Chapter 8: C     \
where it stated that '[i]n asystem of tendering, competition is of the essence. If the tenders submitted by those taking part are not the result of individual economic calculation, but of knowledge of the tenders by other participants or of concertation with them, competit10ll is prevented, or at least distorted or restricted'.210 In Building and construction i„ fa Netherlands,1" the Commission condemned the rules and regulations of the umbrella assnq-ation of the federations of Dutch builders and contractors (SPO) which provided, among other things, for exchanges of information prior to tendering and for collusion on price ten ders for building and construction contracts. Moreover, it shared the demand side of the mar ket between members through the prior designation of successful tendering undertakings and the protection of enrided undertakings. The CFI, which rejected the appeal by the undertakings, stated that concertation by contractors regarding the manner in which they intend responding to an invitation for tender is incompatible with article 81 (1) of the Treaty, even if the invitation sets unreasonable conditions. It is for each contractor to determine independently what it regards as reasonable or unreasonable and to conduct himself accordingly'2" In Pre-insuktedpipes, the members of the cartel had set up asystem through which they could allocate individual projects between themselves by manipulating bidding procedures so that they could decide in advance which producer would be awarded the contract.213 In Raw Tobacco Italy, the processors of tobacco coordinated their conduct in respect of the bids placed at public auctions organised for the purchase of tobacco by the Italian monopoly ATI.214
8.43     Sharing sources of supply Cartel participants may also agree to share out their sources of supply. In Raw Tobacco Spain, the four Spanish processors of raw tobacco agreed each year on the quantities of each variety of raw tobacco that each of them would undertake to purchase from the group of producers.z,s In Raw Tobacco Italy, the tobacco processors also allocated to themselves suppliers and quantities to be purchased.216
(d) Co-ordinated Boycotts, Bans on Imports, Concerted Refusal to Deal
8.44     Carteitists may also agree upon concerted action to prevent the entry of new competitors to the market or upon retaliatory measures against undertakings refusing to comply with their restrictive arrangements.
8.45     Keeping competitors away from the cartel's market In several cases, the Commission has condemned coordinated measures designed to keep competitors away from the market In Meldoc, a particularly serious dimension of the cartel was that it was aimed at protecting
2'o [1973]OJL140/17,28.
2"  [1992]OJL92/l.
™ CvxT-29192 SPO and others v Commission [1995] ECRII-289, para 119.
2« [1999] OJL24/1. The cartel members had agreed that the other producers would submit higher öfters in the tendering procedure. In return, unsuccessful bidders would be granted the opportunity to participate in the project as subcontractors. There was also another 'compensation system' which corrected departures from the quota if the selected producer railed to win the bid. Likewise, a monitoring system was set up for the cases where a producer undercut the allocated 'favourite', and would be called to account and pressurised to withdraw its bid or to increase its price to allow the favourite win.
z<4 Decision of 20 October 2005 (full text of the decision available on DG COMP's web site), paras 126, 150,240.
2" Decision of 20 October 2004 (full text of the decision available on DG COMP's web site), para 67-
*<s Decision of 20 October 2005 (full text of the decision available on DG COMP's web site), paras 240 and 246-249.
770
B. 1      >gy of Cartel Arrangements and Common Features of Collusion
. putch market by blocking imports of cheaper milk from Belgium and Germany, thereby , „jng a clearly adverse effect on consumer interests and obstructing the achievement of one of r he most fundamental objectives of the Treaty, the integration of the economies of the Member Crates.217 A focus on import restrictions was also an important aspect of the cartel agreements in -Welded Steel Mesh. Here, a series of agreements between French producers and producers from Italy! Germany and Belgium who traditionally exported to France included terms setting quo-for import into France. These restrictions on the volume of foreign deliveries to the French market made it possible to enter into a price agreement; the prices charged on the French mar-ÍÍP ket were sec at a level far higher than the average in other Community countries, without this producing an increase in imports as would normally be expected.218 In Luxembourg Brewers, the cartel agreemenrwas intended to keep foreign brewers out of Luxembourg,2'9 In French Beefůie Commission condemned an agreement entered into by farmers' and slaughterers' federations, which had as its object a temporary commitment to suspend all imports of beef.220
Placing certain competitors at a competitive disadvantage Agreements may also be 8.46 reached in order to oblige other parties to adopt a given behaviour, or to place downstream operators at a competitive disadvantage. In Fedetab22^ the Commission condemned the agreement between Belgian and Luxembourg producers of manufactured tobacco not to supply several large distribution firms which did not stock a minimum range of brands. In Electrical and Mechanical Carbon and Graphite Products, cartel participants sold 'rough' carbon blocks in addition to their finished products and found themselves in competition on the downstream market with 'cutters', who bought the rough carbon blocks from them and transformed them into finished products. In order to limit competition from these cutters, the cartel had agreed to keep the cutters in a position of competitive disadvantage by selling the carbon blocks to them at an artificially high price.222
Boycott of reluctant undertakings Action may also be taken to boycott undertakings 8.47 showing reluctance to comply with the collusive arrangements, in Cement, cartel members agreed upon a collective response to what was perceived as the problem posed by the Greek producers, whose conduct was destabilising the cement industry. A 'stick and carrot' approach was devised in order to 'persuade' them to cooperate. The short term punitive actions considered, called 'stick actions', were aimed at defending European domestic markets.223
'" [1986] OJL348/50, para 82.
m [1989] OJfL260/I, para 159.
2,9 [2002] OJ L253/21. In this case, a common defensive mechanism was instituted whereby the parties agreed to consult each other first if a foreign brewer attempted to negotiate asupply contract with one of their tied oudets. Priority would then be allocated to one of the cartel participants, in an attempt to keep the out-lpt as a customer. If that party succeeded in negotiating a new contract with the oudet.it was obliged to compensate the party which had lost the outlet by transfetring an equivalent outlet to him. Other clauses allowed for the exclusion from the cartel of any patty which cooperated with a foreign brewer or distributed its beer.
220  £2003]OJL209/12,para38.
221   [1978]OJL224/29.
m [2004] OJL125/45 (full text of the decision available on DG COMP s web site), paras 154—156. 173 [1994] OJ L343/1, para 25(4). The 'stick' actions consisted of resorting, inter alia, to; administrative
obstacles; quality standards; action by associations; penalising of customers purchasing imported cement; attacks on the export matkets of the producers who were destabilising the market by taking the place of those producers in various countries or by making Greek exports unprofitable; boycotts of shipping lines controlled by the producers who were destabilising the market; adoption of 'guerilla' tactics in the Greek market; and «eking the assistance of international banks to 'convince' the recalcitrant producers to cooperate.
771
Chapter 8: Cartels
8.48
In Pre-insulatedPipes, the Commission found that the cartel participants had devised anii'":$ enforced a sophisticated strategy to eliminate a competitor who had refused to participate^ in their bid-rigging scheme. The award to this undertaking of the largest contract in '^ Germany for ten years sparked off a particularly violent reaction, and a collective boyCOM. 'J\ of this undertaking was instigated. The cartel decided to cease any supply to this undertak ■■ ingand its subcontractors.22''                                                                                                        :J
(e) Exchange of Commercially Sensitive Information                                                             ,;
Under normal competitive conditions, information regarding, for example, capacity, use0f capacity, production levels, customers, prices charged and conditions applied to customers, would not be exchanged between competitors. However, the exchange of such commercially sensitive information usually does take place within cartels, as a means to prepare, implement or monitor the restrictive agreement^) or arrangements. The question is whether such exchanges of information, where they are connected to an identified infringement, are also separately caught by the prohibition of Article 81 EC, and thetefote whether such exchanges are also caught when they take place in complete isolation, ie in the absence of any additional restrictive arrangements.
8.49     A constituent of the 'carteí offence' The exchange of commercially sensitive information is often ancillary to othetper se violations of Article 81 EC, such as price fixing, or market-sharing. It is thus not generally considered separately by the Commission, in that evidence of the exchange of confidential information is used to provide additional proof of the overall cartel arrangements, and the Commission finds a single cartel infringement (implicidy) including the exchange of confidential business information. In such instances, the exchange of information is found to have taken place either by way of preparation and implementation or for the monitoring of the cartel arrangements, and with the same object as was identified for the main (perse) infringement. It is worth noting, however, that the Commission has in many cases singled our exchanges of information as a distinct part of the perse infringement, thereby considering them not merely as a 'facilitating' device but as a separate constitutive element of the 'cartel offence'.225
8.50     A 'cartel offence' in its own right? An important issue, however, is whether the exchange of business sensitive information may constitute an infringement of Article 81 EC on a stand-alone basis. On appeal in Steel Beams, the CFI concluded that, contrary to the Commission's contention during the Court proceedings, the 'information exchange systems' referred to in the decision had been considered by the Commission as separate infringements of Article 65(1) ECSC.226 The CFI subsequently considered that such systems
224   [1999] OJL24/1, paras 98-107.
225  This was the case, for instance, in Cartonboard [1994] OJL243/1, paras 61-64, 105-106 and 134; Seamless SteelTubes [2003] OJL140/1, para 153, third indent; BelgianBeer [2003] OJL200/I, pata 265; Zinc Phosphate [2003] OJ LI53/1, para 215; Methionine [2003] OJ L255/I, para 214; Methylglucamine [2004) OJ L38/18, pata 184; Food Flavour Enhancers [2004] OJ L75/1, para 172; Thread, (para 282) and Industrial Bags (paras 521-523).
«6 CaseT-l4l/94 Thyssen Stahl v Commission [1999] ECR11-347, paras 379-392. In Steel Beams, the Commission had described at length systems of exchange of confidential information, but claimed on appeal before the CFI that it had considered that the disputed information systems did not constitute a separate infringement of Art 65(1) ECSC but formed part of wider infringements consisting, in particular, of price-fixing and market-sharing agreements. The Commission argued that the exchange of information infringed Art 65(1) ECSC insofar as the exchange made it easier fot those other infringements to be committed.
772
B. Typology of Cartel Arrangements and Common Features of Collusion
fe- for the exchange of (confidential) information could indeed be regarded having the restricts rion of competition as their object.227 This was confirmed on appeal by the ECJ.22B It seems I ■ that the Commission has now fully embraced the idea that information exchanges can be }■■■. consideredjWie infringements. In Plasterboard, the Commission regarded the exchange of ;;'.". confidential information as the essential element of the 'cartel offence'. Its finding of a ;;';■. pgf sg infringement of Article 81 EC was based mostly on the existence of several exchanges 'ľ-., of confidential business information. The Commission concluded that there had existed '■': 'between 1992 and 1998 at least in the four major European plasterboard markets a complex, ■:   continuous agreement having as its object the restriction of competition'.229 Interestingly, the r   Commission pointed out that the parties had at certain times knowingly exchanged incorrect information, which confirmed that the exchange of information had had a restrictive ■   object as it was actually used as a monitoring mechanism.230
Conclusion    In answering the question of whether the concerted practice of information    8.51 exchange can be illegal in itself, the first issue is therefore to see whether the exchange of information can be demonstrated to have as its object the restriction of competition.231 If such an object cannot be demonstrated, then other relevant factors may need to be considered.
227 The CFI stated in particular that 'the information which the undertakings received under the arrangements in question was capable of appreciably influencing their conduct, by reason of the fact that each undertaking knew that it was being kept under close surveillance by its competitors and that it could, if necessary, react to the conduct of its competitors, on the basis of considerably more recent and accurate data than those available by other means' and that 'data, indicating the very recent market shares of participants and not publicly available, are by their very nature confidential data, as confirmed by the fact that interested undertakings could receive rhe data distributed by the secretariat only on a reciprocal basis' (para403). The CFI concluded that 'the information exchange systems in question appreciably reduced the decision-making independence of the patticipating producers by substituting practical cooperation between them for the normal risks of competition' (para 406).
"s Case C-194/1999 P Thyssen StahlAGv Commission [2003] ECR 1-10821, paras 59-90.
"» [2005] OJ LI 66/8 (full text of the decision available on DG COMP's web site), para 434. The Commission stated in particular at paras 449 and 450 that 'an information exchange constitutes per se an infringement of Article 61(1} EC if the requirement of independence according to which each trader must determine independently his Conducton the market is undermined as a result. This requirement of independence will without a doubt be affected if the exchange takes place in a highly concentrated market and if it reduces the risk of uncertainty for the trader. These two conditions are manifestly met in the present case inasmuch as the market is an oligopolistic one and the uncertainty has disappeared since the patties agreed [...] to put an end to the aggression reigning on the market, this wish being manifested, moreover, on a number of occasions on the relevant markets. The Commission considers therefore that the system constitutes an infringement of Article 81(1) EC.
230  The Commission stated at para 452 that'[tjhis body of evidence permits the Commission to assert that the [...] agreement concerning the exchange of information [...] was intended to enable participants thetein to monitor the conduct of their competitors at least on the relevant markets and constitutes a manifestation of the common wish of the parties to restrict competition on the plasterboard market in the four major European markets. In so doing, the Commission characterises the agreement as restrictive or competition within the meaning of Article 81(1) EC, being a particular manifestation of the complex, continuous agreement having as its object the restriction of competition on the plasterboard market at least in the four major European markets'.
231   In Steel Beams [1994] OJ LI 16/1, the Commission had described at length, in its decision, systems of exchange of confidential information, but claimed on appeal before the CFI that it had considered that the disputed information systems did not constitute a separate infringement of Art 65(1) ECSC but formed part of wider infringements consisting, in particular, in price-fixing and market-sharing agreements. The Commission argued that the exchange of information thus infringed Art 65(1) ECSC insofar as the exchange made it easier for those other infringements to be committed. See C-194/1999 P Thyssen Stahl AG v Commission [2003] ECRI-10821 para384.
Chapter 8: Cartels
These are, inter alia: the nature of the information exchanged, the level of aggregation and tht age of the data, and the concentration of the industry, The Commission has accepted that depending on an examination of the individual case, exchange of information schemes can be held to be lawful.232 It may indeed be legitimate for companies to pool information about the market in which they operate, as a certain degree of transparency may lead to a more e£H cient functioning of the market. The CFI confirmed the validity of this approach ia John Deere v Commission.133 That said, and as mentioned above, exchanges of information that create an artificial level of transparency and thus a reduction in the uncertainty about com petitors' behaviour in the market are generally considered anti-competitive. For a detailed analysis of the compatibility of exchange of information schemes with EC competition law see Chapter 7 of this book, at paras 7.355 and following.
(f)  The Problem of Tacit Collusion'
8.52     Can economic evidence of the parallel conduct of certain undertakings in the market, in the absence of any proof of collusion, enable the Commission to determine the existence of a 'cartel' prohibited under Article 81 ?234
8.53     The theory of tacit collusion According to economic theory, it is possible for firms in an oligopolistic market to tend towards a monopoly-type price equilibrium in the absence of any kind of explicit collusion. Owing to the interdependence that characrerises firms in an oligopolistic context, where each market operator takes into account the anticipated reaction of its competitors for the purpose of determining its own market strategy, firms may de facto find that they are coordinating their market behaviour without engaging in any form of consultation. What is thus described by economic literature as 'tacit collusion' consists of each undertaking aligning its commercial behaviour with the conduct of the price leader (or of the first mover) by imposing prices increases as soon as the latter does so.
8-54 Does tacit collusion amount to cartel behaviour? The question of whether tacit collusion amounts to cartel behaviour, ie whether it constitutes a per se infringement of Article 81 EC, raises complex issues that still remain somewhat unresolved today. The notion of concerted
23> Steel Beams [1994] 0] LI 16/1. The Commission has however chosen not to address these within the context of the Guidelines on horizontal cooperation agreements (sec Commission Guidelines on theapplica bility of Arc 81 of the ECTreaty to horizontal cooperation agreements, [2001] OJC3/2, para 10). Interestingly, the Commission stated that che Guidelines are 'only concerned with chose types of cooperation which polen dally generate efficiency gains', implying that this is not the case for exchange of information schemes.
J" CaseT-35/92/oAn Deere v Commission [1994] ECR11-957. The CFI found at para 51 chat on a truly competitive market transparency between ccaders is in principle likely to lead to the intensification of competition becween suppliers, since in such a situation, the fe.ee thac a trader takes inco account information made available to him in order to adjust his conduct on the market is not likely, having regard co the atomized nature of che supply, co reduce or remove for che other traders any uncertainty about the foreseeable nature of its competitors' conduct'.
23* This question is of a different nature from the question whether, in the absence of material evidence of an agreement that is restrictive in object or of a concerced practice, evidence of an anticompetitive intention can be deduced from economic evidence of parallel conduce in the market. The laccer question relates to the burden of proof that the Commission must discharge to establish the existence of a common will to restrict competition. As to the former question, it raises the issue whether, in cetcain market conditions, it is possible fot the Commission to conclude that undertakings have engaged in illegal behaviour in respect of cheir ability to jointly restrict competition, by tacitly colluding, ie by aligning their behaviour with their competitors by anticipating and/or following their behaviour.
774
B. Typology of Cartel Arrangements and Com mou Features of Co Huston
ractice as defined by case law does indeed seem very close to the notion of'tacit collusion, u that tne íu'mnient of the corresponding test ('conscious substitution of practical cooperation between undertakings for the risks of competition') does not seem to require much evidence of explicit coordination. However, It seems that a finding of concerted practices nevertheless requires an effective breach of the obligation for a firm to determine its market behaviour independently that must go beyond mere tacit collusion.
Indeed, the Court in Suiker Unie, having reiterated that each economic operator must 8.55 independently determine the policy which it intends to adopt in the market, stated that 'this requirement of independence does not deprive economic operators of the right to adapt themselves intelligently to the existing and anticipated conduct of their competitors'. The Court went on to confirm that 'it does however strictly preclude any direct or indirect contact between such operators, the object or effect whereof is either to influence the conduct on the market of an actual or potential competitor or to disclose to such a competitor the course of conduct which they themselves have decided to adopt or contemplate adopting in the market'.235 The reference to the term contact' (whether indirect or not) seems to rule out the possibility that the notion of concerted practice includes 'tacit collusion.
In its Wood Pulp II judgment, the ECJ gave additional useful indications as to the value of 8.56 economic evidence for the purpose of establishing cartel behaviour. In Wood Pulp,l3B the Commission had found that several producers had engaged in unlawful behaviour which, the Commission submitted, was essentially established by the fact that the firms' quarterly price announcements were near-simultaneous and identical. On appeal, the Court annulled the Commission decision on the grounds that the latter had not established an infringement of Article 81 EC to the required legal standard.237 The Court stated on this occasion that parallel conduct could not be regarded as furnishing proof of concertation unless that collusion constituted the only plausible explanation for such conduct, recalling the right of economic operators to adapt themselves intelligently to the existing and anticipated conduct of their competitors.23a
Conclusion    In the light of the case law discussed above, it seems that tacit collusion can-    8.57 not, as such, be regarded as cartel behaviour. The Community judicature seems ready to accept, albeit under extremely strict conditions (which, it is submitted, are extremely difficult for the Commission to satisfy), that economic evidence may be resorted to in order to establish that prior concertation tookplace between competitors, in cases where that cannot
as Joined Cases 40-48, 50, 54-56, 111 and 113-114/73 Suiker Unie and Others v Commission [1975] ECR 1663, paras 173-174.
a* [198530JL85/1.
237 Since che Commission had no documents which directly established the existence of collusion between the producers concerned, the Court stated thac it was 'necessary to ascertain whether the syscem of quarterly price announcements, the simultaneity ot near-simultaneity of the price announcements and che parallelism of price announcements [...] constitute^] a firm, precise and consistent body of evidence of prior concertation' (Joined cases C-89/85, C-104/85, C-114/85, &116/85, C-117/85 and C-125/85 to C-129/85 A. Abistrom Osakeyhtiö and others v Commission [1993] ECR 1-1307, para 70).
i3a To discharge the burden of proof of cartel conduct, the Commission was required to establish, 'caking account of the natute of the products, the size and thenumber of che undertakings and the volume of the market in question that the alleged parallel conduct of che undertakings could not be explained otherwise than by concertation'. The Court found that the Commission had failed to do this (paras 71 and 72).
775
Chapter 8: Cartels
be established through material evidence of actual
contacts. But one should remain carefoj''''' not to interpret the case law as implying that 'tacit collusion' can be legally qualified asa ca"'^ tel behaviour. The Wood Pulp //judgment addresses the question of the standard of pro c-h applicable to the demonstration of classical cartel behaviour, rather than the possibility of'-'? tacit collusion per se being caught under Article 81 EC. If the Commission is not able tn'; establish to the required standard that contacts between competitors have taken place it -seems that parallel conduct in the market will not be deemed unlawful, as uncertainty-' about the future conduct of each undertaking in the market is preserved. This is so, even if ■ economic theory might suggest that the conditions for tacit cootdinatíon do exist.
(2) Common Features of Collusion 8.58   Analysis of Commission decisions regarding cartels reveals that collusive schemes have many common operational features. Classic patterns of cartels concern their establishment, organisational aspects and the manner in which collusion is conducted on a day-to-day basis. (a) Factors Conducive to the Establishment of Cartels
Cartels will often develop in markets which have a number of characteristics that are favourable to collusion. A specific event affecting the market will generally convince cartel instigators that concerted reaction is needed. In turn they will often convince others to . engage in collusion, thereby acting as a catalyst.
Oligopolistic markets Collusion is made easier in oligopolistic markets. In Italian Flat Glass II, the Commission noted that the flat-glass market was dominated by a tight ofigop-oly.239 Similar features can be found in Alloy Surcharge,2*0 Belgian Beer2A} and Zinc Phosphate, where the five cartel members controlled, in spite of their small size, 90 per cent of the EEA market.*42 In French Beer, a non-aggression pact was made between the two main market players who agreed that neither of them should become dominant.243 Cartels have sometimes been formed after a process of intense concentration has taken place, as in Carbonless Paper2** or Plasterboard245 They may nevertheless involve a relatively high number of players. This is workable in homogeneous product markets, in particular when strongly established trade associations are able to ensure adequate coordination, as in Cartonboard 2i6 or Cement26,1
8.61    Collusive business values and established communication channels    Sector-specific values may result in operators looking upon the control of competition between them in a
8.59
8.60
** [1989] O} 133144, para 12.
240   [1998] OJ LlOO/55. The stainless steel market was described as a highly concentrated market.
241   [2003] OJ L140/1. The cartel was set up by the two main actors in the market, who together controlled over 70% of the entire market.
i« [2003JOJL153/1.
2« [2005] OJ L184/57 (full text of the decision available on DG COMP s web site).
2U [2004] OJ LI I5/I. The Commission observed that the production of carbonless papers in Europe bad become i ncreasingly dominated by a relatively small number of major producers as smaller suppl iers had withd rawn from the market. At the time of the infringement, the cartel members controlled between 85 and 90% of the EEA market for carbonless paper, with the four biggest players controlling around 60% of the market (paras 14-16).
2AS [2005] OJ L166/8 (full text of the decision available on DG COMP's web site), paras 48-51.
2« [1994]OJL243/I.
"> [1994]OJL343/l.
'   t
B. Typology of Cartel Arrangements and Common Features of Collusion
«ftive way- Such values may be diffuse, but have been clearly exposed in certain carrel cases, ich &$ Agreements between manufacturers of glass containers1** or Building and Construction dustry if the Netherlands.249 The Austrian Banks cartel is probably one of the best ill ustra-nris of the influence of specific business values on cartel conduct. In a sector long charac-rised by a very high level of public ownership and considerable oversupply, market players cartel agreements (the 'Lombard club') as a welcome means to combat 'destructive, [t-throat competition, or free competition, as it might simply be termed.350 Established communication channels between competitors may also be conducive to cartel behaviour. This may result from the existence of long established and carefully structured trade associations (Cartonboard,251 Cement)252 from a (legitimate) regulatory framework providing 0pp0rtunities for discussions between competitors (Belgian Beer),2S3 or simply from privi-■;" wed personal contacts between individuals (Fine Art Auction Houses).254 - Exogenous triggering events    Cartels may simply arise from a desire to extract supra   8.62 competitive profits but they are often triggered by an exogenous event resulting in con-. certed response from the industry. The Quinine cartel was formed when the suppliers of this substance (used for producing medicines against Malaria) were faced with a situation of considerable oversupply following the end of the Second "World War.255 In European Sugar Industry,256 the cartel was set up as a concerted attempt to neutralise the effects of the creation of a European common market for sugar. In BNIC, the cartel of cognac producers was triggered by the severe imbalance of supply and demand resulting from the breaking out of the oil crisis in 1973, as 80 per cent of the sales of this luxury good were
«» [1974] OJ L160/1. The 'International Fair Trade Practice Rules Administration' (IFTRA) had established a considerable number of rules to be applied by European manufacturers of glass containers which qualified and prohibited as 'unfair practices' conduct that was really just competitive behaviour.
MS [1992] OJ L92/1. The trade association, SPO, which gathered 28 associations of undertakings representing over 4,000 members of the building sector in the Netherlands, had adopted a 'Code of Honour' and its object was 'to promote and administer orderly competition, to prevent improper conduct in price tendering and to promote the formation of economically-justified prices' (paras 3, 12,47—48).
250 [2004] OJ L56/1. Episodes of more or less unrestricted competition used to be described by the banks as 'hyperactivity' and bank charges were seen in this context, not so much as a factor of competition, but more as 'a joint earnings opportunity', to the detriment of customers. Despite repeated warnings, before the accession of Austria to the EC, that the system would be deemed illegal, banks did not terminate their agreements and the Lombard Club remained in existence, as the banks considered that cartels 'had always been part' of banking and they therefore did not in the least intend to change this now simply because of the applicability of European antitrust law (paras 6,37)-»' [1994JOJL243/1. ™ fl994]OJ 1343/1. "3 Í2003JOJL140/1.
** [2005] OJ L200/92 (full text of the decision available on DG COMP's web site). Such factors explain why recidivism is not rare. Several undertakings have been condemned repeatedly on the basis of the same type offsets. This may also explain the phenomenon of'contamination' from one product market to a neighbouring one. Thus, links can be established between the Graphite Electrodes, Specialty Graphite and Electrical and Mechanical Carbon and Graphite Products cartels, between the Amino Acid and Food Flavour Enhancers cartels, and between the Citric Acid [2002] OJ L239/18 and Sodium Gluconate cartels.
KS [1969] OJL192/5.NewpIantationsofcinchoma had been established in Africa and Asia, resulting in a strong rise in the supply of bark and the situation became even more critical when the US government decided to dispose, at auction, of very large stocks accumulated over the previous years- The producers reacted by agreeing on the price of bark, allocating between themselves purchases of the US stocks, prorecrtng their respective domestic markets and sriaringout others. 256 [I973IOJL140/17.
Chapter 8: Carrels
destined for export."7 In Steel Beams, the cartel resulted partly from a desire to follow uD on the anti-crisis measures initiated and operated by the Commission in the 1980s, which came to an end in 1988.258 In Ferry Operators—-Currency surcharges^ Channel ferry com-panies colluded in response to the devaluation of the pound in September 1992. Similar examples can be found in Alloy Surcharge,250 French Beef 261 and Raw Tobacco Spain.151
The determining role of instigators Particular instigators often play a determining role in the formation of cartels. In Zinc Producer Group, the Commission noted that 'the initiative [. . .] came from the Canadian and Anglo-Australian zinc companies, whilst the Continental European companies joined only after some hesitation 2S3 In PVCII, the cartel originated in a collusive plan formulated at ICI's, and the discussions and consultations which followed.26* In Pre-insulated Pipes, the Commission found that ABB, viewing the price-cutting policies of its competitors as 'irresponsible' in view of the weak state of che market, initiated a series of meetings with competitors in order to find a 'solution'.265 Amino Acidsi6e is also illustrative of typical instigator behaviour. In 1989, ADM decided to start producing lysine when it became aware that two other undertakings were about to set up production facilities in North America. ADM's production facilities for lysine doubled the worlds capacity and it was known that the company had strong financial means and access to cheap raw materials. ADM sent signals to the incumbent producers that, though it intended to be a big player in the lysine market, it would prefer coordination to price war. In order to convince the incumbent producers of the seriousness of its intentions and the penalties of not agreeing, ADM granted its competitors the opportunity to inspect its production plant, and commenced significant sales at low prices. This caused the incumbent producers to drastically lower their prices in an attempt to keep market shares. They finally agreed to cooperate with ADM.
257 [1982] OJ L379/1- The same phenomenon can be observed in Fine Arts Auction Houses, where the cartel was formed at a time when the international art market was undergoing a period of recession, in the context of the economic downturn of the early 1990s (see para 78).
»a [1994]OJLll671,para308.
iss [199710JL26/23.
160 [1998] OJL100/55. The cartel was triggered by a surge in the price of nickel, an important component of stainless steel, at a time when the price for alloys and stainless had fallen sharply (see para 20).
261   [2003] OJL209/12. The defensive carte! was triggered by a sharp fall in the slaughterhouse entry prices that farmers were receiving for cattle, reaching levels that were even lower than during the second outbreak of the 'mad cow' crisis, despite the Community adjustment measures (see paia 184).
262   Decision of 20 October 2002 (full text of the decision available on DG COMP's web site). The agreement between the tobacco processors on maximum prices for the purchase of tobacco leaves resulted from the opening of this activity to competition. Until 1990, one public company had held a legal monopoly over the processing of raw tobacco in Spain and had negotiated the purchase price. In the mid-90s, three new players entered the market, triggering a rise in the average buying price and placing Spanish tobacco at a competitive disadvantage on export markets. The processors began discussions on the setting of a maximum purchase price'to avoid price escalation' (see para 84).
*« [I984]OJL220/27,paraI5.
264 [1994] OJL239/14. Two planning documents amounting to a blueprint for the cartel were found at
ICI, the first proposing a new framework of meetings to administer a revised quota system and price fixing scheme, and the second recording the generally favourable reaction of other producers to the ICI proposals.
2« [1999] OJL24/1, para 30.
2« [2001JOJL152/24.
778
B- TyP0-- <J of Cartel Arrangements and Common Features of Collusion
8.64
(h) Orean^sat^ona^'J^sPects °f cartels
. complex, often multinational, multilateral and secretive schemes, cartels are difficult to gerate and considerable effort is frequendy devoted to ensuring their success. Their intrinsic instability adds to the difficulty of maintaining their functionality. Cartel participants often have to make significant efforts to design the 'modus operandi' of their collu-ion Cartels generally involve frequent meetings and delegation of specific coordinating •oles to their members. In some cases these tasks ate carried out by trade associations or fiduciary companies, which may in in certain circumstances be set up exclusively for the urpOSe. Great efforts are also made to ensure the concealment of the collusive practices, as the participants are aware of the illegality of their behaviour.
(i) Regular Meetings
Akey feature of most cartels is the need for regular contact and meetings to administer the col- 8,65 lusive behaviour. Whilst the development of electronic communication tools might be expected to render physical gatherings less necessary, it nevertheless appears that the complexity of the objectives pursued, the multilateral dimension of collusion and the need for confidence building all mean that physical meetings are still indispensable. Such gatherings are often convened in international hotels, where anonymity can be respected, or in other convenient places such as airports, where company representatives gather for this sole purpose before taking their plane home. Cartel meetings are often scheduled, for opportunistic reasons, before or after official gatherings of industry representatives, such as meetings of trade associations or . international fairs. Of course, such meetings can later provide an apparendy innocent explanation for having been present at a particular location at the same time as competitors.
From informal contacts to highly institutionalised schemes Certain cartels may only 8.66 need sporadic, informal meetings between top executives or sales managers (Belgian Beer,251 Zinc Phosphate).2ۧ However, in view of the technical complexity of implementing restrictive schemes, cartels may require the organisation of a large number of meetings at different levels, from the highest executive circles down to the level of sales managers in local areas. International cartels are often organised along sophisticated lines. The Commission's cartel decisions are full of vivid descriptions of sophisticated cartel structures.269 Although it
»' [2003JOJL140/1.
i«8 [2003JOJL153/1.
JM In Cartonboard [1994] L243/1, the cartel was headed by a 'President Working Group' (PWG) of managing or commercial directors, which met live or six times ayear.The PWG reported to a 'President conference' consis tingofdl producers. Below the PWG, a'Joint Marketing Committee'0MC) at marketing manager level also included nearly all producers. The JMC prepared and briefed the PWG and executed its directives. It met once a month. Besides the JMC, an Economic Committee also included most producers and reported to the President Conference' and the 'JMC on the stare of the market in various countries. In Pre-insulated Pipes [1999] OJ L24/1, the supervisory body was known as the 'directors' club', or the 'Elephant group' and consisted of the chairman and managing directors. As for the 'marketing' or 'contact' groups, they consisted of sales managers and administered the cartel arrangement under the supervision of the directors' club. They were set up in each country and their task was to assign individual projects and coordinate the collusive bidding procedure. In Specialty graphite, decision of 17 December 2002 (full text of the decision available on DG COMP's web site), there were four different levels of meeting: 'Top level meetings', attended by the top executives of the companies; 'International Working Level meetings' (or 'International' meetings), consisting of experts in the senior management; 'Regional (European) meetings and Local (national) meetings which were meant to implement the principles agreed at the International meetings and were attended by local managers.
779
Chapter 8: Cartels
operated within a single Member State, the Austrian Banks cattel provides one of the mot striking examples of the degree of institutionalisation reached by certain cartels. Th Commission stated that 'the agreements were comprehensive as regards their contents highly institutionalised and closely interconnected, and covered the entire country, "down to the smallest village" [...]. For every banking product there was a separate committee on which the competent employee at the second or third level of management sať.270
8.67    Multiple working levels Cartels ate generally characterised by a laddet of several working levels. Top level meetings are usually convened once or twice a year. They act as the cartel 'board' and provide the highest, 'political level' input.27' High level technical meetings normally include top ranking sales executives. They are aimed at administering the cartel arrangements, under the supervision of the top level (Pre-insulated Pipes),272 by translating the broad objectives defined in operational goals and agreeing on the techniques to attain them.273 Regional or local meetings constitute the lowest end of the cartel and ensure their smooth day-to-day running.27A
8.68    'Pre-meerings' and 'ad hoc meetings' are also a frequent feature of cartels. They are generally used to agree in advance upon common positions between undertakings sharing the same kind of interests, or as follow-up to'plenary' meetings, in Polypropylene, the so-called
170 [2004] OJ L56/l.The cartel structure included the Lombard Club, specialist committees, special committees and regional committees. The top-level body (Lombard Club) met every month and was composed of senior representatives of the largest Austrian banks. In addition to matters of general interest that wete clearly neutral from a competition point of view, [they] discussed changes in interest rates, advertising measures, etc. One level down consisted of product-based specialist committees. The most impottant of these were the 'Lending Rates Committees' and the 'Deposit Rates Committees', which, as their names suggest dealt with lending and deposit interest rates and were convened either separately ot jointly. There was a con srant now of information between these committees, in particular with the Lombard Club at the top. Both the Lombard Club and the Vienna Lending and Deposit Rates Committees sent out signals to the diverse and numerous 'regional committees', which held regular meetings in evety province of Austria. In some provinces even the hierarchical sttucture of 'Lombard' and specialist committees was replicated. The views expressed in the ptovinces flowed in the opposite direction back to meetings of the 'Federal Lending and/ot Deposit Rates Committees', in which bank representatives from Vienna met with their opposite numbets from the provinces and whose decisions were in principle valid for the whole of Austria,
771 They ensure adherence to thecollusivescheme(V(/3MiMi) and take major strategic decisions as regards the cartel (Cartonboard). Top meetings are also aimed at ensuring the stability of the cartel by resolving inter nal conflicts {Electrical and Mechanical Carbon andGraphiteProducts).Theydefine the main principles of col laboration for subsequent implementation at lower managerial level {Specially Graphite) and agree on quotas at a global level. They may sometimes go as far as allocating individual market shares in the national markets or agreeing overall price increases {Pre-insulated Pipes). Conversely, 'Summit' meetings may also serve to take notice of agreed price levels and other decisions raken in technical meetings {Electrical and Mechanical Carbon and graphite products).
«7 [1999]OJL24/l.
273   In Specialty Graphite, 'International Working Level meetings' consisted of experts in the senior man agement, who discussed at this level the classification of the products in different categories ot groups and established European minimum prices for each group. In Vitamins, 'Marketing directors' meetings took place at three different levels, from the worldwide level to the regional level. Their role was to take operational deci sions regarding implementation of the agreement, to monitor sales, to ensure respect of the quotas and to implement price increases.
274   In Specialty Graphite 'Local meetings' were meant to implement the principles agreed at international technical meetings and wete attended by local managers. In Electrical and Mechanical Carbon and Graphite Products [2004] OJ L125/45 (full text of the decision available on DG COMP's web site), local meetings were organised on an 'ad hoc' basis in each country and dealt with the implementation of the cartel decisions in each country and discussed local customers* accounts (see paras 74-77).
B. Typology of Cartel Arrangements and Common Features of Collusion
hie four' met in restricted session the day before each 'bosses' meeting'. These so-called meetings' provided a forum in which the four major producers could agree a position etween themselves prior to the full meeting. The idea was that the four major producers, „ith some 50 per cent of the market in their hands, could, by adopting a united approach, ncourage moves towards price stability.275 Similar practices have been identified in •gamlessSteelTubes,276 Graphite Electrodes277 and Citric Acid.276 Commission decisions also .„hjight numerous examples of bilateral meetings aimed at sorting out specific problems egarding, for instance, compliance with agreed quotas or the mutual allocation of specific
customers. Bilateral meetings are also often used to prepare for negotiations in multilateral
meetings and to secure the support of 'allied' undertakings.379
fit) The Role of Trade Associations and Fiduciary Companies
Many Commission decisions show that trade associations or fiduciary companies play a   8.69 central role in the establishment and, in certain cases, the management of cartels. In nearly all cases, the activities of a trade association or of a fiduciary organisation processing market data on behalf of competitors in a given market are part of the general context in which the collusive trade associations or fiduciary companies operate.
Liability of trade associations Cartel decisions adopted in the eatlyyears of enforcement 8.70 reveal that, at that time, trade associations were often directly involved in anticompetitive activities, if not created exclusively for anticompetitive purposes. However, the Commission did not always work on the basis that the associations themselves should be held personally responsible for the infringements found, as illustrated in Vegetable Parchment,™ Cast Iron and Steel Rolls28'' or Zinc Producer Group}*1 It soon became the
3» [1986]OJL230/l,para68.
376 [2003] OJ L140/I. European producers used to coordinate their positions during preparatory meetings, before taking part in the general meerings (see para 59).
W  [2002] OJLlOO/1, 'European meetings'were convened in the absence of the Japanese cartel members.
378 [2002] OJ L239/18. Representatives from the two US underralcings met on at least on 10 separare occasions to prepare posirions prior to multilateral meetings (para 66).
279   In Food'FlavourEnhancers [2004] OJ L95/1, the Commission noted that 'bilateral' meetings were also used to influence the outcomr of the 'general' competitors meetings. For example, two undertakings met the day before the multilateral meeting because one of them wanted to secure the support of the other in relation to a price increase to be applied to a customer (para 135)-
280   [1978] Oj L70/54. Nearly all cartel participants were members of a trade associarion named the 'Genuine Vegetable Parchment Association' (GVPA), which played a central role in the cartel. The GVPAsec-retariat collected all invoices for export sales issued by its members and established statistical tables setting out the tonnages exported by each producer. It also received the price schedules established by manufacturers during the genera! assembly, such as the rates of increase and the dares on which the new prices were to become applicable and sent them to the other members. However, rhe Commission addressed (he prohibition decision only to the GVPA members {paras 36-40).
381 [1983] OJL317/1. The creation of the'International Rolls Manufacturers Association' (IRMA) was purely instrumental in the implementation of the cartel agreements. Carte! members agreed roset up a more elaborate organisation for rhe purpose of coordinating the notification of individual enquiries and the fixing of quotation prices. IRMA was at the core of the cartel and all strategic decisions were taken within the framework of the association's meetings. However, the Commission did not include IRMA among the addressees of the prohibition decisions, only its members.
287 (1984] OJ L220/27. The establishment of the cartel could not be dissociated from the creation of the 'Zinc Producer Group' (ZPG), an institutionalised organisation which tan the cartel in practice for 13 years. It is worth noting that the Commission did not address the prohibition decision to ZPG, but only to undertakings which were members of the ZPG and active in the Community. The Commission stated that '[t]he
ChapterS: Can. ,
Commission's practice, however, to hold trade associations directly responsible for cartel infringements alongside its members. In Belgian 'Wallpaper, the trade association, whose 'internal regulation was held to be a factor in the infringement, was an addressee of the decision although no fine was imposed on it.283 The same approach was followed m Fedetab 2B4 and in Italian Flat Glass I.2as In Steel Beams, trade association Eurofer had facil-itated the implementation of infringements of Article 65 ECSC by its members by organ-ising an exchange of some of the necessary information. The Commission stated that although Article 05(5) provided only for fines against undertakings, an infringement com mitted by an association would expose the undertakings which belonged to it to the risk of a fine.286 With BNIC,2S7 the Commission started to impose fines on trade associations themselves. Trade association BNIC,ZBB which played a central role in the cartel, was deemed to be an association of undertakings and was held directly responsible for the infringement. It was the addressee of the prohibition decision and the fine was imposed on it (and not directly on its members). In Roofing Felt, the Commission found that trade asso ciation 'Belasco'was not involved in all aspects of the cartel's operation, but in one of its most serious aspects, namely the system of quotas. The Commission found that although Belascos members were also members of the cartel, the association itself had to be held responsible, independendy of its members, for its involvement in operating the cartel m A fine was imposed on Belasco and the Commission decision was upheld on appeal!90 As a further example, the Cement case provides a particularly clear illustration of how directly trade associations can be involved in cartels.291
ZPG must be seen as a worldwide cartel whose centre of gravity was always outside the European Community As well as the firms with which this Decision is concerned, a large number of other firms were involved in it which did not, arid do not, have anyregistered office or subsidiary in the Community and did not operate to any significant extent or only sporadically in the Community. For the ZPG, which controlled about 80% of the world zinc market, the European Community was only one of several regional markets [...]. The initiative to form the ZPG and the driving force which carried it along came from n on-Community firms. A report [. . .] written [. . .] on the foundation of the ZPG [. . .] speaks of a "diktat" by the British, Australian and Canadian companies. However, the situation of the zinc market was such that, even in the absence of the restrictive agreements, these third-country companies would not have made any significant contribution to improving competitive conditions in the Community, so that thcit participation in the ZPG had only indirect effects on the Community'. It was therefore appropriate to consider Art 81 EC as applying only to the undertakings which were indirectly involved in the manipulation of competition within the Community.
«a [1974J OJL237/3.
»« [1978JOJL224/29.
J8S [1981] OJ L326732.The Commission found that three associations of Italian glass producers had been directly involved in the cartel as their internal tegulations, which were binding on their members, were at the heart of the restrictive arrangements. These associations weie addressees of the Commissions prohibition decisions, together with their members, but no fine was imposed in recognition of the short duration of the infringement.
2B6 [1994] OjLllß/1 para 317. No additional fines were imposed however, since individual undertakings
were already being fined.
=" [1982]OJL379/l.
288 'Bureau National Interprofessionnel du Cognac'.
«s [1986] OJL232/15,parasll4, 115-
290 Case 246/86 Belasco and others v Commission [1989] ECR 2117.
i9i [1994] OJ L343/1. In addition to fining individual undertakings, the prohibition decisions imposed heavy fines on the European Cement Association (Cembureau) and eight national cement associations. Although officially entrusted with fully legitimate goals, Cembureau was at the cofe of the cartel agreements. Through the collection and dispatch of information and the organisation of regular meetings between 'head
782
B, Tyt. Jgy of Cartel Arrangements and Common Features of Collusion
ver time, cartels have drawn lessons from the Commissions practice and trade associa- 8.71 ns are today less overtly involved in anti-competitive activities. However, recent decisions show that cartels continue to originate in the contacts developed within the context of trade associations and that trade associations continue to be used as legitimate covers for illegal activities, as shown in Amino acids,292 Citric Acid293 or Carbon/esspaper,294 In Industrial Tubes, the Commission noted that towards the late 1980s, the producers, organised within the trade association CuprocUma, extended the scope of their cooperation to competition issues. CuprocUma meetings were held twice a year, and provided a regular opportunity to discuss and fix prices and other commercial conditions for industtiaJ tubes once the official agenda of the meetings had been concluded.295
The role played by certain fiduciary companies in facilitating collusion    The collection and    8.72
processing of (individual) market data to obtain a reliable picture of the size and evolution of the market is an entirely legitimate activity provided that it does not lead to the disclosure of individualised figures to competitors (see Chapter 7 of this book, paras 7.355 and following). However, undertakings may be tempted to go beyond what is allowed, exchanging individual figures in order to set up and monitor collusion. In certain cases, cartel members organise the collection and exchange of data themselves.296 However, the Commission's decisions over the years show that in other cases the cartel delegates this activity to a fiduciary company or a notary's office. In such cases, the fiduciary often carries out an en tirely legitimate activity, whereas the cartel participants exchange individual figures between themselves.297
delegates' [the representatives of the member undertakings), Cembureau organised the collusion on prices and established what was known as the 'Cembureau agreement or principle of non transhipping to internal European markets', in the implementation of which it actively participated. In addition to the heavy fines imposed on its members, Cembureau itself incurred a fine of EUR 100,000. This fine was however subse-quendy annulled by the CFI.
131 [2001] OJL152/24. In order to avoid arousing suspicion over the meetings they attended, the cartel members created an 'amino acid working group' of the wholly legitimate European Feed Additives Association, whose sole purpose was to provides 'perfect cover' for their price-fixing meetings. See JM Griffin, An inside look at a cartel at work: Common characteristics of international cartels', in Konkutrentsverket/ Swedish Competition Authority (ed), Fighting Cartels—why and how? (2001) p 43,
39J [2002] OJ L239/I8. Cartel members used the cover of the legitimate meetings of the Brussels-based trade association ECAMA to hold their meetings, which would typically take place the evening prior to the official ECAMA meeting (see para 87).
294   [2004] OJ LI 15/1. General cartel meetings were convened under cover of the official meeting of the trade association AEMCP. Cartel discussions first took place during the trade association meetings, but it was subsequendy decided that official and collusive activities should be separate. Cartel meetings were then held before or after official meetings (see paras 83-86).
295   The Commission noted that '[djiscussions concerning price, customers, individual sales volumes and market shares mostly took place on the second day of the Cuprocf ima meeting session after the official agenda had been discussed' (paras 78 and 82).
296   See zgMethylglucamine [2004] OJ L38/18, paras 76-82), Choline Chlorids [2005] OJ L190/22 (full text of the decision available on DG COM's web site), para 92.
297   In Peroxygen Products [1985] OJ L35/1, the three cartel participants which supplied the French marker exchanged detailed information on their respective production and sales. This was effected every month through a body known as the 'Chambre Syndicate de i'Eau Oxygenic et des Persels'. This body collected statistics from each of the producers and issued them with composite figures for the whole French market. In addition to the monthly statistics issued by the 'Chambre Syndicate', the producers exchanged further and more detailed information on matters not covered by the 'official' exchange. The data available through the 'Chambre Syndicate' did not include details of sales to individual customers. However, the individual producers exchanged the data with each other (see para 30). See also, eg Sodium Gluconate znd Zinc Phosphate [2003] OJU53/1.
783
Chapter 8: Cartels
However, there are a number of examples where the fiduciary companies actively take parf in the illegal exchange of detailed competitor data. In Italian Cast Glass, a fiduciary com pany was entrusted by the cartel members with the task of supervising the proper implemen. tation of the cartel agreement at issue."8 In Cast Iron and Steel Rolls, the cartel agreement contained an obligation to pass on all customer enquiries immediately to a neutral office in Zurich (Switzerland). It was that office's responsibility to notify all interested parties of the other producers who had received the enquiry. The CQmperitors would then contact each other to establish suitable prices.299 In Cartonboard, the cartel also used the services of a fiduciary company located in Zurich. This company provided the secretariat for the Paperboard product board and was present at all cartel meetings.300
8.73 Towards a systematic sanctioning of fiduciary companies' direct involvement For a long time, the Commission had tended not to hold fiduciary companies responsible for the cartel infringement to which they contributed. However, in Italian Cast Glass, evidence of the fiduciary involvement was so strong that the Commission considered that the fiduciary company had been a party to the anti-competitive agreement and addressed the decision to it.30' No fines were imposed, however. Nevertheless, it appears that the Commission's attitude towards cartel facilitators is no longer so lenient. In Organic Peroxides, the Commission imposed a symbolic fine on a Zurich-based fiduciary for its blatant involvement in the implementation of cartel activity, thus signalling that such behaviour is at risk of being severely punished in the future.302
298 [1980] OJ L383/19. The fiduciary company had to draw up monthly summaries of shipments, con raining information on average prices for each product and each undertaking, and to carry out on their prem ises verifications of the information provided, induding verification of the quantities dispatched and of their consistency with the store accounts and with the general accounts. If it deemed it appropriate, the fiduciary could request addirional information and even carry out random checks on the physical quantities compated with the accounting quantities indicated or appearing in the store accounts.
las [1983] OJ L317/1- The Commission decision contains considerable evidence of the office's direct involvement in the cartel, which took detailed minutes of the cartel arrangements made at the cartel meetings and implemented concealment measures (see paras 10,31),
300 [1994] OJ L243/1. The Commission noted with scepticism, however, that the fiduciary's claims that it had not kept any record of the cartel working groups, contrary to what its practice had been for the other committees. In that case, the fiduciary was entrusted with the task of running a very sophisticated statistical system which was used to assist in the determination and monitoring of individual market shares and in the analysis of capacity utilisation.
3°" [1980] OJ L383/19. The Commission stated that '[i]n assessing whether the agreements concluded between [the cat tel members] on the one hand and Fides on rhe other were liable to restrict competition on the market in cast glass, despite the fact that the first three companies are manufacturing undertakings and Fides a service undertaking, it must be borne in mind that Fides enabled and consciously assisted the implementation of the restrictions of competition which were the very purpose of the agreements, and consequently it is jointly responsible for the resulting restrictive effects' (see point II A4 of the decision).
302 (2003] OJ L6/1 (full text of the decision available on DG COMP's web site). See also IP/03/1700. In this case, the Commission found, inter alia, that the fiduciary had stored the initial cattel agreement in a safe; organised the meetings of the cartel members; taken charge of tcimbutsing the travel expenses of the participants in otdet to avoid traces of the cartel meetings in the undertaking's accounts; produced, distributed and collected the so-called 'pink' and 'red' papers with the agreed market shates which were, because of their colours, easily distinguishable from other meeting documents and which were not allowed to be taken outside the fiduciary's premises; acted as a moderator in case of tensions within the cartel; and even instructed all participants on the legal dangers of the meetings and of measures to take to avoid detection (see patas 91-105).
784
B. Typology of Cartel Arrangements and Common Features of Collusion
t) Concealment Measures
view of the blatant illegality of their behaviour and the severe penalties incurred if they are    8.74 icovered, cartels generally deploy considerable efforts to avoid detection. Concealment asures of all types can be found in cartel decisions.
„jvering the cartel's tracks A basic security measure is to ensure that cartel meetings, and the 8.75 matters discussed at them, remain secret. To that end, many cartel meetings have been convened in hotels in large international cities, and reservations for meeting rooms and accommodation made under false names. In Amino Adds,303 the cartel members took care to stagger their arrival and departure times for the cartel meetings so as not to arouse suspicion by having the entire group enter and leave the room at the same time. In Food Flavour Enhancers, cartel meetings were sometimes limited to just a few undertakings who acted on behalf of the other competitors in order to limit the risk of detection.304 In many cases, cartel meetings have been organised just before or after official meetings of a legitimate ttade association meeting, or during a trade fair.
Paperless policy Cartel members also often endeavour to refrain from taking written 8.76 notes during cartel meetings or from taking such notes away from the meeting room, or they ensure subsequent destruction of any minutes or notes. Jn Cartonboard, the Commission noted that there were, surprisingly, no notes of certain working groups whilst other, more innocent meetings were documented in great detail305 There are many examples of instructions to destroy all notes after the end of the meeting.306 In SAS/Maersk, several measures were taken to avoid keeping a full written record of the points on which the cartel members had agreed.307 In Industrial Tubes, the unofficial meetings were conducted without documentary support, implying that normally no minutes or agenda were drafted. The Commission found evidence of a cartel document specifying the security rules as 'no paper, no document, only with diskette'. In Organic Peroxides, a fiduciary company produced and distributed data on agreed market shares in the form of'pink' and 'red' papers which were easdy distinguishable from other papers and were collected up at the end of the meetings.308
Storing cartel documents in private homes    The use of private homes as places of storage   8.77 of cartel documents, or as safe' mailboxes for the exchange of sensitive information, is also
'»a [2001JOJL152/24.
304 [2004] OJ L75/Í ■ The Commission reports that in one high level meeting, a Japanese company represented all other Japanese members. This company stated that 'it would look suspicious if the Japanese producers all went to a Korean resort together'. In Organic Peroxides [2005] OJ Ll 10/44 (full text of the decision available on DG COMP's web site), a Swiss-based fiduciary company reimbursed travel expenses to avoid leaving traces in the cartel members accounts and in Graphite Electrode: (2002] OJ L100/1, expenses for meetings were paid in cash with no explicit reference to those meetings.
3<" [1994JOJL243/I.
306  See, for example, Graphite Electrodes [2002] OJ LlOO/1. In Copper Plumbing Tubes (decision of 3 September 2004, full text of the decision available on DG COMP's web site), the Commission noted that if any-onesent notes to one of the cartel participants, this person was reminded by phone to destroy the paper (see para 129).
307  Thus, in SAS/Maersk [2001] OJ L265/15, the Commission uncovered a cartel document reading: '[t)he parts of the documents that infringe Article [81 (1)], although they cannot be agreed upon and cannot be put on paper, we presume that because some people will not be present (in the future), these parts will have to be written anyway and be put in escrow in the offices of lawyers from both sides' (see para 89). All material on price agreements, market-sharing agreements and the like had to be destroyed after the meetings and any controversial material on PCs had to be deleted.
™ [2003] OJ L6/I (full text of the decision available on DG COMP's web site),
Chapter 8: Cartels
a classic feature of hard core cartels. In SAS/Maersk, the Commission found the minutes a an interna! meeting of Maersk reporting a statement from a company's representative tha all material on price agreements, market-sharing agreements and the like had to be destroyed before going home today. Anything that might be needed, had to be taken home'.309 In Graphite Electrodes, for the specific purpose of keeping these contacts secret a sales manager had a special dedicated telephone installed at his home with a fax machine 3<o In Vitamins, the cartel members made great efforts to keep track of incriminating docu ments and destroy them, including the conduct of international audits to verify that such documents no longer existed.31'
8.78     Using code names The use of code names to designate cartel members is also a frequent feature of cartel documents. For example, in Cast Iron and Steel Rolls, the Commission notes that the cattel decided, as a precaution in case of cartel investigations, to use acode in place of the abbreviations of members' names previously used by the fiduciary company. In Graphite Electrodes, a sophisticated system of code names was used to conceal the real iden tity of both undertakings and individuals.3 n In Belgian Beer, the cartel agreements wete referred to by cryptic names such as 'project Green or 'Universite de Lille'.313 In Industrial Tubes, a coding system was used to hide the identity of the producers in the cartel documents and in spreadsheets concerning target prices.
8.79     Other measures may also be taken to hide specific cartel implementation measures, In Food Flavour Enhancers, the so-called counter-purchases' between competitors, which were part of the collusive scheme, were made under sophisticated covers: a cartel member purchased the product of another cartel member through the distributor of a third one.314 Emergency measures may also be taken in view of a growing danger of discovery. In Graphite Electrodes, one of the cartel leaders warned all other cartel participants that on-the-spot investigations from the Commission could be expected. With that in mind, all relevant files were reviewed and incriminating documents were destroyed or kept in safe places, such as private homes.315 In Vitamins, as the result of a recent criminal investigation in the US involving ADM, the cat tel members agreed on a principle of'complete security'. Direct contact with the subsidiaries in the US would be suspended and contact would be made with the headquarters.316
ms [2001] OJ L265/15, para 89- In Electrical and Mechanical Carbon and Graphite Product! [2004] OJ L125/45 (full text of che decision available on DG COMP's web site), there were instructions not to keep any cartel document in the undertakings' premises or even at home but to destroy them after use. In the first years of the cartel, communication was conducted through 'mailboxes' and there were instructions to use the telephone rather than raxes.
310 [2002} OJ LlOO/1, In the same cartel decision, the Commission notes that telephone conversations regarding prices were made from a mobile phone registered in Switzerland (see para 59).
3,1 [2003] OJ L6/I. Certain documents that had to be kept were copied onto computer disks and hidden in the eaves of one employee's grandmothers house. See 2003 OECD Report on Hard Core Cartels.
3,3 [2002] OJ LÍ 00/1. SGL was named 'BMW', UCARgot the name of 'Pinoť and the group of Japanese companies were dubbed 'Cold', a derivation from the first letters of their individual code names 'Chivas Ocean,'Lawn' and'Dr/.VAW Carbon was known as'Wave'. Individuals were also given code names, such as Artemis', 'Moustache' and 'Taurus' (see para 59)-
™ [2003JOJH40/1.
"* [2004]OJL75/l,para83-
™ [2002]OJL100/I,para33.
äi« {2003]OJL6Yl,seepara452.
786
B. Typix-jJ of Cartel Arrangements and Common Features of Collusion
Metrical and Mechanical Carbon and Graphite Products a 'security committee' was even i ted in í 998, after the opening in Europe of an investigation into other activities of the jartel members.3'7
,i fanning the Business of Collusion
Resides purely organisational aspects, it is worth describing the common features of rhe man- 8.80 auernerir of collusive activities. The elimination or reduction of competition can be a complex task requiring a particular form of management. In spite of their joint intention to collaborate, a degree of rivalry and distrust will always remain between cartel members. The undercurrent of distrust in cartels means that they require a certain degree of leadership and coordination. Devising sophisticated cartel management tools is often necessary. In spite of such measures, cartels nevertheless appear to be characterised by an inherent instability.
(i) Managing the Elimination ofCompetition "$          Gett'ng°tner markst operators on board    Efforts are often made by large cartel members    8.81
* to convince unruly players to join the cartel. In Vitamins, the Commission noted several examples of efforts used by the ringleaders to convince other participants to enter into the restrictive agreements. In the Vitamin B5 cartel, a representative of Roche travelled to Tokyo to convince Daiichi to limit its exports to other markets as this was pushing down the price. Since Daiichi doubted that the cartel would succeed without BASF, Roche promised to invite BASF to a trilateral meeting convened to promote a collusive scheme regard-. ing vitamin B5.31B In Industrial Tubes, the Commission quoted a cartel participant as stating that small participants were invited to join cartel meetings 'to have better control, because with a fairly small market share these companies still had a big impact on prices'.319
Achieving comparability of the products An acceptable degree of comparability of the prod- 8.82 ucts sold by each member appears to be relevant to most cartels. Significant efforts may be dedicated to this. In Sodium Gluconate, discussions concerned the proper way to define the product ■ and the opportunity to include in the calculation of market shares a partial substitute, as certain producers of that product were suspected of free riding on the other cartel members.320 In Specialty Graphite, in order to be able to fix prices according to equivalent categories of products, the parties established a complex 'Product Grouping Standard'. This classification of grades was devised in accordance with the product applications.321 In Electrical and Mechanical Carbon and Graphite Products, cartel participants had devised a highly sophisticated method for the calculation of the price of the products with refeience to a number of objective factors such as the price of raw materials, the size of the product or the number of components it included.3a
3" [2004] OJ L125/45 (full text of the decision available on DG COMP's web site), paras 81-86.
3la  [2003] OJL6/1, para 297.
3,9 [2004] OJLI25/50, para 89.
«° Paras 111, 135.
3*1 Decision of 17 December 2002 (full text of the decision available on D G COMP's web sice), para 99.
"2 [2004] OJ L125/45 (full text of the decision available on DG COMP's web site).This formula, called the 'barrme', was meant to enable each cartel participant to calculate the price of its products in a way that guaranteed a perfect uniformity of the prices notwithstanding the differences in the specifications of the product. This effort towards a total harmonisation of the prices was the subject ofmany technical meetings of the cartel (see paras 91-97).
787
Chapter 8: Cartels
8.83     Collection of data and circulation of detailed implementation documents The day-to, day management of cartels often involves a sophisticated exchange of important quantiti-, of data such as sales volumes, price lists, tables of market shares or annual 'budgets', either for planning fresh agreements or monitoring existing ones. Cartels often rely on the leem. mate statistical activities of a trade association to obtain the basic data on which they rely to exchange further information and to make their decisions {Sodium Gluconate, Zinc Phosphate)?11 However, the collection and processing of data may be done by the cattel members themselves, with a view to saving costs and/or minimising the risks of detection.324 Such cartel documents may be very sophisticated, as illustrated by numerous cattel decisions.525 In Vitamins-, extremely detailed documents were maintained and circulated All three major European producers supplied to the Commission tables and spreadsheets created and used for the purposes of calculating, reviewing and agreeing the sales quotas of vitamins A and E for each regional and national market.326
8.84     'Coordinators1 and 'market leaders' Cartels often show a high degree of coordination and contact within the network. Specific coordination measures may also be taken. In Pre insulated Pipes, a retired business executive with close personal connections to the ringleader of the cartel was engaged as a consultant to act as the 'coordinator' of the cartel.327 A contact group' also played a fundamental role in the coordination of the cattel. The Commission noted that this complex mechanism enabled the German contact group to monitor some 1,400 to 1,500 projects per year which were over a DEM 50,000 threshold'.328
8.85     The appointment of'market leaders' of particular regional or national markets often serves to ensure the coordination of cartel measures, such as collective price initiatives. The mar ket leaders are supposed to take a 'first mover role in any price movement in the markets for which they have been designated, In Vegetable Parchment™ the cartel members fixed prices for 'free export markets', that is countries in which no member manufacturer was established. Once the cartel members had fixed the date and rate of the price increase 'price leaders' in each market sent to the trade association secretariat, for circulation to the other members, the applicable price schedules.310 In Graphite Electrodes, the 'home producer' was
"3 [2003JOJL153/1.
3M In Sodium Gluconate, the cartel participants had ceased to resort to a Swiss fiduciary but resumed discussions on this point as mutual suspicion of cheating was growing (paras 119, 195, 196). In Citric Acid 12002] OJL239/J8. each participant reported the tonnage sold in each region to the secretariat of the (rotating) president of the trade association, which assembled the data and reported them back to the members by telephone, broken down by firm and b/ region (para 100).
325  See eg, in Citric Aciď [2002] OJ L239/18, the description of certain cartel documents at para 101. Very detailed market share and price increase tables were circulated in Graphite Electrodes [2002] OJL100/l,and Carbonless Paper [2004] 0]LU5n.
326   [2003] OJ L6/1. The Commission described these documents as consisting of worksheets or support documents used to fix the annual 'budget' for each producer ort a country-by-country basis and charts comparing the actual sales of each producer with their respective 'budgeted volumes', ie their quota for each regional and national market both on an annual basis and for interim periods (see paras 190-191).
327   [1999]QJL24/l,pata33. «8 Para 72.
»» [1978]OJL70/54. 33° Para 40.
B. Typology of Cartel Arrangements and Common Features of Collusion
„ de responsible for fixing the price on its domestic territory, and the other cartel members re expected to follow its price initiatives.331 Similar practices are described in Sodium tconate*3,2 and Industrial Tubes.333
(it) Enforcing Cartel Arrangements between the Participants
As regards the risk of'free riding' inherent in any collusive schemes, the enforcement of the   8.86 restrictive arrangements by the members is generally a major source of concern for companies involved in cartels. Monitoring schemes aimed at ensuring compliance are therefore frequently set up, and deviations from the agreements ate often subject to peer pressure or even penalties.
Monitoring schemes The monitoring of compliance with the cartel agreements may simply 8.87 take the form of an informal agreement between cartel members to provide each other with evidence of the implementation of the measures taken.334 However, in many cases, the monitoring takes a more structured form. Compliance is often verified by cross-checking the individual data supplied by cartel members with global market data processed by a coordinator.335 The monitoring of compliance may even take the form of more intrusive methods and may lead to coercive measures. In Preserved Mushrooms, cartel members had agreed that on request, each party would allow the other to check sales, production and dispatch documents; a joint accounting organisation would also be appointed to collect such datafbr the common information of the parties.336 In Italian Cast Glass,337 a fiduciary company
sal [2002] OJL100/1. One undertaking was appointed as the market leader for the US and certain parts of Europe, whilst another was responsible for the rest of Europe. Japanese producers were market leaders in the Far East (see para 50).
3äí A market leader was appointed in each country and was in charge of setting the target price. The market leader was chosen on the basis of its thorough knowledge of the country and of its influence in the said country (see para 89).
"3 [2004] OJ L.125/50. 'Market leaders' were appointed, who were responsible for certain Member States. The market leader was the member of the cartel with the highest sales in a given country. Market leaders were in charge of monitoring customer visits, of gathering information in their respective territory and deciding upon target price changes. A cartel participant stated that 'the mission of the cartel leader [was] to protect the interests of each member as agreed'. He had to be informed of all visits to customers and to be immediately informed of the negotiations. 'Only the market leader [could] change the targets if necessary and [bad to] Inform immediately all the companies involved. [...] In case of disagreement between a member and a market leader, the market leader would take the final decision (see para í 08).
334 In Carbonless Paper, the monitoring of compliance with the price-fixing agreements was conducted through the exchange of the standard letters announcing price increases to customers. The Commission found in the premises of one undertaking a bundle of standard letters from a competitor to its customers, together with a card sent 'with the compliments of another competitor (paras 102-103). In IndustrialTubes, compliance was monitored through exchanges of detailed information between members on deliveries, market shares, customers and prices, primarily within the framework of the cartel meetings, as well as by fax, e-mail and telephone.
J3S In Specialty Graphite {decision of 17 December 2002, full text of the decision available on DG COMP's web site), the frequent exchanges of shipment records among competitors allowed a detailed monitoring of sales and the detection of possible deviations from the cartel instructions. A trade association or a fiduciary company may be entrusted with the task of collecting and aggregating individual sales data. Monitoring may also be carried out by a cartel member. In Sodium Gluconate, sales monitoring was initially conducted on the basisofthestatisticssupplied by aSwiss fiduciary, but to save cosrs, one under taking wasput in chatgeof collecting sales data afew days before the multilateral cartel meetings instead (paras 92—93). Similar schemes were iden-tifiedin CitricAcid[2002] OJ L239/I8 (paras 100-101) And Amino Acids [200\] OJ L152/24 (paras 224-227).
»« [l975]OJL29/26,para4(d).
3" [1980JOJL383/19.
Chapter 8: Caru
was entrusted by the cartel members with the task of drawing up monthly summaries of ship, ments and carrying out, in the premises of the undertaking, verification of the information provided. The fiduciary couid even carry out random checks on the physical quantities com. pared with the accounting quantities indicated or appearing in the store accounts. In Roofing Felt, an accountant was appointed and was in charge of monitoring compliance with quotas and prices.338 In French Beef, the Commission found that three farmers' federations had used physical force to set up mechanisms to verify that the agreement was being applied, such as illegal 'inspections' to establish the place of origin of meat.339
Penalties and compensation measures Cartels frequently agree on compensation measures or even penalties to be applied to members who have not complied with the restrictive agreements. In Roofing Felt, a member in breach of its obligations could be ordered to pay a standard penalty into a common fund, failing which the sum could be deducted from the security the member had lodged in the guarantee fund.3"0 In Pre-insulated Pipes, the performance of each producer against quota was ascertained and discrepancies would be regularised either in the allocation of the following year's quotas or by payment of compensation.3'" In Citric acid,3*2 a compensation scheme was agreed as a corollary to the quota agreements, in order to penalise those companies selling above their assigned sales quotas and at the same time compensate those that did not reach theirs.
Threats, boycotts and coordinated attacks on competitors In addition to the enforcement of compliance schemes Voluntarily' agreed upon by the members, it is not uncommon for cartels to take steps aimed at forcing reluctant market players to align their commercial behaviour, so that the anticipated results of the cartel are not undermined In Roofing Felt, the cartel members took or planned concerted action against other manufac turers and importers to discourage them from pursuing a price-cutting policy, and to take customers away from them..3"3 In Electrical and Mechanical Carbon and Graphite Products, cartel members exchanged information on their competitors and engaged in concerted actions in order to persuade or force them to cooperate.344 In Carbonless Paper, threats were
338 [1986] OJ L232/15. To facilitate monitoring of quotas, members were required to make monthly returns to the accountant stating their purchases and movements of stocks of raw materials and finish«! products and their exports. They also had to keep numbered invoices on both sales and purchases and to keep full accounts, with supporting documents, for tegular or extraordinary inspection. The accountant also administered the arrangements for penalising those who exceeded their quotas and acted as secretary to the general meetings (sec para 20).
33* [2003] OJL209/12,paras30,38, 80,126, 133, 173.
3« [1986]OJL232/15,para21.
*» [1999] OJL24/1, para 65.
341 [2002] O J 1.239/18. It had been decided that if a company exceeded its assigned quota in any one year, it would be obliged to purchase product from the company or companies with sales below their quotaduring the following year. According to the Commission, this led to several important transactions between companies, especially as one company tended to fall shott of its quota, whilst another one tended to remain ahead of it (seeparas 88 and 102-111).
343 [1986] OJL232/15, para 59.
*" [2004] OJ L125/45 (full text of the decision available on DG COMP's web site). On several occasions small local producers were persuaded to take part iniocai cartel meetings. In other instances, competitors were forced to withdraw from the market, or were even taken over by the cartel members, who subsequently ensured that their subsidiary complied with the cartel rules (see para 167).
790
B- TyA      ry of Cartel Arrangements and Common Features of Collusion
made against recalcitrant competitors. A representative of the ringleader stated that he would 'personally take care' of those who would not play by the rules. The company in Question, which enjoyed market power, threatened to destroy any competitor that adopted competitive attitude. It was reported to have executed this threat with regard to an Italian competitor.345 Measures to 'convince' companies to cooperate may even take the form of (threats to take) initiatives aimed at having their behaviour sanctioned by public authorities. In French-West African Shipowners' Committees, the committees deployed efforts to ensure that the African public authorities adopted rules protecting their activities and providing for penalties on members exceeding their quotas and on independent lines operating without the approval of the committees.346 In Amino Acids,3*7 and Citric Acid, recalcitrant third-country producers were threatened with anti-dumping actions.348 Targeted concerted actions against competitors may also be taken. In Pre-insulated Pipes?™ considerable efforts were devoted to taking a small but active competitor out of the market. In Citric Acid, the increasing availability of Chinese production in the European market and the need for a more forceful stance by the cartel members to maintain their level of sales were the subject of discussion at the cartel meetings. A price war against the Chinese competition was decided upon, at whatever price was necessary.350
(iii) The Instability of Collusive Schemes
Cartels are unstable by nature. They are subject Co the economic cycles of the markets in 8.90 which they operate, all the more so because they often occur in sectors with a high degree of cyclicality, such as commodity markets, where surges in demand result in massive expansion in capacity which ineluctably trigger crises of overcapacity. Instability is also inherent m collusive schemes, where supra competitive profits may only be attained through collective discipline, but where the temptation to poach customers by undercutting collusive prices is high, as this may result in even higher benefits for the individual firm. Despite the efforts deployed to ensure mutual control, tensions within cartels are therefore frequent, and mutual grievances often result from the eagerness or the lack of discipline of certain members.351
Claims forhigher quotas    Cartel participants may threaten the other participants by adopting    8.91 competitive behaviour in order to obtain more favourable treatment within the cartel. In Sodium Gluconate, a previous cartel had collapsed because a participant had decided to
3« [2004] OJLl 15/1, paras 104-106.
346   [1992] OJU34/1, para25.
347  [2001] Oj L152/24. The two ringleaders envisaged economic sanctions against a Korean producer who did not comply with the cartel agreement. In particular they threatened it with an anti-dumping action in Europe if it did not limit sales quantities in this area. The Korean producer subsequently agreed to comply with the cartel agreement (see paras 88,330).
3flB [2002] OJL239/18-Withregard to the growing competition from Chinese producers, the cartelmem-bers decided that on forthcoming trips by representatives of the cartel members to China, the local producers were to be threatened with an anti-dumping complaint (see para 122).
a« [1999]0JL24/I.
350  [2002] OJ L239/18. Target customers were allocated to each cartel member. This catalogue of target customers came to be known as the 'Serbia List' and was the subject of regular monitoring and discussion.
351   On theinherent instability of cartelsand the determining factors of their stability, see PZ Grossman (ed), How Cartels Endure and How they Fail—Studies of Industrial Collusion (2004). See also MC Levenstein and VY. Suslow "What determines Cartel Success?' (2002), available at SSRN: <http://ssrn.corn/abstract=2994l 5>.
791
Chapter 8: Cartels
expand its production capacity by constructing new plant. After a period of intense pn war, the cartel was reformed after the undertaking which had led to the collapse manaetU to obtain a higher quota which took into account its expanded capacity. In Vitamin Takedas entry into the cartel arrangements for vitamin B2 provoked intense negotiation regarding annual quotas, with Takeda alleging that Roche had made a mistake in the calcn lation of market shares and demanding a higher allocation. The Commission noted rhar 'Takeda's volume aspirations were clearly a cause of irritation for Roche*.352 In Citric Acid disputes also arose because certain undertakings were unhappy with the quotas that had been allocated to them.353
8.92 Mutual suspicion of cheating Grievances frequently result from suspicions that certain cartel members are not complying with their obligations. In Sodium Gluconate354 ^ Citric Acid,355 for instance, Tensions arose regarding the monitoring of sales data, as certain cartel participants were accused of cheating. Food Flavour Enhancers provides numerous examples of tensions between cartel participants. For instance, one parricipant would no longer agree to cooperate on pricing unless another participant agreed to buy certain quantities of the product from it.356 Grievances regarding alleged non compliance were also numerous.357 Similar episodes can be found in Carbonless Paper356 and Austrian Banks.359
8-93 Leniency programmes as an additional factor of instability On top of the natural instability of cartels, the adoption of leniency programmes by a large number of competition enforcement authorities has added an additional and significant source of distrust between cartel members. In the EU, the chance of obtaining full immunity in exchange for reporting an unknown cartel to the Commission has considerably increased the incentive for cartel members to denounce their peers before another participant does so. It is likely that the racrs that will be set out in future cartel decisions will illustrate this new and important source of tension within collusive schemes.
«2  (2003]OJL671,para280.
a« [2002]OjL239/18.para97.
«« Para 93.
355 Jungbunzlauer was seen to be 'causing problems' in che group because it did not strictly adhere to the agreement at alí times and was perceived to be 'ill-disciplined' by the other cartel participants (para 117).
i« [2004] OJL75/1, para 102.
357   Paras 114, 116.
358   [2004] OJ Li 15/1. The Commission reports that a cartel participant who doubted the accuracy of the figures submitted by another asked for, and obtained, the right to directly verify the figures in question at the premises of that undertaking (see para 106).
359   [2004] OJ L56/1. The case of one bank cutting interest 'without prior notice' prompted strong and immediate reactions from the other cartel members, who immediately convened the relevant cartel committee. The Commission reports that '[f]eelings ran high at the meeting and the undertaking in question, which had already not complied with other cartel measures and which had not provided appropriate information to competitors on the measures taken, was fiercely attacked by the other banks. It was "the unanimous opinion" that there was "no justification" for the measure taken and that it "contradicted the stated objective of all the relevant committee meetings" ie "not to go along with" any cuts in interest rates on loans- Any "such senseless competition" was downright dangerous'(see para 172).
792
C. Investigating Hardcore Cartels
C.  Investigating Hardcore Cartels
'- fl) The Initiation of a Case: Information Sources
The information that may prompt the Commission to start an investigation can come from 8.94 a variety of sources. It is widely known that a prime case-generator is the Commissions Leniency Notice.360 Under the Leniency Notice, undertakings that self-report a cartel infringement and provide sufficient evidence to the Commission can receive immunity ftom fines.36' The Commission also has other means of tracing cartels: an investigation may result from the Commission's own monitoring of markets, from information Transmitted to the Commission by Member State authorities or by third country agencies; it may also originate from submissions by complainants (private customers or industrial clients), or from whistleblowers (for instance (ex-)employees who wish to report wrongdoing). Each of these possibilities is discussed in greater detail below. Given the important place of the Commissions Leniency Notice as an investigative tool in cartel proceedings, a hill In-depth section is dedicated to a discussion of all aspects of its functioning.362
(a) Market Monitoring and Information from Other Investigations
The Commission may put itself on the track of collusive behaviour through so-called 'mar- 8.95 ket monitoring'. The term relates to the informal gathering of information and finding of possible illegal behaviour in a given market, based on indications, first and foremost of an economic nature, through an analysis of structural, behavioural and performance indicators. The Commission may search for particular occurrences or trends that go against the expected economic development of those markets, such as a sudden (upward) swing in the movement of prices or strong price differentials between geographic markets. Furrhermore, the Commission can examine the mechanisms by which particular markets operate and as a result of which (price) transparency is achieved.
Unless a market monitoring exercise leads to the actual acquisition of evidence of a hard- 8.96 core cartel (which is relatively unlikely), the results may still provide the Commission with a sufficient basis for issuing requests for information. The question, however, is whether the Commission will be able to obtain sufficient information through market monitoring to enable it actually to carry out (surprise) inspections.363 In any case, it is difficult to see how the Commission can obtain sufficient leads without actually requesting information from
360  The Commission stated in 2004: 'In recent years, most cartels have been detected by the European Commission after one carte! member confessed and asked for leniency, [.. .]'■ See the brochure EU competition polity otuž the consumer, European Commission, 2004, available at <http://www.pubiications.cu.int/>. (It should be noted that the Leniency Notice docs not require a specific confession (as the brochure states) by the undertaking requesting immunity or a reduction of fines. On the specifics of the Leniency Notice, see further para 8.104.)
361   Ifthe Commission issues a final prohibition Decision with fines after the conclusion of its investigation, the undertakings whose (conditional) immunity is confirmed in that Decision will nevertheless be named in it, with all relevant facts about their participation. Also, a fine will be calcukred for it, from which it is then granted immunity.
3« Seeparas8.105-8.257,below.
363 It should be noted that the mere fact that an industry appears not to be performing in acompetitive manner, cannot, by itself, constitute proof of an infringement, unless—according to Woodpulp—thete exists no other explanation for the malfunctioning of the market than an agreement ora concerted practice between competitors.
793