The Sisyphean quest for a clear test – on ‘by object’ methodology, gin & tonics, and Budapest Bank
Kluwer Competition Law Blog
May 4, 2020
Please refer to this post as:, ‘The Sisyphean quest for a clear test – on ‘by object’ methodology, gin & tonics, and Budapest Bank’, Kluwer Competition Law Blog, May 4 2020, http://competitionlawblog.kluwercompetitionlaw.com/2020/05/04/the-sisyphean-quest-for-a-clear-test-on-by-object-methodology-gin-tonics-and-budapest-bank/
Sisyphus was a Greek mythological figure who was condemned to roll an immense boulder up a hill only for it to roll down every time it neared the top. Repeating this action for eternity. The quest for a clear and operational methodology for finding a restriction of competition ‘by object’ feels similar sometimes. Every time Member State courts roll their questions up the hill to the CJEU, we await in anticipation the clarity that the preliminary ruling will bring about this time. We keenly dissect the judgment once handed down and debate & discuss its implications. However, once the next case comes around before the European Commission or elsewhere, we quickly realize that, on reflection, the judgment didn’t really provide much operational clarity after all. The boulder has rolled down again and we repeat.
The CJEU’s judgment in case C-228/18 Budapest Bank is a good example of this. The judgment provides for very interesting discussions – particularly on whether looking at the ‘counterfactual scenario’ is required for finding a ‘by object’ restriction. But, like cases before it, it provides little practical & operational clarity.
What is settled in ‘by object’ methodology?
We accept that to find a ‘by object’ restriction, you must complete the three-step exercise of analysing 1) the content of the agreement, 2) its objectives, and 3) its legal and economic context.
When examining context, the CJEU says that you must consider the nature of the goods or services affected, as well as the real conditions of the functioning and structure of the market.
A few additional angles are also beyond dispute. Firstly, the parties’ (subjective) intentions are not a mandatory factor in the analysis but authorities can take them into account. Authorities seem to enjoy this prerogative as it means internal correspondence showing the parties’ intentions can only be damning and never exculpatory. Secondly, and typically cited as the first precedent by companies fending of accusations of wrongdoing, Cartes Bancaires provides that the ‘object box’ is narrow and covers only the most obvious anti-competitive practices.
Based on these factors, the competition authority is to check whether the companies’ behavior “reveals a sufficient degree of harm to competition that there is no need to examine its effects”. While accepting that you cannot establish anti-competitive behaviour in the abstract, the framework is not particularly operational for objectively predicting exactly when a full effects analysis is needed. The framework’s shortcomings in this regard are fortified by the many elements still up for debate.
What is still debated?
Despite decades of case-law interpreting the same 63-year-old Treaty provision, the competition law community is still debating some very significant questions on how to find a restriction ‘by object’ – including the following to name a few:
- Does the finding of a ‘by object’ restriction require determining the ‘counterfactual scenario’ and comparing it to the (anti-)competitive nature of the behaviour?
- How far do authorities need to go when looking at the ‘legal and economic context’ of an agreement? In other words, exactly at what point does necessary analysis of context stop and unnecessary analysis of effects start?
- To what extent must competition authorities consider redeeming or pro-competitive features of an agreement as relevant context for their ‘by object’ analysis under Article 101(1) as opposed to a defense that is open to the company to raise and prove under Article 101(3)?
These questions may overlap. But they are all very significant and sometimes decisive when businesses face accusations of wrongdoing from competition authorities. An official from a competition authority was once asked what the difference is between a restriction ‘by object’ and ‘by effect’, to which he famously replied: “About 3 years…”. The fact is that if competition authorities are unable to bring a case as ‘by object’, the alternative (the counterfactual scenario, if you like…) is often not bringing the case at all due to resources. It’s ‘by object’ or nothing. Both authorities and advisors representing companies under investigation are well aware of this which is why we have the ongoing battles of what authorities can get away with putting into the ‘object box’.
Is the counterfactual scenario a thing in ‘by object’ cases?
Counterfactual analysis is most prominent in merger control. This is because merger control is probably the most scientific of the antitrust disciplines in that it examines the impact of a merger on a pure cause & effect basis. Counterfactual analysis is also prevalent in follow-on cartel damages as, once again, the question of cause & effect (causation) is paramount. The exercise is far less explicit in behavioural cases whether it be under Article 102 or most Article 101 cases. However, the fact that we do not observe express counterfactual analysis in many behavioral cases is not evidence that it is not there in a more subtle or indirect form.
Another distinguished competition law blog has addressed this question very competently. Nevertheless, here are the best arguments, as I see them, FOR as well as AGAINST counterfactual analysis being required in ‘by object’ cases.
The case FOR counterfactual analysis in ‘by object’ cases
The most persuasive argument in favour is perhaps derived from common sense – a virtue that should be considered on occasion even by lawyers. When applying common sense, it is impossible to divorce counterfactual analysis from the purpose and context of any agreement. Take as an example absolutely any agreement or transaction in your private life:
- What is the purpose & context of getting a job and signing an employment agreement? Well, absent getting a job (the counterfactual scenario), I would have no money to pay rent or buy food.
- What is the purpose & context of renting a flat and signing a tenancy agreement? Well, absent renting a flat, I would be sleeping on the street or, even worse, at my parents’.
- What is the purpose & context of buying fresh lemons at the supermarket? Well, but for the lemons, I would be enjoying my lock-down gin & tonic with an inferior slice of cucumber. And so on…
In fact, as many of us know from more or less tedious training sessions on negotiation, counterfactual analysis forms such an essential part of the purpose & context of any agreement that it even has a name in negotiation theory: the Best Alternative To a Negotiated Agreement (or BATNA).
It therefore makes no sense to say that analyzing the counterfactual scenario is irrelevant when investigating the content, purpose, and legal & economic context of an agreement or concerted practice. And doing so displays a blatant ivory tower approach to competition enforcement.
As for case-law endorsing the relevance of the counterfactual in ‘by object’ cases, paras. 82-83 in Budapest bank is probably the best we have. Here, the CJEU holds that indications of upward or downward pricing pressure in the absence of the arrangement on multilateral interchange fees cannot be disregarded in the assessment of whether the arrangement is a restriction ‘by object’.
The case AGAINST counterfactual analysis in ‘by object’ cases
If one were to try and push back on the arguments above, one could raise the following:
Firstly, if counterfactual analysis were required to find a ‘by object’ restriction in the same way as a restriction ‘by effect’, the CJEU has gone to great lengths to not say so explicitly in its decades of case-law. The CJEU quite consistently reserves the very express requirement to analyze the counterfactual for when describing how to find a restriction ‘by effect’ – like it also does in para. 55 of Budapest Bank.
Secondly, Lundbeck and other pay-for-delay cases are difficult to square with requiring full counterfactual analysis in ‘by object’ cases as doing so would essentially turn pay-for-delay cases into ones of patent infringement. If the European Commission were required to find the counterfactual in Lundbeck, it would have had to run a full simulation on whether Lundbeck would have won a patent infringement case against the generic manufacturers. Because only by running such a simulation could the European Commission verify whether the generics could legally have entered the market absent the agreements with Lundbeck. If there is no legal entry, there is no competition for the pay-for-delay agreements to restrict. Of course, the CJEU is still to give its final word in this case.
Thirdly, some may take issue with a strict requirement to apply the counterfactual when taken to its logical extremes. Here a horizontal and vertical example:
- Horizontal example: Could cartelists claim that there is no infringement as, absent the cartel, they would have merged into one single undertaking? And to make the example more interesting, adding the assumption that the merger would fall below any filing thresholds (a merger of local driving school instructors, for example). In other words, can the cartelizing, local driving school instructors claim that, while price fixing is their first choice, there is in fact no competition to restrict as they would have otherwise merged?
- Vertical example: Can the supplier claim that the RPM he has imposed on his non-genuine agent is no infringement as he, absent that arrangement, would have simply tweaked the agreement so it falls within genuine agency while still controlling the agent’s pricing?
When does analysis of context become analysis of effects?
Put bluntly, there is no answer to this question. At least none that helps to provide a clear and operational test for predicting exactly when an agreement restricts competition ‘by object’. As Advocate General Bobek put it in paras. 49-50 of his opinion in Budapest Bank, the difference between analyzing context and effects is one of degree rather than kind. And, as he says, there is no standard type of analysis or set level of depth and meticulousness that an authority has to adopt to carry out the analysis of context.
In his opinion in Toshiba, Advocate General Wathelet argued that there is in fact two categories within the ‘object box’ warranting different intensities of analysis of legal & economic context. While agreements expressly referred to in Article 101(1) need only ‘light touch’ investigation of context, an “…atypical or complex [agreement] requires a more thorough analysis of the economic and legal context of which they form part, although that analysis does not go so far as an examination of the effects of the agreement”. In practice though, this does not provide much guidance as an agreement between competitors containing some element of pricing alignment – like Budapest Bank and other MIF cases – can be atypical and complex at the same time as theoretically falling within the category of agreements expressly referred to in Article 101(1).
Are redeeming features a matter for Article 101(1) or (3)?
When battling with competition authorities on behalf of companies under investigation, authorities will often take the position that any pro-competitive features to the agreement is for the companies to raise and prove under Article 101(3). Advisors on the other hand will refute this arguing that redeeming and pro-competitive features form part of the economic context that the authority must consider before finding a restriction ‘by object’ under Article 101(1).
In his opinion in Budapest Bank, Advocate General Bobek endorses the latter position when stating that any time an agreement appears to have ambivalent effects on the market, an effects analysis is required. And that when a possible pro-competitive economic rationale for an agreement cannot be ruled out without looking at the actual effects on the market, that agreement cannot be classified as restrictive ‘by object’. An approach seemingly endorsed by the CJEU in the judgment itself.
Maybe we are not meant to have a crystal clear test…
The CJEU’s guidance in Budapest Bank is indeed interesting and helpful – not least as it holds that a competition authority can only find a ‘by object’ restriction once it has demonstrated based on ‘reliable and solid experience’ that the agreement is harmful by its very nature. However, determining what constitutes ‘reliable & solid experience’ is arguably just as subjective an exercise as any guidance the CJEU has given before. The cynical view would be that the CJEU is simply unpacking existing vague language describing the ‘by object’ test and repacking it in language that, albeit different, is equally vague.
Perhaps the drafters of the Treaty of Rome in 1957 never intended for us to have a crystal clear test for finding a ‘by object’ restriction. Maybe lack of predictability is simply the price for the fact that you cannot determine a restriction of competition in the abstract. In any case, no doubt the immense boulder will continue to roll down and continue to be pushed up the hill again.