European Union: Cartels and Leniency

In summary

This article highlights the European Court of Justice’s key findings on the issue of ‘by object’ infringements in the Budapest Bank judgment of 2 April 2020. It explains the core elements that the European Commission and national antitrust authorities must consider when categorising an infringement as ‘by object’ and analyses the new requirements for authorities to review the counterfactual situation and potential pro-competitive effects.

Discussion points

• Clarification of criteria required to categorise ‘by object’ infringements
• Certain areas in the current jurisprudence on ‘by object’ infringements confirmed by the European Court of Justice in the Budapest Bank judgment of 2 April 2020
• New criteria added by the Budapest Bank judgment of 2 April 2020
• More robust analysis that competition authorities will have to carry out to find a ‘by object’ infringement

Referenced in this article

• Article 101 of the Treaty on the Functioning of the European Union
• European Court of Justice
• Advocate General Bobek
• Budapest Bank judgment of 2 April 2020
• Generics judgment of 30 January 2020
• Cartes Bancaires judgment of 11 September 2014
• Maxima Latvija judgment of 26 November 2015

Introduction

In recent years, the European Commission (the Commission) has come under constant criticism for pushing the envelope with the categorisation of conduct nowhere near a hard core cartel as a ‘by object’ infringement, usually to circumvent a more robust analysis when determining whether a restriction exists. To be sure, infringements commonly considered the most serious, such as price-fixing, bid rigging and market or customer allocation, have traditionally been classified as ‘by object’ violations of competition law by the Commission with little work to prove the conduct was an infringement ‘by object’. Since such a classification relieves the Commission from further examining the effects of this type of conduct, the Commission’s evidential burden as backed by the EU courts is considered low, but still requires that the conduct is by its own nature clearly detrimental to competition.

This application of the standard by the Commission, and subsequent assessment by the European Court of Justice (ECJ) and the General Court (GC), has led to calls for increased clarity. The ECJ has recently responded in a number of its judgments by tightening the scope of conduct that can be considered a ‘by object’ infringement and reaffirming the high evidential burden with which the Commission must comply before making such a classification. The start of this trend can be found in the 2014 landmark judgment in Cartes Bancaires¹ and, most recently, the Budapest Bank² judgment of 2 April 2020. This article summarises the ECJ’s key findings on the issue of ‘by object’ infringements in Budapest Bank, including the new aspects incorporated into the Commission’s necessary analysis when identifying a ‘by object’ infringement – the counterfactual and potential pro-competitive effects.

Budapest Bank: background

The Budapest Bank judgment was delivered in response to a reference for a preliminary ruling from the Hungarian Supreme Court, in the context of an agreement between seven banks to introduce uniform multilateral interchange fees (MIF), applicable to Visa and MasterCard.³

In 2009, the Hungarian Competition Authority (HCA) found that the MIF agreement was a restriction of competition both ‘by object’ and ‘by effect’, and fined the seven banks, MasterCard and Visa. On appeal, the Budapest High Court annulled the decision, and the HCA subsequently appealed to the Hungarian Supreme Court, which in turn referred several questions to the ECJ. One of these asked for the conditions under which an agreement, such as the MIF agreements, could be deemed to amount to a restriction ‘by object’.

On 2 April 2020, the ECJ issued its preliminary ruling, detailing the relevant steps required to characterise conduct as a ‘by object’ infringement. The ECJ confirmed well-established case law principles and helpfully clarified in a new sophisticated manner, the key factors of the Commission’s ‘by object’ analysis, providing a clear road map for characterising a restriction under article 101 of the Treaty on the Functioning of the European Union (TFEU) as one ‘by object’. The Budapest Bank judgment complements another important ruling delivered in January 2020, Generics,⁴ in which the ECJ responded to similar questions from the UK Competition Appeal Tribunal.

Concept of ‘by object’ infringement must be interpreted restrictively

The ECJ has repeatedly stressed this principle in previous cases,⁵ and it is again confirmed in Budapest Bank⁶ and Generics.⁷ The ECJ in Budapest Bank also confirmed the need for competition authorities to support their infringement findings with clear evidence, given in particular the burden of heavy sanctions imposed for infringing conduct.⁸ In other words, the judgment suggests enforcers must adduce great evidential support for their infringement findings, to become arguably commensurate with sanctions faced by parties.

The ECJ can therefore be expected to continue reviewing Commission decisions with increasing scrutiny to ensure that ‘by object’ categorisations are based on unassailable reasoning. As Advocate General (AG) Bobek recalled in his Opinion in Budapest Bank, ‘the existence of alternative legal boxes is no licence for vagueness, in particular when imposing heavy administrative sanctions’.⁹

Analysis of agreement’s nature and content of agreement must be based on prior experience

An often repeated requirement for the classification of infringements of ‘by object’ is that there is a ‘sufficient degree of harm to competition for the view to be taken that it is not necessary to assess their effects’, on the basis that some forms of coordination can be regarded by their very nature as anticompetitive.¹⁰ This basic rule is an ingrained part of the ‘by object’ analysis, but the specific criteria for this has not always been clear. In particular, the Commission has been criticised for using the category for cases that are very distant from hard core cartels. More light has been shed on such criteria in Budapest Bank.

No other plausible purpose

As part of the analysis of whether an agreement amounts to an infringement ‘by object’, an authority can take into consideration the objective purpose of the agreement, but only if that purpose is established effectively.¹¹ In Budapest Bank, for example, the ECJ referred back to the Hungarian court the determination of which of the MIF agreements’ purposes were actually established. Equally, the ECJ in Generics confirmed that the underlying purposes of an agreement will only render the agreement an infringement ‘by object’ if the purposes show that the agreement has no plausible explanation other than the restriction of competition.¹²

Robust and reliable experience

Although this point has been raised in earlier cases,¹³ there is now improved guidance as to what may or may not constitute sufficient experience for the purpose of proving an infringement ‘by object’ in relation to conduct and markets that are complex or unique. In Budapest Bank, the ECJ emphasised that findings of ‘by object’ infringements must be based on robust, reliable and directly relevant experience.¹⁴ To be sure, in cases of hard core cartels, such as horizontal price-fixing, the Commission will continue to have sufficient experience to allow a ‘by object’ infringement finding, as it is clear that this type of conduct harms consumers.¹⁵ In the same vein, however, the Commission should not lean too heavily on its prior decisions to justify ‘by object’ infringement findings in future cases where the relevant markets, products, services and conduct are complex or unique or the conduct may have a serious pro-competitive angle.¹⁶ On this basis, AG Bobek’s Opinion in Budapest Bank highlights that relevant prior experience may comprise a ‘widespread and consistent practice of European competition authorities and/or of the courts of the Member States supporting the view that agreements such as that at issue are generally harmful to competition’.¹⁷ Interestingly, the ECJ found that this threshold was not met in Budapest Bank itself – the ECJ was unable to find evidence of such practice in a similarly ‘unique and atypical’¹⁸ market, and the HCA’s reliance on simply the ‘European Commission’s practice’, namely, the ‘judgments of the EU Courts in MasterCard’¹⁹ was insufficient. These judgments and related Commission decisions (which also dealt with MIFs) did not provide sufficient general and consistent experience that could be applied to the present case. Instead, they were lacking in uniformity and riddled with inconsistencies as identified by AG Bobek in his Opinion.²⁰

The Commission’s five previous decisions against MasterCard in relation to similar MIF agreements varied greatly and did not demonstrate a decisive position as to whether the conduct amounted to a restriction ‘by object’. In fact, a ‘by object’ infringement was found in only one instance (in other instances, a finding of an infringement ‘by effect’ was made or the article 101(3) exemption applied). 21 The inconsistency in the previous practice is why in Budapest Bank, the Commission was required to conduct a detailed examination of the effects so as to confirm, for example, whether the conduct did have the effect of establishing a minimum threshold applicable to service fees. It is therefore imperative for competition authorities to exercise caution when relying on purported precedents, such as those of the Commission, given the potential lack of uniform practice. Indeed, as noted by AG Bobek in Budapest Bank, an enforcer’s own understanding of a certain category of agreements will evolve over time in parallel with increased knowledge and experience, creation of more sophisticated analytical tools, and a development of economic perspectives.²²

Sufficient economic evidence needed that conduct is harmful

Budapest Bank also highlights the need for authorities to adduce evidence of economic consensus to support ‘by object’ findings. To that end, studies or reports prepared by independent authors based on methods recognised by international economic principles are likely to carry significant weight. This is particularly likely to be so in cases in which the EC does not have relevant decisional experience – here, economic reports are likely to be helpful to the Courts to assess whether the conduct amounts to a restriction ‘by object’ or is in fact wholly legitimate when viewed in the overall economic context of the given market. This is particularly important since the notion of a restriction of competition is, as noted by AG Bobek, ultimately an economic concept.²³

Intention is a relevant but non-determining factor for characterisation of an infringement as ‘by object’

The ECJ in Budapest Bank supports the often-cited view that intention is not a necessary factor in determining whether an agreement between undertakings is restrictive, but that the fact that it is not necessary does not mean it cannot be taken into account.²⁴ Interestingly the ECJ closely examined the parties’ intention in its Budapest Bank judgment. However, in practice, subjective intention is ordinarily irrelevant to a finding of a ‘by object’ infringement and is instead reviewed as part and parcel of the other factors in the ‘by object’ analysis (both those listed above and the pro-competitive effects and counterfactual discussed further below).

Key additions in Budapest Bank that tighten ‘by object’ infringement standard

Aside from the legal clarifications noted above, Budapest Bank adds sophistication to the legal standard for establishing a ‘by object’ infringement – first, the need to examine the pro-competitive aspects of the conduct and second, to examine whether the conditions of competition would have been worse absent the conduct (the counterfactual). These points arise as part of the often repeated standard of analysis with which the Commission must comply when reviewing an agreement and determining whether it amounts to an infringement ‘by object’ – namely, a review of the specific circumstances at hand,²⁵ and a practical examination of the nature of the goods or services affected, together with the conditions of functioning and structure of the market.²⁶ Arguably, this standard of analysis is still somewhat subject to the discretion of authorities, given the continued lack of a uniform approach in the conduct of such an analysis in practice.²⁷ However, and as clearly recalled by AG Bobek in his Opinion, this step of the analysis is a ‘basic reality check’ that should not amount to a de facto ‘by effect’ analysis.²⁸

The lengthy and very detailed analysis of the economic and legal context carried out in Budapest Bank and Generics continues to demonstrate that the analysis of an agreement in light of article 101(1) of the TFEU is complex and case-specific. The importance of the actual context is particularly reflected in Generics, in which the ECJ highlighted the unique elements of the medicines sector (including its particular barriers to entry, legislative control of pricing, and price sensitivity related to delay of entry by competitors) and conducted a close review of the pay-for-delay settlement agreements at issue to understand whether they were tantamount to market-sharing agreements (and therefore restrictions ‘by object’).²⁹

New addition 1

Pro-competitive aspects of conduct must be considered when evaluating whether conduct is sufficiently harmful to amount to a restriction ‘by object’

Importantly, the ECJ underscores for the first time in Generics and Budapest Bank³⁰ that the
pro-competitive effects of an agreement are relevant to the contextual analysis to determine when an agreement is characterised as a ‘by object’ restriction. In Generics, the ECJ stated, ‘where parties to that agreement rely on its pro-competitive effects, those effects must . . . be duly taken into account for the purpose of its characterization as a “restriction by object”’.³¹

This is a novel expansion of the legal test, and the ECJ has never previously been as categorical about this requirement – by contrast, in previous cases, the pro-competitive effects were assessed rather as part of the article 101(3) of the TFEU balancing exercise to evaluate whether the pro-competitive benefits outweigh the anticompetitive harm of the restriction (ie, at the exemption rather than characterisation stage). This was, for example, the GC’s approach in its HSBC judgment in September 2019,³² in which, in complete contrast to Budapest Bank and Generics, it stated that pro-competitive effects can only be taken into account in assessing the application of article 101(3) of the TFEU (unless the conduct relates to ancillary restraints).

Importantly, even if the effects on the market are merely ambivalent, or if one cannot rule out a possible pro-competitive economic rationale for an agreement (in the case of Budapest Bank, the reduction of interchange fees) without examining the actual effects, then that agreement cannot be classified as a ‘by object’ restriction.³³ Conversely, if there are pro-competitive effects, this does not mean that the agreement cannot be considered as a ‘by object’ restriction – instead, the pro-competitive effects must be considered as part of the ‘by object’ assessment when they are capable of calling into question whether the relevant conduct revealed a sufficient degree of harm to competition. In other words, in instances where there are both pro-competitive and anti­competitive aspects to the conduct, there cannot be a ‘by object’ infringement if the pro-competitive aspects call into question the alleged harm to competition.³⁴

In any case, the key point to take away is that this further analysis of pro-competitive effects must be conducted in a ‘by object’ context when those effects call into question the alleged harm to competition. In detail, the ECJ in Generics found that, if pro-competitive effects are demonstrated, relevant and specifically related to the agreement concerned, they must be considered if they are ‘sufficiently significant’ so as to justify a reasonable doubt as to whether the agreement caused a sufficient degree of harm to competition and therefore amounted to a ‘by object’ infringement.³⁵

New addition 2

Counterfactual must be taken into account to determine whether conduct had an anticompetitive object

In Budapest Bank, the ECJ clarified that the Commission must also conduct a full counterfactual analysis at the ‘by object’ stage,³⁶ by determining what the competitive situation would have been absent the agreement (for instance, by considering whether a manufacturer would have entered the market, or whether fees for a certain service would have been lower). In this case, the ECJ concluded that if there were important indications that absent the agreement, the conditions of competition would have been worse, this could prevent the characterisation of the conduct in question as a restriction ‘by object’. Earlier, in Generics, the ECJ explained that the counterfactual must not be deduced from a ‘purely hypothetical possibility’ but should be based on ‘real and concrete possibilities’,³⁷ while equally noting that there is no need to show that the counterfactual will in fact happen.³⁸

Budapest Bank therefore provided a concrete example of how an assessment of the counter­factual is relevant at the ‘by object’ stage.³⁹ This assessment of the counterfactual has not previously been as clearly applied within the cartel context. 40 As a result, this part of Budapest Bank signifies a key new development in ‘by object’ analyses in cartel cases. The Budapest Bank approach on this point also differs from the previous approach in Lundbeck,⁴¹ in which the GC held that the counterfactual analysis was impracticable and unnecessary in a ‘by object’ assessment. The pending appeal of this judgment before the ECJ may, however, present an opportunity for an affirmation of the more explicit endorsement of counterfactual analysis at the article 101(1) of the TFEU stage in Budapest Bank,⁴² tightening the evidentiary and analytical burden of the Commission as a result.

Conclusion: key takeaways for future ‘by object’ analyses

The ECJ’s guidance in Budapest Bank introduces additional novel elements to the ‘by object’ standard. In particular, when conducting a ‘by object’ analysis, the Commission must examine:

• the agreement’s nature and content to decide whether it falls within a category of agreements that are commonly presumed to be inherently anticompetitive, making such a finding only when supported by a sufficiently reliable and robust wealth of experience.⁴³ This experience may not necessarily be identifiable through prior decisional practice of competition authorities (including the Commission); and
• if competitive harm is presumed, the economic and legal context relating to the agreement to determine whether the presumption is negated. This includes an assessment of the counterfactual and of any potential pro-competitive effects arising from the agreement.⁴⁴

It is clear that, in practice, assessments will generally be case-specific and authorities such as the Commission should not form decisions at face value without examining the specific goods, markets and potential benefits to the parties in question, as well as potential pro-competitive effects and what the competition conditions in the market would have been absent the agreement at issue. Notably, the ECJ does not intend for these developments to render the EU competition law ‘by object’ standard akin to a US-style ‘rule of reason’ test.⁴⁵ Instead, the Budapest Bank and Generics judgments are likely to mean that in practice, the Commission:

• will need to conduct a more robust assessment in categorising practices as ‘by object’ infringements and more robust justification will be demanded, particularly in complex and novel markets that have not previously been looked at in detail by the Commission. For example, it will need to conduct a more detailed assessment of how the market works and gather more information going to the heart of an agreement’s purpose, particularly when it comes to newer markets and atypical arrangements; and
• may scale back reliance on precedents that are not substantially similar, even when based on similar fact patterns, or be forced not to rely on those decisions at all. As a result, practices involving unique or newer markets, such as multisided markets, or technology-based products and services, are likely to encounter fuller analysis and may be less likely to be considered ‘by object’ infringements while experience continues to evolve in these areas. In that vein, jurisprudence emanating from other jurisdictions with similar rules⁴⁶ may, as hinted by AG Bobek, be informative to decisions in the European Union.

In conclusion, the novel aspects introduced into the ‘by object’ analysis through Budapest Bank build on the previously less rigorous approach taken by the Commission in ‘classic’ cartel cases (ie, relating to traditional goods or services markets in which the same sort of overtly infringing behaviour takes place over time). It is hoped that this tightening of the case law around ‘by object’ classifications signals a move away from a mere consensus approach, whereby certain activities are simply accepted to be anticompetitive, and towards a more detailed case-by-case analysis taking into account the factors discussed above. This should allow for a better application of the ‘by object’ standard to the more complex and rapidly evolving markets with which the Commission is nowadays often faced in its cartel investigations (ie, in which goods and services are increasingly technology based, and related practices do not sit comfortably within a traditional framework of analysis).

In other words, Budapest Bank signals that the legal rules surrounding ‘by object’ infringements must evolve in tandem with market changes, and clarifies complex pre-existing case law while introducing new requirements to review potential pro-competitive effects and the counterfactual in ‘by object’ assessments. The ECJ’s next moves in the Lundbeck and Servier⁴⁷ appeals will be pivotal in confirming whether this trend is here to stay.

Notes