Global Antitrust Review 2014

163

The Assessment of Selective Distribution Systems
Post-Pierre Fabre
CRISTIANA DE FAVERI

This article develops an analysis of selective distribution systems, which considers their main characteristics, the traditional principles governing their discipline and the recent evolution of the jurisprudence of the European
Courts. In the Pierre Fabre case, the Court seems to have changed its approach to the subject, as, analysing a ‘de facto’ ban on online sales, it passed from the traditional ‘effects’ analysis of these systems to their qualification as restrictions ‘by object unless objectively justified’. The legal result is unaltered as the usual criteria of competition analysis are still adopted, but new perspectives could have been potentially opened in their interpretation. Furthermore, the article also takes into account the Court’s assertion that ‘the aim of maintaining a prestigious image is not a legitimate aim for restricting competition’, which represents another aspect of the judgment in apparent contrast with the settled case law. In both instances, it presents possible explanations consistent with the tradition. Finally, it underlines the significance of the ruling in relation to the controversial topic of internet selling.

I. Introduction
The topic of selective distribution systems has been explored in many authoritative academic works.1 Yet, it is still important and highly debated, not only because
1

See Alison Jones and Brenda Sufrin, EU Competition Law (OUP 2014); Maher M.
Dabbah, EC and UK Competition Law (CUP 2004); Richard Whish and David Bailey,
Competition Law (7th edn, OUP 2012); Joanna Goyder, EU Distribution Law (Hart
Publishing 2011); Vivien Rose and David Bailey (eds), Bellamy and Child: European
Union Law of Competition (Sweet and Maxwell 2013), Joanna Goyder and Albertina
Albors-Llorens, Goyder’s EC Competition Law (5th edn, OUP 2009); Valentine Korah, An

Selective distribution systems post-Pierre Fabre

164

many aspects are not entirely clear, but also because recent peculiar developments in the jurisprudence of the European Court of Justice (CJEU) have fed the discussion. A Selective Distribution System is a network of vertical agreements between a supplier and its distributors.2 Resellers (wholesalers and/or retailers) are admitted to the system only on the basis of their acceptance of certain obligations, necessary in the view of the supplier to create and maintain a prestigious brand image or to provide customers with a qualitative service. As such, they might consist, for instance, of pre-sales or after-sales services, technical qualifications and training of the staff, equipment or location of the premises, stock and inventory of contract goods, or even the commitment not to sell competing products of inferior quality in order not to damage the image of the brand. Indeed, the system is usually adopted in relation to products of high value and technological complexity or with a luxurious image to convey and the shopping experience of customers must be preserved. The supplier provides its products only to authorised members of the system, who, in turn, undertake not to resell to non-authorised resellers.
The system is designed to enhance the efficiency and quality of the distribution chain, but it inherently restricts the economic freedom of its components and
Introductory Guide to EC Competition Law and Practice (9th edn, Hart Publishing 2007);
Giorgio Monti, EC Competition Law (CUP 2007); David Bailey, ‘Restrictions of
Competition by Object under Article 101 TFEU’ (2012) 49 Common Market Law Review,
559; Pablo Ibanez Colomo, ‘Market failures, transactions costs and article 101(1) TFEU case law’ (2012) 37 European Law Review, 541; Jorren Knibbe, ‘Selective distribution and the ECJ's judgment in Pierre Fabre’, Case Comment (2012) 33 European Competition Law
Review, 450; Giorgio Monti, ‘Restraints on Selective Distribution Agreements’ (2013) 36
World Competition, 489; Louis Vogel, ‘Efficiency versus Regulation: The Application of
EU Competition Law to Distribution Agreements’ (2013) 4 Journal of European
Competition Law & Practice, 277; Andreas Themelis, ‘After Pierre Fabre: the future of online distribution under competition policy’ (2012) 20 International Journal of Law &
Information Technology, 346; Okeoghene Odudu, ‘Interpreting Article 81 (1): the object requirement revisited’ (2001) 26 European Law Review, 379; Saskia King, ‘The Object
Box: Law, Policy or Myth?’ (2011) 7 European Competition Journal 2, 269; Cosmo
Graham, ‘Methods for determining whether an agreement restricts competition: comment on Allianz Hungaria’ (2013) 38 European Law Review, 542; Csongor Istvan Nágy, ‘The
Distinction between Anti-competitive Object and Effect after Allianz: The End of
Coherence in Competition Analysis?’ (2013) 36 World Competition 4, 541; Alison Jones,
‘Left Behind by Modernisation? Restrictions by Object under Article 101(1)’ (2010) 6
European Competition Journal 3, 649.
2
A vertical agreement is described, by article 1(a) Commission Regulation No 330/2010,
OJ [2010] L 102/1, as an ‘agreement or concerted practice entered into between two or more undertakings each of which operates, for the purposes of the agreement or the concerted practice, at a different level of the production or distribution chain, and relating to the conditions under which the parties may purchase, sell or resell certain goods or services’. Global Antitrust Review 2014

165

typically creates serious concerns about its compliance with competition rules. In particular, its compatibility with article 101(1) TFEU is often questioned. For many years the jurisprudence of the European Courts developed on the basis of few and very specific criteria and gave economic operators and legal practitioners a (not always consistent) line of reasoning for self-assessment and avoidance of infringements of the EU competition law. Yet, very recently the CJEU seems to have adopted a different approach and started to depart from previous case law, in contrast with the methodology followed by the EU Commission.
In this light, the article begins in section II with a brief overview of the main characteristics of Selective Distribution Systems, focusing on their anti and procompetitive aspects and their reciprocal interaction. In section III, it analyses the principles traditionally governing the application of the competition analysis by the
European Courts and the Commission to their discipline, preceded by the preliminary consideration of the fundamental distinction between ‘restriction by object or by effect’. Section IV focuses on the recent developments in the jurisprudence of the European Court of Justice and on the Pierre Fabre case in particular, as well as on the judicial assertion that ‘the aim of maintaining a prestigious image is not a legitimate aim for restricting competition’.3 It starts with a short report on the factual background of the proceedings and, then, it examines the peculiarities and the importance of the judgment for competition analysis in general and Selective Distribution Systems in particular.
The article underlines and explains not only the CJEU’s approach to the controversial topic of online sales, but also its puzzling new qualification of
Selective Distribution Systems as restrictions on competition ‘by object’ ‘in the absence of objective justification’, rather than ‘by effect’, and its assertion that ‘the aim of maintaining a prestigious image is not a legitimate aim for restricting competition’. The last two statements, indeed, seem to contrast with the settled case law on the subject matter, but it is also possible to provide an interpretation consistent with the tradition.
II. The economics of selective distribution systems
This specific type of distribution can be described according to the wording of article 1(1)(e) of Regulation 330/2010, as:
(…) a distribution system where the supplier undertakes to sell the contract goods or services, either directly or indirectly, only to distributors selected
3

Case C-439/09 Pierre Fabre Dermo-Cosmétique SAS v Président de l’Autorité de la concurrence [2011] ECR I-9447, para 46.

Selective distribution systems post-Pierre Fabre

166

on the basis of specified criteria and where these distributors undertake not to sell such goods or services to unauthorised distributors within the territory reserved by the supplier to operate that system.4
The provision aims at determining the boundaries of the Verticals Block
Exemption, but it is generally adoptable as an appropriate reference, because it captures the essence of the system.
1. The anti-competitive aspects of selective distribution
A Selective Distribution Agreement may vary widely, case by case, containing different terms and clauses in relation to the various circumstances under which it is adopted. Yet, two elements are consistently considered: the position of the supplier, which undertakes to provide its products only to selected distributors, and the role of the distributors that commit not to resell the products to non-authorised dealers. The contract goods and services must be resold only to end users or to other members of the network, designed to be of ‘closed character, making it impossible for non-authorised dealers to obtain supplies’.5
The obligation on the appointed resellers is regarded as necessary to the effectiveness of this particular marketing method (so as to ensure that the specified criteria for the selection of the resellers are satisfied).6 However, it is important to underline that, according to the case law of the European Court of Justice,7 the
‘imperviousness’ of a Selective Distribution System worldwide is not a condition of its validity under the EU law, contrary to what happens in some Member States
(e.g. under the German law, implying that ‘an un-authorized dealer can obtain the goods covered by that system only by participating in the breach by an authorized dealer of his contractual obligations’).8 As a consequence, it might happen that official retailers have to compete with other non-members, legitimately dealing with original products which have been purchased at a lower price in another
Member State where selective distribution is not applied, and made available through parallel import.9 Yet, some authors caution that it would be hard to convince the Commission and the European Courts of the necessity of the system
4

Cf n 2.
Guidelines on Vertical Restraints, OJ [2010] C 130/1, para 178.
6
Case 26/76 Metro v Commission (No 1) [1977] ECR 1875, para 27; Jones, Sufrin, (n
1),806; Joanna Goyder, EU Distribution Law (Hart Publishing 2011), 135.
7
Case C-376/92 Metro v Cartier [1994] ECR I-00015, para 28; Whish, Bailey (n 1), 645;
Case T-88/92 Groupement d'Achat Edouard Leclerc v Commission [1996] ECR II-1967, para 115; Goyder (n 6), 139; Vivien Rose and David Bailey (eds), Bellamy and Child:
European Union Law of Competition (Sweet and Maxwell, 2013) 462 (footnote n 387).
8
Metro v Cartier, n 7, para 28.
9
Goyder, Albors-Llorens (n 1), 234.
5

Global Antitrust Review 2014

167

in relation to the nature of the product, if it is not applied in all jurisdictions in which goods and services are sold.10
The burden not to resell the purchased goods to whomever the new owner wishes represents a clear restriction of its economic freedom. However, it is considered compatible with competition law rules when the overall efficiencies of the system outweigh the anti-competitive aspects. This is believed to happen when the specific limitations identified by the Commission and the CJEU are respected.11
Case law and decisional practice show that selective distribution systems have been adopted for many categories of goods and services12 (without firm rules and predictability for each of them),13 but they are usually preferred for the marketing and sale of final products of high quality and value, technologically complex, or branded with a luxury or prestigious image to promote and protect.14 In relation to these products, they are of particular benefit for consumers.15
2. The pro-competitive aspects of selective distribution
Selective distribution allows suppliers to maintain a certain degree of control on the distribution chain, even if implemented through independent undertakings. The manufacturer may ensure that the outlets do not deviate from the image it wishes to project and that they provide customers with quality pre-sales and after-sales services. In the meantime, it may avoid costs, such as the need for specific expertise, knowledge of local markets or financial and commercial risks that other forms of distribution, such as vertical integration or agency relationships, entail.
In addition, selective distribution systems may protect the authorised distributors’ investment and promotional efforts against other retailers’ temptation to ‘free-ride’.
Indeed, the margin for free-riding among them becomes consistently reduced, if the contract products can be sold only by authorised members, selected on the basis of uniform criteria and, thus, subject to the same obligations and requirements (in terms of equipment and appearance of their premises, technical qualifications and
10

Jones, Sufrin (n 1), 810; Goyder (n 6), 139.
Korah (n 1) 315; Goyder, Albors-Llorens (n 1), 237.
12
The topic will be further considered in the following sections of this work. However, to have a general view please refer to the main text-books quoted in the footnote n 1, among others, please refer to Bellamy and Child, where it is highlighted the acceptability of
Selective Distribution concerning cameras, televisions, hi-fi products, computers, high quality watches and clocks, jewellery, glass crystal, ceramic tableware, newspapers, but not mass watches, tobacco products, or plumbing equipment. Also, it has been accepted in relation to motor vehicles, dental prostheses, perfumes and cosmetics.
13
Bellamy and Child, (n 1), 456; Goyder, Albors-Llorens (n 1), 240.
14
Guidelines on Vertical Restrains (n 5,) para 174.
15
Giorgio Monti (n 1) 370.
11

Selective distribution systems post-Pierre Fabre

168

ability to provide professional advice on the point of sale, pre-sales services and promotional activities). The system could also protect distributors’ relation-specific investments against the possible misbehaviour of the manufacturer itself (the so called ‘hold-up problem’), providing guarantee of long-term supply through appropriate arrangements.16 This protection serves as an incentive. Furthermore, purchasers of expensive products are usually sophisticated. Selective distribution is particularly apt to meet their high expectations, so as to increase demand and enhance sales for the advantage of all the parties involved.
3. Inherent contradiction and inextricable connection
Contradictions are inherent in selective distribution due to the restrictions, on the one hand, imposed on the economic freedom of the undertakings involved
(particularly significant in this respect is the obligation on distributors not to resell contract products to non-authorised members of the system) and on the other hand, the potentially high benefits it entails for the efficient distribution of goods and services on the market. Indeed, selective distribution promotes reduction of costs and greater control by the manufacturer, improved services and promotional efforts by distributors for the benefit of customers. Enhanced efficiency implies increase in sales and profits for both suppliers and distributors, while it also ensures consumer satisfaction.
In this context, the compliance of selective distribution systems with competition law has been unsurprisingly questioned. In consequence, specific criteria of evaluation for the balancing of the contrasting aspects of selective distribution systems have been developed.17 The rationale has often been identified with the original lack of Block Exemptions for this category of vertical agreements and the consequent willingness of the European Court of Justice to establish specific criteria ‘to test its legality’.18 These allowed undertakings to escape the wide application of article 101(1) TFEU by the Commission, avoiding the long and complex procedure of notification and individual exemption under article 101(3)
TFEU, in sectors in which Selective Distribution System was not only commonly adopted, but also appreciated as highly beneficial for the efficiencies developed.19
Indeed, the early case law contributed to the understanding of the inextricable connection between the restrictive effects and efficiencies of selective distribution and to the determination of the boundaries of its compatibility with competition law. 16

Guidelines on Vertical Restrains (n 5) paras 185, 107(a), 107(d).
Metro (No 1) ( n 6), para 21.
18
Monti (n 1), 370.
19
Ibid, 371.
17

Global Antitrust Review 2014

169

The system has been considered appropriate only for the distribution of certain kinds of products, as it satisfies the request for higher quality by demanding purchasers. Customers of sophisticated goods, such as technically complex or branded products, have high expectations in terms of technical advice and assistance, pre and after sales, and in terms of image and appearance of the shopping environment, and are willing to pay a higher price in order to have it.20
For this reason, if the inter-brand rivalry on the market is lively, the reduction of intra-brand competition that selective distribution implies is not of great concern.
Customers who prefer lower prices and lower services can simply buy products from different suppliers. Consequently, the system may promote inter-brand competition. However, if there is ‘cumulative effect’, due to the presence on the relevant market of ‘parallel networks of selective distribution’, the risk of foreclosure of certain types of distributors increases, together with the possibility of collusion among suppliers.21
A manufacturer imposes on its distributors specific obligations, in order to ensure their compliance with requirements that it feels necessary for appropriate marketing and sale. The willingness to accept and implement them determines the admission to the system. The restriction not to supply non-members with contract products is necessary to obtain its functionality, provided that it only aims at ensuring the compliance of resellers with proportionate, objective, selection criteria
(and not, for example, at boycotting specific outlets, such as price discounters).
III. The pre-Pierre Fabre analysis of selective distribution systems under article 101 TFEU
The analytical approach suggested by the Commission for the assessment of vertical restraints in general and Selective Distribution Systems in particular is practical, as it implies, first, the evaluation of the applicability of a Block
Exemption Regulation and in case the agreement is not covered by a Block
Exemption Regulation, the need to establish whether it falls within article 101(1)
TFEU.22 Indeed, it must be remembered that, from the Commission’s point of view, when the presumption of legality does not apply because the market-share thresholds are exceeded, the agreement is not presumed illegal. An assessment pursuant to article 101(1) TFEU is required and if an infringement of that provision is established, the selective distribution system may still benefit from an individual exemption based on article 101(3) TFEU. In this case, the undertakings concerned must provide evidence of the fulfilment of the four cumulative conditions and the
20

Goyder, Albors-Llorens (n 1), 232.
Guidelines on Vertical Restraints (n 5) paras 176-178; Case 75/84, Metro v Commission
(No 2), [1986] ECR 03021, para 40.
22
Verticals Guidelines, n 5, para 110.
21

Selective distribution systems post-Pierre Fabre

170

exemption may apply as long as they are fulfilled, then it ceases.23 In principle, all agreements may benefit from an exemption under article 101(3) TFEU.
1. Appreciable restriction of competition by object or effect
A brief examination of the meaning and practical relevance of the distinction between restrictions by object and restrictions by effect is appropriate before considering the traditional approach of the EU Courts and the Commission to the competition analysis of selective distribution systems. Indeed, only agreements which have as their object or effect the prevention, restriction or distortion of competition are prohibited by article 101(1) TFEU and their overall impact on competition must be appreciable. While the expression ‘prevention, restriction or distortion’ is usually regarded as pleonastic and aimed at covering all possible nuances of interference, the distinction ‘object or effect’ is of utmost importance for the consequences it implies. It can be understood with reference to the case law of the European Courts and the Commission’s Guidelines on the application of
Article 81(3) EC Treaty [now 101(3) TFEU].24
‘Object or effect’ clearly identifies an alternative.25 The object must be considered first and if a restriction ‘by object’ is established, it is not necessary to demonstrate any actual effects on competition in order to determine an infringement of article
101(1) TFEU.26 Nor is it possible to prove the absence of any anti-competitive effects to escape the prohibition. The finding is definitive.27 The violation is presumed and efficiencies may be claimed and evaluated only under article 101(3)
TFEU.28 Only if the agreement does not have a restrictive object, its anticompetitive effects must be assessed.29 In both cases the burden of proof is on the party claiming the breach, but its weight is different as the inquiry into the actual effects on competition is complex and costly with uncertain outcomes.

23

Guidelines on the application of Article 81(3) of the Treaty, OJ [2004] C101/97, para 44.
OJ [2004] C101/97, paras 19-27.
25
Case 56/65, Société Technique Minière, [1966 ] ECR English Special Edition, p 249.
26
Case 56 and 58/64, Consten and Grundig, [1966] ECR English Special Edition, p 342.
27
David Bailey, ‘Restrictions of Competition’, (n 1), 559.
28
Unless the restraint is ‘objectively necessary for the existence of an agreement of a particular type or nature or for the protection of a legitimate goal, such as health and safety, and therefore falls outside the scope of article 101 (1)’, Guidance on restrictions of competition ‘by object’ for the purpose of defining which agreements may benefit from the
De Minimis Notice, 25 June 2014, p 4.
29
Société Technique Minière, n 26.
24

Global Antitrust Review 2014

171

An agreement is classified as restriction by object when its very nature has the potential to harm competition, causing injury to its proper functioning.30 However, this must be established with regard to a number of factors and, in particular, the content of its provisions, its objectives and the economic and legal context of which it forms part.31 The extent of their application is the subject of considerable debate. The subjective intention of the parties may be relevant too, but not necessary.32 Exemplifications may be found in article 101(1)(a-e) TFEU,33 in jurisprudence
(e.g., minimum resale price maintenance and absolute territorial or customer protection) and in Commission Block Exemption Regulations, Guidelines and
30

Case C-209/07, BIDS v Commission, [2008] ECR I-08637, para 17; Case C-226/11,
Expedia Inc v Authorité de la Concurrence, 13 December 2012, para 36. The agreement is considered ‘so likely to be harmful that the harm can be presumed’ on the basis of its nature, but also on experience (acquired in light of various factors, such as previous EU decisions, empirical research, experience of another jurisdiction and policy judgments, too), cf. Bailey, supra n 28, p 559, 564, and 101(3) Guidelines, n 24, para 21. The author highlights two other reasons for the prohibition of object restrictions, to promote legal certainty (enabling firms to understand the legal consequences of their conduct and, consequently, increasing the deterrent effect of competition law) and to facilitate the enforcement of article 101 (not having to prove actual effects).
31
Case C-501/06, GlaxoSmithKline v Commission, [2009] ECR I-09291, para 58; Case C32/11, Allianz Hungária v Gazdasági Versenyhivatal, 14 May 2013, para 36. It has been suggested that the requirement of such an assessment blurs the distinction between restrictions by object and by effect, as it is unclear the extent of market analysis to be conducted before finding an infringement by object or starting the full effects examination, cf. Jones and Sufrin, n 6, p 213. It has also been noticed that in both cases the definition of the relevant market is required, at least because of the necessity to evaluate (preliminary) the appreciability of possible effects on inter-state trade of the agreement, cf. Whish, Bailey
(n 1), 120. However, some authors specify that, even if it cannot be maintained that market definition is irrelevant in all object cases, it may be possible to find an object restriction without any market definition, cf Bailey n 28.
32
Cf Bailey, n 28; Pablo Ibanez Colomo, ‘Market failures, transactions costs and article
101 (1) TFEU case law’, (2012) 37 European Law Review, 541. Contra, Okeoghene
Odudu, ‘Interpreting Article 81(1): the object requirement revisited’, (2001) 26 European
Law Review, 379, according to whom ‘the ‘object’ criterion is satisfied and thus ‘effects’ need not be demonstrated when parties to a practice intend to restrict competition, this being justified on the grounds of preventative intervention against parties seeking, but yet to succeed, in restricting competition’; indeed, ‘without such intervention, the parties will eventually succeed in their attempts’ (p 388). He also maintains that ‘the intent can be proven in a number of ways’ and that ‘the reason why the Court’s treatment of different types of agreement appears to differ is not because different rules are applied’, ‘but because different factors constitute evidence of intention to restrict competition in different contexts’ (p 389), ‘whilst the same rule applies to all agreements, the probative value of facts and assumptions can vary in different contexts’ (p 390). Accordingly, ‘there is no predefined conduct or type of agreement that always satisfies the object criterion’ (p 389).
33
Yet, the nature of the examples mentioned in article 101(1)(d) and (e) as restrictions on competition by object has been questioned, cf. Bailey, n 28.

Selective distribution systems post-Pierre Fabre

172

Notices (as the hardcore restrictions are considered by the Commission as restrictions by object), but the lists are not exhaustive. They are considered a starting point, indicating restraints ‘highly likely to be found, in principle’ restrictive by object. However, they may vary, expanding or even narrowing the category of object restraints, not only with reference to a particular type of agreement following reconsideration by the Courts (for example ‘to reflect changes in economic thinking’), but also in a specific case ‘in relation to the specific facts/circumstances’, as the objectives and the context of an agreement are ‘also relevant to the assessment.34
It has been underlined that the instances in which an agreement is likely to be deemed restrictive of competition by object ‘are notoriously unclear’,35 while the
‘hard core’ and ‘underpinning logic’ of the notion is known, the determination of its boundaries is complex and it is felt difficult to identify an operational definition.36 Indeed, while the Commission tends to rely on the form of restraints
(the so-called ‘hardcore’ restraints), the CJEU insists in refusing the concept of a closed category of object restrictions (or ‘object box’), but, confusingly, it adopts a multiplicity of methodologies for their identification.37 As a consequence, sometimes its judgments appear contradictory.
It has been suggested that an alternative reading of the object/effect divide, more economics-oriented, could explain the jurisdictional approach.38 Considering that the private transactions examined under article 101 (1) TFEU may be an effective means to address market imperfections (such as market power, negative/positive externalities or information asymmetries) or transaction costs, many restrictions that appear anti-competitive are instead a source of efficiency gains.39 Consistently, a specific restraint will be deemed restrictive of competition by object not whether it has a particular form, nor whether it can be presumed to have anti-competitive effects, but only where ‘in the light of the nature of the agreement, and the context in which it is concluded’ it cannot be plausibly explained on efficiency grounds.40
This ‘default, efficiency-based methodological approach’ is allegedly not applied in two instances, concerning resale price maintenance and restrictions of parallel trade.41 Yet, these agreements are considered restrictive of competition by their very nature, irrespective of efficiency claims (in consideration of the fact that
34

Jones, Sufrin (n 1), 206, 212.
Colomo, n 33, 541.
36
Ibid, 546.
37
Ibid, 546, 547.
38
Ibid, 546, 548.
39
Ibid, 543, 544.
40
Ibid, 549.
41
Ibid, 552, 553.
35

Global Antitrust Review 2014

173

market integration is an EU competition law objective worthy of being protected in itself), only ‘in principle’, on the basis of a rebuttable presumption. The default methodology still applies in case of sufficiently compelling efficiency explanations
(for example, where, according to the analysis of the nature and context of the agreement, it is understood that efficiencies can only be achieved through such restrictions).42 However, some authors contend that the concept of anti-competitive object is not economic, but legal. In particular, it has been maintained that, in principle, ‘even a legal counsel with a rudimentary understanding of economics’ should be able to assess whether the agreement is anti-competitive or not, without the need to consider circumstances ‘beyond the four angles of the contract’.43 According to this opinion, the legal and economic context may be relevant only for the understanding of the meaning and function of the agreement and the anti-competitive object must be distinguished in relation to pre-determined categories, identified and developed by the jurisprudence on the basis of two considerations: the nature of agreements and the judicial experience. The only task of the analyser is to read the specific arrangement at stake and check whether it is blacklisted or not.44 There is no necessity of case-by-case analysis, otherwise the advantages of predictability and certainty would be lost.45 Notwithstanding the common belief in the legal and not economic nature of the term ‘object’ under article 101 (1) TFEU, other authors challenge this category-building or ‘object box’ approach as, they assert, it is not fully reflective of the case law of the CJEU. It is more consistent with statements of policy by the Commission and, therefore, it is better regarded as a guidance.46
Indeed, they identify at least two key methods in the jurisprudence of the Courts, the ‘orthodox’ (based on the object box concept) and the more analytical (effectsbased, as ‘established’ in Société Technique Minière, STM, and ’subsequently confirmed in Consten’ and Grundig).47 Allegedly, the CJEU currently follows a hybrid approach, according to which certain agreements can be regarded as
42

Ibid, 552.
Csongor Istvan Nágy, ‘The Distinction between Anti-competitive Object and Effect after
Allianz: The End of Coherence in Competition Analysis?’, (2013) 36 World Competition
4, 544, 555.
44
Ibid, 542, 544, 546, 555.
45
Ibid, 555. However, according to the author, the recent ruling of the CJEU in Allianz
Hungária, n 32, seems to contradict the traditional concept of anti-competitive object.
Widening the interpretation of the term ‘context’, it introduces a case-by-case assessment of the agreement, based on a ‘truncated’ effects-analysis, which, allegedly, ‘lifts the borderline between ‘object’ and ‘effect’ cases and eliminates the merits of legal certainty, predictability and simplicity’, p 558, 559, 561, 562.
46
Cf. Saskia King, ‘The Object Box: Law, Policy or Myth?’ (2011) 7 European
Competition Journal 2, 276, 270.
47
Ibid, p 277-9. Supra, n 26 and n 27.
43

Selective distribution systems post-Pierre Fabre

174

automatically restrictive of competition. Yet, the list is not exhaustive, nor is there an irrebuttable presumption ‘owing to the analysis of the ‘legal and economic context’ required under STM’ for each of them.48 Consequently, any restraint can be a restriction by object and, on the other hand, even ‘hardcore’ restrictions can be excused under the article 101(1) TFEU assessment.49 It must be noted that, according to this opinion, ‘the assessment of the ‘legal and economic context’ in accordance with the test in STM’ is a flexible process and the level of economic analysis depends on the facts of the case.50 Indeed, the extension to be given to the contextual examination of agreements appears to be the main element on which the opinions of commentators diverge.51
Whatever definition is adopted, when an agreement does not restrict competition by object, it must be assessed whether it has restrictive effects, applying a full
48

King, (n 47) 277, 291.
Ibid, p 291.
50
Ibid, p 295.
51
In this sense, cf. Cosmo Graham, ‘Methods for determining whether an agreement restricts competition: comment on Allianz Hungaria’, (2013) 38 European Law Review,
542. The author underlines that ‘what a court or competition authority is meant to do when considering the legal and economic context is unclear, except that it must mean something different from the exercise undertaken in relation to determining whether or not the agreement has the effect of restricting competition’, p 544. He mentions Allianz Hungaria, n 32, as illustrative of this uncertainty, since it seems to require a broader inquiry into the factual circumstances of the case than usually assumed, blurring the distinction between object and effect, p 545, 547. However, the judgment does not seem to represent a rethinking of the previous case law, but an alternative approach to be applied in a limited number of cases, when dealing with agreements which do not fall within the ‘object’ category identified by the jurisprudence of the courts. Consequently, it seems to suggest ‘a more searching form of inquiry’ for ‘novel arrangements’, whereas the ‘truncated in principle analysis’ is to be used with agreements that have already been considered object restrictions in the past, p 547, 549, 550. Allegedly, though, even if this could be a desirable development, it is unlikely that it will be repeated in later cases, given the ‘very traditional approach’ of the CJEU to object agreements, p 551. Also Alison Jones, ‘Left Behind by
Modernisation? Restrictions by Object under Article 101(1)’, (2010) 6 European
Competition Journal 3, 649, highlights the persistent formalism in the approach to ‘object’ restraints. Indeed, despite the modernisation process, which led the Commission to a more economics and effects-oriented interpretation of the EU competition law rules and of article
101 TFEU in particular, the ‘case-by-case analysis of the impact on competition of every agreement’ is still considered non-appropriate, while ‘some sorting of agreements, according to their likely competitive effects’ is recognised as necessary, p 654.
Accordingly, the Commission still identifies restrictions of competition by object as
‘hardcore’ restraints in its Guidelines and Block Exemption Regulations on the basis of the judicial experience only, whereas the case law provides that regard must also be had to the content, objectives and legal and economic context of agreements, p 656. This suggests that not only can the ‘object’ category be expanded, but also narrowed, p 657, 661, 664.
However, the jurisprudence on this latter point is considered hard to rationalise and it seems that the re-evaluation of ‘past precedents’ may happen only in ‘exceptional circumstances’, p 664, 668.
49

Global Antitrust Review 2014

175

economic analysis.52 There is, indeed, no presumption in this context. By contrast, it shall be determined if it has possible restrictive effects on actual or potential competition, so that ‘on the relevant market negative effects on prices, output, innovation or variety or quality of goods and services can be expected with a reasonable degree of probability’.53 The extensive analysis required must be conducted with the use of a (double) counter-factual, considering what the competition situation would have been in the absence of the specific agreement and of the contractual restraints, so as to assess the impact on inter-brand and intrabrand competition respectively.54
The Commission underlines that negative effects are likely when at least one of the parties has some degree of market power and the agreement contributes to its creation or exploitation. Details of its full competition analysis for vertical restraints are set out in the Verticals Guidelines.55
a) The balancing of anti and pro-competitive effects
It must be mentioned that a persistent inconsistency exists between Commission and European Courts in the interpretation of the effects analysis to be conducted under article 101(1) TFEU. Indeed, according to the Commission it should encompass only anti-competitive effects, whereas the balancing of offsetting efficiencies should take place more appropriately under article 101 (3) TFEU. 56 On the contrary, the case law shows, on the whole, a different approach, being open to the evaluation of both anti- and pro-competitive effects under article 101 (1)
TFEU.57
The reasons for this divergence can be traced back to the pre-modernisation era, when the Commission adopted a literal, broad interpretation of article 101 (1)
52

Case C-234/89, Delimitis v Henninger Bräu, [1991] ECR I-00935, paras 13 ss.
101(3) Guidelines, n 24, para 24.
54
Ibid, para 18. Cf. case T-328/03, O2 (Germany) v Commission, [2006] ECR II-01231, paras 71 ss.
55
101(3) Guidelines, n 24, paras 25, 27. Verticals Guidelines, n 5, paras 97, 110-121.
56
101(3) Guidelines, n 24, para 11.
57
Even if dissenting voices exist, cf. case T-112/99, Métropole Télévision (M6) v
Commission, [2001] ECR II-02459, para 74. In the sense of the text, Jones and Sufrin, n 6, p 196. Cf. Metro (No I), n 6, paras 20-22; Société Technique Minière, n 26, p 249-250; case
258/78, Nungesser v Commission, [1982] ECR 2015, paras 54-58; case 27/87, ErauwJacquéry v La Hesbignonne, [1988] ECR 1919, paras 8-11; case 42/84, Remia v
Commission [1985] ECR 2545, paras 19-20; Case 161/84 Pronuptia v Pronuptia [1986]
ECR 353, paras 15 ff; Case 23/67 Brasserie de Haecht (No 1) [1967] ECR 407, p 414-416; case C-250/92, Gøttrup-Klim e.a. Grovvareforeninger v Dansk Landbrugs Grovvareselskab
AmbA, [1994] ECR 5641, paras 31-45; case C-234/1989, Delimitis v Henninger Bräu,
[1991] ECR 935, paras 13-27.
53

Selective distribution systems post-Pierre Fabre

176

TFEU, had exclusive power of granting individual exemptions on the basis of an ex ante notification and only block exemption regulations by category existed.58 The jurisprudence of the CJEU developed, partly, as a remedy to the consequences of this procedure, allowing the evaluation of at least some pro-competitive aspects of allegedly infringing agreements even by the national authorities, in the absence of notification and absent the relief of block exemptions.
When Regulation 1/2003 entered into force,59 on 1 May 2004, the notification system was abolished and National Competition Authorities and Courts were allowed to apply article 101 TFEU in its entirety, so including article 101 (3)
TFEU. As a consequence, the procedural problematic aspects were overcome.
Additionally, the Commission’s approach to article 101 (1) TFEU changed, with the adoption of a less formalistic, more economic-oriented methodology.60
Yet, the debate is still heated, also because of the practical implications in terms of burden of proof and efforts and costs connected. Indeed, it lays on the person claiming the breach under article 101 (1) TFEU, but on the defendant under article
101 (3) TFEU.
It is worth underlining the inter-connection between the respective role of article
101 (1) TFEU and article 101(3) TFEU. A wide understanding of the former, inevitably exerts an influence on the latter. This poses the risk, in the
Commission’s view, to cast article 101 (3) TFEU aside, or at least to divert it from its purpose, ‘which is to provide a legal framework for the economic assessment of restrictive practices’.61 More precisely, if the analysis of anti and pro-competitive aspects is made under article 101 (1) TFEU, article 101 (3) TFEU risks to be used only to exempt agreements restrictive by object, or restrictions by effect justified by non-economic reasons.62
b) The appreciability condition
As already mentioned, the CJEU clarified that the prohibition of article 101 (1)
TFEU applies only when agreements restrict competition to an appreciable extent.63 In the Völk case, this requirement seemed to concern both restrictions by object and by effect.64 For this reason, taking into consideration its non-binding
58

Regulation 17/1962, OJ [1962] 13/204, arts 9 and 4.
Regulation 1/2003, OJ [2003] L 1/1, art 1, 3, 5, 7-10.
60
Between 1996 and 2004, cf. Jones and Sufrin, n 6, p 193.
61
White Paper on the Modernisation of the Rules Implementing Articles 85 and 86 of the
EC Treaty [101 and 102 TFEU], OJ [1999] C132/1, paras 56-57.
62
Jones and Sufrin, n 6, p 199.
63
Case 22/71, Béguelin Import Co. v S.A.G.L. Import Export, [1971] ECR 949, para 16.
64
Case 5-69, Völk v Vervaecke, [1969] ECR 295, para 57.
59

Global Antitrust Review 2014

177

nature, the previous Commission ‘De Minimis’ Notice (stating that the presumption of non appreciability could not be applied to agreements containing hardcore restrictions) was usually interpreted as simply not dealing with the problem of identifying when restraints by object were of no significance.65 Yet, subsequently, in the Expedia case, the CJEU stated expressly that an agreement with an anticompetitive object constitutes, by its nature and independently of any concrete effect it may have, an appreciable restriction on competition.66
On the 25th of June 2014, the Commission issued a revised ‘De Minimis’ Notice, where, taking into consideration the recent judgment of the CJEU, clarified that restrictions by object can never benefit from its safe harbour, as they cannot be considered minor and always constitute an appreciable restriction of competition.67
Contrary to its predecessor, it does not contain any list, but it mentions some examples (fixing of prices, limitation of output or sales, allocation of markets or customers) and makes reference to the hardcore restrictions contained in any current and future Commission Block Exemption Regulation, as expressly considered to generally constitute restrictions by object. Additionally, the Notice is accompanied by a Staff Working Document, named ‘Guidance on restrictions of competition ‘by object’ for the purpose of defining which agreements may benefit from the De Minimis Notice’, which lists the restrictions described as ’by object’ or
‘hardcore’ in Commission Regulations, Guidelines and Notices, with examples taken from case law and decisional practice.68
Consequently, only restrictions ‘by effect’ may be presumed non-appreciable by reference to (unchanged) market share thresholds, 10% for the parties to horizontal agreements (aggregate) and 15% for each party to a vertical restraint. As before, the 10% provision applies in case of difficult classification of the agreement and all the market thresholds are reduced to 5% where the cumulative effect of parallel networks of agreements restricts competition (unlikely, if they cover less than 30% of the relevant market). The agreement continues to be considered not appreciable even if the market shares of the parties outgrow the thresholds, by no more than 2% points, during two successive calendar years.

65

Notice on agreements of minor importance which do not appreciably restrict competition under Article 81(1) [101(1)] (De Minimis Notice), OJ [2001] C368/13, para 11.
66
Expedia, n 31, para 37.
67
Notice on agreements of minor importance which do not appreciably restrict competition under Article 101(1) (De Minimis Notice), OJ [2014] C 291/01, paras 13, 2.
68
SWD
(2014)
198,
,
accessed 28 November 2014.

Selective distribution systems post-Pierre Fabre

178

The Notice is not binding, but meant to give guidance (also to national courts and competition authorities), without prejudice for the CJEU’s interpretation of article
101 (1) TFEU. Anyway, it creates legitimate expectations and the Commission cannot disregard them without reasons. The accompanying document expressly states to be intended not to be an obstacle to any developments of case law and decisional practice.

2. The EU Courts’ traditional approach to selective distribution systems
As already mentioned, it is traditionally understood that the compliance with certain specific qualitative criteria determines the legality of selective distribution systems and since the 1970s the jurisprudence of the European Courts developed peculiarly around them. They are the so-called Metro criteria, after the name of the case where first established.69
In their presence, the system was considered productive of efficiencies capable of counterbalancing the inherent restrictions affecting the economic freedom of dealers and the reduction of intra-brand price-competition. Indeed, in the Metro case, the CJEU expressly stated that price competition is not generally emphasised in such systems, however it is not the only effective form of competition, which may be reconciled with objectives of different nature.70
On the contrary, quantitative criteria of selection were usually believed not appropriate to save the system, as potentially conducive to additional anticompetitive inefficiencies, in direct contrast with the EU objectives (protection of competition and market integration). They more directly limit the number of dealers, leading to potential territorial partitioning, without producing counterbalancing pro-competitive effects.71
However, even a ‘qualitative’ Selective Distribution Agreement may infringe article 101 (1) TFEU, in case of cumulative effect due to the presence of similar networks on the relevant market. This principle, developed in the Metro No 2 case, does not automatically imply a prohibition, depending on the overall actual net

69

Metro (No 1), n 6, paras 20, 21.
Metro (No 1), n 6, para 21.
71
Verticals Guidelines, n 5, paras 7, 175.
70

Global Antitrust Review 2014

179

effect on the competitive situation.72 The Commission quantifies its significance in the Verticals Guidelines.73
a) The Metro criteria
The subsequent interpretation by the case law specified the meaning and boundaries of the Metro criteria, adding further qualifications. The Verticals
Guidelines mention them as follows:
1 - the nature of the product must necessitate a selective distribution system, to preserve its quality and ensure its proper use;
2 - resellers must be selected on the basis of objective criteria of a qualitative nature, laid down uniformly for all potential distributors and not applied in a discriminatory manner;
3 - the criteria must not go beyond what is necessary.74
In relation to the first criterion, it has been noticed that many goods have been taken into consideration by the EU Courts and the EU Competition Authority and there are not precise rules, only trends.75 Usually, three categories of products are mentioned, two well-known (technologically complex and luxury/branded goods), as their need for high-quality shopping environment, promotion and assistance is evident, and one more peculiar (newspapers), where the requirement for Selective
Distribution System is connected with the characteristics of the distribution, due to the short shelf-life.76
It has been held that the nature of the product encompasses not only its material composition, but also the aura of prestige and luxury that characterises its brand image and represents what is wanted by customers and promoted by manufacturers and distributors.77 Furthermore, it appears from the case law that the necessity of a
Selective Distribution System cannot be maintained where public rules already exist and provide goods with the necessary protection, so that private contractual rules become superfluous.78 This may happen when EU or national rules concern
72

Metro (No 2), n 21, paras 40, 41. Cf. Metro (No 1), n 6, para 50.
Verticals Guidelines, n 5, para 179.
74
Verticals Guidelines, n 45, para 175.
75
Bellamy and Child, n 67, p 456; Jones and Sufrin, n 56, p 810.
76
Cf. Metro (No 1), n 6, para 20; Metro (No 2), n 21, para 54; Leclerc, n 7, paras 107-117;
Case 243/83, SA Binon v Agence Messageries de la Presse, [1985] ECR 2015, paras 31-33.
77
Case T-19/92, Groupement d'Achat Edouard Leclerc (YSL) v Commission, [1996] ECR
II-01851, paras 114-123.
78
Case 31/80, L'Oréal, [1980] ECR 3775, para 16.
73

Selective distribution systems post-Pierre Fabre

180

the admission to the trade or conditions of sale of certain products, such as pharmaceuticals. The second Metro criterion concerns the selection of resellers, making an important distinction between qualitative and quantitative parameters. This influences the legitimacy of the mechanism, but the dividing line is considered difficult to trace.
Mainly, qualitative criteria concern the objective suitability to distribute the specific type of product (e.g. technical qualifications, apt premises, setting or location, after-sales services), in contrast with a more direct limit on the potential number of dealers (such as specific amounts of stocks, minimum annual turnover, etc., but with some overlaps).79 It has been said that the latter aims at protecting approved members from other retailers meeting the same qualitative criteria.80
The criteria must be laid down uniformly for all potential distributors and applied in a non-discriminatory way. In principle, all dealers meeting them should be admitted and able to obtain the contract goods. A refusal to supply those who meet the criteria discriminates and infringes article 101 (1) TFEU, if made in agreement with other members of the system (otherwise it represents a unilateral conduct, legitimate unless abusive of dominant position). However, the penalty may only consist of a fine, not an order to supply.81
Interestingly, the CJEU held in a recent case that, in order to avoid a risk of conflation, the reciprocal characteristics of qualitative and quantitative criteria must be distinguished. In particular, the content of quantitative criteria may be verified, but it is not necessary that they are ‘objectively justified and applied in a uniform and non-differentiated manner’.82 The context of the judgment was very peculiar, connected with a question of applicability of a specific exemption regulation for the motor vehicle sector, but the importance of the statement allows its use as a general reference.83
Finally, the third Metro criterion expresses the necessity of proportionality between selection criteria and contract goods. Indeed, it requires that the restrictions on which the Selective Distribution System is based (i.e. selection criteria of resellers) do not go beyond what is essential to the protection of quality and proper use of the product. Some authors underlined its increasing importance in the evaluation of the
79

Verticals Guidelines, n 5, para 175.
Jones and Sufrin, n 6, p 811.
81
Case T-24/90, Automec v Commission, [1992] ECR II-2223, para 51.
82
Case C-158/11, Auto 24 SARL v Jaguar Land Rover France, 14 June 2012, paras 30, 32.
83
Louis Vogel, ‘Efficiency versus Regulation: The Application of EU Competition Law to
Distribution Agreements’, (2013) 4 Journal of European Competition Law & Practice 3,
277.
80

Global Antitrust Review 2014

181

legality of the system, as the category of suitable goods by nature has not clear confines and seems subject to unpredictable expansion.84
It appears from the case law of the CJEU that, when the Metro criteria are not satisfied, Selective Distribution System infringes article 101 (1) TFEU. However, as previously mentioned, according to the Commission’s view expressed in the
Vertical Guidelines, further economic factors might be analysed.
3. The Commission’s approach
It has been underlined by authoritative doctrine with general reference to the category of vertical restraints, which Selective Distribution Agreements belong to, that: Presently, there is a broad consensus among economists that vertical restraints are neither always harmful nor always beneficial, and that a determination of whether a vertical restraint reduces economic welfare requires an appraisal of all relevant facts (…).
The few empirical economic studies that have been carried out suggest that in general vertical restraints are efficiency enhancing, but do not allow one to identify a priori all necessary and sufficient conditions to be able to identify when vertical restraints are beneficial (…).
In sum, for each pro-competitive justification for a vertical restraint
(facilitating entry, resolving the free-rider problem, or removing ‘hold-up’ risks) there can be an adverse effect on distributors and competing manufacturers, suffocating intra- and inter-brand competition.85
Accordingly, it is commonly agreed that an absolute aprioristic answer (i.e.
‘always’, or ‘never’, or ‘it depends on specific conditions’) to the question of their compatibility with competition law is inappropriate. They should be assessed considering facts and circumstances of each particular case. After the so-called modernisation of the late 1990s, this position has also been assumed by the
Commission, as confirmed in its Guidelines,86 based on the principles developed by the peculiar case law of the time.87 Here, with specific reference to Selective
84

Goyder, Albors-Llorens, (n 1), 240-1.
Monti, n 15, p 348; cf. Damien Neven, Penelope Papandropoulos and Paul Seabright,
Trawling for Minnows: European Competition Policy and Agreements between Firms
(CEPR, 1998), p 17.
86
Verticals Guidelines, n. 5, paras 175-185, cf. paras 110-121.
87
The reference is to the Metro doctrine. Although the Commission does consider those specific criteria when assessing the possible anti-competitive effects of Selective
85

Selective distribution systems post-Pierre Fabre

182

Distribution System, it emphasises the risk of reduction in intra-brand competition, usually outweighed by sufficient inter-brand competition, unless the number of similar networks on the market is high with consequent detrimental development of the so-called ‘cumulative effect’. Then, it mentions the risk of foreclosure of certain types of distributors and of softening of competition and facilitation of collusion between suppliers or buyers. Yet, additionally, it should not be forgotten the risk of creation of obstacles to market integration, one of the main objectives of the EU law88.
According to the Guidelines, competition risks should be evaluated in the light of article 101(1) TFEU, when the agreement is not exempted under the Block
Exemption Regulation, undertaking a full effects analysis in relation to various factors, which may have an influence in each specific case. Yet, recent developments in the jurisprudence of the European Court of Justice on Selective
Distribution System cast some doubts on this methodology.89
a) Factors of analysis
The relevant factors identified may consist, firstly, in the criteria adopted for the selection of distributors (qualitative or quantitative). This is a very significant point, to be firmly underlined, as it represents the recognition of the Metro criteria, developed, as above mentioned, by the early case law in order to discern the legitimacy of Selective Distribution System. It shows that this doctrine is deeply rooted in the analysis of the topic and bound to persist in its influence on the subject. The Commission confirms that a preliminary distinction between purely qualitative and quantitative Selective Distribution System is required, as the first is
‘in general considered to fall outside article 101 (1) TFEU for lack of anticompetitive effects, provided that three conditions are satisfied’:
-‘the nature of the product must necessitate a selective distribution system’, to ‘preserve its quality and ensure its proper use’;
-resellers must be selected ‘on the basis of objective criteria of a qualitative nature’, ‘laid down uniformly’ for all potential distributors and ‘not applied in a discriminatory manner’;

Distribution System under article 101(1) TFEU, it does not rely exclusively on them. They only represent some of several factors to be examined, relevant but not conclusive.
88
Ibid, paras 100 and 7.
89
Pierre Fabre, n 3, para 39; the point will be developed further in the following sections.

Global Antitrust Review 2014

183

-the criteria ‘must not go beyond what is necessary’.90
Secondly, the assessment of the agreement requires the market position of the parties to be considered. If the position of the supplier is strong, the loss of intrabrand competition may be problematic, whereas the presence of buying power may increase the risk of collusion between dealers and lead to the foreclosure of other, more efficient, retailers, e.g. through the imposition of preferred selection criteria on the supplier to their advantage.91
Then, the position of the supplier’s competitors has to be evaluated. As mentioned, the presence of strong competitors may imply that inter-brand competition is sufficient to outweigh the loss of intra-brand competition. Yet, the overall effect on competition may vary according to the number of similar networks operating on the same market. If only one supplier adopts this model of marketing, even the operation of a quantitative selection may not create net negative effects, at specified conditions (i.e. when the nature of contract goods requires the system and the selection criteria applied are necessary for their efficient distribution).
However, when the majority of competitors apply the same system, they may cause a significant loss of intra-brand competition, possible foreclosure of certain types of distributors and increased risk of collusion between the major suppliers.
Cumulative effects are believed non-problematic where Selective Distribution
System covers less than 50% of the relevant market, or the aggregate market share of the five largest suppliers (CR5) does not exceed 50%. When both the conditions are not satisfied, it should be assessed whether all the largest five suppliers apply the Selective Distribution System. A positive finding may give rise to competition concerns. However, the individual contribution to a cumulative effect is not considered significant whether the supplier’s market share is below 5%.
The foreclosure of suppliers is not regarded as a typical problem of Selective
Distribution System, as, usually, dealers are allowed to resell competing products.
It may happen when non-compete obligations combine with a dense network of authorised distributors, or cumulative effects. Alternatively, even in the absence of such clauses, the drawback may be caused when the main suppliers adopt nonpurely-qualitative criteria of selection, imposing certain quantitative requirements, such as minimum shelf-space in favour of their products. 92
Additionally, the analysis under article 101 (1) TFEU of a suspicious agreement should consider entry barriers, in particular when facing foreclosure of nonadmitted resellers. In case of branded products, as it is usual within a Selective
90

Verticals Guidelines, n 5, para 175.
Ibid, paras 177, 181.
92
Ibid, paras 177-179, 183.
91

Selective distribution systems post-Pierre Fabre

184

Distribution System, they are likely to be relevant, as time and consistent investments are said to be required to launch a new brand or conquer a different satisfactory source of supply.93 Finally, the maturity of the market should be appreciated (as the more mature it is, the more problematic the loss of intra-brand competition and the possible foreclosure could be), together with all those other factors that may be of relevance in a specific case.94

b) Efficiencies
The Commission also highlights some of the efficiencies the Selective Distribution
System may realise. These are economies of scale in transport and consequential savings of logistical costs (described as only marginal), contribution to the solution of a free-rider problem or to create a brand image (mainly in connection with new, complex or ‘experience’ and ‘credence’ products, which present qualities difficult to judge before and sometimes even after their consumption) and, at specified conditions, also the protection of relationship-specific investments made by authorised dealers.95
It believes that their appraisal should take place more appropriately within the framework of article 101 (3) TFEU.96 However, this is a controversial point, as the case law of the CJEU often adopted a different approach, balancing anti and procompetitive effects under article 101 (1) TFEU.
IV. The Pierre Fabre judgment: competition analysis of selective distribution systems revisited?
1. Background
On the 13rd of October 2011 the CJEU issued an important judgment on the subject matter, in the case C-439/09, Pierre Fabre Dermo-Cosmétique SAS v
Président de l’Autorité de la concurrence. The proceedings concerned a reference for preliminary ruling on the interpretation of Article 81 (1) and (3) EC Treaty
[now, article 101 (1) and (3) TFEU] and the Verticals Block Exemption Regulation

93

Ibid, para 180.
Ibid, paras 184, 110, 121.
95
Ibid, para 185.
96
Ibid, para 122. Cf. Guidelines on the application of Article 81(3) of the Treaty, OJ [2004]
C101/97, para 11.
94

Global Antitrust Review 2014

185

2790/1999,97 with particular attention to the provision of article 4 (c), excluding the application of the exemption to agreements having as their object the restriction of active or passive sales to end users by members of a selective distribution system operating at the retail level of trade, without prejudice to the possibility of prohibiting a member of the system from operating out of an unauthorized place of establishment.
The reference was made under article 267 TFEU by the Court of Appeal of Paris, 98 in an action for annulment or amendment brought by Pierre Fabre DermoCosmétique SAS against a decision99 by the French Competition Authority (NCA).
This sanctioned a clause contained in the Pierre Fabre DC’s Selective Distribution
Agreements which was considered a de facto general and absolute ban on internet selling to end users by authorised distributors.
Pierre Fabre Dermo-Cosmétique was a company in the Pierre Fabre group, producing and distributing cosmetics and personal care products through various subsidiaries and under different brands on the French and European market. The products were sold mainly through pharmacies, even if they were not classified as medicines and, consequently, not covered by monopoly.
Its Selective Distribution Agreements stipulated that sales had to be made exclusively in a physical space and at the presence of a qualified pharmacist,100 de facto excluding all forms of internet transactions. During the NCA’s administrative proceedings, Pierre Fabre DC insisted that the products, by their very nature, required personalised advice from a qualified pharmacist and that this was the reason of the ban of their sale via the internet.101
In its final decision, the National Competition Authority underlined that the ban on internet sales was a limitation of the distributors’ commercial freedom and of consumers’ choice, which prevented sales to final customers not located in the
‘physical trading area’ of the authorised distributors. Consequently, it assessed the prohibition as a restriction of competition by object, additional to those limitations typically contained in a Selective Distribution System (selection of dealers and prohibition to supply non-approved distributors).

97

OJ [1999] L 336/21.
Decision of 29 October 2009, received at the CJEU on 10 November 2009.
99
Decision No 08-D-25 of 29 October 2008.
100
Arts 1.1 and 1.2 of the general conditions.
101
Cf. Pierre Fabre, n 3, opinion of Advocate General Mazak, para 7.
98

Selective distribution systems post-Pierre Fabre

186

The block exemption was not applied, notwithstanding the market-share threshold was not exceeded. Indeed, even if the prohibition of internet selling was not expressly mentioned in the 1999 Regulation, it was considered equivalent to a ban on active and passive sales by members of a Selective Distribution System, contemplated (and not permitted) by article 4 (c). Furthermore, it was stated the inapplicability of its claimed exception (possibility of prohibiting approved dealers from operating ‘out of an unauthorised place of establishment’), as the National
Competition Authority believed that the internet is not a place where goods are marketed, but an alternative means of selling.
Finally, the French Competition Authority stated that Pierre Fabre had not demonstrated the existence of the conditions for an individual exemption under article 101 (3) TFEU, in particular rejecting its claims concerning risks of counterfeiting and free-riding between pharmacies and consumer’s welfare.
Consequently, it held article 101 TFEU infringed and ordered the removal from the selective distribution contracts of all terms equivalent to a ban on internet sales, while making express provision of this optional method of distribution.
On the 24th of December 2008, Pierre Fabre DC challenged the decision and brought an action for its annulment or amendment. The Court of Appeal of Paris decided to stay the proceedings and referred the following question to the CJEU for a preliminary ruling:
Does a general and absolute ban on selling contract goods to end-users via the internet, imposed on authorised distributors in the context of a selective distribution network, in fact constitute a ‘hardcore’ restriction of competition by object for the purposes of Article 81(1) EC [Article 101(1) TFEU] which is not covered by the block exemption provided for by Regulation No
2790/1999 but which is potentially eligible for an individual exemption under Article 81(3) EC [Article 101(3) TFEU][?].102
The CJEU subdivided it into three questions:
-whether the contractual clause at issue in the main proceedings amounts to a restriction of competition ‘by object’ within the meaning of Article 101(1)
TFEU,
-whether a selective distribution contract containing such a clause – where it falls within the scope of Article 101(1) TFEU – may benefit from the block exemption established by Regulation No 2790/1999,
102

Pierre Fabre, n 3, para 31.

Global Antitrust Review 2014

187

-whether, where the block exemption is inapplicable, the contract could nevertheless benefit from the exception provided for in Article 101(3)
TFEU.103
2. Peculiarities of the judgment
The CJEU begins the analysis of the submitted question with a reference to the concept of ‘hardcore restrictions’, mentioned by the referring court, and specifies that it is not a term included in the Treaty, or in the 1999 Block Exemption
Regulation.104 It is presumably a way to reaffirm the necessity of a rigorous approach in the analysis of the anti-competitiveness of a contractual clause, in contrast with the practical methodological approach promoted by the EU
Commission which starts with the exam of the applicability of a Block Exemption before assessing the infringement of article 101(1) TFEU.105
Advocate General Mazak explains in his opinion that hardcore restrictions are often confused with restrictions of competition by object, whereas they are distinct legal concepts .106 ‘Hardcore’ is a notion functional to the application of the Block
Exemption Regulation, used by the EU Commission to indicate the restrictions black-listed and leading to the exclusion of the agreement from its scope.107 They often result in a finding of restriction by object, but there is not such a legal presumption. On the contrary, restrictions by object may not be established using abstract formulas. They require individual assessment, which may be abridged sometimes, but never dispensed with.
Indeed, in the Pierre Fabre judgment, the CJEU recalls that an agreement infringes article 101(1) TFEU when its ‘object or effect’ restricts competition and that, given the alternative nature of the requirement, confirmed by settled case law, a proper assessment requires, as a preliminary step, the consideration of its purpose and economic context. In fact, if an anti-competitive object is established, it is superfluous to consider the effect.108
Consistently, the evaluation of the contractual clause as a restriction by object implies the preliminary analysis of its content (sales of cosmetics and personal care
103

Ibid, para 33.
Ibid, para 32. However, the concept is mentioned in the 2010 Regulation No 330, n 2, art. 4.
105
Verticals Guidelines, n 5, para 110.
106
Pierre Fabre, n 3, opinion of Advocate General, paras 24 ss.
107
In this case the reference is to the (2000) Verticals Guidelines, OJ [2000] C 291/1, cf. paras 46 ss.
108
Pierre Fabre, n 3, para 34. Cf. Société Technique Minière, n 26, para 249;
GlaxoSmithKline, n 32, para 55.
104

Selective distribution systems post-Pierre Fabre

188

products made in a physical space with the presence of a qualified pharmacist),109 objectives (de facto prohibition from any form of internet selling, which, by de facto excluding a method of marketing that does not require the physical movement of the customer, considerably reduces the ability of a distributor to sell to customers outside its contractual territory)110 and economic and legal context
(Selective Distribution).111
The CJEU concludes that the clause is liable to restrict competition by object, unless justifiable112 and it points out the parameters for the further analysis of
‘justifiability’ by reminding the characteristics of Selective Distribution
Agreements.113 It asserts that they necessarily affect competition and that, however, it has always been recognised in the case law the existence of requirements that may legitimate them, justifying a ‘reduction of price competition in favour of competition relating to factors other than price’.114 Consequently, if they aim at a
‘legitimate goal capable of improving competition in relation to factors other than price’, they are in conformity with article 101 (1) TFEU. The CJEU clearly identifies such requirements with the Metro criteria.115
It is a duty for the referring court to examine the possible justification of the specific clause in the light of those criteria, but the CJEU provides some further guidance on the interpretation of the EU law. It underlines that, in the specific case, the qualitative requirement for the selection of resellers is undisputedly satisfied
(the second Metro criterion), but that it must still be established whether the restrictions pursue ‘legitimate aims in a proportionate manner’ (first and third
Metro criteria).116
It suggests that the test may not be satisfied. Indeed, in relation to the nature of the products, a ban on internet sales may not be appropriately justified by the alleged aim to provide individual advice to customers and ensure their protection against the incorrect use.117 It lacks proportionality, as specific precedents in the Internal
Market case law indicate with reference to non-prescription medicines and contact

109

Pierre Fabre, n 3, paras 35, 36.
Ibid, paras 35, 37-38.
111
Ibid, paras 35, 39.
112
Ibid, para 47.
113
Ibid, paras 39, 40.
114
Cf. Case 107/82, AEG v Commission, [1983] ECR 3151, para 33
115
Pierre Fabre, n 3, para 41.
116
Ibid, paras 42, 43.
117
Ibid, para 44.
110

Global Antitrust Review 2014

189

lenses.118 Nor might it be justified by the other alleged aim, concerning the maintenance of the prestigious image of the products.119
3. The importance of the judgment for competition analysis under Article 101
(1) TFEU in general and for Selective Distribution System in particular
The Pierre Fabre judgment confirms the alternative nature of the ‘object or effect’ requirement of restrictions on competition under article 101 (1) TFEU and the fact that, in drawing the line between them, the CJEU does not rely on the mere form of restraints, but, in coherence with settled case law, it adopts a case-by-case approach, considering their content, objectives and legal and economic context.120
It has been suggested that also the ‘emerging principle of objective justification under article 101 (1) TFEU supports the consideration that the form of an agreement cannot be decisive when determining its object.121
If a restriction by object is consequently established, the infringement of article
101(1) TFEU is presumed without any demonstration of their actual effects on competition. Only whether a restriction by object cannot be assessed, a thorough examination of anti-competitive effects should take place.
The distinction has acquired enhanced relevance by virtue of the recent judgment in the Expedia case, where the CJEU expressly stated that ‘an agreement that has an anti-competitive object constitutes, by its nature and independently of any concrete effect that it may have, an appreciable restriction on competition’.122 As clarified by the EU Commission in the 2014 ‘De Minimis’ Notice, a restriction of

118

Cf Case C-322/01 Deutscher Apothekerverband [2003] ECR I-14887, paras 106-107,
112; case C-108/09, Ker-Optica, [2010] ECR I-12213, para 76.
119
Pierre Fabre, n 3, paras 45-46.
120
Pierre Fabre, n 3, paras 34-35.
121
Bailey, n 28,561, 580-1. The author underlines that the concept of ‘objective justification’, used in the Pierre Fabre judgment, can be found in many judgments under article 102 TFEU and under the free movement provisions, but it appears to be the first time it is adopted under article 101(1). In this context, he highlights, the principle seems in contradiction with previous case law, according to which an agreement may have a restrictive object even if it also pursues other legitimate objectives. He suggests that a reconciliation is possible considering the principle as a means to determine when a prima facie restrictive conduct falls outside article 101(1), rather than to determine whether a restriction by object can be saved by it. However, it is appreciated that the same principle can be found in the Verticals Guidelines, n 4, para 60. According to Colomo, n 33, p 550, the same principle of ‘objective justification’ may also confirm the ‘efficiency’ interpretation of the object/effect divide, as suggestive of ‘a solid understanding’ by the
CJEU ‘of the importance of some restraints to preserve the value of the intangible property that is associated with the products covered by the selective distribution’.
122
Expedia, n 31, para 37.

Selective distribution systems post-Pierre Fabre competition by object can never presumption.123 benefit from its

190

‘non-appreciability’

However, the Pierre Fabre judgment bears a particular meaning in relation to
Selective Distribution System. Not only it officially recognises the importance, in this context, of the use of the internet as a method of marketing the contractual products alternative to traditional sales, but it also presents two relevant elements which are surprising and, if intentional, may have the potential of overruling the traditional case law on Selective Distribution System, or at least its traditional interpretation: the statement that ‘agreements constituting a selective distribution system’ are to be considered, ‘in the absence of objective justification, as
‘restrictions by object’’ (para 39) and the assertion that ‘the aim of maintaining a prestigious image is not a legitimate aim for restricting competition and cannot therefore justify a finding that a contractual clause pursuing such an aim does not fall within article 101(1)’ (para 46).
a) Selective distribution systems as restrictions by object?
Indeed, Selective Distribution Systems used to be considered ‘restrictions of competition by effect’, where the Metro criteria, developed by the case law of the
CJEU, were adopted as a peculiar mechanism for conducting the effects analysis, balancing anti-competitive and pro-competitive aspects in order to assess the infringement of competition law under article 101(1). On the contrary, in the Pierre
Fabre judgment, they are qualified as ‘restrictions by object, unless objectively justified’. In this context the Metro criteria are still used, but as a mechanism to establish the existence of an objective justification.124
The perspective seems inverted, although the substance and legal result appear the same.125 Selective Distribution Systems are still assessed under article 101 (1)
TFEU, in the light of the traditional criteria. However, if the new interpretation were confirmed, new developments might potentially arise, for example in connection with a different interpretation of the nature of the objective justification suitable for their analysis (e.g. public, instead of private).126

123

2014 ‘De Minimis’ Notice, n 68, paras 13, 2.
Cf. Jones and Sufrin, n 6, p 815.
125
Giorgio Monti, ‘Restraints on Selective Distribution’ (n 1), 489.
126
In this sense, cf. Pierre Fabre, n 3, opinion of Advocate General, para 35, ‘In my view, the legitimate objective sought must be of a public law nature and therefore aimed at protecting a public good’ and Bailey, n 28, p 580 (footnote n 147), where it is maintained that paras 40-41 of the Pierre Fabre judgment appear ‘to explain the Metro doctrine as an example of an objective justification for a selective distribution system’.
124

Global Antitrust Review 2014

191

Yet, it could be possible to identify the reason of the change in a simple mistake, perhaps a linguistic oversight, also because it is contained in only two lines at the end of paragraph 39. Some authors have already referred to it as a probable ‘slip of the tongue’.127 Indeed, a certain degree of confusion seems to arise from the
‘putting the clause into its context’ exercise made by the Court, which might cause the attribution to Selective Distribution Agreements in general of qualifications appropriate for the sole contractual clause (‘restriction by object unless objectively justified’) and vice versa (Metro criteria within the concept of ‘objective justification’). b) The aim of maintaining a prestigious image
The same might be in relation to the second assertion mentioned (‘the aim of maintaining a prestigious image is not a legitimate aim for restricting competition’), which is expressed in absolute terms and seems referred to the
Selective Distribution System overall.128 Otherwise, it is not easily understandable how the CJEU could overrule the settled case law on this topic in only few lines, without explanation.
Indeed, Selective Distribution Systems are traditionally used for the marketing of luxury and branded products and it is commonly accepted that ‘the quality of luxury goods (…) is not just the result of their material characteristics, but also of the allure and prestigious image which bestows on them an aura of luxury’, so that
‘an impairment to that aura of luxury is likely to affect the actual quality of those goods’.129 Additionally, it has been acknowledged that: the concept of the 'characteristics' of luxury cosmetics, within the meaning of the judgment in L'Oréal, cannot be limited to their material characteristics but also encompasses the specific perception that consumers have of them, in particular their 'aura of luxury'.130
A possible alternative interpretation could be that this aim is not a sufficient justification for a ban on internet sales, in general, as the latter could lead ‘to partitioning of markets along national lines’, ‘regarded as a serious restriction of competition under EU law’.131 Yet, more convincingly, it could be understood that, according to the CJEU, the aim of maintaining a prestigious image is not an
127

Monti (n 126)501.
Cf. Jorren Knibbe, ‘Selective distribution and the ECJ's judgment in Pierre Fabre’, Case
Comment, (2012) 33 European Competition Law Review 10, 450; Monti, n 126, p 502.
129
Case C-59/08, Copad v Dior, [2009] ECR I-3421, paras 24, 26.
130
Leclerc, n 7, para 109. Cf. in this sense, Pierre Fabre, n 3, Opinion of Advocate General
Mazak, 3 March 2011, paras 44-45, 54.
131
Cf Knibbe, n 129, p 450-1.
128

Selective distribution systems post-Pierre Fabre

192

appropriate justification for a general and absolute ban on internet selling in the specific case, in relation to the nature of the contract goods (cosmetics and personal care products). Consequently, it would simply allude to a specific lack of proportionality of the restriction adopted in relation to the products concerned, on the basis of the same reasoning followed for the alleged aim of ensuring their proper use.132
This interpretation appears more consistent with the wording of the first answer to the referring court: in the context of a selective distribution system, a contractual clause …, resulting in a ban on the use of the internet for those sales, amounts to a restriction by object … where, following an individual and specific examination …, it is apparent that, having regard to the properties of the products at issue, that clause is not objectively justified.133
c) De facto ban on online sales
After the specification of the criteria to be adopted for the assessment of a relevant restriction of competition, the CJEU faces the possible application of the Block
Exemption Regulation. It notices that the market-share threshold is not exceeded, but the contractual clause represents a forbidden restriction. Indeed, by de facto prohibiting internet selling as a method of marketing, it has at least the object of restricting passive sales to consumers ‘wishing to purchase online and located outside the physical trading area’ of the authorized distributors.134 It does fall within the provision of article 4 (c), contemplating restrictions of active or passive sales to end users by members of a Selective Distribution System operating at the retail level of trade.
It is also established that its specific exception (admitting restrictions on nonauthorized places of establishment) is not pertinent, as the notion of ‘place of establishment’ concerns ‘only outlets where direct sales take place’. Thus, it does not include the internet, nor the place from which its services are provided. A broader interpretation (encompassing the latter) is not necessary, as, it is stated, an undertaking may always have recourse, on an individual basis, to the exception of article 101 (3) TFEU in order to protect its rights.135

132

Cf Jones & Sufrin, n 6, p 815 (footnote 266).
Pierre Fabre, n 3, para 47. (emphasis added)
134
Ibid, para 54.
135
Ibid, paras 56-57.
133

Global Antitrust Review 2014

193

Finally, the CJEU confirms that the Block Exemption does not apply, but that it is open the possibility to benefit from the exception provided for in article 101 (3)
TFEU. However, it reminds, the assessment must be conducted by the referring court. It may not provide further guidance on this, not having sufficient information.136 (i) Significance of the judgment
In the Pierre Fabre case, the CJEU officially recognises the importance for
Selective Distribution System of the internet, as a method of marketing and selling alternative to traditional sales. It opens many opportunities to both distributors and consumers, promoting cross-border trade and shopping and, in the meantime, contributing to the attainment of the EU objectives (single market integration and consumer welfare). For this reasons, its development is actively promoted by the
EU Commission.137
Distributors may reach a higher number of customers beyond geographical barriers and provide them with enhanced services (e.g. in terms of availability of information and products for comparison and purchase, at a distance and without time restrictions), stimulating price competition. Consumers may benefit from wider choice and often lower costs.
The alleged risks of counterfeiting, free-riding and even image downgrading exist
(in particular for certain products, such as cosmetics, where the ‘image counts more than substance’ and it is considered easier to maintain in a physical environment).138 However, they may be counteracted by appropriate security and quality enhancing measures, as Advocate General Mazak underlines in his opinion.139 The judgment, more than introducing novelty, appears to satisfy a need for legal certainty in relation to the long-lasting debate concerning the possibility that suppliers prevent or control online sales by their authorised distributors. As noted by the referring Court of Appeal, ‘neither the Commission’s Guidelines nor its observations were binding on the national courts’.140
It establishes that an absolute and general ban on internet selling represents a restriction of competition by object unless objectively justified, infringing article
136

Ibid, paras 48-50.
Goyder, n 6, p 141; Vogel, n 84, p 277.
138
Monti, ‘Restraints on Selective Distribution’ (n 1) 499.
139
Cf. n 107, paras 38-40.
140
Pierre Fabre, n 3, para 30.
137

Selective distribution systems post-Pierre Fabre

194

101 (1) TFEU. It confirms its equivalence to restrictions on active or passive sales to end users, contemplated in the Block Exemption Regulation as hardcore restraints (article 4(c)). It allows the same prohibition to apply, specifying that the internet is not a (virtual) place of establishment, but a method of selling.
Furthermore, recalling the need for proportionality in the assessment of the legality of the qualitative selection criteria, in relation to the nature of the product and their legitimate aims, it does not simply prohibit a ban on internet selling, but suggests that its control may be allowed. In the words of Advocate General Mazak, a general and absolute ban on internet sales imposed by a manufacturer on its distributors is likely to be proportionate ‘only in very exceptional circumstances’.
However, ‘appropriate, reasonable and non-discriminatory conditions’ may be imposed, ‘thereby ensuring the quality of the presentation and distribution of the goods and services advertised and marketed by that means’.141
Indeed, the CJEU agrees with the line of reasoning already expressed by the
Commission in both its 2000 and 2010 Vertical Guidelines, where it is stated that, in principle, every distributor must be allowed to use the internet to sell its products.142 The judgment gives this rule binding power as a principle of law.
More precisely, the Verticals Guidelines state that, within a Selective Distribution
System ‘the dealers should be free to sell, both actively and passively, to all end users, also with the help of the internet’.143 However, the supplier may impose quality standards for the use of an internet site to resell contract products and it may require that its distributors have ‘one or more brick and mortar shop or showroom’ as a condition for their admission and, consequently, before they may operate online.144 This implies the possibility to prevent distributors from selling exclusively by internet,145 adopting the so-called ‘click and brick’ method of distribution.146 Any obligation which dissuade dealers from using the internet to reach more and varied customers by imposing criteria for online sales which are not overall equivalent to those imposed for brick-and-mortar shops is considered as a hardcore

141

AG Mazak, n 107, para 61.
2010 Verticals Guidelines, n 5, para 52.
143
Ibid, para 56. Furthermore, it is specified that, in general, ‘having a website is considered a form of passive selling, since it is a reasonable way to allow customers to reach the distributor’, but sometimes the use of the internet may ‘lead to active selling’, e.g.
‘online advertisement specifically addressed to certain customers’, paras 52, 53.
144
Ibid, para 54. Cf. Jones, Sufrin (n 1), 830
145
Goyder, n 6, p 145; Bellamy and Child, n 7, p 458.
146
Monti, n 126, p 497.
142

Global Antitrust Review 2014

195

restriction. They are not meant to be identical, but their dissimilarities should be justified by the different nature of the two distribution methods.147
(ii) Dissenting voices
The strict link established by Commission and CJEU between the treatment of online distribution and conditions of the physical environment has also been criticised.148 Indeed, it has been stated that, given the distinctive differences between internet selling and traditional sales (different market environments, trends and ways to operate), the lack of an autonomous competition policy for online trade may compromise its potentiality for development. The adoption of the same instruments of regulation has been seen as a defective approach.
These opinions have challenged the very foundation of the CJEU’s judgment in the
Pierre Fabre case. The internet is not seen as a simple method of marketing goods, alternative and complementary to traditional sales, but as an additional (virtual) market space. Accordingly, the question is no longer whether, but how distributors may use the internet. Although, someone may object that this has always been the real question.
V. Conclusion
The analysis of the characteristics of selective distribution systems reveals their inherently contradictory nature. On the one hand, such systems may be deeply restrictive of the economic freedom of the undertakings involved. On the other hand, they can also be highly beneficial for the efficient distribution of certain goods and services in the market. Therefore, it is not surprising that the compatibility of selective distribution with competition law, and in particular with article 101 TFEU, has been often questioned.
The early case-law of the CJEU on selective distribution systems has contributed to the understanding of the inextricable connection between the anticompetitive and the procompetitive aspects of this type of distribution. By the same token, the early jurisprudence on this subject has also helped define the boundaries between lawful and unlawful selective distribution systems, by identifying specific criteria for their evaluation, the so-called Metro criteria. Where these criteria are satisfied, a selective distribution system is considered as producing efficiencies capable of

147

Verticals Guidelines, n 5, para 56.
Cf. Andreas Themelis, ‘After Pierre Fabre: the future of online distribution under competition policy’, (2012) 20 International Journal of Law & Information Technology 4,
346
148

Selective distribution systems post-Pierre Fabre

196

counterbalancing the restrictive effects on competition arising from the restraints on the economic freedom of dealers.
Indeed, in selective distribution systems the manufacturer typically imposes on its distributors specific obligations in order to ensure their compliance with requirements that it feels necessary for the appropriate marketing and distribution of its product. Acceptance of these requirements conditions admission to the selective distribution network. An essential restriction accompanying such admission is the obligation to refrain from supplying non-members. This restriction is necessary in order to ensure that non-members will not be able to supply the contractual goods. Because of its restrictive nature, selective distribution has been considered as appropriate only for the distribution of certain types of products.
In recognition of the efficiencies that selective distribution may generate, the
European Commission adopted an economic approach and in its Verticals
Guidelines it identified a list of factors to be considered in relation to the compatibility of a selective distribution system with article 101 TFEU. However, recent developments in the jurisprudence of the CJEU have cast some doubt on the appropriateness of the Commission’s methodology. Indeed, in October 2011 the
CJEU published its judgment in Pierre Fabre in the context of a reference for a preliminary ruling on the interpretation of articles 101(1) and 101(3) TFEU regarding a clause contained in Pierre Fabre’s selective distribution agreements that stipulated that the sale of contract products must be made exclusively in a physical space in which a qualified pharmacist must be present, thereby imposing an absolute de facto ban on internet selling to end-users. The CJEU took the view that such a clause amounted to a restriction of competition by object to which the
Verticals Block Exemption Regulation was not applicable and that it did not satisfy the conditions of article 101(3) TFEU.
The Pierre Fabre judgment is particularly important for the future competition analysis of selective distribution systems. Although the CJEU recognised the significance of internet as an alternative method of marketing to traditional sales, its conclusions raise considerable concern to the extent that the judgment signifies a shift from the traditional classification of selective distribution systems as restrictions of competition by effect to their categorisation as restrictions of competition by object. As a result of this shift, the traditional Metro criteria are now to be used as a basis for establishing the existence of an objective justification, rather than for conducting an effects analysis. For this reason, Pierre Fabre may pave the way for future developments, relating for instance to the types of objective justifications that may be successful in the context of such an analysis. In any event, this article submitted that the shift in the Court’s approach could perhaps be

Global Antitrust Review 2014

197

attributed to a simple linguistic mistake or ‘slip of the tongue’ that was made when the Court considered the clause in question in its context.149
In addition, a further implication of Pierre Fabre relates to the CJEU’s statement that ‘the aim of maintaining a prestigious image is not a legitimate aim for restricting competition’. However, selective distribution is traditionally used for the marketing of branded products for which it is commonly accepted that their quality is associated not only with their characteristics but also with their allure and prestigious image. In this regard, a more convincing understanding of the CJEU’s position might be that although the aim of maintaining a prestigious image is not an objective justification for a general and absolute ban on internet selling, in some circumstances this may be the case. In this case, the CJEU’s position might be explained as simply finding the restriction in question disproportionate. Indeed, the
Court did not simply prohibit a ban on internet selling. Instead, it recalled the need for proportionality in the assessment of the legality of the qualitative selection criteria having regard to the nature of the product and their legitimate aim.
This article tried to decipher the real meaning of Pierre Fabre comparing the ‘new approach’ to selective distribution systems with their traditional analysis under EU competition law. However, in view of the possible implications of the judgment for the assessment of selective distribution agreements, it cannot be excluded that a future clarification by the CJEU may be in order.

149

Supra, n 128.