Third-generation competition law

Abstract

This note outlines three key areas of transformation that competition law has gone through in the past few decades and considers the direction it should take from here. The idea is to juxtapose the role of ex ante regulation with ex post competition law interventions in getting to grips with novel types of harms that the digital economy has brought about and that increasingly pertain to consumers, the market, and society as a whole. There are essentially two debates around the role of competition that have influenced the manner in which the law contends with data-driven markets. One relates to the goals of competition law.1 Much legal doctrine has assumed that we need one single theory to make competition policy consistent—we may refer to this as first-generation competition law. Another relates to the means by which competition law achieves its goals.2 Here, legal doctrine has focused predominantly on entitlements and market structure—we may call it second-generation competition law. The issues underlying these debates have recently been catalysed by several high-profile empirical investigations that observed rising concentrations of corporate power3 and the associated claim that competition law should be employed to restructure markets, by breaking big technology companies up or by divesting them into separate units.4 Regardless of whether one leans towards ex ante regulation or ex post competition law interventions in dealing with problems of concentration and their harmful economic and social effects, the result of these debates is a set of shared assumptions which have produced the coding structure that has become the node between theory and practice. Analysts have internalized these assumptions, although different camps put their own normative twists on them. The problem is that competition law’s coding structure has become redundant: it no longer represents the deep-seated realities of data-driven markets. In highly concentrated marketplaces, competition law and regulation inevitably work together, depending on and colliding with one another, but neither entirely holds sway. Against this backdrop, competition theory and doctrine can no longer serve one single goal nor can they continue to rely exclusively on entitlements or market structure. And yet much legal scholarship still turns upon these debates.5 The Digital Markets Act (DMA) is perhaps the most pertinent example of this trend.6 Data-driven markets require a distinct frame for conceptualizing novel types of harms—third-generation competition law. This is especially relevant for stalwarts of ex ante regulation who steadfastly maintain that issues around product quality in digital markets are outside the purview of competition law. By insisting that competition law pursues one single goal—usually in terms of low price (and high output)—this perspective severely prejudices any potential reflection as to how the law contends with issues posed by markets with a zero price and frustrates the ability to arrive at an effective resolution of such issues. Apologists of ex post competition law interventions must also adjust their views. There are good reasons behind their belief that decentralization matters; however, they frequently insist on an increasingly antiquated notion of how competition operates, in that the only disciplining mechanism in terms of leading actors’ behaviour consists in affording consumers an ‘exit’ option, that is, an opportunity to switch between different alternative offers. The image of the perfectly competitive market, for instance, is exit-centred to the extent that sellers are unable to displease their customers because consumers will swiftly shift from one supplier to another if they are dissatisfied with the products and services they receive. And if switching turns out to be impossible or unworkable, second-best alternatives such as separations or divestitures are thought to reinstate the ability of consumers to switch, pushing concentrated power out to the ends of the Internet and defying any consolidated arrangement involving small groups of dominant undertakings, as though no alternatives prevailed. This note will briefly survey these debates, but the emphasis is on what lies ahead—what ought competition theory and doctrine to resemble in the digital economy? Without a doubt, any answer to this question is unlikely to assume the sophistication and precision that have marked previous debates; however, this is precisely the point of this note. Disputes around data-driven markets regularly turn upon the major Internet platforms’ illegal acquisition of small competitors, their character as monopolies, and their preference for their own related products through network algorithms, resulting in harm to competing producers. Moreover, there are concerns around the platforms’ size and reach, which highlight the risk of suffocating progress, intensifying inequality, and exacerbating social and political polarization.7 To address these harms, policymakers regularly maintain that competition law should be employed to break the companies up or to divest them into previously acquired entities. Proposals such as these, however, contain a complex tension: because the dominant platforms’ activities present risks in economic and non-economic terms, competition law either needs to be repurposed, to deal with a broader set of values, or is required to refrain from contending with non-economic issues altogether—issues that are better dealt with by laws designed specifically to target privacy, data security, fake news, and so forth. The problem with this view is that distrust in the sheer size of big technology platforms, and the reach of their activities, obscures the network benefits that these platforms generate and that they may augment by expansion in the future.8 The tension is in large part a consequence of the debate around the goals of competition law: whether competition law needs to be focused exclusively on consumers, competitors, or both, or whether it should also attend to the social and political harms of market power. First-generation competition law cannot adequately capture the realities of digital markets that implicate seemingly non-economic values such as privacy, diversity, or editorial integrity. It presupposes claims that are based upon price and output when in fact these claims relate mostly to (various types of) product quality.9 Bringing competition theory and doctrine up to date would require theorists to acknowledge that competition law cannot be understood as an avenue to promote one single set of values. Instead, it should be seen as a practical device to underpin consumers’ calls and to align the economic goals of leading actors with the views of their own customers. To be sure, it is perfectly reasonable to argue that competition authorities and courts should focus on one single goal because they lack the required knowledge to make rules that can govern unknown future circumstances in dynamic settings. In the digital economy, however, decisions taken by authorities and courts inevitably affect the policies adopted by dominant incumbent actors, thereby shaping the broader legal and economic landscape. For this very reason, authorities and courts with limited knowledge may adopt instead a blend of different enforcement strategies available—because they are sometimes ignorant. Although this may at first seem counterintuitive, adopting a blend of different enforcement strategies enables authorities and courts to confront incentives to rule broadly while at the same time working towards greater convergence of different pertinent values such as privacy, diversity, editorial integrity, and competition.10 First-generation competition law inexorably collides with this point. That is precisely why stalwarts of regulation and apologists of decentralization regularly pit the values of competition against those associated with special purpose regulation. The camps’ common supposition is that we need regulation or competition, and the mere issue that persists is to assess which realm a particular problem is best associated with. And while debates around the goals of competition law may have shifted our perception as to whether a particular issue belongs to regulation or competition, these debates have barely influenced the manner in which we contemplate the workings of concentrated markets. These quarrels notwithstanding, with the newly enacted DMA, ex ante regulation and ex post competition law interventions now govern together in a space that has relentlessly been negotiated and contested. Contrary to what stalwarts of regulation would suppose, the DMA does not mean that regulation will prejudice competition, that it will impinge upon the continued application of Articles 101 and 102 of the Treaty on the Functioning of the European Union (TFEU) to the behaviour of big technology platforms or that competition law will gradually become redundant. Nor does it mean, as apologists of decentralization would maintain, that we should no longer be required to reconsider the assumptions of first-generation competition law. The DMA creates a regulatory regime that is aimed at ensuring fairness and contestability and when it regulates it does not hold unfettered sway any more than competition law wields unencumbered control when it is engaged.11 To be sure, as an ex ante regulatory instrument governing core platform services offered by entities that qualify as ‘gatekeepers’, the DMA reigns supreme, but in effect, it creates a space of regulatory overlap in which the actual choice is not between ex ante regulation or ex post competition law interventions; rather the choice is both as well as and, and this assumes some significant degree of common ground that accompanies that interdependence.12 Digital markets arguably distort the classic distinction between production, distribution, and consumption. Decisions taken by consumers in the digital economy are increasingly made passively, through implicit or explicit product matching and personalized recommendations rather than through active search. Take artificial intelligence-based algorithms as an example. Such algorithms recommend and purchase products based on spoken, written, or inferred requests obtained from users of digital platforms or devices such as mobile phones, speakers, and smart assistants.13 These algorithms benefit consumers in that they facilitate swifter and more complex transactions, and diminish consumers’ search and transaction costs. But they also work to entrust the ability of consumers to actively search—including their capacity to pick and choose—to big technology firms themselves and place consumers’ decision-making power into the hands of producers. This creates opportunities for misaligning the preferences of consumers with the goods and services actually sold, causing harm to customers, competitors, and the market (including society) as a whole.14 To remedy such harms, competition law usually ensures that consumers have an opportunity to choose between different alternative options. This is the equivalent of the conventional idea in which efficient markets afford consumers an ability to switch between distinct providers so that they can pursue their own commercial transactions at their individual discretion. But in concentrated markets, consumers typically are no longer able (or indeed willing) to choose effectively between alternatives. This is why, in theory at least, competition law is supposed to break big technology companies up—to reinstate the ability of consumers to choose between different alternative offers. Second-generation competition law grew out of its fixation on consumer switching as the only disciplining mechanism in terms of market actors’ behaviour. Even in highly concentrated markets, most of the time, there are thought to be alternatives, particularly where competition by other actors is just one click away. If switching nonetheless ends up being impracticable or turns out to be ineffective, competition law requires structural separation, to reinstate consumers’ choices, and to lessen the adverse effects of dominant incumbent actors on competitors. The main cognates of this account are regulatory arrangements that seek to preserve consumer switching through data portability or interoperability measures.15 Such measures seek to make it easier for participants to compete and to restore the autonomy of consumers, including their ability to choose. As to the predominant role that switching plays in promoting competition, second-generation competition law has not attained a middle ground in the debate around ex ante regulation and ex post competition law interventions; instead, most policy proposals fit into two groups. One is about entitlements: regulation should afford consumers a right to switch—an ability to choose between different options, put in place by mandated interoperability or established rights to data portability, which are geared towards both animating rivalry/lowering barriers to entry, and allowing new firms to compete. The other is about market structure: policymakers are concerned that the marketplace itself does not provide consumers with adequate opportunities to switch. The solutions they present almost inevitably involve forcing big technology companies to separate from (some of) their essential services to reinstate consumers’ choices and to lessen the adverse effects of dominant incumbent actors on other players. Both approaches seek to re-establish competition, by rendering users’ data portable and competing firms’ services interoperable. Portability provides consumers with an entitlement to switch; structural separation and/or interoperability afford consumers the ability to choose between different alternative offers by dint of adjustment to market structure. As with the shared assumptions underpinning first-generation competition law, second-generation competition law is difficult to square with the realities of data-driven markets. It rests on the notion of outright regulation or sheer and an perception of how data-driven markets As concerns the and structural are not the only at not the most for consumers to In concentrated marketplaces, consumers are most to have their by some in leading actors’ decision-making as to the on which those actors or their Here, consumers’ preferences can best be put into if users and their in this a to leading actors’ product The of such typically through an avenue of it ensures that in concentrated consumers can a of market to a significant second-generation competition law has the notion that entitlements and market structure are what most for data-driven markets, law in fact that competition authorities and courts work to consumers to leading actors’ product For in the into the that the most effective to competition and consumers’ not to break the up but to require to into the users to out of the data In a the European in upon the to that competing services no than its own its search of consumer preferences into leading actors’ has also been to be adequate in to claims against the to or and are with to dominant As the fact that second-generation competition law does not account for the manner in which data-driven markets it is to that in data-driven markets, and to consumers to from to Take the in In this the understood that been and that has in a of digital companies on consumers’ from platforms to As a separation of the would only have opportunities for entities to compete in and to confront the underlying incentives that the an alternative avenue of it consumers to of terms and of forcing the platform to control of the and of data to consumers, consumers in a to control the of data Second-generation competition law has the that of consumer of its focus on entitlements and market it has not any space for consumers to some control product and sellers by to them. If we the coding structure of second-generation competition law, competition theory and doctrine would be required to what may be the of an extent of manner in which authorities and courts and effectively the policies of leading actors on of consumers, and thereby work to consumers’ views into the market when other of are or In the DMA itself a of to the behaviour of that not only to promote consumer switching but that also to this regulatory overlap and are thought to be the rule rather than the law, for its will have to put emphasis on entitlements and market structure and more on of consumer the role consumers in shaping leading actors’ policies from their rather than to the part they in them from does the DMA for competition theory and The DMA an emphasis on the regulatory but the is more In effect, the DMA that the time is to the that has arguably the of competition theory and competition law. This is an issue that legal scholarship has yet to for both and normative In terms, stalwarts of regulation and apologists of decentralization have been in a the and benefits of but they have on one single between means and Both camps have supposed that competition goals and that the most problems are best exclusively by of regulation. That turns out to be The DMA has what most theorists thought to be the part of the debate around regulation and This is because competition law also a regulatory this holds for and As concerns the competition law a significant structural role in that a of in Articles and DMA from behaviour that competition authorities and courts previously and that relate to that is best understood in terms of Here, competition law regulates in that it is in the is competition law ensures that the DMA DMA that the may a market for the purpose of whether one or more services the digital should be to the of core platform services in DMA or for the purpose of that the contestability of core platform services or that are and which are not effectively by the in its the any relevant of Articles 101 and 102 digital markets well as any other relevant market investigations DMA may some in new in with the the DMA that competition law will a role in its and new rules are and are the result of Here, the in which an issue relates either to ex ante regulation or to ex post competition law interventions ends up being than previously been the DMA (and new will continue to in part on that around of market that is, around how to a significant role in governing markets is for consumers as a Moreover, competition law assumes a role in of the new The in for made it that a extent of market concentration inexorably incentives for services to be towards the and from the preferences of consumers, the market, and society at these leading search social and may through their algorithms and the of other actors, with in terms of product production, and In to be the not only to adjust the and but it also the manner in which policymakers and theorists the underlying It this on of consumers and decision-making on the of the to the for the on in and related and and and/or adopted by in different The DMA several in which by consumers, put in place through has been to ex post competition law interventions with ex ante regulation to contestability and fairness in digital markets. and in concentrated are difficult to they need to be and from through authorities and courts on of consumers as the of the of new laws those and The DMA that taken by authorities and courts can break ground for novel policies that leading actors to consumers’ views into their when and have gradually become more taken on of consumers does not require a debate of issues on an but a in a and in are out through a mechanism of in which and reasonable is such a opportunities for by the law, are in terms, and the main actors are and that their views ex The between and their is distinct from that which authorities and courts on of consumers ex that in is based on the of a rather than on mere assumptions or And it is for consumers to a more perception of how the policies of leading actors work in especially in of the and of data-driven markets. who to with a view to a through are to have a better of the circumstances of a and they may hold of a broader of Moreover, consumer may in in to behaviour and in to new so that authorities and courts can consumer more than the in a competition law also more ends in of the on its the of regulation groups of actors with different of knowledge and from those in competition law including of of and the focus of regulation is typically on that are from the effects of based on actual the are for the most part they are on from past investigations and are up by the shared and of authorities and courts that seek to in a manner that is for incumbent actors to In this the DMA to a of good in markets, that core platform services offered by are to of the arrangement that the DMA is based upon by and large from law that has been put in competition law. It the of consumers in the of leading actors’ to incentives and to underpin consumers’ calls rather than to promote one single set of values. One that the why there are different camps is that different camps hold distinct of how markets That to the point as to why competition law calls for a need to the camps competition law is different from what either but it is also different from what either significant The DMA offers to a middle ground between the camps in to As the DMA, investigations by competition authorities and courts will a Although the DMA in to on competition law is also to the regulatory This is where competition law and the DMA are to in and to different law may not be but it will also be for shaping the new is where the digital economy and competition law will Such may the of entirely new or a towards different market And where it is that market investigations DMA will to the of new core services and of competition are the competition investigations the for new most significant that need to be understood they are taken into the regulatory law is not or by instead, competition law To be sure, there are concerns the application and of the in the DMA But for of the such from competition law and thereby its stalwarts of regulation nor apologists of decentralization are to view this arrangement as But nor will either be to the DMA represents a to debates around data-driven markets. these to the competition theory and doctrine need to be in them are a of assumptions that no longer the realities of data-driven markets. In such markets, regulation and decentralization are neither the DMA nor competition law instead, they govern for about a of this has entirely been by competition theory and It is time to them up to date with the of competition law.