Livv
Décisions

GC, 5th chamber, September 9, 2010, No T-155/06

GENERAL COURT

Judgment

PARTIES

Demandeur :

Tomra Systems ASA, Tomra Europe AS, Tomra Systems GmbH, Tomra Systems BV, Tomra Leergutsysteme GmbH, Tomra Systems AB, Tomra Butikksystemer AS

Défendeur :

European Commission

COMPOSITION DE LA JURIDICTION

President :

Vilaras

Judge :

Prek, Ciuca (Rapporteur)

Advocate :

Ryan, Midthjell, Frey

GC n° T-155/06

9 septembre 2010

THE GENERAL COURT (Fifth Chamber),

Background to the dispute

1 Tomra Systems ASA is the parent company of the Tomra group. Tomra Europe AS coordinates the business of the European distribution subsidiaries within the group. The distribution subsidiaries concerned by the present case are Tomra Systems GmbH in Germany, Tomra Systems BV in the Netherlands, Tomra Leergutsysteme GmbH in Austria, Tomra Systems AB in Sweden and Tomra Butikksystemer AS in Norway (hereinafter referred to, together with Tomra Systems ASA and Tomra Europe AS, as 'the applicants'). The Tomra group produces automatic recovery machines for empty beverage containers (reverse vending machines ('RVMs')) - which are machines for the collection of used beverage containers which identify the container by reference to certain parameters, such as shape and/or bar codes, and calculate the amount of the deposit to be reimbursed to the customer. The group also provides RVM-related services throughout the world. In 2005, the Tomra group had a turnover of around EUR 300 million and 1 900 employees.

2 On 26 March 2001 the Commission of the European Communities received a complaint from Prokent AG ('Prokent'), a Germany company also active in the empty-beverage-container collection sector and the supply of related products and services. Prokent requested the Commission to investigate whether the applicants had abused a dominant position by preventing it from entering the market.

3 On 26 and 27 September 2001, the Commission inspected the premises of Tomra Systems GmbH, in Germany, and of Tomra Systems BV, in the Netherlands. At the Commission's request, the European Free Trade Association ('EFTA') Surveillance Authority inspected the premises of Tomra Systems ASA and its subsidiaries in Norway. Next, the Commission requested information from Tomra Systems ASA and from a number of its competitors and customers, pursuant to Article 11 of Council Regulation No 17 of 6 February 1962, First Regulation implementing Articles [81 EC] and [82 EC] (OJ, English Special Edition 1959-1962, p. 87).

4 On 23 December 2002, in a letter to the Commission, the applicants stated that they would put an end to the exclusivity agreements and would no longer apply the loyalty rebates.

5 On 30 March 2004, the applicants presented a competition compliance programme for the Tomra group, which was to apply from 1 April 2004.

6 On 1 September 2004, the Commission sent a statement of objections to Tomra Systems ASA, Tomra Europe AS and subsidiaries of the Tomra group in six States forming part of the European Economic Area (EEA), to which the applicants responded on 22 November 2004. The hearing took place on 7 December 2004. On 19 April 2005, the Commission issued requests for further information, which the applicants provided on 25 April and 3 May 2005.

The contested decision

7 On 29 March 2006, the Commission adopted Decision C(2006) 734 final relating to proceedings under Article 82 [EC] and Article 54 of the EEA Agreement (Case COMP/E-1/38.113 - Prokent-Tomra) ('the contested decision', a summary thereof being published in the Official Journal of the European Union (OJ 2008 C 219, p. 11)). In that decision it found that the applicants had infringed Article 82 EC and Article 54 of the EEA Agreement in the period from 1998 to 2002 by implementing an exclusionary strategy in the national RVM markets in Germany, the Netherlands, Austria, Sweden and Norway, involving exclusivity agreements, individualised quantity commitments and individualised retroactive rebate schemes, thus foreclosing competition on the markets.

I - Relevant market

8 With regard to the relevant product market, the contested decision states that for the purposes of its assessment the Commission started from the premiss that there was a specific market for 'high-end' machines or other systems, including in particular all RVMs that can be installed through a wall and connected to backroom equipment, and an overall market encompassing both high-end and low-end equipment. The Commission decided, however, to use the wider market definition as a working basis, since it yielded more favourable figures to the applicants' advantage.

9 With regard to the relevant geographic market, the Commission found in the contested 'decision that the conditions of competition had not been harmonious across the EEA in the period under consideration and that the relevant geographic markets were national in scope.

II - Dominant position

10 In the contested decision, the Commission, after finding, inter alia, that the applicants' market shares in Europe had continuously exceeded 70% in the years before 1997, that after 1997 they exceeded 95% and that, on all the relevant markets, the applicants' market shares were many times larger than those of their competitors, concluded that the Tomra group was a dominant undertaking for the purposes of Article 82 EC and Article 54 of the EEA Agreement.

III - Abuse

11 The contested decision states that the applicants devised a strategy having an anti-competitive object or effect, both in their practices and in internal discussions within the group. The Commission asserts in the decision that the applicants sought to preserve their dominance and market share through means such as preventing new operators gaining market entry, keeping competitors small by limiting their growth possibilities and weakening and eliminating competitors by way of acquisition or other methods. That strategy was, according to the contested decision, implemented through the conclusion, between 1998 and 2002, of 49 agreements between the applicants and a certain number of supermarket chains, which took the form of exclusivity agreements, agreements containing individualised quantity commitments and agreements establishing individualised retroactive rebate schemes.

12 It is also apparent from the contested decision that, although the agreements, clauses and conditions contain a number of different features, such as explicit or de facto exclusivity clauses, undertakings or promises, on the part of the customers, to purchase quantities corresponding to a significant proportion of their requirements or retroactive rebate schemes related to the customers' requirements or a combination of those features, they must all, in the Commission's view, be seen in the context of the applicants' general policy directed at preventing market entry, market access and growth opportunities for existing and potential competitors and, ultimately, at driving them out of the market so as to create a situation of virtual monopoly.

13 First of all, according to the contested decision, exclusivity clauses, because they require customers to purchase all or a significant part of their requirements from a dominant supplier, are, by definition, liable to have a foreclosure effect. In the present case, as the applicants were dominant on the market and as the exclusivity clauses were applied to what the Commission considered could be regarded as a substantial part of the total demand, the Commission concluded that the exclusivity agreements entered into by the applicants were capable of having, and in fact had, a market-distorting foreclosure effect. It was stated in the contested decision that, in this instance, there were no circumstances that could exceptionally justify exclusivity or similar arrangements and that the applicants had also failed to justify their practices by cost savings.

14 The contested decision then adds that discounts granted for individualised quantities corresponding to the entire or almost entire demand have the same effect as explicit exclusivity clauses, in that they induce the customer to purchase all or virtually all its requirements from a dominant undertaking. The same is true of loyalty rebates, in other words, rebates that are conditional on customers purchasing all or most of their requirements from a dominant supplier. In the Commission's view, it is not decisive for the exclusionary nature of the agreements or conditions concerned whether the purchase-volume commitment is expressed in absolute terms or as a percentage. So far as the applicants' agreements were concerned, the contested decision states that the stipulated quantity targets constituted individualised commitments which were different for each customer, regardless of its size and purchasing volume, and which corresponded to the customer's entire requirements or to a large proportion of them, or even exceeded them. The contested decision adds that the applicants' policy of tying their customers, in particular their key customers, into agreements that aimed at excluding competitors from the market and denying them any chance of growth is evident from the documents relating to the applicants' strategy, their negotiations and the offers made by them to their customers. Given the nature of the RVM market and the characteristics of the product itself, in particular the transparency and relatively foreseeable demand of each customer for each year, the Commission found that the applicants had the market knowledge necessary to make a realistic estimate of each customer's approximate demand.

15 Furthermore, concerning the rebates, the Commission noted in the contested decision that the rebate schemes were individual to each customer and the thresholds were linked to the total requirements of each customer or a large part thereof. They were established on the basis of the customer's estimated requirements and/or purchasing volumes achieved in the past. The incentive for buying exclusively or almost exclusively from the applicants was, according to the contested decision, particularly strong where thresholds such as those applied by the applicants were combined with a system whereby the achievement of the bonus or a more advantageous bonus threshold benefited all purchases made by the customer during the reference period and not exclusively the purchasing volume exceeding the relevant threshold. Under a retroactive scheme, a customer which has started purchasing from the applicants - a very likely scenario given the applicants' strong market position - thus has a strong incentive to reach the required threshold in order to reduce the price of all its purchases from the applicants. That incentive increased, according to the contested decision, the closer the customer came to the threshold in question. For the Commission, the combination of a retroactive rebate scheme with a threshold or thresholds corresponding to the customer's entire requirements or a large proportion thereof represented a strong incentive to buy all or almost all the equipment needed from the applicants and artificially raised the cost of switching to a different supplier, even for a small number of units. The Commission therefore concluded that, in accordance with case-law, the rebate schemes at issue had to be qualified as a means of loyalty building and thus as loyalty rebates.

16 Finally, the contested decision points out that, although - as was held in Case T-203-01 Michelin v Commission [2003] ECR II-4071 ('Michelin II'), paragraph 239, and Case T-219-99 British Airways v Commission [2003] ECR II-5917, paragraph 293 - for the purpose of establishing an infringement of Article 82 EC, it is sufficient to show that the abusive conduct of the undertaking in a dominant position tends to restrict competition or, in other words, that the conduct is capable of having that effect, the Commission completed its analysis by considering the likely effects of the applicants' practices on the RVM market. In that regard, the contested decision states that during the entire period under consideration, namely 1998 to 2002, the applicants' share on each of the five national markets under consideration remained relatively stable. Over the same period, their competitors' position remained rather weak and unstable. One successful competitor, the complainant, left the market in 2003 after managing to acquire an 18% share of the German market in 2001. Other competitors that had demonstrated the potential and ability to acquire larger market shares, such as Halton and Eleiko, were eliminated because they were acquired by the applicants. In addition, in the Commission's view, the applicants' exclusionary strategy, as it was operated by them throughout the period from 1998 to 2002, had an effect that is demonstrated by changes in the tied market share and the sales of market players. Moreover, according to the contested decision, some customers started purchasing more of the competing products after the expiry of their exclusionary agreements with the applicants. In addition to there being no cost efficiencies such as to justify the applicants' practices, there was in the present case no benefit to consumers either. The contested decision states that the price of the applicants' RVMs did not fall after the sales volume had increased and that, on the contrary, prices stagnated or even rose during the period under investigation.

IV - Fine

17 The contested decision states that the assessment of the gravity of the applicants' infringement must take account of the fact that they had deliberately employed the practices in question in the context of their exclusionary strategy and also of the geographic scope of the infringement, namely the fact that it encompassed five EEA States: Germany, the Netherlands, Austria, Sweden and Norway. Conversely, account must also be taken, in the Commission's view, of the fact that the infringement did not cover the whole of the reference period on all the national markets under consideration and that within each of those markets its intensity could well have varied over time.

18 In particular, the contested decision specifies, at recital 394, that the infringement concerns the following territories and periods:

- Germany: 1998 to 2002

- Netherlands: 1998 to 2002

- Austria: 1999 to 2001

- Sweden: 1999 to 2002

- Norway: 1998 to 2001

19 The Commission concluded that it was a serious infringement and set the basic amount of the fine at EUR 16 million, taking as its basis the five-year period running from 1998 to 2002. The starting amount of the fine was increased by 10% for each full year of the infringement. Finally, the contested decision states that there were no aggravating or mitigating circumstances.

20 The enacting terms of the contested decision read as follows:

'Article 1

[The applicants] have infringed Article 82 [EC] and Article 54 of the EEA Agreement in the period 1998-2002 by implementing an exclusionary strategy in the national [RVM] markets in [Germany, the Netherlands, Austria, Sweden and Norway], involving exclusivity agreements, individualised quantity commitments and individualised retroactive rebate schemes, thus foreclosing competition on the markets.

Article 2

For the infringement referred to in Article 1, a fine of EUR 24 million is imposed on [the applicants], jointly and severally.

...'

Procedure and forms of order sought

21 By application lodged at the Registry of the Court on 14 June 2006, the applicants brought the present action.

22 As the composition of the Chambers of the Court changed, the Judge-Rapporteur was assigned to the Fifth Chamber, to which the present case was accordingly allocated.

23 On hearing the report of the Judge-Rapporteur, the Court (Fifth Chamber) decided to open the oral procedure. The parties presented oral argument and replied to the Court's oral questions at the hearing on 14 January 2010.

24 The applicants claim that the Court should:

- annul the contested decision;

- in the alternative, annul or substantially reduce the amount of the fine;

- order the Commission to pay the costs, including those incurred by them in providing a bank guarantee to cover their obligation to pay the fine.

25 The Commission contends that the Court should:

- dismiss the action;

- order the applicants to pay the costs.

Law

26 The applicants put forward six pleas in law. The first five pleas seek, in essence, annulment of the contested decision and the sixth seeks annulment or reduction of the fine. The first plea alleges that the Commission used manifestly incorrect and unreliable evidence in finding that the applicants conducted a strategy of foreclosure. The second plea alleges manifest error of assessment by the Commission in holding that the applicants' practices were capable of eliminating competition, and failure to state reasons. The third plea alleges manifest errors in the Commission's assessment of whether the practices did in fact eliminate competition. The fourth plea alleges manifest error of law in the characterisation of the applicants' practices as unlawful per se. The fifth plea alleges a manifest error on the part of the Commission in holding that non-binding commitments could infringe Article 82 EC. Last, the sixth plea alleges breach of the principles of proportionality and non-discrimination in the imposition of the fine.

I - The claim for annulment of the contested decision

A - First plea: use of manifestly incorrect and unreliable evidence in finding a strategy of foreclosure and in proving the existence, and determining the content, of certain agreements between the applicants and their customers

27 This plea is expressed in two parts. First, the applicants claim that the contested decision does not contain any reliable evidence demonstrating that they designed a strategy to foreclose competition. Second, the applicants claim that the contested decision is based on incorrect and unreliable evidence to prove the existence and content of at least 26 of the 49 agreements to which the contested decision refers.

1. First part: lack of reliable evidence to demonstrate the existence of a strategy of foreclosure

(a) Arguments of the parties

28 First, the applicants dispute the Commission's use of their internal correspondence as evidence. On that basis, the applicants claim that there is no connection between the documents assembled by the Commission and that they are taken wholly out of context. The applicants further claim that the Commission omitted evidence which shows instead that they intended to compete normally with their competitors. The applicants assert that the contested decision ignores the documents which express their intention to use legitimate competitive methods.

29 Second, the applicants assert that the Commission failed to consider in the contested decision whether the Tomra group had been successful on the RVM market between 1998 and 2002 because, from 1997 to 2001, it was the only supplier of RVMs equipped with a 'new revolutionary technology'. That, in the applicants' submission, constitutes an error such as to entail annulment of the contested decision: their competitive advantage was their technology and it was on the basis of that advantage that they decided to operate on the market.

30 Third, the applicants observe that, by relying on the allegedly anti-competitive agreements themselves as evidence of an exclusionary strategy, the Commission becomes trapped in a circular argument, because at several points elsewhere in the contested decision it refers to the applicants' exclusionary strategy in order to show that the same agreements were anti-competitive. Consequently, those agreements cannot serve as evidence of an exclusionary strategy. Even if the Commission had been able to show instances of agreements which infringed Article 82 EC, it would still not have offered any explanation of the way in which that fact confirms the existence of a group-wide strategy of excluding competition between 1998 and 2002.

31 Fourth, the applicants assert that the Commission, in the contested decision, on the one hand, did not consider that their patent infringement proceedings or their acquisitions formed part of the infringement, but, on the other hand, considered that those proceedings and acquisitions represented factors demonstrating the exclusionary strategy followed by them. However, the Commission did not adduce the slightest evidence showing that the applicants' protection of their patents or their practices of cooperating with or acquiring other companies demonstrated an exclusionary strategy. In particular, the applicants argue that, according to a consistent line of decisions in which it was held that the existence of intellectual property rights does not infringe competition law, unless it is vexatious, an attempt to uphold a patent or other intellectual property right before the national courts cannot infringe Community competition law.

32 The Commission disputes the arguments advanced by the applicants.

(b) Findings of the Court

33 First of all, it should be noted that the contested decision, at recital 97 et seq., summarises what the Commission considered to be the applicants' anti-competitive strategy. It makes the following statement in that regard:

'[The] strategy [of the Tomra group] was based on a policy that sought to preserve its dominance and market share through means such as ... preventing market entry, ... keeping competitors small by limiting their growth possibilities and ... finally weakening and eliminating competitors, by way of acquisition or otherwise, especially those competitors that were deemed to have the potential to become more serious challengers. To obtain this objective [the] Tomra [group] employed various anti-competitive practices, including exclusivity and preferred supplier agreements, as well as agreements containing individualised quantity commitments or retroactive rebate schemes.'

34 The contested decision goes on to say that '[the] Tomra [group's] overall strategy is not only confirmed by the different practices employed by the group, but was also discussed extensively within the group on various occasions, be it at meetings and conferences or in correspondence, for instance, e-mail'.

35 Thus, the Commission, after mentioning the various anti-competitive practices employed by the applicants, was right to examine the applicants' internal documentation. Such documentation may indicate whether the exclusion of competition was intended or, on the contrary, suggest another explanation for the practices under consideration. In this instance, the applicants' internal correspondence allowed the Commission to place their practices in context and to substantiate its own assessment of those practices. It should also be pointed out that the Commission's findings in the contested decision are never based on just one of the applicants' documents taken in isolation but on a whole series of different items.

36 First, as regards the assertion that the contested decision shows that the Commission ignored documents which reveal an intention on the part of the applicants to employ legitimate competitive methods, it is perfectly legitimate for the contested decision to concentrate primarily on the applicants' anti-competitive conduct and not on their lawful actions, since it was precisely that conduct which it was the Commission's task to investigate. However, it should be pointed out that the contested decision does not conceal that certain of the applicants' internal documents also mention other quite legitimate methods of competing (see, for example, recital 100 to the contested decision).

37 Second, as regards the applicants' alleged technological advantage, it would have had no impact on the Commission's findings if that advantage had been referred to in the contested decision. The applicants do not show in what way the technology which they have developed could have served to justify their practices. Moreover, if that technology was really so markedly superior to that of their competitors that customers would never have purchased the latter's products anyway, it becomes even more difficult to explain the use of exclusivity agreements, quantity commitments and other individualised rebate schemes.

38 Third, in relation to the alleged circularity of the Commission's reasoning in the contested decision so far as the relationship between the anti-competitive agreements and the strategy of foreclosure is concerned, it must be stated that, according to settled case-law, the concept of abuse is an objective concept referring to the behaviour of an undertaking in a dominant position which is such as to influence the structure of a market where, as a result of the very presence of the undertaking in question, the degree of competition is already weakened and which, through recourse to methods different from those which condition normal competition in products or services on the basis of the transactions of commercial operators, has the effect of hindering the maintenance of the degree of competition still existing in the market or the growth of that competition (Case 85-76 Hoffman-La Roche v Commission [1979] ECR 461, paragraph 91, and Case T-210-01 General Electric v Commission [2005] ECR II-5575, paragraph 549).

39 Applying that line of case-law, the Commission, at recital 107 et seq. to the contested decision, established that the applicants' practices, considered in their context and in conjunction with a series of factors, including the applicants' internal documents, were liable to foreclose competition. Consequently, contrary to the applicants' submission, the Commission by no means relied exclusively on the applicants' intention or policy in order to substantiate its finding of an infringement of competition law.

40 Finally, with regard to the applicants' patent infringement proceedings and their acquisitions, it is sufficient to note that the contested decision, at recitals 106 and 107, makes clear that those practices are not part of the abuse of a dominant position. They are thus merely relevant facts that put the applicants' practices in context but have no impact on the finding of an infringement.

41 The first part of the first plea must therefore be rejected.

2. Second part: the use of incorrect and unreliable evidence in proving the existence, and determining the content, of certain agreements between the applicants and their customers

42 This part may be divided into four subsidiary parts. The first subsidiary part concerns the exclusivity agreements before 1998, the second subsidiary part concerns the agreements designating the applicants as 'preferred, main or primary supplier', the third subsidiary part concerns individualised quantity commitments and individualised retroactive rebate schemes and, last, the fourth subsidiary part concerns certain of the agreements relating to four of the five countries examined in the contested decision, agreements which, in the applicants' submission, were incoherently assessed by the Commission.

(a) The exclusivity agreements before 1998

Arguments of the parties

43 The applicants emphasise that 9 of the 21 exclusivity agreements pre-date the period covered by the contested decision (1998 to 2002) and cannot therefore have contributed to the foreclosure of competitors during that period. The applicants conclude that those nine agreements ought not to have been referred to in the contested decision and ought also to have been excluded when the fine was calculated.

44 Even though, in the defence, the Commission asserts that it did not take certain of those agreements into account in its assessment, the applicants question why they are listed at recital 296 to the contested decision as though they were relevant.

45 The Commission disputes the arguments advanced by the applicants.

Findings of the Court

46 In this connection, nine of the exclusivity agreements (namely the agreements with AS Butikkdrift for 1995 and 1996, Kiwi Minipris Norge for 1996, Køff Hedmark for 1996, Rema 1000 for 1996, AKA/Spar Norge for 1997, Rewe Wiesloch and Rewe Hungen for 1997, De Boer Unigro for 1997 and Samenwekende van den Broek Bedrijven for 1997), referred to at recital 296 to the contested decision as agreements infringing Article 82 EC and Article 54 of the EEA Agreement, relate to a period preceding the period covered by the contested decision (1998 to 2002). It is therefore clear that the Commission made an error in referring to those agreements at recital 296, a fact which is in any event acknowledged by the Commission itself in its pleadings.

47 It should, however, be observed, first, that the calculation of the fine did not take into account any fact pre-dating 1998 and, second, that the Commission maintains, without being contradicted by the applicants, that the agreements applicable before 1998 were never taken into account in its assessment of the part of demand that was not contestable by the applicants' competitors and that the findings on the applicants' foreclosure strategy are therefore wholly independent of those nine agreements.

48 A reading of the contested decision as a whole removes any ambiguity in that regard (see, for example, recitals 134, 159, 166, 242, 264, 269, 394, 417 and 418 to the contested decision) and makes clear that the Commission at no time took into account any infringement prior to 1998. It follows that this complaint cannot be accepted.

(b) The agreements designating the applicants as 'preferred, main or primary supplier'

Arguments of the parties

49 The applicants assert that the Commission 'automatically' characterised the agreements whereby the customer undertook to retain them as 'preferred, main or primary supplier' as exclusivity agreements, although the terms are too vague for the agreements to be classified as exclusive in contract law. Furthermore, although the customers concerned purchased RVMs from the applicants' competitors while the alleged exclusivity contracts were in force and those customers had stated that the agreements were in fact non-exclusive, the Commission regarded them as exclusive.

50 In the applicants' submission, the Commission, in the contested decision, failed to analyse whether an enforceable exclusivity right had been created from the point of view of national contract law. The contested decision contains no analysis whatsoever of the contracts by reference to national law. In the contested decision and the Commission's requests for information issued prior to its adoption, loose understandings, which do not create binding and enforceable contractual obligations, are treated in the same way as formal enforceable contracts. The applicants contend that, while that approach may be appropriate in the context of a cartel for the purposes of Article 81 EC, it is not appropriate in the context of an alleged exclusivity agreement for the purposes of Article 82 EC. Unless an exclusivity obligation can be enforced under national contract law, it will not prevent a competitor from selling to the customer or prevent the customer from accepting an offer. Likewise, it cannot create any ex ante deterrent effect on customers.

51 Furthermore, the contested decision relies, in support of the Commission's findings in relation to the contractual status of the documents, on irrelevant evidence such as internal memoranda of the applicants, press releases or the Tomra group's annual report, which are unilateral declarations of that group to which the customer had not subscribed. The Commission failed to analyse the evidential value of such statements by reference to national contract law.

52 Last, the applicants contend that the agreements with the two purchasing organisations Superunie (the Netherlands, 2001) and ICA/Hakon (Sweden and Norway, for the period 2000 to 2002) did not legally bind the members of those organisations to purchase RVMs solely from them and that there is no evidence that those agreements provided 'pressure or strong inducement' for independent retailers to buy all their RVMs from the applicants.

53 In particular, as regards the agreement with Superunie, the applicants maintain that the contested decision fails to offer any evidence to show how this agreement could nevertheless commit the independent outlets which were members of Superunie to purchase 130 RVMs from them. Instead, the Commission reversed the burden of proof in arguing that 'there [was] no evidence suggesting that the individual members did not feel bound ...' by the agreement. If the members had been offered a better deal from one of the applicants' competitors, the Commission has not, in the applicants' submission, explained why the members would still decide to purchase from the applicants under a non-binding agreement with the central level of their organisation.

54 The Commission disputes the arguments advanced by the applicants.

Findings of the Court

55 It is appropriate first of all to take issue with the applicants' statement that the Commission 'automatically' characterised the agreements whereby the customer undertook to regard the applicants as 'preferred, main or primary supplier' as exclusivity agreements. Recitals 114 to 122 and recital 286 et seq. to the contested decision in any event attest to the contrary inasmuch as they explain in detail the approach and the findings of the Commission in that regard.

56 Furthermore, although it is true that certain customers sought to include in their 'preferred supplier' agreements a clause allowing them to purchase competing machines for testing purposes, that reinforces the conclusion that those agreements were intended to be exclusive and that the possibility of purchasing competing machines was an exception limited purely to the testing of those machines.

57 Moreover, the Commission, in the contested decision, characterised the preferred supplier agreements as exclusive on the basis of the available evidence of the parties' intentions. That evidence shows that the agreements were in fact intended to be exclusive and were understood to be exclusive, irrespective of the question as to whether they were enforceable under national contract law.

58 The agreement with Royal Ahold, for example, describes the Tomra group as 'main supplier'. However, the president of Tomra Systems ASA had stated during negotiations with that customer (see recital 139 and footnote 267 to the contested decision):

'It is therefore our preference that [the] Tomra [group] be appointed the "exclusive" global RVM service provider to Ahold. Words other than exclusive could be crafted, which would carry out the basic intent of the parties. However, putting aside [choice] of words, the deal as we all the time [have] discussed it is that [the] Tomra [group] is to have the right to place machines ... at any new store requiring RVMs and ... upon [expiry] of existing agreements at any store now serviced by another RVM provider.'

59 As to the applicants' claim that the Commission did not, in the contested decision, analyse the exclusivity of the agreements on the basis of the applicable national law, it should be recalled that there is no need for the practices of a dominant undertaking to tie purchasers by a formal undertaking in order for it to be found that those practices amount to an abuse of a dominant position within the meaning of Article 82 EC. It is enough that those practices give customers an incentive not to turn to competing suppliers and to obtain all or most of their requirements exclusively from the undertaking concerned (see, to that effect, Hoffmann-La Roche v Commission, paragraphs 89 and 90).

60 In that connection, the agreements concerned not only describe the Tomra group as 'preferred, main or primary supplier' but they also contain quantity commitments or progressive retroactive rebates conditional upon purchase of a given volume. The agreements for 2000 to 2002 with the Netherlands group Royal Ahold and with ICA/Hakon/Ahold in relation to Sweden and Norway are examples of this type of agreement.

61 Finally, with regard to the agreements entered into between the applicants and central purchasing organisations such as Superunie and ICA/Hakon, it should be noted, first, that the applicants do not dispute that those agreements were binding on the purchasing organisations which had entered into them and, second, that the question as to whether the agreements also had an impact on the purchasing behaviour of their members does not depend on a formal analysis.

62 Indeed, the Court shares the Commission's view that, when the conditions negotiated depend on target quantities to be purchased by the central organisation as a whole, it is inherent in the negotiation of that type of agreement that the agreement will encourage members of the organisation to make purchases with a view to achieving the target set.

63 Moreover, the fact that the purchasing target set by the agreement concluded with ICA Ahold/Hakon was reached (see recital 171 to the contested decision) shows how the central purchasing organisation had the power to influence the behaviour of independent retailers.

64 Furthermore, the agreement with Superunie expressly mentioned each of the different members and the number of machines which each of them was supposed to purchase.

65 Finally, it should be noted, as the Commission has rightly pointed out, that the case-file contains ample evidence that compliance with the agreement was closely monitored and that pressure was brought to bear on individual retailers.

66 In that regard, mention should be made, for example, of the letter of 16 February 2001 from Tomra Europe to ICA Ahold concerning the agreement of 13 October 2000, which expresses the applicants' concern about ICA's purchasing patterns under the agreement and recalls that ICA 'has undertaken to do their utmost at a central level to support [the] Tomra [group] across its network of stores and encourage its franchisees to accelerate old machine replacements and remain 100% loyal to this agreement'. The letter states that 'central support from ICA has clearly not been effective so far' and that ICA's communication of the incentives under the agreements to its franchisees has not been sufficient. Tomra Europe thus called upon ICA to take urgent steps to implement a plan in accordance with the agreement.

67 In light of the foregoing, the applicants' complaint must therefore be rejected.

(c) Individualised quantity commitments and individualised retroactive rebate schemes

Arguments of the parties

68 The applicants claim that there is no proof that they were able to estimate customer requirements accurately.

69 They assert, in the first place, that the Commission acknowledges in the contested decision that customers did not inform the applicants of the size of their total or near-total RVM requirements in the contract period.

70 They maintain, in the second place, that if demand for RVMs was non-recurring and irregular, as stated in the contested decision, then the number of RVMs which a customer purchased during the preceding year could not enable the applicants to estimate what the same customer's total or near-total demand would be the following year.

71 The applicants also dispute the Commission's assertion that the demand of each individual customer is easy to predict 'as it is created or increased by the introduction of mandatory deposit systems'. On that point, the applicants assert that none of the five States in which the alleged infringements took place introduced a mandatory deposit system between 1998 and 2002.

72 In the third place, the Commission indirectly accepted that the applicants were unable to estimate their customers' total or near-total RVM requirements by deleting throughout the whole of the contested decision, as business secrets, all figures relating directly or indirectly to past annual purchases of RVMs.

73 Last, the applicants assert that it was rare for quantity commitments and objectives to coincide with customers' actual purchases. Actual purchases were either significantly below or significantly above the alleged quantity commitments, while the customer also purchased RVMs from the applicants' competitors. In support of that assertion, the applicants submit a report drawn up by economists, who set out the actual purchases made by each relevant customer and compared them with the alleged quantity commitments.

74 The applicants submit that the empirical findings in that report show that actual purchases of RVMs were consistently higher than contracted quantities. That is consistent with the fact that the applicants were able to offer 'revolutionary technology', which customers were eager to have at their outlets. It is not, however, consistent with the contested decision (see recital 123), where it is stated that purchases envisaged in the agreements corresponded 'entirely or almost entirely' with the number of RVMs which the customer eventually purchased during the contract period.

75 The Commission disputes the arguments advanced by the applicants.

Findings of the Court

76 So far as the individualised quantity commitments and the individualised retroactive rebate schemes are concerned, the applicants claim that there is no proof that they were able, during the period under consideration, to estimate customer RVM requirements correctly. In their submission, the unlawfulness of those types of practices depends precisely on the ability of the supplier to estimate customer requirements.

77 It must be stated as a preliminary point that the Commission, in the contested decision, in fact found that the quantity commitments and the rebate schemes had been individualised for each customer and that the thresholds related to the customer's total requirements or to a large proportion thereof (see, for example, recital 319 to the contested decision).

78 It may thus be stated that, as the applicants maintain, the contested decision, in relation to the argument concerning quantity commitments and rebate schemes set out therein, rests on the fact that the applicants were able to individualise this type of agreement by knowing what each customer's requirements were. The applicants' argument is based on the fact that in reality they were unable to estimate customer requirements precisely and thus that the Commission could not refer to individualised agreements. It should thus be established whether the Commission made an error in this regard.

79 First, it should be noted that the customer sometimes indicated its expected future requirements, as was the case for example in relation to the agreement with Rimi Svenska. Moreover, in their reply to the statement of objections, the applicants point out that in contract negotiations 'it [was] normal and necessary for both parties to have an approximate idea of the quantity, i.e. number of units which is likely to be the subject of the contract'.

80 Second, the applicants' assertion that the Commission found that targets were established on the basis of past purchases alone is incorrect. On the contrary, the Commission, correctly, held in the contested decision that, in view of the characteristics of the RVM market, the demand of each customer was relatively easy to predict. In order to provide for the future needs of their customers, the applicants had various elements at their disposal: information provided by the customers themselves, customer purchases during the previous year or earlier years, transparent data on the most relevant factors (number and size of outlets, existence or expectation of a deposit system) and the applicants' own market research based on their thorough knowledge of the market (see recital 298 to the contested decision).

81 In particular, with regard to the applicants' argument that none of the five countries concerned by the contested decision introduced a mandatory deposit system between 1998 and 2002, it must be stated that demand for RVMs increased in anticipation of the introduction of a deposit system, as proved to be the case, for example, in Germany in the period from 2000 to 2001, even though the system was not actually introduced until the end of 2002 (see recitals 188, 219 and 221 to the contested decision). Likewise, 'voluntary' systems for the collection of used beverage containers, such as that existing in Norway, also had an obvious and predictable impact on demand for RVMs (see recital 242 to the contested decision).

82 It follows that, contrary to the applicants' claim, 'non-recurrent' and 'irregular' demand may - as in the present case - none the less be readily foreseeable.

83 As for the confidential treatment, in the contested decision, of information provided by customers about purchases from competing suppliers, it cannot be concluded on that basis that the applicants were not able to estimate their customers' demand. The customers' request for confidentiality merely indicates that they did not wish to disclose to the applicants their purchases from competitors.

84 Finally, as regards the lack of systematic correlation between the alleged quantity commitments and customers' actual purchases, it must be stated that the study provided by the applicants rests on a flawed reading of the contested decision. In fact, the contested decision states that the contracts in question generally corresponded to all or a high proportion of customers' actual requirements in a certain contract period and not that quantity commitments must have corresponded precisely with total actual demand as observed ex post facto (see recitals 102, 108, 123, 124 and 127 to the contested decision).

85 Furthermore, the contested decision makes clear that when the targets set by the agreements did not correspond to the customer's total requirements, they corresponded at least to somewhere between 75% and 80% of its total demand (see, for example, recital 159 to the contested decision).

86 In any event, the statistical study submitted by the applicants appears to confirm the premiss on which the contested decision is based. The applicants' ex post analysis shows, in substance, that actual purchasing volumes are, in most cases, slightly above the volumes provided for in the quantity commitments. That finding is borne out by the comparative table provided by the Commission, in which quantity commitments and rebate thresholds are compared with customers' actual purchases.

87 In view of the foregoing considerations, this complaint must also be rejected.

(d) The assessment of certain agreements entered into in Germany, the Netherlands, Sweden and Norway

88 The applicants maintain that, for four of the five countries in which the infringement was committed, most of the agreements mentioned in the contested decision were incoherently assessed. Those agreements should therefore not have been taken into account by the Commission and that ground is sufficient to justify annulment of the contested decision. The applicants do not dispute the analysis of the agreements in Austria. The first plea in law does not therefore concern that market.

89 It is therefore appropriate to examine below the complaints relating to the agreements concerning Germany, the Netherlands, Sweden and Norway.

Germany

90 The applicants assert that more than half the agreements concerning Germany on which the Commission relies did not exist or did not include any exclusivity clauses, quantity commitments or retroactive rebates.

91 The Commission disputes the arguments advanced by the applicants.

- Edeka Bayern-Sachsen-Thüringen (1998-1999)

92 The applicants claim that this agreement was not exclusive and that it provided only that the customer should purchase RVMs from them in a concentrated manner. However, the customer was allowed to test competing machines and to purchase them if they offered significant advantages.

93 The applicants further submit that the Commission has adduced no evidence that the agreement could be characterised as exclusive.

94 It must be stated that, contrary to the applicants' assertion, the agreement was conceived at the outset as an exclusivity agreement. That argument is borne out by an oral contact between the two companies that was referred to in an e-mail circulated within the German subsidiary of the Tomra group.

95 Furthermore, when an agreement is not exclusive the customer retains freedom of choice and can purchase from any competitors whatsoever. There is normally no need for the customer to prove that the competitor has 'significant advantages', as is the case here.

96 As to the argument that the agreement in question provides that Edeka Bayern-Sachsen-Thüringen can actually test the products of the applicants' competitors, if a customer is to order machines from a new supplier, it must have tested them over some time; a period of exclusivity is therefore not incompatible with the retention of the right to test competing machines. Moreover, a contract that is not exclusive does not, as a rule, contain such a clause.

97 It must therefore be stated, as the Commission has correctly pointed out, that a contract does not normally contain clauses restricting or directing the customer's purchase options. Clauses such as that allowing the customer to test competing equipment over a limited period or a clause allowing the customer to obtain supplies from competitors only if there is a significant advantage are not an indication that the agreement is not exclusive.

98 This complaint must therefore be rejected.

- Edeka Handelsgesellschaft Hessenring (1999)

99 The applicants challenge the way in which the Commission analysed the 1999 agreement and claim that it produced no evidence to support the assertion that the DEM 2 million threshold, which had to be reached in order to obtain a supposed individual discount of 0.5%, represented an individualised rebate scheme based on the total or near-total requirements of the customer. The applicants add that that threshold was not reached by Edeka Handelsgesellschaft Hessenring in 1999. Thus, even if such a discount clause depending on the amount of purchases did exist, it could not have been applied.

100 In that regard, it must be noted, as the applicants have pointed out, that the Commission does not establish in the contested decision that the agreement with Edeka Handelsgesellschaft Hessenring represented an individualised rebate scheme based on all, or almost all, the customer's requirements. In the absence of other evidence, the fact that Edeka Handelsgesellschaft Hessenring did not purchase products for an amount exceeding the specific threshold needed to qualify for the rebate appears to confirm that.

101 It should, however, be pointed out that, even though the Commission did not prove that the agreement was individualised, it cannot be denied that what is at issue is an agreement for a progressive and retroactive rebate. Further, in cases in which the customer had purchased below the threshold, the Commission asserts, without being contradicted by the applicants, that it included in the 'tied share' only actual purchases from the dominant undertaking during the period of the agreement.

102 Accordingly, this complaint cannot be accepted.

- Edeka Baden-Württemberg (2000)

103 The applicants claim that the agreement at issue cannot be characterised as exclusive. In their submission, it confirms an order concerning 1.7% of Edeka's outlets. The agreement contains no other information concerning the possibility that it might be an exclusivity agreement.

104 The applicants' argument is not persuasive inasmuch as the Commission twice adduced evidence that the agreement was exclusive. In support of its argument, the Commission submits an internal note of 24 September 2000, which refers to an existing exclusivity agreement ('bestehenden Exclusivvertrag').

105 Furthermore, concerning the argument that the Commission referred to Edeka as a whole, it must be noted that that is not the case. The Commission specifies that the agreement is limited to Edeka's new outlets.

106 Accordingly, this complaint must be rejected.

- COOP Schleswig-Holstein (2000)

107 The applicants claim that there was no exclusivity agreement. They submit that all that the Commission had was a letter to COOP from them confirming the purchase of 25 RVMs by COOP. There was no exclusivity agreement, particularly since, in the applicants' submission, COOP purchased only seven RVMs. The applicants conclude that the parties did not consider that they were bound by that letter.

108 The applicants further submit that the letter of 10 March 2000 contained no provision that prevented COOP from obtaining supplies from their competitors.

109 However, the evidence provided by the Commission, according to which the letter of 10 March 2000 from Tomra Systems GmbH to COOP concerned a 'framework exclusive agreement', is sufficient to show that the agreement in question was indeed an exclusivity agreement.

110 For that reason, this complaint must be rejected.

- Netto

111 The applicants maintain that the progressive bonus included in the agreement was not delivered, since the customer had not ordered the number of RVMs necessary to obtain the bonus. In order to have been entitled to two free RVMs, Netto would have had to order 150 RVMs, whereas, according to the applicants, the order was for only 109 units in 2001 and 126 units in 2002.

112 The applicants further submit that the target was not reached, since actual purchases were significantly below the target, and the fact that the agreement was unilaterally extended by them beyond the date initially provided for is irrelevant, since the customer can have had no expectation that they would extend it, which precludes there having been any deterrent effect in this instance.

113 With regard to the agreement with Netto, the Court notes that, although it is the case that the thresholds were not reached, the contested decision states, at recital 202, that the agreement had been extended, presumably with the aim of allowing the customer to reach the thresholds thereafter. The applicants have not shown that that is not the case. Accordingly, this complaint must be rejected.

- Rewe Wiesloch and Rewe-Hungen (1997)

114 The applicants deny that they were the exclusive supplier of RVMs to those two organisations, and assert that the contract for the supply of RVMs was awarded to Halton in 1997.

115 There is no need, however, to adjudicate on this complaint, since the agreement falls outside the period considered in the contested decision.

- Rewe Hungen (2000)

116 The applicants argue that the agreement is not contrary to Article 82 EC, since the order that Rewe placed with them was significantly below (less than 50% of) the customer requirements. Moreover, in the applicants' submission, the volume actually purchased was significantly above the target agreed between the parties.

117 It must be stated that the fact that a customer does not exceed a contractually agreed threshold does not detract from the additional inducement given by the rebate until the target volume is reached. Furthermore, when calculating the size of the volume which was not available to competitors, the Commission took into account only the volume up to the threshold (here 20 machines), and regarded the rest of the purchases by the customer as contestable demand.

118 Consequently, this complaint must be rejected.

119 In view of the foregoing, it must be concluded that the contested decision is not vitiated by a manifest error of assessment so far as the agreements entered into in Germany are concerned.

The Netherlands

120 The applicants dispute the Commission's assessment of four contracts examined in the section of the contested decision dealing with the Netherlands. In their submission, the contested decision incorrectly characterises those contracts, which could not be relevant in respect of an infringement of Article 82 EC.

121 The Commission disputes the arguments put forward by the applicants.

- Albert Heijn (1998-2000)

122 The applicants claim that their unsigned order confirmation dated 30 October 1998, which refers to the fact that Albert Heijn had placed a telephone order for 200 RVMs, contains no information to support the assertion that Albert Heijn was required to purchase 200 RVMs or that the price would change if the customer chose to purchase a lower volume of RVMs, or that that quantity constituted the customer's total or near-total requirement. In fact, they claim that the Commission acknowledges that in April 2000 Albert Heijn had purchased only 121 RVMs from them.

123 It must none the less be pointed out that the document in question expressly states that Albert Heijn was required to purchase 200 RVMs: 'Albert Heijn obliges itself to buy 200 Tomra T600 automatic machines before 31 December 2000, with an extension until (and including) 31 March 2001'.

124 It should also be noted that, as the Commission has pointed out, the applicants submitted that agreement to it with their response of 14 March 2002 to the Commission's request for information under Article 11 of Regulation No 17 and that, in their reply to the statement of objections, the applicants did not at any time claim that the agreement, which was included with the statement of objections, was unsigned or had never existed.

125 In addition, with regard to the fact that, as at April 2000, Albert Heijn had purchased only 121 RVMs from the applicants, it should be noted that at that time a full year still remained before the deadline specified in the agreement. In view of those factors, there are no grounds for concluding that the agreement was not respected.

126 It should be added that the existence of that agreement is confirmed by the Royal Ahold Global Master agreement, examined below, which was a broader exclusivity agreement, entered into with Royal Ahold, the group to which Albert Heijn belonged. Clause 4.2 of that agreement refers to the previous agreement with Albert Heijn and recalls that 'under the agreement dated 30 October 1998' Albert Heijn is still 'obligated to purchase an additional 79 machines from [Tomra Systems BV]', which corresponds to the difference between the initial commitment to buy 200 machines and the 121 machines already purchased.

127 In view of the foregoing, it must be concluded that the document in question is evidence that a contract existed under which Albert Heijn was obliged to purchase 200 RVMs from the applicants.

- Royal Ahold (2000-2002)

128 The applicants deny that the Royal Ahold Global Master agreement was exclusive and maintain that there is nothing in the agreement to prevent Royal Ahold from purchasing competing machines. They claim that clause 1.2 of the agreement explicitly states that Royal Ahold is free to purchase from other suppliers and that Royal Ahold is not required to terminate any existing agreements with other RVM suppliers. The applicants maintain that they were in fact supposed to be only the 'primary supplier' and not the exclusive supplier.

129 The applicants further contend that the Commission, in the contested decision itself, acknowledges that Royal Ahold purchased RVMs from other suppliers during the contract period. Furthermore, they maintain that the statements referred to in the defence and their indications aimed at explaining the agreements do not, under the applicable contract law in the present case (that is to say, New York (United States) law), constitute evidence that would convince a court to enforce exclusivity against Royal Ahold. The fact that Royal Ahold made purchases from competing suppliers raises the question of what deterrent effect that contract may actually have had.

130 The applicants' arguments concerning the agreement with Royal Ahold cannot be accepted.

131 The documents cited in the contested decision confirm that that agreement was exclusive. The applicants' press release of 13 April 2000 states, for example, that '[the] Tomra [group] and ... Royal Ahold have entered into a global agreement that makes [the] Tomra [group] the exclusive provider of reverse vending machine technology and services during a period of three years' (see recital 139 to the contested decision). It should also be noted that the applicants, in their reply to the statement of objections, expressly recognised this.

132 It is certainly the case, as the Commission admits, that the agreement does not oblige Royal Ahold to terminate existing agreements with other RVM suppliers before they have run their term. However, clause 1.2 of the agreement provided that purchases of further RVMs from competitors were 'not prohibited', 'provided, however, that the term relating to such additionally installed machines shall not exceed the longest remaining term in existence at the retail account at which additional machines are installed'.

133 Clause 1.2 of the agreement in question thus provides that agreements with other suppliers were to be phased out and that contracts with competitors which would last beyond the longest remaining term at the level of each individual outlet would not be acceptable.

134 Accordingly, this complaint must be rejected.

- Lidl (1999-2000)

135 The applicants claim that, as regards the agreement of April 1999, the Commission omitted from the evidence the fact that Lidl's order states explicitly that Lidl Nederland GmbH does not commit itself to an exclusivity agreement.

136 The Commission also misrepresents the evidence where it states that Lidl's intention was to purchase 'at least' 40 RVMs, since the order letter states merely that Lidl intended to purchase 40 RVMs. The Commission also acknowledged that Lidl bought only 21 RVMs from the applicants in 1999.

137 The applicants observe that at recital 142 to the contested decision the Commission refers to an agreement in 2000 'to replace 44 old Halton and 33 old Tomra machines with 77 new Tomra RVMs by the end of that year'. No minimum quantity is specified. The letter merely confirms, rather, that Lidl had ordered 77 RVMs to replace the same number of old RVMs in its stores. The Commission acknowledged that Lidl actually bought 82 RVMs from the applicants in 2000. In the applicants' submission, the letter clearly demonstrates that Lidl was asking them to replace its machines because their advanced technology could be more specifically adapted to Lidl's needs.

138 As regards the agreement of April 1999, the Court finds that the contested decision does not characterise the agreement with Lidl as exclusive. The agreement is described under the heading 'Exclusivity and quantity commitments' at recital 142 to the contested decision and under the heading 'Quantity commitments and unilateral conditions relating to specific quantities' at recital 302 thereto. As to the applicants' statement that Lidl had purchased only 21 machines in 1999, it must be noted that that does not alter the fact that the agreement concerned obliged the customer to purchase 40 machines over a two-year period and that the contractual period of two years was not yet completed.

139 The applicants' submissions relating to that agreement must therefore be rejected.

140 As regards the agreement signed on 29 September 2000, suffice it to state, first, that the applicants do not dispute that the customer had committed to purchasing 77 machines by the end of the year and, second, that the contested decision did not take that agreement into account in the calculation of the market share which could not be contested by the applicants' competitors (see recital 163 and footnote 335 to the contested decision).

141 Consequently, this complaint cannot be accepted either.

- Superunie (2001)

142 The applicants claim, in essence, that Superunie is a central purchasing organisation governed by Netherlands law, that its members take their purchasing decisions independently and that an agreement concluded with an organisation of that type, which contained a commitment to purchase a minimum of 130 machines over a period of one and a half years, is not binding on the members.

143 The applicants' arguments cannot be accepted. In that regard, it is appropriate to refer to the points made in paragraphs 61 to 66 above in relation to agreements with purchasing organisations.

144 In light of the foregoing, it must be held that the contested decision is not vitiated by any manifest error of assessment, so far as the four Netherlands contracts raised by the applicants are concerned.

Sweden

145 The applicants maintain that the contested decision incorrectly characterises most of the agreements entered into in Sweden and that it is thus vitiated by a manifest error.

146 The Commission disputes the arguments advanced by the applicants.

- ICA Handlares (Sweden) and Hakon Gruppen (Norway) (2000-2002)

147 The applicants claim that, as this agreement is an attachment to the agreement between the Tomra group and Royal Ahold, it cannot be characterised as exclusive since the Royal Ahold agreement is not itself characterised as exclusive.

148 The applicants add that both ICA and Hakon are 'central administration offices' for wholly independent outlets. Thus, the applicants maintain that, even if the agreement in question had been exclusive, there would be nothing to prevent the outlets from obtaining supplies of RVMs from the applicants' competitors.

149 With regard to the argument concerning the characterisation of the agreement as exclusive, it should be noted that the Royal Ahold Global Master agreement has been examined above and that the conclusion drawn was that it was an exclusivity agreement (see paragraphs 128 to 133 above). The Commission was thus fully entitled to conclude that, as the ICA agreement was annexed to that agreement, it was by definition exclusive. It is also apparent from the file and in particular from the 'Proposal for an extra overriding discount in Norway and Sweden' that the Commission was correct in its assertion that the agreement also contained a clause giving entitlement to an additional discount at the time of purchase of RVMs. The Commission characterised the agreement as abusive because it did not contain just an exclusivity clause. A 10% rebate was paid to ICA, in addition to the specific rebate scheme put in place in Sweden, if it undertook to purchase a minimum of 1 100 new RVMs in Sweden and Norway over a period between 2000 and 2002. The object of the agreement was therefore not only to establish an exclusivity clause but also to ensure the customer loyalty of ICA and Hakon by means of a rebate that was paid above a fixed quantity of RVM purchases.

150 As regards the applicants' argument concerning the independence of the sales outlets, it is sufficient to refer to the points made in paragraphs 61 to 66 above in relation to central purchasing organisations.

151 For those reasons, this complaint must be rejected in its entirety.

- Rimi Svenska (2000)

152 The applicants assert that there was no quantity commitment arising from the April 2000 block order because Rimi Svenska had, during the period covered by the agreement, purchased only 23 RVMs for SEK 2.6 million, whereas the retroactive rebate was payable only if purchases exceeded SEK 7.5 million.

153 The applicants observe, in the reply, that the specified target was not reached in the context of the wider contract with ICA either.

154 Those arguments may not be taken into account, since the agreement with Rimi Svenska was replaced in October 2000 by a broader agreement, entered into with ICA Ahold, of which Rimi Svenska is a subsidiary, which provided for the same 10% discount but for more flexible conditions in order to obtain the discount. Rimi Svenska therefore did not lose the discount to which it was entitled under the previous agreement, which also made provision for a discount if it designated the Tomra group as primary supplier.

155 The letter of 2 November 2000, annexed to the application, cannot be ignored in this connection, given that it confirms that Rimi Svenska obtained a partial retroactive rebate and that the remainder would be paid to it in November 2000. The document also states that the block order is terminated and replaced by the ICA agreement, which, as noted in paragraph 154 above, was broader. It should also be stated that the applicants' assertion that, under the wider contract with ICA, the specified target was not reached either is not substantiated.

156 Therefore, this complaint must also be rejected.

- Spar, Willys and KB Exonen (Axfood group) (2000)

157 The applicants maintain that the Commission has adduced no evidence to show that the parties had an agreement under which Spar was 'entitled' to retroactive rebates. They claim that in their reply to the statement of objections they had specifically referred to the Axfood statement, which explains that Spar and Willys made their purchases under an agreement concluded between D-Gruppen and the Tomra group in 2000, which did not offer volume-based rebates.

158 The Court notes in that connection that the Commission, in the contested decision, states that the agreement in question contains retroactive rebates in return for the purchase of a given quantity of the applicants' products. The applicants submit that the Commission has not substantiated its findings.

159 However, the applicants appear to be advancing contradictory arguments. Initially, when they provided information to the Commission, the applicants stated that they had entered into agreements with Spar and Willys containing retroactive rebate clauses. Then, later, the applicants stated that those agreements (i) were either included within a broader agreement which did not include a rebate clause or (ii) had never existed. Finally, they maintained that certain evidence had disappeared.

160 In view of the evidence before the Court, this complaint must clearly be rejected.

- Axfood (2001)

161 The applicants maintain that this agreement was not binding and that Axfood was under no obligation to purchase the quantities concerned. They claim that Axfood purchased only half the agreed quantity.

162 It is common ground that this agreement was not exclusive and did not contain any commitment on quantities.

163 This agreement is cited at recital 314 to the contested decision under the heading 'Rebate schemes'. As recital 178 and footnote 389 to the decision indicate, the agreement provided for thresholds entitling the customer to retroactive rebates, by reference to the number of machines purchased. As the applicants do not challenge that finding, there is no need for the Court to adjudicate on this complaint.

- Axfood (2003-2004)

164 The applicants contend that this agreement was not exclusive because Axfood was explicitly authorised to test competing machines and there was nothing in the agreement to prevent it from purchasing such machines.

165 In that regard, it is sufficient to state that the agreement concerned does not fall within the temporal scope of the infringement established in the contested decision. It is therefore unnecessary to rule on this complaint.

166 In view of the foregoing, it must be concluded that the contested decision is not vitiated by any manifest error of assessment so far as the agreements concluded in Sweden are concerned.

Norway

167 The applicants claim that all the agreements concerning Norway (100% of the applicants' sales in Norway) on which the Commission relies either did not exist or did not contain any exclusivity clauses, quantity commitments or retroactive discounts.

168 The Commission disputes the arguments advanced by the applicants.

- KøffHedmark and Rema 1000 (1996), AKA/Spar Norge (1997)

169 The applicants maintain that the alleged agreements with the abovementioned customers are not exclusivity agreements but mere offer letters quoting prices.

170 The applicants emphasise, moreover, that, as the agreements were concluded in 1996 and 1997, they fall outside the temporal scope of the contested decision.

171 In view of the fact that these three agreements do in fact fall outside the temporal scope of the contested decision, there is no need for the Court to rule on this complaint.

- NorgesGruppen, Hakon Gruppen, NKL (COOP) and Rema 1000 (1999-2000)

172 The applicants first of all complain that the Commission has not separated out the various agreements at issue and, as regards the relevant recitals, that they are 'difficult to read'.

173 As regards the agreement with NorgesGruppen, the applicants claim that the agreement is not exclusive. They assert that the customer's total demand was 1 300 RVMs. However, only 635 RVMs were ordered from the applicants. As the needs were greater than the demand, the applicants conclude that further supplies were obtained from competitor undertakings. Thus the Commission failed to take into account a letter of 13 October 1998 confirming the customer's actual requirements.

174 In addition, the applicants maintain that the Commission itself acknowledged that NorgesGruppen was not bound to purchase a minimum quantity of RVMs from them. The order for those machines and the confirmation explicitly state that the agreement was not exclusive.

175 Last, the applicants observe that the Commission raised a relevant argument with respect to the discount that NorgesGruppen would have had to repay had it purchased fewer than 500 RVMs. However, the reality shows that that was not the case, since 635 units were ordered. The Commission has been unable to demonstrate how the agreement differs from an ordinary volume rebate.

176 As regards Hakon Gruppen, the applicants claim that no minimum purchase threshold was set, since the customer's order confirmation states that the customer could purchase a lower number of RVMs, although it was entitled to a rebate only for machines actually purchased. That evidence is not, they claim, contested by the Commission.

177 Furthermore, the applicants complain that the Commission failed to take account of evidence confirming that the agreement was not exclusive, and in particular of the minutes of a meeting of the 'Panel for the Shops' held on 2 February 1999.

178 As regards NKL (COOP), the applicants claim that the customer was not required to purchase the agreed volume and that the contested decision does not prove that the applicants adjusted the order to suit the customer's individual demand. The applicants likewise assert that NKL also obtained supplies from their competitor Lindco during the period in question.

179 As concerns, last, the agreement with Rema 1000, the applicants claim that the alleged agreement offering discounts on the basis of an order for 200 machines was not signed by the customer and did not oblige Rema 1000 to purchase a minimum quantity of RVMs. They refer to Rema 1000's statement that other suppliers could not compete with the applicants' quality and level of service. The applicants add, last, that Rema 1000 mainly obtained supplies from their competitors.

180 It is appropriate to examine the NorgesGruppen, Hakon Gruppen, NKL (COOP) and Rema 1000 agreements at the same time. In essence, the applicants take issue with the Commission for having characterised these agreements as exclusive and for having maintained that the applicants provided for a progressive rebate in return for an order of a given quantity.

181 There is no need for the Court to adjudicate on the argument concerning the existence of exclusivity agreements, given that the contested decision did not characterise the agreements concerned as exclusive (see recital 302 to the contested decision).

182 With regard to the argument concerning the characterisation of quantity commitments in return for sizeable discounts, the Commission, rightly, specified that what is important in ascertaining whether commitments exist is whether the practices create an incentive not to purchase from competitors.

183 With regard to the agreement entered into with Rema 1000, it is clear that the discount is large in proportion to the number of machines to be purchased (14% for 200 machines), given that, for other contracts, the discount was the same but for a much larger number of purchases (500 machines). Moreover, for all the other contracts, the applicants proposed to grant the discounts prior to purchase of the RVMs and stipulated that, if the target quantity was not reached, the customers would have to pay back the discount in respect of the number of machines that were not ordered. It follows, as the Commission quite rightly points out, that the incentive for the customer is greater than if the discount had been given after each order.

184 Consequently, this complaint must be rejected.

- NorgesGruppen (2000-2001)

185 As regards the NorgesGruppen agreement for 2000 and 2001, the applicants claim that there was no agreement with that customer, contrary to the Commission's assertion. What did exist was an unsigned offer letter.

186 In that regard, it must be stated that the unsigned offer letter attached to the application was produced by the applicants to the Commission in their reply to a request for information, which asked the applicants to indicate all agreements, including those which had been concluded in a less formal manner.

187 It should be added that, on the basis of the material in the file, it may be confirmed that the applicants offered an advance discount on the basis of a target volume of 150 RVMs and supplied the customer at the reduced price (see recital 247 and footnote 547 to the contested decision).

188 Although it is the case that the document attached to the application is an unsigned offer letter, it is nevertheless the case that the discount was granted and that sales were made on the terms set out in that document. It is also apparent that the customer bought nothing from competing suppliers during this period.

189 Finally, the contested decision itself states that the document was an unsigned offer and that the purchasing target had not been met. In the light of those considerations, this complaint must be rejected.

- NKL (COOP) and Rema 1000 (2000-2001)

190 As regards NKL (COOP), the applicants acknowledge that a 10% discount was offered to NKL if purchases exceeded 150 RVMs. However, the applicants assert that NKL did not sign that offer and that the agreement was not finally concluded. The applicants claim that the fact that the customer's actual purchases were below the initial target in the offer proves that there was no binding commitment.

191 As concerns Rema 1000, the applicants adopt the same reasoning as for the agreement with NKL (COOP), inasmuch as a discount was also offered if 70 machines were purchased and that the customer refused the offer. The applicants therefore deny that an agreement was concluded between the parties.

192 The Court rejects the applicants' argument that the fact that the purchases actually made by the two customers were lower than the target initially set in the offer proves that there was no binding commitment.

193 On that point, it must be observed, in the first place, that the contested decision does not state that those two customers (NKL (COOP) and Rema 1000) were contractually bound to purchase a given number of machines. The two companies are mentioned at recital 302 to the contested decision under the heading 'Quantity commitments and unilateral conditions relating to specific quantities'. In both cases, discounts were conditional upon the purchase by the customer of a large number of machines over a period of around one year. In the case of Rema 1000, the rebate scheme provided for an immediate conditional rebate (10% for 70 machines) and an additional retroactive rebate of 3% for 85 machines. The customer purchased 73 machines (see recital 261 to the contested decision).

194 In the second place, the contested decision considers the relatively flexible contractual terms offered, for example, to NKL (COOP) in comparison with other Norwegian customers and takes them into account (see recital 256 and, in relation to the impact of the applicants' practices, footnote 604). In fact, the Commission points out, rightly, that it was not necessarily decisive whether or not a specific target was met as long as no, or only insignificant, purchases were made from competitors (see recital 312 to the contested decision).

195 For those reasons this complaint must be rejected.

196 In light of the foregoing, the Court rejects the applicants' argument that all the agreements relating to Norway were wrongly characterised in the contested decision.

197 The second part of the first plea and, accordingly, the first plea in its entirety must therefore be rejected.

B - Second and fourth pleas: manifest errors of assessment in relation to the question whether the agreements were capable of foreclosing competition, and failure to state reasons

198 The second and fourth pleas in the application have been rearranged under this plea, which is structured in three parts. First, the applicants maintain that the Commission made a manifest error of law in holding that exclusivity agreements, individualised quantity commitments and individualised retroactive rebates are unlawful per se under Article 82 EC and in failing to explain the test or criteria which it used when assessing whether the agreements were capable of restricting or foreclosing competition. Second, the Commission failed to consider whether the contestable part of the RVM market was sufficiently large to enable equally efficient competitors to remain on the market. Third, the Commission's assessment of the capability of the alleged retroactive rebates to foreclose competition is based on incorrect and misleading evidence and assumptions.

1. The alleged per se unlawfulness of the applicants' agreements and the failure to explain the test or criteria used by the Commission when assessing whether the agreements were capable of restricting or foreclosing competition

(a) Arguments of the parties

199 First, the applicants complain that the Commission made a manifest error of law in not including in its legal assessment the market context in which the three types of agreement functioned.

200 The applicants claim that the Commission misconstrued the test set out in Michelin II. They submit that according to Michelin II the contested decision must show that the agreements are 'capable' of restricting competition. That requires an examination of the market context. In effect, nothing, or virtually nothing, would remain of the test set out in Hoffmann-La Roche v Commission if, under Michelin II, the Commission were required to examine only the content of an agreement under Article 82 EC.

201 The applicants further assert that the per se test in the contested decision will lead to the prohibition of a large number of agreements in the internal market, in various sectors, even if they strengthen competition instead of restricting it, depending on the market context. There is no basis in economic theory or business practice for the argument that exclusivity agreements, individualised quantity commitments and individualised rebates always or almost always lead to a restriction of competition when they are used by a company in a dominant position.

202 The applicants claim that by not considering certain factors the Commission did not assess whether, as a matter of law, their practices were capable of having a restrictive effect on competition.

203 The factors which the Commission failed to consider are the fact that the applicants were the only RVM manufacturer able to offer the 'revolutionary horizontal in-feed technology' between 1997 and 2001; the fact that the applicants' competitors could still offer their machines to at least 61% of the total RVM market between 1998 and 2002; the fact that the applicants sold their machines directly to the end customer (the supermarket chains) and that the agreements in question did not prevent their competitors' access to distributors; and, last, the fact that the supermarket chains are professional buyers which were capable of comparing the applicants' RVMs with their competitors' RVMs and thus deciding for themselves which RVMs had the right price, the right quality, reliability and technology and the right level of service.

204 Second, the applicants argue that the contested decision is based on the legal assumption that Article 82 EC requires only that the Commission prove the existence and the form of the agreements and that it does not contain a sufficient statement of the reasons why any of the 49 agreements was capable of foreclosing competitors from the RVM market.

205 The Commission disputes the arguments advanced by the applicants.

(b) Findings of the Court

206 It is appropriate to recall that, according to settled case-law, the concept of abuse is an objective concept relating to the behaviour of an undertaking in a dominant position which is such as to influence the structure of a market where, as a result of the very presence of the undertaking in question, the degree of competition is weakened and which, through recourse to methods different from those which condition normal competition in products or services on the basis of the transactions of commercial operators, has the effect of hindering the maintenance of the degree of competition still existing in the market or the growth of that competition. It follows that Article 82 EC prohibits a dominant undertaking from eliminating a competitor and from strengthening its position by recourse to means other than those based on competition on the merits. The prohibition laid down in that provision is also justified by the concern not to cause harm to consumers (Case T-65-98 Van den Bergh Foods v Commission [2003] ECR II-4653, paragraph 157).

207 Whilst the finding of a dominant position does not in itself imply any criticism of the undertaking concerned, that undertaking has a special responsibility, irrespective of the causes of that position, not to allow its conduct to impair genuine undistorted competition on the common market (Case 322-81 Nederlandsche Banden-Industrie-Michelin v Commission [1983] ECR 3461, paragraph 57, and Case T-201-04 Microsoft v Commission ECR II-3601, paragraph 229). Likewise, whilst the fact that an undertaking is in a dominant position cannot deprive it of its entitlement to protect its own commercial interests when they are attacked, and whilst such an undertaking must be allowed the right to take such reasonable steps as it deems appropriate to protect those interests, such behaviour cannot be allowed if its purpose is to strengthen that dominant position and thereby abuse it (Case 27-76 United Brands and United Brands Continentaal v Commission [1978] ECR 207, paragraph 189, and Michelin II, paragraph 55).

208 It should also be recalled that, according to case-law, an undertaking which is in a dominant position on a market and ties purchasers - even if it does so at their request - by an obligation or promise on their part to obtain all or most of their requirements exclusively from that undertaking abuses its dominant position within the meaning of Article 82 EC, whether the obligation in question is stipulated without further qualification or whether it is undertaken in consideration of the grant of a rebate. The same applies if the undertaking in question, without tying the purchasers by a formal obligation, applies, either under the terms of agreements concluded with these purchasers or unilaterally, a system of fidelity rebates, that is to say, discounts conditional on the customer's obtaining all or most of its requirements from the undertaking in a dominant position (Hoffmann-La Roche v Commission, paragraph 89).

209 Obligations of this kind to obtain supplies exclusively from a particular undertaking, whether or not they are in consideration of rebates or of the granting of fidelity rebates intended to give the purchaser an incentive to obtain his supplies exclusively from the undertaking in a dominant position, are incompatible with the objective of undistorted competition within the common market, because they are not based on an economic transaction which justifies this burden or benefit but are designed to remove or restrict the purchaser's freedom to choose his sources of supply and to deny producers access to the market (Hoffmann-La Roche v Commission, paragraph 90).

210 With more particular regard to the granting of rebates by an undertaking in a dominant position, it is apparent from a consistent line of decisions that a loyalty rebate, which is granted in return for an undertaking by the customer to obtain his stock exclusively or almost exclusively from an undertaking in a dominant position, is contrary to Article 82 EC. Such a rebate is designed through the grant of financial advantage, to prevent customers from obtaining their supplies from competing producers (Joined Cases 40-73 to 48-73, 50-73, 54-73 to 56-73, 111-73, 113-73 and 114-73 Suiker Unie and Others v Commission [1975] ECR 1663, paragraph 518, and Michelin II, paragraph 56).

211 A rebate system which has a foreclosure effect on the market will be regarded as contrary to Article 82 EC if it is applied by an undertaking in a dominant position. For that reason, the Court has held that a rebate which depended on a purchasing target being achieved also infringed Article 82 EC (Michelin II, paragraph 57).

212 Quantity rebate systems linked solely to the volume of purchases made from an undertaking occupying a dominant position are generally considered not to have the foreclosure effect prohibited by Article 82 EC. If increasing the quantity supplied results in lower costs for the supplier, the latter is entitled to pass on that reduction to the customer in the form of a more favourable tariff. Quantity rebates are therefore deemed to reflect gains in efficiency and economies of scale made by the undertaking in a dominant position (Michelin II, paragraph 58).

213 It follows that a rebate system in which the rate of the discount increases according to the volume purchased will not infringe Article 82 EC unless the criteria and rules for granting the rebate reveal that the system is not based on an economically justified countervailing advantage but tends, following the example of a loyalty and target rebate, to prevent customers from obtaining their supplies from competitors (see Hoffmann-La Roche v Commission, paragraph 90, and Michelin II, paragraph 59).

214 In determining whether a quantity rebate system is abusive, it will be necessary to consider all the circumstances, particularly the criteria and rules governing the grant of the rebate, and to investigate whether, in providing an advantage not based on any economic service justifying it, the rebates tend to remove or restrict the buyer's freedom to choose his sources of supply, to bar competitors from access to the market, to apply dissimilar conditions to equivalent transactions with other trading parties or to strengthen the dominant position by distorting competition (Michelin II, paragraph 60).

215 It may be concluded from that line of cases, as the applicants indeed maintain, that in order to determine whether exclusivity agreements, individualised quantity commitments and individualised retroactive rebate schemes are compatible with Article 82 EC, it is necessary to ascertain whether, following an assessment of all the circumstances and, thus, also of the context in which those agreements operate, those practices are intended to restrict or foreclose competition on the relevant market or are capable of doing so.

216 In the first place, consideration must be given, in this instance, to whether the Commission, in the contested decision, did not pay due account to the context in which the agreements concerned operated and, in the second place, to whether it provided an adequate statement of reasons for its conclusion concerning whether the agreements were capable of excluding competition.

217 In that regard, the Court notes that the contested decision, after examining the structure of the relevant markets and the positions held by the applicants and their competitors on that market and concluding that the applicants had a very strong dominant position (see recitals 12 to 96 to the contested decision), examined each of the applicants' practices individually (see recitals 97 to 133 to the contested decision). The contested decision then devoted long passages to the examination of the ability of those practices to distort competition in the circumstances of the case (see, in particular, recitals 159 to 166, 180 to 187, 218 to 226, 234 to 240, 264 to 277 and 286 to 329 to the contested decision).

218 Furthermore, the contested decision, after looking at the applicants' practices in each relevant national market in conjunction with the size of the customers, the term of the agreements, the development of demand on the relevant market and the percentage of the tied part of demand, established that those practices were capable of impairing the emergence or growth of competition and concluded that there was an abuse where those practices tended to foreclose a significant part of demand. Concerning, in particular, the discount and rebate schemes employed by the applicants, the contested decision illustrates with diagrams the 'suction effect' of certain schemes for each country.

219 The Commission, even though the case-law does not require it, also analysed, in the light of market conditions, the actual effects of the applicants' practices.

220 With regard to the other factors which the applicants claim the contested decision should have examined in order to establish whether their practices were capable of restricting competition, the Court makes the following observations.

221 First, the applicants' alleged technical superiority - as the only manufacturers of RVMs capable of offering 'revolutionary horizontal in-feed technology' between 1997 and 2001 - cannot affect the assessment of whether the agreements were capable of restricting competition. Only the assessment of the applicants' competitive position on the market and thus their dominant position might possibly be affected by that factor.

222 Second, the fact that the machines were sold directly to the end customer is a factor which reinforces the finding of abuse of a dominant position, and not the reverse. Admittedly the agreements concerned did not in theory prevent competitors' access to distributors, but it is none the less obvious that distributors had no interest in buying machines, given that the applicants' agreements prevented their competitors from offering their machines to the end customer.

223 Third, with regard to the fact that the supermarket chains are professional buyers which were able to compare the applicants' RVMs with competing RVMs and make a choice between them, the applicants' conduct was clearly designed to set up schemes which provided an incentive to customers not to purchase from other suppliers and to maintain that situation.

224 Finally, as the Commission has pointed out, the applicants had every opportunity to put forward an objective economic justification of their practices which was not anti-competitive. They could have explained what efficiency gains they thought they might make from their exclusivity agreements, quantity commitments and individualised rebate schemes. However, the applicants do not maintain before the Court that their conduct gave rise to the least discernible efficiency gain, was otherwise justified or resulted in lower prices or any other benefits for consumers.

225 In light of the foregoing, the applicants' argument that the Commission analysed only the content of the relevant agreements and not the context in which they operated must be rejected.

226 In the second place, the complaint alleging that the reasons stated in relation to this aspect of the contested decision were insufficient cannot succeed either.

227 The statement of reasons required by Article 253 EC must be appropriate to the act at issue and must disclose in a clear and unequivocal fashion the reasoning followed by the institution which adopted the measure, in such a way as to enable the persons concerned to ascertain the reasons for the measure and to defend their rights and the Court to exercise its power of review (Joined Cases 296-82 and 318-82 Netherlands and Leeuwarder Papierwarenfabriek v Commission [1985] ECR 809, paragraph 19, and Case C-114-00 Spain v Commission [2002] ECR I-7657, paragraph 62). In the case of a decision adopted pursuant to Article 82 EC, that principle requires that the contested decision mention facts forming the basis of the legal grounds of the measure and the considerations which led to the adoption of the decision (see, to that effect, Case T-340-03 France Télécom v Commission [2007] ECR II-107, paragraph 57, not appealed on that point).

228 In that regard, reference must be made to the points made in paragraphs 216 to 218 above, from which it can clearly be seen that the Commission gave a detailed explanation of the reasons which led it to believe that the agreements in question were capable of restricting or excluding competition.

229 It cannot therefore be denied that the applicants were able to ascertain all the reasons for this aspect of the contested decision. Furthermore, the Court has been fully able to carry out its review of the legality of the contested decision. It follows that sufficient reasons have been stated for this aspect of the contested decision.

230 In light of the foregoing, this part of the second plea must be rejected.

2. 'Insufficient coverage' of total demand for RVMs by the applicants' practices

(a) Arguments of the parties

231 The applicants claim that even if the contested decision had demonstrated that all the agreements in question might have had foreclosure effects, that would prove only that competitors would have been foreclosed from supplying customers which had already concluded those agreements (whose existence is still disputed by the applicants). Competitors would none the less remain free to seek the customers of other undertakings. In order to establish an infringement of Article 82 EC, the contested decision ought to have demonstrated that those agreements covered such a broad part of the market that they were capable of foreclosing a sufficient number of competitors from the whole market, to the extent of causing a significant reduction in competition. The Commission does not explain why the fact that competitors were unable to sell their RVMs to specific customers would lead to their foreclosure from the market as a whole.

232 The applicants argue that the relevant question would be whether a competitor could profitably remain on the market by serving only the contestable part of the market, and that the Commission ought to have determined the size of minimum profitability required to operate on the market concerned. If the contestable demand was sufficiently large, and the scale of profitability sufficiently reduced, to enable a potential competitor to enter or remain on the market alongside the applicants, the Commission ought to have concluded that the applicants' practices were not abusive. Nor did the Commission clearly indicate the market share that would need to be covered by the agreements in order for the agreements to be capable of foreclosing competitors. The contested decision does not provide any objective criterion for determining where the threshold might be found.

233 In the applicants' submission, if the Commission had carried out such an analysis, it would scarcely have been possible for it to prove that the agreements in question could have excluded equally efficient competitors from the market. The applicants emphasise that it was for the Commission to demonstrate in the contested decision that their practices were capable of having an exclusionary effect. In the absence of sufficient reasoning in the contested decision, the applicants are under no obligation to prove the contrary.

234 The applicants observe that the new analysis in the defence, relating to the exclusionary effect of their practices, is inadmissible. The question for the Court is whether the contested decision contained sufficient reasoning on that point. The defendant cannot be permitted, generally, to remedy the errors and omissions in the contested decision by submitting a new analysis and additional evidence during the proceedings before the Court.

235 The applicants reject as irrelevant the assertion that it is not for the dominant undertaking to determine the number of competitors on the market. Different markets may be fully competitive even if the number of competitors varies and sometimes even where there are only two competitors. It was, in their submission, for the Commission to determine, in the contested decision, the minimum viability threshold for undertakings on the relevant market and to determine whether the size of the contestable portion of the market allowed a sufficient number of them to operate in such a way that competition was effective. That was not done in this instance.

236 Last, the applicants claim that their practices did not cover a sufficiently large proportion of the total demand. They contend that the contestable portion of the market was, for each national market, at least 30% and in most cases over 50% and, for the five markets taken together, on average around 61%, that is to say, more than 2 000 machines per year. That figure is higher than the minimum level of sales necessary to ensure the viability of a producer of RVMs, which the applicants estimate to be between 500 and 1 000 units per year.

237 The Commission disputes the arguments advanced by the applicants.

(b) Findings of the Court

238 It should be stated, first of all, that, in essence, the question that arises is whether the Commission, in order to prove the foreclosure of competitors from the market as a whole, ought to have determined the minimum viability threshold necessary to operate on the relevant market and ought then to have determined whether the non-contestable portion of the market (that is to say, the part of demand tied by the applicants' practices) was sufficiently large to be capable of having an exclusionary effect vis-à-vis competitors.

239 In the present case, the Commission, in the contested decision, found that, in the countries and during the years in respect of which the finding of infringement was made, the part of demand foreclosed was 'substantial' or 'not insubstantial' and that, particularly during the 'key years' of growth on each of the relevant markets, it represented a very significant proportion' (see recital 392 to the contested decision). The contested decision did not, however, establish a precise threshold beyond which the applicants' practices would be capable of excluding competitors.

240 The Commission, rightly, held that by foreclosing a significant part of the market, as in the present case, the dominant undertaking restricted entry to one or a few competitors and thus limited the intensity of competition on the market as a whole.

241 In fact, the foreclosure by a dominant undertaking of a substantial part of the market cannot be justified by showing that the contestable part of the market is still sufficient to accommodate a limited number of competitors. First, the customers on the foreclosed part of the market should have the opportunity to benefit from whatever degree of competition is possible on the market and competitors should be able to compete on the merits for the entire market and not just for a part of it. Second, it is not the role of the dominant undertaking to dictate how many viable competitors will be allowed to compete for the remaining contestable portion of demand.

242 In that connection, only an analysis of the circumstances of the case, such as the analysis carried out by the Commission in the contested decision, may make it possible to establish whether the practices of an undertaking in a dominant position are capable of excluding competition. It would, however, be artificial to establish without prior analysis the portion of the tied market beyond which the practices of a dominant undertaking may have an exclusionary effect on competitors.

243 In particular, in the first place, the applicants' practices foreclosed, on average, a considerable proportion (two fifths) of total demand during the period and in the countries under consideration. Consequently, even accepting the applicants' proposition that foreclosure of a small portion of demand does not matter, that portion was far from being small in the present case.

244 In the second place, the applicants' practices often led to a very high proportion of 'tied' demand during the 'key years' in which demand was highest and would have been most likely to attract successful market entries, in particular the years 1999 and 2000 in Austria, 2001 in the Netherlands and 1999 in Norway (see, for example, recitals 163, 219 and 237 to the contested decision).

245 In the third place, the applicants' practices tied end customer, not distributor, demand. Competitors were thus unable to use alternative distribution methods which could have mitigated the effects of the applicants' practices.

246 In light of the foregoing, this part of the second plea must be rejected.

3. The allegedly incorrect and misleading evidence and assumptions used as the basis for the assessment of the capability of the retroactive rebates to foreclose competition

(a) Arguments of the parties

247 The applicants maintain that the Commission's case on retroactive rebates is based on two factors: in the first place, on the fact that customers are not willing to buy more than a small number of machines from a new supplier and, in the second place, on the fact that the retroactive rebates allow the applicants to charge negative or very low prices. The applicants observe that in almost all the examples used by the Commission the prices were never capable of being negative and that in all cases competitors would have been able to earn positive revenues from their sales. They also observe that the Commission did not examine the applicants' costs either in order to establish the level below which prices would be exclusionary or predatory.

248 In the applicants' submission, where retroactive rebates lead to positive prices, it cannot be presumed that they will necessarily be capable of producing exclusionary effects. Such an approach would effectively lead to a per se prohibition of retroactive rebates.

249 The applicants further submit that the contested decision did not evaluate the prices resulting from their rebates against any benchmark or workable objective criterion. It merely asserts that the rebates imposed an ill-defined opportunity cost on competitors and that the resulting prices are therefore, in its subjective opinion, 'very low', whilst the Commission does not define what the latter expression is supposed to mean. The applicants contend that the Commission cannot rely on such assertions and subjective opinions to establish whether rebates are capable of having an exclusionary effect.

250 The applicants claim, moreover, that the finding in the contested decision that the retroactive rebates were capable of having exclusionary effects is based on incorrect diagrams.

251 In two of the seven cases which are cited in the contested decision (Figures 23 and 24 thereof relating to Austria), the Commission relies on diagrams which are inaccurate and misleading. The Commission's argument that competitors would have had to charge negative prices in those cases is incorrect in any circumstances.

252 In four other cases (Figures 15 and 18 relating to the Netherlands and Sweden respectively and Figures 21 and 22 relating to Germany), the Commission ignored the existence of rebates available to customers for sales below the threshold which the Commission used in its analysis. Once the error is corrected, prices are not negative in any circumstances in three of the four cases and are only marginally negative for sales of a single unit in the other case.

253 Contrary to the Commission's assertion, in six of the seven cases competitors would have been able to charge positive prices even if they sold only very small quantities, that is to say, two or three machines.

254 In each of the seven cases, the Commission incorrectly assumed that competitors would have been limited to selling a small number of RVM units.

255 In each of the seven cases, the Commission ignored relevant evidence on the functioning of the market, which undermines its findings. In particular, the Commission ignored revenues from after-sales services and follow-on sales of RVMs. When those revenues are taken into account, competitors could have expected to receive positive revenues, even in the case of sales of RVMs at negative prices.

256 The applicants submit that even if competitors had been obliged to sell only a small number of RVMs (for example, one or two machines), the Commission failed to demonstrate in the contested decision that the applicants' rebates would have been capable of excluding such competitors from the market.

257 The Commission disputes the arguments advanced by the applicants.

(b) Findings of the Court

258 It must be stated, first of all, that this complaint is based on an incorrect premiss. Contrary to the applicants' contention, the fact that the retroactive rebate schemes oblige competitors to ask negative prices from the applicants' customers benefiting from rebates cannot be regarded as one of the fundamental bases of the contested decision in showing that retroactive rebate schemes are capable of having anti-competitive effects.

259 However, a whole series of other considerations relating to the retroactive rebates operated by the applicants underpins the contested decision as regards its conclusion that those types of practices were capable of excluding competitors in breach of Article 82 EC.

260 In the first place, the contested decision finds that the incentive to obtain supplies exclusively or almost exclusively from the applicants was particularly strong when thresholds, such as those applied by the applicants, were combined with a system whereby the achievement of the bonus threshold or, as the case may be, a more advantageous threshold benefited all the purchases made by the customer during the reference period and not exclusively the purchasing volume exceeding the threshold concerned (see recitals 132, 297 and 316 to the contested decision).

261 In the second place, the Commission points out in the contested decision that the rebate schemes were individual to each customer and that the thresholds were established on the basis of the customer's estimated requirements and/or past purchasing volumes.

262 The contested decision refers, in particular, to the fact that a retroactive rebate scheme combined with one or more thresholds corresponding either to the customer's total requirements or a large proportion thereof represented a strong incentive for buying all or almost all the equipment needed from the applicants and artificially raised the costs of switching to a different supplier, even for a small number of units (see recitals 131 to 133, 297, 321 and 322 to the contested decision).

263 In the third place, the Commission finds that the retroactive rebates often applied to some of the applicants' largest customers with the aim of ensuring their loyalty (see, for example, recitals 180 and 240 to the contested decision).

264 Finally, the contested decision states that the applicants failed to show that their conduct was objectively justified or that it generated significant efficiency gains which outweighed the anti-competitive effects on consumers (see recital 391 to the contested decision).

265 It is certainly true that the contested decision illustrates with diagrams (see Figures 15, 18, 21 to 24 and 27) the fact that the applicants' retroactive rebates had an exclusionary effect by driving competitors to ask very low, sometimes negative, prices in respect of the last units before the rebate threshold (see recitals 165, 186, 224, 235, 236 and 268 to the contested decision).

266 However, the Commission, in the contested decision, first, does not state that the rebate schemes automatically resulted in negative prices and, second, does not maintain that showing that is a prerequisite to finding rebate schemes abusive. Moreover, the contested decision does not contain diagrams for each of the discount and rebate schemes employed by the applicants. It includes only one or two diagrams per country which illustrate the foreclosure effect of the applicants' rebate schemes.

267 In that regard, the exclusionary mechanism represented by retroactive rebates does not require the dominant undertaking to sacrifice profits, since the cost of the rebate is spread across a large number of units. If retroactive rebates are given, the average price obtained by the dominant undertaking may well be far above cost and ensure a high average profit margin. However, retroactive rebate schemes ensure that, from the point of view of the customer, the effective price for the last units is very low because of the 'suction effect'.

268 In view of the foregoing, the fact that some diagrams contain errors cannot, on its own, undermine the conclusions relating to the anti-competitive nature of the rebate schemes operated by the applicants. This complaint is therefore ineffective.

269 As regards the applicants' argument that competitors were not restricted to selling a small number of units to each customer, it must be stated that it is inherent in a strong dominant position, such as that occupied by the applicants, that, for a substantial part of the demand, there are no proper substitutes for the product supplied by the dominant undertaking. The dominant supplier is thus, to a large extent, an unavoidable trading partner (see, to that effect, Hoffmann-La Roche v Commission, paragraph 41). It follows that in those circumstances the contested decision was correct in stating that customers turned to alternative suppliers only for a limited portion of their purchases.

270 For the same reasons, it is difficult to concur with the applicants' argument that a competitor may offset the lower prices that it is obliged to charge a customer for units below the threshold by selling additional units to the same customer (above the threshold). In fact, that customer's remaining demand is at best limited, so the competitor's average price will remain structurally unattractive.

271 The same is true of the assertion that a competitor would be able to offset initial losses or low profitability caused by the applicants' practices with after-sales earnings (service and repair). The applicants' large installed base also gives them an obvious advantage in repair and maintenance of their machines so that it is not clear from the applicants' argument how competitors' structurally low margins in the primary market could be offset by profits on the after-sales market.

272 In light of the foregoing, the third part of the second plea - and consequently the second plea in its entirety - must be rejected.

C - Third plea: manifest errors in the Commission's assessment of whether the agreements did in fact foreclose competition

1. Arguments of the parties

273 The applicants maintain that the analysis of the actual effects was an integral part of the contested decision's findings on foreclosure. The passages of the contested decision relating to the 'impact' for each of the five countries confirm that that is so.

274 In their submission, the evidence used by the Commission is contradictory, speculative or irrelevant and, as a result, provides no basis for finding that the agreements did in fact eliminate competition.

275 First, the applicants claim that, in most of the five national markets analysed, their market share fell over the period in respect of which the Commission stated that they were engaging in anti-competitive practices.

276 The applicants contend that the graph presented by the Commission confirms their assertions concerning market share, namely that their market share was declining in three of the five countries concerned, which cannot be regarded as proof of an anti-competitive effect.

277 Second, the applicants dispute the Commission's argument that the position of their competitors in each of the five countries remained weak during the period analysed. They maintain that their competitors won market share in three countries, that their market share remained largely unchanged in Germany and that they lost market share only in Sweden.

278 Third, the applicants dispute the existence of an obvious link between the size of the tied market and their market share in each of the five national markets between 1998 and 2002. Examination of the five national markets as a whole reveals no evidence to suggest that a high tied market share leads to an increase in the applicants' market share. For example, in the Netherlands and in Norway, where the applicants' tied market share was highest, their market share fell, whereas in Germany and Sweden, where there was a lower tied market share, the applicants' market share rose or was stable. Only in Austria did their market share fall more rapidly than in the Netherlands and Norway.

279 The applicants claim, moreover, that there is no statistically significant link between the share of the non-contestable market and their market share for the five countries examined over the period analysed.

280 The applicants observe that the Commission relies on its own subjective interpretation of the evidence and rejects any objective criterion. The Commission maintains that the variables are 'connected' but rejects any attempt to subject its assertion to an objective and statistically robust test. Nor does the Commission adduce any evidence to support its assertion that the applicants' statistical analysis could have been manipulated.

281 Fourth, the applicants dispute the Commission's argument that their prices did not fall, and claim that the Commission ought to have examined the net effective (post-discount) prices and not their list prices.

282 The applicants emphasise that a proper analysis of the data, such as that carried out in the application, shows that prices fell in three of the five countries.

283 Fifth, the applicants claim that the fact that three competitors left the market does not prove that there were anti-competitive effects. Prokent became insolvent precisely when the applicants' allegedly anti-competitive practices ceased; and the applicants' acquisition of Halton and Eleiko contradicts the Commission's theory of negative effects on competition, since, if it were true that the applicants had the ability to contain and exclude their competitors, they would not have needed to buy those two companies in order to remove them from the market.

284 Last, the applicants claim that even if some of the facts adduced by the Commission were correct, they would not necessarily prove that the applicants' commercial prices had an anti-competitive effect. In the first place, the market share of an historic operator and the market position of its competitors may remain stable over time for perfectly legitimate reasons; in the second place, the Commission contradicts itself with respect to the applicants' prices and their evolution over time by characterising their pricing policy as predatory and at the same time accusing them of maintaining high prices; and, in the third place, the fact that one or more competitors leave the market does not prove an anti-competitive effect, but may simply be the result of the normal competitive process.

285 The Commission disputes the arguments advanced by the applicants.

2. Findings of the Court

286 It should be borne in mind that, according to well-established case-law, where some grounds of a decision on their own provide a sufficient legal basis for the decision, any errors in other grounds of the decision have, in any event, no effect on its enacting terms (see, by analogy, Joined Cases C-302-99 P and C-308-99 P Commission and France v TF1 [2001] ECR I-5603, paragraphs 26 to 29).

287 The contested decision clearly states, at recitals 285 and 332, that although, according to case-law, it is sufficient, for the purpose of establishing an infringement of Article 82 EC, to show that the applicants' practices tended to restrict competition or that their conduct was capable of having that effect, the Commission completed its analysis in the present case by considering the likely effects of the applicants' practices on the RVM market.

288 It is thus clear that the Commission did not attempt to base its finding of an infringement of Article 82 EC on that consideration of the actual effects of the applicants' practices on each of the national markets examined but that it merely complemented its finding of infringement with a brief examination of the likely effects of those practices.

289 It must also be stated that, for the purposes of establishing an infringement of Article 82 EC, it is not necessary to show that the abuse under consideration had an actual impact on the relevant markets. It is sufficient in that respect to show that the abusive conduct of the undertaking in a dominant position tends to restrict competition or, in other words, that the conduct is capable of having that effect (Michelin II, paragraph 239, and British Airways v Commission, paragraph 293).

290 In light of the foregoing, the third plea must be rejected as ineffective and it is not necessary to consider whether the evidence adduced by the Commission demonstrated that the agreements in question had actually eliminated competition. In fact, even if the Commission had made a manifest error of assessment, as the applicants allege, in holding that those agreements actually eliminated competition, the legality of the contested decision would not be affected.

D - Fifth plea: manifest error on the part of the Commission in holding that non-binding quantity commitments could infringe Article 82 EC

1. Arguments of the parties

291 The applicants refer to their first plea and assert that most of the 18 agreements (listed at recital 302 to the contested decision) envisaging quantity commitments were non-binding. In the applicants' submission, as in the case of non-binding exclusivity agreements, a non-binding quantity commitment, even if it represents all or virtually all of a customer's requirements, cannot preclude competition. If a customer is not legally bound to comply with a commitment to purchase a specific quantity from a supplier, it will be free to accept better offers from competing suppliers at any time. A non-binding individualised quantity commitment is simply an estimate.

292 The applicants contend that there is no legal basis in Community law for prohibiting a customer from giving his suppliers an estimate of his entire or near-entire needs in a specific period, even if one of the suppliers is in a dominant position. If that is so, the 18 agreements in question are not capable of producing anti-competitive effects and cannot therefore be relied on in the contested decision. If 18 of the 49 agreements on which the contested decision is based are disregarded, the foundation of the contested decision is undermined so decisively that, in the applicants' submission, the decision must be annulled in its entirety.

293 The applicants further submit that, contrary to what is alleged in the defence, many of the agreements do not link the price to the volume purchased but have a single unit price for each machine ordered (Lidl, COOP, etc.).

294 The Commission disputes the arguments advanced by the applicants.

2. Findings of the Court

295 As has been recalled at paragraphs 208 and 209 above, according to well-established case-law, an undertaking which is in a dominant position on a market and ties purchasers - even if it does so at their request - by an obligation or promise on their part to obtain all or most of their requirements exclusively from that undertaking abuses its dominant position within the meaning of Article 82 EC, whether the obligation in question is stipulated without further qualification or whether it is undertaken in consideration of the grant of a rebate. The same applies if the undertaking in question, without tying the purchasers by a formal obligation, applies, either under the terms of agreements concluded with these purchasers or unilaterally, a system of fidelity rebates, that is to say, discounts conditional on the customer's obtaining all or most of its requirements from the undertaking in a dominant position (Hoffmann-La Roche v Commission, paragraph 89).

296 In fact, obligations of this kind to obtain supplies exclusively from a particular undertaking, whether or not they are in consideration of rebates or of the granting of fidelity rebates intended to give the purchaser an incentive to obtain his supplies exclusively from the undertaking in a dominant position, are incompatible with the objective of undistorted competition within the common market, because they are not based on an economic transaction which justifies this burden or benefit but are designed to remove or restrict the purchaser's freedom to choose his sources of supply and to deny producers access to the market (Hoffmann-La Roche v Commission, paragraph 90).

297 In the present case, contrary to the applicants' submission, the Commission, in the contested decision, correctly considered the individualised quantity commitments not only in a purely formal way from the legal point of view but also taking into account the specific economic context in which the agreements in question operated. That was the basis for the Commission's conclusion, in the contested decision, that the agreements concerned were capable of excluding competitors.

298 Individualised quantity commitments, like those to which the contested decision refers at recital 302, which de facto tie and/or induce the purchaser to obtain all or most of its requirements from the dominant undertaking and which are not based on an economic transaction which justifies this burden or benefit but are designed to remove or restrict the purchaser's freedom to choose his sources of supply and to deny producers access to the market, - even if it is accepted that they do not bind the purchaser by a formal obligation - constitute an abuse of a dominant position within the meaning of Article 82 EC (see, to that effect, Van den Bergh Foods v Commission, paragraphs 84 and 160).

299 Even if a number of examples confirm that, as regards quantity commitments and rebates, the applicants allowed a degree of flexibility in relation to due observance of deadlines and targets, that flexibility, which was even applied to certain agreements that the applicants concede were 'binding', in no way diminishes the foreclosure caused by those practices. On the contrary, the Commission stated, rightly, in the contested decision that the exact volume of purchases was less important to the applicants than the customer's loyalty. Indeed, that flexibility helped to maintain the incentive to buy the applicants' RVMs, even with regard to customers which otherwise would not have reached the requisite thresholds (see recital 312 to the contested decision).

300 Moreover, the great majority of quantity commitments which the applicants describe as non-binding are agreements in which they made price and commercial terms conditional upon the customer purchasing a certain volume. Those agreements generally included a discount which was expressly made conditional upon achievement of a target. The customer was not legally bound to reach the target but was required to achieve it in order to obtain or keep the discount. Such agreements are present in this instance, for example those concluded with Axfood (2001), COOP (2000), NorgesGruppen or Hakon Gruppen. Those agreements are similar to a retroactive rebate. The risk of losing the discount retroactively is a strong incentive to the customer to reach the target. The fact that the applicants might ultimately have not required reimbursement of the discount is irrelevant, as is the absence of documented acceptance by the customer of an offer from the applicants. What matters are the expectations of the customer at the time when it placed the orders in conformity with the terms and conditions of the offer.

301 It follows from all the foregoing that the plea alleging a manifest error on the part of the Commission in holding that non-binding quantity commitments could infringe Article 82 EC must be rejected.

II - Claim for annulment or reduction of the fine

A - Arguments of the parties

302 The applicants maintain, in the framework of their sixth plea, that the Commission breached the principles of proportionality and non-discrimination in setting the fine at 8% of the Tomra group's worldwide turnover.

303 In response to the Commission's assertions set out in the defence, the applicants reiterate that the fine of EUR 24 million imposed by the Commission represents 7.97% of worldwide group turnover in 2005.

304 First, the applicants contend that, while the Commission is free to increase fines in order to strengthen their deterrent effect, the Commission's policy must none the less comply with requirements of proportionality, according to which fines for 'very serious' infringements have a stronger deterrent effect than those for 'serious' infringements. That logic is recognised by the guidelines on the method of setting fines, which provide that 'serious' infringements will attract a basic fine of between EUR 1 and EUR 20 million and that 'very serious' infringements will attract a basic fine in excess of EUR 20 million.

305 In the applicants' submission, the Commission followed that logic when it imposed a fine on Microsoft for a 'very serious' infringement. However, they point out, by way of comparison, that the fine imposed on Microsoft represented only 1.5% of its worldwide turnover, although the infringement was considered to be 'very serious'. The applicants take the view that that leads to the illogical conclusion that the Commission considers it more important to create a deterrent effect vis-à-vis them, a group of companies with a turnover of under EUR 300 million, for a 'serious' infringement, than it was to deter Microsoft, one of the five largest undertakings in the world, with a turnover (in 2003) in excess of EUR 30 billion, for a 'very serious' infringement. Likewise, the Commission imposed a fine on AstraZeneca for two 'serious' infringements which, after the need to impose a fine with a deterrent effect proportionate to the relevant profits was taken into account, amounted to only approximately 3% of its worldwide turnover.

306 The applicants observe that the Court of Justice has recognised that any significant change of approach by the Commission requires a detailed explanation. However, the Commission did not explain in the contested decision why the Tomra group, an undertaking which is not even among the 50 largest undertakings in Norway, received a fine representing 'the highest ever percentage of worldwide turnover of a company fined for violating the competition rules'.

307 The applicants maintain, in substance, that the contested decision does not contain a detailed explanation for that significant change in the Commission's approach to fines.

308 Second, the applicants contend that the level of the fine is disproportionate in light of the limited amount of their turnover on the geographic markets in question. The applicants estimate that less than 25% of their turnover is generated in Germany, the Netherlands, Austria, Sweden and Norway, and less than 34% in the EEA as a whole. In their submission, the Court of Justice has accepted that the principle of proportionality can be breached if the Commission ignores the relationship between worldwide turnover and the turnover 'accounted for by the goods in respect of which the infringement was committed'. The Commission could not, for that reason, merely take account of the fact that the infringements did not continue throughout the entire period under investigation on all the national markets concerned.

309 The Commission disputes the arguments advanced by the applicants.

B - Findings of the Court

310 As regards the applicants' argument alleging an infringement of the principle of non-discrimination in that the Commission set the fine at 8% of the applicants' worldwide turnover, it should, first of all, be borne in mind that, when determining the amount of a fine, the Commission must not infringe the principle of equal treatment, a general principle of Community law which, according to settled case-law, is infringed only where comparable situations are treated differently or different situations are treated in the same way, unless such difference in treatment is objectively justified (Case T-220-00 Cheil Jedang v Commission [2003] ECR II-2473, paragraph 104).

311 However, it must be emphasised, in that connection, that the Commission's practice in earlier decisions does not in itself serve as a legal framework for fines in competition matters. The fact that in the past the Commission may have applied fines of a particular level for certain types of infringements does not mean that it is stopped from raising that level within the limits indicated by Council Regulation (EC) No 1-2003 of 16 December 2002 on the implementation of the rules on competition laid down in Articles [81 EC] and [82 EC] (OJ 2003 L 1, p. 1) if that is necessary to ensure implementation of Community competition policy (see, by analogy, Case 100-80 Musique Diffusion française and Others v Commission [1983] ECR 1825, paragraph 109).

312 Furthermore, the gravity of infringements must be established by reference to numerous factors, such as the particular circumstances of the case, its context and the deterrent effect of fines, although no binding or exhaustive list of the criteria which must be applied has been drawn up (Case C-219-95 P Ferriere Nord v Commission [1997] ECR I-4411, paragraph 33). The relevant data, such as the markets, the products, the countries, the undertakings and the periods concerned differ in each case. It follows that the Commission cannot be compelled to impose on undertakings fines which amount to identical percentages of their respective turnovers in cases that are comparable so far as the gravity of the infringements is concerned (see, to that effect, Case T-67-01 JCB Service v Commission [2004] ECR II-49, paragraphs 187 to 189).

313 Since fines constitute an instrument of the Commission's competition policy, that institution must be allowed a margin of discretion when fixing their amount, in order that it may direct the conduct of undertakings towards compliance with the competition rules (Case T-49-95 Van Megen Sports v Commission [1996] ECR II-1799, paragraph 53).

314 In this case, the Court must thus reject at the outset the applicants' argument based on a comparison between the fine imposed on them and the fines imposed by the Commission in other decisions, since, as has been recalled above, the Commission's practice in earlier decisions cannot in itself serve as a legal framework for fines in competition matters. The Commission cannot be compelled to set fines which display perfect coherence with those imposed in other cases.

315 As to the applicants' argument that the contested decision marks a change in policy which warranted specific explanation, that cannot succeed either. The Commission, in setting the fine at issue, complied with its obligations under Regulation No 1-2003 and its own guidelines on the method of setting fines which are not in any event contested by the applicants. The level of the fine set by the Commission does not represent a change in its policy with regard to fines but, on the contrary, a standard application of that policy.

316 As to the allegedly disproportionate nature of the fine because of the limited amount of the respective turnovers of the applicants on the geographic markets in question, the Court observes that, subject to compliance with the upper limit provided for in Article 23(2) of Regulation No 1-2003, which refers to total turnover (see, by analogy, Musique Diffusion française and Others v Commission, paragraph 119), it is permissible for the Commission to take account of the turnover of the undertaking concerned in order to assess the gravity of the infringement when determining the amount of the fine, but it may not attribute disproportionate importance thereto by comparison with other relevant factors (Joined Cases C-189-02 P, C-202-02 P, C-205-02 P to C-208-02 P and C-213-02 P Dansk Rørindustri and Others v Commission [2005] ECR I-5425, paragraph 257).

317 In the present case, the Commission applied the method of calculation set out in the guidelines, which envisages that numerous factors are to be taken into account when the gravity of the infringement is assessed for the purpose of setting the fine. Those factors include, in particular, the nature of the infringement, its actual impact on the market, where this can be measured, the geographic size of the market affected and the necessary deterrent effect of the fine. Although the guidelines do not provide that the fines are to be calculated according to the overall turnover of the undertakings concerned or their turnover on the relevant product market, they do not preclude such turnover from being taken into account in determining the amount of the fine in order to comply with the general principles of Community law and where circumstances demand it (Dansk Rørindustri and Others v Commission, paragraphs 258 and 260).

318 It follows that, while it cannot be denied that the relevant product turnover may constitute an appropriate basis for assessing the threat to competition on the relevant product market within the EEA, it is nevertheless the case that that factor is not the sole criterion by reference to which the Commission must assess - and has indeed assessed in this case - the gravity of the infringement.

319 Consequently, contrary to the applicants' submission, restricting the assessment of the proportionate nature of the amount of the fine set by the Commission to a correlation between that amount and the relevant product turnover would confer excessive importance on the latter. The nature of the infringement, its actual impact, the geographic scope of the market affected and the necessary deterrent effect of the fine are further factors, taken into consideration by the Commission in the present case, which may justify to the required legal standard the amount of the fine.

320 In any event, as the Commission rightly states, the fact remains that the turnover achieved by the applicants in the markets affected by the infringement represents a relatively important part of their worldwide turnover, namely around 25%. As a consequence, it cannot be claimed that the applicants achieved only a small part of their worldwide turnover on the relevant markets.

321 It follows that the plea alleging that, having regard to the Commission's practice in previous decisions and the applicants' turnover in the markets concerned, the treatment which the applicants received was disproportionate and/or discriminatory must be rejected and, consequently, the claim for annulment or reduction of the fine must also be rejected.

322 It follows from all the foregoing that the action must be dismissed in its entirety.

Costs

323 Under Article 87(2) of the Rules of Procedure of the Court, the unsuccessful party is to be ordered to pay the costs if they have been applied for in the successful party's pleadings. As the applicants have been unsuccessful, they must be ordered to pay their own costs and those incurred by the Commission, in accordance with the form of order sought by the Commission.

On those grounds,

THE GENERAL COURT (Fifth Chamber)

hereby:

1. Dismisses the action;

2. Orders Tomra Systems ASA, Tomra Europe AS, Tomra Systems GmbH, Tomra Systems BV, Tomra Leergutsysteme GmbH, Tomra Systems AB and Tomra Butikksystemer AS to bear their own costs and to pay those incurred by the European Commission.