EC, July 24, 2002, No M.2706

THE COMMISSION OF THE EUROPEAN COMMUNITIES,

Having regard to the Treaty establishing the European Community, Having regard to the Agreement on the European Economic Area and in particular Article 57 thereof, Having regard to Council Regulation (EEC) No 4064-89 of 21 December 1989 on the control of concentrations between undertakings (1), as last amended by Regulation (EC) No 1310-97 (2), and in particular Article 8 (2) thereof, Having regard to the Commission Decision of 11 April 2002 to initiate proceedings in this case, Having given the undertakings concerned the opportunity to make known their views on the objections raised by the Commission, Having regard to the opinion of the Advisory Committee on Concentrations (3), Having regard to the final report of the Hearing Officer in this case (4), Whereas:

I. INTRODUCTION

(1) On 27 February 2002, the Commission received notification pursuant to Article 4 of Council Regulation (EEC) No 4064-89 (the Merger Regulation) of a proposed concentration by which the United States-based undertaking Carnival Corporation (Carnival) intends to acquire control within the meaning of Article 3 (1) (b) of the Merger Regulation of the British undertaking P& O Princess plc (POPC) by way of a public bid announced on 16 December 2001.

(2) After examination of the notification, the Commission concluded that the notified operation fell within the scope of the Merger Regulation as amended and raised serious doubts as to its compatibility with the common market and the functioning of the EEA Agreement. On 11 April 2002, the Commission decided, in accordance with Article 6 (1) (c) of the Merger Regulation and Article 57 of the EEA Agreement, to initiate proceedings in this case.

(3) Following a thorough investigation of the case, the Commission has now come to the conclusion that the proposed concentration will not create or strengthen a dominant position as a result of which effective competition in the common market and the functioning of the EEA Agreement would be significantly impeded.

II. THE PARTIES

(4) Carnival is a company active worldwide in maritime cruise operation, offering a broad range of cruise brands through Carnival Cruise Lines (CCL), Holland America Line (HAL), Costa Cruises (Costa), Cunard Line (Cunard), Seabourn Cruise Line (Seabourn) and Windstar Cruises (Windstar). Carnival operates around 40 ships in various geographic areas around the world.

(5) POPC is also a company mainly active worldwide in maritime cruise operations, with around 20 ships, marketed under the following brands: Princess Cruises (Princess), P& O Cruises, Swan Hellenic, Aida Cruises (Aida), Arkona, and A'Rosa. POPC is also active in river cruises. Carnival and POPC are collectively referred to hereinafter as "the Parties"(5).

III. THE OPERATION AND THE CONCENTRATION

(6) On 16 December 2001, Carnival announced a unilateral preconditional public offer to acquire all the shares of POPC. Carnival intends to acquire 100 % of POPC. Consequently, the notified operation constitutes a concentration within the meaning of Article 3 (1) (b) of the Merger Regulation.

IV. COMMUNITY DIMENSION

(7) The undertakings concerned have a combined aggregate worldwide turnover of more than EUR 5 billion (6) (Carnival: EUR 4,05 billion; and POPC: EUR 2,62 billion). Each of them has an aggregate Community-wide turnover in excess of EUR 250 million (Carnival: EUR [500 to 750] *(7) million; and POPC: EUR [500 to 750] * million). POPC achieves more than two-thirds of its aggregate Community-wide turnover in the United Kingdom, but this is not the case for Carnival. The notified operation therefore has a Community dimension within the meaning of Article 1 (2) of the Merger Regulation.

V. PROCEDURE

(8) On 18 March 2002 the Commission received a request from the United Kingdom, pursuant to Article 9(2) (a) of the Merger Regulation, to refer the case as regards the market for the supply of cruise holidays in the United Kingdom to the competent United Kingdom competition authorities. The United Kingdom considered that the concentration would threaten to create a dominant position on such a market. As a result of this referral request, the period for a preliminary examination under Article 10 (1) of the Merger Regulation was extended from one month to six weeks.

(9) On 11 April 2002, the Commission decided to initiate proceedings pursuant to Article 6 (1) (c) of the Merger Regulation finding that the operation raises serious doubts as to its compatibility with the common market in at least the national markets for the provision of oceanic cruises in the United Kingdom and Germany.

(10) On 22 May 2002, the Commission communicated a Statement pursuant to Article 18 of the Merger Regulation (hereinafter: Statement of Objections) to Carnival. Carnival submitted on 18 June 2002 a reply to the Statement of Objections (hereinafter: Carnival's Reply). On 3 June 2002 POPC submitted a reply to the Statement of Objections (hereinafter: POPC's Reply). Carnival and POPC both informed the Commission's Hearing Officer prior to 7 June 2002 that they would not exercise their right to a formal oral hearing.

(11) The Advisory Committee discussed the preliminary draft decision on 22 July 2002.

VI. COMPATIBILITY WITH THE COMMON MARKET

A. INTRODUCTION

(12) The operation relates to the provision of oceanic cruises for vacation purposes (8). This section describes the historic development of the industry and some of its fundamental characteristics.

A.1. INDUSTRY BACKGROUND

(13) The present-day cruise industry has its roots in the conversion of transatlantic liners from a means of transport to leisure cruising, starting in the 1960s. The cruise industry grew rapidly during the 1980s, with the larger cruise companies constructing "megaships", and some smaller companies acquiring second-hand ships. Since the mid-1990s some of the large cruise companies have constructed "post-Panamax" ships, of over 100 thousand tonnes, so-called because they are too large to pass through the Panama Canal. In 2001, the cruise industry comprised about 230 ships worldwide, a number expected to grow to around 260 in 2003. Ships are not tied to particular regions or itineraries, may be moved around the globe, and are acquired and chartered on an international basis.

(14) The four largest cruise operators worldwide are Carnival, Royal Caribbean Cruise Line (RCCL), POPC and Star Group (including Norwegian Cruise Line Limited (NCL). In terms of fleet size, Carnival, with 43 ships, is number one, Royal Caribbean, with 23 ships, is number two and the NCL/Asian Star group and POPC each have 18 ships. Carnival currently has 12 ships on order for the end of 2004. POPC currently has five ships on order for the end of 2004. RCCL has five ships scheduled for delivery by the end of 2004; it has the youngest fleet of any cruise line (average age about five years), and the world's largest ships (Voyager, Explorer, and Adventure, each with 3100 berths). NCL/Star has one ship on order.

(15) In 2001, there were approximately 7 million cruise passengers from North America (72 % of world demand), 1,9 million (20 %) from Europe and 800000 (8 %) from Asia. In the United States cruising has achieved a much higher penetration rate than in Europe (United States: 2,2 %; Continental Western Europe: 0,4 %) (9).

(16) In the last few years, the cruise industry has seen much higher growth rates than have other types of vacations. Growth is forecast to continue more rapidly in Europe than in the United States, but is generally expected to be between 8 % and 20 %.

(17) In the United Kingdom a significant part of this growth was accounted for by the entry into cruising of vertically integrated tour operators, beginning in 1994//95 with Airtours, quickly followed by Thomson, and, more recently, First Choice. These companies together carried about one third of all United Kingdom cruise passengers in 2000.

A.2. BACKGROUND TO THE PARTIES

(18) Carnival was founded as a start-up in the United States of America in 1972. Its current brands are CCL, Costa, HAL, Cunard, Seabourn and Windstar, the latter two offering upmarket luxury cruises. CCL offers American-style cruises. Carnival acquired joint control of Costa with Airtours in June 1997, Carnival acquiring full control of Costa in 2000. Costa is mainly active in Europe in Mediterranean markets, where its share (in passenger numbers) is about [50 to 60 %] * in Italy, [35 to 45 %] * in Spain, and [15 to 25 %] * in France. Costa also represents Carnival's main presence in the German market. HAL, which offers American-style cruises, was acquired by Carnival in the late 1980s. CCL, Costa, and HAL have relatively small shares of the United Kingdom market. Carnival's main entry into the United Kingdom market was through its acquisition of Cunard in May 1998.

(19) The other party to the current transaction, POPC, was demerged from the P& O Steam Navigation Company in October 2000. It has a long history of passenger cruising, going back to the 19th century. POPC operates under the P& O Cruises, Princess, Swan Hellenic, and Aida and A'Rosa brands, and is about to launch a new brand in the United Kingdom, "Ocean Village", targeting a younger, more "informal" customer segment and set to start operations in 2003. P& O Cruises, offering British-style cruises, is the leading brand in the United Kingdom market. Princess is mainly active in the North American market. POPC entered the German cruise market through the acquisition of AIDA in 1999 and of Seetours in 2000.

(20) Carnival's bid for POPC is competing with a rival bid by Royal Caribbean. Royal Caribbean's bid was examined by the Bundeskartellamt, who cleared the case on 8 January 2002 and by the OFT (Office of Fair Trading) who recommended referral to the United Kingdom Competition Commission. The United Kingdom Competition Commission's report was published on 19 June 2002, with the conclusion that bringing together RCCL and POPC would not have a significant impact on competition. Both deals are subject to ongoing anti-trust scrutiny in the United States with the FTC (Federal Trade Commission) being the competent authority.

B. THE RELEVANT MARKETS

(21) There are a number of different characteristics in the nature of the product offered by the Parties and other operators in different countries. For clarity of exposition, the evidence that has led the Commission to consider that the relevant geographic markets are national is outlined first. The discussion of the relevant product markets follows.

B.1. THE RELEVANT GEOGRAPHIC MARKETS

(22) Carnival submits different aspects that could characterise the relevant geographic markets for cruises as national or international.

(23) As regards arguments indicating an "international" market, Carnival [...] * mention (s) that ships are not tied to particular regions or itineraries and may be moved easily around the globe, that many oceanic cruises are marketed worldwide and comprise an international mix of customers and languages, that cruise operators seek to encourage an international atmosphere on their cruises, and that the internet makes it possible to "shop around" between different countries for holidays and to purchase holidays, such as cruises on-line. [Carnival] * also refer (s) to the euro making it easier for customers to compare prices and highlight that ships are acquired and chartered from around the world on an international basis.

(24) Concerning aspects pointing towards a national geographic market [Carnival] * refer (s) to divergent national holiday traditions, tastes and preferences (e.g. in regard to timing, or dining and drinking habits), languages, nationally targeted marketing, differing average prices, and substantial variations in the market presence of particular tour and cruise operators.

(25) In decision No IV/M.334 - Costa Crociere/Chargeurs/Accor (10), the Commission concluded that "the markets for cruises and other package holidays are still predominantly of a national character".

(26) The results of the market investigation undertaken by the Commission indicate that the relevant geographic markets for oceanic cruises are national, as a result of the existence of the following factors:

(a) the market presence of cruise operators varies considerably from one Member State to another. For example, in 2000 Carnival had a [50 to 60 %] * passenger market share in Italy and [0 to 10 %] * in the United Kingdom, whereas POPC had less than [0 to 5 %] * in Italy and [15 to 25 %] * in the United Kingdom;

(b) the penetration of the cruise holiday is different from one Member State to another: in 2000, 1,25 % of British citizens cruised, while the rate is around 0,4 to 0,5 % for Germany, Italy and France, and below 0,2 % for Spain;

(c) distribution channels are different in Member States as both specialised cruise agencies and general travel agents operate on a predominantly national level. Only a small amount of sales are made via the internet and of the internet sites selling cruises, many operate on a national level. In the United Kingdom, there is a strong presence of specialised cruise agencies in addition to general travel agencies;

(d) marketing and promotions are usually run at national level, usually by a subsidiary or by an independent general sales agent, whose function it is to take account of national specificities in marketing and promotion. The content and presentation of promotions are adapted in each Member State;

(e) pricing strategies and price levels are different in each Member State. The Commission has found evidence that the same cruises were sold by cruise operators at prices that could vary depending on the country to which they were marketed;

(f) some cruise brands are only or mainly marketed in certain Member States, for example Cunard and P& O Cruises are focused on the British market, while Aida and A'Rosa are focused on the German market.

(27) Most of the reasons set out in recital 26 would apply equally were the product market to include other holidays as well as oceanic cruises. The conclusion that the relevant geographic markets are national would therefore remain the same even were the relevant product market considered to be that for the provision of leisure travel as proposed by Carnival (see recital 29).

B.2. THE RELEVANT PRODUCT MARKETS

(28) The Parties are involved in the provision of oceanic cruises. Oceanic cruises involve a journey of a multi-night sequence on board a ship undertaken for leisure purposes in maritime waters.

B.2.1. The Parties' views on the product market

(29) Carnival considers that the appropriate product market is the market for the provision of leisure travel, which includes a wide variety of alternative vacation options. It argues that a cruise holiday is substitutable for various other types of vacation such as a stay in a holiday club or an all-inclusive hotel resort, a package holiday, a skiing holiday or a tour in an exotic location.

(30) The target company (POPC) takes a similar but not identical view. It sees oceanic cruises as part of a wider market. It does not, however, maintain that oceanic cruises compete with all other types of holiday (11).

(31) The Commission's investigation has pointed to a number of elements that distinguish oceanic cruises from other forms of holidays across Europe (12).

B.2.2. Oceanic cruises and other (land-based) holidays

Characteristics of cruising and consumers' perceptions

(32) A customer of an oceanic cruise (13) buys a product that is distinct from any other holiday product in that it combines: (a) the experience of the sea; (b) the experience of travelling from one place to another in a multi-night sequence; (c) without the necessity of continuous re-packing; (d) with a specific on-board experience of conviviality that is different from land-based group experiences.

(33) This is reflected in the catalogue of reasons given by consumers for going on a cruise, which is unique to cruising and sets it apart from any other type of holiday, including package holiday alternatives. The Commission's market investigation has complemented the information available in industry sources on the reasons why consumers book a cruise (14). Asked for their consumers' reasons for booking a cruise, travel agents mentioned three top reasons which, taken together, are specific to cruises. These are 1. the attraction of the itinerary and the sequence of destinations, 2. the all-inclusiveness of the services offered on board a ship, and 3. the fact that cruising does not require any packing nor unpacking to go from one place to another, thus constituting a very convenient way of holidaymaking. No other form of holiday can combine these three main attractions. A number of other reasons cited were not cruise specific, such as price, value for money and quality/satisfaction.

(34) POPC claims that coach tours and multi-centre land-based holidays have some similarities in characteristics (15). POPC also points to its own consumer research in which the most important reasons for taking a cruise mentioned are price, itinerary, special occasion and time of the year (16). The price argument is very general, as is, in a travel context, the consideration of "time of the year". That cruises provide a year-round service in fact distinguishes them from seasonally more limited other holiday products. "Special occasion" may be a common characteristic for a number of higher-priced holidays but may also be a distinguishing feature of cruises from many other forms of vacation. Finally, the itineraries of cruises may be regarded as entirely unique to cruises as it is hardly possible to replicate the itineraries of cruises, let alone the experience associated with it, with land-based means of transport.

(35) Carnival's position on the potential distinction between cruise holidays and other holiday products in the United Kingdom is based upon two studies, prepared by IRN Research (17) and Mintel (18).

(36) Mintel were commissioned by Carnival to consider the extent to which cruises compete with land-based holidays, as well as whether specialist cruise operators such as Carnival compete with cruises offered by the tour operators (on the latter see recitals 70 to 93 below). Mintel provided data from a variety of sources to show: (a) that cruises are more and more integrated into holiday packages including land-based components; (b) the falling average age of cruisers and the adaptation of cruise ships to these younger customers' needs; (c) the falling cruise prices in real terms; and (d) that cruisers also take other holidays than cruises.

(37) The Commission does not consider the first argument as indicative of the existence of a single market between the cruise components and the land-based components of such holidays. The relationship between the two aspects could be complementary rather than substitutive in which case the same evidence could be interpreted in exactly the opposite way. Nor does the average age of cruise customers, which is falling rather slowly, point to the existence of one market containing both cruises and land-based holidays as the customer profiles of both in regard to age continue to differ considerably (see Table 1). The data provided by Mintel do not show that prices per day are falling. The increase of the total cruise price bracket below GBP 500 as compared to the price bracket between GBP 500 and GBP 1000 can be explained by the increase in short-break cruises that Mintel itself refers to, whereas price categories above show much less change. Table 1 shows that cruise prices still tend to be substantially different from the prices of the other main types of holidays. Finally, Mintel itself acknowledges that holiday habits of cruise customers differ substantially from the average population, with cruisers being much more likely to take more than one holiday a year. Whether or not these other holidays are cruise holidays does not have any importance for considering whether oceanic cruises and other land-based holidays are in the same market, in that the different ways that people spend their leisure time and disposable income are to be seen as complementary. The Commission therefore does not regard this evidence to be persuasive in terms of market definition.

(38) IRN Research concluded that there is considerable interchange between land-based holidays and cruise holidays and both can be considered as part of the same market. IRN reached this conclusion after analysing information on holidaymakers' perceptions and preferences, as well as on more general industry trends. IRN noted, in particular, that 64 % of adults whose last holiday was a cruise considered a land-based holiday (19). Unfortunately, from the results presented it is not possible to establish whether such consideration involved active comparisons of prices, itineraries, etc. which led to the choice of a cruise or whether respondents decided to go on a cruise before getting more information on cruises.

(39) The opinion of Preussag (the owner of TUI, Thomson and Hapag Lloyd) is also notable as it has ample knowledge of the general holiday market and via Thomson cruises and Hapag Lloyd provides cruises in both the United Kingdom and in Germany. Preussag stated that one of the distinguishing features of the cruise sector was its strong emphasis on the group character of the cruise, which made social factors more important. Certain social groups would want to be among their social peers. This points to the importance of the element of conviviality. Also other elements that were mentioned in recital 32 as objective characteristics of the cruise, were referred to by Preussag as playing an important role for the customers' cruise preference as a specific form of holiday (comfortable mobility, no packing/unpacking, roundtrip by sea).

(40) The Commission considers that consumers' distinct preferences for oceanic cruises are reflected in the information that they seek when first contacting travel agents. When first contacting a travel agency, customers can request a variety of information on different types of holidays. For example, they may request both cruise brochures and foreign package holiday brochures and then decide which type of holiday to take. To explore this issue, the Commission asked travel agents to indicate what information those customers who eventually booked a cruise had originally asked for. The result indicates that only a few customers who finally bought a cruise first requested a foreign package holiday brochure. On the other hand, almost all customers who actually bought a cruise first requested a cruise brochure. This indicates that their decision to go on a cruise had been largely taken before visiting the travel agent.

(41) The result referred to in recital 40 has been supported by information provided by RCCL (20). RCCL commissioned a survey to ask people who had been on a cruise in the last five years whether they had considered another type of holiday at the time they were deciding to take their last cruise (21). The survey found that less than one in five people (18 %) who had taken a cruise had considered taking an alternative type of holiday. This proportion was consistent amongst both those taking premium cruises and those choosing cruises provided by the tour operators. Unfortunately, the researchers did not ask whether the respondents would have considered an alternative had relative prices between cruises and alternatives been any different.

(42) These findings are in line with work carried out in the United States, the Nationwide Tracking Survey, which gathers information about people who have taken a cruise in the preceding 12 months. This survey found that 85 % of cruise customers had decided to go on a cruise before contacting a travel agent (22). This indicates that there is only limited competition and substitutability between cruises and other forms of holidays at the point when customers make contact with travel agents.

Customer groups

(43) Social mix is an important defining feature of the cruise itself. The demographics of the customers are therefore important considerations in determining the structure of demand for oceanic cruises. The demographic spread of cruise customers differs to a considerable extent from other customer groups.

(44) Whilst the demographics of the cruise industry are changing, the typical cruise customer is still much more advanced in age and more affluent than customers of land-based foreign package holidays. One report (23) found that "(t)he average all inclusive consumer tends to be between 20 and 34, in the C1 or C2 socio-demographic grouping (24), and in full-time employment. There appears to be a rough 50/50 split between the sexes and also married and unmarried customers". Contrast this with the average age of the British cruise passenger in 2000 which was still 53,9 years, the fact that only 9 % of British cruise passengers were below 35 years whereas 78 % were above 45 years of age (25).

(45) A survey conducted in the United Kingdom on holidays taken away from home in 1999(26) confirms the main demographic distinctions between the holiday takers of cruises only and of all-abroad holidays or inclusive holidays. This information is summarised in Table 1. This clearly shows how the typical cruiser is different compared to the average all abroad holiday taker, to the average customer of foreign inclusive holidays and to the long-haul vacationer in terms of age (older), family (fewer children), social class (higher), employment (retired) and income (higher).

Table 1 - Demographics by holiday type

<emplacement tableau>

Source:

NOP.

(46) POPC has submitted a study prepared for them by [the POPC-commissioned study] * examining actual and potential demand for cruises in the United Kingdom in 2001. The [POPC-commissioned] * study firstly analyses the broad demographics of cruisers, those who would consider cruising and those who would never consider cruising. These findings support the characteristics that have been described in recitals 43 to 45 above, in particular the finding that cruisers travel less with their children (27).

Prices

(47) The Parties have argued that prices of cruise holidays are similar to the prices of other land-based holiday products.

(48) Table 2 compares average per diem spending of United Kingdom customers on inclusive holidays, on independent holidays and on long-haul holidays with oceanic cruises. The data reveal that the average per diem expenditure for cruises is more than twice as high as average per diem expenditure of inclusive holidays abroad (i.e. foreign package holidays), long-haul holidays and independent holidays abroad.

Table 2 - Average per diem prices: cruises and other holidays, 1999

<emplacement tableau>

Source:

NOP Consumer 1999.

(49) Table 2 also shows that in terms of total spending only long-haul holidays are close to oceanic cruises. This is because long-haul holidays have considerably longer average duration.

(50) Carnival has presented evidence seeking to prove the equivalence of land-based holiday prices and cruise prices. Much of this information presents isolated examples of cruise and non-cruise holidays, which Carnival considers substitutable and which show similar prices.

(51) Both Carnival and POPC have presented a report "Turning the Tide"(28) by the Cruise Information Service, on behalf of the Passenger Shipping Association, the United Kingdom trade organisation of cruise companies. This report seeks to show pricing equivalence between cruises and other holidays. Only one of the five cruises presented is a premium cruise, the other cruises are a river cruise, two economy cruises (one by Thomson, one by Royal Olympic) and one luxury cruise (see recitals 100 and 102 below). It is worth noting that the premium cruise presented is untypical insofar as it presents a cruise that is only half the duration of this ship's average cruise duration.

(52) Carnival has highlighted a number of premium holidays that are potentially close substitutes for oceanic cruises. Evidence provided by operators of land-based all-inclusive holidays does not support an argument that these holidays belong in the same relevant market as oceanic cruises. For example, Carnival mentioned one all-inclusive operator as providing holidays that are close substitutes for oceanic cruises. That operator sees himself explicitly as a luxury all-inclusive resort operator and considered that oceanic cruises would only "to a small extent" be considered by his customers as a substitute. That same opinion was given by a company that provides more down-market all-inclusive resort holidays. Furthermore, the latter operator considered that customers would "not, or to a negligible extent" switch to oceanic cruises for price reasons, indicating that cross price elasticity between land-based all-inclusive resorts and oceanic cruises is low.

Branding, marketing and distribution

(53) Whilst several land-based tour operators have started cruise operations in the United Kingdom economy segment (see recitals 70 to 93 below), the very large majority of cruise companies which provide cruises in the United Kingdom are specialised and do not focus on the provision of other forms of (land-based) holidays. For the most part, therefore, the brands under which cruises are sold in the United Kingdom are cruise-specific brands, and extension of other brands into cruising is not the predominant feature of the industry. This also applies in Germany where the brands under which cruises are sold are cruise specific. Good examples of cruise-specific brands in Germany are POPC's Aida and A'Rosa brands.

(54) Nevertheless, in the United Kingdom there is some brand extension for a few upmarket brands such as "Disney", "Club Med" and "Radisson" (Radisson Seven Seas Cruises) and for brands like Saga and Page & Moy. Taken together these currently account for less than 10 % of cruises in the United Kingdom.

(55) There are some commonalities between the marketing of oceanic cruises and other holiday products. For example both are advertised in brochures and sold through travel agents. However, most oceanic cruises are advertised in specialised brochures. In the United Kingdom there are also a large number of independent travel agencies through which a substantial proportion of oceanic cruises is sold.

Customers' booking patterns

(56) There are indications that booking patterns, contrary to the claims of the Parties, are not entirely comparable. Tables 3a and 3b present data from the NOP 1999 survey summarising information about booking patterns for different groups of holidaymakers. They indicate important differences both in respect of mode/channel of booking (in person, telephoning) and by booking agent (travel agent, tour operator, cruise company).

Table 3a - Booking mode

<emplacement tableau>

Source:

NOP Consumer, 1999; pp. 112 to 117 and 130 to 132.

Table 3b - Booking channel

<emplacement tableau>

Source:

NOP Consumer, 1999; pp. 112 to 117 and 130 to 132.

Views of Travel agents and competitors

(57) In the Commission's market investigation, a large majority of travel agents regarded oceanic cruises as belonging to a product market that is distinct from other forms of vacations. Travel agents also denied the possibility that only specific categories of oceanic cruises, for example, those in the premium segment, form a common product market with other types of vacations.

(58) The Parties' competitors have expressed a range of views. Some companies, among them in particular vertically integrated United Kingdom tour operators offering cruises, see oceanic cruises as part of a wider market. Most of the Parties' closest cruise competitors both worldwide and in Europe, however, view oceanic cruises as constituting at least one separate product market.

(59) Competitors have indicated to the Commission that their pricing, both strategic and, even more so, tactical, takes account of the pricing of other cruise companies only and not of other holiday options available to the consumer.

Empirical evidence on the impact of new capacity on yields

(60) Two consultants' reports have been submitted to the Commission exploring the relationship between yields from cruising, capacity additions and the Parties' resulting conclusions on product market definition.

(61) One was prepared by NERA Economic Consulting on behalf of POPC (29) and shows POPC's yields (30) for cruises, analysing these in the context of the overall capacity offered in the United Kingdom by POPC and other cruise operators, in particular the significant new economy cruising capacity offered by the tour operator companies since 1995. The main conclusion put forward by NERA is that POPC's yields have remained broadly constant over time whilst capacity has increased significantly. This finding is more or less the same for cruises to the Caribbean, the Atlantic Islands and the Mediterranean.

(62) NERA drew the corollary that the large increase in cruise holidays was achieved largely at the expense of other types of holiday and that the absence of "large yield reductions in the face of large capacity increases is hard to reconcile with the proposition that cruise holidays comprise a separate product market". NERA therefore concluded that because yields were unaffected by new capacity coming on stream, cruises can be considered as part of a wider holiday market.

(63) However, this conclusion is not the only possible explanation that can be inferred. In particular, while the finding that POPC's yields were unaffected by the significant new cruising capacity would be consistent with the relevant market being that for all holidays (and therefore the impact of the capacity additions being relatively very small), it would also be consistent with the view that premium cruises in the second half of the 1990s belonged in a separate market from economy cruises (explaining why the capacity additions in the economy market did not affect yields in the premium market). Other explanations are possible and attention should be drawn to NERA qualifying these conclusions with the recognition that the study contains no information on demand shifts. NERA have concluded that "we do not have sufficient data, for example on demand shifts, to permit definitive conclusions on market definition"(31). This therefore limits the suitability of these findings for drawing definitive conclusions on market definition.

(64) Carnival have presented a paper prepared by Margaret E. Guerin-Calvert and Janusz Ordover (32) (the Ordover paper) which draws similar conclusions to those drawn by NERA, albeit for information which relates to the United States market. The Ordover paper has analysed data for five ships operating in the Caribbean, looking at the hypothesis that if cruising were a separate relevant market, then substantial increases in industry capacity could only be absorbed at the expense of significantly depressed profits (33). The Ordover paper examined profits and capacity during 1995 to 2001 for five Carnival ships that sailed in the Caribbean during most or all of that period.

(65) The conclusion reached in the Ordover paper is that significant expansion of capacity in the Caribbean had only a modest effect on profits and prices. The authors concluded that "it would be difficult to believe that such volume of capacity could be absorbed in a Caribbean cruising-only market without depressing profits substantially" and that the results were "consistent with a broader relevant market"(34).

(66) As with the NERA study, the Ordover paper can be criticised because it has only examined changes on the supply-side. In particular the conclusions reached are undermined by the fact that introduction of new capacity is accompanied by actions by the cruise operators to try to shift demand for cruises. For example, information provided by Carnival giving its marketing expenditure in the United States shows that Carnival increased its marketing expenditure considerably over the period of analysis. Carnival's marketing expenditure in 2001 was [...] * higher than in 1995. Between 1998 and 1999 alone, Carnival increased its marketing expenditure by [...] *. These marketing costs have not been taken into account in the analysis of yields, therefore the finding that yields were unchanged may not be robust. Secondly, it is clear that Carnival was investing a considerable amount of money to shift demand. Without further information on these shifts in demand, it is not possible to use information on the effects of changes in supply upon yields to draw conclusions about the scope of the relevant product market.

(67) The Commission therefore does not consider that the conclusions reached by either NERA or by Guerin-Calvert/Ordover provide reliable evidence that cruising forms part of a wider relevant product market. The NERA study is also relevant in terms of the premium/economy cruise distinction (see recitals 69 to 98 below).

Conclusion

(68) In the light of the above, the Commission concludes that the relevant national product markets are the markets for the supply of oceanic cruises. This conclusion applies regardless of whether premium oceanic cruises are considered in relation to the competitive impact exerted by premium land-based holidays or whether all oceanic cruises are considered in relation to the competitive impact exerted by a wider group of land-based holidays.

B.2.3. Premium cruises and tour operator (economy) cruises in the United Kingdom

(69) There are a number of specific features of the cruising market in the United Kingdom, in particular the emergence between 1995 and 1999 of a special type of cruise offered by United Kingdom tour operators, which have made it necessary to examine whether premium cruises could be considered to belong to a separate relevant market from economy cruises. Indeed, when the Commission issued the Statement of Objections, it was of the preliminary view that there was a separate relevant market for premium cruises in the United Kingdom. The following analysis presents the results of the Commission's considerations that have taken place in the light of the responses received to the Statement of Objections.

Product characteristics and consumers' perceptions

(70) It is possible to distinguish three segments within an overall oceanic cruise market in the United Kingdom: luxury, premium and economy. As regards luxury cruises see recitals 101 and 102 below.

(71) Premium and economy cruises share many of the same product characteristics. Nevertheless, there are a number of differences in the characteristics that typify the product offering in these market segments. For example, there are some differences in the hardware that is used (i.e. the ships, premium cruise ships tend to be newer with better facilities) and/or in the "software", which includes the quality of the service, for example, premium cruises tend to offer higher staff to customer ratio(35) and would tend to provide a more formal or elegant environment. There are also differences in the manner in which economy offerings are branded, marketed and distributed, all of which are reflected in the price differences that can be observed between premium and economy cruises.

(72) It is, however, also clear that for each of the factors referred to in recital 71, there is a range of quality within both the premium and the economy segments. For example, within the premium segment there are some cruise brands such as Cunard which offer a very formal product, while others such as RCI provide and market themselves on the basis of a much more relaxed service, with much less emphasis on the formal rituals traditionally associated with premium quality cruises such as the "captain's dinner". Similarly, in terms of the hardware used, there are some premium cruise operators who seek to use only very modern ships, while others, such as Saga, use ships that are older than those used by some of the economy operators.

(73) These factors are reflected in some, but not all, of the evidence presented to the Commission showing the perceptions that consumers have of the different cruise lines. POPC has submitted the results of a survey on the image of cruise lines to the European Commission (36) that asked consumers for their opinions on a set of 19 questions on their views about cruise operators. The Commission's analysis of that information shows that consumers' perceptions of the different premium operators (P& O Cruises, Princess, Cunard, RCI) are similar to each other, likewise their perceptions of Thomson and Airtours are similar to each other, but that there are differences in the perceptions of the premium operators when compared with the economy operators.

(74) Consultancy work carried out by IRN Research on behalf of Carnival does not support this distinction. IRN highlighted that there is a very close relationship between the reasons why people booked a cruise with a cruise line and the reasons why people booked a cruise with a tour operator (37). It should be noted that many of the results are based on samples of less than 100 which limits their statistical reliability.

(75) On a slightly different angle, POPC has submitted a survey it carried out in August 2001 (customer survey) with which it seeks to demonstrate that their own passengers also consider the cruises of a great variety of alternative cruise operators, including the tour operators. Referring to the customer survey, POPC observes [reference to POPC's submission] * that [15 to 25 %] * of the survey respondents who looked at a P& O Cruises brochure or their website also looked at an Airtours brochure or website. In addition, POPC states its belief that when Thomson started cruise operations in 1996, [15 to 35 %] * of its passengers had previously taken a POPC cruise. In POPC's view, this finding indicates that United Kingdom tour operator cruises (TOCs) compete directly with those of POPC (38).

(76) POPC has also provided evidence to show that, apart from routinely monitoring all cruise operators' brochures, including those of Airtours and Thomson, it also surveys the quality of these competitors' operations by sending some of its own employees on holidays with all the other operators, including the United Kingdom tour operators, to conduct benchmarking studies. POPC claims that this is proof it considers that actions by those other operators constrain its own competitive behaviour and that there is no basis for defining a relevant product market for certain cruise operators only (39).

(77) The evidence on product characteristics and consumers' perceptions is therefore mixed. Some evidence highlights differences between premium and economy products, while other evidence suggests that the boundaries between these segments are not clear cut.

Prices of premium cruises and economy cruises

(78) If there were a clear distinction between premium and economy cruises, then this would be likely to be reflected in differences in pricing (40). However, if there was a continuum of prices, then this would be an indication that there is not a clear quality/price distinction between premium and economy cruises and therefore that they belong in the same relevant product market. The Commission has therefore sought to establish whether there is such a clear difference, not just in terms of average pricing, but also whether there is a continuum of prices between the top end of the economy offerings and the lower end of the premium offerings.

(79) POPC has submitted a study by RBB Economics, which argues for the existence of a price continuum between TOCs and other cruises offered to United Kingdom customers. RBB Economics' methodology is based upon comparing average prices for given cabin types (suite/premier, inside and outside) weighted by the number of cabins at each price point, for a number of TOCs and premium cruises offered on the United Kingdom market. It considered a range of cruises from high-end premium to lower end economy cruises. The study shows tables and charts of cruise prices in which the prices of tour operator cruises and premium cruises overlap. Closer analysis of the study reveals that it compares prices of cruises of: (i) different duration (41); (ii) substantially different itineraries (42); (iii) different distance to ports of departure (fly cruises, ex-United Kingdom cruises); (iv) different all-inclusiveness concepts (covering tips and/or drinks or not); and (v) different brochure editions (making it unlikely that a first edition on-brochure discount of one product is available to the consumer at the moment he examines the on-brochure discounted price of the third edition of another product's brochure).

(80) For evidence of a continuum to be meaningful, it must reflect choices that consumers are able to make. The fact that seven- and 14-day cruises for example may have smaller per diem prices does not mean that they are in any way close alternatives for consumers or that an increase in price in a given seven-day cruise will result in a consumer switching to a 14-day cruise costing considerably more. Likewise mixing different brochure editions in the price observations that are made is inappropriate, as it cannot reflect choices that consumers can make at a point in time. For this reason, the Commission has looked only at "same edition" prices. It has found that cruise lengths were not easily adjustable to the 14-day format as actual cruise lengths differed between 10 and 16 days. POPC has provided per diem prices on the basis of a weighted basis both for brochure prices and for on-brochure discount prices. The Commission has therefore segmented the sample given by POPC by: (i) departure date (April, June, August); (ii) cruise duration (seven days, 14 days); (iii) brochure edition (first, second); and (iv) on-brochure discount price/undiscounted brochure price.

(81) The samples included between three and nine individual cruises by both United Kingdom tour operators and premium brands. The Commission then measured the price difference between the most expensive tour operator cruise and the cheapest premium cruise (as a percentage difference to the tour operator cruise's price.

(82) The results of this comparison are as follows: of the 48 resulting price observations, the highest tour operator cruise price was cheaper than the cheapest premium price in 41 cases, in six cases it was not (one was same price). In 15 cases, the difference was less than ± 10 %. Table 4 gives the percentage differences for the first and second brochure editions.

Table 4 - Percentage difference most expensive United Kingdom tour operator and cheapest United Kingdom premium

<emplacement tableau>

Source:

Commission on the basis of data provided by RBB.

(83) Table 4 shows that there are some significant price differences between tour operator cruises and premium cruises if the same cabin types, cruise duration and approximate departure at the same brochure edition are compared, in particular when comparing the first editions. However, it is also clear from this table that not all of the prices of the cheapest premium cruises are significantly greater than the more expensive economy offerings. Indeed, in particular when comparing the discounted prices, there are several examples where the economy cruise is more expensive than the premium cruise. On the basis of this evidence, there does not appear to be a clear break between the prices of economy and of premium cruises.

(84) Furthermore, consumers' choices may not be limited to a choice between the same type of cabin of different ships. In this context, the Parties have claimed that cheaper inside cabins of a premium cruise might be equivalent to outside cabins of economy cruises and that therefore there was an overlap in pricing even if average prices for similar cabin categories differed. If one accepts that the different cabin types are substitutes, the number of overlaps increases. Indeed of the 16 achievable comparisons (four departure dates/cruise durations, multiplied by two prices, discounted/non-discounted, multiplied by two brochure editions) 10 show an overlap between premium inside and economy outside cabins. This overlap would not disappear even if some adjustment were made for the fact that economy cruises contain tipping whereas premium cruises usually do not. In about one third of the comparable cruises referred to in Table 4 there would be an overlap between inside premium and outside economy cabins even if GBP 10 per day were added to premium cruises to adjust for this different tipping policy.

(85) In a telephone survey carried out on behalf of Royal Caribbean on 13 March 2002 substantial price differences between all premium products and all tour operators' economy products were found(43). The survey asked for the prices for a standard outside cabin on seven-day Mediterranean cruises (including flight, port taxes, security charge) for three departure dates in June, July and August 2002 and, with some adjustment for the inclusion or not of tips and drinks, and arrived at price differences between 20 % and 45 % between the United Kingdom tour operators' cruises and the premium cruises (44). An even higher difference was found comparing the prices of one TOC cruise and two premium cruises for 14-day cruises in the Caribbean in January 2003 (45). As the Commission noted in the Statement of Objections, the relevance of this information to the assessment of this operation is limited by the fact that neither P& O Cruises nor Fred Olsen were included in the sample, apparently because they did not offer comparable itineraries.

(86) RCCL has prepared a further survey updating and extending the price evidence it presented before. This survey was carried out on 21 May, 2002 and inquired about the prices of Caribbean cruises of 14 or 15 days duration. Prices were obtained from the cruise lines' reservation systems. The survey included five premium and two economy cruises departing in January/February 2003 and covered the standard outside cabin, the standard inside cabin and the balcony or equivalent deluxe cabin. RCCL adjusted the data to take into account whether or not tipping was included in the price. The results showed that the prices charged by the premium operators were on average 47 % higher for the same duration.(46) A comparison between the cheaper inside cabins of premium ships against the outside economy cabins showed that the premium inside cabins were on average 34 % more expensive than the economy outside cabins. It must, however be noted that price differences between tour operators and premium brands tend to be larger in the Caribbean. Also, the consideration of average values of the two groups of ships may conceal overlaps between economy and premium prices. Looking at prices on a ship by ship basis, it is of note that the closest two offers in the sample (the cheapest premium and the most expensive economy) were within 2 % of each other. Therefore, even on the basis of data provided by RCCL, a continuum of prices cannot be excluded.

(87) The Commission also examined an important United Kingdom high-street travel agent's sales figures which showed a differentiation between TOCs and premium cruises in the United Kingdom. Putting the percentage of a certain brand's sales volume (measured in passengers) in relation with this brand's sales value (i.e. the total price that the customer paid for cruise and flight) for this travel agent there is an important distinction between premium cruises and TOCs. Each of the TOC's value/volume ratio was less than half the value of the premium cruise brands' average ratio in three consecutive years (1999 to 2001). This indicates that the prices of TOCs are significantly cheaper than the average prices of premium cruises. This evidence does not permit any conclusions to be drawn on the existence of a continuum between TOC and premium cruises.

(88) The Commission has verified this result with data from two other important high street retailers for individual years and has found that each individual TOC's value was between 43 % and 73 % of the premium cruise brands' average.

(89) However, in each of these cases there were also premium cruises such as Costa, Festival, RCI and CCL, with altogether not insignificant sales, that "filled the gap" by showing values between the TOCs and the premium average. Considering further that values for each brand are of course average values for all cabin types and all times of the year the existence of some sort of price continuum between TOCs and premium cruises also cannot be excluded on the basis of this data.

Conclusion on premium versus economy segmentation

(90) There are a number of other factors which could be considered in relation to the premium and economy segments, for example, the hardware (see recital 71), branding (see recital 53) and distribution (e.g. that cruise specialists tend to sell mainly premium and luxury cruises). This shows that there are a number of dimensions upon which cruise operators can seek to distinguish their products from their competitors' and on which there exists a variety of offerings available for customers which may be grouped into distinct segments of the oceanic cruise market.

(91) There is not a clear dividing line between premium and economy offerings for any of the factors mentioned in recital 90. The Commission considers that evidence points to a single market for oceanic cruises. If there were a clear enough distinction between the quality of economy and of premium cruises such that they did not exert any competitive impact on each other, then this would be reflected in an obvious discontinuity of prices. Such a discontinuity could not be established with sufficient certainty.

(92) It is also of note that two of the premium cruise operators, RCCL (jointly with First Choice) and POPC either have recently or will shortly launch new brands (Island and Ocean Village respectively) which may further blur the boundaries between the premium and the economy segments. This is also borne out by the fact that economy operators are introducing newer, more modern ships (see recitals 157 to 159 below).

(93) In the light of the evidence presented in recitals 32 to 92 above, the Commission considers that the relevant product market for the purposes of this decision is the supply of oceanic cruises in the United Kingdom.

B.2.4. Premium and Economy cruises - Germany, Spain, Italy and France

(94) There is no clear distinction in Germany between price/quality segments such as premium or economy but a broad spectrum. Generally, the boundaries between German-speaking economy products and premium products are less accentuated than they are in the United Kingdom.

(95) Whilst Mediterranean companies active in Germany at the economy end see differences between their economy products and their premium offering, the fact that both are offered by the same companies under as yet not, or, in Festival's case, not fully, distinguished brands also militates against a separation of premium and economy markets in Germany.

(96) In Spain, Pullmantur with an own cruise product, and Iberojet, in a joint venture with Festival and Trasmediterránea called Spanish Cruise Line, have entered the market and offer low-price cruise products which, however, apart from this economy approach, have little in common with British TOCs. Most notably, what these Spanish products typically offer is not a holiday package. Instead, they provide round trip itineraries from Spanish ports in the Western Mediterranean which, for their Spanish customers, do not need to be combined with charter flights nor "stay and cruise" packages.

(97) In Italy, there was one product, offered by the integrated tour operator Valtur who offered Caribbean "club" cruises on one older ship that may have constituted a parallel with United Kingdom TOCs. However, Valtur's market share was small (4.8 % in 2000), and it has stopped cruise operations. As its inclusion or not in the same market as other cruises does not influence the assessment materially, it is not necessary to investigate this any further. Its pending market exit can be seen as further proof that United Kingdom TOCs have no equivalent in Continental Europe. For economy products offered by Mediterranean Shipping Cruises (MSC) and Festival the same applies as stated for Germany in recital 95.

(98) In France, tour operators like Nouvelles Frontières or Marsans-Transtours have entered the market with full ship, often part-of-the-year charters. Upmarket sail cruises are offered by Club Med and Star Clippers. At the luxury end Compagnie des Iles du Ponant is providing French-speaking cruises on small ships. Apart from this comparatively small French offering the largest part of the market is characterised by cruises supplied by Mediterranean companies and brands like Festival, Carnival's brand Costa, MSC and Royal Olympic. For the same reason as stated in recital 95 there is no distinctness of offer in terms of product market definition between premium and economy products in France.

Conclusion

(99) In the light of the above, the Commission concludes that the relevant product market in Germany, Spain, France and Italy is that for the supply of oceanic cruises.

B.2.5. Other issues relating to product market definition

Whether luxury cruises constitute separate relevant product markets can be left open

(100) For the purposes of this decision it can be left open whether luxury cruises should be considered as belonging to a separate relevant market, as they represent only a small market volume and their inclusion or exclusion would not affect the conclusions reached. On the assumption of a separate luxury market, the operation would not create any competition concerns, as there is no overlap between the Parties' activities. The luxury segment or market accounts for no more than 5 % of cruise capacity (47) in the United Kingdom, Spain and Italy, and 11 % in Germany (48).

(101) Only Carnival is active in the luxury segment or market through its Seabourn and Windstar brands and arguably through the small number of luxury cabins on Cunard's Queen Elizabeth II. POPC is not active in providing luxury cruises.

(102) Due to the relatively small size of the luxury segment, whether or not it is included in a wider oceanic cruises market does not substantially affect the assessment in this decision. For the purposes of this decision, data on luxury cruises has been included in the data for all oceanic cruises.

River cruises and coastal ferry cruises do not belong to the same product market as oceanic cruises

(103) Although POPC, in Germany, also provides river cruises there is no overlap between the Parties in this area and, for the reasons outlined in recital 104 they will be excluded from the following considerations.

(104) The Commission's market investigation confirmed that river cruises and oceanic cruises are in separate markets. This view was expressed almost unanimously by competitors and travel agents/tour operators. These types of cruises not only differ in terms of type of voyage, itinerary and destination, but also in terms of duration and price, an average river cruise being shorter and costing significantly less than an average oceanic cruise (49). Due to these differences, the consumer does therefore not consider river cruises and oceanic cruises substitutable.

(105) For the purpose of this decision river cruises could therefore be regarded as belonging to a different product market from oceanic cruises. As there is no overlap in the Parties' activities in river cruises (only POPC is active in the provision of river cruises through its German subsidiary Seetours/A'Rosa) this market does not need to be considered any further.

(106) A further distinction can be made between coastal cruises and oceanic cruises. The evident distinction that main itineraries of coastal cruises are set in coastal waters has an important effect on the nature of the product. With coastal cruises the scenery of the shoreline provides one if not the main attraction. These cruises therefore tend to move during the daytime and sometimes anchor at night. The reverse is true for the typical oceanic cruise. These cruises also tend to be different in terms of entertainment programme offered. Typical destinations for this type of cruise are the British Isles, the Galapagos Islands and the Norwegian coastline.

(107) A special kind of coastal cruise are coastal ferry cruises such as provided by Norwegian Coastal Voyages/Hurtigruten. These products have as a further distinguishing factor their usually prevailing (but at peak times receding) transport function for the local population accentuating these products' "hop-on/hop-off" character and further changing the on-board experience provided by these cruises. Although the newest set of ships for Norwegian Coastal Voyages take account of an increased tourist importance of their product, the main distinguishing features of coastal ferry cruising remain valid also for these newer products. Coastal ferry cruise ships do not offer entertainment and facilities that are comparable to the ones offered by oceanic cruises and port stops are a function of utility rather than of shore excursions (which may, however, also be provided). The product offered by these ships is therefore substantially different from an oceanic cruise, which may, nevertheless, equally cruise the Norwegian coastline.

(108) For the purpose of this decision coastal ferry cruises are therefore regarded as belonging to a different product market from oceanic cruises. This approach has been confirmed by the views expressed by a very large majority of respondents to the Commission's investigation and conforms to the way Norwegian Coastal Voyages/Hurtigruten see their product (50).

(109) Due to the very limited impact of other coastal cruises (the number of vessels dedicated to these cruises are small), it is not necessary to decide whether coastal cruises in general belong to a different product market.

It is not necessary to decide on oceanic niche cruises

(110) The Commission also invited market participants to express their views on other niche cruises, in particular sail cruises and explorer cruises. Although a majority of respondents saw them as belonging to separate markets the views expressed were not as clear cut as the one expressed for river cruises and coastal ferry cruises. It is not necessary to decide whether these cruises belong to different markets or not.

(111) For the purpose of this decision, the oceanic niche cruises referred to in recital 110 are therefore regarded as part of a wider oceanic cruises market.

B.2.6. Conclusion on product market definition

(112) The Commission considers that there exist separate relevant national markets for the supply of oceanic cruises which are distinct from markets for the supply of coastal ferry or riverine cruises, and also separate from wider markets for the supply of other (land-based) holidays. Furthermore, it considers that it is not possible to distinguish a separate relevant market for the supply of premium oceanic cruises in the United Kingdom.

C. COMPETITIVE ASSESSMENT

C.1. EEA PASSENGER MARKET SHARES FOR 2000 AS ESTIMATED BY CARNIVAL

(113) In the EEA as a whole, the combined oceanic cruises market share of the Parties (measured by the number of passengers) is estimated to have amounted to 32 % in 2000 (Carnival 19 % and POPC 13 %). Table 5 gives the market share to which the operation would give rise (only affected markets listed).

Table 5 - EEA cruise passenger market shares 2000

<emplacement tableau>

(Source:

Carnival estimate).

(114) This table shows that the main overlaps with combined market shares in 2000 of between [25 to 35 %] * and [20 to 30 %] * arise in the United Kingdom and Germany. Beyond that, the Parties are strong in Spain and Italy ([35 to 45 %] * and [50 to 60 %] * respectively), but with very limited overlaps of 2 % or less (51).

(115) No concerns about the merger were raised during the Commission's investigation in relation to the markets for oceanic cruises in Austria and Ireland. Some concerns about the merger were raised in relation to the oceanic cruise market in Belgium and the Netherlands. These were not confirmed in the Commission's investigation. According to Carnival, the Parties' combined market share in each of Austria, Ireland, Belgium and the Netherlands is below 30 %, and the increment from the merger 5 % or less. In those countries, the cruise markets are relatively immature. For example, cruises have achieved a much lower rate of penetration than in the United Kingdom(52). The number of cruises sold in those countries has grown significantly in recent years and, as in Germany and the United Kingdom, this is forecast by industry observers to continue (53). Furthermore, the cruises sold in these countries take a mix of people from different countries. Companies selling United States-style cruises, pan-European-style cruises and European national-style cruises of neighbouring countries are readily able to adjust the nationality mix on-board their ships to increase the number of cruises sold in Austria, Belgium, Ireland and the Netherlands (see also recitals 189 to 195). The Commission therefore does not consider that the operation will have any significant negative impact on competition in Austria, Belgium, Ireland or the Netherlands.

C.2. UNITED KINGDOM

C.2.1. Aspects of product segmentation relevant for brand positioning within the mainstream oceanic cruises market

C.2.1.1. The importance of brands for luxury and premium cruises

(116) There are a large number of criteria defining a particular cruise product. These individual parameters include size of ship, age of ship, size and equipment of cabins, size and decor of public spaces both indoors and outdoors, facilities offered, quality and style of food, quality and style of on-board entertainment, quality and style of cuisine, quality and style of service, on-board language or languages, (distance to) port of departure, length of cruise, itinerary. The relevant category synthesising all these individual criteria and making products distinguishable for the consumer is the brand. This brand can consist of one single ship or of up to around 10 to 15 different ships which are bound together by some common brand image expressing a certain common product standard of service and entertainment.

(117) In terms of the ship's hardware most multi-ship operators will try to achieve a reasonable balance between variety and uniformity within their individual brands and will consider whether a certain (e.g. older) ship still fits with a particular brand image. Multi-ship premium or luxury brands typically also have a small number of largely identical "sister ships" in their fleet limiting the brand's diversity (and, if ordered as new-builds, are cheaper to build). Cruise operators decide on the positioning of their brands via brand management and try to increase the level of brand-recognition with the relevant customer groups.

(118) A number of relevant dimensions for brand positioning are outlined in recitals 119 to 128 below. These criteria are relevant for determining a brand's positioning and the positioning of the brands that are in close competitive relationship with it.

C.2.1.2. Language and other cultural codes of reference: national, pan-European and United States-style cruises

(119) On-board language and other "cultural" criteria, such as food, daytime and evening entertainment play an important role for a certain cruise product's perception by the customer. In all major European cruising countries (United Kingdom, Germany, France, Italy, Spain), there is a distinct perception of what is felt to be a "national-style" cruise. This national-style cruise is determined first and foremost by the dominant on-board language but also includes other distinguishing factors such as preferred cuisine, eating times, drinks, daytime entertainment, evening entertainment, decor of the ship. All of these major European cruising countries have products on offer that are exclusively, or very predominantly, geared towards one specific national taste, in each of these countries there is a significant percentage of customers who prefer going on such a national-style cruise.

(120) Whilst a certain number of customers in all countries investigated have no specific preference, there is also a certain minority percentage of customers in each country who prefer to go on an "international-style" cruise. Basically there are two quite distinct types of "international" cruises on the market: (a) United States-style cruises; and (b) Mediterranean/pan-European-style cruises.

(121) Like British cruises, United States-style cruises are characterised by English as the only on-board language. What unmistakably distinguishes them from British-style cruises and all other cruises styles is that entertainment and decor (Las Vegas style) are geared towards American tastes which makes United States-style products quite distinct. Commonality of language lowers the cultural threshold for United Kingdom passengers. Passengers of other nationalities can expect to find the day's programme and sometimes menus in their national language. Depending on their numbers there may also be one member of the reception crew speaking the European passenger's language who is assigned as a contact person.

(122) Mediterranean pan-European-style cruises try to allow for the tastes of several nationalities. They typically have on-board announcements in several languages, shore excursions split by language groups, varied European cuisine, two sittings for dining to accommodate for different national habits and non-language based entertainment that is acceptable to all nationalities on board.

(123) To date, Mediterranean/pan-European cruises (54) have been quite successful in penetrating a number of European markets, while United States-style products have had more penetration in the United Kingdom and attempts to more strongly penetrate other countries' markets have also been made.

C.2.1.3. Lifestyle/quality criteria and star ratings

(124) A large majority of travel agents and tour operators responding to the Commission's questionnaire confirmed that they were operating with quality or lifestyle criteria like "luxury", "premium" or "economy" in advising their customers.

(125) The Commission's market investigation also revealed a high acceptance of the system of star rating as applied in the annual "Berlitz Complete Guide to Cruising and Cruise Ships"(55).

(126) As in previous decisions, both in the hospitality sector and in consumer goods sectors (56), the Commission regards the above distinctions as relevant for the assessment of brand positioning and the closeness of competition between different products.

C.2.1.4. Classical cruises v contemporary/fun cruises; targeted life-stage and income groups

(127) A distinction that is relevant particularly in premium segment (s) of cruises is the one between contemporary or "fun" cruises and "classical" cruises. Fun cruises stress on-board activities (such as on-board ice-skating and rock-climbing in the Caribbean) and informality. Classical cruises stress a quieter, more relaxed but also more formally stylish (Captain's dinner) approach to cruises. Both European national-style and United States-style products can follow one or the other model although United States-style products are avant-garde in terms of fun-cruises. Whilst classical cruises are limited to the upper premium segment in the United States, classical European cruises can also be found in the lower premium segment, particularly in the United Kingdom and Germany. Pan-European cruises in the premium segment tend to steer a middle course and often combine family-friendliness with a moderately restrained approach towards formality (most dinners are informal but there are one or two formal dinners). Economy segment cruises tend to be limited in respect to the necessary amenities for fun cruises. They may nevertheless have some attributes of traditional cruises such as one or two formal evenings and a captain's reception.

(128) The average age of cruise customers varies from country to country and from brand to brand. Specific age groups or life-stage groups may belong to a particular brand's target group or not. The average age of its adult customers and the percentage of children and teenagers on board are an indication of brand positioning. However, brands usually attempt to appeal to a range of customers. This range typically overlaps with the ranges of customers to which other brands may appeal. The same applies to income groups.

C.2.2. Market structure

(129) The United Kingdom is by far the country with the largest number of cruise customers in Europe. Next to the United States customers, United Kingdom customers constitute the second largest nationality group of cruise passengers worldwide. In Europe, cruising is showing fast growth rates but cruise customer penetration rates are far below those attained in the United States. In the United Kingdom as in other countries, the cruise industry has experienced much higher growth rates than have other vacations over the same period of time. From 1994 to 1998, the number of United Kingdom customers booking package holidays grew by 15 % and of those who spent overseas holidays by 18 %, whereas the growth in United Kingdom cruise passengers over the same period was 135 %(57).

C.2.2.1. Preliminary remark on measuring market share

(130) Carnival has claimed that market shares should be measured by passenger numbers. The Commission regards passenger figures as a convenient but not as the most appropriate means of measuring market share as they do not take into account the varying lengths of cruises. Market shares measured in terms of passenger numbers would overstate the market share of companies with shorter cruises. Market shares in this decision will therefore be measured in passenger cruise days (PCD) rather than in passengers. Value market shares can also be used as an indicator of market share. However, the Commission has not obtained the necessary information from all market participants which would enable reliable calculations of market share by value.

C.2.2.2. United Kingdom Market shares

Table 6 - United Kingdom oceanic cruise market shares 2001, by PCD

<emplacement tableau>

Source:

Commission, market investigation.

(131) The proposed operation would combine the largest player with one of its premium competitors, Carnival (the other premium competitor being RCCL, which is of similar size to Carnival in the United Kingdom). Together the new entity would have [35 to 45 %] * of the United Kingdom market (by PCD). POPC's strongest brand in the United Kingdom is P& O Cruises with [20-30 %] *. Carnival's shares in the United Kingdom are spread over a number of brands with Carnival's leading United Kingdom brand Cunard accounting for [0 to 5 %] *.

(132) The Commission has been unable to estimate reliably the Parties' combined market share measured by value. It would, however, be higher than their PCD share because the Parties operate in the premium segment of the oceanic cruises market and therefore have higher than average cruise prices per day.

(133) The Commission has assessed how changes in fleet composition and intentions regarding ship deployment and passenger mix in the years up to 2004 are likely to affect PCD market shares. The Commission in its assessment has taken into account the announced capacity changes by the Parties and their competitors and has also asked market participants for unannounced capacity changes. This is the time horizon for which the delivery of new ships can be established with a relatively high degree of certainty. The result of this investigation was that Cunard's market share might increase [...] *(58). As POPC is also increasing its capacity (59) above average, the market share of the combined entity could rise slightly in the next two years if it does not retire additional tonnage to that already planned. However, due to the uncertainties inherent in such forecasts the Commission does not consider it appropriate to base its decision on them.

C.2.3. United Kingdom brand positioning

C.2.3.1. Preferences: Language and cultural codes: British/International

(134) A section of United Kingdom customers prefers cruises that reflect their national tastes and preferences and on which their national language is the only or predominant on-board language. The Commission's market investigation has found that, by contrast, a certain percentage of the British population prefer cruises that are felt to be international. There are basically two international offerings, United States-style cruises and Mediterranean pan-European-style cruises. There is also a certain percentage of cruise customers who have no clear preference between the two main groups. The average estimates given by travel agents in the Commission's investigation were around 60 % British-style preference, around 25 % international-style preference and around 20 % without preference (60). The [POPC-commissioned] * study arrives at different figures. Depending on the group considered, this study examines preferences for "Britishness" and arrives at the following results:

only [25 to 35 %] * of "potential cruisers" would prefer a British-style cruise but of these, [50 to 60 %] * would be prepared to pay a price premium. Older potential cruisers are more attracted to British-style cruising ([40 to 50 %] * among those over 65) as are multiple cruisers ([40 to 50 %] *). Above [35 to 45 %] * of P& O previous cruisers prefer British-style ships and above [45 to 55 %] * of a group of other British-style premium brands, including Cunard, Fred Olsen, Page & Moy, Saga, and Swan Hellenic. However, [60 to 70 %] * of previous cruisers and [50 to 60 %] * of potential cruisers, have no preference or prefer United States cruises. Preferences for United States-style ships vary strongly and are between around [5 to 15 %] * (among Cunard's group of "British-style cruisers") and around [25 to 35 %] * (among United States-style previous cruisers) (61).

(135) These results indicate that while British-style cruising is preferred by a section of potential and past United Kingdom customers, an even larger group have indicated no particular preference, while another group would prefer United States-style cruises. As regards Cunard, it should be noted that while it is perceived to be a British brand a large number of their customers are United States customers, particularly on one of its two brand focuses, namely the "transatlantic liner" product currently provided by Cunard's ship the Queen Elizabeth II which will as of 2003 be provided by Cunard's new ship Queen Mary.

C.2.3.2. Next best alternatives

(136) In a survey conducted on behalf of RCCL, researchers asked past cruise customers what they viewed as the closest alternative brand to the one they had chosen (62). The results of this survey are summarised in Table 7. Two points should be emphasised. First, the survey excluded the pan-European and Mediterranean products that represent almost 10 % of oceanic cruises sold in the United Kingdom. Second and more significantly, the questions only asked for information on the customers' next best alternative rather than information on all of their perceived alternatives. The results therefore provide a narrow view on customers' perceptions of the competitive relationship between cruise operators. For example, a P& O Cruises customer who considered Fred Olsen as her next best alternative may also have considered another operator such as Thomson as an alternative. These "secondary" relationships are not captured in the information presented in Table 7. The table gives an indication of customers' alternatives, but does not fully reflect all the alternatives which customers considered when they selected their cruise.

Table 7 - United Kingdom customers' next best alternative

<emplacement tableau>

Source:

Conquest Report on telephone interviews conducted among United Kingdom cruise passengers, May 2002.

(137) The largest percentages of Cunard and P& O Cruises customers view each other as the closest alternative (46 % and 26 %). However, P& O Cruises' and Cunard's customers see a wide range of other competitors as these brands' closest competitor. For P& O Cruises customers, British-style Fred Olsen is seen as the closest alternative by 18 % of respondents and by 13 % of Cunard's customers. Fred Olsen customers considered P& O Cruises (45 %) and Cunard (13 %) as their closest alternatives.

(138) P& O Cruises' and Cunard customers also consider United States-style products as the closest alternative. Overall, 24 % of P& O Cruises' customers see the competitors RCI, NCL or Celebrity (brands owned by the Parties' competitors) as their closest alternatives. If Princess and Carnival CCL are included, this figure rises to 31 %. RCI is the single United States-style brand that is most frequently mentioned as the closest alternative by Cunard and P& O Cruises customers. Customers of United States-style brands consider P& O Cruises and other United States-style brands as the closest alternative. These findings support those of the [POPC-commissioned study] * (see recital 135 above) which show that a significant proportion of potential cruisers in the United Kingdom has no preference between British-style and United States-style cruises.

C.2.3.3. The Parties' position in cultural style segments of the United Kingdom market

(139) POPC is present in the United States-style segment with Princess, which has a share of [0 to 10 %] * in the United Kingdom. The merger will add HAL (perceived to be a close competitor of Princess in the Commission's market investigation) and CCL which each have 1 %. It would also have a slightly enhanced presence in United States-style cruising and a small but unchanged one in Mediterranean/pan-European-style cruising.

(140) As a consequence of the proposed concentration, the new entity would therefore have a stronger position in British-style cruising, however, as shown in the [POPC-commissioned] * study (see recitals 134 to 135 above), customers of British-style cruises also consider other types of cruises as viable alternatives.

C.2.3.4. The Parties' positions in regard to quality segments in the United Kingdom

(141) The Parties offer a wide range of services, from luxury brands like Seabourn and Windstar, over upper premium brands such as Cunard, HAL, Princess, P& O Cruises, Swan Hellenic and lower premium brands like CCL and Costa, to the recent project by POPC to launch an economy brand, Ocean Village. It is therefore particularly in the premium segment (s) that the Parties' activities overlap. In each of the segments, the Parties will be confronted with a number of strong competitors.

(142) In the first Conquest report, United Kingdom customers were asked on their perceptions of price and quality concerning cruise brands. In regard to quality, P& O Cruises was close to Cunard whereas in price perception it was close to RCI(63). It should also be noted that this section of the Conquest report only refers to seven brands and does not consider several of the options available to consumers, such as NCL, Orient, Saga and Celebrity. This limits the relevance of the results.

C.2.3.5. The Parties' positions in regard to "classical" cruises v "fun" cruises

(143) There is some horizontal overlap between Cunard and P& O Cruises in traditional British-style cruising. The main similarities between the Parties' offerings in this regard come from the cruises on board Cunard's Caronia, a 30-year-old liner.

(144) There are also significant differences between the products offered by these two brands. Cunard differs from P& O Cruises in that a significant part of its business concerns Atlantic crossings which are particular to Cunard. These are currently provided by Cunard's liner, the Queen Elizabeth II. The transatlantic crossings provided by this ship are unlike anything offered by any of the other cruise operators and are sold more to United States customers than to United Kingdom customers.

(145) The market share currently represented by the Caronia is less than [0 to 5 %] * of the United Kingdom market. Altogether, Cunard's market share represents less than [0 to 5 %] * of the United Kingdom market.

(146) The Parties' position in relation to British-style cruises may increase [...] *. However, there is a significant degree of doubt about precisely how the new capacity will be deployed and where it will be marketed. It is therefore not possible to establish with sufficient certainty what the impact of new capacity in the United Kingdom might be.

(147) In the United Kingdom segment the Parties will be able to offer a range of products ranging from "fun" cruise offerings such as CCL, to more "classical" products like HAL with Princess' brand profile being situated in between these, stressing both style and informality. In each of these segments the Parties will be confronted with a number of competitors.

C.2.3.6. Brand awareness

(148) In terms of brand recognition by United Kingdom customers, a study commissioned by POPC in 2001 (64) showed that consumers are most likely to be aware of P& O Cruise's brand (values are for cruisers/non-cruisers: [60 to 70 %] */[55 to 65 %] *) followed by Airtours ([25 to 35 %] */[15 to 20 %] *), Thomson Cruises ([20 to 30 %] */[15 to 25 %] *), Cunard ([20 to 30 %] */[25 to 35 %] *), Fred Olsen ([20 to 30 %] */[15 to 25 %] *) and RCI ([20 to 30 %] */[10 to 20 %] *).

(149) The [POPC-commissioned](65) study showed that among those respondents who were likely to cruise between 2002 and 2004, the largest portion of respondents mentioned P& O Cruises as their most likely choice of cruise brand, followed (in order) by Royal Caribbean, Cunard, Thomson, Airtours and Celebrity.

(150) This shows that, P& O Cruises aside, consumers are equally aware of both premium and economy cruise brands.

C.2.3.7. Evidence on closeness of competitors in Cunard's internal papers

(151) [...] *(66).

(152) [...] *(67).

C.2.3.8. Conclusion on United Kingdom brand positioning

(153) It can be seen from the above that P& O Cruises and Carnival's brand Cunard are competitors in the British-style cruise segment. Other brands, whether active in the same segment (such as Fred Olsen) or in a neighbouring segment (such as RCI) are seen as close competitors by consumers.

C.2.4. Competitive constraints

C.2.4.1. Market growth in the United Kingdom

(154) There have been high levels of growth for cruising in the United Kingdom in the 1990s, both in absolute terms and in comparison to the growth rates seen for other types of holiday.

(155) Over the period 1995 to 2001, the average annual growth rate in terms of the number of passengers was nearly 17 %. Figures from the Passenger Shipping Association show that the number of cruise holidays taken by United Kingdom passengers more than doubled over this period. Between 1995 and 1998 these annual growth rates were all over 20 % (1995 - 26 %, 1996 - 22 %, 1997 - 25 %, 1998 - 27 %). Although growth rates fell below these levels in the years 1999 to 2001, - 1999 (13 %), 2000 (1 %), 2001 (3 %), there remain expectations of high levels of growth in the market. For example, two independent sources (68) have given annual growth estimates for the United Kingdom of 8,5 % to 9 % for the years to 2006 and 2010. It should be noted that one of these estimates was published after the events of 11 September 2001. This growth will create incentives for both new entry and expansion from the current operators, and will therefore limit the impact of the barriers to entry.

C.2.4.2. Recent history of entry and expansion in the United Kingdom

(156) The United Kingdom oceanic cruises market has also been characterised by significant entry in the past five to six years. The history of entry into the cruising sector in the United Kingdom is largely a history of tour operators' entry into the economy segment of the United Kingdom market. The tour operator companies now carry 25 to 30 % of United Kingdom cruise passengers. In addition, a number of players have entered the premium segment.

(157) Airtours first offered cruise holidays in 1995, with an 800-berth ship, Seawing, operating under the brand Sun Cruises. Since then, Airtours has expanded its fleet with the addition of three other ships, and now has a total capacity of 4200 berths in summer (69). Its latest ship, the Sunbird is significantly newer and provides better facilities than its other three ships.

(158) Thomson, recently acquired by the German group Preussag, first offered cruises in 1996 and expanded its offer in 1997, 1998 and 1999. It currently operates three ships in summer (70), having reduced its capacity in 2000. May 2003 will see the addition of a new ship to the Thomson Cruise programme, the Thomson Spirit, which originally entered service for the Holland America line in 1983. This may replace one of their existing, older ships.

(159) The third TOC to have entered the market was First Choice who started their cruise operations in 1999 with a ship chartered from Festival cruises, the Bolero. In 2000 it added the Ausonia. Having terminated these ships' charters, First Choice relaunched its product in 2002, in a joint venture with RCCL, Island cruises. Island operates Island Escape, a 12-year-old ship formerly operated by RCCL.

(160) Companies that have entered the premium market in the United Kingdom are players such as Saga in the United Kingdom. Saga began by selling holidays to senior citizens in 1951 and today provides a broad range of services to people aged 50 and over, such as insurance and financial services, home shopping, gas, electricity and telephone services, publishing and radio broadcasting. Saga entered cruising in 1997 with a small ship (587 berths) which it had acquired from Cunard. In 2003 it will add another, slightly smaller ship (previously operated by POPC, 352 berths) which it will apparently charter for part of the year (May to October).

(161) Festival, a Mediterranean player entered the economy cruising segment during the 1990s and the premium cruises segment in 1999. It partly sells through tour operators' brochures. Its United Kingdom premium segment share is now 1 %.

(162) One other player has entered the premium segment of the market: Disney Cruise Line (Disney). Disney operates two ships which are focused on the United States market but which are also sold to United Kingdom and other European customers. Disney is still a relatively small player in the United Kingdom (market share < 0,5 %). Canyon Ranch plans to enter a niche of high value spa cruising around 2004/2005 but will likewise mainly market to United States customers.

(163) Fred Olsen, who currently has about 6 % market share, is a company at the lower end of the premium segment. It operates three ships and has expanded its offering twice in the last 10 years. In 1996 it added a second ship and in 2001/2002 a third. Its net growth in berths, however, stayed behind P& O Cruises' which during the same period added ships in 1995, in 1998 (in replacement for an older one), in 2000 and in 2002 (against one smaller retirement) (net gain in berths: P& O Cruises: 4722 berths; Fred Olsen: 1536 berths).

(164) Between 1995 and 2000, RCCL maintained its market share (71) of 6 to 7 % in the overall oceanic cruise market in the United Kingdom by increasing the number of passengers it carried from 22000 in 1995 to 50000 in 2000. In 2001, RCCL increased its market share to 10 % with an increase in the number of United Kingdom passengers that it carried of nearly 30000. RCCL achieved this market share increase without dedicating ships to the United Kingdom market, but by changing the nationality mix on its ships and increasing the proportion of United Kingdom passengers.

C.2.4.3. Forthcoming developments

(165) The traditional premium players are launching brands that are more closely targeted at the upper end of the economy segment. For example, RCCL has recently launched Island as a joint venture with First Choice. Island started operations in 2002. In a similar vein, POPC will be launching a new brand, Ocean Village, in 2003. The latter brand is set to target a younger, more "informal" customer segment. The two new brands will be operating with relatively recent ships transferred from their parents' activities. Ocean Village, for example, will operate with the Arcadia, which to date is part of P& O Cruises' fleet.

(166) Cunard is planning a number of significant changes to its fleet. These plans are described in detail in footnote 55.

C.2.5. Barriers to entry and expansion

(167) The Parties view entry barriers as being low because ships are readily available and can also be refurbished quickly. In their view, the capital requirements for buying or leasing older ships are not high and access to distribution is easy. They consider that brand awareness is not a barrier as there are many brands in the consumer goods or tourism sectors, which can be extended into cruising. That the United Kingdom market is growing further enhances the possibilities for entry and expansion by other players.

C.2.5.1. Brand awareness

(168) In 2001, Carnival's top selling brands in the United Kingdom (Cunard, CCL and Costa) spent [...] * % and [...] * % of their annual revenue, respectively, on marketing. For Cunard this amounted to GBP [...] *. Of Cunard's marketing budget, [...] * % was spent on newspaper and magazine advertising, a large part of which is attributable to building brand recognition and brand awareness.

(169) The Commission has seen evidence that marketing expenditure for established brands is normally between 5 and 10 % of annual revenue whereas it can be 10 to 15 %, sometimes up to 20 % for brands that wish to expand their operations. Marketing expenditure is usually higher in the educational/discovery segment and in the luxury segment.

(170) Unless new entrants already have some brand equity that can be used to sell cruises, the expenditure on branding may make it more difficult for new players to enter the market. In this context, it is notable that most of the new entrants to the cruising market are companies that have established brands in a range of related fields. The TOCs were providing holidays before entering the cruising market. Saga has been able to successfully extend its established brand (a brand associated with providing a range of services, among them holidays, to elderly customers via direct marketing) into cruising. The example of Island (owned by First Choice and RCCL) and the forthcoming launch of Ocean Village (POPC) shows that new brands can be launched by existing operators and therefore that the costs of launching a new brand are not prohibitive.

(171) Another possibility is entry or expansion from another geographic market. Although the brand recognition of Mediterranean operators such as Festival and Royal Olympic in the United Kingdom is still relatively low, they have upgraded the quality of their offering in the United Kingdom in the last two to three years. This shows that entry and expansion is possible even without high brand recognition.

C.2.5.2. Marketing and distribution

(172) Most sales of cruises in the United Kingdom are made through travel agents. Very few sales are through the internet, and a not insignificant amount of sales are made via telephone marketing. Access to the retail channel of travel agents and active promotion by the sales agent is therefore indispensable to the success of a cruise operator.

(173) Multiple travel agents have indicated that they do not consider that limitations on racking space would hinder access to distribution. As an indication of depth of offering, one major multiple travel agent said that its agencies currently stock up to 13 different cruise lines' products on their shelves. It should also be noted that cruise brands usually make a significant proportion of their sales via independent travel agents and cruise specialists.

(174) Concerns have been raised in relation to the structure of the travel agents' commissions, in particular that they could provide incentives for them to favour the sales of the stronger brands. However, the Commission does not have evidence that this has prevented or could prevent entry or expansion in the market and therefore does not consider that access to distribution constitutes a major barrier to entry into the United Kingdom cruise market.

(175) Cruise suppliers also need to "train the trade" because travel agents are reluctant to recommend cruise products to their customers if they are unfamiliar with the product on offer. However, this is also a cost factor for existing companies, as they need to keep the distributors' knowledge up to date. Furthermore, joint marketing activities with travel agents are a means to encourage promotion of an operator's product. This is a method that both new entrants and existing operators can use.

(176) One particular aspect of marketing and distribution that plays a role is the possession of a customer database and access to repeat sales. Access to repeat sales is significant because once customers have cruised once, they are far more likely to cruise again. Repeat customers are therefore the lowest cost customers to gain. It also appears that repeat customers render a higher than average yield. This has been contested by Carnival but is confirmed by its figures on Cunard (including Seabourn) for part of 2002, Cunard/Seabourn's repeat passengers accounted for [...] * % of passengers, [...] * % of cruise days sold but for [...] * % of revenue generated(72). A past customer database can help to target promotional activities and therefore help to sell the berths at acceptable prices. Cunard's past passenger database encompasses [...] * households in the United Kingdom. Costa's past customer database contains the names of [...] * British customers (73). In addition, POPC has a significant past customer database.

(177) Repeat customers make up a substantial proportion of passengers on cruises. For example, Carnival has estimated Cunard's United Kingdom repeat customer rate to be between [...] * %. The total percentage of all United Kingdom repeat cruisers (i.e. customers that have cruised before with this or another line) on Cunard ships is even estimated to be up to [...] * %. The target company, POPC has also cited high repeat rates.

(178) It has been argued by RCCL that newcomers would not have access to a database of previous cruisers, thereby increasing the costs of selling their cruises. This is because it is less expensive to market to previous customers on whom the companies hold information. In its Reply, Carnival has argued that Conquest, for its report paid for by RCCL was able to purchase the names of approximately 10000 United Kingdom based consumers who had previously indicated that they had taken a cruise in recent years. This indicates that information on previous cruisers (who are potential customers) would also be available to new entrants.

(179) Moreover, many of the recent entrants, such as the TOCs and Disney, have existing operations either in relation to the supply of holidays from which they could gather related information, or as in the case of Saga, in the sale of goods and services to the demographic groups which are prime cruising targets. Saga's entry was further facilitated by the fact that it sold, and still sells, other cruise companies' cruises to its customers. Other companies could take advantage of similar information that they hold.

(180) POPC has highlighted that new entrants could undertake joint marketing campaigns with the travel agents which could enable the entrant to access the data held by the travel agents.

(181) The Commission therefore does not consider that there are significant difficulties in accessing the necessary means of distribution, nor in obtaining the information necessary to enable marketing to be targeted to potential customers in a cost-effective manner. These factors are therefore not considered to represent significant barriers to entry.

C.2.5.3. Access to ships

(182) Cruise operators can either order new ships, buy second hand ships, or they can charter ships and crews from companies who specialise in such chartering. The Parties account for the majority of current orders for cruiseships in the world(74). The capital costs of new ships, especially post-Panamax vessels, are very high, with the cost of new ships running into hundreds of millions pounds sterling.

(183) However, in terms of their competitive effects the magnitude of these costs matters less than the question of whether or not operators could recoup these costs on exit, in other words, whether or not these costs are sunk.

(184) Three factors limit the risk that sunk costs incurred from investing in new ships represent a significant barrier to entry. First, it is not necessary to enter the oceanic cruise market with ships that are newly built. Entry with new ships has taken place extremely rarely. Disney is the only operator currently active in the United Kingdom who has entered with new ships. Second, even if an entrant did invest in new ships, the existence of a market for the sale (and subsequent refurbishment) of these ships suggests that operators who wish to exit the market could recoup a proportion of their investment. Third, the risks of the investment costs being sunk costs would be significantly greater if the oceanic cruise markets were shrinking, rather than exhibiting growth.

C.2.5.4. Economies of scale and finance

(185) Operating a large fleet of ships generates economies of scale. Various cost functions such as overheads (e.g. for technical operations, hotel costs, and shore-side operations), can be spread over a larger fleet.

(186) Economies of scale are not only achievable by operating more ships but also by operating larger and more modern ships. Larger ships can spread various costs (e.g. technical staff) over a larger number of customers. Newer ships are cheaper to operate as they have, for example, significantly lower fuel costs. Entry into a European market with a big new ship, however, is difficult. Even in the United States, by far the biggest market worldwide, only one such entry that has happened in the last 10 years: Disney. These economies of scale are therefore only in very rare and exceptional cases achievable by new entrants. As regards expansion, only the majors and, to a lesser extent, a few second tier players can fully achieve these economies. On the other hand, entrants using second hand ships or charters can reduce their financing costs and risks substantially.

(187) The Parties have contested the argument that there are significant economies of scale, arguing that if there were such economies then one would expect to see a trend towards an increasingly concentrated industry. POPC highlights that over the last few years, the industry has become less concentrated as a result of the activities of the newer operators.

(188) A number of players (e.g. Airtours, Thomson, Saga and Island) have entered with one ship, then expanded by adding further ships. This shows that entry on a relatively small scale is possible, and therefore that achieving economies of scale is not indispensable for successful entry.

C.2.5.5. Switching nationality-mix/capacity between countries

(189) The ability to shift brand marketing from one country to another and the modification of nationality mix on ships is an important feature of this industry. Recent examples have shown that competitors can rapidly expand their operations in a country by shifting capacity on multinationally marketed ships, such as on United States-style ships and on Mediterranean European-style cruises.

(190) In its Statement of Objections, the Commission reached the preliminary conclusion that United Kingdom consumers' preferences for British-style cruises would significantly inhibit the ability of those operators providing either Mediterranean-style or American-style cruises to exert significant competitive pressure on the Parties, because such preferences could inhibit the ability of those operators who operate ships with people from more than one country to increase the proportion of United Kingdom customers.

(191) However, in the light of the Replies received, and of analysis undertaken since the Statement of Objections was issued, the Commission considers that this preliminary conclusion overplayed the significance of such national tastes as a barrier to entry. As already shown in recital 138 above, nearly a quarter of P& O customers considered that the United States-style brands RCI, NCL and Celebrity were their closest alternatives. Evidence provided in the [POPC-commissioned] * study (see recital 134 above) reinforces the conclusion that United States-style and United Kingdom-style cruises are in close competition. The [POPC-commissioned] * study suggests that while there is a significant minority of potential customers who have a preference for a United Kingdom-style cruise, the majority of actual and potential customers are either indifferent or would prefer an American-style cruise.

(192) Furthermore, evidence provided in Carnival's Reply shows a number of examples where there were significant shifts in the nationality mix on board its ships, such as the increases in the proportion of United Kingdom customers on the Cunard ships. For example, in two years, the share of United Kingdom customers on Cunard's Caronia grew by [...] * percentage points. There were also several other examples given of shifts in the proportion of United Kingdom customers of greater than 10 % between 1999 and 2001.

(193) RCCL has recently substantially increased its market share in the United Kingdom by increasing the number of United Kingdom passengers on its ships. It is particularly relevant that the only cost it incurred to achieve this expansion was the money that it spent on marketing. No costs were incurred in relation to either the hardware or the software on the ships, and RCCL achieved this expansion in market share without dedicating a ship to the United Kingdom market.

(194) In addition to expanding by changing the nationality mix, cruise operators could also continue to reallocate whole ships between different countries. Such "ship shifting" is a significant factor. For example, a report published in March 2002 by UBS Warburg (75) shows that in the period 2001 to 2003, 10 ships are being reallocated from the United States market to Europe. These ships represent a shift of 7 % of capacity in the United States market to European markets. Among these ships are POPC's Crown, Regal and Ocean. Another example which highlights the flexibility with which ships can be used is RCCL's Viking Serenade which is being run by the newly launched brand Island Cruises, and which has been running cruises in Europe this spring and summer and which will return to the United States market for 2002-2003 winter season.

(195) The Commission considers that an attempt to raise prices post-merger would be countered by both United States and pan-European players, in particular RCI and NCL, who could increase the proportion of United Kingdom passengers on their ships in the short term. This would not require the establishment of new brands or distribution networks, as these companies are already active selling cruises in the United Kingdom with brands that are recognised and with access to the necessary distribution networks. In the medium term, these companies could build on this presence to dedicate ships to the United Kingdom market.

C.2.6. Conclusion for the United Kingdom

(196) It is clear that the operation will create a combined entity that has a relatively high market share in the market for oceanic cruises in the United Kingdom. However, this alone does not confer dominance and there are several specific characteristics of the oceanic cruises market in the United Kingdom which have led the Commission to conclude that the proposed operation will not create competition concerns.

(197) The United Kingdom market is relatively immature and is forecast to grow rapidly in the next few years. This growth will provide opportunities for existing players to expand their operations and for new players to enter the market. The ability for new entry and expansion has been proven by recent history in the market.

(198) In the British-style segment, the overlap between Cunard and P& O Cruises is limited. Cunard accounts for less than [0 to 5 %] * of the United Kingdom market. Most of this is accounted for by the Queen Elizabeth II liner. A significant part of Cunard's business is in niche products, such as its transatlantic crossings onboard the Queen Elizabeth II, which are unlike the cruises typically sold by P& O Cruises.

(199) Most potential cruisers in the United Kingdom do not limit their choice of cruises to those companies providing "British-style" cruises. Instead, they consider cruises with a range of characteristics provided by a range of operators.

(200) There are no significant barriers to expansion for international operators. In the short term (within a year), these operators can expand by increasing the proportion of United Kingdom customers on their ships and the threat of such expansion will be a significant competitive restraint on the combined entity. In the medium term (in 1 1/2 to 2 years), they could further enhance their position in the United Kingdom by dedicating ships to the United Kingdom market.

(201) The Commission therefore concludes that the operation will not create or strengthen a dominant position on the United Kingdom market for the provision of oceanic cruises as a result of which effective competition would be significantly impeded in the common market and the EEA.

C.3. GERMANY

C.3.1. German market shares and capacity

C.3.1.1. German market shares for 2000 by passengers

(202) Carnival estimated the Parties' market shares for the year 2000 on a maritime cruise market including coastal ferry cruisers such as Hurtigruten (76) are given in Table 8.

Table 8 - Parties' estimated passenger market shares in Germany in 2000

<emplacement tableau>

(Source:Carnival).

C.3.1.2. German market shares 2001 by PCD

(203) The total market volume 2001 in Table 9 was derived from the annual survey on the German cruise market published by DRV, which is considered the most reliable source for the German cruise market (77). An "other" category fills the difference between the total PCDs of the Commission's investigation and the total market volume as stated in the DRV publication (78). This "other" category covers a number of luxury cruise brands (79) and a number of niche players (80).

Table 9 - German PCD market shares 2001

<emplacement tableau>

Source:Commission, based on information supplied by parties and competitors.

(204) Table 9 shows that the Parties' combined PCD market shares for 2001 are [20 to 30 %] *. This is lower than Carnival's estimate for 2000 on the basis of passengers.

C.3.1.3. Current market shares by value

(205) The Parties' current value market shares are likely to be somewhat lower than their PCD market shares, due to the price premium German traditional products currently command(81).

C.3.1.4. Forthcoming capacity changes

(206) The German market is forecast to expand rapidly. G.P. Wild assumes as a minimum 8 % annually, and Carnival's own comments are even based on annual growth rates of 25 % or 39 %(82). The Commission believes that Carnival's estimates are very optimistic. Other market participants have given much lower estimates. However, generally, the expectation of strong growth as such is undisputed. In line with their growth expectations, some market participants are planning to expand capacity considerably. Foremost among these are Carnival and POPC, which by 2004 intend to have six instead of, in 2001, two ships dedicated predominantly to the German market. About half of this capacity increase has already been realised in 2002 in connection with the Parties' brand building exercises (see recital 222).

(207) In 2001, POPC operated two ships with a total capacity of 1686 lower berths that were dedicated to German-style market segment (Arkona - 500 berths; AIDAcara - 1186 berths). In the years to 2004, four additional ships are planned to be added to POPC's dedicated German capacity, whereas the operation of the Arkona was stopped in 2002. The new ships will be partly new-builds, partly transferrals from POPC's United States brand Princess. This would have the effect of increasing POPC's German capacity to a total of 6906 lower berths. POPC's (Seetours') German-style capacity in 2004 can be expected to be composed of the following ships: AIDAcara - 1186 berths; AIDAvita - 1270 berths; AIDAaura- 1270 berths; A'Rosa Blu/ex-Crown Princess -1590 berths; A'Rosa 2/ex-Regal Princess - 1590 berths.

(208) The Carnival ship currently operating exclusively for German customers under Costa's "La Deutsche Vita" sub-brand, Costa Marina, has a capacity of 763 lower berths. In addition, [...] *(83).

(209) A number of "pan-European" Mediterranean companies offering cruises that wish to attract customers from several European nationalities, are likewise increasing their capacity. Festival, MSC, and Royal Olympic, all represented on the German market, have put new ships into service in recent years, or are about to do so in the coming years. The capacity of these operators will rise by above 40 % between 2001 and 2004, going along with a substantial upgrading of their fleet. The effect of this increase will be dispersed over a number of geographical markets as each of these operators only draws a certain amount of passengers on board its ships from Germany.

(210) All United States brands are planning to increase their worldwide capacity between 2001 and 2004. The majority of this capacity increase by United States brands is carried out by the Parties themselves, but a substantial capacity increase is also foreseen for RCCL. To a smaller extent NCL is also increasing its capacity. The percentage of German customers on United States-style ships is currently small.

(211) The strongest group of competitors, by market shares in 2001, was composed of mostly traditional German cruise companies who are not expected to expand their capacity to any comparable extent.

(212) The Commission in its Statement of Objections presented a forecast of market shares for 2004, based on the above capacity changes, which showed that the Parties' combined market share would be significantly higher than in 2001. A forecast prepared by RCCL on the basis of expected passengers in 2005 also arrived at a higher market share for the Parties than they held in 2001.

(213) Carnival and POPC, in their Replies to the Statement of Objections, have questioned the assumptions on which these forecasts were based, in particular in relation to capacity utilisation, the nationality mix and certainty of capacity forecasts.

(214) The Commission accepts that these assumptions introduce significant uncertainty into any projection of market shares as any of these assumptions may not materialise. The Parties and their competitors may change their plans concerning their capacity. Nationality mixes on-board ships may also shift and capacity utilisation cannot be predicted with sufficient certainty for individual operators (84). While the Commission expects the Parties' combined market share to grow as a result of the capacity changes outlined in recitals 206 to 208, given the uncertainty in forecasting the precise effect of these changes, it does not consider it appropriate to base its decision on the market share forecasts presented in its Statement of Objections.

C.3.2. Brand positioning

C.3.2.1. Language and cultural codes: German/International

(215) The largest part of the German oceanic cruises market, around 50 % of the market (85), is accounted for by products offering cruises in German, geared to German national tastes.

(216) The percentage of "international style" cruises is more limited. Among these international-style cruises, pan-European Mediterranean-style cruises, as offered by Carnival's Costa brand and by Festival have seen considerable growth rates in recent years, reaching, together, around 25 % market share.

(217) United States-style cruises still appeal only to a minority of German customers (around 5 %) but may have a certain growth potential as the number of younger repeat cruisers with a better command of English rises (86). It should be noted that RCI and Celebrity currently have the highest market share in this segment

(218) A particularity of the German market vis-à-vis the French, Italian and Spanish markets is that Cunard, offering predominantly a "classical British" product is comparatively well known in Germany and attracts a certain group of German repeat cruisers.

(219) One of the main characteristics of the German market in recent years has been the rising importance and attractiveness for German customers of pan-European Mediterranean offers.

(220) Carnival's Costa brand (in terms of sales jointly with MSC but with a much stronger brand presence) is the segment leader in pan-European cruises for German customers. This strong position is reflected in strong passenger growth between 1999 and 2001. Also Festival, in line with its strong capacity growth in these years, has increased the number of its German passengers considerably, but remains a distant third behind Costa and MSC.

(221) German-style cruises have been for a long time dominated exclusively by traditional German cruise operators, most notably by Hapag Lloyd, Peter Deilmann, Phoenix Reisen and Transocean Touristik but also by Delphin Seereisen, Plantours and, more recently, Hansa Touristik. Most of these companies are small companies owning or frequently chartering one or two mid-size ocean cruise vessels. Peter Deilmann operates two ocean cruise vessels plus a sail cruise ship for Galapagos cruises. Hapag Lloyd, owned by Preussag/TUI, operates four ocean cruise vessels and two sail cruise vessels.

(222) This situation has begun to change with the acquisition of Seetours and AIDA by POPC. The recent creation of a new German POPC brand, A'Rosa, and Costa's decision to dedicate one of its ships to the German market exclusively in 2002, creating a new sub-brand, "La Deutsche Vita" will accelerate this shift in the prevalent segment of the German market away from traditional players to national brands owned by the majors, currently Carnival and POPC.

C.3.2.2. Quality segments and star ratings

(223) Other elements of brand positioning are provided by the categorisation into luxury, premium (upper premium and lower premium) (87) and economy/budget segments. These categorisations, although not identical, are linked with star-ratings such as provided by Berlitz.

(224) In this context it needs to be noted that the Parties' offerings predominantly fall within the premium segment or segments (roughly corresponding to four-star and three-star+). These are the core segments of the cruise offering and it is only in these premium segments that a product can both command a higher margin and attract broad consumer groups (88).

(225) The economy segment of the German market is smaller than in other European countries and is composed on the one hand of a few German companies and on the other hand of Mediterranean companies like Louis, ROC, MSC and Festival who operate two-star and three-star economy ships but most of which also try to penetrate the premium segment (s). In terms of passengers, the segment constituted around 20 %(89) of the overall German oceanic cruise market in 2001, but much less than that in value terms. It is significant that operators present in this market segment are seeking to penetrate the premium segment (s) by ordering new ships.

(226) Of the German oceanic cruise market, 11 % in terms of passengers is occupied by luxury offerings (90). Hapag Lloyd and Peter Deilmann are strong in these luxury segments. In addition to other international companies like Silversea and Radisson, Carnival is also represented in these segments with its "international" offerings Seabourn, Windstar and Cunard(91).

(227) The Parties' future position in the premium segment (s) of the market, constituting around two thirds of the total market (92) can be expected to be relatively strong. Whilst A'Rosa is placed in the upper premium segment the other two brands are placed somewhat lower. However, in spite of the Parties' strengths, companies like Peter Deilmann, Hapag Lloyd, Phoenix, Transocean and other smaller German operators can continue to count on their traditional repeat clientele and may also seek to modernise their product.

(228) Furthermore, competition in the premium segment (s) would come from the pan-European Mediterranean operators having upgraded part of their fleet while remaining active in the economy segment. In this way they could profit from repeat customers' tendency to "trade-up". They have experience in accommodating German tastes and can have German speaking full-ship chartered cruises.

(229) Finally, RCCL's and NCL's ships could constitute an attractive offering in the premium area with their stunning big new ships, which could appeal to a younger more international-minded audience.

C.3.2.3. Classical cruises v fun cruises

(230) Traditional "classical" cruises are offered mainly by the traditional German operators. These products focus on formal dress, more sedate entertainment, sometimes with an educational/cultural touch, and "old fashioned" high quality service. The emphasis on formality diminishes gradually in the lower categories but is never fully abandoned and formal dress is generally required for dining. Entertainment is generally aimed at the older customer to whom this product particularly appeals.

(231) Cruises that appeal to "fun and activity" values were introduced into the German scene by AIDA in the latter half of the 1990s with its clubship concept. The closest equivalent to that style of cruising is provided by the United States ships' so called "contemporary" offering, epitomised by CCL with variations offered by NCL, RCI and Princess. The fun and activity cruises provide a more casual and relaxed cruise experience with wider, more animated entertainment programmes. A variety of restaurants, bars and nightclubs cater to a broad mix of passengers. The larger the emphasis on fun and "club-style", the younger the customers tend to be and the more first-time cruisers tend to be among them. In Germany, AIDA has therefore a very high rate of first-time cruisers and, for Europe, an exceptionally low average age which is comparable to the average ages on United States-style "contemporary ships" whose average customer is in his/her forties.

(232) The positioning of both A'Rosa and Costa/La Deutsche Vita towards these opposite groups of "traditional" versus "contemporary" offerings is significant. POPC, giving up its Arkona brand which had operated in the classical segment until 2002(93) decided to create a new brand that aims at bridging the divide between both extremes. A'Rosa attempts to appeal to the more discerning, experienced, but "young-at-heart" customer (aged late forties to early sixties) offering him/her an experience that is both relaxed and refined, containing fun-elements that are not overly action-oriented. Its target group to some extent overlaps with both the traditional classical cruisers and the contemporary cruisers who in themselves do not constitute homogenous customer groups but represent a whole range of customer profiles.

(233) Costa/La Deutsche Vita on the other hand also addresses both customer groups albeit from a different angle, namely the angle of family-friendliness with an Italian/Mediterranean touch. The range of age groups catered for by the Costa/La Deutsche Vita brand can be expected to be quite broad and in addition to vary according to season, carrying a higher number of families with children in summer. The Costa/La Deutsche Vita brand has a decidedly relaxed, "contemporary" and moderately activity oriented (sauna, gym, jogging) approach. Bridging thus, like A'Rosa, the classical versus contemporary cruise divide, the La Deutsche Vita customer group can in a number of respects be expected to overlap with the A'Rosa customer group.

(234) Customers would find that a set of similar itineraries are offered by La Deutsche Vita and A'Rosa, both of which, in 2002, offer a series of comparable circular cruises out of northern Germany (Kiel, Warnemünde, Hamburg) alternating between Norway and the Baltic at comparable prices. However, customers of these cruises may also find it attractive to switch to the Mediterranean operators some of whose departure ports are in Northern Italy, which is quite easily accessible for a large number of German customers. The Mediterranean companies are also beginning to venture into Norwegian and Baltic Sea cruises (likewise with departure ports often within easy reach of potential German customers).

C.3.3. Dominance assessment - Germany

C.3.3.1. The Parties' market position

(235) The Parties' combined current market shares are currently clearly below a level that could raise concern in an expanding market. However, their shares can be expected to rise (see recitals 206 to 208).

C.3.3.2. Competitors

(236) Two of the Parties' three groups of competitors are also expanding capacity between 2001 and 2004. Both of them are likely to impose effective competitive constraints on the Parties. The United States-style competitors' brands like NCL and RCI can be expected to present a credible alternative to AIDA for some of AIDA's customers. A United States-style brand like Celebrity could constitute an alternative to A'Rosa and La Deutsche Vita's position can be challenged by the Mediterranean operators.

(237) In addition the established German operators, even though they are not increasing capacity, can be expected to constrain the Parties' market power with respect to customers preferring German ships and therefore being inflexible towards switching to international brands. Indeed the Parties must seek to enlarge their customer base at the expense of these operators.

C.3.3.3. The threat of capacity oversupply and the need to fill ships

(238) The basic economics of the cruise industry provide considerable incentives to arrive at 100 % or close to 100 % capacity utilisation. This is due to the fact that, once capacity is set the marginal costs are very low. Nearly all costs are fixed costs and only costs such as for food are variable costs. These costs are largely outweighed by the on-board revenue generated by each passenger. The only inhibition to last minute discounting is therefore the dissatisfaction discounted prices for late bookers can create for customers having paid a higher price earlier, and the resulting dispersion of future yields. The Parties, in line with these economic considerations, are openly committed to operating at 100 % capacity utilisation rate (94).

(239) In order to achieve the goal of very high capacity utilisation companies generally aim to manage their sales curves carefully and try to influence it via tactical pricing and other promotions.

(240) The Commission's market share projection has shown that the market would have to grow at a very fast rate to absorb the new capacity that would be put on the German market (around 20 %) if the planned capacity increases take place even at capacity utilisation rates which are certainly below those targeted by the Parties. German market growth by passengers was 2 % (in 1995); 17 % (in 1996); 11 % (in 1997); 8 % (in 1998); 8 % (in 1999); 15 % (2000); and 3 % (2001) (95). It may be noted that the German market before 11 September 2001, showed signs of accelerating expansion but a market growth of 20 % for three successive years seems very optimistic. Should growth stay behind expectations, it is difficult to predict what alternative scenarios may emerge.

C.3.3.4. Entry barriers and barriers to expansion

History of entry/Potential entry

(241) As regards entry barriers, similar considerations apply as are discussed above in the context of the United Kingdom (recitals 167 to 192 above). In Germany, entry in the economy segment in the past 5 to 10 years has been mainly characterised by entry and subsequent expansion by Mediterranean companies (such as MSC and Festival) selling part of their cruises to German passengers. Some smaller German companies, usually charterers, have also entered (see recital 243 below).

(242) In the premium area, Aida entered in the second half of the 1990s with one ship. Together with Seetours, it was acquired by POPC in 1999/2000. A second Aida ship has been added this year (and a third one will be added in 2003). Also in 2002, Costa (Carnival) created a dedicated German sub-brand (La Deutsche Vita) and POPC-Seetours started a new German brand A'Rosa.

(243) Whereas German tour operators have not shown much propensity to enter the oceanic cruise market, smaller companies, which may or may not be affiliated with tour operators, may continue to enter the market, as happened in 2001 with Hansa Touristik (chartering an old but renovated ship, the "Paloma") and as is foreseen for 2003 with Holiday Kreuzfahrten (operating P& O Cruises' current "Victoria" renamed "Mona Lisa"). Also ADAC Reisen, the travel branch of the German motorists' association, has entered cruising on a small scale as a charterer of cruises offered to its members. Expected market growth will facilitate this entry. At the luxury end of the market, there has been an entry by the operators of the two "Sun Bay" luxury ships. While it is not likely that such entry will directly challenge the Parties' brands, it is still possible that these small-scale entries may further constrain the Parties' market strength.

History of and potential for expansion

(244) In non-German-style premium segments of the German market, United States-style premium brands have had some limited success in Germany, which is exemplified by their low overall market share (around 5 %) and the fact that only two individual brands, RCI and Celebrity have shares larger than 1 %(96). The low speed of their expansion can be seen from the fact that they were already present on the German market in 1994(97). However, these brands can be expected to have the means to participate in growth of the German market, most of them offering new premium ships for "contemporary" cruises. This participation in growth could take the form of an increase in the percentages of German customers onboard their cruise ships, which would be achievable through increased marketing effort.

(245) The Mediterranean companies have been successful in entering and expanding in Germany. They are more firmly established in the economy segment and carry a significant percentage of Germans on their ships. Shifts in nationality mixes onboard Mediterranean ships have taken place in the past and may be expected to take place in response to any post merger price rise. Festival has successfully expanded into the premium segment with its new premium ships.

(246) The German operator Transocean Tours, which charters its ships, has expanded recently. After introducing a 590-berth ship in 1996 (98), a second ship was added in 2002 (of similar capacity, previously operated by POPC-Seetours). The entry pattern of Transocean Tours may be compared with Fred Olsen's in the United Kingdom. Other smaller German operators could follow this route. Expected market growth in Germany will facilitate expansion.

Brand awareness

(247) The traditional mode, still employed by market leaders Hapag Lloyd and Peter Deilmann, is the one-ship brand, which may then be combined with a company masterbrand. This may begin to change with the Parties' activities', promoting (what are planned to be) multi-ship brands, A'Rosa, AIDA, La Deutsche Vita. In 2002, with the beginning of operation of Costa's La Deutsche Vita sub-brand, Costa's marketing expenditure in Germany [...] * on the year before (99). POPC's marketing expenditure for A'Rosa and Aida is likewise considerable. Transocean's two-ship brand Astor/Astoria may be an early sign of a beginning restructuring of the German offering. Further indications are Hapag Lloyd's joint marketing of its two explorer ships and of its chartered sail cruisers "Sea Cloud".

Marketing and distribution

(248) The structure of the German travel agent market is characterised by the absence of specialised cruise shops and by a very high number of travel agencies, most of them grouped in cooperations (100). Access to this distribution is generally seen as easy. There is no indication that access to distribution would be more likely to constitute a barrier to entry after the merger.

(249) While a large part of marketing expenditure goes to advertising and the building of brand recognition, a substantial sum is also spent on other marketing activities such as training and familiarising the trade. Carnival's marketing in Germany will be strengthened by having access to Aida-Seetours' more firmly established sales and marketing network. However, this does not prevent competitors from increasing their marketing efforts themselves.

(250) Some marketing difficulties in Germany can be overcome by marketing via a general sales agent who may market several companies' products and, as a specialised tour operator, can put them into a common brochure. Several big tour operators are packaging cruise products and are producing cruising brochures themselves (FTI-MyTravel through Air Maritime; Thomas Cook through Neckermann and Terramar; TUI packages under its own brand). Marketing through this channel constitutes a viable alternative to own-marketing for cruise companies.

C.3.3.5. Conclusion on competitive constraints

(251) The Commission concludes that competitive constraints to which the Parties are submitted, particularly competition by expanding competitors (through the addition of ships and possible changes of competitors' nationality mixes on board their multinationally marketed ships) as well as the Parties' own need to expand sales in order to fill their own ships are sufficient to outweigh the risk posed by the Parties foreseen capacity increase which can, although this projection is surrounded by a number of uncertainties, be expected to lead to higher market shares. Entry barriers, moreover, are not considered high and will not be raised to a level that cannot be overcome by new entrants.

C.3.4. Conclusion on Germany

(252) The Parties' combined market share was [20 to 30 %]* in 2001 in PCDs. Considerable capacity increases are planned for both Parties' German-dedicated brands in the years up to 2004.

(253) The Commission has found that competitors can also expand their operations. In particular both Mediterranean and United States operators are in a position to change the nationality mix on board their multi-nationally marketed ships. They have access to sufficient capacity and are currently increasing their capacity. Mediterranean operators in particular, have successfully penetrated the German market in recent years and have, as is the case with Festival, achieved an upgrading of their offering. In addition, some smaller German companies, usually charterers, have entered the market.

(254) The Commission has found that entry into and expansion in the German market is not difficult. No substantial barriers to entry would be created subsequent to the proposed concentration. New entry can therefore be expected to fuel competitive conditions on the German market for oceanic cruises.

(255) An element that was found to support both expansion by existing competitors and new entry is the expected strong growth of the market. A further effective constraint on the Parties' ability to raise prices in the years to 2004 is the capacity increase to which they are partly committed.

(256) The proposed concentration will therefore not lead to the creation or strengthening of a dominant position on the German market for the provision of oceanic cruises.

C.4. MEDITERRANEAN MARKETS: ITALY, FRANCE AND SPAIN

C.4.1. Configuration of the markets in Italy, France and Spain: general aspects and Parties' position

(257) The cruise industry in Spain, Italy and France, the so-called Mediterranean or Southern markets, is at an embryonic stage, as revealed by the much lower penetration rates of 0,17 % (Spain), 0,38 % (Italy) and 0,45 % (France) (101). They are thus rather immature markets (compared to United States and even United Kingdom markets), presenting a particularly significant potential for growth and, therefore, an undeniable attractiveness for entry and/or expansion. Whilst constituting separate national markets these markets display a number of common features that makes it possible to discuss them together.

(258) Language and national preferences are important aspects in the marketing of cruises, and English-speaking cruises are at a distinct disadvantage to Spanish, Italian and French speaking cruises. An important part of the cruises offered can therefore be considered as pan-European/Mediterranean-geared products that take into account the minor, although existing, cultural differences between the various Southern nationalities. This focus explains the substantial presence of operators like Festival, MSC and, to a more limited extent, Royal Olympic Cruises (ROC) on all of these markets. It equally explains the brand recognition and remarkable market position of Carnival's Costa brand, which is by far the strongest player in the three countries, particularly in Italy and Spain.

(259) The main cruise players currently operating on the three markets concerned are as follows: (i) Italy: Carnival (Costa and CCL), MSC, Festival, Royal Caribbean (RCI and Celebrity), and ROC; (ii) France: Carnival (Costa and Cunard), Royal Caribbean (Celebrity and RCI), Festival, Transtours, the tour operator Nouvelles-Frontières, ROC and MSC; (iii) Spain: Carnival (through Costa and CCL), Royal Caribbean (RCI and Celebrity), tour operator Pullmantour, Festival (on its own and through tour operator Spanish Cruise Lines, joint venture), MSC and ROC.

(260) Table 10 shows the configuration of the Mediterranean markets in terms of position (shares) of the major players operating therein.

Table 10 - Mediterranean market shares

<emplacement tableau>

Source:information and data submitted in the course of the Commission's investigation by a very large majority of market participants, on the basis of their own sales in 2001.

(261) In this context (on an all-cruises basis), the planned operation technically gives rise to affected markets in Spain, Italy and France. Carnival's position in all three countries is already significant irrespective of the planned operation, particularly in Italy and Spain. It currently has [55 to 65 %] *, [25 to 35 %] * and [25 to 35 %] * of the Italian, French and Spanish markets, respectively (for 2001 by PCD) (102). POPC only contributes a very limited increase of market shares that can actually be considered negligible. Namely less than 1 % on the Italian, Spanish and French markets (based on the Commission's investigation, for the year 2001).

C.4.2. Potential competition in Italy, France and Spain

(262) In the course of the Commission's preliminary market investigation, various companies argued that, irrespective of the non-significant horizontal overlap already mentioned, the planned operation could lessen potential competition in the markets for all cruises in Spain, Italy and France.

(263) Reference was made in this context to the so-called "Mediterranean Joint Venture" (the JV), created by POPC and RCCL as a parallel operation within the broader framework of the POPC/Royal Caribbean merger. The JV would be the vehicle for an entry on a massive scale into the southern markets. The dissolution or any other form of change of control of the JV would trigger a "put option" mechanism entailing a dramatic economic penalty on the exiting party except if some operational criteria are not met by a certain date. Carnival emphasised in the course of the Commission's preliminary proceedings that it views the dissolution of this JV as a precondition for its bid, and believes that it is in the power of POPC management to dissolve the JV without significant cost. The fact that the United Kingdom Takeover Panel (the City of London's takeover authority) granted Carnival a highly exceptional timetable to ensure that its deal need not close until after January 2003 (when the JV could be undone without these penalties) must be taken as a further indication that the dissolution of the JV is also regarded by the Stock Exchange authorities as a possibility and confirms that the JV is in effect a "poison pill" designed to make an alternative bid more difficult. POPC itself has indicated that "the Mediterranean JV was primarily created and structured to provide RCCL with the deal protection that it required as a condition to their entering into the merger agreement with POPC"(103).

(264) Concerns were raised by third Parties that the possibility of entry by POPC was one of the few credible constraints on Carnival's current and future position, and that it would be eliminated at the time when Carnival has planned to increase dedicated capacity, thereby strengthening its position. These third Parties saw Carnival's position as one of dominance.

(265) A closer market investigation has been undertaken in order to obtain conclusive evidence on the indications initially identified by the Commission as to a possible adverse impact in terms of elimination of actual and potential competition. With regard to the expansion in the near future of the English-speaking segment in the southern markets, the results of the second phase inquiry confirm the initial indications with regard to the relevance of factors such as national language, culture and preferences when marketing ocean cruises in Italy, Spain and France. Most customers feel attracted by pan-European/Mediterranean products addressing the various national differences, as already indicated. Accordingly, the best positioned players in the markets concerned are those following this balanced orientation in terms of choice of food, entertainment, timetables, languages of crew (coming from a significant number of countries) and board languages. These considerations lead to the conclusion that the expansion of the oceanic cruises industry in the Mediterranean markets will mainly focus on the pan-European/national segment.

(266) In relation to the existence of feasible alternatives vis-à-vis Costa, the in-depth investigation has called into question the suggestion that POPC is the most credible potential competitor for the Mediterranean markets and has confirmed that competition from established players already operating in Italy, Spain and France is more relevant to the competitive situation than hypothetical entry by POPC.

(267) The criticisms expressed during the preliminary proceedings were based on the assumption that POPC is the most credible potential competitor to Costa. In this respect, it must be noted that:

(i) the joint venture between RCCL and POPC cannot be interpreted as a real signal of POPC's intentions to enter the Mediterranean markets;

(ii) furthermore, POPC's failed attempt to acquire Festival some time ago does not single out the former as a viable potential entrant. Other cruise companies, financial investors and tourism companies, such as tour operators, would be capable of doing so;

(iii) in relation to POPC's alleged expertise in creating a product addressing the various Mediterranean tastes and preferences, it must be underlined that it is non-existent, given that the cruises marketed by this company in Europe are not pan-European/Mediterranean-style cruises, but basically national-style products aimed at satisfying the requests of North European nationals, with no multi-cultural Mediterranean dimension. Furthermore, POPC's German experience is not based on its own contribution, but on the acquisition of Seetours, in Germany.

(268) All these considerations allow to call into question the credibility of the suggestion that POPC is the "best potential competitor". [Statement on POPC's intentions concerning entry] *.

(269) As to capability of other players to penetrate/expand these markets, the presence of operators already established in the area concerned, like Festival (104), MSC, ROC (to a more limited extent) and national players is more relevant to the market situation than a potential entry by POPC. They are well known and have increased their capacity over the last years. Significant branding/marketing efforts are being made and, as indicated above, these operators have the necessary expertise to address the preferences of the Italian, French and Spanish public by offering a Mediterranean product that meets their expectations.

(270) Further expansion by RCCL (which is present in Italy, Spain and France with market shares of [0 to 5 %]*, [5 to 10 %]* and [0 to 5 %]*, respectively, for the year 2001) is a more likely possibility with or without POPC. RCCL's position as number two in the world and the obvious availability of resources to offer the type of product required by the markets concerned testify to this. RCCL's experience with the Splendour of the Seas constitutes a signal of the company's own interest in the southern markets.

C.4.3. Mediterranean Markets Conclusion

(271) Consequently, the Commission concludes that the present transaction will not lead to the creation or strengthening of a dominant position on the markets for oceanic cruises in Italy, France and Spain.

C.5. SHIPBUILDING

(272) The results of the Commission's investigation have not sufficiently confirmed the concerns expressed by third Parties as to the effect resulting from the argued buyer power to be exerted by the merging entities vis-à-vis the shipbuilding companies (105).

(273) It must be noted in this respect that, as revealed by the Commission's inquiry, the construction of large cruise vessels is a concentrated business where only a limited number of players are present, know-how is specific, expertise from different industries plays a significant role and availability of a broad network of subcontractors is essential. These factors would constitute protection with regard to the exertion of buyer power by cruise operators. While a number of the relevant shipbuilders whose main business area is the construction of cruise ships are actually able to produce other ship types such as container ships, ferries, submarines, naval vessels, gas carriers and offshore vessels (106), they have not been able to attract a sufficient number of orders for those ships (where applicable) due to trade conditions in these segments. There is consequently mutual dependency between the cruise operators and the cruise ship builders.

(274) The merger may enable the combined entity to negotiate more favourable conditions for supplies of various kinds, including the supply of ships. However, the Commission has received no evidence that competing cruise companies would be prevented from having access to the shipyards or from receiving deliveries according to agreed timetables. This has not happened in the past and there is no evidence that it will happen in the future. Therefore the investigation has not shown that the merger could lead to any foreclosure effects for other cruise ship buyers.

(275) It must be noted that demand of cruise ships will basically depend on two factors: demand trends and levels of competition in the cruise market. Concerning the first factor, a significant potential for growth has been identified, as already indicated in recital 16. This growth in demand will induce demand for additional tonnage, including new ships and the refurbishment of old ships in order to improve the services offered to customers. As regards the second factor, this decision finds that no dominant positions will be created or strengthened by this proposed operation in the national oceanic cruise markets in the EEA.

VII. SUMMARY

(276) It must accordingly be concluded that the proposed concentration will not lead to the creation or strengthening of a dominant position in any market as a result of which effective competition would be significantly impeded in the common market or a substantial part of it. The operation is therefore to be declared compatible with the common market and the functioning of the EEA Agreement, pursuant to Article 8 (2) of the Merger Regulation and Article 57 of the EEA Agreement,

HAS ADOPTED THIS DECISION:

Article 1

The notified operation whereby Carnival acquires 100 % of POPC is declared compatible with the common market and the EEA Agreement.

Article 2

This Decision is addressed to: Carnival Corporation Carnival Place

3655 NW 87 Avenue

Miami, Florida 33178 USA.

(1) OJ L 395, 30.12.1989, p. 1; corrigendum OJ L 257, 21.9.1990, p. 13.

(2) OJ L 180, 9.7.1997, p. 1.

(3) OJ C 233, 30.9.2003.

(4) OJ C 233, 30.9.2003.

(5) The term "the Parties" is used for drafting convenience: it does not support that the Parties dealt jointly with the Commission.

(6) Turnover calculated in accordance with Article 5(1) of the Merger Regulation and the Commission notice on the calculation of turnover (OJ C 66, 2.3.1998, p. 25). To the extent that figures include turnover for the period before 1 January 1999, they are calculated on the basis of average ECU exchange rates and translated into EUR on a one-for-one basis.

(7) Parts of this text have been edited to ensure that confidential information is not disclosed; those parts are enclosed in square brackets and marked with an asterisk.

(8) As far as POPC's German subsidiary is concerned it also related to the provision of river cruises.

(9) Figure refers to number of passengers cruising per year in relation to total population. (G.P. Wild: Fleet Changes, p. 56).

(10) OJ C 204, 28.7.1993, recital 19.5.

(11) [Reference to POPC's submission]*.

(12) The Commission has, however, been unable to gather data that would enable it to perform any quantitative tests on the boundaries of the relevant market. When such data is available empirical tests such as the SSNIP test can be done to try to establish how customers would react to a change in relative prices of the products in question. SSNIP stands for "small but significant non-transitory increase in price".

(13) In line with common industry assumptions the term "cruises" will only be applied to multi-night maritime or river journeys for leisure purposes.

(14) See G.P. Wild (International) Limited: Maritime Tourism Competing in a World Market, September 2000, p. 71.

(15) [Reference to POPC's submission] *.

(16) [Reference to POPC's submission] *.

(17) IRN Research, investigation of cruises in the United Kingdom holiday market, April 2002.

(18) Mintel Consultancy, a report on the United Kingdom cruise sector, 10 April 2002.

(19) Given the small sample sizes involved, IRN subsequently confirmed that the "true" value of this estimate could be between 50 % and 80 %.

(20) Report of a second survey RCCL had commissioned from Conquest Research, submitted 8 July 2002.

(21) The researchers achieved over 1000 interviews with respondents and claim that the sample was highly representative.

(22) This figure relates specifically to the Nationwide Tracking Survey undertaken for Holland America Line, June 2000.

(23) The Mintel Group: All Inclusive Holidays, Leisure Intelligence, August 2001, p. 8.

(24) These categories represent social grade, social grade is the socio-economic classification used by the market research and marketing industries in the United Kingdom, most often in the analysis of spending habits and consumer attitudes. The social grade categories are defined as: A - professional (non-manual); B - middle managers (non-manual); C1 - all other non-manual workers; C2 - all skilled manual workers; D - all semi-skilled and unskilled manual workers and E - on benefit/unemployed.

(25) Cruise Information Service, Passenger Shipping Association, Annual Cruise Review 2000 (United Kingdom and Europe), p. 9.

(26) The NOP Holiday Survey, 1999 by NOP Research Group Limited.

(27) [Reference to the POPC-commissioned study] *.

(28) Passenger Shipping Association/Cruise Information Service: Turning the Tide; May 2002.

(29) NERA Economic Consulting "POPC/RCCL - Analysis of cruise yields and capacity by Destination", 2 April 2002.

(30) Yields in this study have been defined as the (weighted) average of achieved ticketed revenue per passenger bed days across all POPC's cruises departing to a given destination during the particular month.

(31) NERA Economic Consulting "POPC/RCCL - Analysis of cruise yields and capacity by destination", 2 April 2002, page 12.

(32) Margaret E. Guerin-Calvert and Janusz Ordover "Profits and prices in evaluation of capacity expansion", dated 27 June 2002.

(33) The authors obtained revenue and cost data at ship level for 1995 to 2001 for Carnival Cruise Lines. Two profit measures were presented: gross profits (revenues obtained from cruise passengers minus the cost of sales) and gross operating profits which also deducted the costs of operating expenses.

(34) The Ordover paper, section II.

(35) On this point a competitor has provided evidence on service levels (measured by staff numbers) differing considerably on the same ship before and after being converted to a TOC product. Staffing was reduced by 18 %.

(36) This information was also submitted to the United Kingdom Competition Commission and forms part of its report. It is, however, not publicly accessible.

(37) Response of 26 June 2002, the Cruise line and tour operator figures are positively correlated with a coefficient of more than 0,9.

(38) [Reference to POPC's submission] *.

(39) [Reference to POPC's submission] *.

(40) It is of course possible that two products with very different prices are in competition with each other. However, given the similarities between the essential nature of premium cruises and economy cruises, the closer the prices, the more likely it is that there is a direct competitive relationship between the two segments.

(41) Mainly seven- and 14-day cruise products, mixed with 10-, 12- and 16-day cruise options.

(42) Mediterranean/Scandinavian cruises with Caribbean cruises.

(43) This survey also includes prices quoted by POPC.

(44) The survey covered three premium (four-star) products and four tour operator cruise offerings by all three major United Kingdom tour operators. (Festival's European Stars, Costa Victoria, RCI's Splendour of the Seas; Airtours' Carousel and Sunbird Thomson's Topaz and Island Cruise's Island Escape). Rates are those quoted by cruise lines reservations on 13 March 2002. The Commission notes, however, that the most important United Kingdom supplier of premium cruises, P& O Cruises, is not contained in the premium sample as they did not offer comparable cruises.

(45) This sample included one P& O Cruises cruise (on the Oriana), one RCI cruise (on the Explorer of the Seas) and one Airtours cruise (on the Sundream).

(46) Equalised for tipping costs, the difference is 57 %.

(47) See Travel and Tourism Intelligence, World Cruise Market 2001.

(48) DRV: Der Kreuzfahrtenmarkt Deutschland 2001, p. 13.

(49) These figures refer to Germany, in which the riverine cruise market is particularly developed. See Alf Pollak, Gaby Pollak-Lenke, The German Cruise Market, Census 2000, February 2001, p. 18ff.

(50) The website of this company states in an information on their product for German tourists: "Bitte bedenken Sie: Es handelt sich nicht um eine Kreuzfahrt" (Please consider that this is not a cruise), http://www.hurtigruten.de/html/ frame1.htm.26.

(51) See also Table 10, where the overlap in PCDs for Spain is stated as < 1 % in 2001.

(52) See Passenger Shipping Association (PSA): Annual Cruise Market Digest United Kingdom (Europe) 2000, p. 32; G.P. Wild (International) Limited: Maritime Tourism Competing in a World Market, September 2000, p. 26; and estimates given by Carnival, Annex 7.2.3 of form CO.

(53) G.P. Wild (International) Limited: Maritime Tourism Competing in a World Market, September 2000, p. 27.

(54) The naming of the style should not be confused with the itinerary. Such cruises may also have a Caribbean or Northern European itinerary.

(55) Douglas Ward: Berlitz - Complete Guide to Cruising & Cruise Ships 2002, Princeton 2002. The guide's ratings are based on the familiar system of one to five stars complemented by attributions, or not, of a "plus" (+) qualification that indicates a somewhat higher position of a certain product in the eyes of the rating agent. The Berlitz guide arrives at its ratings by allocating a certain number of points to five aspects of each cruise product: the ship (25 %), accommodation (15 %), cuisine (15 %), service (20 %) and cruise experience (25 %). The basic units of these ratings are individual ships.

(56) As to decisions in the hotels sector see Case No IV/M.1133 - Bass plc/Saison Holdings BV; Case No IV/M.1596 - Accor/Colony/Blackstone/Vivendi; Case No COMP/M.2197 - Hilton/Accor/Forte/Travel Services JV. Decisions in consumer goods sectors: Case No COMP/M.2544 - Masterfoods/Royal Canin; Case No COMP/M.2569 - Interbrew/Beck's, Case No IV/M.794 - Coca Cola/Amalgamated Beverages; Case No IV/M. 623 - Kimberly-Clark/Scott.

(57) G.P. Wild: Maritime Tourism, pp. 73 and 57.

(58) Cunard will add to its fleet the "Queen Mary 2", the world's biggest and most expensive ship, in 2003. Another ship, the [...] * will only be added in 2005 and was not considered in this projection. [...]* capacity will further add capacity even at the assumption of a simultaneous retirement of the Caronia, as the [...]* capacity by far surpasses Caronia's capacity. [...] *.

(59) For POPC the following capacity changes were considered: intra-company transfers from Princess: Oceana, ex-Sea Princess; transfer to Ocean Village: Arcadia. For Princess: addition of newbuilds: Coral Princess, Star Princess, Diamond Princess, Island Princess, Sapphire Princess, Crown Princess (II); intro-company transfers to other brands or retirements: Ocean Princess, Pacific Princess, Regal Princess, Crown Princess (I), Sea Princess. For Swan Hellenic: replacement of Minerva (I) by Minerva (II)..33.

(60) The total of the figures goes beyond 100 % because some respondents understood the third group to be overlapping with the first two, others understood the three groups to be complementary.

(61) [...] *.

(62) Conquest Research: Report on telephone interviews conducted among cruise guests in the United Kingdom, 16 May 2002, p. 5f. The report was commissioned by Royal Caribbean. Respondents were asked to say which cruise line they would have chosen instead of the one they actually travelled with last time. Quite a high percentage (49 %) of respondents chose none of the alternative lines, which is consistent with a high level of loyalty to the line selected on the last occasion. The table excludes those who said they would not choose an alternative.

(63) Conquest Research: Report on telephone interviews conducted among cruise guests in the United Kingdom, 16 May 2002, p. 6 ff.

(64) [...] *.

(65) [...] *.

(66) [...] *.

(67) [...] *.

(68) G.P. Wild (International) Limited, Implications of fleet changes for cruise market prospects to 2010, August 2001 and UBS Warburg Global Equity Research, It's high tide you looked at the Cruise Lines, 19 March 2002.

(69) In winter somewhat less as Airtours charters one of its ships only in summer.

(70) As with Airtours one ship is not chartered during the winter season.

(71) [Reference to POPC's submission] *.

(72) Reply to the Commission's questionnaire of 25 June, page 18. Figures derived from Annex 24.1 [...] *.

(73) Carnival's reply to the Commission's questionnaire dated 25 June, pp. 15 and 18.

(74) According to G.P. Wild, the Parties accounted for 52 % of new-build orders at the end of 2000 (Carnival 32 %, POPC: 20 %), G. P. Wild: Maritime Tourism, page 42.

(75) [UBS Warburg Global Equity Research It's high tide you looked at the Cruise Lines, 19 March 2002] *.

(76) Hurtigruten share is stated as 5,1 %. See recitals 107 and 108. In Table 9 Hurtigruten's shares are not included, in line with the proposed market definition.

(77) DRV: Der Kreuzfahrtenmarkt Deutschland 2001, Eine Studie von Alf Pollak und Gaby Pollak-Lenke, im Auftrag des Deutschen Reisebüro und Reiseveranstalterverbandes e.V. The DRV figures include Norwegian Coastal Voyages/Hurtigruten, for which an allowance of 7 % was made.

(78) This is minus the estimated share for Hurtigruten which is excluded from the Commission's market definition but included in the DRV figures. The (passenger) market share for Norwegian Coastal Voyages/Hurtigruten was estimated to be 5 % by Carnival in its estimate on 2000. These 5 % were deducted from the DRV market volume.

(79) Crystal, Radisson Seven Seas, Silver Sea, Sun Bay and Compagnie des Iles de Ponant.

(80) African SafariiClub, Quark, ADAC, Club Valtur, American Classic Voyages.

(81) This may be due both to these companies' no-discounting policy and to the fact that Hapag Lloyd and Peter Deilmann have a number of ships that are either luxury or upper premium and therefore are also higher value.

(82) Carnival, in its submission "An Assessment of the EC Commission's Article 6 (1) (c) decision" of 14 May 2002 comments on this demand forecast and underlines that the figure of 8 % only reflects a low growth forecast, described by G.P. Wild as "conservative, even pessimistic". Carnival's submission adds G.P. Wild's medium and high growth forecasts, which, if applied to Germany (as is presented as a realistic possibility in Carnival's submission), arrive at growth figures of around 25 % yearly for German demand (i.e. total German demand increases from 2001 to 2004 of 100 % and 156 % respectively).

(83) [...] *.

(84) This is notwithstanding the arguments in recital 238.

(85) See Table 9 "German nationals": [30 to 40 %] * plus POPC Aida and Arkona: [10 to 20 %] *.

(86) One could expect this growth to materialise to a larger extent with "American" destinations such as Alaska, Hawaii, Mexican Riviera, Caribbean, Barbados and New England.

(87) In DRV: der Kreuzfahrtenmarkt Deutschland 2001, "upper premium" is termed "premium" whereas "lower premium" is termed "standard". The understanding of "premium" in this decision therefore covers the DRV categories "premium" and "standard". This survey gives the following price per diem ranges (in DEM, EUR in brackets): Budget: DEM 150 to 249 (77 to 127); Standard DEM 250 to 349 (128 to 178); Premium DEM 350 to 499 (179 to 255); de Luxe above DEM 500 (256). Ibid. p. 13.

(88) By contrast, economy cruises can attract a large number of mainly first-time cruisers but will only have small margins and luxury cruises have high margins but can appeal only to a small group of customers.

(89) DRV: Der Kreuzfahrtenmarkt Deutschland 2001, p. 13.

(90) Ibid. p. 13. The boundaries between luxury and upper premium may sometimes be blurred.

(91) For Cunard, however, only Queen Elizabeth II's Queen's Grill and suites.

(92) DRV: 68 % composed of 23 % upper premium (Premium) and 45 % lower premium (Standard). Ibid. p. 13.

(93) The ship is now chartered by Transocean Tours.

(94) Not infrequently they operate even above as "upper berths", i.e. the third or fourth beds per cabin, are not taken into account for measuring 100 % capacity utilisation.

(95) PCD growth rates are slightly lower as the average length of cruises is sinking slowly.

(96) As RCCL's market share in Germany remains below 5 % and as RCCL has no German-style brand, nor even a European-style brand, Carnival's and RCCL's positions in Germany are not comparable.

(97) The Commission has no information going beyond that date. See DRV: Der Kreuzfahrtenmarkt Deutschland 1994, im Auftrag des Deutschen Reisebüroverband durchgeführt von Alf Pollak., p. 6. CCL, Princess, HAL (market shares in 2001 < 1 %) were represented by a general sales agent.

(98) The charter of another ship ended in 1998.

(99) Carnival's reply to the Commission's questionnaire dated 25 June, p. 22.

(100) The German association of tour operators and travel agents counts 4270 members and states that they would account for three-quarters of the turnover of this sector. (DRV: Fakten und Zahlen zum deutschen Reisemarkt, März 2002). Only six tour operators account for more than 80 % of the tour operator market. A cruise company states, "The German travel agents market consists of about 22000 travel agencies, most of them belonging to cooperations. Only one third of those travel agencies are so-called free agencies which are choosing the products on their own".

(101) Data provided, for the year 2000, by the United Kingdom Consumers' Association.

(102) There is some discrepancy of these market shares with those estimated by Carnival (see Table 6). Some of the discrepancy may be due to the absence of direct comparability of passenger and PCD market shares. Some of it may be due to changes between 2000 and 2001. E.g. Carnival's lower market share in Spain in 2001, may be due to increased competition by national economy operators. On both assumptions of market shares for Carnival the main argument concerning POPC's position remains unchanged.

(103) [Reference to POPC's submission] *.

(104) According to press information, Festival has completed negotiations to secure a relevant package of extra funding (private placement of share capital and bond financing) in order to consolidate its financial strength with a view to widen the company's shareholding base and to expand its fleet. Furthermore, Festival's chairman and CEO has stated that record bookings in the first months of this year will allow to consider 2002 as Festival's "best-ever year".

(105) The following shipyards produce large cruise ships: Chantiers de l'Atlantique (France), Meyer Werft (Germany), Fincantieri (Italy), Kvaerner-Masa (Finland) and Aker MTW Werft (Germany). Lloyd-Werft (Germany) does not produce whole ships but only completes bought-in hulls, as does Mariotti (Italy).

(106) Chantiers de l'Atlantique and Fincantieri produce other types of ships such as naval vessels, ferries, gas tankers and submarines, respectively. Kvaerner Masa-Yards produces gas tankers, Meyerwerft produces gas tankers and ferries and Aker MTW Werft produces, among other things, containerships.