Inductive reasoning is sometimes explained by using either the ‘duck test’ (“if it looks like a duck, swims like a duck and quacks like a duck, then it probably is a duck”) or the ‘elephant test’ (“it is difficult to describe, but you know it when you see it”).
In EU competition law, it sometimes feels that identifying a restriction of competition ‘by object’ is much the same. Using experience and intuition, we (and particularly competition authorities) believe we can tell when an agreement does restrict competition without actually assessing its effects on competition. This does, of course, lead to a risk of a perceived degree of arbitrary decision-making, other than in the case of hardcore cartels, which are now broadly accepted as being anti-competitive.
In its recent judgment, of 11 September 2014, in Case C-67/13P Cartes Bancaires v European Commission, the Court of Justice of the European Union (“CJEU”) had to consider again the difficult question of when an agreement has an anti-competitive object and is therefore prohibited by Article 101(1) TFEU.
In its judgment, the CJEU has provided further clarity on when an agreement has the object of restricting competition. In particular, a ‘restriction by object’ may be found only after it is shown that the agreement, by its wording, objectives and context, displays as sufficient degree of harm to competition so that, by its very nature, they are intended to change, appreciably, the structure of the market. The CJEU also confirms that, where an agreement concerns a two-sided market (such as payment systems) or other related markets, its ability to harm competition must be demonstrated on all such markets.
Whilst the CJEU’s guidance should restrict competition authorities’ increasing tendency to take a ‘short cut’ and find that an agreement or practice restricts competition ‘by object’ (which obviates the often difficult assessment of its effects on competition), there remain ambiguities as to exactly when an agreement restricts competition ‘by object’.
Having annulled the GCEU’s upholding of the Commission’s decision that CB’s new rules restricted competition by object, the CJEU referred the case back to the CGEU for a consideration of whether the rules had the effect of restricting competition.
Cartes Bancaires: background
Cartes Bancaires is one of several cases concerning the application of competition law to card payment systems. Others include Case C-382/12P Mastercard (in which an identically constituted Chamber of the CJEU upheld, also on 11 September 2014, the Commission’s decision that Mastercard’s multilateral interchange fee infringed Article 101(1) TFEU and was not exempted under Article 101(3)) and Visa Europe (in which Visa gave commitments to resolve the Commission’s investigation into its MIFs).
Cartes Bancaires (“CB”) is the national bank card payment system in France, to which French banks are members. In 2007, the Commission found that certain of CB’s new rules infringed Article 101(1).
The Commission objected to new CB rules that imposed: (i) a levy on banks that had significantly greater card issuing activities than merchant acquiring activities; (ii) a membership fee per card issued; (iii) supplementary membership fees on banks that increased significantly the number of cards issued by them; and (iv) further fees on banks that had hitherto been inactive but which subsequently increased significantly the number of cards issued by them. CB had argued that these rules were required to prevent ‘free riding’ by banks that issued cards but did not have any or only limited merchant acquiring activities. However, the Commission considered that these rules had both the object and effect of restricting competition by increasing the costs of and thereby impeding new entry and expansion by card issuers, so limiting price competition between card issuers, to the benefit of CB’s largest members. It refused to grant an exemption under Article 101(3).
In 2012, the General Court (“GCEU”) dismissed appeals by CB and three of its member banks. CB and the banks made a further appeal to the CJEU. The CJEU’s judgment focuses on the GCEU’s finding that the CB rules restricted competition by object.
Cartes Bancaires: restriction of competition by object – applicable legal principles
Article 101(1) prohibits agreements, decisions and concerted practices that have the object or effect of restricting, preventing or distorting competition. Where an agreement has an anti-competitive object, it is not necessary for the Commission to demonstrate that it has in fact had anti-competitive effects.
In Cartes Bancaires, the CJEU reconfirmed when it is permissible for the Commission to find than agreement restricts competition ‘by object’. The same principles apply to the GCEU when hearing appeals against decisions of the Commission, since it must undertake a full review of the Commission’s application of Article 101(1) (see in this regard Case C-386/10 Chalkor).
According to the CJEU, reflecting prior case law (in particular Case C-209/07 Beef Industry Development Society and Case C-32/11 Allianz Hungária), a finding that an agreement (or decision or concerted practice) restricts competition ‘by object’ can be made where it has “a sufficient degree of harm to competition” (see para. 49) such that it can be regarded “by [its] very nature as being harmful to the proper functioning of normal competition” (see para. 50). This will be the case where it has the object of “changing appreciably the structure of the market”, for example – as in BIDS – by causing competitors to leave the market and/or by reducing over-capacity (see paras. 84 and 85).
Cartes Bancaires: the new rules did not constitute a restriction by object
The CJEU found that the GCEU had not applied the correct legal principles in upholding the Commission’s finding that CB’s new rules constituted a restriction of competition ‘by object’. It had failed to ascertain whether the rules had, by their very nature, a sufficient degree of harm to competition and had, instead, applied a lower threshold of whether the rules were “capable… of preventing, restricting or distorting competition” (see para. 57). The GCEU had also incorrectly held that the concept of an infringement by object should not be given a restrictive interpretation (see para. 58).
The CJEU then considered whether the GCEU’s errors invalidated its assessment that the new CB rules constituted a ‘by object’ restriction of competition. It found that the GCEU did not explain how the wording of the rules led to a restriction of competition (see para. 65). It merely inferred that the rules impeded competition by new entrants on the French market for issuing payment cards (see para. 68) so as to sufficiently harm competition and thus constitute a restriction by object (see paras. 69 and 85). Accordingly, the GCEU had failed to establish that the rules had a restrictive object (see paras. 70 and 71) by appreciably changing the structure of the market (see para. 85).
According to the CJEU, the new CB rules at most imposed on certain issuing banks a choice between either limiting the number of cards issued by them or paying a contribution to reflect the benefit to them of the activities of acquiring banks: this was not by its very nature harmful to the functioning of normal competition (see paras. 75). The rules could, of course, still have anti-competitive effects by hindering competition from new entrants, by hindering their expansion or even excluding them from the CB system, but these were not evidence of an anti-competitive object (see paras. 80 and 81).
Cartes Bancaires: restriction by object in two-sided markets and where a legitimate objective is claimed
The CJEU then went on to find a second error in the GCEU’s legal analysis.
Payments systems are two-sided markets, covering both the card issuing and merchant acquiring markets, with interactions between them leading to network effects. CB maintained that the rules were intended to encourage the development of members’ acquisition activities and ensure an ‘optimal balance’ between issuing and acquiring activities, thereby preventing some issuing banks ‘free-riding’ on acquiring banks.
In its judgment, the CJEU confirmed that when coordination concerns more than one market, a restriction by object may only be found after an assessment of whether that coordination is by its nature harmful to competition on all markets to which it relates. In the case of two-sided markets, or vertically-related markets, this must include an assessment of interactions between the markets (see paras. 75 to 79). As payments systems are two-sided markets, the GCEU could not conclude that CB’s rules were by their very nature harmful to competition without assessing their impact on all markets, which it had not done (see para. 74).
The CJEU then went further: where it is accepted – as the GCEU did – that a rule (or presumably other form of coordination) has a legitimate objective (so as to – at least potentially – result in benefits on one market or for an overall system), it cannot be regarded by its very nature as harming competition so as to constitute an object restriction (see para. 75).
Analysis: when does an agreement or practice ‘restrict competition by object’?
It is welcome that the CJEU has cautioned against an overly-expansive application of the concept of ‘restrictions by object’: both the Commission and national competition authorities have sought to expand this concept to adopt infringement decisions without having to assess the actual effects of an agreement on competition. In addition, in Cartes Bancaires, the CJEU appears to have given a narrower interpretation of ‘restriction by object’ than in earlier judgments: in Case C-8/08 T-Mobile Netherlands, the CJEU had held that, for a concerted practice to be regarded as having an anti-competitive object, it was sufficient that it has “the potential” to have or “simply be capable” of having a negative impact on competition (see para. 31 of the T-Mobile judgment and also para. 38 of the Allianz Hungária judgment). In Cartes Bancaires, the CJEU held that the GCEU had incorrectly applied the “potential” test and should have applied a narrower test of being “so likely” to have negative effects on or a sufficient degree of harm to competition.
The CJEU has made clear that this ‘decisional short-cut’ may be taken only where the agreement or practice, in its proper context, will inevitably have significant negative effects on competition; however, even in this case, a mere assertion of negative effects will not suffice and a proper analysis must be undertaken as to why this is the case (although this is, self-evidently, not as comprehensive as a full ‘effects analysis’). If this were not the case, the Commission would be able to prohibit, without a full analysis of their actual effects on competition, a broad range of agreements that do not by their very nature harm cause substantial harm to competition (see para. 58).
However, considerable uncertainty remains. Particularly surprising is the CJEU’s conclusion (in para. 88) that the new CB rules did not constitute a restriction by object even though internal documents seized by the Commission indicated that, by adopting the new rules, the main CB members intended to impede competition by new entrant card issuers and to protect their revenues, so limiting reductions in card fees paid by consumers. Surely measures that prevent market entry and expansion with a view to limiting price reductions, particularly where this is intended by the participants, “change appreciably the structure of the market” as much as naked cartels, output limitations and reductions in capacity? Is this not also the case where the rules are intended to achieve a given ratio between card issuing and merchant acquiring activities? In Cartes Bancaires, the CJEU thought not. In its view, these matters are relevant only an assessment of a measure’s effects (see paras. 82 and 86 to 88).
It is trite and self-evident that hard-core price-fixing and output limiting cartels restrict competition ‘by object’. As the CJEU itself observes in Cartes Bancaires, the reason that this is self-evident is that “experience shows that such behaviour leads to falls in production and price increases, resulting in poor allocation of resources to the detriment, in particular, of consumers” (see para. 51).
Whilst authorities (and courts) should not readily find a restriction ‘by object’, there are clearly many other types of agreement that do constitute ‘object restrictions’: whilst the concept of such a restriction must be interpreted widely or expansively, there are doubtless numerous types of agreement or concerted practice that can, in the appropriate context, satisfy the CJEU’s definition. From BIDS, it is plain that arrangements to cause some producers to exit a market and/or which otherwise reduce capacity in a market are also ‘object’ restrictions: this was confirmed in Cartes Bancaires.
An exchange of information also constitutes a restriction ‘by object’, particularly on highly concentrated oligopolistic markets, if this reduces or removes the degree of uncertainty as to the operation of the market, even if any effect on prices is indirect (see paras. 34 to 37 of the T-Mobile judgment). The CJEU followed this approach in Case C-541/11P Solvay, where it held that even the exchange of historic non-price information accompanied by discussions to ‘stabilise the market’ by entering into a cartel agreement can be a restriction ‘by object’, if it exchanged on a regular basis over a long period even if no such agreement is concluded.
Can other forms of horizontal or vertical coordination legitimately be found to restrict competition by object? In Allianz Hungaria, the CJEU confirmed that a series of parallel vertical agreements can – even in the absence of a horizontal agreement or concerted practice – restrict competition by object, if they significantly disrupt the proper functioning of one of the markets concerned by eliminating or seriously weakening competition. Presumably, resale price maintenance remains an ‘object’ infringement, as will restrictions on internet sales by distributors, unless objectively justified (see Case C-439/09 Pierre Fabre), and other agreements that partition markets through the use of territorial restrictions or other restrictions on parallel trade, such as differential pricing arrangements (Case C-501/06P GlaxoSmithKline) and restrictions included in agreements for the licensing of sports broadcasting rights (Case C-403/08 FA Premier League).
However, beyond these clear examples, the question remains as to how are we to judge whether an agreement or practice has “a sufficient degree of harm to competition” to have the object of restricting competition? As well as relying on experience, this requires a degree of judgment and the undertaking of some analysis: it is not permissible for an authority to simply state, without more, that an agreement has, by its very nature, the requisite negative effects on competition in order to take a short-cut to a finding of infringement. The difficulty of course is that this still leaves room for the application of the ‘duck test’ or the ‘elephant test’, since it is experience that enables us to apply those tests and also the ‘restriction by object’ test.