Judicial review and Article 102 TFEU: undue deference to the Commission? Not so fast
One of the most popular mantras in the EU competition law community is the idea that the Commission ‘always wins’ Article 102 TFEU cases, and that it is all but impossible to see an abuse decision annulled before the EU courts.
I have spent quite a bit of time examining this question (and related ones) lately. I have concluded that there is perhaps a grain of truth in this mantra. As often happens, however, the grain of truth has been distorted so much that it is barely recognisable. Even worse, most claims these days are made without much evidence to back them up.
What do the figures tell us when we look at them?
If you look at the figures about which most people care (challenges brought against Article 102 TFEU decisions and rate of annulment of these decisions) you are likely to be surprised.
You will realise that seeing a Commission decision annulled was actually quite commonplace in the early years (that is, until the late 1980s); much more so, in fact, than conventional wisdom would suggest.
Then, something happened when the General Court was created. It is true that the annulment of Commission decisions became significantly less frequent at the time. This is certainly the origin of the mantra, and the underlying (and possibly only) grain of truth of some narratives.
However, these figures do not mean, in and of themselves, that the EU courts became unduly deferential to the Commission, or that its attitude to judicial review of abuse cases changed. This is the point at which we enter the realm of speculation.
What do the figures fail to tell us?
I keep insisting on this point, but I do not think I can emphasise it enough: the annulment of a Commission decision may be a popular proxy for curial deference, but it is actually a very poor one. The fact that a decision is annulled, or fails to be annulled, does not say anything about the intensity of judicial review.
Why is it that annulments fell when the General Court was created?
The undue deference story is only one possible explanation of the phenomenon. The observed shift in the rate of annulments is also consistent a scenario in which the authority became increasingly risk averse and its decisions became subject to tight scrutiny by the review courts – I am not saying this is what happened, it is just that it is also perfectly plausible.
In fact, if the early years were marked by frequent annulments in Article 102 TFEU cases, it would be reasonable for the Commission to become more risk averse over time, and to pursue only the cases that it is certain to win. With risk aversion, annulments inevitably become less frequent, even if judicial scrutiny is as intense as it can get.
Conversely (and equally importantly), the frequent annulment of Commission decisions is not evidence, in and of itself, that judicial review is intense. Many of the biggest victories for the authority originated in decisions that were annulled.
Just think of Kali & Salz: the individual decision may well have been annulled, but the Commission won on a major issue of principle (what matters for a repeat player): collective dominance was found to fall under the scope of the Merger Regulation.
Bottom-line
The bottom-line of the above is clear: examining the substance of individual decisions is inevitable if one wants to come to meaningful conclusions about whether the annulment (or the absence thereof) is a reliable proxy for deference to the analysis of the Commission.
In particular, it is necessary to develop a way of checking whether the Commission is risk-prone or risk-averse when adopting Article 102 TFEU decisions. This is one question that has kept me nicely busy. Answers? Soon! But I can say that the truth, as far as I can tell, is (surprise, surprise) complex and does not quite fit one particular story or narrative.
The pace of the law
I was thinking, while drafting this post, that some of those who claim that the EU courts are unduly deferential are just impossibly impatient people. These people would like to see the complete overhaul of the EU competition law system from one day to another.
This is not how the law works, and for good reasons. The pace of the law is – and can only be – slow and incremental. It is a mistake to think that courts are deferential simply because their attitude towards legal change takes into account factors beyond the specific case at hand. Predictability and stability also matter.
Legal change may seem slow; eppur si muove: just think of Intel.
And as I write this, I am reminded of the excellent Rebecca Haw Allensworth, who has written (much more) eloquently about these questions. You may want to take a look, for instance, at this paper.
Chillin'Competition 
Good post and I fully agree.
Kinsella, Stephen
11 May 2018 at 12:50 pm
I totally agree. Contrary to the mantra you refer to, I think that Court’s review was largely limited to procedural rather substantive matters given that Commission’s complex economic assessments could be subjected only to limited review (Costen Grundig). In this sense in the early years, the Court was largely supportive of the Commission’s interpretation (Do you have a reference for the figures you cite?). The establishment of CFI, it seems to me, that it triggered an intensification of review (Societa Italian Vetro SpA v Commission; European Night Services; Metropole; Tetra Laval v Commission; Microsoft v. Commission). And gradually we see that both Courts reduce the scope of the complex economic assessment doctrine and exercise an increasingly more meticulous review to such an extent that we may doubt the very existence of the complex econ assessment doctrine (Marc Jaeger, ‘The Standard of Review in Competition Cases Involving Complex Economic Assessments). What do you think?
Stavros (EUI)
11 May 2018 at 2:16 pm
Hi Stavros!
Thanks for your comment, as usual!
On the source for the figures: wait till my book comes out 😉
For the rest: I do not believe I agree. You seem to imply that, in the early days, the EU courts controlled for procedural matters alone. This is not what a careful analysis of the case law shows. Quite the contrary, actually. Do not forget that the Consten-Grundig bit to which you refer concerns Article 101(3) TFEU! And do not forget that there has never ever been limited review on issues of law as such (such as the definition of the legal test).
Pablo Ibanez Colomo
11 May 2018 at 5:23 pm
We have talked a lot about this and we agree. In any case, re-reading the post triggered a thought. Instead of discussing over the phone, as usual, I’ll put this in the open so that others can jump in too:
For the most part, I entirely agree that the Commission did turn risk averse and only pursued cases that it was certain to win. The unclear cases were all addressed via commitments, etc.
At the same time, there are exceptions that confirm the rule. Some particularly important cases were arguably quite risky (e.g. Microsoft), and so are other recent and ongoing 102 cases.
Fans of pendulum theories could posit a reading of the situation whereby (i) the early annulments gave rise to a phase of risk aversion, (ii) the litigation stats and the Commission always wins” 102 cases myth then also also seduced the Commission itself boosting its confidence; (iii) we enter now a new phase of less risk aversion and return back to square one.
Admittedly, another plausible explanation may be simply that the Commission believes very high-profile cases are more likely to be confirmed.
Alfonso Lamadrid
11 May 2018 at 3:08 pm
It has been a while since we last exchanged comments on the blog!
Let me be clear (and perhaps I was not clear enough): the post does not claim that the Commission actually turned risk averse. I was simply mentioning that a risk aversion story could plausibly explain the decline of annulments. I did not write that this is what actually happened.
Once you get around to reading the book (just saying), you will see that the picture, as you suggest, is mixed. Has the Commission occasionally been risk prone in the past two decades? Certainly, is my answer, having looked at the question closely.
Pablo Ibanez Colomo
11 May 2018 at 5:28 pm
one factor which may explain Commission’s 102 track record on appeal, before and after mid/late 1980s, has also to do, in my opinion, with ECJ’s judgment in Michelin I. It was only in 1983 that the ECJ elaborated the notion of special responsibility in 102 proceedings. From that moment on, it was certainly easier to bring cases and justify enforcement, because that concept – coupled with that of competition on the merits and very loose market definitions – served as a Procrustean bed for most if not all claims in a 102 setting.
stefano grassani
14 May 2018 at 5:38 pm
Out of curiosity: have you looked into whether the decentralised enforcement of the competition rules might also contribute to the “Commission always wins”-trend? It seems logical (at least to me) that the greater flexibility in terms of choosing which cases to pursue and ressource allocation could play a role as well.
Charlotte Forno
16 May 2018 at 8:29 am
What a great comment. Thank you!
I have examined the question. Even though we do not have enough cases, my intuition is that it influences the likelihood of deference.
With decentralisation, my impression is that review courts will be marginally less likely to defer to the analysis of the Commission. Why? Under a centralised system, review courts can trust that the expert authority will get it mostly right. Once more authorities and courts start applying the competition law provisions, the errors made by the Commission can have more important and unpredictable consequences. As a result, review courts will be inclined to review more intensely what the Commission does.
This, in fact, is the impression I get from the recent preliminary rulings, where the Court is clearly concerned with consistency and has refined some old doctrines (e.g. Post Danmark I and II). Would you agree?
Pablo Ibanez Colomo
16 May 2018 at 1:29 pm
I do agree with your reading of recent preliminary rulings such as Post Danmark I and II.
Do you think that the Court takes the same approach to preliminary references and actions for annulment when dealing where Article 102 TFEU is concerned? I admit that I haven´t looked at this thoroughly, but it´s not my general impression that judgments rendered by the Court in actions for annulment reflect a concern for consistent application of Article 102 TFEU across the member states.
Charlotte Forno
17 May 2018 at 11:31 am
Hi Charlotte,
We don’t have that many appeals. This said, if there is anything at which the appeal judgment in Intel hints, this is a concern with consistency. In such a brief judgment, the extensive references to Post Danmark I as the ruling setting the general framework are remarkable.
Pablo Ibanez Colomo
17 May 2018 at 1:34 pm
Very interesting post and very interesting discussion. Let us not forget, however, that looking at the Court (or, better, at the EU Courts) as a sort of monolithic and homogenous body that has one and clear policy/approach/attitude is somewhat of a simplification. Obvious truth is that both the CJ and the GC are composed of members with very different backgrounds (previous professional experience, areas of expertise, legal culture etc): that inevitably reflects on their approach as to what judicial review of administrative decisions should be. Much more than what it happens at national level, both in the EU Member states and elsewhere. Add to that that their knowledge of key economic concepts and principles varies (a lot!). So, unfortunately but inevitably, there is a non-negligible level of hazard in how closely (or deferentially) the EU Courts look at specific cases. The GC mostly decides in chambers of 3, the CJ in chambers of 5. The weight of each member of the EU Court sitting in the chamber that is responsible for a case (and, when the case is before the CJ and is decided with an opinion, of the AG in charge) is very significant. It is thus entirely correct to look at trends, to aggregate/combine cases, examine numbers and so forth… but not everything can be logically explained. To use metaphors: some pieces of the puzzle will never fit, or the picture will never have a perfect focus. This should be always kept in mind. Apologies if my post is somewhat too obvious.
Luca
17 May 2018 at 9:44 am
Thanks, Luca!
These are factors that are of fundamental importance! But I guess they raise a different question; given the heterogeneity of the EU courts and their size, how is it that patterns across provisions and over the years are so marked? The observed heterogeneity would suggest much more randomness.
Pablo Ibanez Colomo
17 May 2018 at 1:36 pm
My top off the head answer would be that, even if formally we have not the principle of binding precedent, respect for what is found in past cases (not only the reasoning, but including empty formulas, mere dicta etc) is seen as almost sacred. Even too much, if you ask me. On the one hand, this may be one of the factors explaining the unfortunate reluctance to abandon case-law from the seventies and eighties even when legal and (especially) economic thinking has evolved in another direction. On the other hand, this also explains, at least to come extent, the CRT+C approach that can sometimes be detected in some decisions. Let me explain: a certain statement was reasonable in a given case or context, but then it takes a life of his own and is reproduced in other decisions where it may not be appropriate or correct anymore. So, a certain line of case-law, decision after decision, expands till it is applied even where it is not longer justified. I hope I made myself clear… I typed these thoughts in a rush. I am trying to finalise smng (that you’ll probably read end of July), but still wanted to give you some quick feedback. To be clear: surely the above does not explain it all. There are obviously more serious reasons explaining what you have found (which sounds reasonable to me), and those deserve a thorough investigation …. but, again, I would not discard the (perhaps limited but not-insignificant) contribution to what we could regard, depending on the specific circumstances of each case, as coherence/uniformity/stability/certainty or as rigidity/deference to past cases/superficiality etc that was given by ore trivial factors
Luca
17 May 2018 at 3:33 pm