EC, April 30, 1999, No 1999-485

THE COMMISSION OF THE EUROPEAN COMMUNITIES,

Having regard to the Treaty establishing the European Community, Having regard to Council Regulation (EEC) No 4056-86 of 22 December 1986 laying down detailed rules for the application of Articles 85 und 86 of the Treaty to maritime transport (1), as amended by the Act of Accession of Austria, Finland and Sweden, and in particular Articles 3, 11, and 12 thereof, Having regard to the Commission Decision of 6 April 1994 to initiate proceedings in this case, Having given the undertakings concerned the opportunity to make known their views on the objections raised by the Commission and to present any other comments in accordance with Article 23 of Regulation (EEC) No 4056-86 and with Commission Regulation (EEC) No 4260-88 of 16 December 1988 on the communications, complaints and applications and the hearings provided for in Council Regulation (EEC) No 4056-86 laying down detailed rules for the application of Articles 85 and 86 of the Treaty to maritime transport (2) as last amended by the Act of Accession of Austria, Finland and Sweden (3), Having consulted the Advisory Committee on Restrictive Practices and Dominant Positions in the field of maritime transport, Whereas:

SUMMARY

(1) In this Decision, the Commission considers whether the parties to the Europe Asia Trades Agreement (EATA) have infringed Article 85 (1) of the Treaty in relation to an agreement not to use capacity and to exchange information and examines their application for individual exemption.

THE FACTS

I. The application

(2) On 2 September 1992, in accordance with Article 12 (1) of Regulation (EEC) No 4056-86, the Commission was notified of an requested to exempt under Article 85 (3) of the Treaty, the Europe Asia Trades Agreement (EATA) concerning scheduled maritime transport services for the carriage of containerised cargo from north Europe to the Far East. On 19 September 1997, the Commission was informed that the EATA parties had terminated the agreement with effect from 16 September 1997.

(3) The following shipping lines were party to the EATA:

- CGM Orient SA (CGM),

- Hapag-Lloyd AG (Hapag-Lloyd),

- Kawasaki Kisen Kaisha Ltd (K Line),

- A.P. Møller - Maersk Line (Maersk),

- Malaysian International Shipping Corporation Bhd (MISC),

- Mitsui O.S.K. Lines Ltd (MOL),

- Nedlloyd Lijnen BV (Nedlloyd),

- Neptune Orient Lines Ltd (NOL),

- Nippon Yusen Kabushiki Kaisha (NYK),

- Oriental Overseas Container Line (OOCL),

- P & O Container Line (P & OCL),

- Cho Yang Shipping Co. Ltd (Cho Yang),

- Deutsche Seereederei Rostock GmbH (DSR),

- Evergreen Marine Corp (Taiwan) Ltd (Evergreen),

- Hanjin Shipping Co. Ltd (Hanjin),

- Hyundai Merchant Marine Co. Ltd (Hyundai),

- Senator Linie GmbH (Senator),

- Yangming Marine Transport Corp. (Yangming) (4).

(4) A summary of the application was published in the Official Journal of the European Communities on 6 April 1993 (5) in which the Commission stated that it considered prima facie that the agreement fell within the prohibition in Article 85 (1) but that it had not yet taken a position as to the applicability of Article 85 (3). The Commission informed the former EATA parties on 14 June 1993 that it had serious doubts as to whether any of the four conditions of Article 85 (3) were satisfied.

(5) The Commission received complaints about the EATA from a number of third parties including the British Shippers' Council (BSC) the European Shippers' Council (ESC), the French Conseil National des Usagers des Transports (CNUT) and the Japan Shippers' Council (JSC). The former EATA parties have been given an opportunity to comment on these complaints.

II. The notified agreement

(i) The parties

(6) The former EATA parties are all shipping lines. A number of them are members of the Far Eastern Freight Conference (FEFC) while the remainder are independent shipping lines operating in the same trades (see recital 66).

(7) Hyundai became a party to the EATA on 12 March 1993. The Commission was informed on 23 July 1993 that the East Asiatic Company Ltd and EACBen Container Line Ltd had left the EATA with effect from 30 June 1993. These two shipping lines no longer operate vessels on the trades in question, although for a while the vessels they had operated were maintained in use on the same trades by Maersk. CGM left the North Europe/Far East trade in 1994 and on so doing ceased to be a party to the EATA.

(ii) Summary of the agreement

(a) General provisions

(8) Article 2 of the EATA provided that the purpose of the agreement was to establish a capacity management programme in order to achieve:

" (a) the optimum use of available capacity on deepsea vessels owned, operated or controlled by the [EATA] parties ... and

(b) the improvement of revenue ... to a level consistent with a reasonable rate of return on investment for the services provided, and to maintain the viability of such services in the future ...".

(9) According to the recitals to the EATA (6), this "improvement of revenue" was to be brought about by remedying the "consistently depressed freight rates" referred to in recital (b) to the EATA. Thus, the purpose of the EATA was to bring about an increase in freight rates on services operated by the EATA parties between North Europe and Far East.

(10) One of the ways in which these objectives were to be achieved was by the allocation to each of the parties of a "maximum allowed capacity", i.e. the maximum amount of capacity that each party was allowed to offer to the market for the carriage of goods (Article 5 (a)). These allocations could be made in respect of a particular area, several particular areas or generally (Article 16 (e)). They could be made for some or all of the parties (Article 6 (e)). The agreement included provisions for the calculation and revision of the allocations.

(11) The "maximum allowed capacity" of each party was to be calculated according to the eastbound slots available per vessel declared by each party (Article 16 (b)). In practise this was done by calculating the total which would be available in the absence of the EATA and reducing it by a certain percentage. This maximum was then divided into four periods of account of three months each.

(12) If a party exceeded its maximum allowed capacity in any period of account, it was potentially liable to pay a "capacity charge" calculated by reference to the number of TEUs (7) by which it had exceeded its allocation for the period in question (Article 22). Unused portions of a Maximum Allowed Capacity in any period of account could not be carried forward or transferred (Article 18)).

(b) Administrative provisions

(13) The agreement was administered by a General Policy Committee, a Market Review Committee and a Secretariat (Article 5b)).

(14) The General Policy Committee was made up of representatives of each of the parties and determined the size of the allocations of maximum allowed capacities on the basis of recommendations made by the Market Review Committee (Article 6). It also decided the amount of any capacity charges imposed by reason of a line carrying more than its maximum allowed capacity as well as any other measures necessary to give effect to the objectives of the agreement, such as the exchange of information on market conditions and the size of penalty payments for failure to provide information (Article 6).

(15) The Market Review Committee was also made up of representatives of the parties (Article 7). Its function was to consider market conditions and to report to the General Policy Committee. It was supplied by the secretariat with the declarations made by the parties as to their fleet capacity and monthly liftings (i.e. cargo loaded) (Article 14). The agreement provided that the parties were to supply the secretariat with any data, reports or documents necessary to ensure compliance with the agreement (Article 12).

(16) It was the practice of the EATA parties to provide a declaration of the individual capacity of each of their vessels biannually and in advance. The parties also supplied the EATA secretariat every month with the following details:

(i) name of vessel,

(ii) date of Suez Canal entry,

(iii) maximum declared capacity in TEUs,

(iv) total actual filled slots in TEUs,

(v) non-scope cargo (8) in TEUs lifted,

(vi) percentage utilisation,

(vii) a forecast of capacity for each vessel for the following two months,

(viii) estimated monthly totals for the next four months,

(17) The secretariat was accordingly supplied with details of the previous month's actual liftings as well as a forecast for the next six months' liftings. Each of the EATA parties participated in this exchange of information form its inception in 1992 until May 1997: a number of the EATA parties continued to exchange information until July 1997. CGM ceased to provide the relevant information with effect from October 1994.

(18) The secretariat was also given the task of monitoring compliance with the agreement (Article 20). According to the agreement, the Director-General and staff of the FEFC were to serve as secretary and secretariat of the EATA for the duration of the agreement, unless the parties to the EATA otherwise unanimously agreed (Article 30).

(19) The agreement applied to scheduled international maritime transport services for containerised goods between Northern Europe and Asia via the Suez Canal. For the purposes of the notified agreement, the description "Asia" excluded Pakistan, Sri Lanka, India and Bangladesh and, for certain purposes, the People's Republic of China (Article 3).

(20) Westbound Trades were not directly concerned by the notified arrangements but the EATA parties acknowledged that it was possible that some effect on capacity and freight rates would also be experienced in that direction as a result of the reorganisation of eastbound services reducing the frequency, size or number of calls of vessels (9).

(21) Article 4 of the agreement stated that nothing in the agreement was to have prevented, limited or otherwise precluded the rights of the parties to set their rates independently. This provision should be considered in the light of the comments concerning market structure (recitals 66 to 79).

(22) Any party could withdraw from the agreement by giving less than 90 days' notice (Article 33 (d)). However, notwithstanding such withdrawal, parties remained bound to observe their maximum allowed capacities for a period of up to 12 months unless they ceased to own,operate or maintain vessels as ocean common carriers (Article 18). Failure to comply with the obligation to observe a maximum allowed capacity potentially resulted in the forfeiture of a financial guarantee and other financial penalties (Article 18). New parties were required to receive the unanimous approval of existing parties (Article 33).

(23) Payments in respect of early withdrawal, exceeding maximum allowed capacities or for failure to provide information were to be paid into a fund and distributed equally between the parties at the end of each year (Article 25). Administrative and other expenses incurred in connection with the EATA were to be paid by the parties to the agreement in equal shares (Article 31).

(24) The EATA was said to be of indefinite duration, to be terminated by all the parties to it once the structural problems the parties considered to exist on the northern Europe/Far East trades had been resolved on a lasting basis (10).

III. Implementation of the agreement

(25) On 26 November 1992, the parties agreed to limit the amount of capacity each of them offered for supply for the period from 1 January 1993 to 31 March 1993 by between 6,25 % and 12,25 %, calculated on the basis of a sliding scale so that smaller lines (in terms of those with smaller capacities on the routes in question) reduced the amount they offered to the market by smaller proportions than did the bigger lines. This reduction was set at an initial level of 10 % of the capacity offered eastbound by the parties to the EATA.

(26) On 12 March 1993, the parties agreed to limit further the amount of capacity each party to the EATA offered for supply on the northern Europe to the Far East eastbound trades by increasing the percentage reductions for the period 1 April 1993 to 30 June 1993 to between 11 % and 17 %, also calculated by reference to a sliding scale. The EATA parties claimed that this amounted to an aggregate maximum reduction of 15 %.

(27) The same aggregate reduction and sliding scale was maintained for the third quarter of 1993. The maximum aggregate reduction was reduced on 5 October 1993 to 5 % for the fourth quarter of 1993, still on the basis of the sliding scale, but on 27 October 1993 was further reduced to zero. Full details of these reductions and the sliding scale are given in Annex II. No reduction was imposed between 27 October 1993 and the date of abandonment of the agreement.

(28) In the second quarter of 1993, five of the EATA parties carried more cargo than permitted under the maximum each had been allocated.

IV. Capacity non-utilisation agreements in liner shipping

(29) A "capacity management programme" is an agreement under which the parties agree not to use a proportion of the space on their vessels for the carriage of goods in a particular trade. The proportion set aside is part of the forecast excess of supply over demand.

(30) The parties agree that this space cannot be used for cargo falling within the geographic scope of the agreement ("scope cargo"), although it may be used for cargo originating from elsewhere and transhipped onto the vessels in the programme ("non-scope cargo"). In this Decision, the expression "non-utilisation" is used to describe agreement not to use space for scope cargo.

(31) There have been three examples of capacity management programmes: the Trans-Atlantic Agreement (TAA), the Trans-Pacific Stabilisation Agreement and the EATA. Each of these agreements has operated on one of the three main world trade lanes and in each case has been made up not only of the conference members in the trade but also the most important non-conference carriers.

(32) The 15 parties to the Trans-Pacific Stabilisation Agreement (11) operated a capacity non-utilisation programme from 1989 to 1995, when it was indefinitely supended. During this period those parties operated some 80 % of the available capacity on the trades between Asia and the United States. The extent of the capacity agreed to be set aside varied from 6 % to 15 %. According to Drewry (12), in 1994 this was the equivalent to the artificial withdrawal of the eqivalent annual capacity of seven Panamax container vessels of 4000 TEU each.

(33) When the parties to the Trans-Pacific Stabilisation Agreement applied to the US Federal Maritime Commission in 1996 to restore the capacity non-utilisation programme, the Federal Maritime Commission launched a detailed investigation following which approval was not forthcoming and in March 1997 the application was withdrawn.

(34) The parties to the Trans-Atlantic Agreement implemented their capacity non-utilisation agreement in August 1994 until September 1994. The agreement was for up to 25 % of space not to be used for scope cargo but in practice only around 15 % of westbound capacity was withdrawn from sale. On 19 October 1994, the Commission adopted Decision 94-980-EC(13) prohibiting the Trans-Atlantic Agreement.

(35) On the Europe/Far East Trades, the members of the FEFC had until 1990 a market-sharing agreement, setting the maximum share of the trade that each member of the conference, or each consortium operating within the conference, could carry. There was no agreement not to use certain capacity.

(36) According to the EATA parties, the share of the FEFC on the north Europe/Far East trades had by 1990 fallen to approximately 59 % and the market-sharing agreement was terminated for the principal reason that,

"the parties to it no longer had a sufficient critical mass in the NE/FE trades to regulate reserve capacity in a way that would contribute to the stabilisation of the trades (14)."

(37) Drewry has estimated (15) that the effect of the EATA in 1993 would over a full year have meant the non-utilisation of some 13 500 TEUs of capacity, the equivalent of around 3,33 4 000 TEU vessels. On the basis of a 63-day round voyage time, each vessel could be expected to undertake 5,8 round voyages per annum. Drewry estimates (16) the fixed costs of each round trip for a 4000 TEU vessel on the Europe/Far East trades in 1996 to be the equivalent of some USD 4,1 million. As Table 1 indicates, the total fixed costs over a 12 month period involved in operating the capacity which the EATA parties agreed not to use would have been in the region of USD 80 million.

Table 1

Estimate of annual fixed costs

<emplacement tableau>

V. The relevant product market

(38) The relevant product market for the purpose of considering the EATA is that of scheduled maritime transport services for the transport of containerised cargo from north Europe to the Far East. This includes on the one hand ports in Belgium, Denmark, Finland, Northern France, Germany, Iceland, Ireland, the Netherlands, Norway, Sweden and the United Kingdom, and on the other hand ports in Hong Kong, Japan, North Korea, South Korea, Malaysia, the Philippines, Singapore and Taiwan.

(39) The parties consider that there are a number of substitutable ways of transporting goods from Northern Europe to the Far East.

(i) The first is that provided by specialised vessels which can carry some of the large-volume homogeneous products moving within the general cargo sector.

(ii) The second is provided by ships in the bulk of specialised sectors which can also carry a number of containers.

(iii) The third is said to be air and air-sea combinations for good requiring faster shipment.

(iv) The fourth substitute is the possibility of transporting goods by the trans-Siberian Railway.

(v) The fifth possibility is offered by carriers operating westbound out of north Europe which can also serve the Far East either by using rail-land bridges across the United States, or by transshipping at US west coast ports onto Pacific services (i.e. Europe/US east coast/US west coast/Far East).

(vi) Finally, the parties consider that services from the Mediterranean and Black Sea represent a significant source of competition to the operators on the north Europe to Far East trades.

(40) For the reasons set out below, none of these is considered to form part of the same market as the market for scheduled maritime transport services for the transport of containerised cargo from north Europe to the Far East.

(41) In the Tetra Pak Case (17), the Court of Justice of the European Communities stated that the stability of demand for a certain product is a relevant criterion for defining a relevant market and that the fact that different products are, to a marginal extent, interchangeable does not preclude the conclusion that these products belong to separate product markets.

(42) The Commission takes the view that, for the vast majority of categories of goods and users of containerised liner shipping, the other forms of maritime transport, including conventional (break-bulk) liner transport, do not offer a reasonable alternative to containerised transport services on the routes falling within the relevant geographic market in this case and that these services constitute one or more markets in their own right.

(43) The effect of marginal competition from other means of transport for certain categories of goods can be limited. This situation arises because liner shipping companies are able to identify shippers of such goods and, because of the differentiated price structure in liner shipping, offer lower prices to such shippers without necessarily affecting prices generally.

(i) Non-scheduled services

(44) Firstly, scheduled maritime transport, or liner, services constitute a separate market from that of tramp services. On the whole the nature of the relationship between shippers on the one hand and shipping lines is quite different depending on whether the latter is providing liner services or tramp services. Liner services are provided on the basis of the operator being a "common carrier": that is to say, the transport service provider offers to carry all the goods brought to it for carriage. On the other hand tramp services are usually supplied on the basis of ad hoc individually negotiated contracts (18).

(45) Charters are only viable for containerisable goods (19) provided the shipper has a sufficiently large cargo, or is able to combine with other shippers for each trip. Furthermore, chartering and bulk or specialised services will not as a rule give the frequency, regularity or dependability of service required by many shippers, nor can they provide the door-to-door service often required.

(ii) Bulk services

(46) It is clear that many bulk commodities can be containerised and that before the advent of containerisation (in the late 1950s) all goods travelled bulk of some description or another. In the present case, in order to determine the competitive condition in the relevant market, it is only necessary to consider the effect of substitutability from carriage in container to carriage in bulk: there is no lasting substitution from container to bulk in the vast majority of cases.

(47) Almost all cargo can be containerised and, over time, it is likely that the degree of containerisation in most maritime markets involving Member States will be very high. In mature markets, such as the northern Europe/US or the northern Europe/Far East markets, the process of change toward containerisation is more or less complete and few, if any, non-containerised cargoes are left which are capable of being containerised.

(48) Furthermore, once a type of cargo regularly becomes containerised it is unlikely ever to be transported again as non-containerised cargo. The reasons for this are that shippers become accustomed to shipping in smaller but more frequent quantities and become accustomed to the fact that once cargo has been loaded into a container, it is easier to ship onwards from the port of delivery to the ultimate consignee using multimodal transport.

(49) Thus, as the degree of containerisation increases, shippers of non-containerised cargoes turn towards containerised services but once those shippers have become accustomed to shipping in containers they do not revert to non-containerised shipping. Such examples of one-way substitutability are not uncommon (20).

(50) Drewry (21) makes a cautious estimate that the containerised share of world general cargo trade has risen from 20,7 % in 1980, to 35,1 % in 1990, to 41,6 % in 1994. Drewry forecasts that by the year 2000, this percentage will rise to 53,8 %. The transformation from bulk shipping to containerisation reflects not only a change in the nature of the goods being shipped (essentially from raw materials to manufactured commodities), but also the inherent characteristics of containerised liner shipping.

(51) These characteristics include the following. Smaller more frequent shipments, as is typically the case with container shipping, lead to reduced inventory costs. Containerised goods are less likely to suffer from damage and pilferage. Containerised goods are easier to ship multimodally. For these reasons, once a commodity has made the transformation from bulk to container, possibly on a route-by-route basis, the differences in the nature of the service being provided mean that once that transformation period is over, a shipper is highly unlikely to revert to bulk shipping.

(52) In this context, it is not important that certain commodities may still travel by both means: the essential question for determining demand substitutability is whether the choice of mode is made on the basis of the characteristics of the mode. Thus, the fact that some steel products may travel by bulk and others by container does not show that the two modes are substitutable since it does not take into account the diverse nature (and value) of steel products nor the delivery requirements of customers. The same is true for the other products for which the parties claim substitutability between bulk and containers.

(53) Even to the extent that reefer containers (i.e. refrigerated containers) may be substitutable for bulk reefer services (22), for the reasons given above this does not mean that bulk reefer services are substitutable for reefer container services. Apart from the advantages offered by liner container services, such as smaller volumes and speed of transfer to other modes of transport, a wider variety of products can travel in reefer containers than can travel as bulk reefer cargoes. Such products include furs and leathers, pharmaceuticals, electronic goods and, because of the steady temperatures and the ability to control ripening, soft fruits.

(54) On this basis, while it is possible that in exceptional circumstances some substitution may occur between break-bulk and container transport, it has not been demonstrated that there is any lasting substitution from container towards bulk for the vast majority of cases.

(55) On the supply side, the parties have argued that break and neo-break carriers could readily convert their vessels so as to enter into the relevant market and for this reason should be regarded as potential competitors.

(56) In principle, any vessel can carry containers. The effect of potential competition from operators of non-fully containerised vessels could only be material if both of the following two conditions were fulfilled. First, it would have to be shown that suppliers of such services could economically compete with the parties on even terms and second that customers regarded carriage on a non-fully containerised vessel as being functionally interchangeable with carriage on a fully containerised vessel.

(57) In considering whether the first of these cumulative tests is fulfilled, it is essential to note that the characteristics and performance of non-fully containerised vessels are significantly different from those of fully containerised vessels.

"It is crystal clear, and almost implicit in the terminology, that cellular containership capacity is more efficient and more productive than non-cellular space when it comes to carrying unitised (i.e. containerised) cargo, and hence of greater significance as far as the supply/demand balance is concerned. Each slot on a cellular vessel will provide more container carrying capacity in any given year than a slot on a non-cellular vessel, since the cellular ship:

- spends less time in port,

- usually possesses a much higher sea-speed,

- operates on regular lines schedules.

While ro-ros may achieve 80 % or more of the productivity of a cellular vessel, a semi-containership or a con-bulker will be considerably less efficient on the grounds of both speed and cargo handling time in port. An overall estimate of the relative productivity ratios for cellular and non-cellular capacity might reasonably be 2:1 (23)."

(58) Apart from the performance characteristics which militate against supply-side conversion, there are a number of technical characteristics. The first of these is the additional expenditure required to carry containers on vessels which were not specifically built as container vessels. Such costs are both one-off in the sense that chains and fittings have to be purchased (according to Dynamar BV, shipping consultants, at a cost of some ECU 150 per slot) and variable in the sense that the cost of labour is higher for stowing containers on non-container vessels than on container vessels. Account need also to be taken of the additional port costs involved in carrying containers on such vessels due to slower towing times and consequentially longer periods spent in port.

(59) The second reason for which the potential capacity of non-fully containerised vessels is less than the parties assert is the fact that the operators of such vessels do not possess the same fleets of containers as do operators of fully containerised vessels. Typically each of the parties has three containers for every vessel slot it operates. Many operators of breakbulk services will own no containers at all. The significance of this is especially important given that "the global box inventory has seldom, and certainly not for the last 10 years or morebeen sufficient to permit all the nominal containership capacity of the non-cellular fleet to be used" (24). This situation is compounded by the fact that the operator of non-fully containerised vessels do not generally possess the same landside facilitites as do operators of full-containerised vessels.

(60) So far as the customers are concerned, the Commission does not accept that the vast majority of customers of the parties would regard carriage on a bulk or neo-bulk vessel as being substitutable for carriage on a fully-containerised vessel. Differences so far as customers are concernded are the absence of scheduled weekly sailings and the fact that in many cases non-fully containerised vessels use different port terminals or berths to the ones used by fully containerised vessels with consequential inefficiencies for multimodal transport operations.

(iii) Air transport services

(61) Air transport services (or combined sea-air transport services) are only appropriate for goods which require faster transport times, the high value of which permits higher transport costs to be incurred.

(62) Given that any shipper is concerned to reduce its transport costs to the necessary minimum, it will only incur the additional expense of air transport services if it is necessary for its goods to be delivered faster than would be possible if they were carried by sea and if the value of the goods is sufficiently high to bear the higher transport costs. In such a case it would not be correct to say that sea transport services were substitutable for air transport services. It should be noted that the vast majority of goods carried eastbound from northern Europe to the Far East (the scope of the EATA) are considered by the EATA parties to be of low value.

(iv) Land transport services

(63) The parties point out in the notification that the trans-Siberian Railway is not an efficient substitute form of transport during the present period of economic and political instability in the Commonwealth of Independent States. It is understood that volumes carried on the trans-Siberian Railway have decreased considerably since the collapse of the Soviet Union. In any event, it is understood that a single 4 000 TEU vessel carries the equivalent of some 50 train loads of approximately 2,4 km in length.

(v) Trans-Pacific services

(64) The parties have furnished no evidence that transshipments at US ports or rail land bridges account for material quantities of goods carried from north Europe to the Far East.

(vi) Mediterranean/Black Sea services

(65) The parties have also argued that services from the Mediterranean and Black Sea represent a significant source of competition to the operators on the north Europe to Far East trades. The parties have failed to give a single example of cargo actually switching from one to the other in either the application for exemption or the Reply to the Statement of Objections. Furthermore, Drewry considers that:

"By turning in the Mediterranean, Europe-Far East ships could save at least two weeks on their average nine-week round voyage time (a 22 % increase in vessel productivity), but this seems unlikely in anything like the foreseeable future, given the infrastructural limitations of the southern ports and the Eurpean rail system(25)."

VI. Market structure

(66) In considering the EATA, it is relevant to consider not only the market shares of the parties to the agreement at the relevant time (as to which see recitals 80 and 81) but also the structure of the market. In particular it is significant that many of the EATA parties were members of the FEFC and that all the parties to the EATA, other than Hyundai, were parties to the Far East Trade Tariff Charges and Surcharges Agreement (FETTCSA). Membership of the EATA parties to each of these agreements is set out in Table 2 (26).

Table 2

Market Structure - 1993

<emplacement tableau>

(i) The Far Eastern Freight Conference (FEFC) (27)

(67) The FEFC is an association of shipping lines which operate liner shipping services on routes between Europe and the Far East. The FEFC does not itself offer any shipping service or enter into contracts with shippers or other customers. It regulates a number of conditions, including tariffs, on the basis of which its member lines offer their services. The FEFC is used by its member lines as a central body for publishing their decisions and communicating with other relevant organisations, including shippers' councils.

(68) In order to come within the terms of the block exemption for liner conferences (28), the members of the FEFC are obliged, according to Article 1 (3) (b) of Regulation (EEC) No 4056-86, to operate under uniform or common freight rates.

(69) As can be seen from recital 80, the members of the FEFC had a 58 % market share in 1991 for traffic between northern Europe and the Far East. In other words, more than 50 % of the relevant market was, in principle, moving between northern Europe and the Far East at uniform or common rates decided upon by the shipping lines constituting the majority of the membership of the EATA.

(ii) The Far East Trade Tariff Charges and Surcharges Agreement (FETTSCA)

(70) The FETTCSA was an agreement dated 5 March 1991 between all the major shipping lines on the Europe/Far East trade (other than Hyundai), including 12 members of the FEFC and six non-conference lines operating in the trades. According to the EATA secretary (29), the FETTCSA emerged as a result of discussions by various shipping lines concerning the forerunner to the EATA, the European Stabilisation Agreement (ESA).

(71) The FETTCSA envisaged discussions by the parties for purposes of calculating and setting charges and surcharges other than sea freight and inland freight by the following means:

(i) the establishment or application of uniform rules concerning the structure and the conditions governing the application of transport tariffs, and

(ii) the establishment or application of inclusive rates and conditions for the organisation and execution of successive or supplementary maritime transport operations.

(72) The FETTCSA provided for the parties to exchange views and information regarding any matter within the scope of the agreement as well as for the administration of the agreement.

(73) According to data from the European Shippers' Council (ESC) and the Conseil National des Usagers des Transports (CNUT) (30) and accepted by the FETTCSA secretariat (31), charges additional to ocean freight such as surcharges often amount to some 35 % of the total transport cost to shippers' or as much as 60 % of the actual freight rate. The Japan Shippers' Council (JSC) has commented that charges and surcharges are much higher than corresponding expenses and therefore constitute part of the lines' overall revenue.

(74) An agreement which allows competitors to discuss the manner in which prices are set and the elements which are to be included in prices is likely to have an effect on freedom to set prices. It amounts to an attempt to fix price levels and not just methods of calculation and causes those prices to be different from what they would be otherwise.

(75) The parties to the FETTCSA terminated that the agreement following the adoption of a Statement of Objections by the Commission in 1994.

iii) Combined effect of the FEFC, FETTCSA and EATA

(76) The express purpose of the EATA was to allow the parties to it to increase their freight rates (see recital 9 of this Decision). Although some members of the FEFC were not party to the EATA, the system of price-fixing practised by the FEFC and the fact that the most important members of the FEFC were parties to the EATA would have ensured that all members of the FEFC operating within the geographical scope of the EATA would have benefited from rate increases whether or not they were party to the EATA.

(77) The combination of the FEFC and the two agreements between the FEFC and independent shipping companies, the EATA and the FETTSCA, is likely to have had a significant effect on price competition between the members of the FEFC (who charge common or uniform rates and do not therefore compete on price) and the non-conference shipping lines. This is because both the EATA and FETTCSA were likely to have caused prices to be raised in a coordinated manner.

(78) Apart from the direct effect the EATA would have had ou the freight rates of the EATA parties, there is also the possibility that the freight rates of independents which were not party to the EATA would also have been increased as a result of the operation of the agreement. This possibility, which would amount to normal and legitimate commercial behaviour on the part of the non-EATA-shipping lines, arises from the fact that those independent lines take into account the FEFC's tariff(32) in setting their prices and the FEFC tariff will itself will have been raised as a result of the operation of the EATA.

(79) The EATA and the FETTCSA contained provisions appointing the Director-General and staff of the FEFC as secretary and secretariat of the respective agreements which are identical in all material respects. Their role of administering the three agreements was likely to reinforce the strong links and increase the exchange of information between the conference members and non-conference members.

VII. Market shares

(80) According to EATA estimates supplied to the Commission in the application for exemption (and therefore relating to the period before the EATA came into effect), the EATA parties had a market share of some 86 % of all eastbound liner traffic from north Europe to the Far East in 1991. This leaves 14 % of the market nominally unaffected by the agreement. The figures include an allowance for traffic carried on the trans-Siberian Railwy. For the reasons given in recital 63, the Commission does not consider that the trans-Siberian Railway forms part of the same market but because of the relatively unimportant quantities (approximately 2 %) carried in the manner it has not been considered necessary to adjust the figures applied by the EATA so as to exclude the trans-Siberian Railway.

Table 3

Volumes and market shares 1991

Containerised cargo on liner vessels eastbound Northern Europe to Far East

<emplacement tableau>

(81) The EATA parties have estimated that the comparable figures to those given for the EATA parties' market share in 1991 in Table 3 were 83,5 % in 1993, 80 % in 1994 and 78 % in 1995.

VIII. Freight rates

(82) The EATA is not a conference agreement (see recital 180) and has no direct mechanism for agreeing on the implementation of freight-rate increases. However, freight-rate increases announced by the Eastbound Management Agreement (EMA-an operating arm of the FEFC which sets FEFC rates for eastbound services) since the implementation of the EATA were USD 100/200 per TEU/FEU in April 1993 and a further USD 75/100 per TEU/FEU from July 1993. In November 1993, after the reduction in the amount of capacity supplied in the market was reduced to zero, it was announced that the EMA would increase rates by a further USD 150/225 per TEU/FEU with effect from 1 January 1994.

(83) According to Drewry (33), average eastbound freight rates on the Europe/Far East trades increased as shown in Table 4.

Table 4

Europe/Far East Average Freight rates 1992 to 1995

<emplacement tableau>

(84) On 7 June 1993, the Member Lines of the Asia Westbound Rate Agreement (AWRA - another operating arm of the FEFC covering the westbound trades to Europe and the west Mediterranean) announced that they had agreed to apply a "rate restoration" of USD 150/300 per TEU/FEU with effect from 1 July 1993. A rate restoration in this context is understood to mean an agreement to reduce the level of discounts across the board, i.e. it is not an increase in the conference tariff rate (which is known as a general rate increase (GRI)) but an attempt to raise actual prices toward the published rate.

(85) According to the EATA secretary, the EATA parties discussed at their meeting in Paris in March 1993 a proposal to extend the EATA to give the parties authority to discuss rates and this proposal did not lead to any decision but some Members considered that were was merit in pursuing the matter further(34).

IX. Capacity utilisation

(86) Capacity utilisation on the north Europe/Far East eastbound trades is affected by the fact that deadweightlimitations can reduce a ship's carrying capacity by up to 20 % so that 80 % capacity utilisation represents, for practical purposes, full vessels. It is also affected by the fact that seasonal variations mean that demand is considerably greater in the third and particularly in the fourth quarters of an calendar year than in the first and second quaters (35).

(87) These factors mean that volumes and revenues would be lower on eastbound trades than on westbound trades even if the degree of capacity utilisation and freight rates was the same in both directions. Moreover, the fact that demand is significantly higher towards the end of a calendar year means that a degree of under-utilisation of capacity is inevitable in the early part of the year if lines wish to have sufficient capacity to meet demand later in the year. The tendency towards lower revenues on eastbound trades is reinforced by the fact that many goods shipped eastbound from north Europe to the Far East are of lower value than goods being carried in the other direction and are generally more sensitive to transport costs.

(88) Another factor which may help to explain apparently low eastbound revenues is the wish of lines to operate "near-similar capacity" vessels. In other words lines seek to ensure that each vessel in a string of vessels is of a similar size. The consequence of this is that the overall capacity of a string of vessels is governed not just by the regularity and speed of service desired by the line but also by the fact that vessel size will be determined by the highest level of demand anticipated for any vessel in the string. Thus, for example, capacity put in place to meet demand in November will still be in place in January when demand is likely to have slackened.

(89) A further factor affecting capacity utilisation on the eastbound north Europe/Far East trades is the very significant amount of repositioning of containers which takes place. The repositioning of empty containers from one part of the word to another (or from one region to another) results from the imbalance between trade flows so that, for example, more full containers are carried westbound from Japan than eastbound to Japan: the result is that the excess has to be carried back empty. According to figures supplied to the Commission by the EATA parties, some 18 % of all containers carried eastbound by EATA parties in 1992 ( including non-scope cargo) were empty. The corresponding figure for 1991 was 24 %.

(90) The Commission understands that at the relevant time, westbound capacity exceeded eastbound capacity by a considerable degree (36) and that this was brought about as a result of reorganisation of services to meet the imbalance between eastbound and westbound demand. This was achieved by offering eastbound round-the-world services using smaller and fewer vessels then westbound round-the-world services.

(91) Table 5 sets out capacity utilisation figures averaged across each of the EATA parties and over the four-year period 1989 to 1992.

Table 5

EATA parties' capacity utilisation 1989 to 1992

<emplacement tableau>

(92) Cumulative figures for the period January/August 1993 showing eastbound capacity utilisation excluding empty containers show that during this period the EATA parties averaged a utilisation level of 80,2 %. When non-scope cargo was taken into account this figure increased to 83,2 %. The capacity utilisation for August 1993 was 86,3 % (excluding empty containers).

(93) One of the economic advisers to the EATA parties has written that:

"As a general observation, however, conversations with ship operators induce the author to believe that round voyage break-even load factor can vary from 55 to 80 %...(37)."

(94) Furthermore, the increase in both eastbound and westbound capacity available in the four years preceding the implementation of the EATA closely matched theincrease in demand. Table 6 takes as its basis the amount of capacity offered and the volume of cargo carried by the EATA parties eastbound and westbound in 1989. In order to make a comparison with subsequent years each base figure has been converted into 100.

Table 6

Increase in capacity measured against increase in demand 1989 to 1992

<emplacement tableau>

(95) Table 6 demonstates that over the four years 1989 to 1992 taken as a whole capacity eastbound increased at the same rate as demand and that capacity westbound has also grown as a similar rate to the growth of demand. Accordingly it may be deduced from Table 6 that the argument of the EATA parties that eastbound capacity had grown in excess of eastbound demand is not substantiated.

(96) The Commission understands that demand on the eastbound leg was well in advance of expectations for the fourth quarter of 1993 and that as a result the capacity non-utilisation programme of the EATA was "temporarily" suspended (see recital 27), never to be reintroduced. The assertions of the EATA parties as to the structural nature of the alleged overcapacity on eastbound northern Europe to the Far East services (see recital 24) were accordingly unsubstantiated. In any event, the relevance of these assertions is considered further at recital 227.

(97) Moreover, the assertion that the overcapacity at that time was structural in nature is contradicted by the arguments put forward by the parties in the application for individual exemption. To show that any such overcapacity is structural in nature, the parties would have to demonstrate that it could never in its lifetime be efficiently used. However, the EATA parties argued precisely the opposite:

"Even allowing for the present degree of overcapacity ... capacity will have to grow substantially over a 10-year period."

"Taking a 10 year view, the maritime transport industry will have to meet substantial demands for additional capacity, as well as a certain level of replacement, at high new-building prices (38)."

(98) In the light of these comments, which the Commission has no reason to doubt, it is considered that the parties assertions that there existed a structural problem of overcapacity on the northern Europe/Far East trades have not been demonstrated to be well-founded.

(99) Finally, according to Drewry, the supply/demand balance on the north Europe/Far East trades looked as follows in the period 1992 to 1997.

Table 7

North Europe/Far East supply/demand balance

1992 to 1995

<emplacement tableau>

(100) The figures given for demand in Table 7 exclude military traffic and relay/transshipment cargo moved via main trade ports as well as empty containers. They therefore underestimate actual vessel utilisation. The figures given for capacity are calculated after deduction of EATA cap in 1993 and 20 % slot reduction due to deadweight limitations.

(101) Table 7 demonstrates not only continuous eastbound and westbound growth in demand but also demonstrates, as further illustrated in Table 8, that during the period in which the EATA was in operation, the increase in supply easily outstripped demand. Accordingly, in so far as there were any problems of overcapacity, it may be deduced that these would have been caused by the introduction of new capacity and not the existence of overcapacity at the time of implementation of the EATA.

Table 8

North Europe/Far East supply/demand balance

Eastbound 1992 to 1997 (million TEUs)

>PIC FILE= "L_1999193EN.003601.EPS">

Source EATA.

(102) Two further conclusions may be drawn. First, there is no obvious reason for the capacity freeze being applied only to eastbound services, that is to say exports from northern Europe. Secondly, there is no reason to believe that the capacity freeze was necessary in 1993 when it was not necessary in 1994 to 1997. Table 9 confirms these conclusions.

Table 9

Capacity utilisation 1993 to 1995

<emplacement tableau>

(103) Table 9 also responds to the argument of the EATA parties that empty containers should not be included in the figures for demand on the grounds that, on the whole, they do not produce revenue. Whatever the merits of this argument, it is indisputably the fact that empty containers reduce the amount of space available on vessels for full containers. Accordingly Table 9 calculates the amount of available capacity by subtracting from the total capacity the number of empty containers carried.

X. The notion of stability

(104) A liner shipping conference brings stability to the trades it affects by fixing a uniform tariff which serves as a reference point for the market. Prices set in this way are likely to remain unchanged for a longer period of time than if they are set by individual lines. This reduction in the price fluctuations which would be expected in a normally competitive market may benefit shippers by reducing uncertainly as to future trading conditions (39).

(105) The rate stability envisaged by Regulation (EEC) No 4056-86 has the consequential effect of assuring shippers of reliable services. Liner services are by their nature regular in the sense of an evenly spread timetable. Reliable services are those which are of a reasonable quality, such that the shippers goods come to no harm, and are at the same price irrespective of which day and which line is chosen to carry the cargo. Reliability in the supply of transport services is the maintenance over time of a scheduled service, providing shippers with the guarantee of a service suited to their needs.

(106) For these reasons, the customers of the members of liner shipping conferences are considered to obtain a fair share of the benefits arising from the restrictions of competition brought about by liner shipping conferences. Therefore, provided they remain subject to effective competition, agreement between the members of a liner shipping conference as to the rates they charge benefits from group exemption.

(107) However, the EATA parties rely on the thesis that the liner shipping market is so different from all other markets for goods and services that it must be exempt from the normal rules of competition which apply to those other markets. They argue that, "a competitive equilibrium does not exist in liner shipping" and that "if a competitive equilibrium does not exist, attempts to attain it via competition policy will be in vain" (40).

(108) The EATA parties would like the notion of stability to amount to the assurance that any particular shipping line on any particular trade should be guaranteed sufficient on its capital that its owners should not be tempted to invest that capital elsewhere.

(109) Their main argument is that since the liner shipping industry has considerable fixed but avoidable costs, the existence of reserve capacity gives rise to short-run price competition at levels close to marginal cost, resulting in the withdrawal of capacity as operators either move their vessels to more profitable trades or go bankrupt. In theory (41), this could lead to a shortage of capacity which would cause a large increase in prices drawing innew capacity and operators into the market. Capacity would then increase until an adequate level of reserve capacity was reached (a level necessary for the provision of reliable services) and the cycle would begin again.

(110) These potentially large fluctuations in price and available capacity are a form of market failure: the market is inherently unstable. In order to break out of that cycle, strict price discipline is necessary, preventing lines from offering services at levels which the advocates of this thesis regard as too low (that is to say, "destructive price competition"). That, in their view, is recognised by Regulation (EEC) No 4056-86, which grants exemption for certain price-fixing activities of liner conferences.

(111) Moreover, simple price-fixing is not sufficient; additional measures such as capacity non-utilisation agreements are necessary in order to limit the volume of goods which each line is allowed to transport, so as to eliminate the temptation for lines to transport additional goods at a price close to marginal cost. Such measures, they say, do not give conferences the power to raise prices as much as they wish since supra-competitive prices would draw hitherto potential competitors into the market.

(112) This analysis is an instance of the arguments made in many competition cases by cartel members seeking to draw a distinction between "fair" and "destructive" competition (42). In fact, most of the factors relied on in support of the thesis that there is no equilibrium in the liner shipping market are not peculiar to liner shipping. The existence of reserve capacity is common to most capital-intensive industries, where there are large sunk costs but marginal variable costs of production are low. In certain circumstances, there is an incentive to lower prices in order to increase turnover, by means of an increase in volume.

(113) The arguments of those who put forward the thesis of inherent instability are based on two very controversial theories which Pirrong(43), Sjostrom and Davies (44) in particular have sought to apply to sea transport; the core theory and the theory of contestable markets. The hypotheses necessary for the application of these theoretical models (the presence of fixed but avoidable costs, the impossibility of adjusting capacity in accordance with demand and suicidal conduct of shipowners in setting their own prices) appear unrealistic to most commentators. It is thus hardly surprising that the work of the economists who apply these heterodox theoretical models has not been accepted by most analysts of sea transport (45).

(i) The core theory and the theory of contestable markets

(114) According to the core theory (46), the maintenance of adequate reserve capacity in order to provide regular and reliable service in spite of demand fluctuations can lead shipowners to forget the raison d'être of the reserve capacity (which ought to remain unused unless there is an exceptional increase in demand at constant prices, for example seasonal demand), so that they decide to lower their prices in order to fill their reserve capacity by attracting additional customers. Such a commercial strategy aims at maximising the use of ships.

(115) The decision to lower rates leads other shipowners to enter into a price war which leads to very low price levels, close to or at the short-run marginal cost of transporting an additional container of cargo, since shipowners are not able quickly to adapt capacity to demand. It is in this way that the cycle of excessive swings in prices and service quality described above begins, and this constitutes the inherent instability of the industry.

(116) Without going into the question whether the core theory is anything more than a theory without application, it is clear in any event that the hypotheses which underlie the core theory are not applicable to the liner shipping industry.

(117) First, the concept of reserve capacity corresponds to the problem of the indivisibility of factors of production in regular transport services (47). Thus, from the point of view of the individual shipping line, the capacity costs of ships are common costs to all cargo carried and cannot be allocated to individual consignments. They are therefore not included in the calculation of what a particular freight rate should be and the charging floor becomes the direct handling costs. The view that capacity costs cannot be allocated is made more complicated by price discrimination between commodities with the result that one commodity may pay five times as much as another for the same service.

(118) However, as Jansson and Schneerson point out, there is a general problem of part-load transport markets, both for passengers and freight where, if part-loads are relatively small (for example a single container on acargo liner), a problem of indivisibility may arise leading to the conclusion that the capacity cost should not be included in the marginal cost. The extent that this ever has been a problem, it is clear that the means to minimise its negative effect on profitability now exist and are well understood by the liner shipping industry. The definition of reasonable prices at the margin is generally achieved by the use of the principles of "yield management" which are well known in the fixing of air fares. Today, the main shipping lines use these principles of marginal pricing in order to avoid any pollution of their normal rates by sales at the margin (48).

(119) The 1991 Drewry report (49) pointed out that prices at the level of marginal variable cost would be in the vicinty of USD 20 to 400 per 40-foot container (depending on whether equipment was owned or leased), which is well below any rate recorded at the time. The analysis in the Drewry report shows the absurdity of the idea put forward the proponents of the destructive competition theory that certain occasional rates on marginal shipments (at a level approaching marginal variable cost) will contaminate all rates. Shipping lines, like other businesses, realise the impossibility of making a profit on their operations by offering rates which on average are lower than average total cost.

(120) Secondly, the hypothesis of capacity fluctuation is essential to the application of the core theory. However, it is inconsistent with the existence of large sunk investments on the part of the shipping lines. Investment in, for example, container ships for each service, port and terminal facilities, and the sales and administrative structures necessary to build up a satisfactory customer base cannot be realised withour loss if a line suddenly decides to withdraw from a route. Since withdrawal from a route normally takes place at a time of difficulty, the line must expect to sell its assets at very low second-hand prices.

(121) The collapse of US Line in 1986 is a signal example of sunk costs in ships. It has been estimated that about 40 % to 50 % of the residual value of investments was not recovered in that case (50). The investments are so large that lines have no interest in withdrawing from a market as soon as prices start to slide. Consequently, potential competitors must also expect a vigorous reaction from lines already on the market in the event of any change in the competitive structure of the trade. Accordingly, the existence of significant sunk costs limits the risk of hit and run entry predicted by the core theory.

(122) Thirdly, the core theory suggests that the withdrawal of capacity results in a diminution of service quality (frequency and capacity), then a price rise which draws new competitors into the market. This hypothesis reflects the view that the services and capacities operated by lines on the market are rigidly defined, with no flexibility for adaption to demand. This hypothesis corresponds to an incorrect view of the stability of services, according to which the stability of transport means the protection of all existing services and requires protection against all competition, which is considered destructive. Such a hypothesis does not correspond to the circumstances of the main liner routes.

(123) For the above reasons, the Commission does not accept that the core theory is applicable to the study of the liner shipping industry. Moreover, in addition to the fact that the core theory does not provide a satisfactory theoretical framework, it should be observed that the specialists of that theory are not able to propose specific solutions. Dr Pirrong(51) states "Other forms of institutions may also rectify empty core problems. These include monopoly, long-term contracts and vertical integration".

(124) Accordingly, a commercial strategy based on a differentiation of service quality according to the client and the conclusion of individual service contracts makes it possible to resolve any problem of "inherent instability" in the industry. Since these commercial strategies are well-known to shipping lines, it must be concluded that the core theory, even if it were applicable, provides no justification for cartels.

(125) Shipowners also reply on the theory of contestable markets in order to argue that the existence of potential competition guarantees efficient services at competitive prices. The threat of the sudden entry of competitors subjects the lines present on the market to certain constraints of efficiency. The applicability of this theory to liner shipping is, however, disputed by many economists (52).

(126) In addition to the existence of significant sunk costs which limit profitable entry to the market, it should beobserved that a condition of the applicability of the theory of the contestable market is that there should be no entry or withdrawal, the threat of potential competition being sufficient. According to the contestability theory, it is only if potenial competition does not exert any real competitive pressure that high prices or supply-side inefficiencies may arise and attract new entrants hoping to profit from these inefficiencies.

(127) Jankowski has pointed out that a large number of incidents of entry and exit from the market, most of which are unprofitable and result in failure, suggests a lack of contestability of the market. In his 1986 study (53), Dr Davies argued that liner trade between Europe and Canada was contestable precisely because of the number of entries observed; he did not, however, investigate the profitability of these movements. The conclusions of Dr Davies concerning the contestability of the market are therefore questionable.

(128) It should be added that most if not all of the incidents of entry and withdrawal observed by Dr Davies concerned the redeployment of ships by lines present on neighbouring trades. No analysis of these redeployments was done in order to determine whether these were in fact hit-and-run entries.

(129) The concept of mobility and flexibility in the positioning of ships is accepted by most specialists and by shipowners themselves (54). The observations of Dr Davies concerning entry to and exit from the Europe-Canada trade also support the idea of flexibility in redeployment of ships rather than any real contestability of the market:

"All the above analyses of entry and exit have been made on the basis of services creation or extinction, not of companies. The movement of vessels by a firm from one trade to another must give rise to the exit and entry of a service - something which in itself will influence the competitive environment in each trade - but it may not alter the number of companies involved in the wider market environment. [Table 4] shows the turnover of companies supplying the liner services on Canadian trades, but again the same problems arise: a company that is new to Canada may not be new to the rest of the world, and an exit from Canada may imply not a death but a movement to a non-Canadian route. Indeed, of the 49 exits listed in Table 4, only six were caused by the complete bankruptcy of the company involved: the remainder were brought about by a service movement or, in the case of the further six, by take-over or merger (Abbott et al. 1984). Similarly, almost all the 60 firms which were new to Canadian trades over the period were not newly established but existing foreign corporations that were redeploying their tonnage, presumably in response to perceived profit opportunities" (55).

(130) In his Tarporley lecture, Professor Gilman has also criticised the idea of contestability and the idea of inherent instability on the main world routes in the following terms:

"However, one does not actually need empirical evidence relating to losses in individual cases to criticise the idea of low sunk costs. The basic argument comes down to a question of industry structure as this controls relationships between markets. The mainstream trades consist of three large markets, the Atlantic, the Pacific and the Europe/Far East route. The availability of capacity for entry, and the level of sunk costs related to exit, from any one of these, will clearly depend on conditions on the other two. If the three routes were in anything like equilibrium, the capacity simply would not be available for the instant and total replacement of the incumbents in any other market. As ships began to move out of the other two, rates would go up, the extent depending upon demand elasticities, and the process would quickly come to a stop. So even in the complete absence of barriers to entry and exit, the market would not be contested to the extent of total replacement.

Very large scale entry via newbuildings would also be quite impossible. To replace all of the ships on the Europe Far East route for example would cost in excess of USD 10 billion. And from the decision to build to the full capacity coming on stream would take up to five years on an optimistic view. Ship prices would also escalate, (which would mean that the potential new entrants could not obtain their ships at the price paid by the incumbents) and the attempt to build on this scale would set in train a major disequilibrium in the world shipping industry.

Concentration in global ownership of large containerships also affect to the possibility of entry. Theworld fleet is dominated by some 20 very large carriers, many of whom operate on two, or even three, of all mainstream routes, so they basically count as incumbents. The trend over the last two decades has been towards concentration and this expected by many to continue over the next 10 years. New entry into this big league is likely to be quite limited, and the pool of potential new entrants is very small indeed.

Turning to exit, opportunities even for limited redeployment of vessels would depend on market conditions in the sector targeted. If one of the three mainstream routes were over-tonnaged and the other two were strong, there would be opportunities to redeploy some vessels. However, if all routes were suffering the effect of world recession then opportunities would rather be limited. The point here is that, to the extent that markets are affected by a common set of influences, there will be relatively easy exit (for a moderate number of vessels) when market conditions are strong, and much more difficult exit when markets are weak. Thus relative ease of exit is most likely to be available when it is least likely to be of value. Very large scale redeployment would be impossible, and the surplus capacity involved would have to go into lay up.

It is clear from the above analysis that there are a number of implicit assumptions about the structure and performance of a set of related markets, behind the idea of a single market is small relative to size of the industry as whole, the second that the global set of markets is operating profitably, and the third that the ownership structure is diverse enough to provide a pool of potential new entrants. It is only in these circumstances that capacity sufficient to replace that of incumbents could quickly and easily be found and could equally be easily reabsorbed should the incumbents retaliate. For individual liner markets, like the mainstream sectors, which are large relative to global industry size, one cannot very easily envisage a low level of sunk cost except for modest incursions. To come back to current realities, incursions, of moderate scale in a market in which there is no ease of exit can result in ferocious competition, particularly as transport capacity is a perishable commodity.

Physical mobility of assets is directly related to economic mobility in some cases, including that of the transport industries. But it does not guarantee a low level of sunk costs. In other cases output is transportable, and the assets can quite happily stay in one place. A manufacturing company in a single location, but with a diversified strategy which enables it to cover a range of world markets, might well be able to switch its output (and therefore the deployment of much of its fixed capital) between markets, with ease, without moving it an inch."

(131) Even this analysis of the contestability understates the distinction between the mobility of vessels and the contestability of markets. This is because maritime companies offering multimodal transport services need to make on-shore investments in such areas as management and marketing. Such assets are considerably less mobile than vessels, if at all. In any event the sunk costs of relocating or terminating management and marketing functions can be considerable.

(132) The economic analysis of liner transport is thus a complex area of study. The present discussion leads to the conclusion that arrangements between shipping lines on rates (conferences) or capacity (consortia), or in cartels which are even more restrictive of competition, in particular those which combine rate fixing with capacity management, cannot be analysed solely and simplistically on the basis of the core theory or the contestable market theory.

(ii) Destructive competition

(133) Certain shipowners distinguish between two types of destructive competition: type A, which can exist in industries with certain characteristics, including marginal costs well below average costs, excess or unused capacity, and the presence of sunk costs; and type B, which can exist in a situation where there is no competitive equilibrium, such as that described by the core or contestable market theories.

(134) The foregoing discussion has shown that the conditions required for type B destructive competition are not met, inter alia because of the presence of sunk costs and the absence of profitable hit and run entries. As for type A destructive competition, Professor Yarrow (an economic adviser for the EATA parties) accepts that it does not require particular measures as regards the application of the rules of competition, in part because:

"Most economists with a specialism in competition policy, myself included, would not, however, regard such an outcome as itself a sufficient argument for exemption of price agreements from the general provisions of competition law. A central reason for this is that, while firms may suffer protracted accounting losses, customers benefit more than firms suffer: it is economically efficient for prices to be lower than average costs in circumstances of excess capacity."

(135) Consequently, it is impossible to distinguish that situation of "destructive competition" from normal competition. Furthermore, the theories of inherent instability disregard certain characteristics which run counter to them. For example, the possibility of redeploying ships on the main world routes as described by Professor Gilman allows supply to adapt itself to considerable fluctuations in demand in order to avoid prolonged overcapacity and find a balance between supply and demand. Dr Reitzes summed up the situation in the 1989 US Federal Trade Commission report as follows:

"Ocean liner markets fail to exhibit the high market-specific sunk costs (as opposed to firm-specific sunk investments) that are a key condition of destructive competition (there is also little evidence in support of the proposition that shipping markets are unsustainable natural monopolies vulnerable to inefficient small-scale entry). None of the empirical studies of this industry has been performed at a sufficient level of sophistication to generate useful insights into this issue. For an illustration of how this might be approached, see Evans and Heckman (1984). Ships are mobile assets that, in some circumstances, may be transferred from less profitable geographic markets in response to fluctuations in demand. The FMC report notes that carriers in certain regional markets can easily alter their port call patterns in response to changing market conditions (FMC-report, p. 165). Furthermore, carriers and shippers can negotiate long-term contracts to minimise the risks associated with uncertain demand and supply conditions (while the theory of the core stresses 'avoidable costs' rather than sunk costs, the difference does not appear important in this case). Because ships are mobile assets, and because long-term contracting is available, ocean carriers have latitude in deciding where and whether to operate their ships. That latitude suggests that carriers will operate their ships on routes least burdened by excess capacity, making destructive competition unlikely."

(136) The achievement of a balance between supply and demand through the operation of the market ensures that the interests of transport users will be taken into account in the determination of service levels and in maintaining rate stability. If however, cartel agreements make the supply of transport services more rigid, the interests of users, the efficienty of services and the stability of rates are endangered. E. Benathan and A. Walters gave the following conclusion in their first study on cooperation in the liner shipping industry:

"Pooling, as practised in the shipping industry for example, is generally regarded, however, as the most anticompetitive form of cartel organisation. Our findings are consistent with this view in so far as competitive pricing is concerned. Since it is more flexible than quota regimes, revenue pooling should give greater stability to cartels. It allows a fuller and more continuous exploitation of profit opportunities by the cartel as a whole. Hence it may avoid the disruptive crises of the more rigid quota system. If, on the other hand, a cartel is to be permitted (as in the case of shipping cartels in most maritime countries) then pooling does permit a better allocation of traffic within the cartel. It allows the expansion of efficient low-cost firms in the cartel, and promotes the contraction of the inefficient ones. It should, therefore, be less inimical to technical progress in the liner trade than simple quota systems (56)."

(137) Consequently, where the adaption of supply of services and rates to demand is prevented by cartel agreements, in particular stabilisation agreements involving a freeze on capacity use and the imposition of artificial rate discipline, the stability and efficiency of services, and thus the interests of users, are endangered. In such circumstances the offer of capacity to the market may be reduced artificially by a partial freeze of capacity use which may lead to large rate increases or at least to the maintaince of an artifical rate level which does not encourage the elimination of less efficient services and any excess capacities (that is to say, capacity in excess of a reasonable reserve which is necessary in order to provide a service which corresponds to users' needs).

(138) In relation to "stabilisation agreements" including a freeze on the use of part of capacity, the 1991 Drewry report (57) made the following remarks:

"Sub-optimal utilisation need not, of itself, be synonymous with low profitability provided that rates are kept as reasonable levels and costs are contained. Indeed the structural inevitability of over-tonnaging in an open, competitive trade will almost certainly produce a 'demand gap ratio' of between 15 % and 35 %. Clearly, an acceptable level of profitability is more likely if that gap is nearer 15 % than 35 % but, given some form of trade organisation (whether it be a conference or a stabilisation agreement) to regulate capacity and/or rates, the market can be manipulated at any reasonable level of demand gap ratio."

(139) The main risk in restrictive cartel agreements is that competition between transporters will be limited to competition on service quality. The lines present on the trade will then be drawn into a race to operate more and more ships of larger and larger capacity (the Averch-Johnson effect). The Problem of monopoly-induced waste is considered by Scherer and Ross (58):

"Price-fixing agreements, tacit oligopolistic collusion, and monopoly pricing can also stimulate the wasteful accumulation of excess capacity. There are four main mechanisms.

First, offering ample reserve capacity provides another kind of non-price rivalry advantage, for example, as travellers patronise airlines with the most flights and seats available at the last moment, or as industrial buyers favour suppliers who were able to meet their demands in unusually tight gray markets. Second, when cartel sales quotas are allocated in proportion to capacity, as they were under the US crude oil prorationing system until the early 1970s, investment in excess capacity to get a higher quota is encouraged. Third, excess capacity may be carried to strengthen the credibility of a monopolistic group's entry deterrent. And fourth, monopolistic pricing cushions the survival of capacity in secularly declining industries.

There is reason to believe that the relationship between monopoly power and certain of these propensities is non-linear. Thus, ocean shipping cartels that perfected their monopoly controls through controls over entry, investment, and scheduling were less prone toward costly excess capacity or 'overtonnaging' than the losser 'open' cartels serving US routes."

(140) As an illustration of this phenomenon, the 1991 Drewry report (59) has described in the following manner the history of trans-Atlantic trade in the 1980s:

"All the conference lines operate on a seven-day schedule in all of their various services, and this frequency is matched by all the leading independents, leaving only some of the smaller operators with lower quality schedules. In 1987 exactly half of the 46 separate services were operated at seven-day intervals, so there has been a major advance in the quality of service generally being provided. This suggests it is becoming increasingly difficult for less frequent services to secure acceptance in the market."

(141) Such an increase in capacity, which is logically accompanied by a decline in rates, cannot continue without an increase in rates to make up for the increase in product costs. Supply moves further and further away from the point of equilibrium and users pay for this race towards service quality through rates which are higher than they wish. The cycle continues until the point where market forces cannot be resisted. A return to commercial reality on the trade then causes an abrupt adjustment in rates which may cause carriers to leave the market and affect the supply of capacity. The cycle is clearly a source of instability for the market. It is not however, a problem of inherent instability or empty core, but simply the consequence of the disturbance of normal market conditions as a result of the misuse of the market power held by the shipping lines cartel(60).

(142) Furthermore, the solutions put forward in Regulation (EEC) 4056-86 to bring about the stability recognisedby that Regulation, that is to say rate stability, are not and were never intended to deal with any problems brought about by liner shipping operators as a result of uneconomic investment decisions.

XI. Complaints

(143) The British Shippers' Council (BSC) wrote to the Commission on 20 April 1993 to express its serious concern that the EATA would create the conditions to enable the EATA parties to eliminate all effective competition in the market. The BSC stressed its concern that the impact of the EATA, in combination with the impact of the Trans-Atlantic Agreement, would be to harm European exporters in their two major exporting markets, namely the United States and the Far East.

(144) The BSC also stressed that the objective of the EATA, being to maintain control over prices and capacity arrangements in the market, did not fulfil the conditions of Articel 85 (3) of the Treaty since the Agreement was not in the interests of users and did not provide any service or economic benefits to consumers. The BSC urged the Commission to reject the EATA's application for exemption.

(145) The French Conseil National des Usagers des Transports (CNUT) wrote to the Commission on 27 April 1993 to state its view that the EATA did not fulfil the conditions of Article 85 (3) for the following reasons:

(a) the purpose of the EATA was not to achieve stability but to restrict competition between the conference members and the independents,

(b) the EATA did not aim to bring about a lasting reduction in capacity and the non-utilisation of capacity meant that shippers would continue to have to meet the costs of all capacity, and

(c) users would not benefit from the EATA but would face rate increases which would harm their ability to export.

(146) The European Shippers' Council (ESC) wrote to the Commission on 28 April 1993 to express its concerns about the EATA. The ESC stated its view that in the case of the EATA, none of the four conditions of Article 85 (3) was satisfied. It stressed that the fact that many of the parties to the EATA were members of the FEFC was of major significance. The result of this was that rate increases put into effect by the FEFC would be followed not only by the EATA parties who are not members of the FEFC but also by other parties such as non-vessel operating common carriers.

(147) The Japan Shippers' Council (JSC) wrote to the Commission on 25 May 1993 to express its concern that the EATA would lessen competition at the expense of shippers and consumers. The reason given by the JSC for this view was that a capacity control agreement between parties with an 85 % market share (see recital 80 of this Decision) would lead to excessive price rises. The JSC also expressed concern that a decrease in eastbound capacity would lead to a resulting decrease in westbound capacity. The JSC also stated that since Japanese shippers are still subject to strict loyalty contracts binding them to use conference carriers exclusively (61) and are therefore unable to use independent shipping lines, any restrictions on westbound capacity would affect them acutely.

LEGAL ANALYSIS

XII. Article 85 (1)

(i) Restriction of competition

(148) For the reasons set out below, the Commission considers that the provisions of the EATA relating to capacity non-utilisation and the exchange of information fell within the scope of Article 85 (1) of the Treaty.

(149) The agreement relating to the non-utilisation of capacity (as described in recitals 8 to 24 of this Decision) and the exchange of information concluded between the parties to the EATA in respect of their maritime transport activities was an agreement between undertakings within the meaning of Article 85 (1) of the Treaty.

(150) The purpose of the EATA was to reduce price competition between the parties to the EATA by artificially limiting the liner shipping capacity made available to shippers wishing to have goods transported from north Europe to the Far East. The EATA parties acknowledge that the effect of the EATA was "to arrest the rate of decline in average freight revenue" (62).

(151) The EATA had as its object or effect the prevention, restriction or distortion of competition because it allowed the limitation or control of production within the meaning of Article 85 (1) (b). In particular it allowed the members of the EATA to restrict substantially the competitive capacity of each one of them vis-à-vis the others by limiting the volume that each one offered to the market. Because of the relationship between supply and price, the EATA also had an effect on prices.

(152) The very high market shares of the parties to the EATA (some 86 % in 1991, see recital 80 of this Decision) meant that such prevention, restriction or distortion of competition was likely to have been appreciable. The restrictive effect of the EATA was likely to have been reinforced by the provision for financial sanctions against lines which carried greater volumes of cargo than they were allowed under the term of the EATA (see recitals 12 and 28 of this Decision).

(153) The EATA must also be considered in the light of comments made concerning market structure (see recitals 66 to 79 of this Decision). That is to say, the EATA cannot be considered without also taking into account the restrictions of competition which resulted from the FEFC (the purpose of the FEFC being to maintain or increase tariffs to a level higher than they would otherwise be) and which resulted from the FETTSCA whilst it was in force.

(154) Moreover, every month the EATA secretariat supplied the EATA parties, through the Market Review Committee, which was made up of representatives of the individual lines, with the following information:

(i) maximum declared capacity in TEUs,

(ii) total actual filled slots in TEUs (on a monthly basis),

(iii) non-scope cargo in TEUs lifted,

(iv) percentage utilisation,

(v) forecast of capacity for each vessel for the following two months,

(vi) estimated monthly total for the next four months.

(155) Each of the EATA parties participated in this exchange of information from its inception in 1992 until May 1997: a number of the EATA parties continued to exchange information until July 1997. The information exchanged was not aggregated but clearly stated to which EATA party it related. Thus, every month for five years each EATA party learned exact details of its principal competitors' actual liftings of scope and non-scope cargo, their capacity, their level of capacity utilisation as well as their forecasts of non-scope cargo and capacity. The supply of this information would have ensured compliance with any decision on capacity non-utilisation.

(156) During the course of the administrative procedure in this case, the EATA parties claimed confidentiality vis-à-vis the complainants and other third parties in respect of their individual levels of capacity and capacity utilisation. It is clear that the reason for this was that they did not wish their customers to have access to information which was of a commercially sensitive nature with the potential to have an effect on prices. Information which is commercially sensitive vis-à-vis customers is also likely to be commercially sensitive vis-à-vis competitors. The sensitive nature of the information confirms the anticompetitive context in which the exchange took place.

(ii) Effect on trade between Member States

(157) The EATA parties argued in their application for individual exemption (63) that it was "wholly improbable" that the EATA would have any appreciable effect on trade between the Community and the third countries covered by the relevant trades because of the insignificance of increases in freight rates in comparison with the overall delivered price in the third countries concerned of the goods carried.

(158) The EATA parties also consider that "it is axiomatic that any effect on trade between Member States, on the one hand, and a third state, on the other, is not an effect on trade between Member States for the purposes of Article 85 (1) and must therefore be disregarded for the present purpose" (64). They concluded that the EATA consequently fell wholly outside the scope of Article 85 (1).

(159) According to the case-law of the Court of Justice, the test of effect on trade between Member States is met whenever it is possible to foresee with a sufficient degree of probability, on the basis of a set of objective factors of law or fact, that the agreement or concerted practice in question may have an influence, direct or indirect, actual or potential, on the pattern of trade in goods or services between Member States (65).

(160) It is not necessary to make a finding that inter-State trade is actually being affected at the present time. The condition concerning the effect on trade must be deemed to the fulfilled where it is established that intra-Community trade has actually been affected or that it is potentially affected to a significant degree (66).

(161) In considering whether the EATA was capable of affecting trade between Member States, it must be emphasised that the relevant markets which were directly affected relate to the provision of transport and intermediary services and not the export of goods to third countries (67).

(162) It is well established that arrangements to share markets or to set prices (including target prices) or allocate quotas between firms from different Member States has an effect on inter-State trade because such agreements, in addition to affecting the structure of competition in the Community, establish a form of private regulation that runs counter to the Treaty's objective of eliminating customs duties, quantitative restrictions on imports and exports and all other measures having equivalent effect (68). An effect on inter-State trade may be established when an agreement has the effect of partitioning national markets within the common market (69).

(163) The EATA parties make the argument in the Reply to the Statement of Objections (Annex 12 at paragraph 6) that, because Regulation (EEC) No 4056-86 does not apply to inland transport, the services offered by the EATA parties do not concern goods physically passing between Member States. This argument is wrong since it confuses the legal question with the factual question of whether a reduction in competition between undertakings in different countries is likely to affect trade between Member States.

(164) The Commission considers that the EATA was capable of appreciably affecting trade between Member States in the following ways.

(165) The EATA involved shipping lines operating in at least seven Member States and restricted competition between such lines in respect of the services each of them offered and the price charged. The restrictions placed on the use of capacity were restrictions on the service offered and were put in the place for the purpose of reducing price competition.

(166) The elimination or diminution of price and service competition between these companies was likely to reduce significantly the advantages which would accrue to the more efficient of them. This was likely to affect in turn the normal pattern of losses and gains of market share which would have been expected in the absence of the EATA. This restriction of competition between shipowners operating in many Member States consequently influenced and altered trade flows in transport services within the Community, which would have been different in the absence of the EATA.

(167) These changes in the normal pattern of competitive behaviour by which more efficient companies are likely to enjoy increases in market share may also have influenced competition between ports in different Member States, by artificially extending or diminishing their catchment areas (70), and the market shares of shipping lines operating out of those ports. In particular, shipping lines operating out of more efficient ports would have been unable to pass on to their customers cost savings resulting from improvements made in port efficiency. The effect that the EATA had on the normal play of competitive forces may also have changed the capacity available at each port and caused deflections of trade between points in Europe and ports in northern Europe from some ports to other ports, thereby being capable of having affected trade between Member States.

(168) The effect on the supply of maritime transport services described in the preceding paragraphs is likely to have had a consequential effect on the supply of services ancillary to the supply of maritime transport services. Such services include the services of freight forwarders, port services, land transport services, and stevedoring services. The effect on these services would principally have been brought about by the alteration in the flow of transport services between Member States.

(169) The Commission thus considers that the EATA affected trade between Member States in relation to the supply of maritime transport services and the supply of services ancillary to the supply of maritime transport services.

(170) By limiting the capacity offered by each party to the EATA in the eastbound direction and by intending to maintain or increase freight rates, the EATA may have reduced (or restrained increases in) the flow of traffic between Europe and the Far East and by doing so may have affected trade between the Member States. This arose in part from the fact that some products exported form northern Europe to the Far East are especially sensitive to increases in freight rates and also in partfrom the very high market shares of the EATA parties which reduced the availability of competing transport services.

(171) Where there are goods which would have been exported from a Member State to a third country in the absence of the EATA but which, as a result of increased transport prices, were sold from that Member State into other Member States, the competitive position of intra-Community exporters already selling their goods in those other Member States may have been affected.

(172) Accordingly the restrictions concerning transport services had an indirect effect on trade in goods between Member States. This effect was likely to have been particularly noticeable in those Member States where the EATA parties had an especially high market share.

(173) It has been held that a cartel agreement which fixed the price of a semi-finished product (eau de vie) which was not itself normally exported but which constituted the raw material of another product which was exported throughout the Community did affect trade the finished goods between Member States (71). The Commission considers that an agreement such as the EATA which was intended to affect the price of transport services for exported goods may likewise have affected trade in those goods between Member States.

(174) This is in line with the judgment of the Court of Justice in the Commercial Solvents case where it was pointed out that the expression in Article 86 of the Treaty containing the requirement to demonstrate an effect on trade between Member States:"... is intended to define the sphere of application of Community rules in relation to national laws. It cannot therefore be interpreted as limiting the field of application of the prohibition which it contains to industrial and commercial activities supplying the Member States" (72).

(175) The Commission therefore considers that the EATA also had an indirect effect on trade in goods between Member States since it may have had an effect on goods exported from Member States to third countries.

(iii) Conclusion in relation to Article 85 (1)

(176) The agreement relating to the non-utilisation of capacity and the exchange of information between the parties to the EATA in respect of their maritime transport activities was an agreement which restricted competition within the meaning of Article 85 (1) of the Treaty.

XIII. Article 85 (3)

(i) Article 3 of Regulation (EEC) 4056-86

(177) Article 3 of Regulation (EEC) No 4056-86 grants exemption from the prohibition in Article 85 (1) of the Treaty to the members of a liner conference in respect of the fixing of uniform or common freight rates and any other agreed conditions with respect to the provision of scheduled maritime transport services. It also grants exemption to a limited number of other activities if one or more of these is carried on by the members of a liner conference in addition to fixing prices and conditions of carriage for maritime transport services.

(178) In the Decision on the Trans-Atlantic Agreement (73), the Commission explained that the regulation of capacity within the meaning of Article 3 (d) of Regulation (EEC) No 4056-86 has always been understood to allow:

" (i) capacity adjustments to facilitate the organisation of conference members' sailings and calls, in order to improve the regularity, reliability and frequency of services to all the ports served, and

(ii) capacity adjustments to take account of seasonal (or short-term) fluctuations in demand ...."

(179) The Commission added that:

"Capacity regulation is exempted by Article 3 (d) where it consists of temporary or short-term adjustments in the amount of physical capacity available, such as the withdrawal of a vessel or a reduction in the frequency of a service to meet a seasonal reduction in demand. Article 3 (d) does not exempt capacity non-utilisation agreements because their only effect is to raise the level of prices and they do not involve any improvement of the services offered".

(180) Under the group exemption, conferences are allowed both to operate under a common or uniform tariff and to engage in a number of other specified restrictions of competition including the regulation of the carrying capacity offered by each member of the conference. The EATA did not constitute an agreement or arrangement within the framework of which the parties to the EATA operated under uniform or common rates. The EATA was accordingly not a liner conference within the meaning of Article 1 (3) (b) of Regulation (EEC) No 4056-86. The block exemption contained in Article 3 of Regulation (EEC) No 4056-86 was not applicable since it only applies to liner conferences as defined in Article 1 (3) (b).

(ii) Conditions for grant of individual exemption

(181) In considering whether the conditions for exemption are fulfilled in respect of any agreement between undertakings, decision by association of undertakings or concerted practice, the Commission must fully take into account the fact that it is one of the fundamental principles of Community law as laid down in the Treaty that the Member States and the Community are to act in accordance with the principle of an open market economy with free competition, favouring an efficient allocation of resources, and in compliance with the principles set out in Article 3a of the Treaty.

(182) Notwithstanding the abandonment of the EATA, it is in the Community interest to consider whether that agreement fulfilled the conditions for individual exemption for the following reasons.

(183) First, the parties to the EATA, which have continued to argue that the conditions for exemption are fulfilled, are likely to benefit from the increased legal certainty arising from a formal Commission decision concerning the practices in question, particularly as a large number of them have been the addressees of Commission decisions in the past relating to findings of infringement of Community competition law.

(184) Secondly, Regulation (EEC) No 4056-86 provides that parties do not need to notify an agreement or arrangement for it to be granted exemption by the Commission: other liner shipping companies may therefore benefit from the increased legal certainty arising from a formal Commission decision concerning the practices in question.

(185) Thirdly, national courts and authorities may benefit from a clear statement of the Commission's position in the event that any third party seeks to obtain redress under national law for any harm it has suffered as a result of the practices in question.

(186) Finally, in view of the practice of the Commission of increasing the penalties imposed in the case of recidivist infringement of Community competition law, it is important that a formal decision be adopted in this case for the purposes of future enforcement action.

(187) The Court of First Instance has ruled that: "it should not be overlooked that whenever an exemption under Article 85 (3) of the Treaty is sought, it is incumbent on the applicant undertaking to prove that it satisfies each of the four conditions laid down therein, and to set out in the A/B application form its position on each of those conditions .... It should also be recalled that, owing to the cumulative nature of the conditions required, the Commission is entitled at any time before the definitive adoption of the decision to find that any one of the four conditions is not satisfied" (74).

In the light of the Commission's finding that the first condition of Article 85 (3) is not fulfilled, it is not strictly necessary to examine whether the other three conditions were fulfilled. However the Commission has undertaken this exercise in the interest of providing greater legal clarity and in order to address the arguments raised by the EATA parties.

(188) Although a high market share does not automatically rule out the possibility of a grant of an individual exemption, the burden of demonstrating that the four conditions of Article 85 (3) are satisfied is an important consideration in the case of the EATA given that the EATA parties had a combined market share of some 86 % (see recital 80) in the year immediately preceding the application for exemption.

(189) The parties argued that the EATA was necessary to maintain the structure of the industry on the trades in question. As a result of the acquisition of considerable new capacity in the mid-1980s coupled with subsequent decline in demand, there was said to be a large degree of structural over-capacity (75).

(190) It was said that unless the decline in freight rates was reversed, the severe competition resulting from the alleged imbalance between supply and demand would result in failure and a contraction in the industry. Over the medium to long term, world trade would expand to absorb the then current capacity and the large amount of new capacity then on order. However, if the industry contracted at that time to match current demand at current prices, it would have been unable to expand as rapidly as necesssary to meet anticipated demand and there might have been problems of undercapacity with a consequential effect on freight rates.

(191) According to the parties, while the conference system still provided certain benefits to shippers, it could no longer provide the necessary trade and price stability. The EATA provided this trade and price stability and helped to maintain a viable industry which continued to invest, to make a smoother adjustment to the cost profiles that it would face in the mid 1990s and to offer adequate efficient scheduled maritime transport services.

(192) The parties argued that the EATA did not directly control rates and would cause no permanent reduction in capacity which would also reduce westbound capacity and affect FDW (fixed day, weekly) and RTW(round-the-world) services. It would restrict eastbound capacity in a way which would not simply lead to a transfer of market share to independent shipping lines.

(193) The parties claimed that since the arrangements did not include pricing, or operations and service quality, they would not eliminate competition, in spite of the fact that they covered a substantial part of the trade.

(a) Improvement of production or distribution or promotion of technical or economic progress

(194) Concerning the first condition of Article 85 (3), the EATA parties have argued that:

"A healthy industry would need to respond (to the needs of the second half of the 1990s) with an active investment programme, meeting the needs and anticipated needs for more capacity with new orders, and allowing for an acceleration in the rate of scrapping in order to accommodate new technological requirements. An impoverised industry would become much more risk-adverse and inclined to make do with existing capital stock" (76).

(195) The EATA parties argued in the application that by countering the effects of overcapacity and low rates of return, the EATA would enable them to continue to invest in new capacity, new containers and technological advances such as EDI (electronic data interchange). In particular, it was said that the successful operation of the EATA would enable the parties to order new tonnage of the largest and most efficient variety (77).

(196) In the reply to the Statement of Objections, the parties altered their argument to one based on stability and the notion of destructive price competition, arguing that destructive price competition needed to be prevented in order to allow the parties to make a sufficient return on capital to contemplate making further investments on the north Europe/Far East trades.

(197) The reasons for which the Commission does not accept the parties' arguments as to stability and destructive price competition are set out in recitals 104 to 142. So far as investments are concerned, it is true for all industries, and not just shipping, that the money available for investment is always potentially increased where prices and revenues are artificially increases by price-fixing arrangement. The fact that companies are making profits does not necessarily lead to investment and certainly not to any particular kind of investment (78).

(198) Instead of attempting to resolve any substantive problems of overcapacity, whether structural or temporary, the EATA allowed shipowners to maintain their capacity on the northern Europe/Far East trades at a level higher than was necessary to meet demand. The parties produced no evidence that the EATA would help to ensure that in the long-term the level of capacity would be better adjusted to meet the level of demand.

(199) Furthermore, the temporary freezing of capacity does not encourage the removal of older capacities on a real and lasting basis but principally brings about an increase in freight rates by temporarily reducing the supply of capacity to the market. The Commission considers that such an approach could not address the long-term structural problems of the industry which the parties alleged to exist in the application for exemption.

(200) It is also possible that rather than encouraging the introduction of new technology, the restrictions on competition resulting from the EATA stifled such introduction by reducing the competitive advantages which would otherwise accrue. This situation arises from the fact that the diminution or elimination of price competition which was the purpose of the EATA was likely to have prevented shipping lines from passing on cost savings resulting from new technologies to their customers. Equally the fact that more efficient lines were less likely to benefit from their efficiencies and were less likely to increase market share as a result means that efficient lines were less likely to invest in new technologies.

(201) Finally, there is no evidence that the services operated by the EATA parties during the period in which the EATA was in operation were in any way impoved by the agreements not to use part of the parties' capacity and to exchange information.

(202) Furthermore, the parties have argued that the sole purpose of the agreement to exchange information was to enable them to put into effect the capacity non-utilisation agreement and have put forward no reasons why this agreement could have contributed on its own to improving technical or economic progress or to improving production or distribution of maritime transport services. Since the capacity non-utilisation agreement did not have this effect, it may be inferred that neither did the agreement to exchange information.

(203) For all of these reasons, the Commission considers that the capacity non-utilisation agreement and the exchange of information between the EATA parties did not contribute to improving technical or economic progress or to improving production or distribution of maritime transport services.

(b) The question whether consumers receive a fair share of the resulting benefits

(204) The EATA parties stated that:"... capacity management agreements such as the EATA would provide support to the existing structure since they would tend to have an upward effect on rates which currently do not provide an adequate return on capital" (79).

(205) The parties considered that any increase in rates which came about as a result of the operation of the EATA would benefit shippers because it would allow the carriers involved to make a smoother adjustment to future cost profiles, to continue with new investments and to provide an adequate efficient scheduled maritime transport service.

(206) At paragraph 4.3.3 of their application for individual exemption, the EATA parties claimed that:

"... any increase in ocean freight rates resulting from the EATA will be insignificant in comparison with the overall delivered price, in the third countries concerned, of the goods carried."

(207) In the reply to the Statement of Objections, the parties again focused on what they consider to be stability and argued that the benefit shippers derive from stability is sufficient for the second condition of Article 85 (3) to be fulfilled.

(208) As discussed at recitals 8 and 9, the purpose of the capacity management programme was to alter the alleged imbalance between supply and demand in the eastbound leg of the trades in question so as to allow the EATA parties to increase, collectively and individually, their freight rates. This is a benefit to shipowners and not to shippers.

(209) In the short term, the purpose of the EATA was to raise prices and freight rates, as is demonstrated by the increases announced with effect from 1 April 1993, 1 July 1993 and 1 January 1994 (see recitals 82 to 85). These increases were directly contrary to the interests of shippers, who were obliged to reflect them in their selling prices or their margings without benefiting from any advantage in terms of frequency, regularity or reliability of service.

(210) The EATA prevented the use of part of existing capacity for certain cargoes (i.e. those falling within the geographic scope of the EATA) but did not eliminate it. This action did not reduce transport costs and made clients carry the burden of unutilised capacity.

(211) The proportion of fixed operating costs (capital, labour, energy and insurance) to variable costs is very high; this is especially true of capital costs which are, according to the EATA parties, a particularly significant element of overall costs. The capacity management programme withdrew capacity from the market which over a full year would have been the equivalent of almost three and a half 4 000 TEU vessels sailing empty. As described at recital 37, the fixed costs of operating this amount of capacity would have been very significant.

(212) The direct effect of an artificial reduction in capacity utilisation (as opposed to a permanent reduction in capacity) is to share fixed operating costs amongst a smaller number of containers. The EATA had no effect in reducing fixed operating costs. A reduction in capacity could benefit shippers if the cost of transportwas reduced, i.e. if capacity was really withdrawn from the northern Europe/Far East trade by the progressive withdrawal of certain vessels or certain operator currently present.

(213) The Commission has received complaints about and objections to the EATA from bodies representing a significant number of consumers of maritime transport services. It has received no indications from shippers that they consider the EATA to have been of benefit to them. There is no evidence that conditions on the north Europe/Far East trades prior to the implementation of the EATA were such that shippers were in danger of not having access to the reliable services which Regulation (EEC) No 4056-86 recognises to be a benefit to them.

(214) An agreement such as the EATA and in particular the capacity non-utilisation agreement and exchange of information, the purpose of which was an increase in freight rates with no corresponding increase in service quality, cannot be regarded as allowing consumers a reasonable share of the benefit. Accordingly, the Commission considers that the EATA did not allow a fair share of benefits which it brought about to be passed to consumers.

(c) Indispensability of restrictions of competition

(215) The parties argued in the application for exemption that the EATA was the loosest and least restrictive cooperative arrangement that could be made to deal with the prevailing conditions of overcapacity and low rates. It was also claimed that the combination of overcapacity and depressed freight rates leading to uneconomic rates of return constituted a "threat to the future financial viability and stability of scheduled liner services from northern Europe to Asia" (80).

(216) At paragraph 6.1.10 of their application for exemption, the EATA parties argued that only the EATA could provide the stability envisaged in Regulation (EEC) No 4056-86. This appears to have been because the market power of the FEFC was said to be no longer sufficient to regulate the use of capacity through the conference structure. Consequently restrictions on use by the members of the conference alone would have resulted in a loss of market share to non-conference members. The EATA parties also argued(81) that the EATA was considerable less restrictive of competiton than traditional liner conferences.

(217) The EATA parties have also argued that if a conference which fixes prices and regulates capacity fulfils the conditions of Article 85 (3)of the Treaty, it follows that an agreement such as the EATA which only regulates capacity must necessarily fulfil the conditions for exemption.

(218) The Commission considers that the EATA must be considered in the context of the market in which it operated in cosidering the question in indispensability. To this end, the Commission regards as significant the facts that the majority of the parties to the EATA were members of the FEFC and that all but one of them were parties to the FETTSCA because of the cumulative effect of the restrictions of competition arising under the three agreements.

(219) As explained at 66 to 79, the likely effect of the EATA was determined to a very considerable extent by the structure of the market in which it operates. The combination of the two types of restrictive agreement, restrictions on supply arising under the EATA and restrictions on price arising under the FEFC, is likely to have had an effect that is highly restrictive of competition.

(220) The assertion that the EATA was less restrictive than a traditional liner conference is untrue in the context in which the EATA was operating: this conclusion could only be reached if the combined effect of the EATA, FEFC and FETTSCA was ignored.

(221) In any event, it should be noted that the Commission does not accept that an agreement between conference and non-conference shipping lines in a particular trade is necessarily less restrictive of competition than a conference agreement. This arises from the fact that a more flexible kind of arrangement than a conference agreement may be intended to extend the market power of conference members by making it possible for shipping lines which do not wish to operate under common of uniform rates to become party (82).

(222) The group extension for liner conference agreements contained in Article 3 of Regulation (EEC) No 4056-86 is granted on the basis that conferences are subject toreal or potential effective competition. The stability envisaged by the Regulation therefore be capable of existing under conditions where there is effective competition. Consequently, the Commission does not accept that the existence of competition precludes that kind of stability. Indeed, if that kind of stability could only exist where there was an absence of effective competition, then the group exemption for liner conference agreements could not fulfil the conditions of Article 85 (3) of the Treaty.

(223) The stability envisaged by Regulation (EEC) No 4056-86 assures shippers of reliable services. Reliable services are those which are regular, in the sense of an evenly spread timetable, of a reasonsable quality, such that the shippers' goods come to no harm, and at the same price irrespective of which day and which line is chosen to carry the cargo. A conference also brings stability to the trades it affects by fixing a uniform tariff which serves as a reference point for the market.

(224) In the Commission's notice in the Irish Club Rules case (83), it is stated that during discussions with the representatives of the parties to the agreement in question, the Commission informed those representatives that: "the Irish Club Rules could not be exempted on the basis that they had a stabilising effect, which was aknowledged as a sufficient condition by Regulation (EEC) No 4056-86 only for liner conferences; an agreement of this type could not produce the stabilising effect referred to in the eighth recital to Regulation (EEC) No 4056-86 and which results principally from cooperation between shipping companies on freight rates leading to the adoption of a common tariff (a feature absent in the case at issue)".

(225) The Commission considers that scheduled and reliable maritime transport services would have continued to be provided with or without the existence of the EATA.

(226) So far as the argument that almost the entire trade had to be cartelised in order to avoid a loss of market share from the conference members to non-conference lines, it must be emphasised that consumer choice plays a fundamental role in ensuring that markets operate efficently. To the extent that the EATA was intended to deprive shippers of the choice between the members of the conference and an independent carrier, it must be concluded that its object was to prevent the efficient division of market share between conference members and non-conference lines that would have been more likely to result from the free play of supply and demand than from the EATA. Even if this was indispensable as regard the purposes of the EATA, it would certainly not have been in the interests of consumers nor have fulfilled the other conditions of Article 85 (3).

(227) In any event, the EATA parties have not demonstrated that at all relevant times overcapacity actually existed on the trades covered by the EATA. If the problems which the EATA parties alleged existed did not actually exist, then the restrictions of competition that the EATA entailed could not be indispensable to meet those unproven problems.

(228) Even if overcapacity existed (see recitals 86 to 93) and was causing shipowners considerable losses as alleged, it has not been proved that the quality of service was seriously threatened.

(229) Capacity agreements between conference and non-conference members which limit the supply of a service cannot be considered indispensable to achieve the objective of stability within the meaning of Regulation (EEC) No 4056-86 when such agreements operate in combination with direct price-fixing agreements as is the case with the EATA.

(230) Accordingly, the Commission considers that the restrictions of competition arising from the capacity non-utilisation agreement and the exchange of information did not fulfill the third condition of Article 85 (3).

(d) Elimination of competition for a substantial part of the services in question

(231) The parties claimed that since the EATA did not include pricing, or operations and service quality, they would not eliminate competition, in spite of the fact that they covered a substantial part of the trade.

(232) This assertion fails to recognise that the non-utilisation of capacity under the EATA was intended to have an effect on prices by reducing the supply of transport services to the market. This must be viewed, once again, in combination with the price restrictions under the FEFC and in the light of the very high market shares of the parties to the EATA.

(233) It is clear from Regulation (EEC) No 4056-86 that external competition to conferences is an essential factor so far as the granting of the group exemption is concerned. An agreement such as the EATA which enables the members of a conference to extend their market power by engaging in anticompetitive behaviour in common with shipping lines which are not members of the conference is, in general, intended to eliminate competition for a substantial part of the services provided in the market in question.

(234) In the present case, although is likely to have been the intention of the EATA parties, the fact that the capacity non-utilisation programme was only put into effect for a relatively short period of time and the fact that the first three conditions of Article 85 (3) of the Treaty are not, in any event, fulfilled, leads to the conclusion that it is not necessary for the Commission to adopt a formal position as to the fourth condition of Article 85 (3).

(235) Finally, the EATA parties argue that:

"Since it is clear that a liner conference comprising all the members of EATA would qualify for the block exemption under Regulation (EEC) No 4056-86, the EATA parties can only conclude that an agreement with less extensive cooperation between the parties would also satisfy the fourth condition of Article 85 (3) (84)."

(236) The Commission does not accept this argument, which contains at least two flaws. First, it is tantamount to arguing that the absence of a market share limit in the group exemption for liner conferences effectively disapplies the fourth condition of Article 85 (3) of the Treaty: this is not a reasonable interpretation of the Regulation. Secondly, it ignores Article 7 of Regulation (EEC) No 4056-86 which expressly places an obligation on the Commission to withdraw the benefit of the group exemption in circumstances where effective competition is absent.

(iii) Conclusion as to the applicability of Article 85 (3)

(237) Examined in its full economic context, and, in particular, in the light of the other restrictions of competition in which the EATA parties were engaging, the Commission finds that the EATA did not meet the conditions for exemption set out in Article 85 (3) of the Treaty.

XIV. Conclusions

(238) The EATA fell within the scope of the prohibition contained in Article 85 (1) of the Treaty of agreements between undertakings which may affect trade between Member States and which have as their object or effect the prevention, restriction or distortion of competition within the common market. It did not satisfy the conditions of Article 85 (3).

(239) The duration of the infringemnt was from September 1992 to May 1997. CGM ceased to participate in the infringement with effect from October 1994,

HAS ADOPTED THIS DECISION:

Article 1

The agreement between the following former members of the Europe Asia Trades Agreement (EATA) relating to the non-utilisation of capacity and the exchange of information constituted an infringement of Article 85 (1) of the Treaty:

- CGM SA,

- Hapag-Lloyd Container Linie GmbH,

- Kawasaki Kisen Kaisha Ltd,

- A.P. Møller - Maersk Line,

- Malaysian International Shipping Corporation Bhd,

- Mitsui O.S.K. Lines Ltd,

- Neptune Orient Lines Ltd,

- Nippon Yusen Kaisha,

- Oriental Overseas Container Line,

- P & O Nedlloyd Container Line Limited,

- Cho Yang Shipping Co. Ltd,

- DSR-Senator Linie GmbH,

- Evergreen Marine Corp. (Taiwan) Ltd,

- Hanjin Shipping Co. Ltd,

- Hyundai Merchant Marine Co. Ltd,

- Yangming Marine Transport Corp.

Article 2

The application for a declaration that Article 85 (1) of the Treaty is inapplicable to the EATA is refused.

Article 3

Each of the undertakings referred to in Article 1 is hereby required to abstain from any similar agreement or practice in the future which has the object or is capable of having a similar effect to the infringement referred to in Article 1.

Article 4

This Decision is addressed to the undertakings listed in Annex I.

(1) OJ L 378, 31.12.1986, p. 4.

(2) OJ L 376, 31.12.1988, p. 1.

(3) Regulation since repealed and replaced by Commission Regulation (EC) No 2842-98 (OJ L 354, 30.12.1998, p. 18) and Commission Regulation (EC) No 2843-98 (OJ L 354, 30.12.1998, p. 22).

(4) CGM SA is the legal successor to CGM Orient SA. DSR-Senator Linie GmbH is the legal successor to Deutsche Seereederei Rostock GmbH and Senator Linie GmbH. In 1996, CGM was purchased by Compagnie Maritime d'Affretement. In 1997, Nedlloyd Lijnen BV and P& O Containers Limited merged to form P& ONedlloyd, Hapag Lloyd AG transferred its containerised liner shipping activities to Hapag Lloyd Container Linie GmbH and Hanjin acquired control of DSR/Senator.

(5) OJ C 97, 6.4.1993, p. 2.

(6)

"WHEREAS:

(a) The Parties operate scheduled international maritime transport services between North Europe and Asia;

(b) the economic and trading environment in the Eastbound North Europe to Asian Trades has resulted in a combination of overcapacity and consistently depressed freight rates;

(c) the economic imbalances described in recital (b) above have led to consistently uneconomic rates of return on investment for the parties ...

(d) the said economic imbalances and their consequences as described in recitals (b) and (c) above are continuing to deteriorate, leading to a threat to the future financial viability and stability of scheduled liner services from North Europe to Asia;

(e) ...".

(7) TEU/FEU is the industry standard abbreviation for (

"20-foot equivalent unit/40-foot equivalent unit") and refers to the size of the containers.

(8) Non-scope Cargo is cargo lifted from ports outside the geographical scope of the EATA.

(9) See paragraph 4.2.7 of the application for exemption. See also the comments of the Japan Shipper's Council at recital 147 of this Decision.

(10) See application for exemption at paragraph 1.5.

(11) Members: APL, Evergreen, Hanjin, Hapag-Lloyd, Hyundai, Maersk, Mitsui OSK, Nedlloyd, NOL, NYK, OOCL, P& OCL, Sea-land, Yangming.

(12) Drewry, Global container markets, London 1996, p. 72.

(13) OJ L 376, 31.12.1994, p. 1.

(14) See reply to the Statement of Objections at paragraph 2.51.

(15) Drewry, Global container markets, p. 73.

(16) Drewry, Global container markets, p. 162.

(17) Case C-333/94 P, Tetra Pak International v. Commission [1996] ECR I-5951, at paragraphs 13, 14 and 15.

(18) See the definition of "tramp vessel services" in Article 1 (3) (a) of Regulation (EEC) No 4056-86, which emphasises the freely negotiated nature of tramp service freight rates

(19) It should be noted that containers generally measure 20 feet of 40 feet by 8 feet on the outside and that vessel slots are sized to match. Accordingly cargoes which are larger in width or length than these dimensions use up slots in an inefficient manner.

(20) Such oneway substitutability is not restricted to shipping: for example, although soft drinks are not a substitute for bottled waters, it is not necessarily the case that bottled waters are not a substitute for soft drinks - see Commission Decision 92/553/EEC in Case No IV/M.190 - Nestlé/Perrier (OJ L 356, 5.12.1992, p. 1).

(21) Drewry Global container markets, pp. 38 to 48.

(22) According to Mats Jansson, President of Unicool and Cool Carriers, "The reefer container capacity deployed is still limited and the negative impact so far on the demand for specialised reefers is small" (Fairplay, 3 July 1997).

(23) Drewry, Global container markets, pp. 69 and 71.

(24) Drewry, Global container markets, p. 70.

(25) Drewry, Global container markets, p. 76.

(26) EACBen was a party to all three agreements. DSR/Senator and Yangming became members of the FEFC in 1996. Hyundai became member of the FEFC in 1998.

(27) The FEFC includes the Europe/Japan and Japan/Europe Freight Conferences, Hong Kong/Europe Freight Conference, Philippines/Europe Conference and Sabah, Brunei and Sarawak Freight Conference.

(28) Article 3 of Regulation (EEC) No 4056-86

(29) See letter from the EATA secretary to the Commission dated 19 October 1993.

(30) See letter from CNUT to the Commission dated 27 September 1991.

(31) See letter from the FETTCSA secretariat to the Commission dated 19 October 1993.

(32) "It is generally accepted that the FEFC tariff, in the Europe to Asia trades acts as an industry reference point, and independents compete aggressively against the FEFC by offering rates, which are set at a discount from the FEFC tariff". Per FEFC reply to the statement of objections in Case IV/33.218 - DSVK/FEFC, 31 March 1993 at p. 23. That case led to Commission Decision 94-985-EC (OJ L 378, 31.12.1994, p. 17).

(33) Drewry, Global container markets, p. 110.

(34) See letter from the FETTSCA secretariat to the Commission dated 19 October 1993 referred to in note 31.

(35) Drewry, Container market profitability to 1997, London, December 1992, p. 87.

(36) According to figures supplied to the "The American shipper" by NYK and published in the August 1993 edition, westbound capacity (2,42 million TEU) at that time exceeded eastbound capacity (1,75 million TEU) by some 38 %.

(37) 'An analysis of cost and supply conditions in the liner shipping industry', J. E. Davies, The Journal of Industrial Economics, June 1983, at p. 425.

(38) See application for individual exemption at paragraphs 2.5.46 and 2.5.47.

(39) See the Commission's proposal for a Council Regulation laying down detailed rules for the application of Articles 85 and 86 of the Treaty to maritime transport (OJ C 282, 5.11.1981, p. 4) and the accompanying memorandum. COM (81) 423 final.

(40) See reply to the Statement of Objections at paragraph 1.6.

(41) No examples have ever been put forward by the EATA parties of shortages of capacity arising from the withdrawal of vessels from the trade.

(42) See for example the judgement in Case T-29/92 SPO v. Commission [1995] ECR II-289, at paragraph 294, where the Court of First Instance of the European Communities considered that no distinction could be made between normal and destructive competition. The appeal was rejected as manifestly inadmissible by Order of the Court of Justice in Case C-137/95 P [1996] ECR I-1611.

(43) Dr Pirrong, Core theory and liner shipping markets, Journal of law and economics, 1992, footnote on p. 11.

(44) Competition, contestability and the liner shipping industry, 1986, p. 310.

(45) See for example, Jankowski in "The development of liner shipping conferences", International Journal of transport economics, October 1989; Jansson and Schneerson in their book Liner shipping economics, 1987; the report of the US Federal Trade Commission, "An analysis of the maritime industry and the effects of the 1984 Shipping Act"; November 1984, the analysis of the US Department of Justice, "Analysis of the impact of the Shipping Act of 1984", March 1990; the lecture given by Professor S. Gilman at Tarporley in February 1994.

(46) Proponents of the theory of the empty core argue that a competitive equilibrium is only sustainable if there is no coalition of consumers and producers (actual or potential) that can be made better off by deviating from the proposed equilibrium and forming its own agreement. According to the theory, any outcome where firms have excess capacity, or wehre the market would not clear unless firms operate at output levels that exceed their minimum average costs, could potentially produce incentives for opportunistic behaviour by a coalition consisting of an existing firm, or a potential entrant, and a group of consumers. This opportunistic behaviour arises as firms try to eliminate excess capacity, or as potential entrants lure customers away from existing firms that offer prices in excess of average cost. In such circumstances, firms will exit the industry rather than engage in price competition that, for any significant period of time, leads to prices that are below the level consistent with the recovery of all costs. Thus, the market is constantly unstable as firms continually enter and leave the market in response to existing price conditions.

(47) See for example Jansson and Schneerson, see footnote 45, at Chapter 10.2: "Ccommon costs and indivisibility".

(48) See the article "Sea-Land's computer wars" in Containerisation international, August 1995, and the article "Market share isn't everything" in American shipper, July 1995; see also Drewry "Strategy and Profitability" in Global Container Shipping, London 1991, p. 104-106. For example, the article in American shipper states in relation to Atlantic Container Line: "the company developed a contribution model which works off equipment flows. The model serves as a cargo-acceptance guideline. As a result sales staff are much more aware these days of the overall value of a business prospect". See also "The liner industry: structural changes and future outlook", Industrial Bank of Japan Quarterly Survey (1995 IV) p. 43: "There has been a fundamental shift from the profit management system by ship or route to management by cargo unit. The profit for each cargo unit yields the net contribution of each container per voyage. The introduction of unit management has resulted in one container having the same meaning that one ship had in the past".

(49) See footnote 48, at pp. 105-6.

(50) See in particular the Gilman lecture cited in footnote 45.

(51) See footnote 45.

(52) See for example Jankowski, "Notes and comments: competition, contestability and the liner shipping industry", Journal of transport economists and policy, May 1989, or Jankowski, "The Development of Liner Shipping Conferences", cited in footnote 45 and the Gilman lecture cited in footnote 45.

(53) See footnote 44.

(54) See inter alia, the US Federal Trade Commission report, cited in footnote 45, at p. 20; Jansson and Schneerson, cited in footnote 45, at Chapter 10.2; and the article "Sea-Land's computer wars", cited in footnote 48.

(55) See footnote 44.

(56) 'Revenue pooling and cartel', p. 173. See also 'The economics of ocean freight rates', Bennathan and Walters, 1969, Praeger.

(57) See footnote 48, at p. 69.

(58) Scherer und Ross, Industrial market structure and economic performance, 1990, Houghtlin Mifflin, at p. 674.

(59) See footnote 48, at p. 120.

(60) See also Jansson and Schneerson, cited in footnote 44, at Chapter 10.2 and Annex A.

(61) Such arrangements are subject to the obligations on shipping lines contained in Article 5 (2) of Regulation (EEC) No 4056-86.

(62) See the reply to the Statement of Objections at paragraph 2.102.

(63) See application for individual exemption at paragraph 4.3.3.

(64) See application for individual exemption at paragraph 4.3.4.

(65) Joined Cases 56 and 58-64 Consten and Grundig [1966] ECR 322, at p. 341.

(66) Case C-41-90 Höfner and Elser v. Macrotron [1991] ECR I-1979, paragraphs 32 and 33, and Case T-65-89, BPB Industries and British Gypsum v. Commission [1993] ECR II-389, at paragraph 134.

(67) Joined Cases T-24-93, T-25-93, T-26-93 and T-28-93, Compagnie maritime belge transports and Others v. Commission [1996] ECR II-1201, at paragraph 205.

(68) Commission Decision 86-398-EEC in Case IV/31.149-Polypropylene (OJ L 230, 18.8. 1986, p. 1), at paragraph 30.

(69) Commission Decision 92-262-EEC in Case IV/32.450 - French-West African shipowners' committees, (OJ L 134, 18.5.1992, p. 1) at paragraph 43.

(70) See the sixth recital of Regulation (EEC) No 4056-86. See also the Compagne maritime belge transports case cited in footnote 67, at paragraph 202.

(71) Case 136-86 BNIC v. Aubert [1987] ECR 4789, paragraph 18. Similarly, the Court ruled under Article 92 in Joined Case 67, 68 and 70-85 Van der Kooy v. Commission (Dutch natural gas prices 1) [1988] ECR 219, at paragraph 59, that subsidisation of the price of natural gas to Dutch glasshouse crop producers by 5,5 % affected trade between Member States because of the importance of energy costs (25 to 30 % of the selling price) and of the market share (65 %) and the exports (91 %) of the firm receiving the State aid.

(72) Joined Cases 6-73 and 7-73 Istituto Chemioterapico Italiana and Commercial Solvents v. Commission, [1974] ECR 223, at p. 252. The EATA parties have argued in the reply to the statement of objections (Annex 12 at paragraph 13) that jurisprudence of the Court of Justice under Article 86 is not relevant to the present proceedings under Article 85. In the Commission's view, the test of an effect on trade between Member States is identical for the purposes of Article 85 and Article 86. Furthermore, it is clear from the two paragraphs following the one quoted above that the principles being discussed by the Court in Commercial Solvents are applicable to Article 85 as well as to Article 86:

"32. The prohibitions of Articles 85 and 86 must be interpreted and applied in the light of Article 3 (f) of the Treaty, which provides that the activities of the Community shall include the institution of a system ensuring that competition in the common market is not distorted, and Article 2 of the Treaty, which gives the Community the task of promoting 'Throughout the Community harmonious development of economic activities'. By prohibiting the abuse of a dominant position within the market in so far as it may affect trade between Member States, Article 86 therefore covers abuse which may directly prejudice consumers as well as abuse which indirectly prejudices them by impairing the effective competitive structure as envisaged by Article 3 (f) of the Treaty.

33. The Community authorities must therefore consider all the consequences of the conduct complained of for the competitive structure in the common market without distinguishing between production intended for sale within the Market and intended for export. When an undertaking in a dominant position within the common market abuses its position in such a way that competition in the common market is likely to be elimated, it does not matter whether the conduct relates to the latter's exports or its trade within the common market, once it has been established that this elimination will have repercussions on the competitive structure within the common market."

(73) Trans-Atlantic Agreement, Commission Decision 94-980-EC of 19 October 1994 (OJ L 376, 31.12.1994, p. 1 at recitals 365 and 366).

(74) Case T-66-89 Publishers Association v. Commission [1992] ECR II-1995, paragraph 69.

(75) The EATA parties have, somewhat confusingly, described the alleged overcapacity as both structural and cyclical.

(76) See application for individual exemption at paragraph 6.1.22.

(77) See application for individual exemption at paragraph 6.1.28.

(78) See the Interim report of the multimodal group at paragraph 65, published in March 1996 by the Office for Official Publications of the European Communities (ISBN 92-827-6964-X).

(79) See application for individual exemption at paragraph 2.5.43.

(80) See footnote 4.

(81) See reply to the Statement of Objections at paragraph 6.3.4.

(82) See the Decision on the Trans-Atlantic Agreement, cited in footnote 13, at recital 345, et seq.

(83) Notice pursuant to Article 23 (3) of Council Regulation (EEC) No 4056-86 concerning Case No IV/33.677 Irish Club Rules, (OJ C 263, 29.9.1993, p. 6 at point).

(84) See the reply to the Statement of Objections at paragraph 3.77.

ANNEX I

CGM SA 22 Quai Galliéni F - 92158 Suresnes Cedex Paris Hapag-Lloyd Container Linie GmbH Rosenstrasse 17 D - 20079 Hamburg Kawasaki Kisen Kaisha Ltd Hibiya Central Building

2-9 Nishi-Shinbashi 1-Chome

Minato-Ku

Tokyo 105 Japan A.P. Møller - Maersk Line 50 Esplanaden DK - 1098 Copenhagen K Malaysian International Shipping Corporation Bhd. 2nd Floor Wisma Misc

No 2 Jalan Conlay

PO Box 10371 50712 Kuala Lumpur Malaysia Mitsui O.S.K. Lines Ltd 1-1 Toranomon 2-Chome

Minato-Ku

Tokyo 105-8688 Japan Neptune Orient Lines Ltd 456 Alexandra Road

No 06-00 NOL Building

Singapore 119962 Republic of Singapore Nippon Yusen Kaisha Yusen Building

3-2 Marunouchi 2-Chome

Chiyoda-Ku

Tokyo Japan Oriental Overseas Container Line 30th-31st Floor Harbour Centre

25 Harbour Road

Wanchai Hong Kong P& O Nedlloyd Container Line Limited Beagle House

Braham Street

London E1 8EP United Kingdom Cho Yang Shipping Co. Ltd Cheong Ahm Building

85-3 Seosomun-Dong, Chung-Ku

Seoul Republic of Korea Evergreen Marine Corp. (Taiwan) Ltd Evergreen Building

No 166, Sec. 2, 330 Minsheng E. Road

10444 Taipei Taiwan

Republic of China

Hanjin Shipping Co. Ltd 25-11 Yoido-Dong

Youngdeungpo-Ku

Seoul 150-010 Republic of Korea Hyundai Merchant Marine Co. Ltd Mukyo Hyundai Building

92 Mukyo-Dong, Chung-Ku

Seoul Republic of Korea DSR-Senator Linie GmbH Martinistraße 62-66 D - 28195 Bremen Yangming Marine Transport Corp.2 271 Ming De 1st Road

Chidu

Keelung

Taipei 206 Taiwan

Republic of China

ANNEX II

EATA capacity limits 1993

<emplacement tableau>

Sliding scale - 1993

<emplacement tableau>