Introduction

The EU Directorate General for Competition (DG COMP) has issued new draft guidelines on the application of Article 102 of the Treaty on the Functioning of the European Union (TFEU), which prohibits dominant companies from excluding competitors from the market through predatory pricing, margin squeeze, exclusive dealing and refusal to supply as well as other abusive behaviors. These new draft guidelines (DGL) represent a major effort and include a significant update and upgrade on its predecessor version of 2008. Much has happened in the last 16 years, but overall, the new draft seems at first sight less revolutionary than expected after the dramatic technological changes we have witnessed over the last few years.

To assess and comment on a new document requires identifying a benchmark, a standpoint for review. For antitrust hipsters who see in any dominant position a threat to democracy and freedom, the new draft may still not go far enough. For adepts close to Chicago School of thinking, interference with dominance means a punishment of success proven by efficiencies achieved—efficiencies that are not quantifiable but inherent in the company’s success. However, we must remember that Chicago’s tolerance for high concentration levels was justified by the possibility of ex post enforcement to correct abuses.

The angle of review taken here is neither woke nor libertarian—rather we focus on practicability and legal certainty. Ideally, guidelines provide workable guidance to the practitioner and the company concerned. Ideally, guidelines are balanced documents that aim at identifying clear Dos and Don’ts The reality here is different, closer to “Don’ts and Dos but…”.

Before going into the substance of the DGL, we salute its novel methodology in structuring. We identified three parts:

• Section 4.2 provides guidance on certain categories of conduct that have been the object of judgments by Union courts, namely conducts subject to “specific legal tests.” Paragraph 47 explains that the case law of the Union courts has developed specific analytical frameworks to establish whether certain types of conduct by dominant undertakings infringe Article 102 TFEU. A footnote refers to paragraph 130 of the European Super League judgment.¹ When a given conduct meets the conditions set out in a specific legal test, that conduct is presumed to be potentially abusive because it falls outside the scope of competition on the merits and is capable of having exclusionary effects. The specific legal tests relating to five types of conduct are outlined in Section 4.2.
• Section 4.3 discusses specific types of conduct for which no “specific legal test” has been developed but for which the Union courts have provided guidance on how to apply the general legal principles set out in Section 3.
• Sections 2, 3 and 5 set out the general principles underlying the Commission’s assessment of dominance (Section 2), abuse (Section 3) and objective justifications (Section 5, which is a short section of only three paragraphs).

For the advisor, this somewhat simplifies the task. Sections 4.2 and 4.3 basically summarize the existing case-law and should be read first. If these sections do not cover a new form of conduct, Section 3 reminds us that any novel business practice can become the Commission’s next case. The message is that even outside of case law precedents, antitrust enforcement is looming above strong companies’ heads. Adding to that, in the European institutional setup, the investigator is also the first-instance judge; hence, we realize the challenge that companies and advisors face in ensuring compliance. With that in mind, we have limited our comments below to these general principles and point to a few instances where the enforcer’s margin of appreciation appears particularly if not overly broad.

The DGL, limited to exclusionary abuses, set out to ensure that markets remain open and dynamic, creating new opportunities for innovative players, including small and medium-sized enterprises and startups to operate on a level playing field with other players. They aim at ensuring innovation and an efficient allocation of resources, thereby contributing to sustainable development and enabling strong and diversified supply chains—a prerequisite for resilience and long-term prosperity. The competitive harm identified includes higher prices, a deterioration in the quality of goods and services, a reduction in innovation as well as a limitation of consumers’ choices. In view of digitization, which makes strong network effects and “winner-takes-all” dynamics increasingly widespread, “it is important that Article 102 TFEU is applied vigorously and effectively” (§4).

The DGL emphasize the need that Article 102 TFEU should be applied in a predictable and transparent manner so that companies can operate freely in the internal market. We agree but there is still much vagueness in the language. The DGL are certainly transparent in that they leave no safe haven whatsoever to potentially dominant companies. Virtually every conduct can amount to an abuse, suffice its capability to produce exclusionary effects, which as we know does not require such effects to ever materialize. Accordingly, Article 102 TFEU applies to all commercial practices by dominant undertakings which may directly or indirectly harm consumer welfare, including by undermining an effective structure of competition (§5). The DGL are also intended to guide national courts and national competition authorities in their application of Article 102 TFEU but will not necessarily influence the application of stricter national law. The advocated case-by-case analysis is a double-edged sword given the multiplicity of applicable criteria, allowing the enforcer to reach a desired result in virtually all cases.

Dominance

The definition of dominance has not changed: It is a position of economic strength which enables the undertaking to prevent effective competition by giving it the power to behave to an appreciable extent independently of its competitors, customers and consumers. Dominance is not precluded by a “certain degree” of competition, as long as the undertaking can act to an “appreciable extent” as it desires.

Defining the relevant market with the help of the new Market Definition Notice,² which has brought some novelties in relation to digital markets, requires identifying the competitive constraints exerted on undertakings when offering products (or services) in a certain area. In doing so, an already existing market power should be taken into account. Thus, market definition and the assessment of dominance “may therefore be interrelated” (§20), which may facilitate the finding of dominance.

Once dominance has been established, its degree is only relevant for the finding of potential exclusionary effects. Dominance derives in general from a combination of several factors that, taken separately, are not necessarily determinative (§24), another twist facilitating a finding of dominance.

Generally, very large “market shares” (= +50%) are in themselves evidence of dominance, which, however, may also be found at lower market shares. The DGL concede that in fast-growing markets with short innovation cycles, a high market share may be ephemeral, but in practice this prognosis may be difficult to make. Are high market shares sustained for more than two years (cf. §32) an indicator of dominance or a warning that technological change may be looming on the horizon?

Forms of “entry barriers” are well explained, and special attention is given to “platform markets,” where network effects can make it more difficult for entrants to simultaneously attract sufficient users on both sides of the platform, in particular when users single-home (§31).

“Countervailing buyer power” (which differs from general bargaining power and rather refers to the ability of customers to easily switch to competitors), if sufficiently strong to exercise competitive constraints, may deter or defeat an attempt by the undertaking concerned to exercise market power. However, buyer power of a limited customer segment does not suffice (this is clear!), and it is less likely present where buyers are dispersed (§33).

“Collective dominance” requires (no news here) that two or more legally independent entities present themselves or act together on a particular market as a collective entity from an economic point of view. Competition between these does not need to be completely eliminated, or that the undertakings adopt identical market conduct; it is sufficient that the abusive action “can be identified as one of the manifestations of such a joint dominant position.” This requires a look at the “economic links or factors giving rise to a connection,” which may result from an agreement, from structural or other links (e.g., personal ties), “other connecting factors,” or from (a rather open-ended) economic assessment of the structure of the market in question and the way in which the undertakings in question interact on the market (ability to monitor adherence, existence of a credible deterrence mechanism, absence of competitive constraints). It suffices to merely find that the characteristics of the market facilitate the adoption of a common policy (§36).

Abuse

Dominant companies have a “special responsibility” not to abuse their dominance. The concept of abuse is an objective one, which only means that there is no requirement of bad intent (although that helps in finding abuse). There is no objective definition of the objective concept of abuse; all it takes is a finding that conduct, whatever its form, “departs from competition on the merits” and is “capable” of having exclusionary effects. These two concepts are conceptually different, but the same facts can be relevant for their finding (§46)—not that this would amount to double counting.

What is “competition on the merits”? Does it include hyper-competition? Dominant companies may of course protect their own commercial interests, if they are attacked, provided that their purpose is not to strengthen a dominant position or to abuse it (§49). The defense must fall within the scope of competition on the merits, which covers conduct within the scope of normal competition on the basis of the performance of economic operators and which includes some pro-consumer elements. Not surprisingly for the learned reader, the intention to compete on the merits, even if established, is insufficient to prove the absence of abuse—all the more as certain ordinary practices become objectionable in the hands of dominant players (§52).

It is here that the “specific legal tests” developed by the Union courts come into play with regard to potentially abusive conduct. In addition to the Union courts’ guidance on conditional and multi-product rebates, self-preferencing and access restrictions (Section 4.3), these tests include (section 4.2):

• Exclusive dealings
• Tying and bundling
• Refusal to supply
• Predatory pricing
• Margin squeeze 

The practices covered by the “specific legal tests” in Section 4.2 are deemed outside competition on the merits (§53). So-called “naked restrictions,” defined as conduct with zero economic interest for the dominant company other than the restriction of competition, is also deemed outside competition on the merits. For other cases the Union courts have considered relevant factors. They include, for example:

• Discriminatory conduct
• Self-preferencing
• Abrupt and abnormal changes of conduct
• Use of misleading information
• Consumer-hostile conduct
• Violations of other laws affecting parameters for competition
• Conduct blocking as efficient competitors by leveraging a dominant position (§55)

Section 4.2-type “pricing practices” such as predatory pricing or margin squeeze require a price-cost test to establish departure from competition on the merits. Its outcome can also inform the establishment of potential exclusionary effects. But even “ordinary” conduct, such as pricing above average total cost, can depart from competition on the merits depending on market dynamics, the degree of dominance and “the specific features of the conduct at stake” (§57).

Capability of exclusionary effects assessment: Some forms of conduct are presumed to lead to exclusionary effects: these include the “naked restrictions” and the practices listed in Section 4.2. In the first case the presumption is very difficult to rebut; in the second case the quality of defense submissions increases the “evidentiary burden” on the enforcer side. However, the evidentiary assessment “must give due weight to the probative value of a presumption” (§60 b)).

In other cases, the Commission must establish a conduct’s capability to produce exclusionary effects (in short a “capability assessment”). These effects must be more than hypothetical but do not have to have materialized. This capability assessment is fact-specific; it is sufficient to show that the conduct was capable of removing the commercial uncertainty relating to competitive entry or expansion (the reference to commercial uncertainty, a feature of horizontal collusion analysis, is surprising here). A conduct in place for a sufficiently long period of time may have produced “actual” exclusionary effects. However, the “absence” of actual exclusionary effects “cannot in itself” disprove its capability to produce such exclusionary effects, since this may be due to extraneous factors (e.g., unrelated market changes, inability to implement an exclusionary strategy, unexpected reaction of third parties, or natural cataclysms—not mentioned in the DGL). In the absence of actual effects, the dominant undertaking must demonstrate that their absence was indeed the consequence of the fact that that conduct was unable to produce such effects (§64)—it is not clear whether this is intended to imply a reversal of the evidentiary burden in Section 3 and Section 4.3. cases.

While the identified potential exclusionary effects need to be attributable to the conduct at issue, that conduct does not need to be their sole cause. “It is sufficient to establish that the conduct contributes to increasing the likelihood of the exclusionary effects materializing on the market.” (§65).

The following sections, both inspired by the Google Shopping case, are those where the DGL appear to fail in clarity and maybe also in substance: Conceptually, a capability analysis requires a “counterfactual” comparison with the situation absent the conduct, e.g., by comparing the market situation “before and after” the implementation of the conduct (§66). It is wholly unclear how a such a counterfactual can be a suitable test to identify potentially exclusionary conduct not requiring actual effects, in particular, if the implementation is recent but the conduct is hit by a presumption. Such a test may be suitable in specific circumstances, but it should not be a general feature of the capability assessment.

Further, in certain cases, it may be appropriate to use as a counterfactual “an alternative hypothetical scenario where the conduct would be absent and where certain likely developments in the market are also taken into account.” This is said to be appropriate where market developments would with sufficient likelihood have occurred independent of the conduct. In such cases, given the “difficulty to develop credible assumptions, it is not necessary to account for all possible changes and combinations of outcomes and circumstances that could have arisen absent the conduct. It is sufficient to establish a plausible outcome amongst various possible outcomes. In any event, such comparison may not be required in particular where the conduct of the undertaking has made it very difficult or impossible to ascertain the objective causes of observed market developments.” (§67). Again, without a detailed understanding of the Google Shopping case, it is impossible to understand this condensate of confusing language that can cover myriad unspecified situations. It does not satisfy the requirement of legal certainty, nor is it justified as a generally applicable criterion for the purpose of the capability assessment.

A single substantive legal standard applies to establish the capability to produce exclusionary effects on both dominated and non-dominated markets. However, in a dominated market, the assessment should duly take into account that competition is already weakened (bias). The capability assessment must take into account all facts and circumstances relevant to the conduct at issue (position of dominant undertaking and its competitors, market condition, share of affected market, position of customers, evidence of exclusionary strategy, market developments) and should aim to establish, on the basis of specific, tangible points of analysis and evidence, that the conduct is at least capable of producing exclusionary effects. For each criterion, the DGL follow the omnipresent “keep all doors open” technique (in French “en même temps”). Example: the higher the share of the affected market, the greater the capability of exclusionary effects. At the same time, even a small share of affected market can be capable of having exclusionary effects, for instance (= not limited to) where the customers or the market segment targeted by the conduct are of strategic importance for entry or expansion.

Finally, the DGL list factors that are not required for the capability assessment—actual harm to competition; establishing the profitability of the conduct at issue; direct consumer harm; showing that affected competitors are as efficient as the dominant undertaking. Somewhat surprisingly, it states that it isn’t necessary to show that the potentially exclusionary conduct is enabled by the dominant position (§74). This goes back to old case law but somehow dissonates with §65 that requires the potentially exclusionary effects to be attributable to the potentially abusive conduct.

Objective justification

The DGL conclude with a short (too short?) section on the objective justification of allegedly abusive potentially exclusionary conduct. To be “objectively justified,” the conduct must be “objectively necessary” (so-called “objective necessity defense”) or produce efficiencies that counterbalance, or even outweigh, the negative effect of the conduct on competition (so-called “efficiency defense”). 
The objective necessity may stem from legitimate commercial considerations such as the placing of out-of-the-ordinary orders or maintaining technical product performance. As to public health, safety or other public interest considerations, however, it is not the dominant undertaking’s task to play compliance watchdog. Measures taken must be proportionate.

The efficiency defense resembles that of Article 101 (3) TFEU. It requires demonstrating that the exclusionary effects are counterbalanced or outweighed by efficiencies also benefiting consumers. There must be a link between the consumer benefits and the potentially exclusionary effects, and the conduct is proportionate.

The burden of proof being on the dominant undertaking, the defense requires a cogent and consistent body of evidence: Insufficient would be vague, general and theoretical claims or claims relying exclusively on the dominant undertaking’s own commercial interests. Needless to say, this standard is higher than that of the capability assessment. And in any event, it is the Commission that decides whether the standard has been met. While on paper the Union courts have the last word, in practice this means five or more years later, and as Article 102 TFEU-cases are among the economically most complex, the discretionary margin granted to the Commission reinforces its shield.

Conclusion

The DGL are methodologically innovative in respect of the structure adopted. In terms of substance, it resembles a “restatement of the case law” rather than an ambitious policy document. In that the DGL appear to stay behind the recent vertical and horizontal reforms, where the texts produced included some bold innovations and novelties. This impression strongly contrasts with the current enforcement reality that characterizes the digital and technological markets.

We should read the DGL against the background of Pablo Ibanez Colomo’s findings (in The New EU Competition Law, 2023) that the Directorate-General for Competition’s enforcement dynamics have evolved from an evidence-based, open-ended “new economic approach” towards a result-driven enforcement style that uses sweeping remedies and commitments to reshape industrial products, services and business models.

It is somewhat regrettable that the Commission on paper hides behind the Union courts, as if there were no policy and enforcement dynamics, while at the same time keeping all doors open to allow these very dynamics to be unleashed behind an appearance of antitrust orthodoxy.

In the opening scenes of Martin Scorsese’s Casino, we hear the narrator explain: “The Casino always wins.” In Article 102 cases, it is the Commission that always wins, at least almost. The DGL are sufficiently broad and vague to allow the DG COMP to play a winning game when intended.

1. “That demonstration, which may entail the use of different analytical templates depending on the type of conduct at issue in a given case, must however be made in the light of all the relevant factual circumstances (...), irrespective of whether they concern the conduct itself, the market(s) in question or the functioning of competition on that or those market(s). 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 (see, to that effect, judgment of 19 January 2023, Unilever Italia Mkt. Operations, C 680/20, EU:C:2023:33, paragraphs 42, 51 and 52 and the case-law cited).” (Case C-333/21, 21 December 2023).↩
2. The not-so-trivial pursuit of defining markets: A brief guide to the EU’s revised Market Definition Notice.↩