Already in the early 1960’s, the Court of Justice made clear that the application of competition law depends on contextual analysis that takes a wide range of economic and legal factors into account. Modern economics provides useful tools to deal with competition matters. The European Commission increasingly relies on these ‘mainstream’ economics in its assessment of competition cases, and courts have to make up their own mind on the merits of the Commission’s complex assessments and of the economic concepts on which the Commission relied to that effect.

What kind of judicial control are the Union courts supposed to exercise over these complex assessments?  Under the current system set up by Article 263 TFEU, judicial review by the General Court, which has the final say on the interpretation of the facts of the case, is limited to the review of the legality of the Commission’s decision. In its case law, the Court of Justice has traditionally used formulae that suggest a certain degree of deference vis-à-vis the Commission’s assessment.


The use of the manifest error test has been criticised as a sign that the General Court would be unwilling to scrutinise the Commission’s assessments and, hence, that it would be too deferential toward the Commission. This paper, available here, seeks to identify what the manifest error test is supposed to mean, and argues that this test is compatible with full judicial review in cases involving complex economic assessments. It does so as follows:

Section II reviews the principles governing legality review under European law.

Competition infringement decisions are subject to the system of administrative review applicable to EU acts more generally. Under this regime, the General Court has jurisdiction to review these acts on grounds of lack of competence, infringement of an essential procedural requirement, infringement of the Treaties or of any rule of law relating to their application, or misuse of powers. Legality review is limited to assessing the legality of the administrative conduct under scrutiny and does not allow the review court to assess its expediency, nor to substitute its own views to those of the administration. As a rule, legality is assessed only on the basis of the facts and evidence of which the administration was aware when it took its decision. Furthermore, legality review is linked to the pleas and arguments that the parties put forward in their application contesting the decision.

Unlimited jurisdiction allowing allows the Union courts to substitute the Commission’s decision for their own is restricted, as regards competition cases, to fines and penalty payments.

Section III discusses the nature of complex economic assessments.

Assessments are characterised by a degree of uncertainty that facts do not have. The larger the degree of uncertainty, the more speculative the administration’s assessment becomes. As a rule, review of such assessments should not make the life of the administration impossible: courts cannot require certainty where it does not and cannot exist.

It follows that evaluating the legality of assessments presupposes a certain degree of leeway for the administration. In this context, the question arises as to the degree of uncertainty which society is prepared to accept. A higher degree of uncertainty may be tolerable when the administration makes prospective assessments (e.g. on the effects of mergers) or engages in regulatory tasks (e.g. when controlling State aid). By contrast, society is less likely to accept uncertainty in the repressive sphere, i.e. in situations where public authorities threaten to impose heavy sanctions, as is common in competition infringement decisions.

As concerns competition law, the most complex assessment is undoubtedly whether a conduct restricts competition caused by the conduct under scrutiny. This question can rarely be answered in the abstract, but will most often require an in-depth contextual analysis. However, the need for contextual analysis is not necessarily a distinctive criterion to identify complex economic assessments. The application of legal concepts – such as the notion of undertaking, the presence of collusion, the effect on trade or the impact of a conduct in the internal market – also require a contextual analysis, albeit of a non-economic nature. However, the case law does not tolerate any degree of uncertainty as to the interpretation of these legal concepts, which are therefore subject to the Courts’ full review. In this regard, it is important to distinguish between legal concepts, on the one hand, and more economic concepts, on the other.

Section IV discusses the legal formulae applicable to complex economic assessments.

The author identifies two formulas concerning the intensity of the Courts’ review in cases involving complex economic assessments.

  • The first formula focuses on manifest error, and sets forth that: ‘The review by the European Union judicature of the complex economic assessments made by the Commission is necessarily limited and confined to verifying whether the rules on procedure and on the statement of reasons have been complied with, whether the facts have been accurately stated and whether there has been any manifest error of assessment or misuse of powers’ (from GSk).
  • The second one reads as follows: ‘Not only must the Community Courts, inter alia, establish whether the evidence relied on is factually accurate, reliable and consistent but also whether that evidence contains all the information which must be taken into account in order to assess a complex situation and whether it is capable of substantiating the conclusions drawn from it’ (from Tetra Laval).

The latter formula recognises that the Commission enjoys a measure of discretion, especially with respect to assessments of an economic nature, and that the courts must take account of that margin of discretion when reviewing the Commission’s assessments. However, this margin of discretion does not mean that the Union courts must refrain from reviewing the Commission’s interpretation of information of an economic nature. According to the Tetra Laval formula, the courts’ task is threefold. They must check whether the facts are correct, whether all relevant facts have been collected, and whether this evidence is sufficiently robust to substantiate the legal conclusions which the administration draws from it.

What both formulas have in common is that the Union courts are not allowed to substitute their own economic assessment for that of the Commission. However, this does not mean that the review courts should not carry out an in-depth review of the Commission’s decision and the underlying choices.

Section V looks into how the same formulae are applied depending on the contest.

EU courts tend to use the same formulae in a wide range of situations. This indiscriminate use of formulae does not imply, however, that the legality review is carried out in the same manner in the various areas of EU competition law. On the contrary, the intensity of review differs considerably according to the rules at stake.

Review is more intense when antitrust infringements are at stake. When enforcing the EU’s competition provisions, the Commission investigates, prosecutes, decides and sanctions. Moreover, under this system the Commission has consistently increased the level of fines over the years. At the same time, the EU courts (and the European Court of Human Rights) have increasingly imposed due process requirements on competition investigations, reflecting their quasi-criminal nature. In the exercise of their judicial review powers, the Union Courts attach considerable importance to the burden of proof, which under Article 2 of Regulation 1/2003 rests upon the Commission. This goes against the ‘normal’ system of legality review, where the applicant bears the burden to prove that the contested decision is incorrect. It therefore suffices for the applicant to cast doubt in the judge’s mind as to whether or not the Commission correctly acquitted itself from that task. In other words, and as noted in ICAP, the applicant companies benefit from a presumption of innocence: it is up to the Commission to prove the infringement by putting forward firm, precise and consistent evidence. Where the Court still has doubts, the benefit of that doubt must be given to the undertakings accused of the infringement.

As regards penalties, the competent courts is allowed to substitute the Commission’s decision with their own. The law reflects not only Strasbourg case law – particularly the Menarini decision – but also the special position which firms have when they are fined by the Commission. In light of this special position, these firms are allowed – against the general principles governing legality review – to raise arguments and rely on facts that they did not mention during the administrative procedure, and to put forward evidence which the Commission did not have when it adopted the contested decision.

Sections VI and VII consider the practical implications of this.

The ECJ’s case law on the presumption of innocence has practical implications for the review of complex economic assessments in the context of infringement decisions. The Commission is free to choose the theory of harm and the corresponding models, parameters and evidence when making such assessments. Indeed, according to the manifest error test it is not for the General Court to substitute its views for those of the Commission. However, this ‘deference’ does not mean that the General Court will not carry out an in depth review of the Commission’s assessment. The court will not only review that assessment in light of the three-pronged Tetra Laval test, but will also determine whether the answer to these questions is sufficiently convincing to eliminate any doubts which the General Court could have as regards the Commission’s finding. It follows that, when the court is in the presence of two equally plausible assessments of a complex economic situation, the assessment put forward by the ‘applicant’ prevails. Moreover, even if the Commission’s assessment is more plausible than the one presented by the applicant, it will still not suffice to establish the existence of an infringement if the evidence is not sufficiently convincing and consistent to dispel the doubts which could still exist as regards the applicant’s involvement in the infringement.

This robust test finds its origin in the presumption of innocence, and does not apply to the other areas of EU competition law, i.e. the rules dealing with the control of concentrations and State aid. Legality review of complex economic assessments in these domains is limited to the application of the Tetra Laval formula without the two adjustments regarding the possibility to submit new evidence during judicial proceedings and the burden of proof.


As someone who wrote a doctoral dissertation that delved deeply into the modes of legal reasoning adopted by the European courts, I understand why these courts adopt generic formulae that are, to a degree, ambiguous. Yet, I still find it striking how often EU judges write academic pieces trying to elucidate and clarify earlier judgments – see, for another instance of this, here.

This paper’s argument is that, despite the European courts having adopted a single general formula to govern complex assessments where the administrative decision-maker (i.e. the European Commission) enjoys a margin of discretion, in reality the courts apply different tests to different types of administrative decisions. In itself, this is a very illuminating and solid paper – but, personally, I do not see any good reason as to why the European courts could not have made that clear in their judgments.

On a more analytical note, I think that the paper could have benefitted from discussing the appropriate standard of proof, instead of just focusing on the allocation of the burden of proof. This is a matter which is implicit in the different ‘standards of review’ that the author identifies for antitrust, on the one hand, and merger control and state aid, on the other. It is true that – in line with Continental evidence rules – the author advances a standard of proof that focuses on the conviction of the judges. However, discussions concerning the presumption of innocence lend themselves naturally to disquisitions of the required level of persuasiveness that the evidence must have to dislodge this presumption. In this respect, the author seems to advance a standard akin to ‘reasonable doubt’ – i.e. the evidence should be sufficiently convincing and consistent to dispel the judge’s doubts – but such a formula does little to disclose what such evidence might be. There is some interesting literature on this, which seems to suggest that the standard varies depending on the type of practice sub judice. This is something I would have liked the author to have engaged with.  It is also something that comes up in the piece below.

Author Socials A weekly email with competition/antitrust updates. All opinions are mine

What do you think?

Note: Your email address will not be published