producer is likely to restrict competition or is likely to spur more innovations.

All of this requires investment in research staff which have not only a technological understanding but also a business understanding of digital ecosystems, including an understanding of the dynamics of users’ demand for ecosystem service—i.e. the demand trends.

Fourth, when dealing with competition issues among applications within an ecosystem there is a need to understand and take into consideration the interaction between these issues and the competition between ecosystems in order to be better able to differentiate what is an anticompetitive practice from what is a procompetitive practice.

Fifth, there is a need to further research the role of data in competition among platforms and of innovation in digital services. In particular, it would be useful to explore the characteristics of cases where access to the data gathered by an ecosystem or by a platform is necessary for competitors of this ecosystem to be able to compete effectively as well as to define/explore the kind of data we are talking about. One should also explore the conditions under which interoperability and data portability would unambiguously enhance competition, the cases where they could actually make it more difficult for entrants to grow, and finally the conditions under which the quantity and quality of data gathered by ecosystems would be affected by the conditions put on the use of these data. There is also the related question of the scope for convergence on the issue of whether privacy is a relevant element of quality which can or should be dealt with by competition authorities.

Sixth, there is a need to better assess the origin, the nature, and the importance of the biases of consumers digital which may help dominant platforms in an ecosystem to foreclose the producers of competing applications. In many cases the vague mention of the existence of such biases is insufficient to assess the competitive implications of these biases or to design appropriate remedies. In particular, it would be useful to ask what we know about the implications of consumer biases in various sectors, how competition authorities have dealt with this issue in non-digital sectors, and what lessons we have learned from these experiences (or from the disregard for this issue in other sectors).

Fulfilling this (admittedly large) research agenda is a prerequisite if we want to ensure that competition law enforcement and the design of the regulatory environment of digital services are well founded analytically, relevant, and predictable. An exigent intellectual effort is the only way to ensure that competition authorities will avoid the risks of inadvertently giving in to the political pressure of economic populism or ideology or issuing misguided decisions which may be ineffective or, even worse, restrict competition or innovation.

Funding

None declared.

Footnotes

References