Chillin'Competition

Yesterday’s post addressed the way in which the Servier saga (see in particular Case C‑176/19 P and Case C‑151/19 P) refined and clarified the interpretation of what amounts to a ‘by object’ infringement.

The saga also sheds light on the analytical framework that applies to the assessment of anticompetitive effects under Article 101(1) TFEU (the ‘by effect’ stage, if one prefers).

First, the judgments (in particular the one in Case C‑151/19 P) confirm that the divide between actual and potential effects refers to the temporal dimension of the analysis. In other words, the notion of actual has to do with with the observable impact of a practice (what actually occurred). Contrary to what is sometimes suggested, this term does not imply a higher threshold (relative to ‘potential effects’).

The threshold of effects remains the same, whether we consider actual or potential effects. What changes (and paras 313-335 in Case C‑151/19 P are a good illustration) is that in the latter case the analysis is prospective (potential) and in the former it is retrospective (actual).

Second, actual effects may be taken into consideration in the overall assessment even when the analysis is prospective. As the Court put in in para 321 in Case C‑151/19 P: ‘events subsequent to the conclusion of that agreement may be taken into account in order to assess that situation‘.

When the actual operation of the market can be observed, this evidence may shed light on the potential of the practice to harm competition (even if it is not conclusive). The Court’s position is also in line with that expressed in Servizio Elettrico Nazionale (para 54 of Case C-377/20).

Third, the purpose of the counterfactual is to determine whether there is a causal link between the practice and any actual or potential effects (para 317 in Case C‑151/19 P: ‘The purpose of that “counterfactual” method is to identify, in the context of the application of Article 101(1) TFEU, the existence of a causal link between, on the one hand, an agreement between undertakings and, on the other, the structure or functioning of competition on the market within which that agreement produces its effects […]’).

In other words, it is only possible to conclude that effects are attributable to a given behaviour by measuring them against the relevant countefactual (that is, the ‘but for’ scenario revealing how the market would have evolved in the absence of the practice).

As the Court puts it, the counterfactual ‘[…] makes it possible to ensure that characterisation as a restriction of competition by effect is reserved for agreements displaying not a mere correlation to a deterioration in the competitive situation of that market, but for those agreements that are the cause of that deterioration‘ (ibid; emphasis added).

The points above are not strictly new, but are valuable, to begin with, because of the structure the Court provides and the careful drafting. It is a clear and effective articulation of the role of the counterfactual in the analysis of effects and in establishing causality.

It is also valuable in that it is line with Advocate General Kokott’s Opinion in Google Shopping. As explained here, she pointed out (para 172 of the Opinion) that the issue of the counterfactual must not be conflated with the temporal dimension of the analysis.

The Servier saga proves Advocate General Kokott’s point. It illustrates, in concrete terms, that the counterfactual is also relevant when evaluating whether the alleged potential effects of a practice are indeed attributable to it.