The “Acceptable” Cartel? Horizontal Agreements Within EU Competition Law: Introduction

It is uncontentious that hard core cartels are always contrary to competition law, an observation that holds true in every system that has antitrust provisions regulating restrictive agreements. Yet it is equally accepted that literalism is not a hallmark of contemporary antitrust policy. Accordingly, although competition law is inherently skeptical of coordination between would-be competitors, not every horizontal combination is treated as a hard core cartel as such. The prohibition of hard core cartels thus encompasses a focus on both form and substance. To put the point somewhat glibly, it is not enough that a combination looks like a cartel and even walks like a cartel; it needs to “quack” like one too. This straightforward observation opens the door to what may be termed “acceptable” cartels: instances of horizontal coordination that would appear on their face to be plain vanilla cartel behavior, but where the broader context provides a degree of legitimation. How we go about drawing the line between acceptable and unacceptable horizontal coordination from the perspective of both competition law and practice is, however, the rather thornier question that is addressed in this special issue. This problem has come to the fore in the past decade in the context of Article 101 of the Treaty on the Functioning of the European Union (TFEU), which regulates anticompetitive agreements,

concerted practices, and decisions between associations of undertakings with an effect on competition in the internal market. 3 Several factors have brought the issue to prominence, most notably the deliberate departure from formalism embodied in the European Commission's embrace of a so-called more economic approach to competition enforcement 4 and the move to self-assessment by defendant undertakings under the Article 101(3) exception rule. 5 The result has been a flurry of jurisprudential development, involving both Commission enforcement practice and decision-making by the Court of Justice, which has already attracted considerable academic attention. 6 Even as the case law in this area becomes ever more dense and rich, the precise delineation of "acceptability" in the context of nominal cartel behavior arguably remains as elusive as ever. It is this task-namely, throwing greater light on those factors that inform the antitrust assessment of horizontal coordination in a variety of circumstances-which occupies the authors of the articles in this symposium issue.
Although this is a question of compelling contemporary relevance under Article 101, the prompt for the symposium is the centenary of a rather earlier case, McEllistrim v. Ballymacelligot Cooperative Society, 7 which addressed the problem from the perspective of the common law restraint of trade doctrine. McEllistrim concerned the restrictive membership rules of a milk cooperative, which essentially prohibited farmer members, on a permanent basis, from selling their milk to competing dairies. Perhaps unexpectedly for a local issue arising in a remote area of rural Ireland, the dispute reached the House of Lords in 1919. It is something of an irony, when viewed through the lens of contemporary competition law, that the restraint of trade doctrine was historically more suspicious of, or at least inhospitable to, vertical restraints than to what today would be considered cartel conduct. 8 As Morten Hviid explains in his contribution which reimagines the McEllistrim case from the perspective of Article 101, the majority in the House of Lords focused on the purportedly vertical relationship between each farmer member as individual supplier and the cooperative society as the single purchaser of his milk output. Viewed from this perspective, the restrictive rules were held to constitute an unreasonable restraint of trade, limiting the freedom of farmer members to sell to the most advantageous purchaser: a contentious ruling at the time in a still largely agrarian society increasingly exposed to the effects of trade competition in agriculture. Yet the design of the rules-and, indeed, their 3. Article 101 TFEU in fact comprises three interlinked limbs: a prohibition rule in Article 101(1), proscribing forms of coordination between firms that have the "object or effect" of restricting competition; a sanction of voidness in Article 101(2); and an exception rule in Article 101(3), which exempts from application of the prohibition rule certain arrangements which fulfil four cumulative conditions that aim at, in essence, identifying the presence of sufficient consumer welfareenhancing countervailing efficiency benefits.  underlying competitive impact-were of course horizontal, involving in effect a collective effort to pool resources and not to deal with competitors. In McEllistrim, only a single judge, Lord Parmoor, made this connection, and his minority judgment hints at a more economically informed understanding of the cooperative form which is more attuned to contemporary competition analysis. As Hviid argues, the ostensibly restrictive nature of the arrangement thus did not translate readily into harm to either competition or consumers, demonstrating how a de facto cartel may nonetheless be acceptable in antitrust policy terms. The claim that ostensibly restrictive horizontal coordination may in practice be desirable as a means to achieve welfare-enhancing outcomes is a recurrent theme in recent enforcement at EU level. 9 The judgment of Cartes Bancaires, in 2014, was arguably a watershed moment, in which the Court of Justice emphasized most clearly the crucial distinction between prima facie restrictive practices and those which actually entail harm to competition. 10 The upshot, in that case, was that the mere fact that a horizontal agreement contained a clause that was designed precisely to constrain entry into a certain market segment did not qualify it for designation as a "by object" restraint within Article 101(1), because the restraint could be explained by reference to its potential beneficial impacts across a multisided market. That did not preclude the possibility, however, that the restraint may nonetheless be found to have an appreciable harmful effect on competition in practice. 11 This theme has been developed further in the Generics preliminary ruling, which involved an alleged "pay for delay" arrangement between potential competitors in the pharmaceutical sector. 12 Here, the Court expounded the role of procompetitive effects, nominally the purview of exemption under Article 101(3), in determining the existence of a restriction of competition under Article 101(1). It again stressed that harm to competition is the primary concern of the prohibition rule in Article 101(1), and thus recognized a role for procompetitive effects at that stage, albeit limited to the extent that such effects may call into question "the objective seriousness of the practice concerned." 13 There is rather less recent jurisprudence on the Article 101(3) exception rule, a perhaps inevitable consequence of the switch to self-assessment in 2004. There have, however, been calls for greater positive guidance from the Commission, not least because the existing case law may set an unduly demanding threshold for exemption of horizontal coordination on this basis. 14 In this regard, Hviid's article again attempts to situate the factual context of the McEllistrim case within the framework of the Article 101(3) exemption criteria, offering an assessment of whether contemporary competition law may prove more receptive to the ostensibly restrictive (and ultimately prohibited) behavior at hand.
By positing that certain forms of horizontal coordination are acceptable from a competition policy perspective, the symposium theme naturally raises the question of what makes obvious cartel conduct so unacceptable in antitrust terms. Two contributions in this issue address this question from distinct standpoints. Angus MacCulloch explores the subject of cartel wrongfulness, examining how the representation of cartel conduct as a type of public wrong may inform the outer parameters of the "object box." By considering several more peripheral examples of cartel conduct within Article 101namely, information exchange and cartel facilitation-MacCulloch persuasively makes the case that the absolute prohibition of hard core cartel behavior is targeted at the protection of the market as a public institution, and not merely, for instance, at behavior that diminishes overall welfare. In doing so, his work identifies a clear metric by which to measure the acceptability or otherwise of horizontal coordination, and thus a means by which to delineate or rationalize the rather amorphous notion of a cartel found within EU law.
Taking a more doctrinal approach, Niamh Dunne attempts to characterize the hard core cartel concept as it emerges from recent practice under Article 101. Adopting a loose definition of cartels as horizontal coordination over key competitive parameters, undertaken for purely private profitmaximizing purposes and typically involving a degree of secrecy or deception, her contribution similarly focuses on the outer parameters of what is recognized to constitute cartel behavior. Highlighting an apparent disjuncture between anti-cartel enforcement and the evolving jurisprudence on the object/effect divide within Article 101(1), the aim is to bridge the gap between theory and practice, again in an effort to define more clearly the underlying public policy concerns at issue.
Moving beyond theory, moreover, recent case law has emphasized the extent to which the broader legal and economic context can feed into the assessment of allegedly anticompetitive coordination. 15 This is of acute importance in the context of horizontal agreements in particular. An issue of obvious contemporary concern is the potential treatment of putatively restrictive agreements between competitors entered into in an effort to fight the climate crisis. 16 While few would doubt the worthiness of the underlying objective, questions remain of the extent to which the competition rules should and will apply with full force to constrain such private action. Another contentious issue is whether the competition rules apply to combinations of so-called gig economy workers: typically independent contractors and thus "undertakings" for the purposes of Article 101, yet often among the most precarious in the labor force, and thus in most acute need of collective action. 17 Again, it is difficult to defend an undiluted application of the competition rules to such activities; yet drawing a clear line between those independent contractors who are prohibited from cartelizing, and those who are in effect permitted, is a difficult task. 18 The contributions in this symposium issue tackle two high-profile problems in this vein. Kati Cseres explores an important yet underresearched question, namely the application of the EU competition rules within the agricultural sector. The Common Agricultural Policy (CAP) continues to dominate farming markets in Europe, providing extensive carve-outs from application of Article 101 for various forms of coordination between agricultural producers. Yet as the CAP itself moves toward a more market-oriented outlook, the Court of Justice has confirmed that its derogations should be construed as restrictively as possible, thus setting up an uneasy tension between the economic and social objectives of the CAP. Cseres explores the source and implications of this tension, alongside the somewhat dubious remedy provided by recently enacted Directive 2019/633 on unfair trading practices in the food supply chain. Her detailed and thoughtful contribution illustrates the lack of easy answers here and again highlights the inherently fact-specific nature of determining what makes a prima facie cartel acceptable from a competition policy (or indeed, broader social policy) perspective.
Finally, in a very different context-that of the digital economy-Pieter Van Cleynenbreugel offers an innovative means to address algorithmic collusion or so-called robot cartels. Rather than arguing for an extension of the competition rules to provide a bespoke solution to this anticipated future competition problem, Van Cleynenbreugel suggests that one may already exist in the guise of the somewhat neglected "decision of an association of undertakings" concept within Article 101(1). Contending that a purposive-but by no means implausible-interpretation of the concept could encompass the application of autonomous algorithms by online platforms, he uses this as the basis from which to argue for the development of an effective co-regulatory strategy to protect competition in the e-commerce sphere. Thus, the symposium issue concludes on what is hopefully a positive note: delimiting the boundaries of acceptable horizontal coordination from an antitrust perspective may be difficult in certain instances, but it is not an insurmountable task. We hope that the articles in this issue assist readers in better understanding and undertaking this exercise in future.