On the notion of restriction of competition: What we know we don’t know and what we don’t know we don’t know.

Alfonso Lamadrid de Pablo (Garrigues) and Professor Pablo Ibáñez Colomo (London School of Economics and College of Europe) have written together a paper entitledOn the notion of restriction of competition: what we know and what we don’t know we know.[1]

The issue mentioned in the title is both fundamental and unsettled. Thus, one must applaud the authors for taking up this topic. Below are my comments.

After briefly discussing the subject matter and the perspective of the paper (1.), I will comment on how the authors deal with the delicate problems of proving restrictive effects (2.) and defining the notion of restriction of competition (3.).

In the conclusion I will reflect on the perceived gap between the authors’ acute criticisms and the surprising leniency of their final assessment (4.).

As usual, I will mainly focus on the points of disagreement.

I naturally invite any visitor to start by reading the paper (which happens to be available online) and to use the comment section (below) to discuss my article.

1. The subject matter and the perspective of the paper.

1.1. Positivist approach.

As from the beginning, the authors make it clear that their paper follows what they call a “bottom-up” approach[2]. In other words, they want to state the law as it is and not as they think it should be. In clear, they adopt a positivist standpoint.

This approach is very useful. However, in my view, the authors unduly downplay the importance of the “top-down” (prescriptive) approach. Indeed, these perspectives must not be confused, but both are necessary. The law should be at once clear and sound.

Let us note that the paper strongly relies on an article whose author, following the opposite approach, claims that his aim is not to defend a particular view as the only “correct” reading of the law but to develop a consistent theory based on some selection of the case law.[3]

My comments will pertain to what the law is and to what it should be.

1.2. What the authorities say and what they do.

The authors also explain that they did not stick to the courts’ words. Indeed, as they euphemistically write it, “there is sometimes a difference between what the EU courts say and what they actually do” and “some of the formulas used by the Court fail to capture the determinant factor of the ruling.” [4]

I am afraid that such a critical method is indeed necessary. It cannot be denied that the courts and competition law agencies often limit themselves to repeating empty and contradictory slogans. Examples abound.

As a result, a commentator cannot trust the formulas that enforcers use when trying to explain how they decide their rulings. One needs to read between the lines in order to reconstruct their reasoning (or lack of).

As a side note, I must say that I do not know of any other legal discipline where there is such an abyss between what courts say and what they really do. What should be thought of such a discipline will be briefly discussed in the conclusion.

1.3. Explaining the outliers.

Of course, the authors admit that there are some cases which do not fit with their rationalizations. In such cases, they try to point out the outliers and explain them.

Once again, this is a legitimate approach, provided that one conclusively shows that the cases which are not compatible with the stated theory really qualify as mere outliers (which requires that they be comparatively rare).

I am inclined to think that the authors do not always succeed in this exercise, but I will not undertake here to challenge them on this field. I will instead focus on the internal cogency of their statements.

1.4. Only 101§1 or 101§3 as well?

I am not sure about the precise subject matter of the paper. As far as it deals with article 101 of the TFEU, does it pertain only to paragraph 1, or to paragraph 3 as well?

For the record, article 101§1 provides that agreements which restrict competition are prohibited, while article 101§3 specifies that these agreements can be exempted under some conditions if they contribute to economic progress (to make it short).

There are two ways of describing these two provisions.

One is to say that §1 defines what a restriction of competition is (and prohibits it), while §3 would exempt agreements which are known as anticompetitive but display some virtues unrelated to competition. Thus, a practice could be both anticompetitive and legal.

The other reading is that some agreements are exempted under paragraph 3 because, after further analysis, they reveal to be procompetitive. In this view, a truly anticompetitive practice could not be exempted and the concept of a restriction of competition would be defined by combining both paragraphs.

The definition of a “restriction of competition” should vary depending on which of these readings is used.

Now, at first, it would seem that the authors only deal will §1 and retain the first reading. Indeed, they refer thirty-seven times to article 101§1 and only three times to article 101§3.

However, there are several elements which tend to contradict this analysis.

Firstly, they open their paper by writing that “In EU competition law, conduct is deemed lawful or unlawful depending on whether it restricts competition.”[5] Now, unless the paper also deals with 101§3 and thus adopts the second reading, this is not true.

Secondly, the authors make “efficiency gains[6] and “weighing up[7] enter into their definition of a restriction of competition, while these points are normally examined under article 101§3.

Thirdly, the paper holds that the definition it suggests applies to both article 101 and article 102 (and the merger regulation). However, article 102 does not follow the same architecture as article 101. It does not include any paragraph setting out an exemption for economic progress.

As a consequence, the same concept of a restriction of competition can be used for both articles only if the second reading of article 101 is adopted.

2. How to prove a restriction of competition?

Article 101§1 prohibits agreements “which have as their object or effect” the restriction of competition. It is a settled (although contestable) case law that agencies do not need to prove that a practice is a restriction by effect if they can show that it is a restriction by object.

However, there is a considerable confusion about how to define and distinguish between a restriction by object and a restriction by effect. It is well known that this distinction is unclear. Many papers take up this issue and conclude that they cannot make sense of the case law.

Thus, we could say that we know that we do not know how to tell the difference between both notions.

Now, the authors of the paper undertake to solve this issue. I argue that they succeed only partially and will mainly discuss the points with which I disagree.

2.1. The restriction by object.

2.1.1. The authors’ theory.

According to the authors, who give credit to another writer for elaborating this theory[8], a practice can be considered as a restriction by object when:

  • it is not capable of having any procompetitive effect ; and
  • it is capable of having an anticompetitive effect.

These two conditions are cumulative.

This clarification deals with what needs to be proved in order to show a “by object” restriction.

However, the authors are less clear regarding another issue, namely what elements need to be taken into account in order to prove the aforementioned criteria.

2.1.2. Assessment of the context vs. case-by-case assessment.

One of the most debated questions is which circumstances an enforcement agency is supposed to examine in order to determine whether there is a restriction by object.

The courts generally consider that an anticompetitive object must be inferred from both the “very nature of the practice” and “the legal and economic context.”

These injunctions seem contradictory. Indeed, it is difficult to explain how something could result at once from the “very nature” of the practice and from its “economic context.”

In other words, it is not clear whether a case-by-case assessment must be carried out or whether the practice can be considered in the abstract.

I am afraid that the authors do not clearly answer this question. Indeed, they seem to state that a restriction by object must be analyzed in its economic context but without carrying out a case-by-case assessment.

Let us quote them.

First, they state that “by object” restrictions must be assessed in their (economic and legal) context:

If one pays attention to what the EU courts actually do – and not so much to generic statements and formulas – it appears that they typically start by considering the rationale behind the practice, that is, the reasons why a firm would resort to it in the legal and economic context of which it is part. If this analysis leads to the conclusion that there are no credible pro-competitive reasons for its implementation, the EU courts typically conclude that it is restrictive of competition by object.[9]

The need to consider the nature of the behaviour and the context of which it is part knows no exceptions.[10]

Nevertheless, they write that

When a practice is deemed restrictive by object, it is presumed to have a net negative impact on competition without it being necessary to assess its anticompetitive effects on a case-by-case basis. Such a case-by-case assessment becomes necessary, on the other hand, where conduct falls under the ‘by effect’ label.[11]

However, the context which must be examined is necessarily the context of the case. So, how could we do one thing (examining the economic context) without doing the other (doing a case-by-case assessment)?

As discussed by another author,

[L]et us go back to Société Technique Minière. In this case, the Court refers to the “economic context.” At first glance, this notion plants seeds of doubt. It seems to require a genuine case-by-case market assessment, which is a priori incompatible with the idea of presumption which underpins the anticompetitive object.[12]

The link between the taking into account of the context and the necessity of doing a case-by-case assessment is obvious in the following excerpt of the paper, which deals with a “by effect” restriction:

Where the net impact of a practice is found to be ambivalent (that is, if it may be positive or negative depending on the economic and legal context), on the other hand, they require a case-by-case assessment.[13]

If “by object” restrictions too must be analyzed in their context, it means that their impact is ambivalent until the assessment of the context has been done.

If, on the one hand, the distinctive criterion between the two categories is that “by effect” restrictions require a case-by-case assessment, and, on the other hand, it turns out that “by object” restrictions too require this kind of analysis, the difference between both categories seems to vanish once again.

2.1.3. Is it “safe” to presume that a practice has a net negative effect while it is not proved that it is likely to have such an effect?

The authors write that:

Where the threshold of capability is met, and the practice is not a plausible source of efficiency gains, it is safe to presume that it will have a net negative impact on competition. Put differently, it is safe to label the practice as restrictive by object.[14]

This can be interpreted in two different ways.

Either the authors mean that it is logically safe to jump from these facts (capability of causing negative effects and incapacity to generate positive effects) to the conclusion that it must have a net negative effect, because they think that the conclusion safely follows from the premises.

Or the authors consider that it is safe because, even if it is not sure that such a practice would entail a net negative effect, there would be no harm in prohibiting it.

Both interpretations are problematic.

It is clear that the first is incorrect. Indeed, it is obvious that the mere capability of causing a negative effect, even associated with the incapacity of generating a positive effect, is not a reasonable basis to conclude that there is a net negative effect.

The second interpretation, which could be analyzed as a prescriptive approach, is problematic from a public liberty point of view. The (disowned?) father of modern criminal law, Cesare Beccaria, rightly wrote that

To what a situation should we be reduced, if every thing were to be forbidden that might possibly lead to a crime? We must be deprived of the use of our senses. For one motive that induces a man to commit a real crime, there are a thousand which excite him to those indifferent actions, which are called crimes by bad laws.[15]

One could argue of course that it is safe to prohibit a practice which is unable to generate a procompetitive effect because, as Hovenkamp said, “little is lost by prohibiting socially useless behavior.”[16]

However, such a statement is chilling. Little is lost, really? What about liberty? Is it a negligible thing?

Outside of antitrust’s parallel legal world, it would seem horrible to prohibit a behavior under the guise that it could not be of any interest for society. Indeed, in a free society, as opposed to a totalitarian state, people are not required to actively promote the social welfare.

It seems that the authors are somehow conscious about this issue. Indeed, in their paper, they alternate between saying that “it is safe to presume that it cannot have a net positive impact on competition[17] and that “it is safe to presume that it will have a net negative impact on competition.”[18]

Moreover, Hayek warned about of the hidden cost of thinking that we can safely prohibit a behavior, because we cannot know of which benefit this prohibition will deprive us:

Since the value of freedom rests on the opportunities it provides for unforeseen and unpredictable actions, we will rarely know what we lose through a particular restriction of freedom. Any such restriction, any coercion other than the enforcement of general rules, will aim at the achievement of some foreseeable particular result, but what is prevented by it will usually not be known. […] As in the particular instance we shall hardly ever know what would be the consequence of allowing people to make their own choice, to make the decision in each instance depend only on the foreseeable particular results must lead to the progressive destruction of freedom. There are probably few restrictions on freedom which could not be justified on the grounds that we do not know the particular loss they will cause.[19]

In addition, as a matter of fact, it is wrong to assume that a practice which cannot have a positive effect on competition cannot increase welfare.

Carl Menger once wrote that “the general principle of all economic exchanges of goods, according to which both parties must derive an economic advantage from an exchange, maintain[s] its validity unimpaired in the case of monopoly.”[20] As far as I know, this statement has never been refuted.

2.1.4. On the vanity of judicial “experience” about restrictive effects.

The courts think they can rely on “experience” to determine whether a practice can have pro-competitive effects.

The authors report for instance that

The plausibility of the pro- and anticompetitive effects of an alleged infringement cannot be evaluated in the abstract. The analysis of this question needs to consider the economic and legal context of which the practice is part, as well as the lessons of experience and economic analysis.[21]

The threshold of capability is met, it would seem, where it is plausible, in light of experience and economic analysis, that the practice under consideration will have negative effects in the economic and legal context of which it is part.[22]

The authors do not much comment on this point, but it seems useful to say some words about it.

Indeed, it appears that the courts are confusing the stare decisis doctrine with the scientific experimental method. The experimental method works only if each particular experiment can be assessed on its own merits. In contrast, if the experimenter is bound (even partially) by previous experiments, a series of consistent cases, no matter how long, would be of no help to find truth.

Furthermore, experience can be gained only if it is possible to objectively ascertain the result of each experiment. Ludwig von Mises clearly explained this point:

The method of trial and error is applicable in all cases in which the correct solution is recognizable as such by unmistakable marks not dependent on the method of trial and error itself. If a man mislays his wallet, he may hunt for it in various places. If he finds it, he recognizes it as his property; there is no doubt about the success of the method of trial and error applied; he has solved his problem. When Ehrlich was looking for a remedy for syphilis, he tested hundreds of drugs until he found what he was searching for, a drug that killed the spirochetes without damaging the human body. The mark of the correct solution, the drug number 606, was that it combined these two qualities, as could be learned from laboratory experiment and from clinical experience.[23]

Now, it is hard to understand how the courts could ever have gained enough experience to know that a definite practice can be safely presumed to generate net harmful effects, because they cannot even explain what a harmful effect is supposed to be in the first place.

In Mises’ words, they cannot say what the “unmistakable marks” of a restrictive (or pro-competitive) effects are. Actually, they do not even seem to care about it.

As the authors of the paper write (although without specifically linking it with the ability for judicial authorities to gain “experience”),

In the context of Article 101 TFEU, [the Court] has consistently prioritised agreements that it considers to be restrictive by object. This preference is not without consequences. […] It may […] be difficult to understand how to assess the effects of a practice if most agreements pursued by the authority are deemed to be ‘by object’ prohibitions.[24]

Courts and agencies can surely learn by experience that such or such practice is generally prohibited in the case law and so will be similarly prohibited in the future. Having any experience about whether these prohibited practices were indeed harmful is an altogether different issue.

More on this later.

2.2. The restriction by effect.

If a practice cannot be classified as a restriction “by object,” agencies can still try to show that it is a restriction “by effects.” The authors discuss what needs to be established under the latter rubric.

2.2.1. Likelihood of gross negative effects or likelihood of net negative effects?

The authors state that a given practice is to be analyzed as a restriction “by effect” only if it is at least capable of having positive effects (otherwise it would be treated as a “by object” restriction).

Thus, in order to prohibit such a practice, it is insufficient to show that it is capable of having negative effects. According to the authors, it is necessary to demonstrate that, after further inquiry about the negative side of the balance, it is likely to have such effects. Likelihood is a higher level of probability than capability.

However, a practice which is likely to have negative effects and capable of having positive effects can still have a net positive (or neutral) effect if it turns out, after further enquiry about the positive side of the balance, that it is also likely to have positive effects.

The question is thus whether the prosecuting agency can limit itself to demonstrating that the practice is likely to have gross negative effects or if it has to prove that is likely to have net negative effects.

The authors seem to consider that it is necessary to show net negative effects:

The findings of this paper suggest that it is possible to provide a tentative definition of what amounts to a restriction. In essence, it can be defined as a practice that is presumed to have a net negative impact on firms’ ability and incentive to compete. In other words, the notion refers to instances in which the anticompetitive effects are presumed to weigh more than any pro-competitive gains that may result from it. In the context of Articles 101 and 102 TFEU, conduct may be found to restrict competition by object or effect.[25]

The latter hypothesis would require the agency to do a balancing of the good and bad effects. So far, the European Commission has always refused to assume this probationary burden.

2.2.2. Actual, future, and potential effects.

The authors devote a section of their paper to the issue of what they call the “temporal” dimension.

They explain in this section that a “potential” effect can have two meanings:

  • Within the first meaning, a “potential” effect is opposed to an “actual” one, the former being described as an effect which has not “already” occurred[26] ;
  • Within the second meaning, a “potential” effect is an effect whose probability is assessed “in the abstract” on the basis of a “theoretical possibility.”[27]

The paper indicates that the meaning retained by the case law is not the second one, which seems to imply that the courts use the first one.

However, there seems to be a confusion between “potential” and “future.”

Saying that an event has not “already” happened implies that one is sure that it will happen one day, although one does not know when. In such a case, the event is “future” and not merely “potential.”

On the contrary, to say that an event is “potential” means that one does not know whether it will actually happen in the future. This is not just a “temporal” issue.

By creating a false dichotomy, the paper seems to indicate that firms can only be punished for negatives effects for which one can be sure they will happen at some point of the future.

But doing so, it portrays competition law as less derogatory to usual law than it actually is.

3. How to define a restriction of competition?

Handbooks on competition law generally confuse the issue of proving a restriction of competition with the one of defining what a restriction of competition actually is.

As a consequence, they only discuss whether it is sufficient to prove a potential restriction, or whether the restriction must be significant, etc.

These discussions generally assume that the notion of restriction of competition is perfectly clear.[28]

Yet, the definition of the notion is actually even more obscure than the probationary issues. Thus, we may say that most of us do not know that we do not know what a restriction of competition is.

As the authors state,

After more than 50 years, there is no such thing as a generally accepted definition of the notion of restriction of competition. [29]

[T]he Court has not had many opportunities to provide guidance about how the analysis of appreciable effects is to be performed in practice. It is submitted that the lack of clarity about this question relates not so much to the issue of appreciability but to the notion of effects itself. The case law requires that effects must be appreciable, but has not expressly defined what an effect is.[30]

They rightly describe this point as a “crucial issue.”[31]

3.1. What a restriction of competition is not (or not necessarily).

3.1.1. It is possible to infer a set of guiding principles from the case law.

According to the authors,

a restrictive effect can mean many different things, from a disadvantage to rivals and/or customers to a negative outcome in terms of welfare. While the Court has not formally defined what a restrictive effect is, it is possible to infer a set of guiding principles from the case law.[32]

Let us see what these principles are.

3.1.2. Not welfare.

The authors write that

it seems difficult to dispute that a restrictive effect within the meaning of EU competition law does not necessarily amount to a detriment to consumer welfare. The Court has consistently held that the point of the discipline is not merely to protect consumers directly (as it does when action is taken against exploitative and collusive conduct). When action is taken to address exclusionary conduct, consumers’ interests are protected indirectly, by preserving the market structure. Thus, it is not (or not always) necessary to show that an exclusionary strategy has led, or is likely to lead, to consumer harm.[33]

This reasoning seems to be flawed.

Indeed, saying that the point of the discipline is not to protect consumers directly does not imply at all that protecting consumers is not at the heart of the discipline or that it does not require to prove any consumer harm.

To the contrary, if words have a meaning (which admittedly is not sure in competition law), it means that an exclusionary practice can be punished only if it is proven that it harms consumers, albeit indirectly.

At best, it is possible to argue that this proof is established by the mere fact that the “market structure” has been “harmed.” This would be an incorrect theory of harm, but as least this would not be internally inconsistent.

Of course, it is admissible to say that this is just another example of how the Court often improperly rationalizes its own behavior.

3.1.3. Not freedom of action either.

The authors also state that

The analytical approach traditionally favoured by the Commission, which was based on the impact of practices (or other corporate transactions) on firms’ freedom of action, appears to have been all but abandoned. The Commission itself departed from this definition during the 1990s, following a major shift in its policy. The gap left by this shift has not been conclusively closed by other conceptual frameworks.[34]

But what was a restriction of “freedom of action,” actually?

It appears that this notion was totally nebulous, Marxist-inspired, and very alien to the classical conception of freedom.

This latter consisted, within the economic field, in the right for firms to use their property as they desired, for instance by concluding contracts (or by refusing to do so), and to set their prices as they saw fit, provided neither the bodily integrity nor the properties of anyone else were physically injured. Pecuniary externalities (pure economic losses) were considered perfectly legal.

On the contrary, “freedom of action,” as this concept is used in competition law, seems to be injured as soon as someone’s economic interests are negatively affected, or when a contractual tie is formed, or as soon as someone does not manage to enter into such a tie with a suitable partner (i.e., when a firm is “foreclosed” from a market).

For example, the authors write that “limit[ing] access to outlets or supplies[35] or negatively impacting “customers’ ability and incentive to obtain supplies from rival producers[36] constitute restraints on freedom of action.

Now, if one considers that someone’s freedom is harmed as soon as his incentives are modified, every economic action is restrictive of freedom, because all economic actions are interdependent.

Thus, as the authors write it, “In a sense, every practice has an effect on competition.[37] This definition of freedom was so wide as to become nonsensical and even counterproductive. So, I would not complain about its (alleged) abandonment. I would rather criticize its hidden survival.

3.1.4. Not any injury to the “market structure.”

According to the authors,

the Court has held that not every effect on the market structure necessarily amounts to an infringement. In its Article 102 TFEU case law, it has ruled that EU competition law is only concerned with the impact of practices on equally efficient rivals, at least so as a matter of principle.[38]

It would indeed be a good thing if competition agencies were exclusively concerned with the forclusion of as efficient competitors (although it would not be appropriate to analyze this as a sufficient condition for prohibiting a practice).

Indeed, as the authors write,

it is natural that the exclusion of less efficient competitors is not deemed problematic in EU competition law – at least not as a rule. The ability of a firm to attract customers away from rivals because its products are better, or cheaper, is typically the very manifestation of competition and should not as such be subject to scrutiny.[39]

However, it would be difficult to agree with the authors’ characterization of the state of the law.

To buttress their claim, they refer to a paragraph of the Post Danmark II case where the Court held that

it is sufficient to demonstrate that there is an anti-competitive effect which may potentially exclude competitors who are at least as efficient as the dominant undertaking.[40]

The authors specifically quote this sentence.

There seems to be a confusion between what is necessary and what is sufficient. What is sufficient is not always necessary. To say the least, the quoted sentence does not support the claim that competition law is only concerned with the impact of practices on equally efficient rivals.[41]

Moreover, in other paragraphs of the same case, the Court specifically held that

The as-efficient-competitor test must thus be regarded as one tool amongst others for the purposes of assessing whether there is an abuse of a dominant position in the context of a rebate scheme.[42]

[T]he application of the as-efficient-competitor test does not constitute a necessary condition for a finding to the effect that a rebate scheme is abusive under Article 82 EC. In a situation such as that in the main proceedings, applying the as-efficient-competitor test is of no relevance.[43]

Of course, it could be alleged that these pronouncements are specific to rebate schemes (although it would not explain why the authors specifically quoted this case) and that, outside of this field, competition law would be “only concerned with the impact of practices on equally efficient rivals.”

The problem is that this condition is often entirely dropped by the courts and commentators, which focus on the existence of an “exclusionary effect” without enquiring further about whether the excluded firms were as efficient.

3.2. What a restriction of competition would be.

3.2.1. What the authors think a restriction of competition is.

To begin with, the authors write that

It would seem that a restrictive effect within the meaning of the case law is somewhere between a disadvantage to rivals and harm to consumer welfare.[44]

Actually, this is not a very useful or even intelligible statement. One can conceive that a restriction of competition would consist in a disadvantage to rivals causing a harm to consumers, but what does it mean to say that it is “between” a disadvantage to rivals and harm to consumer?

Fortunately, the authors go on with specifying their analysis. They hold that

Restrictive effects seem to exist where a practice has reduced, or is likely to reduce, firms’ ability and incentive to compete on the relevant market. Thus, the focus of the analysis should be placed on the impact of the practice on the competitive pressure to which firms would otherwise have been subject.[45]

Let us notice that the authors resort to the verb “to compete” in their definition of a “restriction of competition.” Now, the very meaning of “competition/to compete” is also much debated.

Thus, read literally, this definition does not seem to take us very far. The authors do not expressly opt for any side of the debate on whether “competition” must be understood in its lay meaning (rivalry) or in a more technical sense (maximizing allocative efficiency).

It is rather surprising that the authors did not clearly answer a question that they clearly raised.

However, it appears that they tend to lean on the “rivalry” side.

3.2.2. Criticisms of the definition set out by the authors.

a) Is there any difference between the authors’ definition and the freedom of action paradigm?

To begin with, let us recall that, as the authors mentioned,

The analytical approach traditionally favoured by the Commission, which was based on the impact of practices (or other corporate transactions) on firms’ freedom of action, appears to have been all but abandoned.[46]

The authors do not seem to mind the slightest this change of paradigm.

In the new paradigm, according to them,

it has become abundantly clear that a restriction in the freedom of action of rivals or suppliers is not sufficient to establish a significant impediment to effective competition. A concentration can curtail the freedom of action of rivals, suppliers and customers in various ways. […] Such consequences, however, do not in themselves justify remedial action. If the transaction has no impact on rivals’ ability and incentive to compete, it does not lead to a significant impediment to effective competition.[47]

As one can see, the authors consider that, in addition to a restriction of the freedom of action of rivals, suppliers and customers, it is necessary to show a restriction of competition within the meaning defined in the paper, namely a negative impact on “rivals’ ability and incentive to compete.”

But, very curiously, the authors also explain that a restriction of someone’s “ability and incentive” is just another name for a restriction on “freedom of action:

The factors that need to be taken into account when evaluating the exclusionary impact of a practice were identified by the Court in Post Danmark II. In addition to the effects on customers’ ability and incentive to obtain supplies from rival producers (that is, on their freedom of action), it is necessary to examine, inter alia, the nature of the product, the features of the relevant market or its coverage.[48]

This reasoning appears inconsistent.

Moreover, the authors incidentally reveal, on the one hand, that they have a very fuzzy idea of the meaning of their own definition and, more importantly, that they do not propose any real change compared to the old, derided paradigm of “freedom of action.”

They appear to be just pouring old wine into new bottles.

b) What about efficiency?

Earlier, I mentioned that the authors considered that:

  • a restrictive effect within the meaning of EU competition law does not necessarily amount to a detriment to consumer welfare ;
  • a restriction of competition was “somewhere between a disadvantage to rivals and harm to consumer welfare.”

These two apparently opposite statements seemed to imply that their definition would be connected in some roundabout way to consumer welfare.

However, the definition they set forth (“Restrictive effects seem to exist where a practice has reduced, or is likely to reduce, firms’ ability and incentive to compete on the relevant market”) does not include any direct hint about it.

Indeed, it is noteworthy that they refer to the ability and incentive “to compete” and not “to serve consumers better.”

Yet, the authors try (rather hesitantly) to reintroduce the efficiency component at a later stage. Indeed, they write that

Insofar as the practice is a source of efficiency gains, their incentives to compete are likely to be, if anything, enhanced.[49]

As one can see, the reference to efficiency is rather hesitant (“if anything”). Moreover, it cannot be accepted.

It is to be noticed that the authors care not about the effect of rivalry on efficiency, but about the effects of efficiency on rivalry. It implies that maximum rivalry (or “competitive pressure”) remains the ultimate objective.

Now, it is absurd to focus on rivalry as an end in itself. Indeed, cooperation is as fundamental to the economy as competition. There is absolutely no reason to privilege one over the other.

Some eminent writers like Robert Bork[50] and Frank Easterbrook[51] have written about this issue more than thirty years ago. It is a pity that the paper ignores their sharp objections.

I know that the authors announced from the outset that their approach would be analytical and not prescriptive. Still, even if their definition really reflected the case law, it would have been better, I think, to emphasize how absurd it is.

In addition, as will be shown in the next section, efficiency gains cannot entail a net increase in firms’ ability and incentive to compete.

c) Rivalry, as opposed to consumer welfare, is necessarily a zero-sum game.

Rivalry, as opposed to serving the consumers well, is a zero-sum game. Indeed, the more one firm’s (or group of firms’) incentive and ability increases, the more other firms’ decreases.

That rivalry is by essence a relative phenomenon has been aptly explained by Armen Alchian:

Positive profits accrue to those who are better than their actual competitors, even if the participants are ignorant, intelligent, skillful, etc. The crucial element is one’s aggregate position relative to actual competitors, not some hypothetically perfect competitors. As in a race, the award goes to the relatively fastest, even if all the competitors loaf. Even in a world of stupid men there would still be profits. [S]uccess (survival) ac-companies relative superiority […].[52]

Thus, the definition suggested by the authors is conceptually flawed. There cannot be any net increase in firms’ ability and incentive to compete.

It should be noted that the authors, who refer to a “net” decrease, do not specify how to do the balancing.

If they had, they would probably have realized that what they suggest to do is impossible. In practice, adopting this criterion would allow judges to decide competition law cases according to their whims.[53]

d) What about the “as efficient competitor”?

Let us recall that, according to the authors, “EU competition law is only concerned with the impact of practices on equally efficient rivals, at least so as a matter of principle.”[54]

However, if their definition were read literally, a negative impact on a less efficient firm’s ability and incentive to compete would qualify as a restriction of competition.

How to explain this apparent inconsistency?

The point is that the authors analyze the criterion of the “as efficient competitor” through the lens of the causality requirement. They write that:

the foreclosure of less efficient rivals is an expected and necessary consequence of the operation of the competitive process. Typically, the exclusion of such rivals is not attributable to the practice implemented by a dominant firm, but to their inability to attract consumers. As such, it does not trigger intervention under Article 102 TFEU.[55]

[I]ntervention is only justified where there is a causal link between the practice and the anticompetitive effects. Thus, it is natural that the exclusion of less efficient competitors is not deemed problematic in EU competition law – at least not as a rule. The ability of a firm to attract customers away from rivals because its products are better, or cheaper, is typically the very manifestation of competition and should not as such be subject to scrutiny.[56]

Because the causality requirement goes generally without saying, the authors do not need either to expressly mention the “as efficient component” in their definition and, as least in theory, they do not have to connect this issue with a welfare analysis.

Now, it should be noted first that this way of explaining why the forclusion of less efficient competitors is (or should be) legal does not seem shared by the courts or other commentators. The latter often approve of the “as efficient” criterion, but they do not seem to link it to the causality requirement.

Besides, I really like the idea that the direct causal factor which leads firms to being excluded is the choice of consumers, and not the behavior of competitors.

However, this is true for almost all the cases where a firm is excluded. It does not hold true only for the case where the forclosed firm was less efficient.

As Richard Posner put it:

Turning to practices more likely to be genuinely exclusionary than tying, we’ll want to distinguish between those that require the cooperation of a customer to be effective and those that don’t. Predatory pricing […] depends on the purchasers’ willingness to buy from the predator (or the intended victim) at the predatory price; blowing up a competitor’s plant, procuring a patent from the Patent Office by fraud, or harassing a competitor with baseless litigation does not.[57]

Even when a firm is driven out of the market by means of so-called “predatory” prices, the immediate cause of the exclusion if the choice of the consumers. A novus actus takes place between the “predatory” behavior and the exclusion.

Moreover, if ones goes to a deeper level of analysis and enquire about what induces consumers to to exclude a firm, it appears that both legal and illegal exclusions are indirectly caused by the behavior of competitors (as Hovenkamp said, “Superior efficiency is the world’s greatest entry barrier, except perhaps for government entry restrictions.[58]).

Of course, it is possible to try to distinguish between firms’ behaviors which rightfully induce consumers to refuse buying from the victims (ex: prices above costs) and those which wrongfully cause the same result (ex: prices under costs). Actually, that is what competition law does.

However, by doing so, one leaves the objective field of causality in order to pass a normative judgment about the proper means of competing.

To sum up, causality cannot explain why competition law allows (or should allow) more efficient firms to exclude less efficient ones.

e) Competition on the relevant market.

Lastly, it should be noted that the authors specify that restrictive effects exist when their is a net reduction of firms’ ability and incentive to compete “on the relevant market.”[59]

Once again, if words have a meaning (a big “if” in competition law), this implies that a practice which incites firms to compete on another market qualifies as an illegal restriction.

This is logically so even if, on the whole, firms’ ability and incentive to compete in the economy would have « increased. » Thus, this requirement appears to be nonsensical.

Moreover, it is in complete contradiction with the theory of allocative efficiency, which implies that there may be too many firms in a relevant market. For instance, according to Laffont and Tirole, “Entry may be socially […] inefficient because it creates a duplication of fixed costs.”[60]

One may wonder why the authors included this requirement. They do not refer to any positive case law demanding it.

One may think that it comes from the Commission guidelines on article 101§3, which provides that only the benefits which accrue on the relevant market can be accounted for the granting of the exemption.

However, it has been shown that this requirement is patently absurd. Thus, one may regret that the authors of the paper include it in their definition of a restriction of competition, in particular in the absence of any case law compelling them to do so[61].

4. Conclusion: The authors’ very low expectations.

One may sympathize with the authors’ goal of clarifying the notion of restriction of competition and of adopting an essentially positivist approach. One may also agree with how they point out the many perceived inconsistencies of the case law which are in need of explanation.

However, on several important points, it is hard to be convinced by the answers that the authors give to the issues they bring up. Moreover, even if all of their explanations were correct, one could not avoid being surprised by their leniency toward the deficiencies of their discipline.

Indeed, the authors quietly explain that it is not possible to trust how competition law enforcers rationalize their own rulings and that it is necessary to go beyond the language of the courts. I do not know of any other legal discipline where this happens on such a regular basis.

They also state after “After more than 50 years, there is no such thing as a generally accepted definition of the notion of restriction of competition[62] and they sum up the findings of their paper by saying that “it is possible [sic] to provide a tentative [sic] definition of what amounts to a restriction.”[63]

Similarly, after writing that “The ability of a firm to attract customers away from rivals because its products are better, or cheaper, is typically the very manifestation of competition,”[64] they casually observe that “It would seem [sic] that not all [sic] practices that make it more difficult for rivals to compete are prohibited.”[65]

Thus, “it would seem” that “not all” practices which are “the very manifestation of competition” are prohibited by competition law. How to report so placidly that one is not sure that a legal discipline does not fully frustrate its official goal?

Yet, at the end of their paper, they conclude that “authorities, courts and practitioners are equipped with sufficient elements to assess the restrictive object and/or effects of a practice[66] and that “the case law is a sound and valuable starting point to assess the impact of a practice on competition.”[67]

How to reconcile the authors’ comments regarding many objective deficiencies of this field of law and the leniency of their final assessment?

It seems that they have very low expectations regarding the rationality of this discipline, as if competition law were a parallel legal world where all the usual quality requirements are adjusted downward or discarded.

The problem is that competition law is a branch of criminal law (or at least repressive law) on the basis of which severe punishments are inflicted to firms by some politico-administrative junta.

Any other body of criminal laws so intellectually vacuous and so at odd with the rule of law would have long been put by the European Court of Human Rights into history’s dustbin.

***

[1]     Alfonso Lamadrid de Pablo and Pablo Ibáñez Colomo, “On the Notion of Restriction of Competition: What We Know and What We Don’t Know We Know,” October 8, 2016 (hereinafter referred to as “On the Notion of Restriction of Competition.” Forthcoming in Damien Gerard, Massimo Merola and Bernd Meyring (eds), The Notion of Restriction of Competition: Revisiting the Foundations of Antitrust Enforcement in Europe (Bruylant 2017).

[2]     On the Notion of Restriction of Competition, aforementioned, p. 4.

[3]     Luc Peeperkorn, “Defining ‘by object’ restrictions,” Concurrences, n° 3-2015, pp. 40-50.

[4]     On the Notion of Restriction of Competition, aforementioned, p. 28.

[5]     On the Notion of Restriction of Competition, aforementioned, p. 2.

[6]     On the Notion of Restriction of Competition, aforementioned, pp. 21, 22, 24…

[7]     On the Notion of Restriction of Competition, aforementioned, pp. 21 and 44.

[8]     See Luc Peeperkorn, “Defining ‘by object’ restrictions,” aforementioned.

[9]     On the Notion of Restriction of Competition, aforementioned, p. 24.

[10]    On the Notion of Restriction of Competition, aforementioned, p. 6.

[11]    On the Notion of Restriction of Competition, aforementioned, p. 44.

[12]    Nicolas Petit, “La rationalisation au long cours de la restriction de concurrence “par objet” dans la jurisprudence de la Cour de justice de l’Union européenne,” Actualité juridique Contrats d’affaires, octobre 2015, p. 424 (my translation). Quoted in the paper, p. 4.

[13]    On the Notion of Restriction of Competition, aforementioned, p. 21.

[14]    On the Notion of Restriction of Competition, aforementioned, p. 35.

[15]    Cesare Beccaria, An Essay on Crimes and Punishments, Liberty Fund, chapter XLI, p. 74.

[16]    Herbert Hovenkamp, The Antitrust Enterprise: Principles and Execution, Harvard University Press, 2005, p. 53.

[17]    On the Notion of Restriction of Competition, aforementioned, p. 21.

[18]    On the Notion of Restriction of Competition, aforementioned, p. 35.

[19]    Friedrich Hayek, Law, Legislation and Liberty, Routledge Classics, 2013, p. 55.

[20]    Carl Menger, Principles of Economics, Ludwig von Mises Institute, 2007, p. 211.

[21]    On the Notion of Restriction of Competition, aforementioned, p. 44.

[22]    On the Notion of Restriction of Competition, aforementioned, pp. 34-35.

[23]    Ludwig von Mises, Human Action – A Treatise on Economics, Fox & Wilkes, San Francisco, 1963, p. 704.

[24]    On the Notion of Restriction of Competition, aforementioned, p. 18.

[25]    On the Notion of Restriction of Competition, aforementioned, p. 44.

[26]    On the Notion of Restriction of Competition, aforementioned, p. 33.

[27]    On the Notion of Restriction of Competition, aforementioned, p. 33.

[28]    See, for a typical example, Luc Peeperkorn, “Defining ‘by object’ restrictions,” aforementioned.

[29]    On the Notion of Restriction of Competition, aforementioned, p. 2.

[30]    On the Notion of Restriction of Competition, aforementioned, p. 39.

[31]    On the Notion of Restriction of Competition, aforementioned, p. 39.

[32]    On the Notion of Restriction of Competition, aforementioned, p. 39.

[33]    On the Notion of Restriction of Competition, aforementioned, p. 39.

[34]    On the Notion of Restriction of Competition, aforementioned, p. 2.

[35]    On the Notion of Restriction of Competition, aforementioned, p. 14.

[36]    On the Notion of Restriction of Competition, aforementioned, p. 15.

[37]    On the Notion of Restriction of Competition, aforementioned, p. 32.

[38]    On the Notion of Restriction of Competition, aforementioned, p. 40.

[39]    On the Notion of Restriction of Competition, aforementioned, p. 40.

[40]    On the Notion of Restriction of Competition, aforementioned, p. 40, footnote 158 (“See above, note 40. See also AKZO (n 73), para 72; TeliaSonera (n 74), para 64; and Post Danmark II (n 13), para 66 (‘it is sufficient to demonstrate that there is an anti-competitive effect which may potentially exclude competitors who are at least as efficient as the dominant undertaking’)”).

[41]    The other cases referred to in footnotes 158 (or 40) do not support this claim any more.

[42]    Judgment of the Court, case C‑23/14, 6 October 2015, Post Danmark A/S v Konkurrencerådet, paragraph 61.

[43]    Post Danmark, aforementioned, paragraph 62.

[44]    On the Notion of Restriction of Competition, aforementioned, p. 41.

[45]    On the Notion of Restriction of Competition, aforementioned, p. 41.

[46]    On the Notion of Restriction of Competition, aforementioned, p. 2.

[47]    On the Notion of Restriction of Competition, aforementioned, p. 14. Let us note that this paragraph dealt with “the field of merger control.”

[48]    On the Notion of Restriction of Competition, aforementioned, p. 15.

[49]    On the Notion of Restriction of Competition, aforementioned, p. 42.

[50]    Robert Bork, The Antitrust Paradox – A Policy at War with ItselfThe Free Press, 1993, pp. 58-59.

[51]    Frank H. Easterbrook, The Limits of Antitrust, Texas Law Review, Volume 63, Number 1, August 1984, pp. 2 and 13.

[52]    Armen Alchian, “Uncertainty, Evolution, and Economic Theory,” Journal of Political Economy, Vol. 58, No. 3 (Jun., 1950), p. 213. Quoted by Bork, in The Antitrust Paradox, aforementioned, p. 121.

[53]    Which admittedly is already what they do.

[54]    On the Notion of Restriction of Competition, aforementioned, p. 40.

[55]    On the Notion of Restriction of Competition, aforementioned, p. 43.

[56]    On the Notion of Restriction of Competition, aforementioned, p. 40.

[57]    Richard Posner, Antitrust Law, The University of Chicago Press, 2001, p. 207.

[58]    Herbert Hovenkamp, Federal Antitrust Policy, West Group, 1999, p. 524.

[59]    On the Notion of Restriction of Competition, aforementioned, pp. 41-45.

[60]    Jean-Jacques Laffont and Jean Tirole, « Cartelization by Regulation », Journal of Regulatory Economics; 5:111-130, 1993, p. 112.

[61]    I cannot exclude that there is no case law demanding it. I just state that the authors do not mention any.

[62]    On the Notion of Restriction of Competition, aforementioned, p. 2.

[63]    On the Notion of Restriction of Competition, aforementioned, p. 44.

[64]    On the Notion of Restriction of Competition, aforementioned, p. 40.

[65]    On the Notion of Restriction of Competition, aforementioned, p. 40.

[66]    On the Notion of Restriction of Competition, aforementioned, p. 43.

[67]    On the Notion of Restriction of Competition, aforementioned, p. 45.

Publicités

Laisser un commentaire

Entrez vos coordonnées ci-dessous ou cliquez sur une icône pour vous connecter:

Logo WordPress.com

Vous commentez à l'aide de votre compte WordPress.com. Déconnexion / Changer )

Image Twitter

Vous commentez à l'aide de votre compte Twitter. Déconnexion / Changer )

Photo Facebook

Vous commentez à l'aide de votre compte Facebook. Déconnexion / Changer )

Photo Google+

Vous commentez à l'aide de votre compte Google+. Déconnexion / Changer )

Connexion à %s