Collective Dominance

A preliminary step before defining a collective dominant position is to define what a dominant position is. This is because, to determine that there is a collective dominant position, it is not only necessary to show that there are several economic agents acting together, but also to show that this collective entity is dominant. The test to demonstrate this collective dominance is the same as that used to determine whether there is individual dominance (Nazzini 2011, 367). On this point, three main conceptualisations have been offered (for more details, see CeCo’s Glossary on Dominant Position).

First, dominance is understood as substantial market power (see CeCo’s Glossary) (that is, the power to increase prices and reduce output). Second, it is understood as the “commercial superiority” in which a firm finds itself vis-à-vis its competitors (which allows the economic agent to exclude its rivals from the market or, at least, to restrict their expansion, distorting the competitive process). Finally, it is defined as a mere jurisdictional criterion. Here it is argued that the concept of dominance would be a sort of threshold, comparable to the role played by market share when block exemptions are at stake. Thus, it would be assumed that, below a certain threshold, such as market share, the prohibition of abuse of a dominant position does not apply.

i. Formal clarifications

As to its legal enshrinement, this concept is present in the regulation of both Chile and the European Union (among other jurisdictions). In Chile, Article 3 of Decree Law 211 (DL 211), both in its first paragraph and in its second paragraph letter b), recognises the concept of abuse of a collective dominant position. In this regard, Article 3, paragraph 1, which sets out the general offence, states that “anyone who performs, individually or collectively, any deed, act or agreement which prevents, restricts or hinders competition, or which tends to produce such effects” shall be sanctioned. Subsequently, paragraph 2 letter b) of the same article sanctions “the abusive exploitation by an economic agent, or a group of them, of a dominant position in the market (…)”.

For its part, in the case of the European jurisdiction, Article 102 of the Treaty on the Functioning of the European Union provides that: “Any abuse by one or more undertakings of a dominant position within the internal market or in a substantial part of it shall be prohibited as incompatible with the internal market in so far as it may affect trade between Member States (…)”.

It is worth highlighting that the regulations which sanction abuse of a collective dominant position do so separately from agreements or concerted practices. Therefore, it is necessary to distinguish bilateral or multilateral conduct, such as agreements and concerted practices (which require communication between the parties), from unilateral conduct, such as abuse of an individual or collective dominant position (which does not require communication between the parties). Thus, the statement that a conduct is unilateral is antithetical to the statement that it is collusive: both descriptions cannot be true at the same time (Black 2008, 121). It is therefore necessary to clearly distinguish abuse of a collective dominant position from the various anticompetitive agreements, since if there were any kind of agreement, whether express or tacit, the situation would fall within the ambit of the rules sanctioning anticompetitive agreements (Martínez and Folguera 2006, 77).

According to Oliver Black, collective dominance would be a “joint coordinated action, that is done knowingly and without communication”. This situation occurs when the following conditions are met:

• “(1) X does Ax;
• (2) Y does Ay;
• (3) X believes that Y does Ay;
• (4) Y believes that X does Ax.
• (5) X, in doing Ax, relies on Y doing Ay;
• (6) Y, in doing Ay, relies on X doing Ax”;
• (7) X, in relying on Y doing Ay, has the objective Ox;
• (8) Y, in relying on Y doing Ay, has the objective Oy;
• (9) Ox = Oy;
• (10) X knows (5)-(9);
• (11) Y knows (5)-(9)”. (Black 2005a: 185-187).

By contrast, joint coordinated action (i.e. non-unilateral) would have the following two additional conditions (which include communication):

• “(12): (10) is true in part because Y communicates (6) and (8) to X;
• (13): (11) is true in part because X communicates (5) and (7) to Y” (Black 2005a, 187).

It should be noted that, unlike the European Union, the concept of collective dominance is not expressly regulated in the United States. Consequently, resorting to the doctrine and case law of that country is problematic. Thus, as stated by Justice Ricardo Paredes in his dissenting opinion in Judgment No. 176/2021:

“(…) the idea that a dominant position cannot be associated with more than one undertaking or economic agent became entrenched in the literature mainly on the basis of the two first sections of the Sherman Act of the US; section 1, which prohibits agreements that restrict competition between undertakings, and section 2, which prohibits monopolising or attempting to monopolise a market” (§ 3).

Indeed, the wording of Section 1 (agreements) and Section 2 (monopolisation) of the Sherman Act distinguishes the number of infringing parties. By definition, only one undertaking can monopolise a market, while the prohibition of agreements between undertakings requires the presence of the collusive element (whether express or tacit). In these terms, a collective dominant position could not, in principle, be understood as falling under either of these two provisions of US law (Folguera and Martínez 2006, 79-80).

ii. The evolution of the concept in European doctrine and case law

Since in cases of abuse of a collective dominant position there is no exchange of information between the parties (otherwise we would approach an agreement or concerted practice), the question arises as to what conditions make collective dominance possible. This challenge has been addressed mainly by European case law, which has evolved through different stages.

Indeed, two stages can be distinguished in this case law: some distinguish between a stage of “restrictive interpretation” and one of “expansive interpretation” (Folguera and Martínez 2006, 85), while others distinguish between a “structural” stage and a “behavioural” one (Petit 2011, 1). Broadly speaking, under the restrictive or structural interpretation, the concept applied only to legally distinct entities that, by virtue of a series of links (e.g. structural, contractual or commercial), belonged to the same corporate group. Subsequently, under the expansive or behavioural interpretation, oligopolies were also included (Folguera and Martínez 2006, 85), which implied that the finding of collective dominance came to depend on the ability of various firms to coordinate their behaviour (Petit 2011, 1).

The existence of different stages explains why, throughout their decisions, the European authorities covered both situations in which two or more economic agents were connected in such a way that they adopted the same conduct or acted as a single economic agent due to structural or commercial links, and later referred to situations of tacit oligopolistic collusion (Nazzini 2011, 359-360).

a) The first stage of case law: restrictive thesis

The first stage included the cases Flat Glass (1992 § 2, 358), Almelo (1994 § 43-43), DIP (1995 § 26), and Irish Sugar (2001, § 46). According to this line of case law, to assert the existence of a collective dominant position, it is necessary to ascertain the presence of links that lead to the adoption of the same conduct in the market or a common policy.

Structural links could consist of interlocking directorates, cross-shareholdings, and participation in joint ventures or trade associations. Commercial links, in turn, could be cross-licensing of technology or reciprocal exclusive distribution agreements (Nazzini 2011, 364).

The case law saga can be explained as follows (this is based on the paragraphs already cited from the judgments mentioned above). The Flat Glass case introduced the requirement of commercial links for a collective dominant position to exist. Almelo, for its part, relaxed the test and stated that there had to be links of some kind between the firms. That is, the links did not have to be only commercial. In turn, in DIP it was stated that, to verify collective dominance, the agents had to adopt the same conduct in the market. Irish Sugar, for its part, emphasized that these links had to create a collective entity that could act independently of customers and other competitors.

It is worth highlighting that this line of case law does not carry as much weight today. This is because, nowadays, the key to collective dominance is considered to lie in oligopolistic interdependence between firms (Folguera and Martínez 2006, 76; Ibáñez Colomo 2018, 265), a concept that is not emphasised in this first stage of European case law. Thus, Petit states that those who still support it “are stuck in time” (Petit 2011, 4).

That said, other authors, such as Nazzini, draw from this case law two types of hypotheses under which a non-oligopolistic case of collective dominance may be configured. The first hypothesis is that there may be horizontal non-oligopolistic collective dominance. This would arise when two or more economic agents act as a collective entity in the same market and have the capacity to harm competition due to the structural or commercial links that bind them, or due to direct or indirect contacts between them (Nazzini 2011, 364). In this first hypothesis, a two-step test should be applied: (i) determining whether there are two or more economic agents acting as a collective entity, and (ii) determining whether they are collectively dominant (Nazzini 2011, 364).

On the other hand, Nazzini’s second hypothesis is that there may be vertical non-oligopolistic collective dominance. This may arise if two economic agents at different levels of the supply or production chain act as a collective entity. Here, the test is also two-fold: (i) determining whether there are structural or commercial links between two or more economic agents that give them the ability to act as a single collective entity, and (ii) determining whether such entity is dominant (Nazzini 2011, 369). This is explained by the fact that it is possible for two economic agents in a vertical relationship not to be so closely involved with each other as to be a single economic agent, but still to have the ability and incentive to act as a single entity (Nazzini 2011, 369).

b) The second stage of case law: expansive thesis

The line of case law just described was gradually left, starting with Gencor (1999) (Folguera and Martínez 2006, 86; Nazzini 2011, 362). In this case, it was stated, in a paragraph that deserves to be quoted at length, that:

“there is no reason whatsoever in legal or economic terms to exclude from the notion of economic links the relationship of interdependence existing between the parties to a tight oligopoly within which, in a market with the appropriate characteristics, in particular in terms of market concentration, transparency and product homogeneity, those parties are in a position to anticipate one another’s behaviour and are therefore strongly encouraged to align their conduct in the market, in particular in such a way as to maximise their joint profits by restricting production with a view to increasing prices. In such a context, each trader is aware that highly competitive action on its part designed to increase its market share (for example a price cut) would provoke identical action by the others, so that it would derive no benefit from its initiative” (§ 276).

This judgment provided an initial glimpse of the emphasis that was to be placed on oligopolistic market dynamics in future case law on the concept. Thus, following Gencor, the capacity of two or more agents to act as a collective agent came to depend not only on commercial or structural links, but also on the ability of firms in an oligopolistic market to coordinate their behaviour tacitly (Nazzini 2011, 364).

This line was consolidated over time. First, in the Compagnie Maritime Belge case, where the European Court of Justice held that the finding of a collective dominant position would depend on an “economic assessment”, and in particular on an assessment of the market structure in question (2000, § 45).

Subsequently, this test was fine-tuned in Airtours (2002 § 3-6), where the Court of First Instance of the EU stated that a collective dominant position may exist when

“each member of the dominant oligopoly, as it becomes aware of common interests, consider it possible, economically rational, and hence preferable, to adopt on a lasting basis a common policy on the market with the aim of selling at above competitive prices, without having to enter into an agreement or resort to a concerted practice within the meaning of Article 81 EC and without any actual or potential competitors, let alone customers or consumers, being able to react effectively” (§ 3)

The Court formalised the test through certain conditions (§ 4) which, given the importance they have taken on in later cases, will be detailed below. First, all members of the dominant oligopoly must be able to know the behaviour of the other members to verify whether they are adopting the same line of action or strategy. On this point, some commentators have noted that the structural conditions for firms to reach an understanding on the terms of coordination are transparency, stable and inelastic demand, and inelastic supply (Nazzini 2011, 317; in more depth: Nazzini 2011, 375-382).

Second, the situation of tacit coordination must be sustainable over time, that is, there must be an incentive not to deviate from the common line of conduct. Indeed, only if all members of the dominant oligopoly maintain parallel behaviour can they benefit from it. This requirement includes the existence of retaliatory measures if one of the agents departs from the common line of action. According to Nazzini, for this to be possible, economic agents must have a direct interest in the profits of the other, either through cross-shareholdings or through sufficiently symmetric demand and cost functions (Nazzini 2011, 372; and in more depth: 382-383).

Third, it is necessary to establish that the foreseeable reaction of competitors — current and potential — and of consumers would not call into question the expected results of the common line of action.

With Airtours, case law stabilised, so subsequent judgments have built on the test established therein. To that extent, in Piau (2005, § 111), the conditions established in Airtours were confirmed and later, with Impala (2008), the way in which the Airtours test was to be applied was clarified. In this case, the European Court of Justice emphasized that, to analyze whether a scenario of collective dominance exists, it is necessary to avoid a mechanical application of the test that would consider each of the Airtours factors in isolation (§ 125). It was therefore established that, to verify the existence of oligopolistic collective dominance, it is necessary to carry out a systematic and integrated analysis of the market (Nazzini 2011, 375).

The impact of Airtours is not limited to subsequent cases, but has also affected merger operations. In fact, the Commission’s Horizontal Merger Guidelines took up the parameters set out in Airtours and even reproduced parts of them verbatim (Folguera and Martínez 2006, 89).

It’s important to recall what was stated earlier: one must distinguish a dominant position from the abuse of such position. This is because, although the former may be undesirable, it is lawful, whereas the latter is undesirable and unlawful. Therefore, meeting the requirements of a collective dominant position is a necessary but not sufficient condition for there to be an abuse of a dominant position.

As a preliminary note, it is worth pointing out that, as the Tribunal de Defensa de la Libre Competencia (TDLC) itself stated in Judgment No. 189-2023, “the concept of abuse of a collective dominant position has not been the subject of specific consideration in national case law, that has served as the basis for an infringement action based on Article 3, paragraph two, letter b), or letter c) of DL No. 211” (C° 165).

Notwithstanding the foregoing, the concept of collective dominance was indeed discussed in the “Conadecus v. Telefónica and others” case, the “Uber-Taxi” case, and the “Cryptocurrency Intermediaries v. Banks” case. In the first of these cases, Conadecus, a consumer association, filed a complaint against Telefónica Móviles Chile S.A. (Movistar), Claro Chile S.A. (Claro) and Entel PCS Telecomunicaciones S.A. (Entel), various telecommunications companies. It claimed that they had infringed Article 3 of DL 211 (it only referred to this provision), by bidding in the “Public tender for granting Public Service Concessions for Data Transmission in the frequency bands 713-748 MHz and 768-803 MHz”, exceeding the spectrum caps that an operator competing in the mobile telecommunications market in Chile may lawfully hold. It is worth noting here that Conadecus did not accuse the companies of having committed an abuse of a collective dominant position, but only accused them of having carried out a conduct that prevents, restricts or hinders competition or tends to produce such effects (in the wording of Article 3 of DL 211).

Conadecus argued that various authorities, including the Supreme Court, had previously established a maximum limit (cap) of 60 MHz on the total amount of spectrum that each mobile telephony operator may hold. It added that, even if the calls for tender did not refer to the cap, the defendants should have taken it into account. This is because those who hold a dominant position would bear a special duty of care, which would require them to ensure that their conduct does not harm competition.

Faced with this, in Judgment No. 154/2016, the TDLC decided to dismiss the complaint. Among other reasons, it pointed out that the defendants, considered individually, did not hold a dominant position (C° 46). This was because: (i) although each of the defendants held a significant proportion of the tendered spectrum, none exceeded 30%; (ii) there were other companies that could offer facilities in the wholesale market for access to mobile networks; and (iii) although the defendants had high market shares in the retail market for analogue and digital mobile telecommunications services, that percentage was not sufficient in itself to define them as dominant (C° 36).

Following a motion for review, the Supreme Court reversed the TDLC’s judgment and upheld Conadecus’ claim. To reach this decision, among other aspects, the Court considered that there was a collective dominant position. This was because:

“there are several factors, both structural and behavioural, that determine that a firm finds itself in a dominant position, among which the existence of barriers to entry (…) should be highlighted. Indeed, the three defendants jointly control 83% of the market, a circumstance that determines the existence of an oligopolistic market, in which the existence of market power must be analyzed not only by isolating the individual market share figures, but also by examining the joint market power of the three firms accused of carrying out anticompetitive conduct, since a dominant position may not only belong to a single firm, but may also be held by two or more competing firms which may, potentially, coordinate their commercial policies without there being an express agreement to that effect. The high degree of concentration in the industry, coupled with the existence of barriers to entry in the wholesale market, clearly reveals the existence of risks to competition” (C° 13) (emphasis added).

Thus, for the Supreme Court, it would seem sufficient that there be an oligopolistic market together with barriers to entry for a collective abuse of a dominant position to be configured, disregarding other factors highlighted by European doctrine and case law, such as market transparency or the incentives to adopt the same conduct (Budnik 2019, 130).

Another relevant case was the one colloquially known as “Uber-Taxi”. In this case, Sindicato Chile Taxis accused the defendants Cabify, Uber Chile and Easy Taxi of conduct constituting: (i) unfair competition, due to the infringement of certain tax and labour provisions and for taking advantage of the taxis’ reputation to divert customers; (ii) predatory pricing; and (iii) abuse of a dominant position through price-fixing.

In Judgment No. 176-2021, the TDLC dismissed the complaint because, in its view, the claimants had failed to provide evidence showing that the defendants have a dominant position or that they have a reasonable expectation of achieving one.

What is interesting for the purposes of this glossary is the divergence between the judgment and the concurring opinion of Justice Paredes. This concurring opinion was prompted by the fact that the TDLC stated that “by definition only one firm can have a dominant position in a market” (C° 76). In response to this statement, Justice Paredes, in his concurring opinion, emphasized that:

“economic theory, since the work of Chamberlin, E. (…), has had a decisive influence, changing the analysis of which agents can abuse their dominance and showing that there can indeed be dominance by more than one firm, without this requiring collusion or parallelism” (Concurring opinion, points 4 to 5) (emphasis added).

Thus, whereas the TDLC adopted a narrow concept of collective dominance, Paredes adopted a broad definition of that concept.

Subsequently, this case reached the Supreme Court. There, the Court, with one Justice dissenting, decided to dismiss the motion for review and uphold the TDLC’s judgment.

Finally, the “Cryptocurrency Intermediaries v. Banks” case also addressed this issue. By way of context, it is worth noting that since 2015 cryptocurrency companies began opening accounts with some banks. However, in 2018 these banks began sending letters announcing the (unilateral) closure of these current accounts, citing regulatory reasons. In reaction, the exchanges attempted to open accounts with other banks, none of which agreed to do so.

Therefore, on 20 April 2018, Buda filed a complaint alleging an infringement of Article 3 of DL No. 211, for conduct consisting of an exclusionary abuse of a collective dominant position through the closure of accounts (against Banco Itaú and Banco Estado), and for refusal to open bank accounts (against Banco Bice, Scotiabank, Banco de Chile, BBVA, Santander, Banco Internacional, BCI and Banco Security) (case file C 349-2018). For its part, on 1 June 2018, Orionx filed a complaint against Banco Estado, Banco de Chile, Banco Bice, Itaú, Santander and Scotiabank, alleging conduct consisting of an abuse of a collective dominant position, with the aim of eliminating or reducing potential competition in the relevant market for digital means of payment (case file No. C 354-2018). Subsequently, on 23 April 2018, CRYPTOMKT filed a complaint for infringement of Article 3 of DL No. 211, on similar grounds (case file No. C 350-2018).

The complaints by CRYPTOMKT and Orionx were joined to the proceedings initiated by Buda (case file C 349-2018).

In general, the defendant banks argued that they had acted solely in compliance with sectoral regulation. Owing to this, they had to close the accounts, since there were risks of money laundering through cryptocurrency intermediation. This was because it was difficult or almost impossible to track the transactions carried out through them.

Lastly, it should be mentioned that on 19 August 2022, BCI, Buda and CryptoMKT filed a settlement agreement with the TDLC. In it, BCI stated that it had refused to open current accounts for the claimant based on compliance with the rules on the prevention of offences such as money laundering and terrorist financing (“ML/TF”). On 29 August 2022, the settlement was approved. Subsequently, on 12 September 2023, Banco Bice and Buda filed a settlement agreement on similar terms, which was approved by the TDLC on 12 October 2023. Finally, on 21 December 2023, the TDLC handed down a final acquittal judgment, dismissing the claims brought by the cryptocurrency firms.

Although in Judgment No. 189-2023 the TDLC dismissed the claim, it did address the concept of collective dominance. On this point, unlike in the Uber/Taxi case, it did acknowledge that “abuse of a dominant position could be exercised not only by one, but by two or more distinct and independent economic agents” (C° 161).

Having said that, it went on to analyze the requirements to establish an abuse of a dominant position. Among these, it mentioned those set out in Airtours (C° 167, 170), which suggests that the Tribunal has adopted the European line of reasoning on this matter. However, it is worth noting that the TDLC ultimately adopts a somewhat eclectic stance. This is because it also refers to “the requirement that there be ‘economic links’ between competitors in order to establish a collective dominant position” (C° 168).

In short, the TDLC currently appears to adopt an intermediate position between what was considered in the first phase of European case law and the second phase.

Given its importance, the Airtours case will be covered (its doctrinal elements have already been discussed). It involved two phases, an administrative one and a judicial one, and concerned a merger between two British companies, Airtours plc (Airtours) and First Choice Holidays plc (First Choice), two tour operators and providers of package holidays.

In the administrative phase, the European Commission decided to prohibit the transaction. It based its decision (2000) on the fact that the transaction would create a collective dominant position in the UK market for foreign package holidays to nearby destinations, as a result of which effective competition would be significantly impeded in the common market. This was because it would increase market concentration (§ 192-194).

Faced with this, in the judicial phase, the Court of First Instance (2002) annulled the Commission’s decision. In this regard, a relevant aspect was that it considered that, in relation to the forward-looking analysis of the market that must be carried out to assess an alleged collective dominant position, such position cannot be considered solely from a static point of view at a particular point in time, but must also be analyzed dynamically (§ 6, 192). Thus, according to this decision, it is not enough for there to be market concentration; it is also necessary to analyze the specific market dynamics. Therefore, several conditions must be met (see above, analysis of Airtours) that go beyond market concentration (§ 62).

One of the most noteworthy aspects of this judgment, then, was that it clarified that the fact that several agents have a high market share within a concentrated market is not sufficient for a finding of collective dominance to be made immediately (Nazzini 2011, 367).