On 10 November 2022, the European Court of Justice (ECJ or the Court) delivered its judgment in another private enforcement case, interpreting Art. 5 of Directive 2014/104/EU (the Damages Directive or the Directive). It answered the preliminary question raised by a Spanish First Instance Commercial Court that the disclosure of information or documents which do not yet exist and must be produced by the defendant in cartel damages proceedings ex-novo is not per se precluded by the Directive. It is subject to a proportionality test to be performed by the referring court.
The second chamber (reporting judge: Nils Wahl) widely followed AG Szpunar’s opinion, delivered earlier this year. The judgment quotes the AG’s opinion on multiple occasions. Whereas the facts of the case and the legal reasoning of the AG have already been presented in great detail by this blogpost of Alvaro López Usatorre and Lena Hornkohl, the following analysis aims to illustrate and analyse the Court’s reasoning and its consequences for pending and upcoming cases before national courts which can be asked to order the disclosure of evidence in private enforcement proceedings.
Besides clarifying the temporal applicability of Art. 5 of the Directive, the focal point of this judgment is the literal, contextual and teleological interpretation of Art. 5 (1) of the Damages Directive. Finally, the practical significance and implications for further damages proceedings before national courts and tribunals are summarized in this blog post.
Applicability ratione temporis – once again
The applicability of the provisions of the Damages Directive ratione temporis has been discussed prominently in the course of the Volvo/DAF case (C-267/20). Whilst in Volvo/DAF, the reasoning of the Court was not always straightforward, it is entirely convincing in the present case.
Quoting both the judgment in Volvo/DAF and the AG’s opinion in the present case, the Court left no doubt that Art. 5 of the Directive is a procedural provision in the meaning of Art. 22 (2) of the Directive (cf. paras 33-35 of the judgment).
Noteworthy is not the result but the reasoning which clarifies to some extent what was said in Volvo/DAF. According to the Court, Art. 5 (1) is not a substantive provision because it does “not impose new substantive obligations” (cf. para 34). Although there is an obligation of the defendant under Art. 5 (1) to disclose evidence, this obligation is of procedural nature since it does not alter the “constituent elements of non-contractual liability” (para 33). This is a generalizable characterization of the law of evidence. Although the Court in para 34 quotes the Volvo/DAF judgment, the convincing and clear finding in the present case is partly in conflict with the findings of the first chamber in Volvo/DAF which characterized the presumption of harm (Art. 17(2) of the Damages Directive) as a substantive norm regardless the fact that the constituent elements of liability were not altered by that presumption either. The Court now follows a literal and systematic interpretation of Art. 22 of the Directive without being influenced too much by the importance that the classification of a rule as substantive or procedural can often have on the outcome of a case. This is a welcome development not only for the rights of claimants but also for legal certainty and predictability with regard to other provisions of the Directive.
The scope of Art. 5 (1) – Production of ex-novo evidence not excluded
Coming to the core of the legal issue at hand, the ECJ’s judgment contains a persuasive well-structured argument for a broad interpretation of Art. 5(1) of the Directive. Starting off from a literal analysis (paras 41-49), this is complemented by a contextual interpretation of the Directive (paras 50-53 and 63-64) and lastly examined with regard to the purpose of the norm (paras 54-62).
The reason why the referring court had doubts whether it could order the disclosure of information that would still need to be prepared by the defendants for the purpose of the proceedings is that Art. 5 (1), first sentence, refers to “evidence which lies in [the defendants’] control”. One could argue that only existing evidence is ‘in the control’ of the defendant.
However, already from a literal interpretation, this view can be questioned. The decisive element is the notion of ‘evidence’. As the Court rightfully points out, it is defined in Art. 2 (13) of the Directive. According to Art. 2 (13), that term covers “all types of means of proof admissible before the national court seized, in particular documents and all other objects containing information, irrespective of the medium on which the information is stored”. In other words, everything that can procedurally be accepted as evidence is covered by the term in the directive. The definition could barely be broader.
Thereunder, also data which is not (yet) physically available can be evidence. As the ECJ puts it, the “provision does not necessarily correspond to pre-existing ‘documents’”. (para 41) Nonetheless, the difference between the wording of the first and second sentences of Art. 5 (1) could raise some confusion against this clear and straightforward definition.
According to the ECJ (para 45), following the AG’s opinion, the additional words “in the control” only constitute a factual observation illustrating the perceived information asymmetry. Any other reading of the norm would furthermore not be consistent with Art. 2 (13) of the Directive anymore. It is an unnecessary clarification that the evidence (in the broad sense of Art. 2 (13)) must be in the control of the defendant. Otherwise, it would be impossible for them to disclose it. A disclosure order concerning documents that one does not possess and cannot produce would be clearly unenforceable and thus pointless. However, and this is what the ECJ distinguishes, the control over the evidence does not mean that it is already processed in an accessible document that could simply be handed over.
That the defendants in the present case have the necessary raw data and information at hand which is needed for the disclosure request is not seriously disputed. Therefore, the preliminary question could already be answered in favour of disclosure on the basis of the literal analysis.
For the sake of completeness and a more persuasive answer, the Court holds that these findings are supported by a contextual reading of the norm. Whilst Art. 2 (13) mentioned above merely contains the definition of evidence and therefore was already relevant for the literal analysis, the Court could now draw from a reading of Art. 5 (2) and (3) as well as the relevant recitals of the Directive which are equally informative.
In particular, Art. 5 (3)(b) supports the findings of the ECJ. As the Court correctly analyses the norm, making reference to the proportionality of costs “presupposes, implicitly but necessarily, that the cost of disclosing evidence may, where appropriate, significantly exceed that corresponding to the mere transmission of physical media, in particular documents, in the control of the defendant” (cf. para 53).
Last but not least, the ECJ checks its previous findings against the compatibility with the purpose of the Directive. In this last section on the substance, the Court also touches upon the general role of private enforcement as opposed to public enforcement and the place for the damages directive on the general field of competition law enforcement.
According to the ECJ, making reference to Recital 6 of the Directive, the primary objective of the Directive is to facilitate the enforcement of Art. 101 and 102 TFEU (cf. paras 55 & 62).
Having the principle of equality of arms in civil proceedings in mind (cf. also para 47 of the present judgment), the Damages Directive aims at “a balancing of the legitimate interests of all parties” (para 57).
Applied to the present case, the Court holds that a limited disclosure, only of “unprocessed, pre-existing and possibly very numerous documents” would unduly restrict the right to full compensation (para 61). Without saying it explicitly, the ECJ thereby emphasizes (once again) the effet utile of Art. 101 TFEU, which is safeguarded by rigorous enforcement of violations.
Such a broad interpretation of disclosure of evidence in the sense of Art. 5 (1) of the Directive does not necessarily lead to an imbalance of the legitimate interests of the parties and therefore is entirely consistent. A balanced and fair outcome for each individual case can and should be reached by a cumulative application of the precision test in Art. 5 (2) and the proportionality test in Art. 5 (3) which the ECJ already used as an argument for the contextual analysis and reiterates in the final reasoning (para 64). A balancing of interests with regard to all particularities of a specific case, however, is only possible if no type of evidence is per se excluded from disclosure. Therefore, the ruling of the ECJ leads to higher chances of accomplishing the balancing exercise that the Directive aims at.
The role of national courts and tribunals – a case for the principle of proportionality
Accordingly, the burden of the application to the concrete case rests on the national court. The answer to the preliminary question is clear. The production of documents ex-novo for use as evidence in damages proceedings is not precluded by Art. 5(1) of the Directive.
Whether the applicants finally succeed with their disclosure request is nonetheless uncertain. As the ECJ emphasized, the proportionality test is not a mere formality but a true balancing exercise for the referring court. It will need to examine the concreteness and precision of the request under Art. 5 (2) of the Directive and analyse the interests of the parties under Art. 5 (3). This focus by the ECJ on the principle of proportionality was not at all imperative. It seems to be an attempt, on the one hand, to answer some doubts about the extent of disclosure, which were expressed by the referring court. On the other hand, the ECJ thereby tries to balance the interests of claimants and defendants in the damages proceedings instead of getting itself into an all too claimant-friendly position. After all, the ECJ jurisprudence is already (and still) one of the driving forces and an impellent of private enforcement of competition law.
Whilst the previous commentators on this blog, with reference to the AG’s opinion, already pondered a ‘general principle of disclosure’ as part of the effet utile principle, there is no reference or argument to that regard conceivable in the final judgment. Instead, the repeated accentuation of the principle of proportionality rather suggests a certain vigilance by the Court – without being expressly mentioned – to prevent the criticism of an ‘Americanisation’ of the disclosure rules in the EU.
Possible interests of the claimants could include a certain probability that substantial pieces of evidence are held by the defendants. An information asymmetry in favour of the defendants is generally presumed by the legislator (cf. Recital 15 of the Damages Directive). In principle, the asymmetry decreases in follow-on damages proceedings as more information is accessible to the claimants in the first place. How significant the information asymmetry is in a concrete case is difficult to assess but a possible indicator is the level of detail in the information accessible from the authority’s fining decision and possibly public subsequent legal proceedings.
On the side of the defendants, the (legitimate) interest against the production of pieces of evidence could entail additional costs for the necessary workforce, but also issues of confidentiality or the inseparability of some pieces of requested data or evidence, e.g. the use of a database that would include products which are relevant for the damages proceedings at hand but also information about other products with no direct relevance for the case. Of course, further arguments or scenarios are conceivable.
As the ECJ in its final reasoning in the present judgment clarifies, “the interpretation of the first subparagraph of Article 5(1) of Directive 2014/104 cannot result in defendants to main proceedings taking the place of claimants to main proceedings in their task of demonstrating the existence and scope of damage suffered”. However, this leaves a wide range for judicial discretion and appraisal of facts for the national courts. It just already highlights one of the likely arguments that will be raised by defendants in upcoming damages proceedings.
The ECJ further states that this “reasoning applies, a fortiori, to proceedings in which no penalty has been previously imposed by the Commission or by a national competition authority for unlawful conduct”, i.e. stand-alone actions. As such stand-alone actions are not very common due to the difficult gathering of evidence, and the relevance might be negligible. However, it was already noticed by commentators that this could lead to a more restrictive handling of stand-alone cases. That is not necessarily meant by the short additional phrase in the judgment. It could as well be interpreted in the line of argument presented above that the balancing of interest is de facto influenced by a preceding public enforcement procedure. In follow-on cases, claimants do not have to prove again the infringement. Besides, they always have prima facie evidence on their side that the defendant has incriminating documents or data in their possession which have been sufficient for public enforcement and would likely also be relevant for the assessment of damages. It is simply consequential from these factually different circumstances that the balancing of interests could lead to different outcomes in stand-alone cases – not because of different legal standards but due to a different amount of indicative evidence that applicants typically have in their hands when requesting a disclosure order.
The arguments for and against disclosure will be brought to the attention of the national court by the parties. It is evident that a purposeful balancing of interests is only possible if these alleged interests of the parties are substantiated. As the balancing exercise for the national court cannot be performed with mathematical precision, it will typically not require the taking of evidence for this purpose, but a check of the plausibility of the argument brought before the national court or tribunal.