Op-Ed: “The Missing Link in Article 102 TFEU: Defining the Boundaries of Competition Law Enforcement” by David Pérez de Lamo
I. The Ever-Expanding Responsibility of Dominant Firms
Article 102 TFEU prohibits ‘[a]ny abuse […] of a dominant position’. Therefore, one might logically think that Article 102 TFEU requires a causal link between an undertaking’s dominant position and the alleged abusive conduct (the ‘causal link question’). Yet, this appears not to be the case when the case law of the Court of Justice of the European Union (CJEU) is examined, as will be done infra.
The causal link question might sound trivial at first sight. However, the absence of a causal link, paired with an open-ended notion of abuse, has enabled (and will further allow) competition authorities to creep unrestrainedly into adjacent legal domains via Article 102 TFEU, including data protection (excessive data collection), intellectual property (content ‘scraping’), consumer protection (opaque and arbitrary application of operating rules) and unfair competition (misleading representations and disparagement practices), inter alia. As a result, the potential scope of Article 102 TFEU has been incommensurably expanded, blurring the boundaries of competition law enforcement.
II. The CJEU’s Case Law on the Causal Link Question
The Continental Can Doctrine: Effects Over Source
The CJEU dealt with the causal link question for the first time in Continental Can (C-6/72), where the applicants argued that ‘Article [102 TFEU] reveals that the use of economic power linked with a dominant position can be regarded as an abuse of this position only if it constitutes the means through which the abuse is effected’ (paragraph 19).
The CJEU dismissed the argument, finding that
‘the question of the link of causality […] between the dominant position and its abuse, is of no consequence, for the strengthening of the position of an undertaking may be an abuse and prohibited under Article [102 TFEU], regardless of the means and procedure by which it is achieved, if it has the effects mentioned above’ (paragraph 27).
To find an infringement of Article 102 TFEU, therefore, the CJEU only checked whether the dominant undertaking strengthened its position by acquiring the relevant target company and so fettered competition (paragraphs 26-27). In other words, the CJEU established that in order to find an abuse it is only relevant to verify whether the conduct has anticompetitive effects, independently of the source thereof (no causal link is needed).
In contrast, AG Roemer had plainly disagreed with such an ‘extensive’ reading of Article 102 TFEU, which had originally been proposed by the Commission. In fact, he explicitly noted having ‘serious objections as regards the soundness of the theories put forward by the Commission’ and being ‘convinced of the applicants’ [weighty] arguments’ (page 255). Contrary to the Court’s findings, AG Roamer concluded that
‘one must […] concede that the wording of Article [102 TFEU] with its expression “abuse …of a dominant position within the Common Market”, appears to hint that its application can be considered only if the position on the market is used as an instrument and is used in an objectionable manner; these criteria are therefore essential prerequisites […]’ (page 254).
Hoffmann-La Roche: The Birth and Shaping of the Notion of ‘Abuse’
The causal link question was raised again before the CJEU in Hoffmann-La Roche (C-85/76). The applicants argued that ‘for Article [102 TFEU] to apply there must be a connexion between the dominant position and the conduct of the undertaking concerned; the conduct complained of must spring from the undertaking’s strength […]’ (page 497). This time, however, the applicants’ proposed interpretation was significantly more extensive, adding that the anticompetitive conduct ‘must only be possible by reason of its dominant position’. Under this line of argument, the applicants noted that it was not the case because the rebates were ‘quite usual on the market.’
The CJEU equally rejected the arguments raised by the applicants, giving birth to what became the classic definition of abuse (paragraph 91), illustrating its intrinsic connection with the causal link question. In particular, the CJEU found that the notion of abuse is an ‘objective concept’, focusing again on the effects of the conduct as opposed to the source thereof. This understanding was further reiterated and broadened in the CJEU’s subsequent case law, which established that a finding of abuse ‘is not necessarily a question of the actual effect [, as it is sufficient] to show that […] the conduct is capable of having that effect’ (Tomra Systems, C-549/10 P, paragraph 68; Clearstream v Commission, T-301/04, paragraph 144; and the case law cited).
In this way, the CJEU followed AG Reischl, who had concluded that, even if the causal link
‘may apply to certain situations covered by Article [102 TFEU], such as the matter of imposing unfair prices […], it certainly does not apply in general […]. [As it was stressed in Continental Can,] the criterion is not the exercise of market power but that there is abuse where an undertaking in a dominant position influences the structure of competition by its acts. […] Here too in the same way it is possible to assume an objective concept of abuse’ (page 583).
AstraZeneca: A Final Spin through ‘Dominance’
After Continental Can and Hoffmann-La Roche were handed down in the late 70s, the causal link question did not re-emerge until 2010 in AstraZeneca (T-321/05). The applicants argued that, due to price regulation and the monopoly power of the State when purchasing prescription medicines, ‘pharmaceutical companies cannot exercise market power in respect of price’ (paragraph 225). Therefore, the applicants alleged that the Commission had failed ‘to demonstrate how […] AZ could have hindered competition by behaving independently of its competitors, doctors and patients’ (paragraph 225). The applicants’ reasoning was, in effect, a spin of the same argument raised in the previous cases: questioning the existence of dominance because it could not lead to an abuse, as opposed to claiming that there had not been an abuse because it could not have originated from dominance.
Once again, the Court rejected the applicants’ arguments, reiterating the Continental Can–Hoffmann-La Roche mantra.
‘[A] finding of market power […] is not conditional on the ability of the undertaking to make use of that market power in such a way as to prevent effective competition from being maintained. [I]n order to be classified as an abuse of a dominant position, behaviour does not necessarily have to result from, or be made possible by, the economic strength of the undertaking, since no causal link is required between the dominant position and the abuse of that position’ (paragraph 267).
On appeal, neither the CJEU (C-457/10 P) nor the AG Mazák addressed the causal link question, focusing instead on the broader issue of assessing AstraZeneca’s market power (paragraphs 178-182 and 122-127 respectively). Intriguingly, when recapitulating the reasoning of the General Court from paragraphs 256 to 266, the CJEU seemed to have intentionally omitted paragraph 267 (paragraphs 179-181). This might be a credible indication that the CJEU was not comfortable with endorsing the Continental Can doctrine again, especially when it did not have to address the incidental point.
III. A Prohibition of Abuse and Dominance: A Critical Review
In sum, the CJEU established that Article 102 TFEU does not require a causal link between an undertaking’s dominant position and the alleged abuse to find an infringement. What is sanctionable then, according to the CJEU, is that a dominant company engages in conduct that is in any way capable of leading to anticompetitive effects, even if the alleged conduct has no meaningful connection with dominance. In my view, this is a rather broad interpretation, which is questionable from a series of angles.
It is contrary to the literal interpretation of Article 102 TFEU
As originally noted by AG Roemer in Continental Can, such an interpretation is plainly contrary to the wording of Article 102 TFEU, which prohibits an ‘abuse […] of a dominant position’, as opposed to ‘and a dominant position’ or an ‘abuse by a dominant company’. Similarly, the notion of abuse, in its plainest sense, implies a misuse of a certain vantage position. Indeed, an undertaking cannot abuse any market position. One must also recall that ‘only an obscure text can give rise to an interpretation that deviates from the current meaning of the terms it contains’, that is, interpretatio cessat in claris (K. Lenaerts and J. A. Gutiérrez-Fons, paragraphs 13, 16 and 18). This is not mere wordplay, but an argument that has been raised consistently by a wide range of relevant commentators ever since the late 70s, after the first two judgments were handed down (P. Vogelenzang, T. Eilmansberger, J. Drexl, T. Schrepel, A. Lamadrid, inter alia).
It is the product of fallacious legal reasoning
The CJEU’s Continental Can doctrine, on which all subsequent case law is based, is neither the result of actual reasoning nor is it imbued by special wisdom. As P. Vogelenzang noted, ‘the plaintiff’s appeal to the necessity of a causal relation was not considered by the Court of Justice on its own merits, but was brushed away on grounds of […] general considerations. […]. I find this formula […] rather cryptic’ (page 70). Indeed, the CJEU’s conclusion followed from the mix of a tautology and a non sequitur: ‘the causal link question is irrelevant because a certain conduct may be abusive regardless of the means and procedure’ (paragraph 27 rephrased). As outlined above, the subsequent Hoffmann-La Roche and AstraZeneca judgments did not add any reasoning to support such a doctrine, but legitimised the mantra through mere repetition or tradition.
It is at odds with the classic definition of dominance in United Brands
The Continental Can doctrine is at odds with the classic definition of dominance, as originally conceived in United Brands (C-27/76; equally T. Eilmansberger, page 142, and argument of the applicants in AstraZeneca).
‘The dominant position referred to in [Article 102 TFEU] relates to a position of economic strength enjoyed by an undertaking which enables it to prevent effective competition being maintained on the relevant market by giving it the power to behave to an appreciable extent independently of its competitors, customers and ultimately of its consumers’ (paragraph 65).
As can be inferred from United Brands, the problem with dominance is that (i) it suddenly enables an undertaking to behave in a way it could not before (‘conditio sine qua non’) or (ii) that it intensifies the effects of its behaviour (‘magnifying effect’). In my view, this is precisely how the causal link criterion should be construed. For instance, the Autorité de la concurrence (competition authority) appears to have followed the same two-pronged test, as shown by (i) Google (No 19-D-26, paragraph 467), and (ii) all of the disparagement cases (for example Roche/Novartis/Genentech, No 20-D-11, paragraph 779), respectively.
Moreover, it is only the fact that an undertaking becomes dominant (the legal title) which justifies the imposition of the ‘special responsibility’ (Michelin v Commission, C-322/81, paragraph 57). Non-dominant firms do not have that duty of care. Therefore, the special responsibility – that is, the range of conducts that may be classified as an abuse – must be connected to dominance.
It contravenes the established dictum that dominance is not illegal per se, as well as the principles of non-discrimination, legal certainty and legality
Sanctioning ‘any conduct capable of producing anticompetitive effects’ vastly expands the scope of Article 102 TFEU, up to a point where we are effectively sanctioning dominance per se, in spite of it not being illegal (Michelin v Commission, C-322/81, paragraph 57).
For instance, whenever a dominant firm breaches any laws of any field, it could be sanctioned under Article 102 TFEU, as it will always be capable of producing anticompetitive effects (namely an asymmetric legal burden on competitors). This includes far-fetched activities such as kidnapping the manager of a competitor (P. Vogelenzang, page 67) . However, it is clear that such conduct clearly belongs to the field of criminal law and should not concern competition law, precisely because it has no connection with firm dominance. In other words, being dominant (i) does not place an undertaking in a favourable position to kidnap the manager of a competitor and (ii) does not intensify the effects of such a conduct vis-à-vis non-dominant firms. Dominant firms should not therefore hold ‘special responsibility’ over such practices and be sanctionable under Article 102 TFEU. Otherwise, being in the same position as non-dominant firms, they would be discriminated against for no reason. The same line of reasoning could be followed about more ordinary circumstances, such as not filing VAT returns or not paying salaries, among myriad other possibilities.
Furthermore, competition authorities can sanction dominant firms for conducts that are not legally characterised or precluded (for example, by data protection, consumer protection, IP law, and so on), even if they have no connection with dominance and thus competition law.
Such an extremely expansive interpretation of Article 102 TFEU – already a rather broad and open-ended provision itself – makes it hardly foreseeable or predictable. This is clearly at odds with the principles of legal certainty and legality (Dansk Rørindustri. Joined Cases C-189/02 P, C-202/02 P, C-205/02 P to C-208/02 P and C-213/02 P, paragraphs 202, 219 and 224), which are of ‘particular importance’ in the field of competition law due to the quasi-criminal nature of competition law sanctions (K. Lenaerts and J. A. Gutiérrez-Fons, paragraph 17; ECHR Menarini v Italy, nº 43509/08, paragraphs 41-42; Bonda, C-489/10, paragraphs 36-37).
The CJEU has already required a causal link in cases of leveraging
In Tetra Pak II (C-333/94 P), the CJEU has already required a causal link for cases of leveraging to adjacent markets (paragraph 27) and there is no objective reason to limit that understanding to those cases. In fact, leveraging also occurs within the same market. As the Commission’s Guidance on Article 102 TFEU demonstrates, abuses such as exclusive dealing and conditional rebates consist precisely of leveraging the non-contestable part of the demand (the part of demand over which the undertaking is an ‘unavoidable trading partner’ as a dominant firm), to the contestable part, within the same market (Guidelines, paragraph 39; Post Danmark II, C-23/14, paragraph 40).
One could read Tetra Pak II (paragraph 27) in the sense that Article 102 TFEU presupposes a causal link dominance-abuse in all cases (for example, T. Eilmansberger, page 142, understands it in this way). However, that is not the most reasonable interpretation, given that the CJEU (i) appears to subsequently limit the causal link requirement to cases of leveraging (R. Whish, page 210), and (ii) it made no reference to the related precedents Continental Can and Hoffmann-La Roche. In fact, the latter were reiterated later on by the General Court in AstraZeneca (T-321/05).
IV. Overrule the Continental Can doctrine
For all of these reasons, the CJEU should firmly overrule the Continental Can doctrine and require a causal link for dominance-abuse under Article 102 TFEU. That would not entail an overhaul of the current interpretation of the notion of abuse. Simply, the latter would be more sensible if moderated with a causal link doctrine that would reasonably and foreseeably circumscribe the special responsibility of dominant firms. In the absence of a causal link and given the broad interpretation of abuse, it is only inevitable that the scope of Article 102 TFEU will continue to expand limitlessly and unpredictably, unduly overextending competition law enforcement.
David Pérez de Lamo is a Legal Consultant at a law firm and holds an LL.M. in EU Law and Economic Analysis from the College of Europe (Bruges). He has published various contributions in EU competition law and constitutional law.