The “Public” Wrong of Cartels and the Article 101 TFEU “Object Box”

Abstract

The criminalization of cartel behavior in the UK turned attention to the criminal wrong at the heart of cartels. An understanding of the true nature of the cartel problem can also be used to develop a better understanding of Art 101 TFEU. This article uses the literature on the wrongfulness of cartels to examine how cartel behavior has the “object” of restricting competition within the terms of Art 101 TFEU. It pays particular attention to cases at the periphery of cartel behavior. The literature focuses on the importance of free markets as public institutions. Cartels are perceived as being a species of “cheating” which deserves opprobrium as it goes against the legitimate expectations of market participants. A reexamination of the cartel periphery cases involving information exchange and cartel facilitation using this lens shows a novel understanding of how these cases fit within the Art 101 TFEU “object box.”

Introduction

The “object” and “effect” distinction is an important feature of Art 101 of the Treaty on the Functioning of the European Union (TFEU). The distinction between the two concepts bifurcates the analysis of anticompetitive agreements into those that are prohibited because they have the “object” of restricting competition, and those that only have the “effect” of doing so. “Object” agreements have always been treated as being the more potentially serious of the two categories, and in practical terms, “object” analysis will always precede any analysis into an agreement’s “effects.” As the Court of Justice of the EU first set out in STM, ¹ the tests are alternatives rather than cumulative, but the Court has done relatively little to help define what factors or forms of analysis should be used to determine what agreements or clauses within agreements would fall within the object category.

The debate surrounding the extent of what has become known as the “object box” in Art 101 TFEU has recently been very active. In a series of significant cases, most notably Allianz Hungaria ² and Cartes Bancaires, ³ the CJEU has been required to expand on its rationale for categorizing agreements as being object agreements. The CJEU has set out a more complete definition that refocuses the examination of object agreements on agreements that “reveal a sufficient degree of harm to competition.” ⁴ The first part of this article outlines the cases that have helped shape this new test and the academic debate that has sought to address its significance.

This article then seeks to take that debate in a new direction. Most of the analysis has not considered cartel type activity; presumably as it is well established that cartels, as the “supreme evil” of antitrust, ⁵ are always object agreements. That is undoubtedly true of classic cartel activity on markets. This article seeks to look beyond the classic economic harms of cartels—seen through the monopoly rent or overcharge that cartels can command during their operation—and examine other significant types of harm that can arise from individual cartelist’s behavior. A deeper understanding of the wrongs or harms at the heart of cartel activity, particularly those harms that stem from the making of cartel agreements rather than their implementation, is found in the discussion surrounding cartel criminalization. In order to successfully criminalize cartel activity, it is important to capture the individual actions that make cartel activity the type that deserves deployment of the law’s most serious criminal sanctions. Where is the criminal act that lies at the heart of a cartel? The literature on cartel wrongs has identified a number of important concepts that are shared across several different conceptions of cartels. This article draws on Green’s conception of “cheating” as an example of how moral theories of crime can be applied to cartels. ⁶ These theories explain that it is the “public harm” of cheating—and by analogy also cartels—which justifies its criminalization. These wider public harms will be used to help understand why the making of a cartel agreement is, of itself, behavior that reveals a sufficient degree of harm to categorize it as an object agreement.

In the final part of this article, the wider public harm of cartel activity is used to reexamine a number of recent cartel cases that are outside the classic model of long-term cartel implementation. I refer to these cases a cartel periphery cases, as they are on the edges of cartel liability. The main examples I draw on are information exchange and cartel facilitation. These cases have been more difficult to categorize as those being held liable for cartel activity have not implemented the cartel agreement on the market. This analysis will show why some of those behaviors reveal sufficient public harm to be properly categorized as an object agreement under Art 101 TFEU.

Defining the Extent of the “Object Box”

The classic form of words used in the object box approach ⁷ can be found in cases like BIDS, ⁸ which was repeated in several cases: the distinction between “infringements by object” and “infringements by effect” arises from the fact that certain forms of collusion between undertakings can be regarded, by their very nature, as being injurious to the proper functioning of normal competition. ⁹

While oft repeated, this formulation leaves as many questions as answers—it confirms that there is a category of behavior that has the object of restricting competition “by its very nature” but gives little assistance in deciding what should be in the box. ¹⁰ It is widely accepted that the subjective intention of the parties cannot be the determining factor but rather that a determination of the objective purpose, or aim, of the agreement is the most appropriate means of categorization. This approach can be seen as developing a taxonomy built up from case law in which certain types of restriction are categorized as being either inside or outside of the object box. ¹¹ Such an approach may have benefits in terms of legal certainly; most obviously, if the concept is narrowly defined and there are a small number of largely hard core restrictions to be found inside the box. That being said an approach based on taxonomy is not well aligned with EU competition law’s current trajectory away from legal formalism and the adoption of more economically sophisticated approaches.

The early cases, such as STM ¹² and Consten, ¹³ were in reality more economically sophisticated than was demonstrated by their latter application. ¹⁴ One of the first cases to reopen the contemporary debate on the nature of object agreements was the judgment of the General Court in GlaxoSmithKline. ¹⁵ The General Court suggested that an “abridged but real” competitive analysis should be undertaken to establish whether an agreement would result in detriment to the final consumer and fall within the object category. This is sometimes termed as a “quick look” analysis. This approach was very controversial, and the CJEU did not confirm it. It referred back directly to STM and set out that the courts should consider, “the precise purpose of the agreement, in the economic context in which it is to be applied.” ¹⁶ In order to assess the anticompetitive “nature” of an agreement one should look to, “the content of its provisions, the objectives it seeks to attain, and the economic and legal context of which it forms a part.” ¹⁷ Although the “quick look” approach was rejected, there was still an invitation to take a somewhat wider examination of the context in which an agreement would operate rather than simply look at its objective purpose.

In Allianz Hungária, ¹⁸ the CJEU put forward a number of ostensibly conflicting formulations in short order. First, it suggested that one should look to the “content of the agreement” reveals “a sufficient degree of harm to competition.” ¹⁹ It then went on to repeat the BIDS formulation of certain agreements being, “by their very nature, as being injurious to the proper functioning of normal competition.” ²⁰ Finally, it suggested that to establish to question of object one would need to examine the “real conditions of the functioning and structure of the market.” ²¹ It is entirely possible to see the first and final element as requiring some form of contextual analysis into the existence of harm, whereas the “by its nature injurious” test is a more traditional examination of the purpose or aim of the agreement, perhaps in context, but not necessarily looking at its end result. There was clear potential for confusion as to the form and nature of analysis that was required. Could one simply rely on a long-standing taxonomy of object agreements that were accepted as having the purpose of restricting competition or should we now be seeking to examine whether an agreement would cause “sufficient harm” in the context of a “real” market? ²²

The CJEU, thankfully, had an opportunity to revisit a number of these questions in Cartes Bancaires. ²³ Here, the Court, while still citing BIDS as authority, subtly changed its approach, moving more toward the new wording found in Allianz Hungaria. It set out that, “certain types of coordination between undertakings reveal a sufficient degree of harm to competition that it may be found that there is no need to examine their effects.” ²⁴ Also that, “certain types of coordination between undertakings can be regarded, by their very nature, as being harmful to the proper functioning of normal competition.” ²⁵ It also went on to explain that certain behavior “may be considered so likely to have negative effects” that analysis of its actual effects would be redundant. ²⁶ The shift here is to focus the analysis of object agreements squarely onto “harm” as the guiding principle. This arguably moves the analysis of object agreements away from their objective aim or purpose toward an analysis of their harm. The reference to harm is not to some kind of abridged analysis of effects or quick look but an examination of the “likely” effects of that agreement. If it is “likely” to cause sufficient harm, it will be unnecessary to examine its “actual” effects. This still begs the question of how one is supposed to determine the sufficiency of likely harm? Here, the Court returns to the market context: “regard must be had to the content of its provisions, its objectives and the economic and legal context of which it forms a part.” ²⁷ This includes “the nature of the goods or services affected as well as the real conditions of the functioning and structure of the market or markets in question.” ²⁸ It is also interesting to note that the Court also took time to address the question of the parties’ intent in relation to object. They made it clear that while showing the parties’ intent was not a necessary element, it could however be taken into account.

A number of attempts have been made to explain the nature and rationale of the object box. Before the Cartes Bancaires clarification, Bailey had suggested that the approach acted as a form of presumption which would catch presumably harmful behavior. ²⁹ He also argued that the separate treatment of object agreements was efficient in that it does not waste administrative cost in unnecessary investigations: “the circumstances in which horizontal price fixing will be beneficial are sufficiently rare that it is not worth incurring the public and private costs of an investigation to identify them.” ³⁰ While supportive of the approach, he was more critical with regard to the proper approach to deciding which agreements should be treated as object agreements. He rejected simple taxonomy of agreements or reasoning by analogy from what had been decided before, as it did not “yield an intelligible, generally accepted, and consistently applied theory of restrictions by object.” ³¹ He argued that the starting point of analysis should be the content of the agreement and its ostensible purpose. ³² Drawing on Visa International, ³³ an early interchange fee case, and GSK, he accepted that contextual questions are important but argued that a “quick look” approach would diminish the “certainty and administrability” of the object approach. ³⁴ He suggested that an examination of the context of an agreement would provide a “reality check” to determine an agreement’s real purpose. ³⁵ Bailey also argued that a high likelihood of inherent potential harm would be sufficient to find there was an object agreement. ³⁶

The literature after the Cartes Bancaire judgment shows the continuation of similar thinking. The focus has shifted to the analysis of harm, to respect the change in wording, but the rationale behind the different treatment of object agreements remains similar. Brankin expressed real concern about the Cartes Bancaire test’s shift to harm as a test: “[t]he degree of harm caused by a restriction necessarily depends crucially on the market context in which it operates and could only be assessed on the basis of effects analysis.” ³⁷ He sees harm in the context as primarily being welfare loss, and on that basis that:

It simply cannot, therefore, be maintained that the object/effect distinction is based on the fact that certain restrictions inherently—i.e., “by nature”—cause more harm (or serious harm) than others. ³⁸

When discussing Cartes Bancaires’ new formulation, he takes the view that the harm caused must “justify” the avoidance of effects analysis. ³⁹ Following a discussion of the AG’s Opinion Brankin expressed support for a “net effects” analysis: the assessment whether it is “easily identifiable (i.e. clear) that the unfavorable effects on competition outweigh the procompetitive effects (i.e. has net negative effects).” ⁴⁰ He sees this standard as solving the problem whereby some naked hard core horizontal restrictions, which are object restrictions in the overwhelming majority or cases, are sometimes not harmful when deployed in particular contexts. The examples he uses are the noncompete clause in Maxima Latvija ⁴¹ and information exchange in Bananas. ⁴²

The rationale behind both of these approaches is based on a form of procedural efficiency, or presumption, that certain types of conduct are highly likely to cause competitive harm, either because it was their aim or purpose or it was highly likely to have the “net effect” of doing so. Given that likelihood, it was not an effective use of resources to prove to the requisite legal standard that it would have that effect.

Another “bottom-up” approach has been suggested by Ibáñez Colomo and Lamadrid de Pablo. ⁴³ They recognize the institutional incentives in tackling “clear cut” cases, where a case-specific approach is not necessary. ⁴⁴ Drawing on patterns from the case law, they suggest that object cases are those where there is “presumed to have a net negative impact on competition” ⁴⁵ or to put in another way they ask whether the practice can be “plausibly pro-competitive.” ⁴⁶ If there is no plausible procompetitive reason for the practice, it can be presumed to have anticompetitive purpose. They suggest that the actual practice of the Court reveals that they:

typically start by considering the rationale behind the practice, that is, the reasons why a firm would resort to it in the legal and economic context of which it is part. If this analysis leads to the conclusion that there are no credible pro-competitive reasons for its implementation, the EU courts typically conclude that it is restrictive of competition by object. ⁴⁷

Information exchange is again used as an example—in this instance, information on credit information for borrowers in Asnef-Equifax. ⁴⁸ This is an unusual example, as it was a horizontal exchange of information between competitors that improved the functioning of the credit market by reducing information asymmetry and was therefore plausibly procompetitive.

As we can see from this analysis, there is increasing consensus that there are both institutional and normative reasons why the object approach is enduring. As economic understanding develops, we may see agreements move in and out of the “object box.” It is also, I think, now clear that the context of an agreement is important—as an agreement which may be seen as an object agreement in one context may not been as such in another. This may be the case as, in the particular context of that market, there is a procompetitive rationale underlying that restriction.

The change in the test to make the concept of harm central makes it more difficult to rely on the aim or purpose of an agreement, but if one relies entirely on economic ideas of welfare loss to encapsulate harm, there are also problems. It is easy to see how an implemented cartel is likely to have a net effect of restricting competition, but it is more challenging to see how a nonimplemented cartel or low-level information exchange would have such a net effect. Are they still object agreements, or is some other element required to explain their inclusion?

In the next section, I will use the recent literature on the moral, or “public,” wrongs of cartels from the debate surrounding cartel criminalization to show one way in which we may be able to better understand the context, or presence of harm, in some practices which should be object restrictions.

The Public Wrong at the Heart of a Cartel

One of the most interesting features of the academic debate that has mushroomed in the UK surrounding the criminalization of cartel behavior has been a focus on establishing the wrong that lies at the heart of a cartel. This focus stems from a need, within the criminal law, to identify the innate criminality of behavior to justify the use of criminal sanctions as an appropriate response. In that debate, the focus is not on the economic harm caused by the implementation of a cartel agreement. As criminal law holds a particular individual responsible, it centers on their personal actions within the cartel. That change in focus gives us an opportunity to look at the cartel in a different way. After setting out the nature of that discourse, I will show how it might also be used to further inform the discussion of the harm that stems from object agreements.

Green’s work in relation to white-collar crime has been influential in that he developed a tripartite test to assess whether behavior deserves criminal opprobrium. ⁴⁹ The three elements of his test are (1) culpability, (2) social harmfulness, and (3) moral wrongfulness. ⁵⁰ That analysis can be applied to cartels to show how they fit within that framework. Culpability is the least problematic—it can be dealt with through the concept of intent. The others require a little more explanation but also provide useful insight that we can use in the context of Art 101 TFEU.

Social harmfulness is explained by Green as being some form of lasting setback to another’s interests, ⁵¹ but he notes that in white-collar crime, the nature of harms caused can be problematic, because of their complexity, when compared to traditional street crime. Cartel behavior exhibits many of the problematic aspects that Green identifies: a complex pattern of behavior, the diffuse nature of cartel victims, the mixing of inchoate, and completed criminality, and that criminal behavior takes place alongside legitimate activity. It is however relatively easy to identify that cartels and therefore, the actions of cartelists do result in significant economic harm. However, there is also another important requirement for criminal conceptions of harm: the distinction between public and private harms. The public/private divide is bound up in both conception of harm and wrongfulness. ⁵² It is a feature of the criminal law that it does not concern itself with private matters; a breach of contract may cause harm and may be seen as morally wrong, but it is not a criminal matter. There are private mechanisms for dealing with such problems. The harm should be “the sort of harm that somehow properly concerns the community as a whole rather than just the individual citizens within such community.” ⁵³ Individual overcharges could be seen as private matters, best dealt with through compensation, and therefore, the concept of harm we see as stemming from a cartel must be a wider, more “public” form of harm that should be of concern to the whole community. I shall return to this point once we have considered wrongfulness.

It is in the final element where the greater debate is to be found—that of the “moral wrongfulness” of cartel behavior. A number of different explanations of cartel wrongs have been suggested in the literature. Harding and Joshua suggested “delinquency,” ⁵⁴ Williams has suggested “exploitation,” ⁵⁵ and Harding has also posited “defiant willingness.” ⁵⁶ In a previous article, I suggested that the wrong could be set out as “subverting the competitive process.” ⁵⁷ But the most commonly discussed, and probably most useful, idea that has been drawn upon by several commentators—namely Beaton-Wells, ⁵⁸ Wardhaugh, ⁵⁹ and Whelan ⁶⁰ —is that of “cheating.” The criminal law has a long history of criminalizing various forms of cheating. If the actions of cartelists are a species of cheating, it would explain why cartels are sufficiently morally wrongful to justify the use of criminal sanctions. For the purposes of this article, it will also give us insight into another form of wrong or harm that stems from that activity.

Green’s work is again influential, he defines cheating, as recognized by the criminal law, as covering situation where, “X must (1) violate a fair and fairly enforced rule, (2) with the intent to obtain an advantage over a party with whom she is in a cooperative, rule-bound relationship.” ⁶¹ Many of the early cheating cases involve gambling or games of chance, and one can see how the context becomes a key part of the offense. ⁶² In the context of cartels, we can see that we are not involved in a game of chance, but it is entirely possible to see behavior as being contrary to the “rules of the game.” In the contemporary economy, the rules of the game are the “rules” of the free market. Market participants enter into the market in the expectation that all the parties are operating on a level playing field and making deals to their mutual benefit. As Green puts it:

[t]he basic social dynamic of cheating is as follows: the cheater and his victim are engaged in a mutually beneficial cooperative enterprise, such as a game, a market, or a political contest. In order to make the scheme work, the parties have implicitly agreed to restrain their liberty by adhering to a series of rules. In restricting their liberty in this way, those who share the scheme’s benefits have both an obligation to follow the rules and a right that the others follow them. In violating a rule that others follow, and thereby breaching an obligations to restrain his liberty in the manner agreed, the cheater is able to gain an (unfair) advantage over those who abide by the rules. ⁶³

Thus, we can say that market participants “restrain their liberty” and agree to be bound by all the rules that are associated with the free market. Those rules are designed to facilitate trade. The “cheat” within the market is that, knowing that the other participants are abiding by the rules, the cartelist steps outside the norms of behavior in order to get an unfair advantage. It is the mutuality of the market relationship which makes the behavior morally wrongful. It is not the seeking of advantage, as that is to be expected from those on the market, but it is the deliberate stepping outside the established norms to gain that advantage that is objectionable. The other market participants are cheated.

We can see that this form of wrongfulness can be combined with the ideas of “public” harm to explain the criminality of individual’s actions in cartels. It is not the overcharge moving down through the supply chain that we are concerned with—here we are concerned with the wider harm to the market as a public institution. This was highlighted by Wardhaugh: [t]he cartelist’s primary harm is the effect it has on the market as an institution used by society to distribute goods and services, i.e. as its means of distributive justice. The cartelist’s harm is therefore not the harm it occasions to any participant in the market. ⁶⁴

This harm is felt by all members of the public as it is a harm to an institution, the free market, which the public values as fundamental part of the economic system upon which we all rely. ⁶⁵ That is a different type of harm from the harm to a private welfare interest of any market participant. ⁶⁶ It is the “reasonable expectation” ⁶⁷ of a free market or a “fair environment for exchange” ⁶⁸ that is harmed by the cheat. It is for these reasons that I suggested that the cartel offense should be centered on “subverting the competitive process.” ⁶⁹ The free market, with the competitive process intrinsic to it, is a vital and protected institution.

Art 101 TFEU has always recognized the existence of private harms of cartel activity, but it has shown little recognition of these public harms. I argue that both forms of harm should be taken into account. Cartel harms are not only economic costs to customers and consumers but also harms to confidence in the market as a vital public institution. Article 101 TFEU can recognize this wider public harm in its object analysis, especially now that the object test has recognized harm as it central concern. Now I turn to examine what this would mean for cartel periphery cases under Art 101(1) TFEU.

The “Public” Harm in Cartel Periphery Cases

Cartel agreements that are implemented cause obvious economic effects and, therefore, there is little to be added to the analysis by considering the additional public harm that they may cause. ⁷⁰ In cases where they are less clear-cut downstream welfare effects, it is more likely that public harms will provide useful assistance. In this article, I examine two types of cartel case where I see this new analysis as being potentially useful. These are both examples of what I term cartel periphery cases, as they are on the edge of cartel liability. First, I shall look at information exchanges and then at cartel facilitators.

Information Exchanges

The classic example of an information exchange case is T-Mobile. ⁷¹ In this, preliminary ruling involved a single information exchange at a meeting regarding a reduction of standard dealer remunerations to take effect later that year. The Dutch court referred a number of questions about the application of Art 101 TFEU to such a scenario. The CJEU, after setting out the classic definitions of concerted practices, stated that it was inherent in competition law that “each economic operator must determine independently the policy which he intends to adopt on the common market.” ⁷² This presumption lies at the heart of the “very nature” of competition. It therefore follows that Art 101 TFEU:

strictly preclude[s] any direct or indirect contact between such operators by which an undertaking may influence the conduct on the market of its actual or potential competitors or disclose to them its decisions or intentions concerning its own conduct on the market where the object or effect of such contact is to create conditions of competition which do not correspond to the normal conditions of the market. ⁷³

It is the change to the structure of competition, away from the presumptions of independence and rivalry, which makes information exchange so problematic. The exchange would be incompatible with the rules if it “reduces or removes the degree of uncertainty as to the operation of the market.” ⁷⁴ The number of meetings was also irrelevant as long as the exchange allowed them to take account of the information and determine their future conduct. ⁷⁵ The Court went on to establish that such an exchange was “tainted with anticompetitive object” but did little to explain why it was an object restriction. Because of the formulation of the question from the referring court, the CJEU confirmed that there was no need to show a direct link between the conduct and prices but not why such an exchange fulfilled the object test. Although it was not directly related to the object question, the CJEU did make the point that “it must be presumed that the undertakings taking part in the concerted action and remaining active on the market take account of the information exchanged with their competitors in determining their conduct on that market.” ⁷⁶ It is possible that an undertaking could seek to rebut that presumption and therefore show that they, notwithstanding the information exchange, continue to act independently within the bounds of “normal competition.”

Information exchange was also discussed in Icap. ⁷⁷ Icap was one of the competition cases to come out of the LIBOR market manipulation scandals in the 2000s. In this case, Icap was found to be involved in a series of six market manipulation instances from 2007 to 2010. It was seen to have facilitated the manipulation of the JPY LIBOR market through its London “Cash/Money Market desk” that distributed quotes, including volumes and price, to its participants. It produced a spreadsheet including quotes for daily JPY LIBOR rates which was said to have “decisive influence” over the banks. ⁷⁸ The General Court reprised its jurisprudence in T-Mobile and subsequently confirmed by the CJEU in Dole Food. ⁷⁹ The Court stressed that the competition rules protect not only the “immediate interests” of consumers but also “to protect the structure of the market and thus competition as such.” ⁸⁰ The Commission argued that the information exchange in Icap had “enabled the creation of a situation of ‘asymmetrical information’ to the advantage only of the participating banks, allowing them to offer contracts on better terms than other banks.” ⁸¹ The Court did not go quite as far merely confirming that the exchange “was capable of giving an advantage to the banks concerned.” ⁸²

The difficulty the Court faced in connecting the coordination of JPY LIBOR submissions to rates charged on actual derivatives highlights the usefulness of reconceptualizing the harm away from private welfare interests. If we look at the public harm, in the sense of a breach of the legitimate expectations of the normal operation of the market, we can see that the attempt to steal an advantage, though creating an information asymmetry, is in itself a sufficient degree of harm to form the basis of an object restriction whether implemented on the market or not. The Court recognized that the structure of competition, the market as an institution, is based on uncertainly and independent choices driven by rivalry. An attempt to step outside those “rules” of the game to seek an advantage is clearly the type of “cheating” which the law recognizes as a public harm on which society has an interest in addressing.

Cover Bidding as a Form of Information Exchange

The issue in relation to information exchanges is perhaps best highlighted by using a particular form of information exchange as an example. The practice of “cover bidding” is well known in sectors of the economy where public or private tendering is common; for instance, in construction. Cover bidding is a process where a firm wants to be seen to entering a tendering process, usually because they wish to retain their place on the tender “list,” but they do not want to be successful, as they do not have to capacity to deliver the contract. In order to ensure they do not win the contract, they may contact a potential rival to enquire how high their “cover” bid has to be to ensure it loses. These situations can be isolated “one off” communications, or part of larger ongoing bid rigging arrangements, but for the purpose of this article I shall focus only on the one off instances. It is the information exchange within a “cover bid” that makes it a potential Art 101 TFEU issue; a simple high bid without communication would not be problematic. Cover bids have been found to “object” agreements in a number of cases. ⁸³

The anticompetitive effects of cover bidding have been discussed by Stephan and Hvvid. ⁸⁴ They explained the difficulty in demonstrating that a one off cover bidding situation, as set out above, has a significant anticompetitive effect. If there are a large number of potential bidders, the loss of one potential bidder will make little difference to the bid that any other competitive bidder will choose to put forward. This is especially the case if there are a number of meaningful bidders who do not know about the existence of the cover bid; those bidders will still consider the cover bidder as a potential competitor and will price their bid in light of that potential, although nonexistent, competition. The most damaging outcome for competition is where there is a small pool of potential bidders, and all bidders know that there is one, or more, cover bid. They all know there is less competition and may move their bids upward in response. ⁸⁵ Perhaps the most counterintuitive element of this analysis is that the same impact on bid pricing would arise if, rather that entering a cover bid, the potential bidder made a public declaration that they would not be bidding for that tender. The remaining bidders would similarly be able to revise their bids. The only way the procurer would be able to undo that harm would be if they could replace the lost bidder with a meaningful replacement who could re-exert competitive pressure. The harm in this situation comes from the remaining bidders knowing that there is one less competitor—it makes no difference whether they acquire this knowledge privately or publicly.

This leads us to the counterintuitive conclusion that the worst result for competition may be the result of competition law enforcement. If a potential bidder wants to ensure they remain on the list, but cannot contact a single competitor to establish their cover bid, they may seek to declare themselves publicly as not entering the tender; this would have a greater anticompetitive effect. The economics of tendering show us that it is the potential uncertainty of rival bids that drives the competitive process, and the reduction of potential competition, and the uncertainty it brings, resulting from both cover bidding and public exit. ⁸⁶

While the economic effect of both cover bidding and public exit are the same, I would argue that there is a further “public” harm from private information exchange in cover bidding. The private exchange in cover bidding creates exactly the sort of information asymmetry that the Court addressed in Icap. The tendering process is designed to ensure that all bidders are operating on a level playing field and create a competitive environment to the advantage of the procurer. Private information exchanges between bidders create an information asymmetry in which the procurer is disadvantaged—thinking that they have received a significant number of competitive bids—when in reality they have received a lower number. A public exit at least has the benefit of the procurer being able to factor that into their procurement decision. As Bennet and Collins have noted, the concern is not that information is exchanged, but how it is exchanged. ⁸⁷ They suggest that public information—if made available to everyone at no cost—increases transparency and is more likely to benefit consumers. ⁸⁸ There is a risk that public announcements could be part of wide anticompetitive arrangement, but the announcements themselves are only problematic as part of that wider scheme. The public/private divide is useful in relation to cover bidding as if information is public, it is available to all the market participants and removes the concern that the private information exchanges lead to information asymmetries causing public harms in the market.

Cartel Facilitators

The classic example of a cartel facilitation comes from AC Treuhand. ⁸⁹ The Heat Stabilisers cartel was a very traditional cartel in many respects, but the role played by AC Treuhand was distinctive. AC Treuhand was a Swiss consultancy firm and was not active on the cartel market. Their role was organizing meetings, supplying data on sales on the relevant markets, and offering to act as a moderator in the event of tensions between producers. For this role, it was paid a fee. It was the first consultancy firm, providing services to a cartel, to be fined but only the very modest sum of €174,000. AC Treuhand challenged the Commission finding on the basis that it was not foreseeable that mere cartel facilitation, where the consultant had no presence on the cartel market, could be seen as being part of the offending agreement as they had no “concurrent intention to conduct themselves on the market in a particular manner,” ⁹⁰ and no direct link with the restrictions of competition identified by the Commission.” ⁹¹ It also argued that as it did not “relinquish its independence in its commercial conduct,” ⁹² it could not be seen as being in a concerted practice. The CJEU rejected those arguments simply relying on the fact that Art 101 TFEU has no explicit wording to restrict its application to those who are active on the market. It stated that Art 101 TFEU would capture an undertaking who, “intended to contribute by its own conduct to the common objectives pursued,” ⁹³ and that there was no requirement to show “mutual restriction of freedom of action.” ⁹⁴ The CJEU was concerned that a reading of Art 101 TFEU that did not capture facilitation would prohibit the Court from putting an end to active contribution toward the unlawful scheme. ⁹⁵ While the Court made it clear that facilitation could fall within Art 101 TFEU, it did not do much to explain why it did, other than the obvious assertion that it would give better effect to the prohibition if facilitation was prohibited. There was also no consideration as to whether facilitation has the object of restriction competition or merely the effect of doing so.

Cartel facilitation came under scrutiny again in Icap. ⁹⁶ This case was post-Cartes Bancaires, but the question whether facilitation was an object restriction in itself was not specifically addressed. ⁹⁷ The General Court set out the position the Court adopted in relation to facilitation in AC Treuhand and sought to apply it to the complex facts in Icap. Icap argued that it could not be shown that it had sufficient knowledge of the banks’ intention to coordinate their JPY LIBOR panel submissions. They also argued there were various mechanisms through which banks could become aware of others’ submissions. The Commission argued that it had proved that Icap was aware, or should have been aware, that its actions contributed toward the competition law infringements. The evidence relied on by the Commission was significantly less damming than that found in AC Treuhand. In essence, it was one or two conversations that indicated that the banks were aware of each other’s positions; there were no meetings attended or offers of assistance made. In relation to some of the infringements, for instance, the UBS/RBS 2008 infringement, the evidence of knowledge had potential other explanations or interpretations and was not enough to support the infringement finding. ⁹⁸ The Court stressed that a facilitator must have “intended to contribute” to the common objective of the participants, it was aware of the conduct planned, or it could have foreseen and was “prepared to take the risk.” ⁹⁹ It also stressed that “passive modes of participation” are “indicative of collusion” unless the party engages in some “public distancing.” ¹⁰⁰ The General Court then turned to look at the nature of a restriction the connection to a facilitator even though they are not active on the market. It made clear that there was no need to show a mutual restriction between parties ¹⁰¹ as long as the conduct of one party was affected by the arrangement. ¹⁰² So far, this is not necessarily troubling, although it does suggest a very low bar for the application of Art 101 TFEU. But what follows is more difficult to reconcile with what has been set out above. Going back to AC Treuhand, it stated that to ensure full effectiveness, an “active contribution” would be caught even if that contribution “does not relate to an economic activity forming part of the relevant market.” ¹⁰³ Given that passive contributions were sufficient, in the sense that they bolstered arrangements, is it significant that a facilitator must be “active” where their contribution does not relate to economic activity on the market? Is passive behavior, which can be seen as encouragement of the market active participants, enough to form the basis of an infringement decision? I suggest that this can be read as being understood that a market participant being passive is encouragement of and, therefore, indirect participation in, the infringement, but that a nonparticipant needs to be active before it can be seen as infringing. Going back to AC Treuhand, one can see that their offer of assistance was clearly an active position to bolster the common scheme rather than merely being passive. The participation of Icap in the arrangement in relation to JPY LIBOR was found to be a “complement” to the activities of the banks and that they had therefore actively participated by increasing the probably of the arrangements success. ¹⁰⁴

It is in relation to this potential distinction between active and passive participation in the common scheme that it is useful to consider the public harm of a cartel. Little thought appears to have been given to how facilitation is an object restriction. Most of the discussion seems to treat facilitation as being a form of inchoate liability parasitic on the activities of the participants who are market participants. I take that from the instrumental justification that not capturing facilitation would “negate the full effectiveness of the prohibition.” ¹⁰⁵ I find the lack of an explicit rationale troubling. I would rather a more direct explanation of how the facilitator themselves, through their own actions, either assumes or does not assume, liability for their own particular role within the common scheme.

By looking through the lens of the public harm associated with cartels, we may be able to draw some distinctions between certain types of conduct, when freed from trying to draw unhelpful connections based on the implementation of an arrangement on the market itself. If we consider the public harm of the cartel—the wider harm to the free market as a public institution—we can see why active participation, as in AC Treuhand, justifiably has the object of restricting competition. The very heart of their contribution to the arrangement is harm to the expectation of a free market. They may not be able to themselves change prices or harm the welfare interests of consumers of competitors, but they are clearly contributing to the limitation of competition. This is perhaps not a very controversial assertion, but I argue that conceptualizing the harm in this way makes it much clearer that the basis of the “sufficient harm” is inherent within their conduct alone. It also makes it clear that the harm is manifest as soon as their contribution is made and that it is independent of any implementation upon the market by market participants. Theirs is not an inchoate liability deriving from the actions of others—they play a distinct and separate part in the common unlawful scheme which carries liability. It may, however, still be necessary to draw on the wider cartel scheme to justify the public wrong of a facilitator in one regard. If we look back to Green’s conception of the “cheat,” one of its key features is that the cheat seeks to take an advantage by stepping outside the norms of market behavior. What is the advantage that the facilitator seeks? It is clear that the cartel gains an advantage by having the services of the facilitator. ¹⁰⁶ It is much less clear what form of advantage the facilitator gains from its behavior. We shall return to the question of advantage shortly.

Active participation is I think therefore potentially clear, but passive facilitation is more problematic to bring within the rationale of a form of public harm. If a nonparticipant in the market is aware of cartel conduct and does not publicly distance themselves from it can they be liable for the common scheme simply because they are seen to “encourage” the behavior, giving confidence and succor to the participants who are aware of their passivity? I would suggest not. I think the distinction is best explained by drawing a distinction between the passivity of a market participant and the passivity of a facilitator. I think it is correct that a market participant who is aware of the cartel activity, but is not an active participant in meetings, can still be seen as a beneficiary of the common scheme. They are essentially the recipient of an information exchange, as they are aware of the cartel scheme, and will be able to adapt their market behavior in response to that information. In the terms of public harm, they are aware that the market is no longer operating as a free market based on competition, and they take advantage of that information as the normal uncertainty of the market is reduced. The only way they can protect themselves from liability is through the positive act of public distancing. It is difficult to see how the facilitator has taken an advantage from their passive awareness of the cartel scheme. If we go back to the advantage gained through cheating, it is difficult to see how, even though the facilitator may be aware that the cartel are operating outside the normal rules of the market, they take advantage of that knowledge. The participants can take advantage on the market, but a facilitator’s advantage must come from elsewhere.

If it is difficult to see where an entirely passive market nonparticipant can be found to be a facilitator, the important question becomes what is “active.” In both AC Treuhand and Icap, the Court found that the facilitator, in one form or another, made a contribution, or least acted in complement, to the participants’ scheme, by offering mediation or publishing information. I can however see that lower levels of activity might also be categorized as being active. In their complaint, Icap advanced an argument that their behavior was merely trying to satisfy the wishes of their customers, ¹⁰⁷ it did not wish to contribute to the scheme. This is probably the cartel facilitator’s equivalent of a “following orders” defense and was rightly rejected. It is, however, perhaps useful to think of the cartelists as the facilitator’s customers or clients—in that sense what advantage does the facilitator get from the cartel? Here I would suggest that the facilitator is keeping their clients happy in order to keep their custom. The facilitator, by providing a service to the cartel, is likely to be asked to continue in their role and continue to receive their remuneration. I would argue that this in itself is sufficient advantage for the facilitator to be seen as a cheat and therefore actively contributing to the wider public harm of a cartel.

Conclusions

There are three things that I argue we can take from reconceptualizing the “sufficient harm” required to indicate the existence of an object restriction in the terms of the public harms of cartels. While the genesis of much of this debate has come from the criminal law, the cartel phenomena are essentially the same for all, and there is no reason that Art 101 TFEU cannot learn from an improved understanding of cartels.

Firstly, much of the debate about the nature of object restrictions has relied on some form of procedural efficiency argument; that it is not necessary to examine the effects of agreements when harm is highly likely to stem from them. I would argue that if one looks to the public harms of cartels, we can see that a significant public harm arises from cartels as soon as the cartel comes to an arrangement. There is no need to rely on any presumption in that regard. Implementation may cause another form of harm, in the sense of economic welfare effects, but a serious public harm has already occurred. By stepping outside the legitimate expectation of a free market to gain an advantage, the cartelists have caused a significant harm to the market as an institution. That in itself is grounds for finding the existence of an infringement. It may be important for questions regarding trade between Member States and the size of a fine to look beyond this public harm but not simply to establish the existence of an object restriction.

Before public harm analysis is applied to information exchanges, we can see that connecting a simple exchange of information with eventual welfare effects on the market is potentially difficult. It is clear from the case law that the Court was convinced that the reduction in uncertainly that followed an exchange was problematic, but little attention had been given the why it could be said to lead to sufficient harm. The protection of the market as a public institution gives a clear answer. The creation of information asymmetry to give advantage to the cartelists, and which is not shared with other market participants, is a clear form public harm. This gives good justification to treating such exchanges as object restrictions. This tallies with many of the issues that the Court has raised in its case law—such as protection of the “normal conditions of competition”—and clarifies when and how the harm comes about. It also, as indicated above, removes the need to rely on presumptions of future conduct to found the infringement.

In the analysis of cartel facilitation, we saw a similar clarification. Focusing on the public harm explains the nature of the active participation required by the facilitator. That active participation is separate and distinct from the actions of the market participants in the cartel. The facilitator’s liability is not simply an inchoate liability from the main cartel arrangement. Their responsibility for the infringement can be seen as being a distinct form of public harm. Through their active facilitation, they may not influence the market directly, but they do take active steps to gain an advantage for the cartel in relation to other market participants. Their actions therefore can be seen as direct harm to the market as a public institution. It is arguable that an entirely passive facilitator cannot be seen to have gained an advantage from the cartel and may escape liability. It may, however, still be possible to argue that providing services to clients are known to be acting outside the norms of the market, and taking a fee in return may be enough to indicate that you are taking an advantage from the cartel stepping outside the “normal rules” of competition.