Guidelines without Guidance: Roundtable discussion on the draft Article 102 guidelines

Jorge Padilla and Damien Neven summarise their respective views on the European Commission’s (the Commission’s) draft Guidelines on Article 102. They set out (1) why they agree that draft Guidelines should be welcomed; (2) their contrasted opinions on the usefulness of the two-step test, and in particular the notion of competition on the merits; (3) their interpretation of the introduction of rebuttable presumptions; and (4) how the draft Guidelines, if adopted, would affect their guidance and advice to clients.

The views expressed are those of the authors only and do not necessarily represent the views of Compass Lexecon, its management, its subsidiaries, its affiliates, its employees, or clients. 

What is your opinion on the Commission’s decision to issue new guidelines?

Context

In 2008, the Commission published its guidance paper (‘2008 Guidance’) on how it enforces competition rules relating to Article 102 of the Treaty on the Functioning of the European Union (‘Article 102’).[1] It signalled the regulator’s direction of travel, moving away from a formalistic approach, towards using an effects-based approach to assess abuse of dominance cases.

In March 2023, the Commission amended the 2008 Guidance on its enforcement priorities and announced it would write new guidelines on exclusionary abuses that would consolidate the Union Courts’ case law.[2] This was followed shortly by a policy brief, which set out a broader aim for the Guidelines of establishing “a dynamic and workable effects-based approach to abuse”.[3]

The draft version of these guidelines (‘draft Guidelines’) was published on 1 August 2024 with a view to adopting the new guidelines in 2025.

Jorge Padilla

The 2008 Guidance was a positive contribution to the assessment of exclusionary abuses. It set out an economically grounded approach to identifying the circumstances where one would identify a conduct as abusive based on the effects of that conduct.

However, it has faced several criticisms, which I think have three strands:

1. Some believe the 2008 Guidance no longer aligns well with case law, which has evolved since its introduction.[4]
2. Others thought that the guidelines provided a cogent and consistent framework in principle, but argued that it was not workable in practice.[5] In particular, assessing the risk of anticompetitive foreclosure and estimating its likely effects is too difficult in practice.
3. Finally, and most prominently, some critics – notably the neo-Brandesians – argued that the effects-based approach promoted by the 2008 Guidance had made enforcement of Article 102 more difficult. It meant cases took longer to conclude and, critics argued, tilted the balance in favour of dominant firms. Ultimately, they believe the 2008 Guidance contributed to the alleged increase in market power over the period it was in force.[6]

I agree that the 2008 Guidance is not perfect and I think an update is warranted. The first of these criticisms is fairer than the others. It is not at all clear that the case law is evolving towards the 2008 Guidance and, in some cases, it is moving in the opposite direction. For instance, TeliaSonera on margin squeeze, Google Shopping with respect to self-preferencing, and the last General Court ruling in Qualcomm predation are all decisions that differ from the 2008 Guidance paper in substance.[7] These draft Guidelines explicitly set out to reflect the new case law, [8] which I think is a laudable aim and at least that intention should be welcomed by all stakeholders.

However, I am much less sympathetic to the other criticisms, and the idea that we need a radical departure from the effects-based approach grounded in economic principles which underpinned the 2008 Guidance.[9] It is true that cases take longer than under a formalistic approach; however, that does not justify tearing up the rulebook.[10] I have set out elsewhere my detailed criticisms of the neo-Brandesian approach and my support for an effects based approach, [11] but suffice to say: I consider counterrevolution towards a form based approach would be an overreaction based on faulty logic.

So, I commend the Commission for its intent – if conducted effectively, it could yield something that benefits all parties and the collective good. But I’m concerned about its execution, and caution against overreaction.

Damien Neven

In its draft Guidelines, the Commission has attempted to consolidate the existing case law of the Union Courts and summarise it in a way that parties can use as “guidance”. Given the number of cases since the Commission published the 2008 Guidance, that is a significant and challenging task.

The draft Guidelines do not achieve that task satisfactorily. They do not provide an overarching framework to effectively guide an assessment of the conduct of dominant firms, and (unlike the 2008 Guidance) they hardly use economic principles to inform such assessments.

The case law of the Union Courts on Article 102 is “partially unclear and incoherent”, which makes any attempt to consolidate it inherently difficult.[12] Either the Commission had to confront some of the ambiguities or contradictions in the case law, or select a side and make significant choices with respect to definition of an abuse and conditions that are required to find an abuse.

From my reading of it, they have chosen the latter approach: selecting some cases and ignoring others. They also fail to address issues like the troubling inconsistency between the treatment of refusal to supply and margin squeeze; such that a refusal to supply is currently effectively treated more leniently than less exclusionary practices such as margin squeezes, since the requirement of indispensability applies only to the former.[13] This is despite firms being less incentivised to conduct margin squeezes, and margin squeezes having potentially less anticompetitive effects.

As I have discussed elsewhere, the Commission is also strengthening the standard with respect to rebates contingent on exclusivity.[14] I think that is not justified even if I acknowledge that this is an open debate.[15] In the draft Guidelines, the Commission has ignored this ambiguity and instead opted to select a side. The recent judgement in Intel is however taking a different position which adds to the confusion.[16]

Is the introduction of the two-step test, and its emphasis on 'competition on the merits', useful for assessing abuse?

Context

In assessing an exclusionary abuse, the draft Guidelines set out two necessary conditions:

• the conduct departs from competition on the merits, and
• the conduct is capable of having exclusionary effects.

This is a departure from the 2008 Guidance, which instead focusses on whether a conduct leads to anticompetitive foreclosure, and refers to competition on the merits only in passing.[17]

The “two-step” test has generated significant discussion, and in particular the emphasis on “competition on the merits”.[18] The role of competition on the merits has eluded a conclusive definition since its introduction in Hoffman-LeRoche and is therefore viewed by some as an “an irritant in the case law”.[19] The Commission’s decision to place it in the forefront of the draft Guidelines has therefore proved controversial.

Damien Neven

Conceptually, I think that the two-step test is unnecessary to the assessment of abuse. In particular, the central role given to the concept of competition on the merits is not helpful. It neither clarifies nor guides parties on what the standard for an abuse is.

If one reviews the second step first, the assessment of exclusionary effects, then according to the case law (Servizio Elettrico Nazionale) this is a conduct that “has the effect of hindering the maintenance of the degree of competition still existing in the market or the growth of that competition”.[20] Any assessment should therefore analyse “the degree of competition” with and without the conduct.

We can interpret the “degree of competition” in one of two ways. If it is understood as referring to specifically the competitive constraints imposed by less efficient firms, then this criterion might come into conflict with the principle that Article 102 is not intended to protect such firms.[21] In this case, a conduct would be prohibited even when it reduces competition from less efficient firms and thereby protect them. To avoid this, a second criterion is necessary. Since Hoffmann-LaRoche, the Union courts required that adverse effects must result from “recourse to methods different from those which condition normal competition”, which is also referred to as “competition on the merits”.[22]

However, if the “degree of competition” is understood more generally, in relation to the intended outcome of competitive interactions – consumer welfare – then incorrectly protecting less efficient firms is no longer a concern. An abuse is a conduct that leads to a less competitive outcome than an alternative conduct, and therefore leads to lower consumer surplus.[23] Less efficient firms might then indeed be protected but only when their marginalisation would deteriorate the competitive outcome in the long term.

The draft Guidelines are not clear on what exactly competition on the merits entails, but the definition provided appears to equate it with any conduct that increases consumer benefit.[24] This would make it equivalent to the same criteria as “maintenance of degrees of competition” outlined above.

This is not clarified by its discussion on assessment. For competition on the merits, the draft Guidelines consider whether an as-efficient-competitor would be unable to adopt the same conduct.[25] However, in the very the next paragraph, [26] it adds that the outcome of an As-Efficient-Competitor test is also relevant to assess whether it leads to exclusionary effects.

From this perspective, competition on the merit is superfluous. Its utility has indeed been questioned in recent case law. In Servizio Ellettrico Nazionale, Advocate General Rantos stated that: “To my mind, demonstrating that a dominant undertaking used means other than those which come within the scope of ‘normal’ competition is not a requirement that needs to be assessed separately from the restrictive effect of the conduct.”[27]

There is also some case law that does not give prominence to the concept and focus solely on exclusionary effects to the detriment of consumers. For instance, in Post Denmark II: “It follows that only dominant undertakings whose conduct is likely to have an anticompetitive effect on the market fall within the scope of Article 82 EC. … Such an assessment seeks to determine whether the conduct of the dominant undertaking produces an actual or likely exclusionary effect, to the detriment of competition and, thereby, of consumers’ interests”.[28]

Overall, the prominence given by the draft Guidelines to competition on the merits is unlikely to provide useful guidance to all stakeholders involved. In my opinion, the Commission would be better served by a single test: whether the conduct led to a less competitive outcome than an alternative conduct, and therefore to lower consumer surplus.

Jorge Padilla

Respectfully, I disagree with Damien. As I have maintained for a number of years now, competition on the merits is a useful concept if interpreted correctly by the Commission and the courts.[29]

In my opinion, the only economically correct way to assess a conduct as an exclusionary abuse is to consider abusive only those unilateral actions that, firstly, distort the competitive process and secondly, reduce long-run consumer welfare – where those two conditions are cumulative.[30]

It is important to note that I am not saying that society should ignore conduct that harms consumer welfare but does not distort competition; I am saying that addressing that type of conduct is a job for regulation, not competition policy.

I believe that the Commission’s two steps, if interpreted correctly, effectively assess my cumulative criteria above:

1. Competition on the merits: This should, in my view, be understood as asking whether a dominant company’s actions distort the competitive process. If an actual or hypothetical as-efficient competitor is able to replicate those actions then consumers’ interests will be protected by effective competition and, therefore, the conduct should not be regarded as abusive.
2. Capable of having exclusionary effects: A unilateral action is only likely to reduce long-run consumer welfare through foreclosure or marginalising effective competition – or, as the draft Guidance says: “capable” of reducing it.[31] Therefore, these two are again equivalent in my view.

To illustrate this, it is helpful to think of a given conduct, such as rebates. If we observe the exclusion of an inefficient company then the second step of the test is confirmed. However, to confirm that this was indeed an abuse, it is correct and proper to check if those royalties were able to exclude an as-efficient competitor. If it could, then that is outside the bounds of competition on the merits and it is an abuse of a dominant position. If not, then this is not an abuse: removing inefficient companies is what competition on the merits does.

This distinction may seem immaterial, but there are realistic scenarios where one step applies and the other does not. For example, if a dominant company's investments in quality improvements lead to the marginalisation or exit of competitors, those investments should not be seen as harming competition if equally efficient competitors could raise the funds to replicate the same strategy. Similarly, if a dominant firm sells damaged or broken goods below cost to clear stock, this may not harm competition, as it does not necessarily drive other firms out of the market.

Ultimately, we cannot compel a dominant company to maximise consumer welfare through competition law. Instead, I believe we should compel a dominant company not to distort the competitive process in a way that harms consumers. It should not be able to behave in a way that leverages its dominance (i.e., something that others cannot replicate) to achieve outcomes it could not achieve otherwise.

But a dominant firm has two safe harbours: if its action does not cause harm through foreclosure or marginalisation, or its action is replicable by an as-efficient competitor, then it cannot be abusive. I believe competition on the merits is an effective tool for that assessment, as is the Commission’s two-step framework.

Damien Neven

Let me first observe that the way in which Jorge defines competition on the merits is not quite the way in which the Commission defines it, but essentially focuses on just one criterion put forward by the Commission.[32] Indeed, the approach put forward by Jorge would greatly clarify the distinction between the two legs of the test. However, even taking the definition at face value, I would still have two concerns.

The first is that less efficient firms can provide a valuable competitive constraint, both in the short term and long term, but we may lose that constraint if we apply a test defined in this way. That is because it makes a finding that the dominant firm’s conduct cannot be replicated by as efficient competitors a necessary condition of an abuse. With such a test, there is no limiting principle on the conduct that forecloses less efficient firms. So, irrespective of the question of whether this approach would be consistent with the case law, it means we must accept that some conduct is not abusive, even when it constrains competition and harms consumers.

The second concern relates to the absence of a counterfactual in the application of the test. This is particularly relevant in situations in which a competitor can become as efficient and replicate the conduct of the dominant firm in the long term. Even if a competitor could become as efficient given the conduct of the dominant firm, one might ask about appropriate counterfactuals involving alternative paths that competitors might take to become as efficient. For instance, is it that with higher margins in the short term they could have invested more or been more resilient to future shocks, or is some other alternative? A full analysis of effects would address that issue, as it requires the specification of a counterfactual. But then, as Jorge points out, once we assess the harm relative to a counterfactual, that extends the special responsibility of dominant firms possibly beyond what might reasonably be imposed.

Even if a replication test is only part of a more general analysis of anti-competitive effects, imposing the requirement of a replication test might still be appropriate from a legal and economic perspective as it provides structure and discipline to the assessment of anticompetitive effects.

Jorge Padilla

I fully agree with Damien’s description of a correct effects analysis. But from a pragmatic view, and especially in terms of ensuring compliance, then competition on the merits is a useful step.

However, I should stress that the definition of competition on the merits in the draft Guidelines is not optimal, and I hope that in any subsequent revisions it is clarified. In particular, the draft Guidelines argue that a conduct that is profitably replicable by an as-efficient competitor can be regarded as abusive “under specific circumstances”.[33] I do not understand how this is plausible, even if the conduct is capable of producing exclusionary effects.

Nor can I fathom what the stated “specific circumstances” could be. The three listed in the draft guidelines are so general that they could only lead to unwarranted intervention.[34] Whilst it is true that above-cost prices can produce exclusionary effects, this is only in circumstances in which that strategy is not replicable by an as-efficient competitor.[35]

The core of my argument is that Article 102 cannot be interpreted as demanding companies to maximise consumer welfare. We can do that with regulations if we desire, intervening as governments do in other markets. Infringing Article 102 is a quasi-criminal offense aimed at ensuring the competitiveness of markets, and for its enforcement the two-step approach is helpful: if used properly competition on the merits correctly treats those conducts which exclude less efficient competitors without harming long-run consumer welfare as lawful.

How do you interpret the introduction of rebuttable presumptions in the draft Guidelines?

Context

The draft Guidelines categorise conducts by the evidentiary burden each needs to assess dominance. The taxonomy has three categories.

1. Naked restrictions which are by their nature capable of restricting competition;[36]
2. Conduct that is presumed to lead to exclusionary effects, but which a dominant undertaking can seek to rebut the probative value of the presumptions – these include: exclusive supply or purchasing agreements; rebates conditional on exclusivity; predatory pricing; margin squeeze with negative spreads; and certain forms of tying; [37] and
3. Conduct for which it is necessary to demonstrate a capability to produce exclusionary effects.[38]

The case law of the Union Courts has historically used the terms “by object” and “by effect” to differentiate, respectively, what the draft Guidelines calls ‘naked restrictions’ from those that require an effects analysis to show abuse. This has led to significant debate over whether dividing “by effect” abuses into those with and without a rebuttable presumption of exclusionary effects signals a more formalistic approach.[39]

Jorge Padilla

I am not overly concerned by the introduction of these tests. We have had rebuttable presumptions in practice for a long period of time. Crucially, the draft Guideline deploys these presumptions based on tests, not form. So, I do not think they differ from our current tests: you can have a rebuttable presumption of illegality if the price-cost test is negative in a predation case. Similarly, the “exceptional circumstances” test in Bronner posits a rebuttable presumption of legality.[40]

Damien Neven

I agree with Jorge that the introduction of rebuttable presumptions is just a different way of presenting the existing case law and should not be seen as a major novelty.

Rather, the real issue is that the Commission is trying to reduce the evidential burden required to demonstrate that the party is pursuing a conduct that triggers a rebuttable presumption.

Firstly, it is not clear to me whether the Commission’s taxonomy of conducts and evidentiary burden presented by the draft Guidelines is a helpful way of presenting the case law. The taxonomy is somewhat misleading. With respect to all conducts except exclusive dealing (including rebates contingent on exclusivity) what triggers the “presumption” goes beyond the form of the practice and requires an assessment of circumstances. In this sense, the draft Guidelines tend to reflect the case law as it currently stands, rather than any movement towards assessment by form.

My only concern is with exclusive dealing, and rebates conditional on exclusivity, where the draft Guidelines take a different approach. For exclusive dealing, the draft Guidelines suggest that harm is presumed simply based on the presence of an exclusivity clause, whether explicit or implied. The draft Guidelines do not provide detailed criteria for assessing the potential exclusionary effects of these practices, such as evaluating whether an equally efficient competitor could replicate them or considering the counterfactual. This means the presumption of harm is triggered more easily without requiring the same level of evidence or analysis as other practices. It is also worth noticing that the recent Court judgment on Intel does not support the Commission’s approach on this point.[41]

Secondly, on the evidential burden, I consider that the Commission is selectively trying to reduce the onus on it to demonstrate effects. This is most clear in the As-Efficient Competitor test, where the Commission has shrunk significantly the circumstances in which the test would be considered meaningful compared to the 2008 Guidance, and changed the evidentiary value of the test.[42] Similarly, the soft safe harbour that was provided by the 2008 Guidance in circumstances where the effective price is above cost is also weakened.[43]

Therefore, whilst I consider the Commission’s taxonomy of rebuttable presumptions simply a confused way of organising the case law, I am more concerned by how they use these presumptions. It is also a concern that the Commission is systematically referring to presumptions rather than factual inferences (with the effect that it appears to strengthen the evidentiary burden on the dominant firm to establish the absence of effects).

Jorge Padilla

I agree that what the Commission has done is, in effect, equate what have long been considered factual inferences with legal presumptions. In this way, they are trying to avoid the discipline imposed by courts to look at all the economic context, and therefore lowering the evidentiary burden placed on the Commission. From a legal standpoint the case law contains very few legal presumptions, and hence it may be cleaner to refer to factual inferences throughout the draft, rather than legal presumptions.

If the Commission adopts the draft Guidelines as they are, would this affect your guidance and advice to clients and, if so, how?

Jorge Padilla

No, the draft Guidelines will not change the advice I give to my clients. The draft Guidelines are not the law. They are just an indication of how the Commission views the law. They are not binding on courts, which ultimately is where these cases are decided.

In its current form, the draft Guidelines will not be a useful instrument going forward for companies, agencies, or the courts. It is a distorted summary of what the courts have already said and frankly we need to open the discussion more widely to improve it going forward.

I am sure the draft Guidelines will come in for criticism from many different angles, not all consistent, but I hope some of the most important concerns that we have discussed are reflected in any revisions.

More generally, I welcome debates such as these: what is an abuse is an interesting and knotty question!

Damien Neven

I think that the draft Guidelines as currently proposed are a missed opportunity to provide real clarity to stakeholders in the context of often quite confusing case law. I do not think that what is provided is effective guidance, and therefore they are not a useful tool for how I will frame my own guidance to stakeholders.

The decision to consolidate the case law, and not present any overarching organising principle to collect and make sense of the evidence means the draft Guidelines are a poor tool for understanding the issues at hand, as they are liable to become outdated as soon as the case law evolves (illustrated by the recent judgments on Google shopping and Intel).[44]

More broadly, the draft Guidelines miss an opportunity to incorporate recent developments in economic research and the improved understanding of exclusionary practices since the 2008 Guidance paper was issued.

References