The Draft Article 102 Guidelines: A Somewhat Confused Attempt to Partly Roll Back the Effects-based Approach of the Union Courts

Introduction

Over the summer, the Commission published its long-awaited draft Article 102 Guidelines (‘draft GL’) for comments.[1] The stated intention of the Commission is to base the Guidelines on the case law of the Union Courts (§9). The stated goals of the Guidelines are to allow a vigorous and effective application of Article 102, but also to ensure that Article 102 is applied in a predictable and transparent manner to enhance legal certainty and help undertakings self-assess whether their conduct constitutes an exclusionary abuse under Article 102 TFEU, and to give guidance to national courts and national competition authorities in their application of Article 102 (§4 and 8).

In the run-up to the publication of the draft GL, it seemed that another intention of the Commission was to partly roll back the effects-based approach introduced by the 2008 Article 102 Guidance Paper[2] and which has subsequently been followed by the Union Courts in a series of judgments beginning with Post Denmark I and Intel.[3]

Reading the draft GL, one cannot avoid the impression that the intention to roll back the effects-based approach has ‘won’, but at the unfortunate cost of reducing the coherence, clarity and predictability of the proposed enforcement policy. The result is a document which, while in part clear and concisely written, creates unnecessary confusion, basing itself on a sometimes selective and biased reading of the case law.

A confusing and superfluous two-step test

In §45 of the draft GL, the Commission introduces a conceptual distinction between finding that conduct is not competition on the merits and establishing the conduct’s capability to have exclusionary effects. Based on that distinction, it states that it is generally necessary to apply a two-step test to find an abuse. Hence, finding an abuse would require both a finding that the conduct departs from competition on the merits and that the conduct is capable of having exclusionary effects.

In support of this proposed two-step test, the Commission refers, in particular, to Servizio Elettrico Nazionale.[4] In §103 of that judgment the Court indeed states that the conduct of a dominant firm … may be characterised as ‘abusive’ for the purposes of that provision if it is capable of producing an exclusionary effect and if it is based on the use of means other than those which come within the scope of competition on the merits. That these two conditions are cumulative is not only clear from the conjunction ‘and’ but also from what follows in §103 (Where those two conditions are fulfilled ….).

The question is why the Commission proposes this two-step test? It is certainly true that in most of the case law, including in more and less recent Grand Chamber judgments such as Post Denmark I and European Superleague[5], the Court of Justice has generally used wording like ‘the use of means other than those which come within the scope of competition on the merits’ in conjunction with wording such as ‘conduct which hinders competition and is likely or capable to have anti-competitive effects’. However, this does not necessarily indicate that an explicit two-step test is required. Effectively, in a long line of case law since Post Denmark I, the focus has shifted and is on whether the conduct hinders competition/has actual or likely exclusionary effects to the detriment of consumers:

In Post Denmark I (§24) the test reads: Article 82 EC applies, in particular, to the conduct of a dominant undertaking that, through recourse to methods different from those governing normal competition on the basis of the performance of commercial operators, has the effect, to the detriment of consumers, of hindering the maintenance of the degree of competition existing in the market or the growth of that competition.
Similar wording is found in European Superleague (§129-131): In order to find, in a given case, that conduct must be categorised as ‘abuse of a dominant position’, it is necessary, as a rule, to demonstrate, through the use of methods other than those which are part of competition on the merits between undertakings, that that conduct has the actual or potential effect of restricting that competition by excluding equally efficient competing undertakings from the market(s) concerned …, or by hindering their growth on those markets … That demonstration must, moreover, be aimed at establishing, on the basis of specific, tangible points of analysis and evidence, that that conduct, at the very least, is capable of producing exclusionary effects …. In addition, conduct may be categorised as ‘abuse of a dominant position’ not only where it has the actual or potential effect of restricting competition on the merits by excluding equally efficient competing undertakings from the market(s) concerned, but also where it has been proven to have the actual or potential effect – or even the object – of impeding potentially competing undertakings at an earlier stage, through the placing of obstacles to entry or the use of other blocking measures or other means different from those which govern competition on the merits, from even entering that or those market(s) and, in so doing, preventing the growth of competition therein to the detriment of consumers, by limiting production, product or alternative service development or innovation …
Or more briefly in Post Denmark II (§67 and 69)[6]: It follows that only dominant undertakings whose conduct is likely to have an anticompetitive effect on the market fall within the scope of Article 82 EC. … Such an assessment seeks to determine whether the conduct of the dominant undertaking produces an actual or likely exclusionary effect, to the detriment of competition and, thereby, of consumers’ interests. (referring to Post Denmark I).

On the basis of the case law, it is therefore equally possible, as others – including me – have argued before, that the focus of finding an abuse should be on establishing likely or actual exclusionary effects and should not require an additional and independent finding that the conduct departs from competition on the merits.^[7] In order to fully protect consumers, it is justified to conclude that as soon as the conduct of a dominant firm is likely to have exclusionary effects to the detriment of consumers in the relevant market(s), the conduct automatically departs from competition on the merits. Otherwise, consumer harm could be excused by defining the conduct as being competition on the merits.

However as said, the Commission in the draft GL stresses the relevance of a conceptual distinction between finding that conduct departs from competition on the merits and establishing the conduct’s capability to have exclusionary effects, and advocates a two-step test to find an abuse. This gives the impression that the Commission considers that it is not only theoretically possible but also practically relevant to distinguish four scenarios (see the diagram below): (1) conduct that is competition on the merits and that is not capable of having exclusionary effects, (2) conduct that is not competition on the merits but that is not capable of having exclusionary effects, (3) conduct that is competition on the merits but that is capable of having exclusionary effects, and (4) conduct that is not competition on the merits and that is capable of having exclusionary effects.

	Conduct that is competition on the merits	Conduct departing from competition on the merits
Conduct not capable of having exclusionary effects	1	2
Conduct capable of having exclusionary effects	3	4

It is clear that types of conduct falling in cell 1 are not abusive while types of conduct falling in cell 4 are abusive unless they can be justified based on objective necessity or efficiencies. To focus the concept of abuse on cell 4 is in line with the case law, as described above, which effectively treats conduct departing from competition on the merits and capability to have exclusionary effects as two sides of the same coin.

However, the Commission, by stressing the conceptual distinction and advocating a two-step test to find an abuse, must consider that in practice also cells 2 and 3 are important. If not, if all relevant conduct falls either in cell 1 or cell 4, stressing the distinction and introducing the two-step test is at best superfluous and in practice also confusing. So, does the Commission in the draft GL describe types of conduct which it considers to fall in the two remaining cells 2 and 3?

In the draft GL no description is found of conduct that would fall in either cell 2 or 3. Not in section 3, which presents general principles for determining whether conduct by a dominant undertaking is liable to be abusive and where the distinction and two-step test are put forward, nor in section 4, which sets out principles for determining whether specific types of conduct are liable to be abusive.

The parts of section 3 that attempt to describe what is and what is not competition on the merits (most of section 3.1 and the whole of section 3.2) are, as a result, at best superfluous but also confusing. Just a few examples:

In §51, the Commission seems to acknowledge that competition on the merits and absence of negative effects are two sides of the same coin: The concept of competition on the merits covers conduct within the scope of normal competition on the basis of the performance of economic operators and which, in principle, relates to a competitive situation in which consumers benefit from lower prices, better quality and a wider choice of new or improved goods and services. This is not surprising, as this quote is derived from case law that effectively emphasises the need to show anti-competitive exclusionary effects.
In §49, the draft GL state that … the fact that an undertaking is in a dominant position does not disqualify it from protecting its own commercial interests, if they are attacked. However, this already unclear description of competition on the merits is immediately curtailed by saying that the dominant firm … may take reasonable and proportionate steps as it deems appropriate to protect its commercial interests, provided however that its purpose is not to strengthen its dominant position or to abuse it. In other words, protecting its own interests is no longer competition on the merits if it objectively results in helping the dominant firm to strengthen its dominance or capability to produce exclusionary effects.
In §57, the draft GL try to specify pricing conduct that is competition on the merits: Conduct that at first sight does not depart from competition on the merits (e.g. pricing above average total costs (“ATC”)) and therefore does not normally infringe Article 102 TFEU may, in specific circumstances, be found to depart from competition on the merits, based on an analysis of all legal and factual elements, notably: (i) market dynamics; (ii) the extent of the dominant position; and (iii) the specific features of the conduct at stake. What the analysis of all legal and factual elements in the end is supposed to establish is whether the pricing conduct is capable of having exclusionary effects.
Finally, in §58, the draft GL provide the most surprising and confusing example of competition on the merits: if a dominant undertaking is able to argue that the actual or potential exclusionary effects produced by its conduct are counterbalanced or outweighed by advantages in terms of efficiencies that benefit consumers, this means that its conduct amounts to competition on the merits. So, depending on whether the efficiency defence succeeds or fails, the same conduct is or is not competition on the merits.

An attempt to partly roll back the effects-based approach

Given that one of the aims of issuing Article 102 Guidelines is to strengthen enforcement by allowing a vigorous and effective application of Article 102, this begs the question why the Commission is advocating this conceptual distinction and a two-step test for finding an abuse.

Introducing two cumulative conditions would logically make it harder to establish an abuse than if only one of the conditions (showing capability to have exclusionary effects) were sufficient. A two-step test will not strengthen the enforcement of Article 102. Conduct in cell 2 – conduct which departs from competition on the merits but which is not capable of having exclusionary effects – would not be abusive, as the draft GL also seem to acknowledge in §5 and section 3.3. The same applies for conduct in cell 3: conduct which is competition on the merits but which is nonetheless capable of having exclusionary effects, i.e. conduct which produces an actual or likely exclusionary effect to the detriment of competition and consumers. It would not be abusive, as it is competition on the merits, as the draft GL seem to say in §50 (…the Union Courts have established that only conduct that deviates from competition on the merits can constitute an exclusionary abuse within the meaning of Article 102 TFEU – referring to Servizio Elettrico Nazionale §103).

As explained above, an explicit two-step test, with two (at least partly) independent steps, is not necessarily required by the Article 102 case law. So, what could explain this choice and interpretation of the case law by the Commission, which at the same time has not provided any examples of conduct falling into cells 2 and 3? This choice will certainly not enhance legal certainty or help undertakings self-assess whether their conduct constitutes an exclusionary abuse, another stated aim of the (draft) GL.

The only explanation that comes to mind is a reluctance by the Commission to accept what the Union Courts have said consistently since Post Denmark I, namely that establishing an abuse is in essence about establishing actual or likely anti-competitive effects, no more, no less. The draft GL give the impression that the Commission hopes that once a particular conduct can be described as deviating from competition on the merits, this will somehow reduce the evidentiary burden of showing capability or likelihood of having exclusionary effects. The text of the draft GL reflects this reluctance in several ways:

While the case law, as shown by the passages from Post Denmark I and II and European Superleague quoted above, uses interchangeably ‘having likely or actual exclusionary effects’, ‘actual or potential exclusionary effects’ and ‘capability to have exclusionary effects’, the text of the draft GL systematically avoids using the words ‘likely exclusionary effects’, only using ‘capability to have exclusionary effects’. As if by doing so the standard of proof can somehow be lowered.^[8]
A similar remark can be made about the draft GL only referring to exclusionary effects (see for instance the definition in §6), without linking the effects to harm to consumers, whereas the case law – like the 2008 Guidance Paper – makes the important distinction between exclusionary effects and exclusionary effects that harm consumers (anti-competitive foreclosure). This suggests going back from ‘protecting competition’ to ‘protecting competitors’.
In §60, where the Commission describes the rebuttable presumption from the Intel judgment for certain types of (exclusivity requiring) conduct, the Commission suggests, in an unclear sentence, that (even) where the dominant firm is able to rebut the presumption, the evidentiary burden for the Commission to show abuse should still somehow be reduced: Even in the scenario set out in (ii), the evidentiary assessment must give due weight to the probative value of a presumption, reflecting the fact that the conduct at stake has a high potential to produce exclusionary effects, as part of the overall assessment of the body of evidence in the light of all the relevant legal and economic circumstances.^[9]
In §62 of the draft GL it is suggested that, in order to show that a conduct is capable of having exclusionary effects, it is sufficient to show that the conduct was capable of removing the commercial uncertainty relating to the entry or expansion of competitors. This unclear phrasing is much wider in scope than what the General Court stated in §363 of its Lundbeck judgment, to which the Commission refers: Accordingly, by concluding the agreements at issue, the applicants exchanged that uncertainty for the certainty that the generic undertakings would not enter the market, by means of significant reverse payments (recital 604 of the contested decision), thus eliminating all competition, even potential, on the market, during the term of those agreements.^[10]
In the next sentence in §62, the Commission seems to say that actual effects and the actual conduct of competitors cannot negate or put doubt on ‘shown’ capability to have exclusionary effects: Moreover, where it is established that a conduct is objectively capable of restricting competition, this cannot be called into question by the actual reaction of third parties. The Commission refers here to §360 of the AstraZeneca ^[11] However, in that paragraph the General Court speaks of conduct which is objectively of such a nature as to restrict competition, i.e. of what are also called naked restrictions/by object conduct. This obviously does not imply that in general (certain) facts and effects can be disregarded when showing capability to have exclusionary effects.
Something similar is found in §70 c) of the draft GL: While the position of competitors is relevant in the assessment, the finding of capability to produce exclusionary effects cannot be called into question by the actions that competitors may have taken – or could have taken – to limit the effects of the conduct of the dominant undertaking. This a very broad interpretation of §102 of Servizio Elettrico Nazionale, where the Court of Justice made this point in relation to situations where the possible reactions of competitors could (only) limit the harmful consequences of that practice.
In section 4.2.1 of the draft GL (relating to exclusive dealing), the Commission suggests that the Union Courts have treated incentive schemes that are conditional on a customer or a supplier purchasing or selling all or most of their requirements from/to the dominant undertaking in the same way as an obligation to purchase or sell all or most of a customer or a supplier’s requirements from/to the dominant undertaking. Specifically, the Commission implies that a price-cost test/the as efficient competitor test has no role to play in assessing the capability to exclude of, for instance, loyalty rebates. This goes directly against what the Court of Justice has recently said about the assessment of all forms of pricing conduct in §79-80 Servizio Elettrico Nazionale, one of the judgments which the Commission otherwise refers to most often. In its judgment the Court of Justice says that for pricing conduct, … which includes loyalty rebates, low-pricing practices in the form of selective or predatory prices and margin-squeezing practices, it is clear from the case-law that those practices must be assessed, as a general rule, using the ‘as-efficient competitor’ test, which seeks specifically to assess whether such a competitor, considered in abstracto, is capable of reproducing the conduct of the undertaking in a dominant position.^[12]

In view of the case law – but possibly more importantly because also (1) economically and in terms of effects there is no fundamental difference between exclusivity and non-exclusivity rebates, (2) the practical difficulty of drawing the boundary between exclusivity and non-exclusivity rebates and (3) in view of the fact that the possible practical difficulties in applying the AEC test depend on much more than a dichotomy between exclusivity and non-exclusivity rebates^[13] – it is not only not justified but also unfortunate and can only produce a policy bias to single out one type of pricing conduct where the AEC test would generally not play a role in the assessment. Instead, the Commission would do better to accept, as required by the case law overall, that the AEC test is in general a useful tool to assess the capability of pricing conduct to harm competition and consumers.

In §95 of the draft GL, the Commission states that in certain circumstances, it may be possible to conclude that, due to the specific characteristics of the markets and products at hand, the tying has a high potential to produce exclusionary effects and those effects can be presumed. This leaves unclear when this rebuttable presumption kicks in. The only ‘explanation’ is the following vague sentence in a footnote: This is notably the case in situations where the inability of competitors to enter or expand their presence in the tied market is likely to directly result from the tying conduct due to the absence of clearly identifiable factors that could offset the exclusionary effects.
In §112 of the draft GL, the Commission extends the Intel rebuttable presumption to all scenarios of predation, whereas the predation-related case law only applies the presumption for pricing below average variable cost/average avoidable cost (AVC/AAC).
In §143-145 of the draft GL, the two-step test again creates unnecessary confusion. First it is said that [T]o demonstrate that a conditional rebate scheme departs from competition on the merits, it may be appropriate to make use of a price-cost test. Subsequently it is said that the use of a price-cost test may not be appropriate in certain cases. In these latter cases, the assessment of whether the conduct departs from competition on the merits will be carried out on the basis of the general principles set out in section 3.2. However, as already explained, that section does not provide useful principles. The real test is whether the pricing has the capability to have exclusionary effects. There is no good reason to separate the assessment of the pricing of the dominant firm and the use of a price cost test from the assessment of whether the pricing is capable of having exclusionary effects.
While the case law since Post Denmark I consistently refers to the central role of the AEC test when assessing pricing conduct (see for instance the quotes above from European Superleague and Servizio Elettrico Nazionale), the draft GL seem to intentionally downplay and sometimes deny the AEC test’s obviously important role in finding an abuse. This appears most clearly in the assessment of conditional rebates (section 4.3.1, §145). First it is said that [T]he fact that even a hypothetical as-efficient competitor would be unable to compensate the loss of the rebates as demonstrated by means of a price-cost test can be a consideration for finding capability to have exclusionary effects. However, the text continues by saying that … the fact that a hypothetical as-efficient competitor would be able to compensate the loss of the rebates is not necessarily a relevant factor showing that the rebates scheme is incapable of producing exclusionary effects. This is because the conduct’s capability to have exclusionary effects needs to be assessed in relation to the existing actual or potential competitors of the dominant firm, rather than in relation to hypothetical competitors. In other words, no (soft) safe harbour whatsoever for pricing above cost in case of conditional rebates. If competitors happen to be less efficient, the AEC test has no role or meaning anymore? As said, this is in stark contradiction with the central role that the case law gives to the AEC test for the assessment of pricing conduct.
In that same section of the draft GL (§144), it is also said that The use of a price-cost test may not be appropriate in cases where …. the emergence of an as-efficient competitor would be practically impossible, for instance, because of the dominant undertaking’s very large market share or the presence of significant barriers to entry or expansion in the market, or the existence of regulatory constraints. In these circumstances, even a less efficient competitor may also exert a genuine constraint on the dominant undertaking. However, also in such circumstances, it would still make sense to apply a price cost test, if only to estimate how much the rebated price remains above cost; if the rebated price is only slightly above cost, it is more likely that it may reduce the possible constraint from less efficient competitors, but if the rebated price is substantially above cost this is not likely.
Section 4.3.2 of the draft GL on multi-product rebates says that for their assessment … the guidance provided by the case-law in relation to exclusive dealing and conditional rebates, depending on the cases, applies by analogy. This ignores the logical link between multi-product rebates and tying and bundling.

Some concluding remarks

Do the above comments and criticisms mean that nothing should be retained of the draft GL? No, certainly not. While sections 3.1 and 3.2 are superfluous and confusing, and parts of them would be better integrated in the sections on the assessment of the capability to have exclusionary effects, other sections, as already said in my introduction, are well written, clear and concise. While sections 3.3 and 4 certainly need changes and improvements as indicated above, this is not to say that those sections, as also section 5, are not overall well written.

In some sections, the conciseness may have gone a bit too far and some more explanation would be useful. For instance, in §118 of the draft GL, regarding the application of the price-cost test, it is said: Furthermore, it may be appropriate to account for opportunity costs of the dominant undertaking. This short sentence somewhat hides the point that predation can sometimes also be found when the dominant firm prices above cost, i.e. in the case of pricing above average total cost/long run average incremental cost (ATC/LRAIC). Some more explanation, for example that this may be the case if the pricing conduct would not pass the (no) economic sense test, in other words if it can be shown that the dominant firm is sacrificing profits while avoiding a loss, would be helpful for a better understanding.

Section 2 on the assessment of dominance is also overall well written and contains a useful new sub-section on collective dominance and tacit coordination. However, this section, if adopted as it currently stands, will seriously undermine legal certainty and predictability, and again for no good reason. Most notable, in §26 the text speaks of the presumption of dominance above 50% market share and refers to a very low 10% soft safe harbour: … very large market shares … are in themselves – save in exceptional circumstances – evidence of the existence of a dominant position. This is the case in particular where an undertaking holds a market share of 50% or above. Dominance may also be found in cases where an undertaking has a market share below 50%. … Market shares below 10 % exclude the existence of a dominant market position save in exceptional circumstances. Although the last sentence is only in a footnote, the text gives the impression not only that dominance is presumed above 50% market share, but also that dominance may be likely between 10% and 50%, depending on other factors, and that only below 10% is there a safe harbour, and even there notwithstanding the possibility to find dominance in exceptional circumstances. This is not only pretty ridiculous in view of the 10% market share threshold in the De Minimis Notice[14], but this scaremongering is also not necessary and will only create legal uncertainty. In the 2008 Guidance Paper it was said, based on case experience until then, that a finding of dominance was unlikely below 40% market share. Since then, the Commission has done a good number of Article 102 cases and in none of them, as far as I know, has dominance been found at market shares below 50%, while in most cases the market shares were well above 70%.

To conclude, while it is certainly possible to issue Article 102 Guidelines based on the case law that allow for vigorous and effective enforcement, as well as ensuring that Article 102 is applied in a predictable and transparent manner, enhancing legal certainty and helping undertakings self-assess whether their conduct constitutes an exclusionary abuse, the current draft GL are, to put it mildly, not there yet.

* The opinions expressed are strictly personal. While I previously worked for the European Commission, nothing in this chapter represents the views of the European Commission, DG COMP or any other institution, entity, person, etc. I have no interests to declare except my own interest as a European consumer. All errors and omissions are mine.

[1] Draft Guidelines on the application of Article 102 of the Treaty on the Functioning of the European Union to abusive exclusionary conduct by dominant undertakings

[2] Guidance on the Commission’s enforcement priorities in applying Article 82 of the EC Treaty to abusive exclusionary conduct by dominant undertakings. For this intention, see in particular the Communication amending the Guidance and Staff Policy Brief.

[3] Judgment of 27 March 2012, Post Danmark, C-209/10, EU:C:2012:172 and judgment of 6 September 2017, Intel v Commission, C-413/14 P, EU:C:2017:632

[4] Judgment of 12 May 2022, Servizio Elettrico Nazionale and Others, C-377/20, EU:C:2022:379.

[5] Judgment of 21 December 2023, European Superleague Company, C‑333/21, EU:C:2023:1011.

[6] Judgment of 6 October 2015, Post Danmark, C-23/14, EU:C:2015:651.

[7] See, for instance, opinion of Advocate General Rantos, delivered on 9 December 2021, in case Servizio Elettrico Nazionale and Others, C-377/20, EU:C:2021:998, §61. As for me, Luc Peeperkorn, Conditional pricing and the AEC test: A happy marriage or an awkward couple?, Concurrences No. 2-2019, §8.

[8] The exception is §169 of the draft GL, where the Commission describes the conditions for a successful efficiency defence as follows: … that the efficiency gains likely to result from the conduct under consideration counteract any likely negative effects on competition and on the interests of consumers in the affected markets. As if the Commission suddenly realised that a balancing of negative and positive effects would normally require the same standard of proof to be applied for both types of effects.

[9] The Commission is clearly trying to make the most of the rebuttable presumption, thereby ignoring the advice given by J. L. da Cruz Vilaça when he was a judge at the ECJ, in his article The intensity of judicial review in complex economic matters—recent competition law judgments of the Court of Justice of the EU, Journal of Antitrust Enforcement, 2018, Vol. 6, Issue 2, p. 183: “I think the regulator would be well advised in the future not to expect any special indulgence for relying on a presumption of any kind of infringement per se of Article 102 TFEU when seeking to prove the anticompetitive character of such a system.”

[10] Judgment of 8 September 2016, Lundbeck v Commission, T-472/13, EU:T:2016:449, paragraph 363.

[11] Judgment of 1 July 2010, AstraZeneca v Commission, T-321/05, EU:T:2010:266, paragraph 360.

[12] It is important to distinguish between situations where the buyer is required to purchase all or most of its requirements from the dominant firm (exclusive purchasing) and where the buyer may be induced to do so (loyalty rebates). In the case of an exclusive purchasing obligation, there is a commitment on the part of the buyer to purchase only from or at least a particular amount from the supplier, whereas in the case of a rebate scheme, there is a commitment on the part of the supplier to offer different prices depending on how much the buyer purchases.

[13] For more background on these practical difficulties for various types of rebates, see Luc Peeperkorn, Conditional pricing and the AEC test: A happy marriage or an awkward couple?, Concurrences No. 2-2019.

[14] Commission Notice on agreements of minor importance which do not appreciably restrict competition under Article 101(1) of the Treaty on the Functioning of the European Union (De Minimis Notice), OJ C 291, 30.8.2014, pp. 1-4.

________________________

To make sure you do not miss out on regular updates from the Kluwer Competition Law Blog, please subscribe here.

Leave a Reply Cancel reply