AG Rantos’s Opinion in Case C-377/20, Servizio Elettrico Nazionale: a clean framework capturing the essence of the case law (I)
Advocate General Rantos’s Opinion Case C-377/20, Servizio Elettrico Nazionale, was published yesterday (see here for the French version). It is notable for two reasons. First, it effectively captures the essence of the case law following the contributions made by the Court from Post Danmark I to Generics. Second, it creates a framework that brings together all the pieces in a way that completes the picture and addresses some misunderstandings.
The Opinion is also a reminder that the pace of the law is not the pace of policy. The law moves in an incremental way that has little to do with the swings (occasionally dramatic) in enforcement. As Advocate General Kokott once memorably put it, the law, as interpreted by the Court, is not driven by the zeitgeist, but by a more stable undercurrent.
The questions asked by the Consiglio di Stato provide the Court with a great opportunity to engage with some questions that had not been expressly addressed so far. The Opinion is incredibly rich and I do not feel I would do it justice in a single post.
Thus, I will start my discussion with the question of whether an abuse of a dominant position involves an element of ‘impropriety’ or ‘abnormality’. Are practices caught by Article 102 TFEU inherently against competition on the merits? Can ‘normal’, ‘widespread’ or ‘commonplace’ conduct be prohibited as abusive?
Advocate General Rantos suggests an elegant answer, and one that is consistent with the case law. Whether a practice departs from competition on the merits and is qualified as abusive depends on a number of considerations, which vary based on the circumstances of each case and its peculiarities. One size does not fit all.
At one end of the spectrum (see para 62), there are practices that are inherently against competition on the merits, in the sense that they can only be rationalised as a means to exclude rivals (pricing below average variable cost is the perfect example in this sense; Advocate General Rantos adds Lithuanian Railways, in which the anticompetitive object of the conduct is also apparent).
Most potentially abusive practices, in any event, can be rationalised on pro-competitive grounds (and thus can be explained for reasons other than exclusion). They are, in that sense, ‘normal’. In such circumstances, the analysis of the anticompetitive effects becomes the central consideration to determine the legality of the behaviour. By the same token, the question of whether the said behaviour departs from competition on the merits and that of whether it has anticompetitive effects collapse into one and the same issue.
Concerning the analysis of anticompetitive effects (which is understood to mean ‘foreclosure’, or éviction), the Opinion addresses a number of important points (some of which will be discussed in a second post).
First, Advocate General Rantos confirms the principle according to which Article 102 TFEU is about equally efficient rivals. In this sense, the Opinion explains that, absent other circumstances, a practice that can be replicated by an equally efficient competitor is not abusive (para 69). This principle applies irrespective of whether the practice is price-based (say, a set of standardised rebates) or not (say, a refusal to deal).
Thus, there is a difference between ‘foreclosure’ and ‘anticompetitive foreclosure’. This point reflects the idea, introduced in Post Danmark I, whereby ‘not every exclusionary effect is necessarily detrimental to competition‘.
In the same vein, the ‘as efficient competitor test’ is considered to be an expression of that principle (but different from it and thus not to be conflated with it). There are circumstances in which the ‘as efficient competitor test’ is wholly inappropriate (as is true of the specific circumstances of Servizio Elettrico Nazionale or of Post Danmark II).
Second, a finding of abuse presupposes that the practice is capable of having anticompetitive effects, which is a point made explicit in Generics. It has long been clear that there is no such thing as a per se infringement in the EU legal order (para 55). We know from the case law, however, that effects need not be established by the authority or claimant in relation to all practices (AG Rantos mentions loyalty rebates in para 54; pricing below average variable costs is another example).
Third, substance trumps form (para 55). This is a theme that cuts across the case law in EU competition law (and, I would add, EU law at large). From the notions of undertaking and agreement to that of abuse, formal considerations are never decisive from a legal standpoint. The underlying substantive aspects are what truly matters.
Fourth, the evaluation of the anticompetitive effects of potentially abusive conduct is not carried out in the abstract and is not purely hypothetical. It must consider the relevant economic and legal context of which the practice is a part.
In fact, it is sufficient to read the application of the principles to the facts of the case (paras 75-81) to realise how far the analysis goes. The Court invites the Consiglio di Stato to consider, in particular, the following elements:
- (i) the importance and extent of the competitive advantage afforded by the practice (the greater the competitive advantage, the more likely the anticompetitive effects).
- (ii) whether the advantage can be replicated by rivals, in the sense that they can still compete effectively with the dominant firm.
- (iii) the existence of a causal link between the practice and the anticompetitive effects: thus, if the competitive advantages can be effectively replicated, any actual or potential effects would not be attributable to the practice).
The nature of the analysis is entirely consistent with that introduced in cases like Deutsche Telekom, TeliaSonera and Post Danmark I and II. It is also consistent with that conducted in the context of Article 101 TFEU (Delimitis springs to mind when reading the Opinion) and merger control.
The above said, the framework proposed by the Advocate General is invaluable, as there are still few examples in the case law in which the issue is presented in such a thorough manner.
Chillin'Competition 
Hi Pablo. It is indeed “interesting”, even if the facts of the case probably did not require so much sophistry. I find it odd that as example of “judicial authority” for his “replicability” approach outside price cases, he (exclusively) mentions a refusal to supply case (§71) (not surprisingly, there is indeed no judicial authority for that). It is on that basis that the rest of his reasoning (including his conclusion) follows (the replicability is just one of the five elements mentioned in 58-71, but as of 72 it appears to have a very prominent role). The standard for refusal to supply is clearly different from the standard applicable in the “ordinary” cases, so the (judicial) basis for his new approach appears to be rather weak. I thought that the ECJ/GC had so far rejected attempts to rely on Bronner outside the confines of Bronner, but here we go again. This smells of the umpteenth attempt to apply “Bronner type” of reasoning outside Bronner.
It is unclear what he makes of the principle that the mere fact that other non-dominant companies also follow the same conduct (or may do it) does not as such means that the conduct is not abusive (§59), and how he can reconcile both approaches. Unless my text search function is wrong, the “special responsibility” is mentioned last time in 67, and it is nowhere to be seen later in the judgment. It’s gone.
Moreover, the Opinion considers whether the means used (the merits) are lawful only after examining what the effects are (see ends of 79 and 81, for example). If the effect is not anticompetitive then the means are lawful. One would think that the case law examines first the means to achieve the effect, not the other way round. If the means are on the merits, the effects (including any exclusion) do not matter, as the conduct will be lawful. That is what the AG does in the cases where he claims replicability approach does not apply, see note 52 (at the end it is not clear when the new approach will apply, by the way, when reading that footnote). The AG proposal of reversing the traditional order of the analysis (except when he decides that another order applies, see note 52) is unusual, to say the least. Yes, defining competition on the merits as a preliminary point is not easy (and certainly in some cases, such as price-related ones, a more nuanced analysis is needed), but I am not sure the effects of the conduct are decisive on that point, in particular when the intellectual construct is allegedly based on Bronner.
We will see if a chamber of five judges feels confident enough to develop general standards that effectively change the case law on these points, or they rather focus on the facts of the case. I am also surprised by the extensive reliance on the Priority Guidance (it is only in a footnote, very late, that he acknowledges that the Guidance only concerns priorities and does not really bind anyone), a document that does not appear to be effectively used much (and whose importance has been downplayed in prior judgments). To be followed.
Joan
10 December 2021 at 7:11 pm
Thanks, Joan
I agree with you that, where the means are found to be anticompetitive, a prima facie finding of abuse does not necessitate a finding of exclusionary effects (AKZO paras 71-72 are the clear example, if only because subsequent case law like Wanadoo expressly rejected the need to assess the effects of predatory pricing). AstraZeneca, mentioned in fn 52, is another great example in this sense.
On the other hand, I am not sure I agree with your reading of the Opinion. I do not believe it denies the abovementioned case law; in fact, it mentions it is a factor of the analysis.
It just so happens that the Advocate General does not believe that the object of the behaviour at stake in the case is anticompetitive. It concludes that there is nothing abnormal in trying to keep one’s competitors.
This is the background against which I read paras 79-81.
Pablo Ibanez Colomo
10 December 2021 at 8:01 pm
Joan’s reading of the Opion triggers some thoughts:
1) First, Joan says “ne would think that the case law examines first the means to achieve the effect, not the other way round. If the means are on the merits, the effects (including any exclusion) do not matter, as the conduct will be lawful”. I see that Pablo agrees, but do you two believe that this is consistent with paras. 195-196 of the Google Shopping Judgment?
In those recitals the Court essentially says (i) that the Commission inferred “abnormality” from effects; (ii) hthat this may be wrong (“could give rise to queries as to whether the test which the Commission used to characterize the infringement of Article 102 TFEU is consistent with the principle of legal certainty”), but (iii) that the conduct at issue was “abnormal” because in its “particular context” it was capable of diverting traffic.
2) I was intrigued by Joan’s comment that according to the Opinion, the “Priority Guidance””does not really bind anyone”. I went to read the actual footnote (108) and it only says that the Guidance Paper is not binding on national competition authorities and national courts (how could it?). It does not say that the Guidance Paper does not bind its own author (in the light of the case law, how could it not?)
Maguerit
13 December 2021 at 2:38 pm
Thanks so much, Maguerit! Fascinating points.
On the first one, I am also somewhat intrigued by Joan’s interpretation of the case law. I am inclined to say his is a relatively restrictive reading of the scope of Article 102 TFEU. This provision captures more practices than he seems to suggest.
Many (if not most) practices may be deemed abusive even when the means are not anticompetitive. The below-cost prices at stake in Post Danmark I or the standardised rebates in Post Danmak II are good examples.
The Court ruled that these two practices are only abusive where they are found to have actual or potential exclusionary effects. In other words, the effects, not the means, were deemed relevant. In Post Danmark I, the Court was explicit about the fact that the means were not anticompetitive (in a reply to the Danish government submission).
Google Shopping also suggests that the behaviour departs from competition on the merits insofar as it is likely to have anticompetitive effects.
On the Guidance: I agree. Advocate General Rantos mentions that it is not binding on national courts or authorities (wholly uncontroversial).
The question that has not been raised before the Court is that of how, and to what extent, the Guidance binds the Commission (for instance: is there a duty to justify why the Guidance has not been followed in a particular instance?).
Pablo Ibanez Colomo
13 December 2021 at 4:44 pm
Pablo, it’s funny you say I have a restrictive reading of the scope of Article 102. That’s what I always thought about you ;-). Let’s keep the intrigue.
On Guidance, the AG uses it to interpret the substantive law. Not just “priorities”. It is clear that it is not binding on anyone on the substantive law. I thought that was fairly uncontroversial. It is in the very title of the Guidance. See also eg T-827/14, 114: “that communication states that its sole purpose is to set out the enforcement priorities that will guide the Commission in applying Article 102 TFEU to exclusionary conduct by dominant undertakings, and not to constitute a statement of the law (see points 2 and 3 of the communication).” I hope this ends the discussion on that obvious point really.
As an aside: Pablo, have locked up Alfonso in your basement?
Joan
13 December 2021 at 9:40 pm
Joan: if you believe London-based academics live in houses with basements, your perception of the property market over here is not particularly realistic… you will have to look for alternative explanations
Pablo Ibanez Colomo
14 December 2021 at 10:41 am
[…] https://chillingcompetition.com/2021/12/10/ag-rantoss-opinion-in-case-c-377-20-servizio-elettrico-na… […]
My antitrust reading list for 2021 – ICE ON THE MERITS
30 December 2021 at 2:21 pm
Hi Pablo
Thank you for another great post. I would be interested in your views on the meaning of anti-competitive effects with regards to foreclosure: are they limited to only the foreclosure of as efficient competitors? Or can the foreclosure of less efficient competitors also produce anti-competitive effects? Put another way, the ECJ in Intel stated that competition on the merits may lead to the departure of less efficient competitors – but does that mean that the foreclosure of less efficient competitors cannot ever be anti-competitive?
As you say, on the narrow question of the use of an AEC test, it is not mandatory and there are markets where it will not be appropriate (e.g. Post Danmark II). But I am more interested in the wider principle of the meaning of anti-competitive effects: in Post Danmark II the court held that the presence of a *less efficient* competitor might contribute to intensifying the competitive pressure on that market and, therefore, to exerting a constraint on the conduct of the dominant undertaking.
So as a general principle of competition law can we say that, to establish an abuse it is sufficient to demonstrate that there is an anticompetitive effect which may potentially exclude competitors that are at least as efficient as the dominant undertaking (with a limited exception for Post Danmark II-type scenarios where the foreclosure of less efficient competitors may still be anti-competitive)?
Many thanks
Charlie
Charlie
25 February 2022 at 8:11 pm
[…] For instance, Ibanez Colomo is of the view that, unlike the Opinion of AG Rantos (which he considers as a clean framework capturing the essence of the case law), the judgment of the Court seems to be […]
The judgment of the Court of Justice in Servizio (C-377/20) – Part I – The Platform Law Blog
23 May 2022 at 3:58 pm