How to Fix a Failing Art. 102 TFEU

In mid-2024, the European Commission is planning to present a first draft of new guidelines on exclusionary abuses under art 102 TFEU. This initiative is part of an ‘Article 102 TFEU package’ that at least partially undoes the Commission’s 2008 commitment to a ‘more economic approach’ which presupposed evidence of likely anti-competitive foreclosure and consumer harm for the Commission to investigate conduct and, instead, aims at establishing a ‘workable effects-based approach’ to ensure a speedier enforcement.

Indeed, looking at the Commission’s enforcement record in recent abuse of dominance cases reveals a worrying picture: on average, the Commission’s adversarial enforcement proceedings are much longer today than they used to be, and excessively long proceedings are occurring more and more often – some lasting six, eight, or even ten years. This risks to render art 102 enforcement ineffective altogether: as remedies against anti-competitive conduct may come too late, a dominant undertaking may already have created long-lasting barriers to market entry (especially in highly dynamic markets) and public enforcement loses its deterrent effect. Against this background and to showcase the potential (and limitations) of future guidelines on exclusionary abuses, Heike Schweitzer and I attempt to flag out important cornerstones of the substantive law on abuse of dominance, and of the law of evidence that governs art 102 proceedings before the Commission in our forthcoming article.

Our inquiry into the concept of ‘abuse’ under art 102 TFEU departs from the much-cited Hoffmann-La Roche case, in which the ECJ defined an abuse as conduct that (1) has the effect of hindering the maintenance of the degree of competition in the market or the growth of that competition and (2) does so by recourse to methods different from those which condition ‘normal competition’. The finding of an abuse therefore uncontroversially includes a criterion of effects on competition. However, it also includes a second criterion relating to a deviation from ‘normal competition’ or ‘competition on the merits’.

The effects criterion in art 102 TFEU is broader than the focus of the ‘more economic approach’ on foreclosure and consumer harm: the Union Courts’ jurisprudence is clear in stating that evidence of consumer harm is not requisite for establishing an abuse. Rather, the assessment of a conduct’s adverse effects relates to the ‘competitive structure’. What matters are potential effects on barriers to market entry or expansion by creating competitive disadvantages for (actual or potential) competitors and thereby shielding the dominant undertaking from competition. These effects must not be ‘purely hypothetical’; some sort of ‘capability’ to produce such effects will suffice, however. The unwillingness of the Union Courts to define a clear standard of probability for the effects to occur gives this criterion a certain flexibility and keeps it open to the peculiarities of the market context in question.

Assessing the adverse effects an undertaking’s conduct has on the competitive structure cannot be the only criterion required to find an abuse, as competition ‘may, by definition, lead to the departure from the market or the marginalisation of competitors that are less efficient’. The second criterion to find an abuse, a recourse to methods different from competition on the merits, has different facets, some of which are more heavily normatively loaden than others. An example for the latter is the AEC principle: the insight that the foreclosure of as-efficient competitors is sufficient to find an abuse. The criterion of ‘no competition on the merits’ may gain considerable weight in cases where the dominant undertaking’s conduct cannot be explained as performance-based competition (‘naked exclusion’, see below).

Evidentiary Requirements Regarding Effects and Possibilities for Shortcuts

A particularly dizzy area of the more recent case law are the evidentiary requirements regarding the adverse effects a conduct has on the competitive structure. Given the criminal law nature of art 102 proceedings, the relevant facts must be proven before the Courts ‘beyond any reasonable doubt’. This does not mean, however, that the applicable law of evidence would not provide for certain shortcuts. Absent a clear conceptual framework of the evidentiary law requirements, we propose a categorisation of shortcuts that makes the changes visible that evidentiary law in abuse of dominance cases has undergone in recent years. We consider there to be two basic forms of shortcuts: Factual inferences and legal presumptions. They differ from one another both in the underlying rationale as well as in their consequence.

Factual inferences, on the one hand, follow an empirical rationale. They may be drawn only if there is a strong empirical link between the occurrence of two facts. Given proof for the first fact, factual inferences lead to an abbreviated assessment of evidence to reach the full conviction of the second fact. In our case: adverse effects on the competitive structure may be inferred from a small set of facts (eg a combination of a certain type of conduct, some degree of dominance and a certain degree of market coverage) if the economically informed experience of the Court suggests a high probability of effects. If the dominant undertaking then challenges the plausibility of the inference (eg by conducting its own AEC test), the Commission would be required to engage with the evidence brought forward by the undertaking. Legal presumptions, on the other hand, do not only or primarily follow an empirical rationale, but a normative one instead – frequently a reduction of overall error costs and an increase in legal certainty –, and they result in either an irrebuttable presumption or at least a full reversal of the burden of proof. The original handling of loyalty rebates and exclusive dealing in Hoffmann-La Roche may well qualify as such a legal presumption based on a normative rationale. In its 2017 Intel judgment, the ECJ ultimately converted this legal presumption into a factual inference.

This categorisation (and a general openness to other kinds of analytical shortcuts, too) paves the way for a sketch of the Commission’s evidentiary law leeway. It may, first, rely on a novel generation of factual inferences, based on sound economic insight on the probability of anti-competitive effects in specific situations. Second, it may revive the concept of legal presumptions: not based on kinds of conduct alone, as in Hoffmann-La Roche, but rather for cases in which the dominant undertaking’s conduct lacks any pro-competitive rationale (‘naked exclusion’, sometimes discussed as ‘abuse by object’). Effects may be presumed in cases in which conduct holds ‘no economic interest for a dominant undertaking, except that of eliminating competitors’, but also in cases in which the possible harm to competition is clearly out of proportion to some relatively minor pro-competitive benefit. Other analytical shortcuts that do not qualify as factual inferences or legal presumptions may include a lowered evidentiary threshold in cases in which a dominant undertaking benefits from a statutory monopoly or comparable exclusive rights, or in which the allegedly abusive conduct is illegal anyways on non-competition law grounds, eg if it consists in handling personal data illegally under the GDPR.

In light of all this, we consider the Commission’s initiative to adopt guidelines on exclusionary abuses to come at the right moment in time. These guidelines are an important possibility to reduce uncertainties in the substantive interpretation of art 102 TFEU and to set up a coherent evidentiary law framework that ensures speedier enforcement. This may contribute to a new vitality of art 102 TFEU to protect competition.

The complete article can be accessed here.

Simon de Ridder is a doctoral researcher at the Faculty of Law at Humboldt-Universität zu Berlin.

Heike Schweitzer holds a Chair for Private Law and Competition Law and Economics at Humboldt-Universität zu Berlin.