CFI, 5th chamber, extended composition, June 6, 2002, No T-342/99

THE COURT OF FIRST INSTANCE OF THE EUROPEAN COMMUNITIES (Fifth Chamber, Extended Composition),

Facts and procedure

1. On 29 April 1999, Airtours plc, a United Kingdom company whose main activity is as a tour operator and supplier of package holidays, announced its intention to acquire all the shares in the United Kingdom tour operator, First Choice plc, one of its competitors.

2. On the same day, Airtours notified the proposed merger to the Commission pursuant to Article 4 of Council Regulation (EEC) No 4064-89-EEC of 21 December 1989 on the control of concentrations between undertakings (OJ 1989 L 395, p.1, corrected version in OJ 1990 L 257, p. 13), as most recently amended by Council Regulation (EC) No 1310-97 of 30 June 1997 (OJ 1997 L 180, p. 1) (hereinafter Regulation No 4064-89).

3. In its decision of 3 June 1999, the Commission found that the merger gave rise to serious doubts as to its compatibility with the common market and decided to initiate the investigation procedure in accordance with Article 6 (1) (c) of Regulation No 4064-89. 4. On 9 July 1999, the Commission sent the applicant a statement of objections under Article 18 of Regulation No 4064-89, in which it set out the reasons why it took the view, prima facie, that the proposed merger would give rise to a collective dominant position in the United Kingdom short-haul foreign package holiday market. The applicant replied to the statement of objections on 25 July 1999. 5. A hearing was held before the Commission Hearing Officer on 28 and 29 July 1999, pursuant to Articles 14, 15 and 16 of Commission Regulation (EC) No 447-98 of 1 March 1998 on the notifications, time limits and hearings provided for in Regulation No 4064-89 (OJ 1998 L 61, p. 1).

6. On 7 September 1999 the applicant submitted a set of undertakings in accordance with Article 8 (2) of Regulation No 4064-89 in order to allay the competition concerns which had been identified.

7. On 9 September 1999 the Advisory Committee on concentrations met and delivered its opinion on the merger and on the undertakings put forward by the applicant.

8. A meeting was held on 15 September 1999, which was attended by representatives of the applicant and of the Commission, following which the applicant submitted a revised set of undertakings.

9. By decision of 22 September 1999 (Case IV/M.1524 - Airtours/First Choice) (Decision C (1999) 3022 final, published under Number 2000-276-EC (OJ 2000 L 93, p. 1); hereinafter the Decision), the Commission declared that the concentration was incompatible with the common market and the operation of the European Economic Area under Article 8 (3) of Regulation No 4064-89 on the ground that it would create a collective dominant position in the United Kingdom market for short-haul foreign package holidays, as a result of which competition would be significantly impeded in the common market. The Commission stated in the Decision that the undertakings proposed by Airtours on 7 September 1999 would not prevent the creation of a collective dominant position and that the undertakings put forward on 15 September 1999 were submitted too late to be considered at that stage in the procedure.

Procedure and forms of order sought by the parties

10. On 2 December 1999 the applicant brought the present action.

11. Upon hearing the report of the Judge-Rapporteur, the Court of First Instance decided to open the oral procedure and, by way of measures of organisation of procedure, the applicant and the Commission were asked to produce certain documents and reply in writing to various questions.

12. By letters from the Commission of 27 July 2001 and 3 August 2001, and by letter from the applicant of 31 August 2001, the parties complied with the measures of organisation of procedure taken by the Court.

13. The parties presented oral argument and replied to the questions put to them by the Court at the hearing on 11 October 2001. 14. The applicant claims that the Court should:

- annul the decision;

- order the Commission to pay the costs.

15. The Commission contends that the Court should:

- dismiss the action;

- order the applicant to pay the costs.

Substance

16. The applicant relies on four pleas in law in support of its application. The first plea alleges that there were manifest errors of assessment in the definition of the relevant product market and infringement of Article 253 EC. The second plea alleges infringement of Article 2 of Regulation No 4064-89, breach of the principle of legal certainty in so far as the Commission applied a new and incorrect definition of collective dominance in its assessment of the present case, and infringement of Article 253 EC. The third plea alleges infringement of Article 2 of Regulation No 4064-89 - in that the Commission found that the transaction created a collective dominant position - together with infringement of Article 253 EC. The fourth plea alleges infringement of Article 8 (2) of Regulation No 4064-89 and breach of the principle of proportionality inasmuch as the Commission did not accept the undertakings proposed by the applicant.

The first plea alleging errors in the definition of the relevant product market and infringement of Article 253 EC

A - The Decision

17. The definition of the relevant product market in the United Kingdom foreign package holiday industry is the only definition challenged by the applicant. The Decision identifies two separate markets, the market for package holidays to long-haul destinations (long-haul package holidays) and that for package holidays to short-haul destinations (short-haul package holidays). In that connection, it is specified in the Decision that the travel industry considers the long-haul sector to comprise all destinations involving a flight time from the United Kingdom substantially in excess of three hours, other than flights to the islands in the Eastern Mediterranean or the Canary Islands, which may take up to around four hours. As a result, all European (mainland and islands) and North African holiday destinations fall into the short-haul category, in contrast to those destinations in, for example, the Caribbean, the Americas or South-East Asia, in respect of which the flight times are substantially longer (typically twice as long or more) (paragraphs 10 to 13 of the Decision).

18. At paragraphs 16 to 28, the Decision sets out the reasons which led the Commission to conclude that the differences between long and short-haul package holidays are, from the point of view of competition, more significant than the similarities and are such as to justify defining separate markets for the purposes of an appraisal of the concentration notified. Those reasons are the following:

(a) first, for airlines (and, therefore, for vertically integrated tour operators) there is limited scope for substitution between long-haul and short-haul flights, given that there is little scope for using the same aircraft for both short and long-haul destinations and given the operating costs for larger as compared with smaller aircraft and the difficulties that charter airlines (including those of the parties to the merger) must overcome if they attempt substantially to reconfigure [their] fleet as between long-haul and short-haul capabilities, namely, the need to make capital investment, the time necessary to do so and the difficulty of leasing aircraft on a short-term basis, inasmuch as charter airlines (including those of the parties) own most of their aircraft or lease them on relatively long leases (typically a lease of five years) in order to reduce costs, maintain quality and ensure continuity of supply (paragraphs 16 to 18 of the Decision);

(b) second, the fact that from the point of view of the ultimate consumer there are a number of significant differences between short and long-haul package holidays:

(i) the image or idea of the holiday: long-haul packages seem more exotic and therefore appeal to single people or couples without children; short-haul package holidays, for example to Mediterranean resorts, are of more interest to families (paragraph 20 of the Decision);

(ii) the time when holidays are taken: long-haul package holidays are less suited to the needs of United Kingdom consumers travelling en famille who, for the most part, go on foreign package holidays during the summer season (roughly, mid-July to the end of August) so as to coincide with the school holidays (and in some areas, factory closures) (paragraph 20 of the Decision);

(iii) transfer time: longer flight times may deter some consumers from choosing a long-haul package holiday, even if it is comparable in other respects to a short-haul package, for example as regards weather, location, price, visas, medical requirements and the like (paragraph 21 of the Decision);

(iv) lack of price substitutability between short and long-haul destinations: prices are appreciably higher for long-haul package holidays and there is only limited convergence between prices for that kind of holiday and prices for comparable short-haul package holidays. Although prices for the two kinds of holiday, particularly at certain times of the year (for example, when the weather is bad) can sometimes be the same or not very different, that very limited overlap is not sufficient to constrain prices throughout the short-haul market, since the long-haul holidays concerned are regarded as effective substitutes by only a very small proportion of customers (paragraphs 22 to 26 of the Decision).

B - Definition of the relevant product market

19. The Court notes, to begin with, that, as regards the application of Regulation No 4064-89 as envisaged in this case, a proper definition of the relevant market is a necessary precondition for the assessment of the effects on competition of the concentration (see, to that effect, Joined Cases C-68-94 and C-30-95 France and Others v Commission (Kali & Salz) [1998] ECR I-1375, paragraph 143).

20. The definition of the market in the products affected by the merger must take account of the overall economic context so as to make it possible to assess the actual economic power of the undertaking or undertakings in question and, for that purpose, it is necessary first to define the products which, although incapable of being substituted for other products, are sufficiently interchangeable with the undertaking's own products, both as regards their objective characteristics and the competitive conditions and the structure of supply and demand on the market (see, to that effect, Case C-333-94 P Tetra Pak v Commission [1996] ECR I-5951, paragraphs 10 and 13, and Case T-83-91 Tetra Pak v Commission [1994] ECR II-755, paragraph 63).

21. The applicant challenges the definition of the relevant product market given in the Decision. Rather than limiting the relevant market to that for short-haul foreign package holidays, the Commission should have defined it as the market comprising all foreign package holidays, including long-haul packages. The applicant complains that the Commission has departed from previous practice regarding the definition of the foreign package holiday market and maintains that the Commission's assessment of demand-side and supply-side substitutability is incorrect. As a result of that flaw in the Commission's reasoning the Decision is vitiated by manifest errors of assessment and thus an error of law.

22. As regards the Commission's proposition that there is no demand-side substitutability between long and short-haul package holidays, the applicant submits that the Commission's arguments concerning, first, the various product characteristics and, second, the differences in average prices for long and short-haul package holidays are mistaken.

23. It refers, first, to product characteristics and challenges the Commission's contention that long-haul holidays are more exotic, are less suitable for families and involve longer flight times. Thus, short-haul destinations, such as Turkey or North Africa, are more exotic than long-haul destinations, such as Florida or the Dominican Republic, which are more family destinations. Travel time to the resort can be as long for short-haul destinations as for long-haul, since what matters is total travel time, including check-in and transfers, rather than flight time in the strict sense. Finally, the applicant claims that variety in the type of holidays offered by tour operators to take account of different lifestyles (for example, family or non-family) and variety of tastes (in particular, so far as accommodation, food, activities, interests and the like are concerned) exists within both the long-haul holiday segment and the short-haul holiday segment.

24. Second, regarding difference in holiday prices, the applicant argues that it is irrelevant to point out that average prices for long-haul destinations exceed those for short-haul destinations when, as in this instance, the products are clearly differentiated. The applicant also points to a price convergence between the two types of holidays, since some short-haul holidays are in the same price-range as some long-haul holidays.

25. The Court notes that it is apparent from the documents before it that the Commission took account of consumer preferences, average flight time, the level of average prices and the limited interchangeability of the aircraft used for each type of destination in reaching its conclusion that short-haul package holidays belong to a separate market from that to which long-haul packages belong. The Commission came to that conclusion, while not, however, disputing that long-haul package holidays are becoming increasingly popular with consumers or that the market studies cited by the applicant in its reply to the statement of objections (see British National Travel Survey 1998, volume 4, The 1998 Holiday Market, and Mintel, Holidays: The booking procedure, 1997) illustrate the tendency of United Kingdom consumers to go further afield for their holidays and particularly to the other side of the Atlantic. Nor did it question the fact that a substantial number of short-haul holidaymakers have also taken a long-haul holiday in the last five years (36%) and that a much greater number (62%) are very or fairly likely to do so over the next five years, as the applicant has indicated in Table 2.4 in its reply to the statement of objections.

26. The Court must therefore consider whether the Commission made a manifest error of assessment when it concluded that those factors were reasons for defining the relevant product market narrowly and excluding long-haul package holidays, which it did not regard as sufficiently interchangeable with short-haul package holidays.

27. First, concerning the average flight time, the Commission pointed out - and was not challenged by the applicant on this point - the significant difference between the average flight time to long-haul destinations, which is over eight hours, and the average flight time to short-haul destinations, which is usually less than three hours (from the United Kingdom, flights to the islands in the Eastern Mediterranean or the Canary Islands may take up to around four hours). The applicant argues that, in practice, what matters to consumers is not the flight time but the total travel time from the home town to the resort. However, it cannot rely on that argument to play down the indisputable difference between average flight times (three hours on average for flights to short-haul destinations compared with eight hours on average for flights to long-haul destinations), since transfer time from the airport to the resort may in fact also vary, whatever the destination.

28. Second, as regards the importance to be attached to the prices of the two types of holiday and their impact on consumers, the Commission found that differences between the average price of a long-haul package holiday and that of a short-haul package are such as to warrant separate market definition. It should be observed in that regard that the Commission accepts that there is a degree of convergence between prices for the two kinds of holiday. However, it contends that the convergence is not such that the two products may be regarded as substitutes or that the prices of one constrain prices of the other.

29. At paragraph 23 of the Decision the Commission explains its reasons for finding that there was no price substitutability between the two kinds of holiday. It considers the prices offered to the consumer to be significantly higher for long-haul package holidays, as reflected in the information supplied by the applicant in Annex 1a to its response of 29 June 1999 to the Commission's request for information.

30. First, the Commission found that there was a difference of over 100% between the average brochure price of long-haul package holidays for the 1998 summer season and that of short-haul packages. It also considered the question comparing similar packages (14 nights, 3-star, self-catering) in Florida and Spain and found that the latter cost on average about half the price of the former. A similar comparison between Florida and Greece or the Canaries gave broadly equivalent results (a difference of around 30 to 40% for catered accommodation). The Decision gives detailed examples of price comparisons between certain short and long-haul tourist destinations offered in the Airtours brochure, which show that significant price differentials exist between the two kinds of destination.

31. The applicant disputes the relevance of average prices as a means of comparing the effect of prices on consumers' decisions where the products are clearly differentiated. It submits that what is significant for defining the relevant product market is the behaviour of customers at the margin and the question of whether they would be prepared to substitute long-haul package holidays for short-haul packages if the price of the latter were to rise. The Commission acknowledges that average prices do not necessarily reflect prices at the margin but is of the view that, where - as in the present case - the differences are so significant, it is unlikely that a sufficient range of genuinely comparable long-haul package holidays is available at prices which are sufficiently similar to constrain prices of short-haul packages, since the long-haul packages concerned are regarded as genuine substitutes by only a very small proportion of the customers.

32. It is therefore appropriate to consider whether the Commission made a manifest error of assessment in relation to the significance of the margin, that is, the number of customers prepared to react to a price increase in short-haul package holidays by purchasing a long-haul package holiday, as compared with the total number of customers who habitually purchase a short-haul package holiday from tour operators.

33. In that regard, the Court notes in limine that it is common ground between the parties that United Kingdom consumers of foreign package holidays are generally very sensitive to the prices of the products.

34. The Commission's argument is set out at paragraph 24 of the Decision, in which it acknowledges that prices of some holidays at certain long-haul destinations, particularly at certain times of the year (e.g. during periods when bad weather is expected) match or come close to those at the upper end (summer peak, better quality accommodation) of the price/quality scale for short-haul ones. It nevertheless went on to conclude that it is not to be expected that this very limited overlap would suffice to constrain prices throughout the short-haul market, since the long-haul holidays concerned would not be regarded as effective substitutes - either on price or other grounds - by more than a very small proportion of customers.

35. In support of that finding, the Commission points out at paragraph 25 of the Decision that none of the long-haul destinations cited by the applicant in its reply to the statement of objections (Table 2.6) in support of its view on price convergence was in the same price-range as that which it had previously supplied.

36. An examination of Annexes 1a and 2 to the applicant's letter of 29 June 1999 responding to the Commission's requests for information of 15 and 21 June 1999 (documents produced by the Commission in the context of measures of organisation of procedure, see Annex 6b/7b to the first set of documents produced by the Commission) reveals that the Commission was right in holding that the differences in average prices are significant, especially if the comparison relates to the same season (summer or winter). Annex 1a in fact shows that for the summer seasons of 1996, 1997 and 1998, average prices per week for short-haul package holidays were, respectively, GBP 354, GBP 378 and GBP 369, while the corresponding figures for long-haul packages were, respectively, GBP 676, GBP 757 and GBP 781. 37. Furthermore, a review of those documents establishes that the Commission's assessment at paragraph 25 of the Decision is well founded. It is apparent from Annex 2 to the applicant's letter of 29 June 1999 that, for short-haul destinations, the applicant had indicated that typical holidays, for example a week in a 3-star hotel, half board, in Majorca, in July or August 2000, cost GBP 485. The figures are appreciably lower than the figures in Table 2.6 on page 21 of the reply to the statement of objections, which is referred to at paragraph 25 of the Decision. Only the prices of holidays offered for December 1999 to Jamaica (GBP 699), Mexico (GBP 649) and Sri Lanka (GBP 699) are closer to the average figures for short-haul destinations applying for the summer season 2000. 38. Likewise, the documents produced by the applicant bear out the Commission's argument. As stated at paragraph 26 of the Decision, it can be seen that in the BA Holidays advertisement for long-haul package holidays produced by the applicant at the hearing before the Commission (see paragraph 26 of the Decision, footnote 23), four destinations were offered at very competitive prices: Barbados (GBP 399), Tobago (GBP 499), Grenada (GBP 529) and St Lucia (GBP 799). However, as the Commission points out, only the package to St Lucia included food, the other holidays included only the flight and the accommodation. In addition, the prices were low-season prices valid for September and October 1999. 39. It should be added that in its reply of 29 June 1999 to the Commission's enquiries of 15 and 21 June 1999 the applicant cited as an example of one of its typical products a summer holiday in Majorca in a 3-star hotel, costing approximately GBP 485, plus a flight supplement.

40. Further, the applicant acknowledged at the hearing that it publishes separate brochures for short and long-haul package holidays.

41. In those circumstances, the Commission's proposition that only a small proportion of the customers of the main United Kingdom tour operators regard long-haul package holidays as substitutes in terms of value for money for short-haul package holidays cannot be regarded as manifestly incorrect.

42. The other arguments advanced by the applicant do not invalidate that finding.

43. The applicant argues that industry-wide studies treat long-haul package holidays as part of the mainstream. It cites, in particular, Holidays - The Booking Procedure, a study by Mintel, in which the point is made that long-haul has broken into the mainstream holiday market. While based on a desire to travel further and see the world outside Europe, pricing has inevitably come into play as a key element in the consumer's choice. In addition, the Commission should have taken into account the statements of third party tour operators obtained in the course of its enquiry, which also show the growing importance of substitution between long and short-haul package holidays.

44. However, in the circumstances of the present case and with reference to market definition, the fact that the Commission did not consider decisive (i) changing consumer tastes, (ii) the growing importance of substitutability between long-haul package holidays to destinations such as Florida and the Dominican Republic, and short-haul packages or (iii) the growth of the market for long-haul packages over recent years is not sufficient to support a finding that the Commission exceeded the bounds of its discretion in concluding that short-haul package holidays are not within the same product market as long-haul packages.

45. Third, as to the applicant's arguments relating to supply-side substitutability and the interchangeability of the aircraft used on short and long-haul routes, the Commission cannot be criticised for having formed the view that the fact that certain dual-purpose aircraft, such as the Boeing 757, may be used to some extent both for long and short-haul destinations was not sufficiently decisive, given the other findings made concerning demand-side product substitutability, to lead it to adopt a wider market definition. In that regard, it is appropriate to refer, as the applicant itself has done, to paragraph 13 of the Commission Notice on the definition of the relevant market for the purposes of Competition Law (OJ 1997 C 372, p. 5):

From an economic point of view, for the definition of the relevant market, demand substitution constitutes an immediate and effective disciplinary force on the suppliers of a given product, in particular in relation to their pricing decisions.

46. Finally, the applicant cannot rely on a failure to state reasons in relation to the definition of the relevant market.

47. The Commission devoted a significant part of the Decision (paragraphs 5 to 28) to explaining why it considered the relevant market to be limited to the market for short-haul package holidays. The Decision thus discloses, in a clear and unequivocal fashion, the Commission's reasoning relating to the definition of the relevant market, in such a way as to enable the Community Courts to exercise their power of review and the persons concerned to be aware of the reasons for the measure in order to defend their rights (see Case C-350-88 Delacre and Others v Commission [1990] ECR I-395, paragraph 15).

48. It follows that the first plea must be rejected as unfounded.

The second plea alleging infringement of Article 2 of Regulation No 4064-89, breach of the principle of legal certainty and infringement of Article 253 EC inasmuch as the Commission applied an incorrect definition of collective dominance in its appraisal of the present case

49. The applicant complains that the Commission, for the purposes of the Decision, applied a new and incorrect definition of collective dominance, which is set out generally at paragraphs 51 to 56 of the Decision, departing from its previous decisions, from Community case-law and from sound economic principles, and also infringing Article 2 of Regulation No 4064-89. The Commission thereby also acted in breach of the principle of legal certainty and Article 253 EC, inasmuch as the Decision is vitiated by a defective statement of reasons.

50. The Commission denies that it adopted a new approach and maintains that it applied the test for collective dominance already used by it in previous cases and approved by the Court of First Instance in its judgment in Case T-102-96 Gencor v Commission [1999] ECR II-753.

51. It is appropriate to point out that the abovementioned paragraphs of the Decision (51 to 56) are in Part VA of the Decision, in which the Commission sets out, purely by way of introduction and summary, the reasons which led it to conclude that the concentration would give rise to the creation of a dominant position and in which it replies generally to observations made by the applicant during the administrative procedure concerning certain of the characteristics of a collective dominant position.

52. In the introduction to its legal analysis of the concentration, the Commission merely sketches the broad outlines of its findings on the effects of the merger, which are subsequently explained and developed in detail at paragraphs 57 to 180 of the Decision.

53. Since the Decision is a measure applying Article 2 of Regulation No 4064-89 to a specific concentration, the Court must, in its review of the legality of the Decision, confine itself to the position adopted by the Commission in relation to the transaction as notified, that is to say, it must examine the way in which the law has been applied to the facts and adjudicate on the merits of the Commission's findings concerning the effects of the concentration on competition. In this case, the specific findings relating to the impact of the transaction on competition, which led the Commission to conclude that the concentration should be prohibited, are stated and developed in paragraphs 57 to 180 of the Decision and are challenged by the applicant in its third plea.

54. It is therefore necessary to consider, first, the merits of the arguments raised by the applicant in its third plea and, at the same time, to take into account its arguments concerning the Commission's general findings at paragraphs 51 to 56 of the Decision.

The third plea alleging (i) infringement of Article 2 of Regulation No 4064-89 in that the Commission found that the concentration would create a collective dominant position, and (ii) infringement of Article 253 EC

55. By this plea, the applicant seeks to show that the Commission made an error of assessment in deciding that the proposed merger should be prohibited. It claims that the Decision does not prove to the requisite legal standard that the outcome of the transaction would be the creation of a collective dominant position of such a kind as significantly to impede competition in the relevant market. In prohibiting the merger, the Commission thus infringed Article 2 of Regulation No 4064-89. A - General considerations

56. Under Article 2 (2) of Regulation No 4064-89, a concentration which does not create or strengthen a dominant position as a result of which effective competition would be significantly impeded in the common market or in a substantial part of it is to be declared compatible with the common market.

57. Under Article 2 (3) of the Regulation, a concentration which creates or strengthens a dominant position as a result of which effective competition would be significantly impeded in the common market or in a substantial part of it is to be declared incompatible with the common market.

58. Where, for the purposes of applying Regulation No 4064-89, the Commission examines a possible collective dominant position, it must ascertain whether the concentration would have the direct and immediate effect of creating or strengthening a position of that kind, which is such as significantly and lastingly to impede competition in the relevant market (see, to that effect, Gencor v Commission, paragraph 94). If there is no substantial alteration to competition as it stands, the merger must be approved (see, to that effect, Case T-2-93 Air France v Commission [1994] ECR II-323, paragraphs 78 and 79, and Gencor v Commission, paragraph 170, 180 and 193).

59. It is apparent from the case law that in the case of an alleged collective dominant position, the Commission is ... obliged to assess, using a prospective analysis of the reference market, whether the concentration which has been referred to it leads to a situation in which effective competition in the relevant market is significantly impeded by the undertakings involved in the concentration and one or more other undertakings which together, in particular because of factors giving rise to a connection between them, are able to adopt a common policy on the market and act to a considerable extent independently of their competitors, their customers, and also of consumers (Kali & Salz, cited above, paragraph 221, and Gencor v Commission, paragraph 163).

60. The Court of First Instance has held that: There is no reason whatsoever in legal or economic terms to exclude from the notion of economic links the relationship of interdependence existing between the parties to a tight oligopoly within which, in a market with the appropriate characteristics, in particular in terms of market concentration, transparency and product homogeneity, those parties are in a position to anticipate one another's behaviour and are therefore strongly encouraged to align their conduct in the market, in particular in such a way as to maximise their joint profits by restricting production with a view to increasing prices. In such a context, each trader is aware that highly competitive action on its part designed to increase its market share (for example a price cut) would provoke identical action by the others, so that it would derive no benefit from its initiative. All the traders would thus be affected by the reduction in price levels. (Gencor v Commission, paragraph 276).

61. A collective dominant position significantly impeding effective competition in the common market or a substantial part of it may thus arise as the result of a concentration where, in view of the actual characteristics of the relevant market and of the alteration in its structure that the transaction would entail, the latter would make each member of the dominant oligopoly, as it becomes aware of common interests, consider it possible, economically rational, and hence preferable, to adopt on a lasting basis a common policy on the market with the aim of selling at above competitive prices, without having to enter into an agreement or resort to a concerted practice within the meaning of Article 81 EC (see, to that effect, Gencor v Commission, paragraph 277) and without any actual or potential competitors, let alone customers or consumers, being able to react effectively.

62. As the applicant has argued and as the Commission has accepted in its pleadings, three conditions are necessary for a finding of collective dominance as defined:

- first, each member of the dominant oligopoly must have the ability to know how the other members are behaving in order to monitor whether or not they are adopting the common policy. As the Commission specifically acknowledges, it is not enough for each member of the dominant oligopoly to be aware that interdependent market conduct is profitable for all of them but each member must also have a means of knowing whether the other operators are adopting the same strategy and whether they are maintaining it. There must, therefore, be sufficient market transparency for all members of the dominant oligopoly to be aware, sufficiently precisely and quickly, of the way in which the other members' market conduct is evolving;

- second, the situation of tacit coordination must be sustainable over time, that is to say, there must be an incentive not to depart from the common policy on the market. As the Commission observes, it is only if all the members of the dominant oligopoly maintain the parallel conduct that all can benefit. The notion of retaliation in respect of conduct deviating from the common policy is thus inherent in this condition. In this instance, the parties concur that, for a situation of collective dominance to be viable, there must be adequate deterrents to ensure that there is a long-term incentive in not departing from the common policy, which means that each member of the dominant oligopoly must be aware that highly competitive action on its part designed to increase its market share would provoke identical action by the others, so that it would derive no benefit from its initiative (see, to that effect, Gencor v Commission, paragraph 276);

- third, to prove the existence of a collective dominant position to the requisite legal standard, the Commission must also establish that the foreseeable reaction of current and future competitors, as well as of consumers, would not jeopardise the results expected from the common policy.

63. The prospective analysis which the Commission has to carry out in its review of concentrations involving collective dominance calls for close examination in particular of the circumstances which, in each individual case, are relevant for assessing the effects of the concentration on competition in the reference market (Kali & Salz, paragraph 222). As the Commission itself has emphasised, at paragraph 104 of its decision of 20 May 1998 Price Waterhouse/Coopers & Lybrand (Case IV/M.1016) (OJ 1999 L 50, p. 27), it is also apparent from the judgment in Kali and Salz that, where the Commission takes the view that a merger should be prohibited because it will create a situation of collective dominance, it is incumbent upon it to produce convincing evidence thereof. The evidence must concern, in particular, factors playing a significant role in the assessment of whether a situation of collective dominance exists, such as, for example, the lack of effective competition between the operators alleged to be members of the dominant oligopoly and the weakness of any competitive pressure that might be exerted by other operators.

64. Furthermore, the basic provisions of Regulation No 4064-89, in particular Article 2 thereof, confer on the Commission a certain discretion, especially with respect to assessments of an economic nature, and, consequently, when the exercise of that discretion, which is essential for defining the rules on concentrations, is under review, the Community judicature must take account of the discretionary margin implicit in the provisions of an economic nature which form part of the rules on concentrations (Kali & Salz, paragraphs 223 and 224, and Gencor v Commission, paragraphs 164 and 165).

65. Therefore, it is in the light of the foregoing considerations that it is necessary to examine the merits of the grounds relied on by the applicant to show that the Commission made an error of assessment in finding that the conditions for, or characteristics of, collective dominance would exist were the transaction to be approved.

B - The Decision

66. The Decision identifies two types of players on the relevant market (see paragraphs 72 and 75), the large tour operators on the one hand, and the secondary or small tour operators on the other:

- the major tour operators are characterised by their relatively large size - each of them having a market share exceeding 10% (according to the Commission's data, Thomson accounts for 27% of sales, Airtours for 21%, Thomas Cook for 20% and First Choice for 11%, that is, overall for 79% of sales. On Airtours' figures, Thomson accounts for 30.7% of sales, Thomas Cook for 20.4%, Airtours for 19.4% and First Choice for 15%, that is, overall for 85.5% of sales). A further characteristic is that they are all integrated both upstream (operation of charter airlines) and downstream (travel agencies);

- the secondary operators are smaller, none of them having a market share in excess of 5%, and in general they do not own either their own charter airlines or their own travel agencies. Apart from Cosmos (which, since it is linked to Monarch, one of the major charter airlines in the United Kingdom, is exceptional among secondary operators where there is no vertical integration), Manos and Kosmar, which are the fifth, sixth and seventh tour operators accounting respectively for 2.9%, 1.7% and 1.7% of sales, there are several hundred competing small tour operators, none of them accounting for more than 1% of sales.

67. It is apparent from the Decision (see the summary of the Commission's appraisal at paragraphs 168 to 172 of the Decision) that the Commission formed the view that the proposed merger would create a dominant position in the United Kingdom market for short-haul foreign package holidays, the effect of which would be to impede competition significantly in the common market for the purposes of Article 2 (3) of Regulation No 4064-89, and that it would do so for the following reasons:

- the proposed merger would remove competition between the three large players remaining after the concentration (combined Airtours/First Choice, Thomson and Thomas Cook). Because of the structural features of the market and the way that it operates, which is dependent on capacity decisions, and because of the high degree of market concentration (the three remaining large tour operators would have about 80% of the market if the operation took place) (Decision, paragraph 169), they would no longer have an incentive to compete with each other;

- the operation would increase the degree of transparency and interdependence which already exists, with the result that the three remaining large tour operators would have every interest in adopting parallel conduct so far as the decision as to how many package holidays to put onto the market is concerned, reducing capacity below what is required as a result of market trends (Decision, paragraph 170);

- an examination of past competition bears out this conclusion, since it demonstrates that the relevant market already had a tendency towards collective dominance (Decision, paragraphs 128 to 138);

- deterrents or scope for retaliation exist, which are connected with the fact that if one of the three remaining large tour operators decided not to restrict capacity, there would be a risk that the two others would do the same, which would result in oversupply and serious financial consequences for each of the operators (Decision, paragraph 170);

- the smaller operators or new entrants, that is to say current and future competitors, would be further marginalised as a result of the operation, since they would lose First Choice both as a supplier of airline seats and as a potential distribution channel. In any event, those operators would not have the ability to offset any reductions in capacity brought about by the three remaining large tour operators (Decision, paragraph 171).

68. So far as the effects of the merger on effective competition are concerned, the Commission found that the effect of restricting overall capacity put onto the market would be to tighten the market and bring about an increase in the prices and profits of the members of the dominant oligopoly (see, in particular, paragraph 56 and the final part of paragraph 168 of the Decision).

C - The Commission's alleged errors of assessment

69. The applicant argues that, contrary to the Commission's contention, the factors put forward by the Commission in the Decision to characterise the situation as one of collective dominance were not present at the time of the notification and would not occur were the merger to proceed.

70. More specifically, the applicant claims, first, that, given the characteristics of the relevant market, the Commission has not proved conclusively that, were the merger to proceed, the three remaining large tour operators would have an incentive to cease competing with each other.

71. Second, it argues that, even supposing that such an incentive did exist, the absence of any deterrents or adequate means of retaliation would prevent the emergence of the alleged dominant oligopoly.

72. Third, and in any event, smaller operators and new entrants, namely current and future competitors, would challenge any capacity restrictions brought into effect and consumers would react as a result, so that the three remaining major operators would not be able, as a result of the concentration, to act together to any appreciable extent independently of other competitors and consumers.

73. Fourth, the applicant claims that the Commission incorrectly assessed the impact of the merger on competition in the relevant market.

1. Preliminary observations

74. The applicant's first point is that the natural tendency of operators in the relevant market to set capacity cautiously has by no means prevented them from engaging in competition with each other in the past and that there is no reason to believe that proceeding with the proposed merger would put an end to that competition by creating a situation in which the three remaining large tour operators have a collective dominant position.

75. The Decision is particularly elliptical in its description of the competitive situation at the time of the notification. However, it is not disputed that the Commission concluded that the proposed merger would create, rather than strengthen, a dominant position on the market (Decision, paragraph 194). The Commission has confirmed in its pleadings that it does not contend that that there was a situation of oligopolistic dominance at the time of the notification and that what is at issue is the creation, and not the strengthening, of a collective dominant position. Thus, it does not deny that prior to the proposed merger the major tour operators did not find it possible or profitable to restrict capacity in order to increase prices and revenues.

76. It follows that in this instance the starting point for the Court's examination must be a situation in which - in the Commission's own view - the four major tour operators are not able to adopt a common policy on the market and hence do not face their competitors, their commercial associates and consumers as a single entity, and in which they thus do not enjoy the powers inherent in a collective dominant position.

77. In those circumstances, it was for the Commission to prove that, in view of the characteristics of the United Kingdom market for operating short-haul package holidays and in light of the notified operation, approval of the latter would have resulted in the creation of a collective dominant position restrictive of competition, inasmuch as Airtours/First Choice, Thomson and Thomas Cook would have had the ability, which they did not previously have, to adopt a common policy on the market by setting capacity lower than would normally be the case in a competitive market already distinguished by a degree of caution in matters of capacity.

78. It is therefore necessary to examine the pleas and arguments raised by the applicant in this case in the light of the foregoing considerations.

2. The finding that were the merger to proceed, the three remaining large tour operators would have an incentive to cease competing with each other

79. The applicant submits that the finding that if the merger proceeded, the three remaining large tour operators would have an incentive to cease competing with each other is erroneous because the Commission, first, did not take into account, as it should have done, the competition obtaining between the leading tour operators at the time of the notification and, second, made errors in its appraisal of the characteristics of the market on which it relied as evidence that a collective dominant position would be created, in particular, the past and prospective development of demand, demand volatility and the degree of market transparency.

(a) The assessment of competition between the leading tour operators

80. The applicant submits that the analysis of competition obtaining prior to the notification (referred to as past competition) is of fundamental importance in this instance, since the main incentives invoked by the Commission, namely alleged capacity rigidities, are inherent in the way the market normally operates, concern the industry as a whole and would not be affected were the proposed merger to proceed. The relevant market has been operating competitively in recent years and the applicant challenges the Commission's assertion that there is already a tendency towards collective dominance. In particular, it criticises the way in which the Commission took into consideration the alleged tendency towards collective dominance before examining the implementation of the proposed merger and the volatility of historic market shares.

81. The Commission contends that the way in which the market previously operated and the fact that competition obtained in the past are not significant factors, since the Decision is based on the finding that a collective dominant position would be created as a result of the proposed operation, that is to say that market conditions would be altered in such a way that previously examined incentives and conduct would no longer be an appropriate guide in determining how operators would react in the new market situation. Therefore, it argues that the crucial question is whether the proposed operation would alter current market conditions in such a way that the leading operators would no longer act in the same way as they have done in the past. Thus, it does not follow that because the market was competitive with four large tour operators, it would continue to be so if the number were reduced to three. The Commission nevertheless challenges the arguments put forward by the applicant to show that competition among the main operators has been, and will continue to be, keen.

82. The Court observes, however, that one of the questions which the Commission is required to address where there is alleged to be collective dominance is whether the concentration referred to it would result in effective competition in the relevant market being significantly impeded (Kali and Salz, paragraph 221, and Gencor v Commission, paragraph 163). If there is no significant change in the level of competition obtaining previously, the merger should be approved because it does not restrict competition (see paragraph 58 above). It follows that the level of competition obtaining in the relevant market at the time when the transaction is notified is a decisive factor in establishing whether a collective dominant position has been created for the purposes of Regulation No 4064-89. 83. As the applicant has submitted, an analysis of competition prior to the notification is particularly important in this instance, since the purpose of the tacit coordination likely, in the Commission's submission, to follow the concentration, would be to restrict capacity put onto the market by the three remaining integrated tour operators. The restriction would involve more than their natural caution in capacity planning, which the Commission itself considers inherent in the way that the market normally operates.

(i) The tendency towards collective dominance alleged to exist prior to the proposed merger

84. First, it must be observed that, although the Commission devoted one section of the Decision to an examination of Past Competition (paragraphs 128 to 138), a detailed analysis of that section reveals that, in fact, the Commission does not there express an opinion on the degree of competition obtaining in the market. It confines itself to setting out (paragraphs 128 to 138) a number of circumstances or factors which have been observed in the market in the years leading up to the notification and concludes (Decision, paragraph 138) that there is evidence that there is already a tendency towards collective dominance in the market at present (most especially as regards the setting of capacity). However, those passages of the Decision make no mention of any reduced level of competition in the market prior to the notification.

- The fact that the large tour operators take a cautious approach to capacity planning and take particular note of the estimates of the main competitors

85. At paragraphs 135 and 136 of the Decision, the Commission explains that the large tour operators adopt a cautious approach to capacity planning and take particular note of the estimates of their main competitors (the Decision cites, at paragraph 136, certain statements made by senior executives at the leading United Kingdom integrated tour operators, which illustrate that cautious approach to planning). In the preceding paragraph (paragraph 135), the Decision describes an episode which took place in the summer of 1995, which, the Commission contends, illustrates the consequences of oversupply in the market: during the 1994 planning period, all the tour operators overestimated demand for the summer season 1995 and were left with unsold capacity, which had to be cleared by means of heavy discounting, something which led to them suffering heavy losses.

86. The applicant claims that the major tour operators cannot be criticised for adopting a cautious approach to capacity planning, by taking particular note of the estimates of the other leading operators' plans, since the Community Courts have recognised that the requirement of independence does not deprive economic operators of the right to adapt themselves intelligently to the existing and anticipated conduct of their competitors (Joined Cases 40-73, 41-73, 42-73, 43-73, 44-73, 45-73, 46-73, 47-73, 48-73, 50-73, 54-73, 55-73, 56-73, 111-73, 113-73 and 114-73 Suiker Unie and Others v Commission [1975] ECR 1663, paragraphs 173 and 174). It does not consider such caution to be incompatible with aggressive competition to maintain or increase its market share at the expense of its main competitors. In that respect, the applicant cites the statements made by the main tour operators describing their ambitions to grow.

87. The Commission contends that the intentions ascribed to the main tour operators by the applicant reflect the situation prior to the proposed merger and thus relate to different circumstances. It does not contend that there was previously a situation of dominant oligopoly. Furthermore, the aggressive growth to which certain statements refer had in the past been, and was in the future to be, carried out through acquisitions. Lastly Thomas Cook indicated to the Commission that size was no longer its prime concern, but rather profitability (Decision, paragraph 131).

88. The Court notes that the Decision acknowledges at several places that this natural tendency to cautious capacity planning is a feature of the relevant market in its current state, in which there is no collective dominant position restrictive of competition, and that it impinges upon all operators and not just the large tour operators, even though the latter are more specifically concerned (see paragraphs 60 to 66, 97 and 136 of the Decision). Thus, in paragraph 97 of the Decision, the Commission states that the volatility of demand makes it rational to limit planned capacity and then add capacity later, if demand proves to be particularly strong. In this way the suppliers protect themselves against downwards volatility in demand, and, in paragraph 136, that the large operators take a cautious approach to capacity planning, taking particular note of estimates of the other major operators' plans.

89. In those circumstances and as the Commission has not denied that the relevant market was competitive prior to the notification (in particular at the time of the 1995 crisis), the episode which occurred in that year and to which the Decision attaches great weight cannot, as such, constitute evidence that a tendency towards collective dominance already existed in the industry. The fact that during the 1994 planning period operators miscalculated and suffered heavy losses after overestimating demand for the 1995 summer season can be regarded as no more than an example of the risks peculiar to this market, the distinctive operation of which is explained at paragraphs 59 to 66 of the Decision.

90. It is quite apparent from the remarks cited at paragraph 136 of the Decision that senior executives at the large tour operators are aware of the risks inherent in expansionist strategies in the relevant market, particularly because of the lessons learned from the 1995 episode and because matching capacity to demand is crucial to profitability (see paragraph 60 of the Decision). However, those remarks do not give the slightest indication that there is no competition between the main tour operators.

91. Finally, contrary to the Commission's contention (see paragraphs 137 and 138 of the Decision), the fact that to some extent (30 to 40% of the shares) the same institutional investors are found in Airtours, First Choice and Thomson cannot be regarded as evidence that there is already a tendency to collective dominance in the industry. It is sufficient to point out in that regard that, as the Commission itself has acknowledged in its defence (paragraph 73), there is no suggestion in the Decision that the group of institutional shareholders forms a united body controlling those quoted companies or providing a mechanism for exchange of information between the three undertakings. Furthermore, the Commission cannot contend that those shareholders are a further force for cautious capacity management, unless it has examined to what extent they are involved in the management of the companies concerned. Finally, even assuming that it were proved that they are capable of exercising some influence on the management of the undertakings, since the concerns of the common institutional investors with respect to growth (and thus capacity) merely reflect a characteristic inherent in the relevant market, the Commission would still have to establish that the fact that institutional investors hold shares in three of the four leading tour operators amounts to evidence that there is already a tendency to collective dominance.

92. It is apparent from the foregoing that, since it did not deny that the market was competitive, the Commission was not entitled to treat the cautious capacity planning characteristic of the market in normal circumstances as evidence substantiating its proposition that there was already a tendency to collective dominance in the industry.

- The assessment of horizontal and vertical integration characteristic of the market since publication of the Monopolies and Mergers Commission Report

93. The applicant points out that the United Kingdom Monopolies and Mergers Commission (the MMC) examined the state of competition in the relevant market in 1997 and produced a report entitled Foreign Package Holidays: a report on the supply in the UK of tour operators' services and travel agents' services in relation to foreign package holidays (the MMC Report). It maintains that the MMC concluded in that report that the situation in that market was broadly competitive.

94. The Commission submits that the market situation has changed significantly since the MMC Report was published in 1997, not only because of the increase in vertical integration of the main tour operators, as the applicant claims, but also because of the significant horizontal concentration that has taken place.

95. At paragraphs 128 to 134 of the Decision, the Commission cites, as evidence of a tendency to collective dominance, the horizontal concentration and vertical integration which have characterised the United Kingdom foreign package holiday industry in recent years and which have accelerated since the MMC Report was published in December 1997, due in particular to the number of mid-sized operators taken over by the four major tour operators.

96. The Court observes, however, that if those trends are examined closely, it can be seen that the main tour operators' acquisitions of tour operators, airline companies and travel agencies, referred to at paragraph 134 of the Decision, do not entail the kind of major alterations in the market which would invalidate in 1999 the conclusions on competition in the market reached by the MMC towards the end of 1997 and that, therefore, those acquisitions could not be regarded as evidence of a tendency to collective dominance.

97. First, the Court would point out that, as the applicant has noted, the MMC found in its report published in 1997 that the foreign package holiday business was dynamic, that competition was keen and that there were no significant barriers to entry. It reached that conclusion after carrying out a particularly detailed study (over 300 pages) of the state of the travel business and the way in which it functioned. The study was carried out over 12 months of inquiry and drew on large amounts of data and a great number of views provided by all parties with a presence in the United Kingdom foreign holidays business. For the purposes of drawing up its report, the MMC commissioned four market studies from external consultants and the report was completed in November 1997, only a year and a half before the Commission examined the market in the context of the notified transaction.

98. More specifically, in paragraph 1.6 of its report, the MMC expressed itself as follows:

The travel trade has been far from static over the last ten years and the picture continues to change, with a trend towards more vertical integration. Of the major participants who featured in our 1986 investigation into foreign package holidays, only Thomson has retained a prominent position. We have received a great deal of evidence to the effect that competition in the trade is strong and we broadly agree with this view. While concentration has increased over the past five years, it is not at a particularly high level. Profits are not excessive taken year on year. Players come and go. There are no significant barriers to entering either the tour operator or the travel agent market.

99. The Commission has not challenged that analysis in the Decision, although in several places it refers to findings made by the MMC in the report relating to other matters (paragraphs 9, 11, 47, 70, 76, 81, 114, 115, 123, 128, 129, 131, 133 and 134 of the Decision). It follows that it does not challenge the MMC's findings regarding its description of that market, in 1997, as one in which competition was strong.

100. The Commission argues, however, (Decision, paragraph 123) that conditions of competition in the market have changed significantly since 1997, in particular owing to increased concentration and vertical integration, as is stated at paragraph 134 of the Decision. In that regard, the elimination of mid-sized operators represents an important change in the competitive structure and increases the scope for parallel conduct between the main tour operators.

101. However, the Court finds that the horizontal and vertical integration which has occurred in the United Kingdom foreign package holiday industry since the MMC Report was published are less significant than the Commission alleges.

102. As regards horizontal concentration, it is apparent from the documents before the Court (page 33a of the document, at Annex 5 to the application, notifying the Commission of the merger, paragraph 4.18 of the 1997 MMC Report and paragraph 72 of the Decision) that the development, between 1996 and 1999, of the market shares of Thomson, Airtours and First Choice does not prove that their shares of the short-haul market have increased significantly. The graph illustrating the tour operators' market shares (Annex 5 to the application, page 33a), including all destinations, shows that Thomson's sales of foreign package holidays, which accounted for 25% of sales of foreign package holidays in 1996, accounted for no more than 22% of sales in 1998, whilst Airtours' sales represented 16% in 1996 and 1998, and those of First Choice fell from 10% in 1996 to 9% in 1998. The three leading tour operators active in 1997 therefore accounted for 51% of sales of foreign package holidays in 1996 and for 47% thereof in 1998. That finding is borne out if the data relating purely to short-haul destinations are examined. It is apparent from paragraph 4.18 of the MMC Report that in 1997 the MMC had also examined the market shares of the integrated tour operators by reference to the narrow definition of the product market used in the Decision, since the advantages of also examining the market by reference to such a definition had been pointed out by Thomas Cook at that time. It is clear from a comparison of market shares in 1996 (paragraph 4.18 of the MMC Report) and 1998 (paragraph 72 of the Decision) that Thomson's share fell from 33% to 30% or 27%, depending on the source of the figures, Airtours' share moved from 20% to 19% or 21%, depending on the figures used, and First Choice from 12% to 15% or 11%, depending on the figures used. Only Thomas Cook increased its market share significantly, rising from 6% to 20%.

103. It is clear from this that the key element in the consolidation that has taken place in the foreign package holiday trade since 1997 is Thomas Cook, whose status changed in the course of a few years from that of a small operator to that of a major tour operator following several transactions involving growth by acquisitions (acquisition in June 1996 of Sunworld, acquisitions in 1998 of Flying Colours, which accounted for 3% of the foreign package holiday trade, and of Carlson/Inspirations, which accounted for between 1 and 3% of the trade) (MMC Report, table 4.1, p. 76; Decision, paragraphs 131 and 134). By virtue of that growth, Thomas Cook emerged in 1998 as a fourth large tour operator, which was vertically integrated and thus better placed to compete with the other integrated tour operators. That fact cannot be interpreted as evidence of a lack of competition in the market.

104. If the case of Thomas Cook is left aside, the acquisitions of tour operators referred to at paragraph 134 of the Decision essentially concern the purchase of smaller businesses which have not significantly increased the market share of the main tour operators in the foreign package holiday business. Therefore, the elimination of mid-sized operators, which the Commission contends represented an important change in the structure of competition and increased the scope for parallel conduct between the main operators, amounts to the fact that a new major tour operator has emerged - Thomas Cook, whose market share has risen from 6% to 20%.

105. The Commission contends that the large tour operators' increased level of vertical integration, which has also occurred since 1997, is further evidence of the industry's tendency to collective dominance (Decision, paragraph 138). The Court observes that the Decision is inconsistent in this regard, since it is based at the same time on the premiss that a strategy of vertical integration is necessary in order to compete with the large tour operators. Thus, at paragraph 132 of the Decision, the Commission states that First Choice had adopted a policy of vertical integration into distribution in 1998, the aim of which was to protect itself against the business practices of the other large tour operators in order to avoid paying them commission and to obtain better information on market trends. That need to become vertically integrated is one of the key factors in the Commission's conclusion, the Commission taking the view that a collective dominant position would arise in the present case, in particular because the concentration would remove First Choice as a competitor at all three levels of the supply chain (paragraph 168 of the Decision).

106. It follows that the Commission itself recognises in the Decision that an increased level of vertical integration is pro-competitive, inasmuch as it increases efficiency and limits the interdependence of the large tour operators, who promote their own distribution channels over those of the other main operators. The fact that vertical integration has occurred since the MMC Report was published in 1997 therefore cannot at the same time be evidence of a tendency to collective dominance. In addition, the MMC also analysed in its report the growing tendency to vertical integration and concluded that it was something that was as likely to stimulate competition as to dampen it (see paragraph 2.193 of the MMC Report). In particular, the MMC concluded that the anti-competitive effects of vertical integration were slight in 1997, account being taken of the levels of concentration in the industry.

107. It follows that the Commission was wrong in taking the view that the horizontal concentration and vertical integration that has taken place since the MMC Report was published in 1997 made it necessary to disregard the latter's findings on the level of competition obtaining in the relevant market.

108. It is apparent from the foregoing that the Commission erred in concluding at paragraph 138 of the Decision that the factors set out at paragraphs 128 to 137 thereof are evidence that there is already a tendency towards collective dominance in the market at present (most especially as regards the setting of capacity).

(ii) The assessment of the volatility of historic market shares

109. The applicant goes on to cite as proof that the market is competitive the fact that in the past the market shares of the main tour operators have been volatile, dynamic and fluctuating.

110. According to the Commission, no such volatility has recently been observed in the relevant market. The changes in the large tour operators' market shares, cited by the applicant, result from acquisitions and not therefore from their performance on the market. If acquisitions are discounted, the market shares of the main operators have shown very little movement in recent years, which suggests that organic growth is difficult (Decision, paragraph 128 and footnote 86).

111. It is appropriate to point out that for the purpose of determining whether there is a collective dominant position, the stability of historic market shares is a factor conducive to the development of tacit collusion, inasmuch as it facilitates division of the market instead of fierce competition, each operator referring to its historic market share in order to fix its production in proportion thereto.

112. In the present case, the Commission's finding that the market shares of Thomson, Airtours, Thomas Cook and First Choice remained stable over the last five years is predicated on the assumption that growth by acquisition is to be ignored. The Commission takes the view that where changes in market share result principally from the acquisition of competitors, the quota to be allocated can be calculated by adding the market shares of the purchaser and the target and that therefore the problem of operators seeking to align their market share on peaks they have achieved in the past does not arise.

113. However, there is no justification in the present case for excluding growth by acquisition when assessing the volatility of market shares, inasmuch as in the relevant market the size of the undertakings and their degree of vertical integration are significant factors in competition (see, inter alia, paragraphs 73, 75, 77, 78, 99, 100, 114 and 115 of the Decision). In such circumstances, the fact that the large operators have made numerous acquisitions in the past, either before or after the MMC Report was published, may be taken to be indicative of strong competition between those operators, which make further acquisitions to avoid being outdistanced by their main competitors in key areas in order to take full advantage of economies of scale.

114. Moreover, the assumption that growth by acquisition is to be ignored is at variance here with several remarks made by the Commission itself in the Decision, which suggest, contrary to the submissions in its pleadings, that an acquisition by one of the major operators results, for the other major operators, not in the mathematical addition of the market shares of the purchaser and the target, but in a competitive reaction on their part.

115. The Decision thus points out at paragraph 137: [w]hen Airtours' bid for First Choice became known in April this year, an announcement by Thomson that it would defend its market-share position led to an immediate drop in Thomson's share price of 9% on the same day as the announcement, due to fears that the company would start a price war, and Thomson's management were obliged to make considerable efforts to convince institutional investors that the announcement had been misinterpreted and that they had no intention of adding capacity in the market but only of mopping up capacity which would be shed by Airtours/First Choice as a result of the merger.

116. Similarly, the Decision states at paragraph 145 that it is apparently widely believed in the industry that all mergers lead to temporary losses of market share for the protagonists due to defection of some customers and suppliers as a result of elimination of duplication in their programmes. An examination of the documents produced by the applicant concerning the development of historic market shares (Annex 6 to the reply, page 2, see also the table on the development of market shares, page 8 of the application) shows that such concerns are well founded. Thus, following Thomson's acquisition of Horizon in 1989, the market share of the new entity should have been 32% (25% for Thomson and 7% for Horizon), whereas it quickly dropped to around 25%.

117. It is appropriate to point out that an analysis of the data produced to the Court shows that, as the applicant has submitted without challenge from the Commission, if growth by acquisition is included, there is considerable variation in the major tour operators' shares of the foreign package holiday market. This can be seen from the table showing the operators' market shares, set out by the applicant in its notification (reproduced on page 8 of the application). Thus, in 1990, Thomson's market share was 21.81%, that of First Choice was 5.82%, that of Airtours was 4.27% and that of Thomas Cook was 2.13%. In 1994 Thomson's market share was 23.13%, Airtours' 15.52%, First Choice's 5.88% and Thomas Cook's 2.41%. Then, in 1998, Thomson's market share was 19.28%, Airtours' 14.26%, First Choice's 7.47% and Thomas Cook's 11.38%.

118. It follows that the Commission erred in holding that market shares resulting from acquisitions should not be taken into consideration and, accordingly, in concluding that the major tour operators' market shares have remained stable over recent years.

119. Finally, as regards competition obtaining in the relevant market, it should be added that the applicant has claimed, and the Commission has not disputed, that the performances of the main tour operators may vary in a given season (with winners and losers) and may also vary from one season to another. That fact must be regarded as evidence that the market is competitive and consequently militates against any finding of collective dominance.

(iii) Conclusion on the assessment of competition between the leading tour operators

120. It follows from the foregoing that the Commission made errors of assessment in its analysis of competition obtaining in the relevant market prior to the notification. First, it did not provide adequate evidence in support of its finding that there was already a tendency in the industry to collective dominance and, hence, to restriction of competition, particularly as regards capacity setting. Second, it did not take into account, as it should have done, the fact that the main tour operators' market shares have been volatile in the past and that such volatility is evidence that the market was competitive.

(b) The assessment of past and anticipated development of demand, demand volatility and the degree of market transparency

121. A section of the Decision entitled Market Characteristics (oligopolistic dominance) (paragraphs 87 to 127) sets out a number of characteristics which, according to the Commission, make the relevant market conducive to oligopolistic dominance. They include, inter alia, product homogeneity, low demand growth, low price sensitivity of demand, the similar cost structures of the main suppliers, a high degree of transparency, interdependence and commercial links between the main suppliers, substantial barriers to market entry and the insignificant buyer power of consumers. According to the Decision (paragraph 87), those characteristics are already present and would remain present if the proposed merger were to take place.

122. The applicant challenges the findings which led the Commission to conclude that those characteristics are already present in the relevant market and that they would make it conducive to oligopolistic dominance if the merger were to proceed. It argues, in particular, that the rate of demand growth and the degree of demand volatility in the relevant market, as well as the degree of market transparency, are, in the present case, factors which, contrary to the Commission's contention, render creation of a collective dominant position more difficult.

(i) Findings on low demand growth

123. The applicant argues in essence that the Commission made an error of assessment in considering demand growth overall to be weak, whilst both the data produced in the administrative procedure and the fact that demand growth is faster than growth in gross domestic product are evidence to the contrary.

124. The Commission sets out its findings about the level of demand growth in the relevant market at paragraphs 92 and 93 of the Decision.

125. At paragraph 92, the Commission states that [a] recent study for a major tour operator, referred to in response to the Commission's enquiries, noted ... that the overall average annual growth rate (3 to 4% over the decade) was quite low. It is also stated that [d]emand growth for the next two years is expected to be close to zero, according to several industry estimates, but with some recovery in prospect thereafter.

126. At paragraph 93, the Commission goes on to state that based on its investigation in this case, it has reached the conclusion that overall growth of demand in the market for short-haul package holidays will continue to be moderate as has been the case in the 1990s. Finally, in conclusion, the Commission finds that market growth is not likely to provide a stimulus to competition within the foreseeable future.

127. The Court holds that the Commission's findings are based on an incomplete and incorrect assessment of the data submitted to it during the administrative procedure.

128. First, it is appropriate to point out that, in response to a measure of organisation of procedure by which the Court called on the Commission to produce the study referred to at paragraph 92 of the Decision, the Commission stated that a full version of that study had at no point been made available to it during the administrative procedure and that all it could produce to the Court was an extract, which a tour operator had annexed to a response to a request for information. That extract consisted of a single page of a document entitled Forecasting Holiday Demand prepared by Ogilvy & Mather at an unknown date.

129. According to that extract, [t]he GB market for holidays abroad has grown massively over the last 20 years. According to the British National Travel Survey, Britons took nearly 30 million holidays abroad (4-plus nights) - more than treble the number in 1978. Over the last decade the market has grown by an average of 3.7% per annum. The extract also states, in relation to demand volatility, that [w]hile underlying market growth has been persistent, annual growth rates have been far from steady. Annual growth rates of 10% or more are quickly followed by sizeable contractions; that not only is holiday demand more volatile than both gross domestic product and consumer durable spending, it is not fully coincident with the economic cycle (for example, the market grew by over 10% during the depth of the 1980/1 recession); and that [t]he volatility of demand makes forecasting volumes highly problematic.

130. However, it is apparent from a cursory examination of that document that the Commission's reading of it was inaccurate. Thus, at paragraph 92 of the Decision, it states that it also noted that the overall average annual growth rate ... was quite low, whilst no statement to that effect is made in the extract sent to the Court. Conversely, the Commission ignored the emphasis placed by the author of the extract on the massive increase in foreign holiday sales that has taken place over the last 20 years. It follows that the Commission construed that document without having regard to its actual wording and overall purpose, even though it decided to include it as a document crucial to its finding that the rate of market growth was moderate in the 1990s and would continue to be so (Decision, paragraph 93).

131. Second, it is apparent from these passages of the Decision (paragraphs 92 and 93) that the Commission did not take account of the rate of demand growth during the two years preceding the notification, 1997 and 1998, which, however, proved to be important points of reference inasmuch as the effects of the 1995 episode had by then been absorbed by the market. It is clear from data in volume 4 of the 1998 British National Travel Survey (dated February 1999), provided in Annex 9 to the applicant's notification of a concentration, that the foreign holiday sector enjoyed strong growth throughout the decade and, thus, also over recent years. It is apparent from page 113 (and the table on page 112) that the number of departures for foreign holidays rose from 21 million in 1989 to 29.25 million in 1998 (an increase of more than 39.2% over the last decade). After the crisis of 1995, as a result of which the number of foreign holidays fell from 26 million in 1995 to 23.25 million in 1996 (a fall of around 10.5%), the number of foreign holidays rose from 23.25 million to 27.25 million in 1997 (an increase of more than 17.2%) and from 27.25 million to 29.25 million in 1998 (an increase of more than 7.3%). It is specifically stated in relation to 1998 that what is concerned is real growth and not a difference caused by the common practice of rounding up figures where the changes from one year to another are small. The fact that those data also relate to long-haul package holidays does not undermine their probative value as regards the tendency to sustained growth, since that type of holiday has accounted for only a fifth of all holidays in recent years (see page 116 of the British National Travel Survey).

132. The Commission failed to take account of those data in its estimates of the level of market growth; instead, it referred to the trend for the next two years, stating at paragraph 92 of the Decision: Demand growth for the next two years is expected to be close to zero, according to several industry estimates, but with some recovery in prospect thereafter. When it was questioned on this point at the hearing, the Commission replied that the finding was based on an econometric study produced during the administrative procedure in response to a request for information. It is noteworthy that neither the nature of that econometric study, nor its authorship, nor the context in which it was produced were mentioned in the Decision. Lastly, it should be added that the estimate of demand growth close to zero is at variance with the following paragraph of the Decision (paragraph 93), in which the Commission itself recognises that the market for short-haul foreign package holidays is likely to continue to grow and that [i]t may also be that the market will grow somewhat faster than overall GDP growth due to increases in vacation time and general wealth.

133. It is clear from the foregoing that the Commission's interpretation of the data available to it concerning growth demand was inaccurate in its disregard for the fact that the market had been marked by a clear tendency to considerable growth over the last decade in general, despite the volatile nature of demand from one year to another, and that the pace of demand growth has increased during recent years in particular. In that context of growth, and having failed to produce any more specific evidence establishing that the tendency to grow would be reversed in future years, the Commission was not entitled to conclude that market development was characterised by low growth, which was, in this instance, a factor conducive to the creation of a collective dominant position by the three remaining large tour operators.

(ii) Findings on demand volatility

134. The applicant submits that demand volatility renders it more difficult to show that a collective dominant position exists, since volatility adds noise to the market making it more difficult to differentiate between changes in demand caused by volatility in the market and capacity increases brought about by departures from the common policy. The fact that it is so difficult to distinguish between the two types of event clearly suggests that any attempted collusion will be unstable.

135. The Commission recognises in the Decision that there is a degree of demand volatility in the market (Decision, paragraphs 92 and 95). However, it argues (Decision, paragraph 97) that in the present case that volatility does not preclude the creation of a collective dominant position but rather the reverse: [it] makes the market more conducive to oligopolistic dominance. The reason is that the volatility demand in combination with the fact that it is easier to increase than to decrease capacity, means that it is rational for the major operators to adopt a conservative approach (wait and see approach) to capacity decisions. In particular the volatility of demand makes it rational to limit planned capacity and then add capacity later, if demand proves to be particularly strong. In this way the suppliers protect themselves against downward volatility in demand.

136. In any event, at paragraphs 94 to 96 of the Decision, the Commission challenges the arguments put forward by the applicant during the administrative procedure concerning demand volatility and its causes, which are linked to gross domestic product, changing consumer tastes and changing costs (the impact of low-cost airlines). The Commission contends (paragraph 95) that all tour operators are exposed to the business cycle and have to consider the macroeconomic development in their forecast. Therefore it is likely that all tour operators will have similar views as to the market development.

137. The Commission recognises (Decision, paragraph 96 and footnotes 73 and 74) that certain exogenous shocks, such as terrorist attacks on tourists in Egypt or Turkey, may disrupt the planning of tour operators, but it none the less refuses to take them into consideration as a factor likely to make the market less conducive to collective dominance, since such events, which are by nature exceptional, are not peculiar to the short-haul package holiday market but may happen in any market.

138. Lastly, the Commission accepts that the applicant's remarks on the difficulty that such volatility entails for the creation of collective dominance are in accordance with economic theory but asserts that they are not relevant in the present case. According to the Commission, since it is easier to add capacity than to reduce it, operators will tend to be cautious in order protect themselves against possible volatility. Moreover, it is easy to differentiate between a decline in demand and an increase in capacity by another operator, since the actions of the latter can be observed directly.

139. The Court observes in limine that, as the Commission recognises, economic theory regards volatility of demand as something which renders the creation of a collective dominant position more difficult. Conversely, stable demand, thus displaying low volatility, is a relevant factor indicative of the existence of a collective dominant position, in so far as it makes deviations from the common policy (that is, cheating) more easily detectable, by enabling them to be distinguished from capacity adjustments intended to respond to expansion or contraction in a volatile market.

140. In the present case, the Commission acknowledges that a certain degree of demand volatility is a characteristic of the relevant market (Decision, paragraphs 92, 95 and 97). However, several of the documents before the Court indicate that there is a considerable degree of volatility in the market. The extract from the study cited at paragraph 92 of the Decision explains that [n]ot only is holiday demand more volatile than both gross domestic product and consumer durable spending, it is not fully coincident with the economic cycle (for example, the market grew by over 10% during the depth of the 1980/81 recession) and [t]he volatility of demand makes forecasting volumes highly problematic. Likewise, there is evidence of the high volatility of the market in the figures taken from the 1998 British National Travel Survey. After the 1995 crisis, as a result of which the number of foreign holidays fell from 26 million in 1995 to 23.25 million in 1996 (a fall of around 10.5%), the number of trips abroad rose from 23.25 million to 27.25 million in 1997 (an increase of more than 17.2%) and from 27.25 to 29.25 million in 1998 (an increase of more than 7.3%).

141. The Commission contends, however, that that fact is not relevant in the present case as operators tend to be cautious to protect themselves against any volatility.

142. However, the Commission is not entitled to rely on the fact that tour operators, to protect themselves against sudden downward volatility in demand, plan capacity cautiously, preferring to increase it later if demand proves to be particularly strong (Decision, paragraph 97), for the purpose of denying the relevance in this instance of a factor which is significant as evidence of oligopolistic dominance, such as the degree of market stability and predictability. Although it is certainly the case that the caution inherent in the way the market normally operates means that account must be taken of the need to make the best possible estimates of the way in which demand will develop, the planning process remains difficult, because each operator must anticipate (some 18 months in advance because of the market's distinctive features) how demand will evolve - demand being distinguished by its considerable volatility and thus entailing a degree of speculation. Furthermore, the Commission did not regard either the operators' caution or demand volatility to be restrictive of competition in the pre-merger market. Caution cannot therefore be interpreted, as such, as evidence of a collective dominant position rather than as a characteristic of a competitive market of the kind that existed at the time of the notification.

143. Finally, the arguments advanced by the Commission (Decision, paragraphs 94 to 96) concerning the applicant's points cannot be accepted.

144. As regards volatility linked to the business cycle, the Commission cannot just conclude - as it does as at paragraph 95 of the Decision - that it is likely that all tour operators will have similar views as to the market development without producing any evidence in support of that statement, given that capacity is set initially some 18 months before the start of the season (see paragraph 63 of the Decision). At that point, it is not possible to make a precise forecast of how the main macroeconomic variables, such as growth in gross domestic product, exchange rates or consumer confidence, will develop.

145. The Commission's approach to volatility related to exogenous shocks is that tour operators take data relating to market volatility into account when setting capacity (Decision, paragraph 96 and footnotes 73 and 74). That approach amounts to acting in the way that it criticises, namely treating exogenous shocks as endogenous variables by taking them into account in forecasting demand. However, tour operators apparently do not act in that way. That can be seen from the reversals suffered by Thomson in May 1999 when it made heavy losses on its package holiday sales in the Eastern Mediterranean as a result of the war in Kosovo and terrorist threats in Turkey, whilst Airtours, for its part, was not affected, a point made by the applicant, which is not disputed by the Commission.

146. Finally, the Court must reject the Commission's argument that there is no difficulty in differentiating between a decline in demand and an increase in capacity by another operator because the latter can be observed directly. The Court rejects that argument on the ground that an integrated tour operator will, for the reasons set out below in the examination of market transparency, find it difficult to interpret with any accuracy capacity decisions taken by the other tour operators.

147. It follows from the foregoing that the Commission has failed to establish that economic theory is inapplicable in the present case, and that it was wrong in concluding that volatility of demand was conducive to the creation of a dominant oligopoly by the three remaining major tour operators.

(iii) The assessment of the degree of market transparency

148. At paragraph 102 of the Decision the Commission states with regard to transparency: a distinction has to be made between the planning period and selling season, where the catalogues have been launched, [b]ut transparency of the market is high for the four major integrated operators in both periods.

149. At paragraphs 103, 104 and 105, the Commission states that [i]n the planning period the crucial capacity decisions for the coming season are made and that ... the capacity decisions of the four major integrated operators will be transparent for each of these suppliers, for the following reasons:

- none of the major tour operators puts out a completely new programme from one season to the next. Rather, the planning of a future season is based on sales in the previous season, increased or decreased by a forecast of demand for the coming season. Changes compared to the previous season are therefore incremental and the development of the programme of a tour operator is evolutionary. Consequently, by virtue of past experience, tour operators know already before the planning of a season to a large extent what the offerings of the other four integrated suppliers will be for the new season (Decision, paragraph 104);

- each of the four major integrated tour operators has some knowledge of the changes planned by the other three during that period, given that they use the same hotels and avail themselves of the other tour operators' airline companies to obtain or supply capacity or agree swaps of seats or slots (Decision, paragraph 105);

- substantial capacity additions cannot be kept secret, for example the purchase or long-term lease of additional aircraft is necessarily made public (Decision, paragraph 105).

150. At paragraph 105 of the Decision, the Commission finds that for all the above reasons, each of the four major integrated operators would know if, for example, one of the other integrated operators was planning to increase the number of passengers carried and thus the number of holidays it could offer. Each of the four integrated operators is thus well able to monitor the total amount of holidays offered by each of the others.

151. At paragraph 113 of the Decision, the Commission concludes from this that, given the capacity rigidities, the high degree of transparency will make it even more likely [after the merger] that the major suppliers will under-supply the market, leaving more unsatisfied demand than would be likely under a less transparent system (in which there would be more - temporary - oversupply, requiring lower prices in order to clear the products) so allowing them to raise average prices above the competitive level.

152. The applicant submits that the relevant market is not transparent during the planning period. It argues, in essence, that overall capacity decisions consist of a wide range of individual planning decisions concerning each resort and each flight and that changes made to capacity planned by reference to past capacity are significant and very difficult to identify.

153. Nor, in the applicant's submission, is the relevant market transparent during the selling period. It maintains in essence that capacity transparency is not possible without price transparency and that the Commission failed to appreciate the nature of the information available on the computerised holiday reservation systems.

154. The Commission accepts that capacity decisions taken in the planning period are not wholly transparent. However, it recalls the various ways in which information may be obtained, referred to at paragraphs 104 and 105 of the Decision, so far as the four major tour operators are concerned.

155. The Commission contends that during the selling period price transparency is of no importance, since the key determinant of competition in the relevant market is not price but capacity. However, during that period transparency in relation to overall capacity is virtually complete, since each operator is able to calculate the capacity of its competitors on the basis of what is offered in their catalogues and also on the basis of their past programmes.

156. It is appropriate to observe in limine that the fact that a market is sufficiently transparent to enable each member of the oligopoly to be aware of the conduct of the others is conducive to the creation of a collective dominant position.

157. First, the Court observes (i) that the Commission's argument is based on the contention that in this instance the tacit coordination instancing the collective dominant position is focused not on prices but on the capacity put onto the market and (ii) that, as the Commission states at paragraph 103 of the Decision, the crucial capacity decisions for the coming season are taken during the planning period. At paragraph 63 of the Decision, the Commission itself recognises that once the booking season has begun (for example, from about the summer of 1999 for departures in summer 2000), the scope for changes is heavily constrained, due to the inflexibility of many commitments with suppliers and the problems associated with changing dates, flights, hotels, etc. for customers who have already booked. It states (paragraph 62 of the Decision) that Airtours accepts that there is scope for an increase of capacity of up to 10% during that period.

158. The approach thus taken by the Commission is borne out by its assertion (paragraph 108 of the Decision) in response to the applicant's argument that, since each of the large integrated tour operators has to deal with several thousand different prices because of the various programmes offered, tacit agreement on all those prices would be impossible: the Commission asserts that it does not consider an agreement on prices to be necessary in this instance in order to reach a collective dominant position. It adds:

during the selling season, there is little incentive for any of the integrated operators to cut prices in order to gain market share, which is determined by the amount of capacity offered. Therefore, operators have no need to tacitly collude on thousands of prices. Indeed, this point was confirmed by the economic experts of Airtours: pricing behaviour of firms after capacity has been determined is not directly relevant for joint dominance, i.e. the collective exercise of market power.

159. It follows that in this instance it is appropriate to ascertain, first, whether each of the large tour operators will be able, when making its crucial capacity decisions during the planning period, to find out with any degree of certainty what those of its main competitors are. Only if there is sufficient transparency will an operator be able to estimate the total capacity decided upon by the other members of the alleged oligopoly and then be in a position to be sure that by planning its capacity in a given way it is adopting the same policy as them and hence will have an incentive to do so. The degree of transparency is also important for the purposes of permitting each member of the oligopoly subsequently to detect alterations made by the others as regards capacity, to distinguish deviations from the common policy from mere adjustments consequent upon volatility of demand and, finally, to ascertain whether it is necessary to react to any such deviations by punishing them.

160. It is apparent from the applicant's responses (part B.1 and Annexes 5 to 8) to a measure of organisation of procedure taken by the Court that capacity setting for each season is not a mechanical exercise involving no more than renewing capacity from one year to the next, which would be easy for the other tour operators to predict, but instead involves each large tour operator in a very complex task, which takes historical data into account only to a limited extent and which is based principally on a subjective assessment by each operator by reference to a whole range of variables and factors.

161. Specifically, an examination of the data shows that the planning cycle does not simply run from year to year. By way of example, for the summer season 1999 (Year N), running from May 1999 to October 1999, capacity planning starts some 18 months before, in October or November 1997 (Year N-2). During the main planning period, culminating with first brochure issue about April or May 1998 (Year N-1), tour operators have available to them data relating to results for the summer season 1997 (Year N-2) and some data relating to the forthcoming summer 1998 season (Year N-1). Within that time frame, overall capacity planning operates by reference to general and specific considerations which are refined over time. General considerations (top-down considerations) take account of the key factors influencing holiday demand, such as economic activity, exchange rates and consumer confidence. Specific considerations (bottom-up considerations) are based on a detailed analysis of existing product offerings, starting with, for example, consideration of gross and net margins by flight and accommodation unit for each resort. In that connection, each flight (by departure and destination airport and flight slot) is analysed, as are the available destinations and products and consumer demand for particular types of holiday, so that a comprehensive range of short-haul foreign package holidays can be prepared. That range is also supplemented by new product offerings developed by the applicant.

162. The applicant has stated that, given that the products concerned are perishable, it tends during the planning stage to attach greater importance to its analysis of macro-economic factors or specific considerations concerning costs and margins than to its examination of levels of historic demand, since those factors are more likely than previous performance (sales realised and projected for Years N-2 and N-1) to affect disposable income and future demand. Previous performance is, however, also taken into account in planning for the summer season of Year N, inasmuch as it is an indicator of the strengths and weaknesses of what is currently on offer and of where what is offered can be improved.

163. In statistical terms, the table submitted by the applicant (Annex 7 to its response) comparing budgeted and actual sales of its main United Kingdom subsidiary, Airtours Holidays Ltd, for the period 1996-2000 shows the differences between projected capacity for the year being planned (Year N), budgeted capacity for Year N-1 (whose selling season has already begun) and capacity sold during Year N-2 (as that season is already over). The table shows that capacity budgeted by Airtours Holidays Ltd for Year N varies significantly in comparison with either budgeted capacity for Year N-1 (from + 7.5% to + 11.2%, depending on the Year N under consideration) or with capacity sold during Year N-2 (from + 7.5% to + 18.6% depending on the Year N under consideration). By way of comparison, the variations represent a capacity increase which is two or three times greater than the overall average annual demand growth in the market (between 3 and 4%) identified by the Commission at paragraph 92 of the Decision.

164. It is apparent from the foregoing that the crux of the planning process is not simply the renewal of capacity budgeted or sold in the past but is the attempt to predict how demand will develop on both a macroeconomic and microeconomic level.

165. In addition to the factors set out above, it is necessary to mention the practical difficulties, to which the applicant has drawn attention, which make it very difficult to find out what capacity is projected by each of the other large tour operators during the planning period, inasmuch as their decisions on total capacity for a given season consolidate a whole range of individual decisions, taken on a resort-by-resort and flight-by-flight basis and varying from one season to the next.

166. The applicant maintains, without challenge from the Commission, that it serves around 50 destinations from 21 United Kingdom airports, which represents more than 1 000 permutations, and that it varies those permutations appreciably from one season to the next. Thus, for the summer of 1999, Airtours increased its capacity to Fuerteventura by 19%, although it reduced its departures to that destination from Manchester by 13%, while departures from Cardiff were increased by 42%. Similarly, Airtours' capacity to Minorca was reduced by 9%, with departures to that destination from Manchester being reduced by 33% and departures from Scottish airports being increased by 25%. By way of example, within the 3-star/self-catering category, which according to the Decision (paragraph 90) accounts for the large majority of short-haul package holidays, there are differences as to the airport and the departure date, the length of the stay and the resort. It should be observed in this respect that the argument that little differentiation is made for the air component (Decision, paragraph 90) does not alter the fact that decisions relating to airline capacity are taken airport by airport and flight by flight.

167. So, contrary to the Commission's contention, capacity decisions do not involve merely increasing or reducing overall capacity, without taking account of the differences between the various categories of package holidays, which are differentiated by destination, departure date, departure airport, aircraft model, type and quality of accommodation, length of stay and, finally, price. To be able to develop their package holidays, tour operators must take into account a series of variables, such as the availability of accommodation at the various destinations and the availability of airline seats on various dates and at different times of the year. As the applicant has argued, capacity decisions are necessarily taken on a micro level.

168. The Commission's global approach (paragraphs 88 to 91 of the Decision), which regards the total number of package holidays offered by each operator as what is important, thus encounters some significant difficulties on a practical level, since, in order to ascertain total capacity - to the extent that it stems from a miscellaneous set of individual decisions - it is necessary to be able to identify those decisions.

169. It follows that, on the face of it, the complexity of the capacity planning procedure, the development of the product and its marketing is a major obstacle to any attempt at tacit coordination. In a market in which demand is on the whole increasing, but is volatile from one year to the next, an integrated tour operator will have difficulty in interpreting accurately capacity decisions taken by the other operators concerning holidays to be taken a year and a half later.

170. However, despite the fact that each tour operator takes capacity decisions on the basis of a miscellaneous set of factors, it is nevertheless necessary to consider whether, in practice, at the time when total capacity is set, each member of the oligopoly can know the overall level of capacity (number of holidays) offered by the individual integrated tour operators.

171. The Commission alleges at paragraph 105 of the Decision that each of the four integrated operators is thus well able to monitor the total amount of holidays offered by each of the others [during the planning period] and that changes made by each individual operator at that stage may be identified by the other major tour operators as a result of their dealings with hotels or their discussions about seat requirements and availability, the purpose of which is to obtain or supply capacity or to negotiate swaps of seats and slots.

172. However, the Commission fails to prove those allegations.

173. First, it cannot be ascertained from the Decision how much information an integrated tour operator may obtain by virtue of the fact that several such operators may be in contact with the same hotels for the purpose of negotiating and reserving bed stocks. Even supposing that the major United Kingdom tour operators were actually offering the same hotels in their packages, it remains the case that a large number of players are involved in the holiday accommodation trade, both on the supply-side and the demand-side. There is, therefore, a strong likelihood that one of the major tour operators will be using a hotel where there is no chance of its finding that one of its competitors is doing so too. That likelihood is increased by the fact that hotel owners prefer to let their rooms to at least two tour operators, generally from different countries. That strategy, which is referred to in the notification, can be explained by the hotel owner's concern to guard against the risk of a decline in demand for holidays supplied by one of its clients or a decline in demand in one of those countries.

174. It follows that the fact that several integrated tour operators may negotiate with the same hotel does not significantly increase market transparency at the time when capacity decisions are made.

175. Second, the Decision gives scarcely any details as to the extent or significance of any information that may be obtained as a result of the larger tour operators discussing airline seat requirements and availability with a view to obtaining or supplying capacity or negotiating swaps of seats or slots. Since the Decision does not provide clarification on this point, it is scarcely conceivable that any useful information concerning the increase or maintenance of capacity could be gleaned by means of the exchange of airline seats or slots, since, as a general rule, those exchanges are likely to be carried out on the basis of one seat against another or one slot against another.

176. In that regard, in the situation envisaged by the Commission, where significant restrictions keep capacity below estimated levels of demand, integrated tour operators would enter into contracts for fewer airline seats and hotel rooms. It is doubtless much more difficult, as a general rule, to detect and interpret decisions to reduce business than decisions to expand it and, in a context of increasing demand, such restrictive strategies would be particularly difficult to detect. Furthermore, the applicant has, without challenge from the Commission, drawn attention to the fact that decisions to increase capacity significantly and to make the corresponding investments are made public only after initial capacity has been set, with the result that they do not permit precise identification of decisions made during the planning period. Consequently, the Commission's argument at paragraph 105 of the Decision, according to which the purchase or long-term lease of additional aircraft could not be kept secret since such decisions are necessarily made public, cannot be taken into consideration for the purposes of proving that there is market transparency as between the four major tour operators at the capacity planning stage.

177. In addition, decisions relating to the use of air fleets are taken at a late stage in the planning period. According to the data provided by Airtours in its reply to the statement of objections, it is only from the 12th month before the start of the season, namely at the same time as the brochure is published, that Airtours takes its first decisions relating to the use of its airline seat capacity. Decisions relating to capacity purchased from other airline companies are not taken until the following months. Therefore, information obtained in the course of negotiations between the larger tour operators is obtained later than the Decision suggests.

178. Furthermore, according to the information submitted at the time of the notification, Airtours does not depend to any significant degree on the other major tour operators for the purchase of airline seats. Airtours apparently only rarely uses its main competitors' charter airlines. Thus, Airtours Holidays' main suppliers of airline seats for summer 1998 were: Spanair (27.2% of purchases); Monarch (22%); Air Europa (21%); Air 2000, First Choice's airline (9.4%); Airworld, Thomas Cook's airline (8.7%); Air Malta (3.8%), and 12 other airlines (7.9%). Thus, Airtours Holidays does not use, or does so only rarely, Britannia (Thomson's airline, which, as is well known, flies mainly for its parent company), and makes only secondary use of First Choice's and Thomas Cook's main airlines (Air 2000 and Airworld account for 18.1% of the total) (Annex 5 to the application, paragraphs 6.94; 6.119 and 6.122, MMC Report, table 3.6, page 66). The applicant's main customers for airline seats for the 1998 summer season were: First Choice (Unijet) (around 68 000 seats); Monarch (Cosmos) (around 45 000 seats); Jet Direct (around 11 500 seats); Air Travel Group (around 10 500 seats); and Manos (about 10 500 seats); the remaining seats were sold to 20 other operators. Here, too, it can be seen that Airtours Holidays is not, or is only rarely, in contact with the airlines of Thomson (Britannia) and Thomas Cook (Caledonian, Airworld, Flying Colours, Peach) (Annex 5 to the application, paragraph 6.94 and MMC Report, table 3.6, page 66).

179. Therefore, contrary to the Commission's contention, the fact that the major tour operators negotiate between themselves to obtain or supply capacity or arrange swaps of seats or slots does not result in a sufficient degree of transparency at the time when capacity decisions are taken.

180. It follows from all of the foregoing that the Commission wrongly formed the view, at paragraph 102 of the Decision, that market transparency is high for the four major integrated operators during the planning period. Accordingly, it appears that it wrongly concluded that the degree of market transparency was a characteristic which made the market conducive to collective dominance (Decision, paragraph 87), and it is not necessary to examine the merits of its findings concerning the degree of transparency during the selling season, since the crucial capacity decisions for the following season are taken during the planning period and thereafter the scope for increasing capacity is very limited.

(iv) Conclusions on the assessment of past and anticipated development of demand, demand volatility and the degree of market transparency

181. It follows from the foregoing that the Commission's examination of competition obtaining between the main tour operators at the time of the notification was inadequate, and that the Commission made errors of assessment concerning the development and predictability of demand, demand volatility and the degree of market transparency, and that it wrongly concluded that those factors were, in this instance, conducive to the creation of a collective dominant position.

(c) Conclusion

182. It follows from all of the foregoing that the Commission made errors of assessment when it concluded that if the transaction were to proceed, the three major tour operators remaining after the merger would have an incentive to cease competing with one another.

3. The inadequate nature of the deterrents which the Commission alleges will secure unity within the alleged dominant oligopoly

183. The applicant complains that the Commission has not taken account of the fact that, even if there were an incentive for the three remaining large tour operators tacitly to coordinate their capacity strategies after the merger because of the characteristics of the relevant market and the impact of the transaction on that market, the retaliation mechanism or deterrents are not adequate to secure unity within the alleged dominant oligopoly. The fact that there are no effective retaliation mechanisms within the relevant market raises the question as to whether oligopolistic dominance is feasible, inasmuch as the long-term incentive not to depart from the common policy is wanting. The applicant submits that the punishment mechanism must be credible and denies, therefore, that the mere threat of retaliation may amount to a sufficient deterrent, as the Commission appears to suggest at paragraph 151 of the Decision.

184. The applicant does not consider the means allegedly available for retaliation during the season to be credible. As regards the scope for adding capacity in the period up to February prior to the summer season, capacity could be increased by only 10% and no further increases could subsequently be made. Further, in an industry characterised by demand volatility, a capacity increase of 10% is not enough to amount to a real deterrent. Moreover, the extra cost of laying on additional capacity for punitive purposes would not be offset by the benefits which those who had been cheated against would derive from imposing the punishment. In any event, any increase in capacity is extremely problematic because it might be against the interests of those called on to punish, inasmuch as late added capacity, being likely to be of low quality (inconvenient flight times, poor-quality accommodation), is difficult to sell. The applicant also denies that it is possible to use cut-price or directional selling against a competitor as a disciplinary weapon.

185. Lastly, the applicant submits that the means allegedly available to retaliate during the following season are not effective. Given the 18-month period necessary to arrange a large volume of capacity, any cheating detected in the course of one selling season could be punished by large capacity additions only two seasons later. The link between the failure to observe the agreement and the punishment will thus be blurred.

186. The Commission observes, first, with reference to the retaliation mechanism contemplated in the Decision, that, since it does not regard oligopolistic dominance as a cartel, it at no time imagined one of the operators simply threatening to retaliate.

187. It goes on to argue that the prospect of retaliation in the same season can be understood as a real and effective threat, inasmuch as operators can assess capacity placed on the market by their competitors once the first edition of brochures is published, some 12 to 15 months before the travel season (see paragraphs 105 to 107 of the Decision). A 10% increase in capacity is likely to exert a significant downward pressure on prices and cancel out a large part of the gains expected by the deviating operator.

188. As regards the means to retaliate during the following season, the Commission (unlike the applicant) takes the view that it is possible to increase capacity considerably during that season and considers it misconceived to claim that a substantial increase could be made only two seasons later.

189. In response to the applicant's argument that it is irrational of the other operators to incur a risk of general overcapacity and that, therefore, one operator can cheat with impunity, the Commission argues that such reasoning is irrelevant in this case, because it implies that the only possible reaction for the other operators is to abandon market share to the cheat.

190. Lastly, the Commission rejects the argument that the tactics of de-racking the brochures of a deviating competitor and engaging in directional selling designed to put its products at a disadvantage would be ineffective. Even if the Commission relies solely on the figures for sales of other operators' stock advanced by the applicant (16% of its products are sold through Thomson and Thomas Cook), a potential loss amounting to that percentage of sales constitutes a significant threat in an industry, characteristics of which are high volumes and low margins.

191. The Court notes that the Commission adopted a somewhat ambiguous approach in the Decision, since it initially stated that a strict retaliation mechanism founded on coercion is not a necessary condition for collective dominance in this case (Decision, paragraph 55; see also paragraph 150), while also stating that it does not agree that there is no scope for retaliation in this market and that [r]ather there is considerable scope for retaliation, which will only increase the incentives to behave in an anti-competitive parallel way (Decision, paragraph 55; see also paragraph 151).

192. The Court observes, in limine, that, as it has already pointed out (see paragraphs 61 and 62 above), the prospective analysis of the market necessary in any assessment of an alleged collective dominant position must not only view that position statically at a fixed point in time - the point when the transaction takes place and the structure of competition is altered - but must also assess it dynamically, with regard in particular to its internal equilibrium, stability, and the question as to whether any parallel anti-competitive conduct to which it might give rise is sustainable over time.

193. It is thus important to ascertain whether the individual interests of each major tour operator (maximising profits while competing with the whole range of operators) outweigh the common interests of the members of the alleged dominant oligopoly (restricting capacity in order to increase prices and make supra-competitive profits). That would be the case if the absence of deterrents induced an operator to depart from the common policy, taking advantage of the absence of competition essential to that policy, so as to take competitive initiatives and derive benefit from the advantages inherent therein (see, to that effect, Gencor v Commission, cited above, paragraph 227 regarding market transparency, and paragraphs 276 and 281 concerning structural links).

194. The fact that there is scope for retaliation goes some way to ensuring that the members of the oligopoly do not in the long run break ranks by deterring each of them from departing from the common course of conduct.

195. In that context, the Commission must not necessarily prove that there is a specific retaliation mechanism involving a degree of severity, but it must none the less establish that deterrents exist, which are such that it is not worth the while of any member of the dominant oligopoly to depart from the common course of conduct to the detriment of the other oligopolists.

196. In this instance the following deterrents are identified in the Decision:

- the deterrent effect of the mere threat of returning to a situation of oversupply, the 1995 experience showing what could happen if a capacity war broke out (Decision, paragraph 151; see also paragraph 170);

- the scope for increasing capacity by up to 10% during the selling season, at least until February (Decision, paragraph 152);

- the scope for a tour operator to add capacity between seasons and indicate that its conduct is retaliation for a particular action so as to make clear the link between the deviation and the punishment (Decision, paragraph 152);

- the scope for de-racking or directional selling during the selling season to the detriment of an operator who has broken ranks in order to force it to sell a larger share of its holidays at discount prices (Decision, paragraph 152; see also paragraph 170).

197. It must first be observed that the characteristics of the relevant market and the way that it functions make it difficult for retaliatory measures to be implemented quickly and effectively enough for them to act as adequate deterrents.

198. Thus, in a case of deviation or, in other words, cheating, (where, for example, during the planning period one of the main tour operators attempted to turn to its advantage the overall capacity restriction resulting from parallel anti-competitive conduct), the other members of the oligopoly would find it difficult to detect the deviation because the market is not sufficiently transparent, as the Court has already held. It is difficult to detect any deviation at the planning stage, given the difficulties that a large tour operator has in anticipating the capacity decisions of its main competitors with any precision.

199. In that context, the deterrents identified by the Commission do not appear to be capable of coming into play.

200. In the first place, the Court finds that the Commission was wrong in concluding that the mere threat of reverting to a situation of oversupply acts as a deterrent. The Commission refers to the 1995 crisis to illustrate the effects of oversupply on the market. However, it should be made clear that the events of 1995 took place in a context different from that of the present case: then, all operators - regardless of whether they were large or small - boosted their capacity during the 1994 planning period in order to meet the increase in overall demand, which sectoral indicators and the preceding two years' growth suggested would occur. However, in this case the Commission anticipates that there will be a situation in which the three major tour operators, acting appreciably more cautiously than normal, will have reduced capacity below forecast demand and in which cheating has occurred. It is against that background, which differs markedly from the 1995 capacity surplus, that the Court must examine whether a possible return to oversupply acts as a deterrent. Oversupply could occur only one season later and only if the other members of the oligopoly decided to increase capacity above estimates of demand growth, that is very significantly in comparison with the level of under-supply that would exist in the context of tacit coordination envisaged by the Commission.

201. In the second place, the scope for increasing capacity in the selling season cannot act as a deterrent for the following reasons.

202. First, as the Decision itself emphasises, the market is distinguished by an innate tendency to caution as regards capacity decisions (see paragraphs 60 to 66, 97 and 136 of the Decision), given that matching capacity to demand is critical to profitability, since package holidays are perishable goods (Decision, paragraph 60).

203. Second, in this market a decision to depart from the common policy by increasing capacity in the selling season would be taken at a stage when it would be difficult to detect it in sufficient time. Furthermore, even if the other members of the oligopoly managed to expose the deviating conduct, any reaction on their part involving a retaliatory capacity increase could not be sufficiently rapid or effective, inasmuch as it could be implemented only to a very limited extent in the same season - as is implicity accepted in the Decision - and only subject to restrictions, which would become increasingly acute as the selling season progressed (in the best-case scenario, capacity for the forthcoming summer season could be increased by only 10% up until February) (see paragraphs 152 and 162 of the Decision).

204. Lastly, it may be assumed that, since they know that the perpetrators of any retaliatory measures are likely to find it difficult to sell late-added package holidays because of the low quality of such products (inconvenient flight times, poor-quality accommodation), the other members of the dominant oligopoly would be cautious about increasing capacity by way of retaliation. Capacity created in that way does not appear capable of competing effectively with capacity added by the operator which has broken ranks in the planning period, since it is both late and of lower quality. The deviating operator thereby benefits from the advantages associated with having acted first.

205. In the third place, as regards the possibility of increasing capacity in the following season and the fact that capacity can be added between seasons (final part of paragraph 152 of the Decision), it is appropriate to observe that increasing capacity in that way is unlikely to be effective as a retaliatory measure, given the unpredictable way in which demand evolves from one year to the next and the time needed to implement such a measure.

206. In the fourth place, retaliatory action by the other members of the oligopoly at the distribution level (by de-racking or directional selling) would - if Airtours were targeted - affect only 16% of its sales (of which less than 10% are made through Lunn Poly (Thomson) and only 6% through Thomas Cook). As the applicant points out, responses at secondary sources of supply do not represent countervailing forces of significance. Moreover, such retaliation would entail economic loss for its perpetrators, who would have to give up the commission paid by Airtours in respect of sales made in its main competitors' networks of travel agencies. Thus, the deterrent effect of such retaliatory action is not as significant as the Decision suggests.

207. It follows from the foregoing that the Commission erred in finding that the factors mentioned in paragraphs 151 and 152 of the Decision would, in the circumstances of the present case, be a sufficient incentive for a member of the dominant oligopoly not to depart from the common policy.

4. Underestimation of the likely reaction of smaller tour operators, potential competitors and consumers as a counterbalance capable of destabilising the alleged dominant oligopoly

208. The applicant claims that the Commission underestimated the likely reaction of smaller operators (also referred to as independent or secondary tour operators), potential competitors (in particular those offering long-haul foreign package holidays) and consumers as a countervailing force capable of counteracting the creation of a collective dominant position. Such a position may be created only if the major tour operators are in a position to act to an appreciable extent independently of other current or future competitors and consumers.

209. The Commission's response is that in order to overcome a coordinated capacity restriction on the part of the oligopolists, a large number of very small operators would need to increase their capacity significantly, which would not be possible given their existing size. Barriers to market entry and to growth beyond a certain size prevent smaller operators and new entrants from successfully challenging the power of the integrated tour operators and their ability to set capacity at a level lower than that of competitive equilibrium. Thus, secondary operators are not in a position to supply adequate capacity to meet extra demand because they encounter substantial barriers to expansion.

210. The Court observes in limine that, to prove conclusively the existence of a collective dominant position in this instance, the Commission should also have established that the foreseeable reactions of current and future competitors and consumers would not jeopardise the results expected from the large tour operators' common policy. In this case, that implies that where the large tour operators, for anti-competitive purposes, reduce available capacity to a level below what is required to adjust to anticipated trends in demand, such a reduction must not be offset by their current competitors, smaller operators, any potential competitors, tour operators with a presence in other countries or on the long-haul market, or their customers (United Kingdom consumers) reacting in such a way as to render the dominant oligopoly unviable.

(a) The possible response of current competitors: smaller tour operators

(i) Preliminary observations on the issue of the size of the smaller tour operators

211. At paragraphs 77 and 78 of the Decision, the Commission states that the ability of the fringe of smaller suppliers to compete effectively with the four large tour operators is further constrained by their lack of vertical integration and their small size, which means inter alia that they cannot make the same economies of scale and scope as the larger operators.

212. The Court must first observe in that regard that the Commission (defence, paragraph 103) accepts that, as the applicant's expert Professor Neven explained during the administrative procedure, the package holiday industry is one in which alternative business strategies may produce good results and one in which there is little room for operators of intermediate size. According to Professor Neven, undertakings may either operate on a small scale and buy on competitive markets the capacity which they need in order to supply package holidays (airline seats and hotel beds). Alternatively, they may decide to produce a large volume of package holidays. Those undertakings will nevertheless find it risky to buy in large quantities of capacity (particularly airline seats) on competitive markets, which is why it is necessary for them to become vertically integrated, at least in air transport services. That alternative business strategy does not necessarily lead to lower costs and a systematic competitive advantage over smaller undertakings. It is also intrinsically more risky than the strategy of remaining small and buying capacity on competitive markets.

213. However, it must be made clear that the issue here is not whether a small tour operator can reach the size necessary for it to compete effectively with the integrated tour operators by challenging them for their places as market leaders. Rather, it is a question of whether, in the anti-competitive situation anticipated by the Commission, the hundreds of small operators already present on the market, taken as a whole, can respond effectively to a reduction in capacity put on to the market by the large tour operators to a level below estimated demand by increasing their capacity to take advantage of the opportunities inherent in a situation of overall under-supply and whether they can thereby counteract the creation of a collective dominant position.

214. In those circumstances, if the Commission is to establish that smaller tour operators would be incapable of successfully countering the creation of a collective dominant position, it cannot confine itself to pointing out the fact (which is not disputed by the parties) that in the current state of the relevant market, in order to compete effectively with the integrated operators, a secondary operator must reach a minimum size enabling it to operate on a sufficiently large scale and must therefore achieve some vertical integration. The Commission's arguments seeking to stress the difficulties that smaller tour operators have in reaching the minimum size at which they are capable of competing effectively with the four large operators are thus immaterial to an assessment of the ability of smaller operators and new entrants to increase capacity in order to take advantage of the opportunities afforded by product shortages, which the Commission alleges would arise if the operation were approved.

215. Furthermore, as the applicant has pointed out, despite the fact that over the last decade a number of small tour operators have been taken over by the larger ones, small operators still exist in large numbers (several hundred), with continuous regeneration by new players entering the market, and continue to account for a significant part of the market.

216. It is against that background therefore that it is appropriate to ascertain whether smaller tour operators are capable in this instance of putting on sufficient additional capacity to counter any reduction in the capacity put onto the market by the large operators.

(ii) The ability of smaller tour operators to put on extra capacity

217. The Court notes in limine that in the present case the members of the alleged dominant oligopoly do not control individually or collectively the markets for the raw materials or services necessary for preparing and distributing the product concerned. In that connection, it is apparent from the Decision (paragraphs 5 to 42) that, as well as the market for short-haul foreign package holidays, the Commission also examined the effects of the concentration upstream, on the market for the supply of airline seats on short-haul charter flights, and downstream, on the travel agency market, but that it nevertheless did not conclude either that the merger would entail the creation of a collective dominant position by the three remaining competitors on the upstream or downstream markets or that the merged undertaking (Airtours/First Choice) would enjoy an individual dominant position.

218. First, it is noteworthy that the applicant has supplied, without challenge from the Commission, several examples of small tour operators who have put on additional capacity in response to opportunities that have arisen as a result of unexpected developments in the market. In 1996 (following the difficulties associated with the 1995 crisis), the three largest tour operators at that time reduced or froze their capacity, whilst several of the smaller operators underwent significant expansion, for example Virgin Holidays (+ 28%), Kuoni Travel (+ 20%), Direct Holidays (+ 68%) and Sun Express (+ 109%).

219. Second, the applicant has stated, without challenge from the Commission, that smaller tour operators tend to set capacity after the large operators have made their major capacity decisions and that they may still, to a certain extent and like any tour operator, increase their capacity subsequently.

220. Third, it is also apparent from the documents before the Court that several small operators have made it clear that they intend to increase their market share, which suggests that they are, on any view, extremely keen to make the most of any opportunities afforded as a result of the leading tour operators making capacity reductions unconnected with foreseeable trends in demand.

221. At paragraph 85 of the Decision, the Commission replies to the argument put forward by the applicant during the administrative procedure, to the effect that secondary operators like Cosmos and Virgin Sun should be regarded as likely future major competitors, since they intend to expand their business. The Commission replied, stating that none of these companies is likely to be able to challenge the major operators in the foreseeable future - because Cosmos/Monarch is heavily dependent on the large operators as purchasers of airline seats and is not vertically integrated into travel agencies, and because Virgin Sun's operations are at present very small and it does not have its own travel agencies either. Finally, Virgin Sun has had considerable difficulties in contracting for accommodation in key short-haul destinations.

222. However, the points made by the Commission do not lend weight to its argument, in that what matters is knowing how smaller tour operators are likely to react in the future in the event of the three remaining large tour operators reducing capacity put on to the market to below competitive levels. Rather, those points show that the clear intention of those two secondary tour operators is to take advantage of any opportunity afforded by the market.

223. First, the Commission is not entitled to rely on the fact that Cosmos (Monarch) currently tends to favour the large tour operators over the small ones as regards sales of airline seats in order to establish that, were capacity restricted to below a competitive level, Cosmos/Monarch would not put its own interests above those of the members of the alleged dominant oligopoly. In any event, the new managing director of Cosmos has stated that the intention is to increase the firm's market share from 3.5% to 5% in two years. To that end, Cosmos has ATOL licences to carry 1.1 million passengers (Air Travel Organisers' Licence, issued by the Civil Aviation Authority).

224. Second, for the same reasons, the difficulties that Virgin Sun has encountered in contracting for accommodation at certain short-haul resorts would be resolved if the large operators reduced their demand for rooms. It is clear from a letter dated 16 August 1999 from Virgin Sun to the Commission, placed before the Court in the context of measures of organisation of procedure, that it is easy to obtain adequate numbers of beds of sufficient quality in most of the Mediterranean resorts except the most sought-after, such as the Balearics, where the large operators more and more frequently negotiate long-term contracts with hotel owners, making it difficult for smaller operators to find the places required. However, it is also apparent from that letter that tour operators like Virgin Sun remain alert to try and take advantage of any opportunities that might arise in the most sought-after markets. Therefore, if, in any attempted capacity restriction, the large operators were not to use the beds contracted for, the smaller operators would rapidly be able to enter into contracts for them in order to increase the number of package holidays put onto the market. Lastly, it is appropriate to bear in mind that Virgin Sun is the short-haul tour operator recently launched by the Virgin Travel Group and has, since 1999, been offering holidays to the most popular resorts, namely Corfu, the Costa Blanca, the Costa del Sol, Majorca, Minorca, Ibiza, Portugal, Rhodes, Gran Canaria, Tenerife and Turkey, with departure flights from London Gatwick and Manchester, and that the President of the Virgin Travel Group has said that the goal of Virgin Sun is to match Thomson's market share within the next 10 years. Virgin Travel Group has ATOL licences to carry 400 000 passengers.

225. Finally, two competitors on the short-haul package holiday market with substantial financial resources at their disposal, British Airways Holidays (which held 375 000 ATOL licences in 1999) and Kuoni (which sold 230 000 package holidays in the United Kingdom in 1998), would also be capable of increasing capacity quickly if the large tour operators tried to engineer a collective dominant position.

226. Fourth, it is appropriate to mention a study indicating which of a selected number (59) of smaller tour operators are also present at 12 of the most sought-after short-haul destinations served by the large tour operators, which was produced during the administrative procedure and was not challenged by the Commission. The study shows (i) that all those destinations are served by at least four small tour operators; (ii) that the most popular, such as Corfu, Rhodes, Majorca or mainland Spain, are served by a large number of them (20 to 30 small tour operators); and (iii) that several small operators (such as Cosmos, Manos or Virgin Holidays) serve practically all the destinations (see Table 1 in Annex 8 to the application, the report by the applicant's expert, Professor D. Neven, entitled Competition in the UK Foreign Package Holiday Market, an Economic Analysis, July 1999). That study also shows that the small tour operators offer similar products (as regards number of nights and services) at prices that are comparable to, or even better than, prices offered by the larger tour operators.

227. Fifth, and contrary to the Commission's contention at paragraph 83 of the Decision, it is apparent from that study, which the Commission does not challenge, that the smaller tour operators normally manage to obtain accommodation at short-haul destinations on conditions similar to those of the large operators. The study examines 20 hotels in popular short-haul resorts and compares the prices paid by Airtours with those paid by Panorama and Direct, two small independent operators which were subsequently taken over by Airtours, and shows that the prices are similar and that in some cases the smaller operators obtained more favourable terms than Airtours, even though Airtours reserved many more nights than the smaller operators.

228. It follows that, in the case under consideration, the smaller operators would try to put on additional capacity. However, an analysis of whether it would be possible for them to do so calls for a more specific examination of whether adequate access to the markets for airline seats and travel agencies is available to them.

(iii) The small operators' access to airline seats

229. The Decision states that the smaller tour operators do not have adequate access to airline seats and that the merger would make that situation worse (Decision, paragraphs 78 and 79 and the final part of paragraph 83). Their small size means that they cannot obtain the benefits of scale and scope of the large operators. For example, they cannot guarantee a charter airline a complete planeload of passengers (except, perhaps, for a few days in peak season). That increases the risk for the airline that it will have to operate the flight at less than optimum loading. Consequently, the airline is likely to charge the small operators a higher seat price than the larger ones to reflect the higher risk (Decision, paragraph 78). Small operators have stated that they already have difficulty obtaining seats at desirable times (especially weekends) and from the major tourist airports (Gatwick and Manchester). Tour operators (and airlines) have commented that they need to offer departures from both those airports in order to have access to the main customer centres and so provide a credible national operation, otherwise their prospects for expansion beyond that of a small-scale competitor are slight (Decision, paragraph 79).

230. At paragraph 80 of the Decision, the Commission adds that the major operators already have considerable market strength in regard to seat sales to independent operators. For example, a tour operator has commented that Monarch, the only substantial supplier to the independent sector that would remain after the merger, already tends to satisfy the needs of the major operators (which together account for over half of its third-party sales) before considering what to offer the independents, and refused even to discuss the forthcoming year's programme with the operator concerned until it had received notice of the majors' requirements.

231. The Court observes, first, that it is apparent from Table 2 in the Decision (paragraph 159), which sets out the market shares of the main suppliers of airline seats to third parties (data which include all sales to third parties, encompassing sales to and between large tour operators as well as sales to small operators), that after the merger Airtours/First Choice would be in a position to control less than a quarter of the airline seats supplied to third parties and that the three large tour operators as a whole would supply less than half, given that Thomson has only a very limited presence on that market. It follows that the essential needs of third parties for airline seats would continue to be met by players who are independent of the large tour operators. That situation offers certain safeguards to the small tour operators, since only two out of the three large operators have any significant presence on that market and third-party independents represent an important source of airline seats.

232. There is no evidence that that situation would be substantially altered as a result of the concentration, contrary to the Commission's contention that the merged entity consisting of the applicant and First Choice would be likely further to rationalise airline seats and that the small operators' problems would be exacerbated by a reduction in the number of seats available. As the applicant claims, the merger would not adversely affect the availability of airline seats for third parties: if Airtours and First Choice took more of their seat capacity from within the post-concentration group (which would displace third-party tour operators currently flying on Airtours International and Air 2000), there would be a corresponding release of seats on third-party airlines previously taken up by Airtours and First Choice. The Commission itself adopted that line of argument in its decision of 8 March 1999 (Case IV/M.1341 Westdeutsche Landesbank/Carlson/Thomas Cook (OJ 1999 C 102, p. 9), paragraph 36), in which it stated that [t]o the extent that the combined entity Thomas Cook might re-orientate its strategy towards using in-house charter airline capacity (eg that available from Caledonian) rather than purchase capacity from third parties, this would liberate that capacity from third parties and render it available to customers who have hitherto purchased from Flying Colours or Caledonian. The Commission has not put forward any persuasive arguments to show that the logic underlying that argument is no longer valid in this instance.

233. Similarly, as regards the points made at paragraph 80 of the Decision, it is sufficient to note that the applicant has maintained - without challenge from the Commission during the procedure before the Court - that Monarch has given evidence that it does not favour the large tour operators to the detriment of the smaller tour operators and that it has recognised that the Thomas Cook/Carlson merger in fact increased its dependence on third-party tour operators without their own charter airline, as the Airtours/First Choice merger would have done.

234. Second, the Court observes that, as the applicant has submitted, the evidence of Hunt & Palmer, one of the main seat brokers, whose business it is to match supply and demand by selling to tour operators spare capacity that airline carriers wish to sell (Annex 39 to the application), shows that the small tour operators may obtain airline seats for a season (or a shorter period) with weekend departures from four sources: overseas carriers; scheduled airlines; low-cost carriers; independent charter airlines based in the United Kingdom. It should be pointed out that there are at least 15 independent seat brokers in the United Kingdom and that the Commission has not disputed the evidence concerned.

235. The arguments put forward by the Commission to show that those sources of supply are not viable are not persuasive.

236. The first source is overseas carriers based at the destination airport (for example, Spanair, Air Europa or Futura).

237. The Commission's view is that those carriers are not a viable alternative, since they have difficulty in obtaining sufficient slots at convenient times at the major United Kingdom airports, in particular at Gatwick. Moreover, as their fleets are not based in United Kingdom airports, the aircraft must fly to the United Kingdom in the morning and make the return flight in the evening, which means that customers have to fly out late in the evening and return early in the morning. A schedule of that kind considerably reduces the time actually spent on holiday, which is not popular with consumers. However, that argument is rebutted by Hunt & Palmer's evidence, which is to the effect that rotations may be effected from Gatwick.

238. In any event, what is at issue here is whether, against a background of under-supply of package holidays, the smaller tour operators could obtain additional airline seats on reasonable terms and not whether they can gain access to better departure airports and slots. In that regard, it should be observed that the Commission deemed it inappropriate to subdivide the United Kingdom markets for short-haul package holidays and for the supply of seats on charter flights to tour operators more narrowly by reference to, for example, region or departure airport. On the contrary, in that regard the Commission found in the Decision (paragraph 45) that there was relative uniformity of pricing and costs, which suggests that there is a sufficient degree of overlap between the possible regional or local markets for them to be regarded for the present purpose as constituting a single national market on the demand side (if a chain of substitution basis is adopted). The Commission reached that conclusion (paragraph 45) after having pointed out that consumers prefer to leave from an airport closer to their home and that landing fees and other related factors mean that prices for departure from some of the smaller regional airports are often higher than those from the main holiday airports (London Gatwick and Manchester), and having concluded that the premium charged (or the discount offered) is usually relatively small compared to the total holiday cost, in particular when account is taken of the additional cost of the road journey to a cheaper airport further away. The same is true on the supply side, since the Commission believes that tour operators market their products nationally (Decision, paragraph 45) and that the latter do not vary much, either in price or otherwise, for consumers in different regions. Furthermore, tour operators and air carriers can, as a general rule, easily move planes and flights between the various airports, with the exception of Gatwick where availability is limited (Decision, paragraph 46).

239. The Court holds that the Commission's argument that the aircraft used by overseas carriers must usually fly to the United Kingdom in the morning and return in the evening, which is inconvenient for consumers, has no factual basis, since the average flight-time to a European destination is about two hours. Carriers based at destination airports can thus carry out several rotations in one day and can, for example, make a Spain/United Kingdom outbound flight and a United Kingdom/Spain return flight in the morning and a Spain/United Kingdom outbound flight and a United Kingdom/Spain return flight in the evening.

240. Finally, it must be observed that in the course of the administrative procedure the applicant maintained, without challenge from the Commission, that carriers based at destination airports supplied over one million seats in 1998 (the last year for which figures were available at the time of the Decision) for package holidaymakers and for seat-only holidaymakers and that the number of seats supplied by destination-based carriers has grown rapidly over recent years.

241. It follows that, contrary to the Commission's contention, those carriers could play a significant role if the small operators attempted to increase the number of package holidays should the opportunity arise.

242. The second source is scheduled airlines (like Debonair, Flightline or Cityflyer), whose weekend loading is light in the absence of business travellers.

243. The Commission regards seats supplied by scheduled airlines to be of little importance in the United Kingdom, where British Airways uses only a small proportion of its capacity for this type of flight. The reasons for that are higher prices, the fact that scheduled airlines do not have direct flights to holiday destinations, the lack of availability and the inflexibility of timetables.

244. However, those factors are not significant obstacles to small tour operators who wish to increase their capacity. As regards price differentials, it should be pointed out that air transport costs account for only a small proportion of the cost of a package holiday. For example, the cost of a Liverpool-Malaga flight in August on EasyJet was GBP 108 and that of a Stansted-Malaga flight on Go was GBP 140, whilst the cost of a 14-day package holiday to Marbella, Spain, in August was GBP 1 598 from Virgin Holidays, GBP 1 698 from Bath Travel and GBP 1 738 from Airtours (application, Annex 8, Table 2, and Annex 40). The proportion accounted for by transport amounts, in any event, to less than 10% of the price of the package. In that connection, the applicant submitted a table (Table 5 in the expert economic report, Annex 8 to the application) during the administrative procedure, which compared prices charged for scheduled flights and charter flights to different destinations on various dates. That table was drawn up on the basis of data acquired from Panorama, a small tour operator recently taken over by Airtours. It is likely that similar data could be obtained from other small tour operators. It is apparent from the table that the difference in price varies from GBP 20 to GBP 30, which ultimately has only a very minor impact on the price of the package and thus on the competitiveness of smaller operators using scheduled flights. In that regard, the price differential can be accounted for in the main by the obligation to pay departure taxes in the case of scheduled flights.

245. As regards the terms that may be offered by scheduled airlines, particularly as regards days and slots, it is noteworthy that two of the five examples cited by the applicant, on the basis of data obtained from Panorama for the period before it was taken over, show that it is possible to obtain departures on Saturday or Sunday. Likewise, Hunt & Palmer's statement draws attention to the fact that the whole point of a seat broker's business is its ability to find flights departing at the weekend. Furthermore, the Decision itself (footnote 38) reveals that British Airways offers some whole-plane charters at weekends from United Kingdom regional airports, using aircraft not required for scheduled services at those times. As regards departure airports, it is appropriate to refer to the Court's earlier finding in relation to overseas air carriers. Finally, as regards the question whether it is disadvantageous to buy seats in only part of an aircraft rather than a full aircraft, the Court observes that the examples of prices offered by seat brokers, which the applicant produced, show that the price differential is minimal (less than 10%) and that the price of a seat bought as part of an allocation of some of the seats on an aircraft can prove to be less than the price of a seat purchased as part of a full aircraft allocation (see Professor Neven's report at Annex 8 to the application).

246. As to the fact that only a limited number of destinations are covered by scheduled airlines, it is noteworthy that apart from the main tourist destinations in Spain, the scheduled airlines cited by the applicant also fly to the south of France and to Italy. In any event, the overseas carriers appear to be in a position to offset, should the need arise, the scheduled airlines' failure to provide a service to other destinations.

247. It follows from the foregoing that, contrary to the Commission's contention, scheduled airlines can be used by the small tour operators to increase their capacity effectively so as to counter any restrictions imposed by the leading tour operators.

248. The third source is low-cost carriers (such as Ryanair or Go), whose capacity has increased substantially in recent years and which are in a position to offer tailor-made services. The applicant produced a map showing the main destinations to which low-cost carriers fly (application, Annex 40), from which it can be seen that the main resorts on the Spanish Mediterranean are served by at least one airline, and often by two or even three airlines: Barcelona (by Debonair from Luton, EasyJet from Liverpool and AB Airlines from Gatwick), Palma (by EasyJet from Luton and Go from Stansted), Ibiza (by Go from Stansted); Alicante (by Go from Stansted); and Malaga (by EasyJet from Liverpool and Go from Stansted).

249. The fourth source is independent charter airlines based in the United Kingdom (such as Monarch but also European Air Charter, British World or Titan), which also use small aircraft with low running costs. However, given that those airlines represent over 50% of the available capacity in the market for the supply of charter airline seats to third parties (see table 2 in the Decision, paragraph 159), it must be concluded that, contrary to the Commission's contention, they must have adequate capacity to act as a credible source of supply for small operators.

250. Finally, the Commission has not taken into consideration the fact that it is crucial for the integrated tour operators to make sure that their planes are full if they are to ensure that their business is viable. The fact that their fleets represent a very significant part of their fixed costs means that there will be an incentive for the large operators to offer the small operators the seats that will remain empty in the situation envisaged by the Commission, in which the large operators appreciably reduce capacity.

251. It is clear from the foregoing that the Commission was wrong to conclude that smaller tour operators would not have access to airline seats on favourable enough terms to attempt to increase capacity and take advantage of the opportunities afforded by the under-supply that would occur in the anti-competitive environment anticipated by the Commission in the event of the operation being approved.

(iv) Access of smaller tour operators to distribution

252. At paragraphs 81 and 82 of the Decision, the Commission refers to certain difficulties encountered by smaller tour operators, such as the discriminatory conditions to which they are subject in the travel agencies owned by the large integrated tour operators, notably commission rates, directional selling of the latters' products and the way in which brochures and offers are displayed, all of which prevent them from engaging in effective competition with the large tour operators.

253. The applicant challenges the Commission's proposition that access to distribution has been restricted as a result of vertical integration by the leading tour operators because, in the applicant's submission, vertical integration produces an unacceptable foreclosure effect only if horizontal market power exists at the distribution level, which is not the case here.

254. The Commission contends that the main tour operators control all the large national chains of travel agents and the great majority of agency branches. In those agencies other operators are subject to directional selling, that is, preference is given to the parent company's products. That is a substantial barrier to the small operators' market entry and growth. Alternative distribution channels such as distribution via call centres and the Internet are not yet viable substitutes for travel agencies, as can be seen from the efforts made by the main tour operators, such as First Choice, to establish or acquire chains of travel agencies.

255. The Court observes that what is at issue here is not how large an operator needs to be to compete with the main operators for market leadership but whether, in the anti-competitive environment anticipated by the Commission, small operators already present on the market would be able to gain access to the market for distribution on satisfactory terms and sell larger numbers of their package holidays to consumers. It is apparent from a cursory analysis of the Decision that, as the applicant maintains, that would be the case.

256. First, it is noteworthy that, as the Commission has pointed out at paragraph 32 of the Decision, the merger is not likely to lead to the creation or strengthening of a dominant position in travel agency services as a whole, irrespective of whether direct sales of package holidays by tour operators or distance sales (by telephone) are included. As regards numbers of outlets, the parties' combined share of the supply of travel agency services is small (around 15%).

257. Further, it is clear from the Decision (paragraph 81) that nearly 40% of all foreign package holidays sold through travel agencies are sold through independent agencies. Second, the Commission itself acknowledges (paragraph 31) that other methods of distribution exist and are developing well, such as direct sales by telephone or the Internet, and already represent around 20% of total sales of package holidays. That increases the scope for small operators to distribute their products efficiently in a situation where supply is restricted. In that regard, it is significant that Direct Holidays (an independent tour operator taken over by Airtours), which sells all its holidays by direct means, enjoyed marked growth during the period 1995-1996 (a period in which the large tour operators experienced financial difficulties) (application, paragraph 9.18). It should be added in this respect that the applicant has drawn attention to the fact, which is not disputed by the Commission, that during the administrative procedure, the Commission received the following evidence from third parties regarding the feasibility of direct sales as a means of market access (cited at paragraph 3.57 of the reply to the statement of objections at Annex 7 to the application):

(i) Thomas Cook indicated that the current trend is away from traditional ways of booking a holiday through a travel agent in person. The British National Travel Survey shows that indirect booking has grown since the end of the '80s, going from 29% of total bookings in 1992 ... to 34% in 1998. (ii) Thomson expressed the view that the number of consumers who book holidays through non-traditional direct purchase methods is growing, as is the proportion of consumers who regard direct purchases as an alternative to booking through a traditional high-street travel retailer.

(iii) Virgin Holidays said: as a tour operator, we do not have a distribution chain. From our own sales, we have experienced a marked increase in the number of holidays purchased via call centres. We have also seen an increase in the number of holidays that are purchased through teletext agencies.

258. Even if it were established, in relation to the difficulties encountered by small tour operators (mentioned at paragraphs 81 and 82 of the Decision), that the practices concerned actually occur and are not unlawful, they would not appreciably limit the ability of small tour operators to take advantage of any opportunities afforded by the under-supply to which the Commission believes the merger would give rise. In such a situation, it may be concluded that, given consumer expectations and the need to maximise income, travel agencies could not refrain from offering the small tour operators' products on reasonable terms, even though the agencies owned by the vertically integrated tour operators would offer the group's products ahead of competing products.

259. In any event, since nearly 40% of all package holidays sold are not sold in agencies controlled by the large tour operators, smaller operators are likely to gain access to distribution on favourable enough conditions to enable them to sell the capacity that they would add in the event of the main operators deciding to restrict capacity to below competitive levels.

260. It follows that the Commission was wrong to conclude that smaller tour operators would not have access to a channel through which to distribute their products to consumers on favourable enough conditions to enable them to expand their capacity significantly in order to take advantage of opportunities afforded by the under-supply which, in the Commission's submission, would occur were the merger to be approved.

261. It is apparent from all of the foregoing that the Commission underestimated the ability of the small operators to increase capacity in order to take advantage of opportunities afforded by a situation of general under-supply brought about by the large tour operators and thus to counteract the creation of a collective dominant position following the concentration.

(b) Possible reactions of potential competitors: other tour operators

262. It is also necessary to consider whether, were the large tour operators to restrict capacity put on to the market to anti-competitive levels, tour operators in other countries of the Community or in the United Kingdom long-haul foreign package holiday market would be capable of entering the United Kingdom short-haul foreign package holiday market.

263. It is appropriate to recall the wording employed by the MMC in its 1997 report:

Players come and go. There are no significant barriers to entering either the tour operator or the travel agent market (paragraph 1.6). ... if particular types of holidays, holidays from certain airports or holidays at particular times of year were overpriced, then tour operators would be able to move their business into each of those areas and undercut their prices (paragraph 4.15).

264. It should be observed that the Commission, at paragraph 114 of the Decision, recognises, first, that a collective dominant position cannot be sustained in the long term if barriers to market entry into tour operation, charter airline operation and the travel agency business are insignificant and, second, that the MMC's 1997 report broadly concurs with the applicant's view on the lack of barriers to entry to the market concerned.

265. At paragraph 115 of the Decision, the Commission nevertheless notes that since the MMC's report was completed in 1997 there has been substantial consolidation in the industry and considers that henceforth barriers to market entry will be greater (they are likely to have a more significant impact) and that they would increase still further if the proposed merger were implemented. The Commission then submits:

To be sufficient to remove the threat of creation of a dominant position, entry must, clearly, be more than merely possible. Among other things, it must be sustainable, which, in markets such as this one, where scale is an important factor, means that it must be capable of being on, or quickly acquiring, a sufficient scale to offer a real competitive challenge to the dominant suppliers. In the Commission's view, this is unlikely to be the case here.

266. It should nevertheless be borne in mind that, as is the case with current competitors, what is important here is not whether there is scope for potential competitors to reach a sufficient size to compete on an equal footing with the large tour operators, but simply whether there is scope for such competitors to take advantage of opportunities afforded by the large operators restricting capacity put onto the relevant market to below a competitive level. In that context, the Commission cannot contend that, merely because they would have difficulty expanding beyond a certain size, tour operators offering other products (such as long-haul foreign package holidays) or carrying on business in other countries (such as Germany or the Netherlands) could not enter the United Kingdom short-haul foreign package holiday market fast and effectively if the large tour operators decided to restrict competition significantly. In that respect, it is noteworthy that other sizeable European tour operators, such as Neckermann and TUI, are mentioned by the applicant as potential competitors likely to enter the United Kingdom quickly in the event of a capacity restriction or a price increase.

267. Furthermore, in that connection, the Decision does not examine competition at the level of holiday accommodation, although the supply of capacity of that type is very important if the dynamic of the relevant market is to be understood, in particular with regard, first, to the ability of members of the alleged dominant oligopoly to act independently of hotel owners at short-haul destinations, and, second and consequently, to the ability of current and potential competitors to react to a possible reduction in the capacity supplied by the large tour operators. It is unlikely that any hotel beds becoming available following a decision by the large tour operators to restrict capacity will not be immediately booked by other operators. Documents and statements from other operators, submitted during the administrative procedure, show that they are willing to acquire accommodation capacity (see, for example, the letter from Virgin Sun referred to in paragraph 224 above).

268. Therefore, each large tour operator is likely to take account of the risks involved where hotel owners react to an appreciable reduction in bed reservations which does not actually correspond to a reduction in demand but to a decision to restrict capacity for anti-competitive purposes. The large operators might find that beds were not available for subsequent seasons on satisfactory terms or in satisfactory numbers.

269. It is clear from the foregoing that, although the Commission examined barriers to developing beyond a certain size in the market, it did not take account, as it should have done, of the fact that the lack of barriers to market entry is likely to allow potential competitors to gain access to, and offer their products on, the relevant market and, therefore, to take fast and effective action in the event of the large tour operators aligning their capacity strategies so as to give rise to a situation of under-supply.

(c) The possible reaction of consumers

270. In seeking to prove that the oligopoly emerging after the transaction would be able to act independently of consumers, it is necessary to determine what the reaction of United Kingdom consumers would be and to ascertain whether they would be prepared to look for other options if the price of short-haul package holidays were to rise significantly or if there were a dearth of such holidays.

271. At paragraph 124 of the Decision, the Commission explains that consumers have no buyer power and that, if that is taken with other aspects of the market, they have difficulty in comparing competing products from the limited information available in tour operators' brochures, which limits the consumer's ability to offset any anti-competitive features on the supply side.

272. According to the applicant, several market studies show that the majority of holiday makers visit more than one travel agency before making a holiday choice and that for 85% of them price is the single most important factor affecting their purchase decision. Individual consumers are therefore able to vote with their feet and look for a cheaper holiday, thus providing the tour operator with an incentive to price competitively.

273. The Commission contends that it is wrong to claim that in a consumer products market such as the package holiday market consumers have any significant countervailing buyer power.

274. However, it is appropriate to emphasise that the fact that consumers do not have significant buyer power because they act in isolation must not be confused with the question of whether they would be able to react to a price rise brought about by the large tour operators restricting capacity put onto the market to an anti-competitive level. As the applicant submits, it is not disputed that consumers make comparisons before purchasing a holiday. The Commission itself admits, at paragraph 98 of the Decision, that consumers are sensitive to relatively small differences in the prices of similar holidays.

275. In that context, the Commission has underestimated the role that might be played by United Kingdom consumers, who are in a position to try to obtain better prices from small tour operators.

276. Furthermore, it is appropriate to point out that in the context of the first plea it was found that, although the Commission, in the proper exercise of its discretion, concluded that the relevant product market should be narrowly defined, it nevertheless did not question either the fact that long-haul foreign package holidays are becoming increasingly attractive to consumers or the fact that the market studies cited by the applicant in its reply to the statement of objections (Annex 7 to the application: the British National Travel Survey and Mintel Holidays: the booking procedure, 1997) draw attention to the tendency of United Kingdom consumers to go further afield for their holidays, in particular to the other side of the Atlantic. That fact lends weight to the applicant's proposition that demand might partly switch to other types of holidays if there were sufficient price convergence, inasmuch as the studies concerned clearly show that consumer tastes are evolving and that consumers do not appear in any way to regard the Mediterranean coast as the only place to go on holiday.

(d) Conclusion

277. In view of the foregoing observations, the Court concludes that the Commission's assessment of the foreseeable reaction of smaller tour operators, potential competitors, consumers and hotel-owners was incorrect and that it underestimated their reaction as a countervailing force capable of counteracting the creation of a collective dominant position.

5. The assessment of the impact of the transaction on competition

278. The Commission sets out its findings on the effect of the transaction at paragraphs 139 to 147 of the Decision.

279. First (paragraph 139), it argues that the transaction would result in increased concentration, since the combined market share of the three leading tour operators would increase substantially: 83% on the Commission's calculations (85% according to Nielsen), compared with around 70% before the merger (for Airtours, Thomson and Thomas Cook). In addition, the share of the fourth operator (Cosmos) would be less than 5%, whereas the operator currently in fourth position (First Choice) has a market share of 11%. However, it is apparent from paragraphs 139 to 147 of the Decision that the Commission did not regard the fact that the combined market share was high (above 80%) as sufficient to establish that there was a collective dominant position.

280. Second, the Commission maintains (paragraphs 140 and 141 of the Decision) that as a result of the merger First Choice would be lost as a supplier/distributor for secondary operators, which would further marginalise the smaller independent and non-integrated tour operators. However, it is appropriate to point out that when consideration is given to whether a collective dominant position might be created, the assessment of the foreseeable impact of the operation on other competitors in the market must ascertain whether those competitors would be in a position to challenge the stability of the alleged dominant oligopoly. The Court has found that the Commission did not prove that they would be incapable of doing so.

281. Third, the Commission contends (Decision, paragraphs 142 to 147) that the merger would increase transparency and the degree of interdependence between the large tour operators. At paragraph 143 of the Decision, it concludes that the fact that the merger would reduce the number of competitive relationships that are possible among the major operators by half, from six to three, would increase significantly the interdependence of the members of the oligopoly, which would be a further incentive to them to restrict capacity because it would be much more clear to them that competing for market share would mean depressed profits for each of them. The further marginalisation of secondary operators would increase the likelihood of that happening. At paragraph 144, the Commission points out that the reduction in the number of competitive and cooperative bilateral relationships would thus increase the transparency of the market, since it would become much easier for one of the major suppliers to detect efforts to destabilise the market involving, for example, attempts to capture market share. That greater transparency would increase the risk that competitive offensives would create oversupply, which would depress profits and consequently be counter-productive.

282. Consequently, the Commission reached the conclusion (Decision, paragraph 147) that the market structure to which the operation would give rise would make it rational for the oligopolists to restrict supply.

283. However, it is appropriate to bear in mind, as regards the level of market transparency foreseeable after the merger, that the Court has found that the Commission was wrong to conclude that the degree of market transparency was sufficient to allow each of the major tour operators to be aware of the conduct of the others, to detect any deviations from the common policy and to see retaliatory measures for what they are. The Commission has failed to establish that the situation would be any different if there was a move from four major tour operators to three. Although there would certainly be some increase in market transparency after the reduction in the number of competitive bilateral relationships between the major tour operators from six to three, it is nevertheless the case that each of the three remaining major tour operators would continue to find it difficult to anticipate the intentions of the other two in sufficient time and to see any deviations from the common policy for what they are.

284. As regards the finding that the operation would significantly increase the interdependence of the major tour operators, it should be noted that the Commission has been inconsistent: it has argued, on the one hand, that in the relevant market it is necessary to become vertically integrated in order to be genuinely competitive and, on the other, that the fact that each of the integrated tour operators sells seats on charter flights to the others in the upstream market and sells the others' package holidays in the downstream market has an anti-competitive effect inasmuch as it increases their interdependence. In the absence of any evidence to the contrary, there must be a presumption, given the logic of the way the market operates, that vertical integration makes the large tour operators more independent from each other and thus reduces their interdependence.

285. Similarly, the Commission has not explained why what it regards as commercial links (purchase of airline seats from the others and sale of its own products in agencies owned by the others) must be explained solely in terms of strong economic links between the major operators (Decision, paragraph 142) and cannot simply be explained on the ground that it is profitable to maintain those links in a competitive situation, given that the major tour operators are economic units, firmly entrenched in several markets within the industry and that it is in their interest to be profitable and to maximise their revenues in those markets as a whole.

286. The Decision is not specific about the strong economic links between the major operators or the way in which they increase the interdependence of the integrated tour operators. At paragraph 57 of the Decision, the Commission states that the extent and nature of the vertical integration of the major suppliers, the extensive commercial and other links between them are among the characteristics which distinguish the conditions of competition in the relevant product market. Then, at paragraph 71, the Decision is a little more specific about the nature of the links in question. It points out that there are a number of commercial links between the integrated companies, deriving in part from their vertical integration, downstream through the use of each other's travel agency chains and upstream because they share airline capacity to some extent, both through direct purchase from each other and from swaps and consolidation arrangements whereby they maximise the use of their respective fleets. At paragraphs 102 to 113, the Commission sets out a number of considerations under the heading Transparency, interdependency and commercial links. Those paragraphs are devoted to an exposé of the Commission's view of the degree of market transparency. The Commission, in that regard, states that vertical integration and the commercial links between the main suppliers help them to obtain precise and up to date estimates of their market share and those of their competitors. However, those paragraphs explain neither the reasons for the tour operators' interdependence nor what effect is produced by the commercial links arising from vertical integration and the way the market operates in that regard (apart from the fact that there is increased transparency).

287. Next, at paragraph 142, in its assessment of the impact of the merger, the Commission states that there is already a certain degree of mutual dependency between the tour operators, owing to the impact on market conditions of the overall level of capacity put on to the market for a season. The Commission adds that this creates strong economic links between the major operators. However, the kind of economic links to which that passage refers is not specified and the Decision does not explain what the strong economic links are. In any event, the Decision does not seem to be referring in that passage to the commercial links stemming from vertical integration (namely, the fact that each of the large tour operators purchases airline seats from the others and sells the other operators' holidays).

288. It is clear from the foregoing that the Commission did not examine to what extent in the pre-concentration situation the commercial links stemming from vertical integration and from the way the market operates increase the interdependence of the integrated tour operators, other than to point out that they increase the degree of market transparency.

289. As the Commission has not provided evidence to the contrary, there must be a presumption that in the conditions obtaining in the relevant market prior to the concentration the fact that each integrated tour operator buys airlines seats from, and sells its products to, companies owned by a competitor no more constitutes evidence of interdependence than it does of independence. It seems to be merely an aspect of the way the economy operates, where business prevails and the integrated tour operators must attempt to capitalise on capacity and business opportunities in an industry with very high fixed costs and low profit margins. As the Decision points out, the integrated tour operators are present in three markets and, therefore, in three different businesses: short-haul charter flights, short-haul package holiday operation and holiday sales in travel agencies. First Choice even carries on a fourth business, that of seat broking (see paragraph 1 of the Decision). The economic logic underlying a group of undertakings demands that each of the undertakings making up the group strives to be as successful as possible.

290. In that regard, it is appropriate to observe that in its assessment of the impact of the transaction, the Commission does not appear to have considered the effect of the economic logic of the group, namely maximising income by maximising overall profits for the group as a whole. The Decision does acknowledge (paragraph 59) that tour operators' margins are fairly low, of the order of 7% in recent years, and that, by contrast, vertically integrated operators will normally also receive income from their airline and travel agency activities, areas in which (particularly so far as airlines are concerned) the margins may be higher. It also acknowledges that for that reason gross margins on the total operations of the integrated operators may be larger than those on their tour operating activities alone.

291. Since that economic logic tends to favour achievement of the greatest synergies possible, the profitability of the various activities of the group (charter airlines, tour operating and travel agencies) will be correspondingly higher where the benefits of vertical integration are fully exploited.

292. Finally, even if the synergies to which the merger is expected to lead did not exceed 1% of the overall costs of the combined entity (Decision, paragraph 146), there is no evidence that Airtours decided to finance the cost (usually higher in the case of a hostile takeover bid) of the shares in First Choice by relying on that large investment being made profitable by the benefits derived from a sustainable collective dominant position.

293. In view of the foregoing considerations and without a more detailed assessment of the implications of the increased market transparency and interdependence of the major tour operators to which the transaction is likely to give rise, the Court finds that the Commission has failed to prove that the result of the transaction would be to alter the structure of the relevant market in such a way that the leading operators would no longer act as they have in the past and that a collective dominant position would be created.

D - General conclusion

294. In the light of all of the foregoing, the Court concludes that the Decision, far from basing its prospective analysis on cogent evidence, is vitiated by a series of errors of assessment as to factors fundamental to any assessment of whether a collective dominant position might be created. It follows that the Commission prohibited the transaction without having proved to the requisite legal standard that the concentration would give rise to a collective dominant position of the three major tour operators, of such a kind as significantly to impede effective competition in the relevant market.

295. In those circumstances, the third plea must be declared to be well founded and, therefore, the Decision must be annulled, without it being necessary to examine the other complaints and pleas put forward by the applicant.

Costs

296. Under Article 87 (2) of the Rules of Procedure, the unsuccessful party is to be ordered to pay the costs if they have been applied for in the successful party's pleadings. Since the defendant has been unsuccessful and the applicant asked for an order for costs against the Commission, the Commission will be ordered to bear its own costs and to pay those incurred by the applicant.

On those grounds,

THE COURT OF FIRST INSTANCE (Fifth Chamber, Extended Composition)

hereby:

1. Annuls Commission Decision C(1999) 3022 final of 22 September 1999 declaring a concentration to be incompatible with the common market and the EEA Agreement (Case IV/M.1524 - Airtours/First Choice);

2. Orders the Commission to pay its own costs and those incurred by the applicant.