EC, January 7, 2004, No M.2978

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), and in particular Article 8 (2) thereof, Having regard to the Commission's decision of 5 June 2003 to initiate proceedings in this case, Having regard to the Commission's decision of 23 July 2003 not to refer the notified transaction to the competent French authorities, 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 (2), Having regard to the final report of the Hearing Officer in this case (3), Whereas:

(1) On 14 April 2003, the Commission was notified in accordance with Article 4 of Council Regulation (EEC) No 4064-89 ("the Regulation") of a planned transaction whereby the undertaking Lagardère would acquire control, within the meaning of Article 3 (1) (b) of the Regulation, of the assets of Vivendi Universal Publishing ("VUP"), controlled by Investima 10, itself controlled by Natexis Banques Populaires, by means of share purchases ("the notification").

SECTION I: THE PARTIES

(2) Lagardère is a limited partnership with shares quoted on the French stock market. The parent company of the Lagardère group is Lagardère SCA. Lagardère SCA is a company quoted on the Paris Stock Exchange. It is not controlled directly or indirectly by any other firm or institution.

(3) Lagardère controls a group of companies in business world-wide in three main areas: (i) communication/media/publishing, (ii) the motor industry and (iii) high technology. On the communication, media and publishing side, alone concerned by the notification, Lagardère's main activities are grouped in Hachette SA ("Hachette"), wholly owned by Lagardère, doing business in the following areas:

- publishing, through Hachette Livre;

- the press, through Hachette Filipacchi Médias;

- distribution and retailing services, through Hachette Distribution Services;

- broadcasting and multimedia services, through Lagardère Active.

(4) Investima 10, a public limited liability company governed by French law, (4) is a wholly-owned subsidiary of Ecrinvest, itself wholly owned by Segex (Groupe Natexis Banques Populaires). Investima 10 was formed to hold the shares in VUP, which is active in various sectors of the creative publishing business and operates in logistics and distribution.

SECTION II: THE TRANSACTION

(5) In September 2002 Lagardère declared its interest in acquiring VUP's publishing assets in Europe and Latin America (except Brazil), which Vivendi Universal had just decided to put up for sale (5). At the end of October, Vivendi Universal announced that it had accepted Lagardère's offer.

(6) Lagardère's procedure for acquiring these assets had to meet one of the seller's desiderata, which was for a quick operation and quick payment. It was in order to meet this concern for speed that, at Lagardère's request, Natexis Banques Populaires became involved in the process of acquiring the relevant VUP assets.

(7) On 3 December 2002, the Natexis Banques Populaires Group concluded a firm sale contract with Lagardère, enabling Lagardère (via Ecrinvest 4), subject to authorisation from the Commission, to acquire full ownership of the capital in Investima 10, the company owning VUP's assets. The acquisition price for these shares was paid immediately, and in advance, by Lagardère to Segex, a company owning all the shares constituting the capital in Ecrinvest 4, on the same date.

(8) Consequently, the transaction is an acquisition of total control within the meaning of Article 3 (1) (b) of the Regulation.

SECTION III: COMMUNITY DIMENSION

(9) The firms concerned had total worldwide turnover of euro 5 billion (6) in 2001 (Lagardère, [...]* million; VUP, [...]* million). Each of them had turnover in the Community of more than euro 250 million in 2001 (Lagardère, [...]* million; VUP [...]* million), and Lagardère does not achieve more than two thirds of its turnover in a single Member State. The transaction therefore has a Community dimension.

SECTION IV: PROCEDURE

(10) After examining the notification, the Commission has concluded that the transaction is caught by the Regulation and raises serious doubts as to its compatibility with the Common market. By decision of 5 June 2003, it accordingly began proceedings under Article (6) (1) (c) of the Regulation.

I. THE FRENCH REQUEST FOR REFERRAL

(11) On 14 May 2003, the French authorities asked for part of the case to be referred to them under Article 9 of the Regulation. The request concerned the analysis of the effects of the transaction on the markets for publishing general literature titles in large format (French grand format), general literature titles in pocket format (format de poche), the acquisition of copyright for editions in pocket format, school textbooks, educational supporting materials, single-language dictionaries, bilingual and multilingual dictionaries, single-volume general encyclopaedias, and marketing and distribution services provided to publishers. The French authorities argued that each of these markets had a national dimension.

(12) By decision of 23 July 2003, the Commission rejected the French authorities' request.

(13) For the markets in copyright for pocket-format editions, marketing and distribution services, large-format general literature titles, pocket-format general literature titles, single-language dictionaries, bilingual and multilingual dictionaries and single-volume general encyclopaedias, the Commission considered that the condition that there should be a distinct geographic market was not satisfied, and that these markets could therefore not be referred to the French authorities under the second subparagraph of Article 9 (3) of the Regulation.

(14) For the market for sales of school textbooks by publishers to dealers, the Commission considered that all the conditions imposed by Article 9 (2) (a) of the Regulation were satisfied. For the market for sales of educational supporting materials by publishers to dealers, the Commission, at the time when it took its decision under Article 9 of the Regulation, was not able to determine whether the geographic dimension was national or supranational, though the other conditions imposed by Article 9 (2) (a) of the Regulation were satisfied. For these two markets, even if the market for educational supporting materials had a national geographic dimension, the Commission decided to analyse the effects of the merger itself, in accordance with point (a) in the first subparagraph of Article 9 (3) of the Regulation, in view of the close links between these two markets and the remainder of the book business.

II. THE SUSPENSION OF TIME-LIMITS

(15) To enable it to examine the notified transaction the Commission sent Lagardère requests for information under Article 11 of the Regulation, in particular on 12 May,

16 May and 2 July 2003. The time-limits for answering these requests were 14 May, 26 May and (after extension) 25 July 2003 respectively. But Lagardère did not provide all the information requested within the time allowed.

(16) The Commission accordingly adopted two decisions addressed to Lagardère, on 16 June and 8 August 2003, in accordance with Article 11 (5) of the Regulation. Under Article 10 (4) of the Regulation, the time allowed for adopting a decision in accordance with Article 8 of the Regulation was twice suspended from the date of the time-limit set by the request for information to the date when the information requested by the Commission was received, that is to say from 14 May to 11 July 2003 and from 25 July to 26 August 2003.

III. THE STATEMENT OF OBJECTIONS

(17) The Commission sent Lagardère a statement of objections on 27 October 2003. Lagardère replied to this statement of objections in a document sent on 17 November 2003. Lagardère did not ask for a hearing.

IV. THE REMEDIES

(18) On 2 December 2003 Lagardère offered the Commission certain commitments with a view to rendering the transaction compatible with the Common market and the EEA agreement. On 23 December the notifying party submitted a number of amendments regarding the manner of implementation of the commitments.

SECTION V: THE MARKETS

I. PUBLISHING

A. DESCRIPTION

(19) Books follow a circuit from author to reader, the "book chain", with different people involved at different stages - the publisher, the distributor, the marketer, the wholesaler and the dealer. They interact broadly in the manner described in the flow-chart below.

<emplacement tableau>

(20) Upstream, the French-language publisher approaches authors or non-French-language publishers, either direct or via literary agents representing them, who assign rights in their works to the publisher and thus supply the content for the production of a book, which in the majority of works, and in particular in the case of general literature titles, is a narrative. For certain categories of books, such as reference works, school textbooks and guides and manuals, the content is often generated by a team of writers and illustrators, illustrations being acquired in different ways; for these books, the French-language publisher does not generally acquire rights but contributions. This appeal by publishers to authors usually takes the form of orders, which account for the bulk of creative publishing.

(21) The publisher designs and manufactures books and then sells them, generally via a marketer, to dealers on the basis of an agreed discount. The publisher determines the retail price (7).

(22) Margins are mainly taken in the form of discounts calculated on the basis of the list price exclusive of tax (French PPHT) which the publisher or marketer will negotiate with the various dealers. The discounts are based on quantitative criteria linked to the number of titles stocked and qualitative criteria linked to the nature of the services rendered to facilitate book sales. This marketing system is similar in some ways to large-scale distribution, where the real business negotiations are on "back margins" (8).

(23) Marketers and distributors respectively then market and distribute the published book to dealers - booksellers, hypermarkets and wholesalers. The publishers remunerate them for these services on the basis of a percentage of the list price exclusive of taxes they receive, net of returns, that is to say net of copies that dealers have been unable to sell and are entitled to return to the publisher at their own expense and on certain conditions.

(24) Marketers (diffuseurs) approach dealers to promote the books (except for school textbooks) in the catalogue of the publishers for whom they act, who may or may not belong to the same group as the marketer. This service is performed by commercial teams consisting of representatives who present new issues, draw attention to existing titles and engage in other promotion operations. The marketer takes orders from dealers and passes them on to distributors.

(25) Distributors (distributeurs) handle all logistical operations involved in getting books to the customer. This includes stocking titles, registering and checking orders, preparing and sending orders, managing returns, issuing delivery orders and invoices, managing client accounts and recovering amounts due. The distributor acts on behalf of publishers who may or may not belong to the same group as the distributor.

(26) The wholesaler exercises a resale and marketing function (marketing/distribution) which mainly consists of selling full assortments of books and/or full management of book departments ("rack-jobbing", especially in supermarkets) with smaller retailers not specialising in selling books. He takes his supplies direct from the publishers and generally has a sales depot or salesroom where retailers can come and obtain books if they wish to order small quantities or at short notice. Wholesalers are remunerated on the basis of discounts calculated on the list price exclusive of tax.

(27) Retailers are classified in different categories depending on their degree of specialisation in book sales, their turnover, the number of titles they keep in stock and their sales area. They may be class 1, 2 or 3 retailers, hypermarkets (hypermarchés) or supermarkets (supermarchés). Class 1 and 2 retailers are larger bookshops and smaller local bookshops respectively. These retailers are supplied primarily by publishers' vertically integrated marketing/distribution units, whereas class 3 retailers, most of which are small retail outlets and supermarkets, are supplied by wholesalers. Hypermarkets tend to operate via a central purchasing facility for their supplies of book titles. Retailers are remunerated on the basis of a discount on the list price exclusive of tax and sell their books to final consumers.

(28) Book clubs, given certain specific features as to publication date, for instance, tend to be involved in re-issues and are in any case a separate, independent sales circuit (9).

B. TRENDS

(29) Consumption in the book trade has been rising since 2000, generating publishers' sales of euro 2.6 billion on 354 million copies sold in France in 2001 (10). Independent bookshops have seen their business withering, growth being centred above all on hypermarkets and big specialists (Virgin, Extrapole and FNAC). In Belgium, sales of French-language books in 2001 totalled euro 210 million (11). There is a high degree of vertical integration in the trade and a relatively high degree of concentration. The six major French publishers - VUP, Hachette Livre, Gallimard, Flammarion, Albin Michel and Seuil - control more than two thirds of the market measured by sales (12), VUP and Hachette Livre being the unchallenged market leaders.

(30) The four most significant segments of the book market - general literature titles including books in pocket format, children's books, guides and manuals and educational books - account on their own for more than half of all sales and for the highest growth rates since 1994 (13).

C. PLAYERS (14)

C.1. THE PARTIES TO THE TRANSACTION

C.1.a. THE LAGARDÈRE GROUP

(31) The Lagardère group is active in the press industry through its subsidiary Hachette Filipacchi Médias. This is the world's largest magazine publisher, with more than 200 titles totalling more than a billion copies and turnover of euro 2.4 billion, more than half of it on the international market. Hachette Filipacchi Médias owns well-known magazine titles in virtually all sectors of the business - current affairs, television, leisure, travel and the youth market. In France it also publishes dailies. A single company, Interdeco, handles international advertising for 200 Hachette Filipacchi Médias magazines.

(32) On the broadcasting and multimedia side, Lagardère, through its subsidiary Lagardère Active, holds some extremely well-known brands, and ranks in the top category of producers and distributors in France and abroad, not least in the new media.

(33) In the publishing business, the Lagardère group operates via its subsidiary Hachette Livre, which had turnover in 2002 of euro 846 million, making it France's second largest publisher by sales. It also has a strong presence in the UK and Spain, with a total of more than 40 publishing houses.

(34) On the market for sales to dealers, Hachette Livre sells general literature titles, children's books, guides and manuals, art books, school textbooks and educational supporting materials, reference works and academic and professional books. It owns major imprints in all sectors, and well-known series. It also publishes books in pocket format.

(35) On the marketing services market, the Lagardère group owns two marketing companies, Hachette Livre in France and Dilibel in Belgium. On the distribution services market, Hachette Distribution Services and Dilibel have their own logistical facilities.

(36) On the retail sales market, Hachette Livre is active in France both in the mail order and direct selling business and in the retail trade through the bookshop chains Le Furet du Nord, Virgin/Extrapole and Relay. The Lagardère group operates in the book retail business in Belgium through Press Shop/Relay, and in the mail order trade through Hachette Collections. Hachette Livre does direct selling business in Spain, and mail order business in both Spain and Portugal.

C.1.b. VUP

(37) VUP is active in several branches of creative publishing and exercises logistical and distribution functions. In 2002, VUP had a consolidated turnover of euro 968 million, making it the publishing market leader.

(38) VUP is in business in all areas of creative publishing and owns recognised imprints and series in general literature, children's books, guides and manuals, art books, school textbooks and educational supporting materials, reference works and academic and professional books. And VUP is active in pocket-format publishing.

(39) VUP is vertically integrated in publishers' marketing services through Vivendi Universal Publishing Services ("VUPS") and VUPS Benelux. VUP and VUPS have their own logistical facilities.

C.2. OTHER FIRMS

C.2.a. THE MAIN GENERALISTS

(40) The main generalists are publishing houses whose capital commonly remains family-owned. Gallimard, Flammarion, Albin Michel and Seuil are the most important.

(1) GALLIMARD

(41) Gallimard is controlled primarily by the Gallimard family. In the publishing sector, the Gallimard Group operates chiefly in the markets for publishing rights, sales to dealers, marketing, distribution and retail sales. In 2002 Gallimard's consolidated turnover was euro 235 million. With a turnover more than seven times lower than the merged entity in the same year, Gallimard is still the largest competing publisher.

(42) In the market for sales to dealers, Gallimard is active in large-format and pocket-format general literature, guides and manuals, children's books, educational supporting materials, reference works and art books.

(43) In the market for marketing services, Gallimard has two marketing companies that market its own titles and those of other publishing houses. One - the Centre de Diffusion de l'Edition ("CDE") - is specialised in marketing targeted on class 1 customers, and the second - France Export Diffusion - is specialised in marketing targeted on hypermarkets and class 2 outlets.

(44) On the market for distribution services, Gallimard has its own integrated logistical facility, Sodis, which offers its services to houses in the Gallimard Group and others.

(45) On the retail market, Gallimard has six booksellers selling books for a large number of publishing houses, including its own.

(2) FLAMMARION

(46) Flammarion is controlled by the Rizzoli Corriere della Sera group, a multimedia group active in traditional lines of business (press, press advertising, publishing and distribution) and in the new media, which is strongly placed in Italy and on the international market (generally in association with local partners, such as the German publisher Burda, and on the press side International Herald Tribune and the Pearson group). In 2000, the Rizzoli Corriere della Sera group's turnover was euro 1 714 million.

(47) In the publishing business, the Flammarion Group's is active mainly on the markets in publishing rights, sales to dealers, marketing, distribution and retailing. In 2002 Flammarion had a consolidated turnover of euro 216 million.

(48) On the market for sales to dealers, Flammarion deals in large-format and pocket-format general literature, guides and manuals, children's books, school textbooks and educational supporting materials, art books, strip cartoons and books on the social sciences and medicine. Flammarion has its own pocket-format publishing company (J'ai Lu), in which Hachette Livre has a holding of 35.33%.

(49) On the market for marketing services, Flammarion has its own sales force specialising in selling its own books and those of other houses to class 1 and 2 retailers and hypermarkets and wholesalers.

(50) On the market for distribution services, Flammarion has a subsidiary, Union Distribution, which offers its services to houses in the Flammarion group and others.

(51) On the retail sales market, Flammarion has a 23.5% holding in Librairies du Savoir, present in France and Belgium, and operates two bookshops directly in Paris.

(3) ALBIN MICHEL

(52) The Albin Michel group is controlled by the Esménard family. On the publishing side it is active mainly in the markets for publishing rights, sales to dealers, marketing, distribution, school textbooks and book clubs. In 2002 Albin Michel had a consolidated turnover of euro 206 million.

(53) On the market for sales to dealers, Albin Michel deals in large-format general literature, guides and manuals, children's books, school textbooks and educational supporting materials, art books, strip cartoons, encyclopaedias and books on the social sciences. On the pocket-format market, Albin Michel has a 20% holding in the pocket-format publishing subsidiary controlled by Hachette Livre, Librairie Générale Française ("LGF"). (15) Albin Michel's capital holding in LGF, with its "Le Livre de Poche" series, means that LGF is its preferred publisher for pocket-format editions of its large-format titles (16).

(54) Under clause 4 of the memorandum of agreement signed on 14 June 1999 between Albin Michel, Hachette Livre and LGF, [...]* (17).

(55) In the market for marketing services, Albin Michel handles its own marketing for class 1 and has it done by Hachette Livre for all other outlets. Its subsidiaries Magnard Vuibert Editions and Dilisco handle some marketing functions for publishers within the group and others, in particular for school textbooks.

(56) In the market for distribution services, Albin Michel has entrusted Hachette Distribution Services with distribution for all outlets. Albin Michel also distributes its own titles and titles of other publishers, in particular school textbooks, through its subsidiary Dilisco.

(57) In the book clubs market, Albin Michel has a controlling shareholding of 54% in Grand Livre du Mois ("GLM").

(4) SEUIL

(58) In the publishing trade, les Éditions du Seuil is a group primarily active in the markets for publishing rights, sales to dealers and marketing and distribution. In 2002 Seuil had a consolidated turnover of euro 154 million.

(59) In the market for sales to dealers, Seuil is active in large-format and pocket-format general literature, guides and manuals, children's books, educational supporting materials, art books, strip cartoons and academic and social sciences books. A significant portion of Seuil's turnover is achieved on pocket-format books marketed primarily under the Points imprint.

(60) In the market for marketing services, Seuil has its own marketing structure which markets its own titles and those of other publishing houses.

(61) In the market for distribution services, Seuil has its own integrated logistical facility, Distribution Seuil, which handles the primary functions of a distribution firm for houses within the group and others.

C.2.b. OTHER PLAYERS

(62) The other players in the publishing business are smaller French and Belgian publishing houses that generally specialise in a given category, such as the La Martinière group for art books, Bréal, Belin and De Boeck for school textbooks, Economica, Francis Lefebvre for academic books, Dupuis, Glénat and Media Participations (Dargaud) for strip cartoons, Payot/Rivages, Milan, and Les Éditions du Rocher and Actes Sud for general literature titles; they are vertically integrated only seldom, and then only partly (18). While these companies often handle part of their own marketing as regards class 1, they usually subcontract the rest of their marketing and all their distribution functions to subsidiaries of the major publishing houses.

(63) It can be seen, then, that in this structure there are three different categories of players in the market:

- two major groups (Hachette Livre and VUP), capable of seeing to their development fully autonomously since, apart from their core business of publishing, they combine the full range of marketing and distribution activities and also have popular pocket-format series that enable the works they publish to enjoy a second life;

- four medium-sized groups, three of which (Gallimard, Flammarion and Seuil) are vertically integrated (marketing/distribution and pocket-format series) but partly dependent on Hachette Livre and/or VUP for marketing as regards the smaller outlets; the fourth group (Albin Michel) is rather more dependent, as a significant proportion of its books are marketed by Hachette Livre and its pocket-format books are generally published by LGF, an Hachette Livre subsidiary;

- an assortment of small firms, heavily and sometimes totally dependent on the larger publishers for the marketing of their products and the publication of their pocket-format books (19).

II. MARKET DEFINITIONS

A. PRODUCT MARKETS

(64) Analysis of how the market operates reveals different levels of interaction between the supply side and the demand side, as the publishing industry comprises a series of separate markets in which the parties to the concentration act as customers and suppliers in turn.

A.1. MARKETS FOR RIGHTS

(65) It is possible to distinguish several categories of rights, which are vital resources for publishers - on the one hand reproduction rights for images and maps and on the other the publishing rights initially acquired from authors.

A.1.a. THE MARKET FOR REPRODUCTION RIGHTS FOR IMAGES AND MAPS

(66) Books often contain illustrations in the form of drawings, photographic images or maps. The rights to reproduce them are acquired from a vast number of bodies which assign publishers non-exclusive graphic or digital reproduction rights.

(67) Publishers buy these reproduction rights from many different sources, such as museums, authors. societies, geographical institutes or archives in order to use them in the works they publish.

(68) In the notification, the notifying party makes no distinction between these rights according to their nature. However, the precise definition of these product markets can be left open, as none of them is affected by this transaction, no matter how they are defined.

A.1.b. MARKETS FOR PUBLISHING RIGHTS

(69) In the case of an individual work, a book stems from the creative act of a single author who, once he has written his book, holds - simply by dint of his role as creator - exclusive intangible property rights in that work, which are enforceable against all parties. These rights comprise aspects of an intellectual and moral nature and others of an economic nature (20). By virtue of his economic rights, the author may authorise various uses of his work - by means of a publishing contract - and receive payment in return, i.e. royalties.

(70) In the case of a collective work, the true initiator is the publisher, as it is the publisher that defines its characteristics and calls on a number of editors, illustrators or photographers to contribute to its execution. For collective works, authors receive a flat-rate, once-and-for-all payment for all exploitation of the work, which does not really constitute copyright royalties in the technical sense of the term. In the rare cases where the payment is proportional to sales, it is calculated pro rata according to the place which the author's contribution occupies in the work, and is therefore limited. Where collective works are originally published in a foreign language, acquisition of the French-language publishing rights is generally combined with the purchase of manufactured copies that are ready for sale, and the rights are unlimited in time (21). Given the number and diversity of contributions that make up such a work, and the complexity of the layout, it is easier for the French-language publisher to buy books that are ready for sale, his only task being to provide the original publisher with a translation of the texts to be inserted into the layout. (22)

(1) PRIMARY MARKETS FOR FRENCH-LANGUAGE PUBLISHING RIGHTS ("PRIMARY MARKETS FOR RIGHTS")

(71) According to the notifying party, copyright in works originally written in French does not constitute a "market" in the usual competition-law sense, for three main reasons (23).

(72) First, the publisher's choice of author is based mainly on a meeting of minds, mutual trust and a willingness to work together to produce a creative work. Accordingly, the decision to publish new authors writing general literature in French depends above all on the quality of the manuscript, which is assessed according to the judgment and personal taste of the publisher.

(73) Second, an author's preference for a particular publisher is a response above all to the interest that that publisher has shown in the author's literary work and the energy that the publisher is prepared to invest in backing it. So, according to the notifying party, publishing policy and support for the writing process are crucial factors in the author's choice of editor.

(74) Third, the great majority of books published are created on the initiative of publishers. They define a project and commission an author to write the book (24).

(75) The second and third of the notifying party's arguments concern only relations between authors and publishers, and do not therefore apply to the French-language publishing rights for an original work in a foreign language, to the secondary publishing rights for books in pocket format or to marketing through book clubs (25). In these cases the notifying party acknowledges that "rights are covered by a contract not between an author and a publisher, but between two publishers, or between a publisher and a literary agent who basically represents the original publisher" (26). The notifying party therefore implicitly acknowledges the existence of separate relevant markets for the acquisition of the French-language publishing rights for an original work in a foreign language, secondary publishing rights for books in pocket format, and book club marketing rights. It is clear, however, that, even for individual works written originally in French, publishers do compete with each other to buy the publishing rights, so that a separate relevant market does exist even there.

(76) Admittedly, publishing houses do not necessarily compete with each other for first works by unknown authors, as their attitude here tends to be that they are prospecting for new authors whose rights can be acquired cheaply. However, there are authors who have achieved success, at least on the literary scene (27), for whom publishing houses compete head on, in particular by offering advances - sometimes very large amounts (28). In general, authors are first "discovered" by a publishing house that agrees to publish them in the knowledge that it will probably not make money on them immediately. As the notifying party points out, "where manuscripts of first works which are of marginal importance - are sent spontaneously, they are addressed to numerous publishing houses. As the acquisition costs are low, the rights are acquired by the most responsive publishing house" (29). After a certain time - the exact length of which will vary from case to case - some of these authors make a name for themselves in the literary world, without however generating high sales. They then begin to gain public recognition, in particular if they win a literary prize. It is only after they have gained this recognition that publishing houses compete head on, and that authors are offered substantial advances to leave their original publisher. The findings of the investigation carried out by the Commission show that in most cases where authors have changed publisher, they have received a higher advance as part of the deal (30).

(77) Moreover, it appears that the publisher's choice of author is based not solely on the factors advanced by the notifying party (a meeting of minds, mutual trust and a willingness to work together), but also on the potential of that author in terms of literary success and hence future profits for author and publisher. The notifying party acknowledges that forecast profit-and-loss accounts are drawn up for many works in order to estimate expected sales and calculate the amount of royalties to be paid to the author (31), but points out that .in the field of general literature, publishers seek out books which may be successful in bookshops. As success in this field is highly uncertain, there are no set targets for returns (32).

(78) From the authors' point of view, the Commission's investigation among authors and publishers reveals that, apart from mutual trust and a willingness to work together, the criteria applied by an author in choosing a publishing house are the amount of payment (advance and/or proportional payment), the extent of its promotional operations, and the marketing and distribution opportunities at all levels that it offers. According to the findings of the Commission's investigation, both publishers and authors believe that the first two criteria applied by an author in choosing a publishing house are mutual trust (something publishers belonging to Hachette Livre or VUP and other publishers are likely to inspire in equal measure) and the amount of payment (33), even though authors writing originally in French rarely use the services of literary agents (34).

(79) Finally, as regards the role of commissions in creative literature written in French, according to the notifying party's definition, contracts for commissioned work are concluded before the text of the work to be published has been completed (35). The work in question may be a project defined by the publisher and entrusted to an author, or a project submitted by an author and accepted by the publisher, or, lastly, a work expected under the terms of a right of first refusal (36) previously granted by the author to the publisher (37). It follows from this definition that only in the first of these cases would there genuinely be a commission placed by a publisher with an author. In any event, it is hard to see how the case for the existence of a market for publishing rights can be weakened by the fact that a literary creation may arise from a commission: many markets, in the most diverse sectors, operate at the initiative of the customer, in the form of invitations to tender or subcontracting, with the customer sometimes directly involved in actually designing the product (e.g. in automobile equipment or armaments).

(80) Thus there is indeed a market for publishing rights.

(a) Distinction according to the original language of the work

(81) The publishing contract or original publishing contract is the contract by which the author assigns to the publisher the right to manufacture or have manufactured copies of the work for the first time, with the publisher bearing responsibility for publishing, marketing and distributing the work. This contract between author and publisher lays down the scope of the rights transferred and the amount of the author's remuneration, i.e. the royalties that depend on the number of copies sold and any advance.

(82) The notifying party states that, in practice, "most of the time the primary French rights are acquired from authors who are natural persons", whereas in the case of foreign rights, "French publishers have no contact with the authors. The rights in literary works can be negotiated via agents, but the contracts are drawn up between the foreign publishing house and the French publishing house" (38). So when a French-language publisher concludes a contract with an author, it always does so in order to acquire the rights in a work written in French. As regards foreign rights, the original publisher has acquired all the economic rights in the work, and it alone may assign to a third party the rights for a French translation.

(83) Moreover, the notifying party states that, in the case of French rights, "the author assigns to the publisher the exploitation rights for all distribution channels, including the rights for translation, new editions, pocket-format editions, book clubs, etc. and - in recent contracts - digital exploitation" (39). The publisher therefore acquires simultaneously from that author not only the rights for graphic publishing, but all economic rights for the adaptation, reproduction and performance of the work, except the rights for audiovisual adaptation, which must be acquired by separate contract. The rights acquired include a right to publish in large format, a right to publish in pocket format where the first published edition is in large format, a right to sell through book clubs (40) and a right to translation, i.e. the right to assign the work for translation, reproduction and hence publication in any another language, the publisher having acquired publishing rights for all languages. However, in the case of foreign rights, the French-language publisher "has the right to exploitation in a specific language, in this case French, through specific distribution channels, normally the traditional book distribution channels, on specific media - paper, digital media - with special, variable clauses for direct sales, sales to clubs or special editions". The foreign rights acquired are therefore more limited in scope.

(84) French rights are assigned by authors to publishers for the duration of the intellectual property protection (41), whereas foreign rights are acquired for five to ten years only (42).

(85) Once foreign rights are acquired, the French-language publisher has to have the work translated into French and publish it. It may, if it so wishes, reassign certain rights, in particular the right to publish in pocket format (43). Because of the translation costs, the proportional payment for books originally written in a foreign language is on average 2% lower than that for books that are not translations; the advances, however, are markedly higher.

(86) The Commission's investigation reveals that, for foreign rights, competition between publishers most often (and in the case of renowned authors nearly always) takes the form of an auction (44). Financial considerations appear to carry even more weight here than in the case of French rights, as personal relations between the original publisher or literary agent and the French-language publisher are virtually non-existent (45). The notifying party also points out that the French-language publisher chosen for foreign authors "is generally the one that makes the most financially advantageous offer" (46). For example, the average advances paid to acquire French rights are lower than the advances paid to acquire foreign rights by a factor of more than two (47), although foreign rights are acquired for a shorter duration and are narrower in scope. So if the amount of advances paid to French authors were to rise permanently by 5 to 10%, this increase would not be significant enough to cause publishers to stop acquiring French rights and to turn to foreign rights instead.

(87) Finally, publishers acquiring publishing rights may have different publishing strategies and policies for French rights and for foreign rights. For example, Gallimard is an important player in the acquisition of French rights but not very active in acquiring foreign rights. The same applies to XO (48). Conversely, First Editions is an important player in the acquisition of foreign rights but not very active in acquiring French rights. There may also be a certain degree of specialisation within groups. For example, within Hachette Livre, Grasset and Fayard are very much specialised in the acquisition of French rights (49). Specialisation is also found within VUP (La Découverte and Les Presses de la Renaissance) and within Seuil (L'Olivier).

(88) Given the differences in the identity of the person assigning the rights (author or original publisher), the nature, the scope (publishing rights in all languages or publishing rights in French) and the duration of the rights (the term of intellectual property protection or a duration of between five and ten years) and the amount of advances paid, the markets for French-language publishing rights for an original work in French ("French rights") and the markets for French-language publishing rights for an original work in a foreign language ("foreign rights") must be regarded as separate markets.

(b) Distinction according to category of books

(89) The categories of books affected by publishing rights are all individual works. Publishing rights in reality concern only individual works for which publishing houses compete with one another. It rarely happens that rights are acquired for collective works; instead a payment is made for work done in the traditional sense of the term. As the notifying party rightly stresses (50), the true initiator of a collective work is the publisher, as it is the publisher that defines its characteristics and calls on a number of editors, illustrators or photographers to contribute to its execution. Authors receive a flat-rate, once-and-for-all payment for all exploitation of the work, which does not really constitute copyright royalties in the technical sense of the term. In the rare cases where the payment is proportional to sales, it is calculated pro rata according to the place which the author's contribution occupies in the work, and is therefore limited. The following books are essentially collective works (51): reference works, school textbooks (52) and educational supporting materials, guides and manuals (53) and art books. The notifying party argues that children's books are also essentially collective works. However, on the list of the 1 500 best-selling large-format books in 2002 supplied by the notifying party, only 13% of children's books are collective works, compared with 58% of guides and manuals. Most children's books are therefore individual works.

(90) The Commission therefore considers that only general literature titles, children's books, strip cartoon albums and academic and professional works constitute individual works for which publishing rights are acquired (54).

(91) The authors in these different categories are not the same. The authors of strip cartoon albums are mainly illustrators, whereas academic and professional works are generally written by teachers, professors or other practitioners. Although some authors write both general literature titles and children's books, such cases remain marginal.

(92) Moreover, publishing houses generally specialise in a particular category. Dupuis publishes only strip cartoons, whereas Masson and Dalloz specialise in the publication of academic and professional books, Gallimard Jeunesse and Deux Coqs (Hachette) publish no general literature titles and, finally, Fayard and Seuil publish neither strip cartoons nor children's books. Some publishing houses are active in more than one category, but the sectors in question are still closely related in terms of readership. For example, Casterman (Flammarion) publishes strip cartoons and children's books, while Nathan (VUP) publishes educational books and children's books.

(93) Furthermore, the amount of the proportional payment and the advance (if any) differs from one category to another. The average advance given to an author of children's books that are not translations is around euro 6 000, whereas the average advance given to an author of general literature titles that are not translations is over ten times higher (55). The ratio for translated books in these two categories is even higher.

(94) Given the differences outlined in paragraphs 89 to 93, the markets for French-language publishing rights must be subdivided according to the categories of books in question.

(2) SECONDARY MARKETS FOR FRENCH-LANGUAGE PUBLISHING RIGHTS ("SECONDARY RIGHTS MARKETS")

(a) Secondary markets for French-language publishing rights for publication in pocket format ("secondary markets for pocket-format rights")

(i) The notifying party

(95) According to the notifying party, if a book published in large format in French achieves a certain degree of success, the publisher may decide to exploit the pocket-format rights. It may do so itself if it runs a pocket-format series, but it may also decide to assign its rights to other publishers of pocket-format books (56).

(96) In this case, the pocket-format publisher acquires the rights direct from the large-format publisher and not from the author. The two publishers conclude a contract assigning the rights and providing for a payment proportional to sales, which may or may not be accompanied by an advance (57). Pocket-format publishing rights therefore form a special category and hence constitute a market separate from that for primary rights.

(ii) The Commission

(97) The Commission agrees with the notifying party's analysis and therefore considers that the market for pocket-format publishing rights is separate from that of the other markets for rights.

(98) As the notifying party points out, pocket-format publishing rights constitute a second exploitation of an author's literary work. As a rule, the pocket-format publishing rights are assigned by the author at the same time as the primary publishing rights and are not the subject of a separate contract (58). The acquisition of pocket-format rights from authors is therefore not a separate market from the acquisition of primary rights where they are bought at the same time. Once the secondary rights have been acquired, the primary publishing house may then use them itself if it has a pocket-format publishing division or assign them to another publishing house. In this case the royalties are paid by the pocket-format publisher to the primary publishing house and the latter is responsible for paying a share to the author (59). The secondary markets for pocket-format rights are therefore distinguished from the primary markets by the identity of the contracting parties.

(99) If a specific payment is attached to the pocket-format edition, it does not compare with the amount of the advance paid to the author or original publisher under a contract assigning first publishing rights. The advance paid to the primary publisher by the secondary publisher is on average more than 20% lower than that paid by the primary publisher to the author for exploitation in large format (60). Moreover, the proportional payment for a pocket-format edition is around 5% of the list price exclusive of tax and the rate is generally fixed, whereas for the primary publication of a general literature title it is always above 8% (and may in some cases reach 18%), and is generally variable (61).

(100) Moreover, publishing groups generally have subsidiaries specialising in the publication of pocket-format books. Examples are Hachette Livre with its subsidiary LGF, VUP with its subsidiary Univers Poche, and Flammarion with J'ai Lu (62). While publishers of books in pocket format always run publishing houses operating in primary publishing, the opposite is not always true.

(101) In view of the above considerations, the market for the acquisition of pocket-format publishing rights is separate from the other markets for rights.

(102) On the question whether the market for pocket-format publishing rights should be subdivided by the categories of books concerned (general literature titles and children's books), it appears that children's books published in pocket-format series are nearly always first editions rather than a second exploitation of the work (63). The rights acquired are therefore primary and not secondary publishing rights. There is therefore no market for pocket-format publishing rights for children's books (64). Likewise, academic works and strip cartoons are not normally published in pocket format.

(103) It follows that this Decision concerns the market for pocket-format publishing rights for general literature titles only.

(b) Secondary markets for book club publishing rights ("secondary markets for club rights")

(i) The notifying party

(104) According to the notifying party, the publishers of a work in large format may assign exclusive rights to book clubs for a limited period (generally six months), entitling them to print the work or to have it printed, to publish it, and to market it by mail order, distance selling or subscription. In return, the book club pays the large-format publisher club marketing rights with a guaranteed minimum. Contracts of this kind are concluded for a limited duration (65).

(105) In its reply to the statement of objections, the notifying party maintains that the Commission is wrong to consider that the secondary market for publishing rights must be subdivided into a secondary market for club rights and a secondary market for pocket-format rights (66). It argues that pocket-format publishing houses and book clubs are in direct competition for the acquisition of secondary publishing rights, as testified by the exclusivity clauses written into contracts assigning rights.

(ii) The Commission

(106) The Commission, however, considers that book club marketing rights are a special category of rights forming a market that is distinct from both the primary rights markets and the pocket-format rights market.

(107) First, it is clear that, from the author's point of view, having a work marketed through a book club (i.e. to a restricted, predetermined target group consisting of club members) is fundamentally different from having it published in pocket format and hence widely marketed to the general public.

(108) Although club-format publishing rights and pocket-format publishing rights both constitute a second use of an author's literary work, the royalties paid to the publisher of the first edition are not the same: on average they amount to 7 to 8% (67) for club rights compared with 5% for pocket-format rights. Also, royalties are passed on to authors on a different basis and on different terms.

(109) While book clubs guarantee a minimum number of sales (e.g. a guaranteed initial print run), this is not the case with pocket-format editions.

(110) Publishers of books in club format do not always run publishing houses that are active in primary publishing, and Albin Michel is the only "primary" publisher with a stake in a book club. Furthermore, no publisher active in publishing books in pocket format takes part in a book club. The secondary markets for pocket-format rights and club rights are therefore distinguished by the identity of the contracting parties.

(111) Finally, the exclusivity clauses written into contracts by book clubs apply only to sales by subscription, mail order and distance selling, and not to the assignment of pocket-format rights. They are also limited in time to six months (68).

(112) In view of the factors explained in paragraphs 106 to 111, the secondary market for club rights is separate from the secondry market for pocket-format rights.

(113) On the question whether the secondary market for club rights should be subdivided by the categories of books concerned, the great majority of works marketed by book clubs are general literature titles, the only other categories they sell being children's books, guides and manuals, art books and strip cartoon albums.

(114) Although the average level of royalties paid by book clubs to acquire rights from publishers may vary from one category of book to another (69), there seems no need to subdivide the secondary market for club rights into the different categories of books concerned. In any event, this question can remain open, as the present operation neither creates nor strengthens a dominant position, no matter which definition is adopted.

A.2. THE MARKETS FOR MARKETING AND DISTRIBUTION SERVICES PROVIDED TO PUBLISHERS

A.2.a. INTRODUCTION

(115) Publishing differs from many other sectors in that the sale of books is not always integrated into the publishers' own activities, but is often subcontracted to marketers and distributors, who carry out these functions on behalf of others. The Commission's investigation reveals that a publisher must be of a minimum critical size to be able to bear the fixed costs of the commercial representatives who carry out the marketing (70) and of the logistical distribution structures (71).

(116) Because of this barrier only publishers of a certain size have enough turnover to possess their own sales and marketing teams and distribution tools. With the exception of Albin Michel, which has opted to use the services of an external firm for a substantial proportion of its activities (while marketing its books to the biggest bookshops and distributing a large proportion of its own school textbooks), most major publishers are vertically integrated in marketing and distribution.

(117) Publishers with a smaller turnover generally use external marketers and distributors, concluding exclusive contracts with a single service provider which itself markets the books to dealers (72) and takes care of all distribution. In this case, marketing and distribution are services offered to publishers.

(118) Marketers may be remunerated in one of two ways: under a trading contract (contrat de négociant) the marketer/distributor receives an overall discount covering the discount it grants to dealers plus its own remuneration; under a service contract (contrat de prestataire) it receives a percentage commission on sales as payment for its services only.

(119) Analysis of marketing and distribution contracts shows that, by default, the general terms of business applied in selling books to dealers are laid down by the marketer. However, the publisher has a degree of influence over certain factors and is free, at its own expense, to apply one-off, additional discounts or to mount promotion campaigns. It is therefore the supplier that bears most of the business risk in the markets for book sales, although its commercial policy and the way it is implemented are strongly influenced by the marketer.

(120) In its reply to the statement of objections, the notifying party argues that the marketer/distributor is paid on the basis of turnover net of returns and that the scope of the marketer's influence on the publisher's policy is questionable.

(121) The Commission holds to its view that the marketer/distributor acts as a service provider vis-à-vis the publisher, but agrees with the notifying party that it is a very special kind of service provider, which should instead be classified as a partner and shares part of the publisher's business risk. The market investigation also shows very clearly that the marketer has a major influence on the publisher's commercial policy (73).

(1) MARKETING

(122) Marketing (diffusion) consists mainly of the marketing of publishers' books to various categories of book dealers. In practical terms, this means that representatives for a specific category of customers and a specific geographic area visit dealers on behalf of a number of publishers in order to present the new books to be published and explain any promotion campaigns on particular titles or special displays of particular books from the list (i.e. the catalogue of titles already published), etc. To perform this task the representatives are equipped with catalogues, promotional material, media plans and sales figures. As well as promoting particular works they may also take orders from customers. The sales pitch has generally been discussed with the publisher, for example at preparatory meetings where new titles are presented to all representatives.

(123) However, there are notable differences in the way marketing operates depending on the categories of dealers, who may be easy or difficult to reach. Organisation therefore takes the form of a number of commercial teams specialising in different classes of dealers. These classes correspond to a classification of the outlets to be prospected by marketers based on turnover, the number of items listed, the retail space reserved for books, their degree of specialisation and the quality of service and expertise they offer to their own customers. Several markets for marketing services can therefore be distinguished according to the category of dealers. As explained below, these are class 1 and class 2 bookshops, hypermarkets and wholesalers (who in turn market books in class 3 outlets).

(2) DISTRIBUTION

(124) Distribution covers all the logistical operations involved in supplying books to customers, from the taking of orders, through the management of returned copies to the collection of payment. The distributor receives and stocks works as they emerge from the printer, takes and checks orders, issues delivery notes and invoices, prepares and dispatches orders, deals with customer relations (information, disputes, etc), keeps customers' accounts, recovers debts, manages returns, takes books back into stock, renews or pulps returned works and draws up credit notes for customers (74).

(125) The distribution function can be split into two main parts: (75) management of physical flows (a platform for stocking and dispatching works that is made up of warehouses (76)) and management of intangible flows (sales administration and financial management (77)). Within the management of physical flows, the stock function can be separated into short-term and long-term stocks. The volume of stocks is generally very high (one year's stocks on average) because of the difficulties of matching supply and demand. This is compounded by the fact that, before a book is pulped (i.e. destroyed), the author's agreement is needed. However, unlike short-term stocks, long-term stocks are not subject to capacity constraints, and are often largely outsourced by both the merged entity and its main competitors. There are certain features of book distribution which do not occur in other forms of distribution, or are not found to the same degree, such as the extraordinarily high number of product listings (80 000 titles at VUP) and outlets (over 14 000 in France), the recurring phenomenon of new products, which appear at weekly intervals and in very large numbers, the phenomenon of returns (bookshops may return unsold copies on certain conditions), the co-existence of fashionable titles which must be made available quickly across the whole territory, stable titles, and a very large seasonal flow of school textbooks. Managing these physical and intangible flows is a complex operation because of their large number. Moreover, the task of informing dealers and helping them in their management and decision-making also imposes a heavy burden, as most dealers are small outlets, not specialised in books and with little expertise in the trade.

(126) For consignments to the French provinces, Belgium and Luxembourg, many publishers and distributors are served by Prisme (78), the multi-industry platform for centralising parcels (79). Prisme consolidates flows from distributors on a daily basis into a single consignment bound for dealers, thereby reducing transport costs. With [15-25]% of total tonnage in 2002, VUP is Prisme's biggest customer, followed by Hachette with [10-20]% (80).

(3) THE SPECIFIC CASE OF WHOLESALERS

(127) Around twenty years ago there were a number of major wholesalers in France, Belgium and Luxembourg which played a not inconsiderable role in supplying small and medium-sized customers (81). But as publishing houses gradually consolidated into publishing groups, publishers themselves - including Hachette Livre and VUP - developed their own marketing and distribution activities aimed at a wider range of customers. These publishers, which had now also became marketers, imposed a new economic model by offering marketing and distribution services to other publishers. This meant that, instead of selling their products to wholesalers, publishers used marketing and distribution services to sell their books. Wholesalers were unable to compete on the basis of the discounts granted to them by publishers which were now their competitors as well as their suppliers, and saw their customer base and margins decline. Many wholesalers disappeared, the only survivors being confined to a very special type of customer, i.e. small outlets, selling books only as an extra, and supermarkets (82). As a general rule, it is not profitable for each marketer to canvass these customers separately, given the marketing costs, the low turnover expected by the marketer on the basis of its publishers' production, and the barriers to entry analysed below (83). Moreover, only wholesalers are able to offer a comprehensive range to this type of dealer, by bringing together titles aimed at the general reader published by a number of publishers.

(128) As a result there are around twenty wholesalers left in France, Belgium and Luxembourg (84). It emerged from the Commission's investigation that wholesalers were active in the market for sales of books to class 3 outlets (85) and could not be considered as suppliers on the markets for marketing and distribution services. By definition a wholesaler buys books from publishers in order to resell them. The wholesaler is a customer of the publisher, and in theory fully independent from it. In other words, as stressed by the notifying party (86), wholesalers are not providers of services to publishers, but customers of publishers operating on the downstream market on their own account and at their own commercial risk, their remuneration being based on the margin between the discount they receive from publishers and the discount they grant to class 3 bookshops.

(129) It must therefore be considered that there is no market for marketing services for sales to class 3 outlets, but instead a market for marketing services for sales to wholesalers, which then sell books on to class 3 outlets.

(130) The diagram below sums up the position of wholesalers and the various channels explored by marketers/distributors on behalf of their publisher customers.

A.2.b. MARKETING AND DISTRIBUTION SERVICES

(1) DISTINCTION BETWEEN MARKETING AND DISTRIBUTION

(a) The notifying party

(131) The notifying party maintains that marketing services and distribution services constitute two separate markets (87) for the following reasons. First, according to the notifying party, distribution contracts are not necessarily linked to a marketing contract, as marketing/distribution contracts may explicitly include an option allowing the publisher to use its own marketing for a specific set of outlets. Second, the notifying party argues that marketing differs from distribution in that it is an activity much closer to that of publishers and that, consequently, marketing and distribution should not be confused, particularly in terms of the potential entry of competitors. The notifying party also maintains that the cost structures of marketing and distribution activities are very different from one another. Finally, it points out that the commissions paid by publishers for distribution services amount to an average of 8% of turnover, whereas marketing services are charged at an average of 6%.

(132) The notifying party concludes that, although distribution and marketing are related, marketing and distribution services constitute two separate relevant markets.

(b) The Commission

(133) The Commission agrees with the notifying party's conclusion that marketing services and distribution services constitute separate relevant markets.

(134) Although marketing services are generally coupled with distribution services, the latter may be offered on their own. For example, publishers with sufficient turnover to finance their own sales force may delegate distribution of their books while keeping marketing at least partly in-house. Examples include Glénat, Dupuis and Albin Michel, (88) which carry out part of their marketing themselves but subcontract all their distribution to Hachette Distribution Services. Also, on the question of cost structure and pricing, there is a variation in the average payment for distribution (around 8% to 12% of the list price exclusive of tax net of returns) and marketing (around 5% to 7% of the list price exclusive of tax net of returns). Finally, it emerged from the Commission's investigation that certain service providers, such as Sofédis, were active only on the marketing side, while other organisations such as MDS (a subsidiary of Média Participations) acted only as distributors.

(2) DISTINCTION BETWEEN MARKETS FOR MARKETING SERVICES

(a) The notifying party

(135) The notifying party maintains that it is not appropriate to subdivide the market for marketing services into a number of separate segments according to types of dealer, for the following reasons. First, each publisher draws up its own classification of dealers. Second, marketing to small outlets is an activity which is not in itself economically rational, but is more in the nature of an adjunct to the services supplied by the marketer to larger outlets. Furthermore, the extension of marketing activities to cover a larger number of dealers, of whatever type, would not involve significant irreversible costs.

(136) The notifying party also maintains that marketing services should not be distinguished according to the type of works marketed, arguing that general marketers handle books dealing even with very specific fields, such as religious books.

(b) The Commission

(137) As regards the distinction within marketing services according to the type of outlets covered, the Commission feels that segmentation of marketing services according to classes of dealers is appropriate for the reasons set out below.

(138) First, these classes are reflected in the organisational structure of marketing companies (89). The investigation has shown that, for the purposes of organising their teams of representatives, analysing sales, and managing terms of business, marketers generally use a classification into class 1 (large bookshops and specialised superstores such as FNAC (90)), class 2 (local bookshops), hypermarkets, and wholesalers (which in turn serve class 3 - small outlets and supermarkets). These categories are common to the whole publishing industry, which is an indication of the existence of separate markets. The investigation shows that the precise demarcation of each class is specific to each marketer and could vary. This is the case in particular for the borderline between classes 1 and 2, where each publisher has its own classification of outlets (91). However, the aggregate total of classes 1 and 2, covering all customers that are "bookshops", comes to around 3 000 outlets for most marketers. All marketers canvass a similar number of hypermarkets (between 700 and 850) and wholesalers (10 to 15). To be more specific, an examination of how dealers are placed in one or other class throws up a clear distinction in the case of hypermarkets and wholesalers (92), whereas the same bookshop may be classified as class 1 or class 2 depending on the marketer and the nature of its catalogue (in particular whether it markets school textbooks or strip cartoons).

(139) Second, the separation between classes of customer is based on a number of structural features which distinguish very specific types of customer and distinct services (93). As the notifying party points out, "representing literary works to 'class 1' bookshops is a different job from selling children's books to supermarkets" (94). Class 1 comprises the largest bookshops and specialised superstores, such as FNAC or Virgin, which sell not only books but also other cultural items and equipment. Class 2 outlets are typically local bookshops with a smaller stock but a fairly full range (95). Class 1 and 2 bookshops are retailers specialising in bookselling with qualified staff who are able to advise customers, receive representatives and choose titles from the catalogues presented to them. Class 1 bookshops sell all types of books and have a considerable number of items listed. Class 2 bookshops have a smaller but still fairly comprehensive range. The marketer generally offers a high quality of service to class 1 bookshops, in the form of numerous, in-depth visits by representatives who provide sales advice, detailed product information, and selling points. Class 2 bookshops receive a lower level of service for a narrower range of products, with a rapid rotation in certain categories of books, and more limited advice. However, the Commission's investigation has revealed a degree of continuity between the smallest class 1 bookshops and the biggest class 2 bookshops, in terms of sales, number of titles, or surface area given over to books, for example (96). Moreover, the fact that certain bookshops are classified in either class 1 or class 2, depending on the marketer, is evidence of a degree of fluidity between the two categories.

(140) Wholesalers have a markedly different profile, as they buy books to sell them on to retailers. Wholesalers are typically among marketers' biggest customers in turnover terms (97). But they are interested only in products that are likely to suit the customers they work with, which are either small outlets not specialised in books (from small newsagents to petrol stations) or supermarkets, so that what they are looking for is a general range containing a limited number of titles that will allow these outlets' book sections to be managed from beginning to end. The range offered to wholesalers tends to be quite narrow, consisting primarily of guides and manuals, children's books and educational supporting materials, an assortment of best-sellers and pocket-format books. The marketing service provided to wholesalers is therefore specifically geared to the categories of books involved, and to the negotiation of discounts, which have a strong influence on wholesalers' final profits.

(141) Hypermarkets, finally, operate a system of central listing covering a limited number of book titles mainly targeting impulse buyers (98), via a central purchasing facility with which the marketer must negotiate the general level of discounts. Once titles are listed, managers of individual hypermarkets' book departments place their orders, bearing in mind any promotion campaigns that may be running. Not having specialised staff, they offer little in the way of advice to customers. The range offered to hypermarkets usually covers general literature, children's books, guides and manuals, art books and strip cartoons (99). The marketing services supplied to hypermarkets reflect the standardised way in which hypermarkets treat purchases, buying books like any other goods they stock. This entails specific services and negotiations relating to end-of-year discounts, special shelf displays and inclusion in advertising brochures, but much less information of a literary nature. To interest hypermarkets, the marketer must therefore be able to adapt to the special approach of central purchasing facilities. This explains why marketers have created specific structures for dealing with hypermarkets (e.g. FED within the Gallimard group).

(142) Third, contrary to the statements of the notifying party, it is clear that in terms of demand for these services publishers take a different view of marketing to the various classes depending on the category of book they produce, as not all categories of book are marketed in the same way to all classes. Whereas school and academic textbooks are marketed almost exclusively to class 1 outlets, strip cartoon albums are marketed predominantly to hypermarkets (100). Publishers specialising in particular publications consequently have different marketing needs. The importance attached to the various classes will vary according to each publisher's production, thereby reducing the scope for substituting marketing services between classes of outlet. Some publishers may be content with marketing activities targeted on a limited number of bookshops, whereas others need a marketer that covers all outlets, as they produce works with large print runs. For example, whereas a publisher of strip cartoons will place a high value on marketing to hypermarkets, a publisher of school textbooks will be much less concerned. Marketing services are therefore clearly defined by class of outlet and difficult to substitute from the publishers' point of view, as is illustrated by the fact that marketing/distribution contracts always stipulate the types of outlets covered by the services provided (101).

(143) Fourth, there are wide differences in the marketing services provided to the various classes of outlets - and hence in their cost (102). In the case of bookshops, marketing is aimed at book-trade professionals, who also make recommendations to customers. In terms of quality, representatives in marketing teams working at this level must be capable of presenting well-honed commercial and literary arguments concerning the products in the catalogue (103), which demands long and frequent visits. Also, there are many more bookshop outlets than there are hypermarkets and wholesalers. This necessitates a larger number of representatives (104). However, hypermarkets must be visited very frequently in order to provide commercial support and secure special displays. Hachette Livre makes an average of [...]* visits per year per hypermarket, compared with [...]* to class 1 bookshops (105). These differences in the nature of services provided are reflected in differences in costs between classes of dealers. Costs also vary notably between companies. Taking as a baseline a marketing cost per outlet of 100 for bookshops, the cost to Hachette Livre for hypermarkets is [...]* and for wholesalers around [...]* (106). There are also wide differences in terms of turnover generated per outlet. If a bookshop generates an average turnover of 100 for Hachette Livre, a hypermarket generates [...]* and a wholesaler [...]*. In the case of VUP, if a class 1 bookshop generates an average turnover of 100, a hypermarket generates an average turnover of [...]*. For the same reasons, there are substantial variations in margins in respect of the various classes of outlet (107).

(144) Fifth, the different sales channels cannot be substituted for one another from the supply point of view either, as access to the various outlets is not uniform. For one thing, bookshops belong to trade bodies which lay down rules on commercial practice between publishers and bookshops (108). Serving bookshops therefore means being able to apply these commercial rules, particularly as regards automatic distribution and the sale of new titles. Hypermarkets present specific barriers such as the additional expense of product listing, which makes the fixed cost of marketing services slightly higher (by 3-4% on average) (109). Since a marketer cannot market books direct to a hypermarket, it must first present its catalogue to the central purchasing facility. Only if it is then listed will the marketer be authorised to visit hypermarkets to manage restocking and the presentation of items on promotion. Over and above the costs relating to product listing and end-of-year discounts, marketing to hypermarkets requires a strong reputation, which restricts the number of publishers listed by central purchasing facilities (110). In the case of wholesalers the barriers encountered are again related to reputation and catalogue, since the products of interest to wholesalers must match the specific needs of the class 3 outlets they serve.

(145) In its reply to the statement of objections, the notifying party claims that Commission has exaggerated the barriers to marketing to hypermarkets and wholesalers. It argues that an author is listed because of the success of a book rather than the reputation of his publisher. However, the Commission found that listing by hypermarkets and wholesalers was done not on a book-by-book basis, but by marketer/distributor. Its investigation also revealed that the number of suppliers was limited. Finally, hypermarkets lay down particularly strict terms for the listing of suppliers. For example E. Leclerc sets the following conditions: "well-known publisher, supplier with strong financial standing, ability to supply a minimum number of listed items, participation in promotional operations, labelling of parcels in line with our own listing codes" (111).

(146) The Commission accordingly considers that it is appropriate to define separate markets formed by classes of outlet between which conditions of competition differ. To be more specific, given the fluid situation regarding substitutability on the demand side, marketing to class 1 and 2 bookshops can be regarded as forming a single separate market, the remaining markets consisting of marketing to hypermarkets and to wholesalers.

(147) As regards a distinction between marketing services according to the category of books marketed, the Commission agrees with the notifying party that such a distinction is irrelevant. Although certain marketers specialise in a limited number of book categories (112), these differences do not prove the existence of separate markets for marketing services according to book category. This view is confirmed by marketing contracts, which generally make no distinction by book category.

(3) DISTINCTION BETWEEN MARKETS FOR DISTRIBUTION SERVICES

(a) The notifying party

(148) As in the case of marketing services, the notifying party considers that it is not appropriate to subdivide the market for distribution services into several separate markets according to types of outlet or categories of works distributed.

(149) First of all, the notifying party argues that a publisher signs only one distribution contract, relating in general to the services of a single distributor (113). Similarly, on the supply side, no distributor specialises exclusively in a particular type of dealer or category of book. Nor is the nature of a dealer of any consequence for the distributor, which is concerned solely with turnover and hence the financial return on covering that dealer. Finally, the notifying party argues that, for a distributor, there is great scope for substituting one type of dealer for another, as it would have to make little in the way of irreversible investments if it wanted to reach a larger number of dealers.

(b) The Commission

(150) It emerged from the investigation that there were no great differences in the distribution services provided to different types of dealers. As far as physical flows are concerned, there appears to be no fundamental difference between delivering consignments to a wholesaler or to a bookshop. Hypermarkets sometimes demand a specific labelling system or delivery carriage-free, but such extra services can be provided quickly and at no major extra cost. As regards intangible flows relating to invoicing or recovery, the Commission has been unable to identify cases where different computer systems are used for different types of dealer. On the contrary, even though the customer file categorises each customer by class of dealer, no fundamental difference has been found in the accounting treatment of the different categories or in the management of invoices. The different classes of dealer (bookshops, hypermarkets and wholesalers) can therefore be substituted for one another from the supply point of view.

(151) The Commission therefore shares the notifying party's view that the market for distribution services must not be segmented according to classes of outlet or types of works distributed, as the logistical services provided to the various types of dealer are relatively uniform.

(4) DISTINCTION BETWEEN IN-HOUSE AND EXTERNAL SERVICES

(a) The notifying party

(152) The notifying party states that marketing or distribution services carried out in-house and the same services performed on behalf of outside publishers should form part of the same relevant markets, arguing that there is a high degree of substitutability between marketing or distribution on behalf of outsiders and marketing or distribution for the publisher's own account.

(153) The arguments put forward by the notifying party are essentially as follows. First, decisions by publishers to switch from one type of service (e.g. in-house) to another (e.g. external) are evidence of a high degree of substitutability. Second, most marketers and distributors perform - in parallel with in-house services - services on behalf of outsiders which are of a similar technical and economic nature and are treated in the same way as in-house services. Finally, no distinction is drawn between in-house publishers and outsiders in the actual organisation of marketing teams.

(b) The Commission

(154) Marketing and distribution mechanisms are designed mainly for the in-house use of publishing groups. Nevertheless, only the services provided to outsiders constitute a market on which supply is matched to demand.

(155) This view is consistent with the judgment given by the Court of First Instance of the European Communities in Endemol v Commission, in which it held that the "independent production of Dutch-language television programmes was a separate market from the market for in-house productions of the public broadcasters." (114) The Court drew on three arguments: (i) substitution between in-house and external production was only partial; (ii) in-house production was essentially intended for producers' own use; (iii) there was no direct competition between in-house production and production offered to outsiders because of the substantial investment on in-house capacity and the need to obtain a return on it. Applying the first of these tests to the present case, different teams are sometimes given the task of marketing on behalf of outsiders, but a certain degree of substitution is nevertheless possible between in-house marketing and distribution services and services performed on behalf of outsiders; pricing is also generally similar (115). However, the other two criteria are clearly met. Marketing and distribution services are predominantly intended for in-house use, as the market for marketing services on behalf of outsiders accounts for around 20% of all marketing services, and the market for distribution on behalf of outsiders accounts for only around 30% of all distribution services (116). Also, marketing activities and, in particular distribution activities, generate very high fixed costs (117), which, moreover, rise in steps. It is therefore clear that the provision of marketing and distribution services to outsiders is a way of exploiting available capacity rather than a form of direct competition with the capacity assigned to publishing houses of the same group (118).

(156) However, the competitive position on this market for services to outsiders cannot be described in full unless in-house services are taken into account. Given the high fixed costs of a marketing/distribution system (119), such a system will be more profitable if it deals with large volumes. From this point of view in-house sales are of considerable importance, as they allow the marketer/distributor to make cost-effective use of its marketing/distribution tool, and to supply services to outsiders at lower cost, while making itself less vulnerable to the possibility of publishers switching to competing marketers/distributors. So while in-house sales should not be taken into account in calculating shares of the market for marketing/distribution services to outsiders, they must be taken into account in the study of the competitive position of the merged entity (120).

(157) The notifying party considers that in-house sales have no bearing on the competitive position in distribution. The Commission's investigation shows the opposite: they play a major role. For both marketing and distribution, a minimum critical size is required to operate on the market, and the fact is that in-house sales form the basis of the ability to compete on the market for marketing and distribution services to outsiders. In-house sales also have a direct impact on marketing and distribution capacities. The investigation reveals that dealers are sensitive to commercial pressure, as analysed below in the assessment of the transaction's impact on competition on the sales markets. If a marketer has many representatives and has high in-house sales accounting for a large share of dealers' purchases, this will have an impact on its ability to encourage retailers to take certain books and hence on its attractiveness vis- à-vis outside publishers (121). On the distribution side, in-house sales create scope for bigger platforms and/or investment in equipment. Analysis of the costs of various distributors shows that large centres generally have lower average costs, particularly as a result of economies of scale (122). This is another feature likely to attract custom from outside publishers.

(158) On this basis, the Commission considers that the analysis of the markets for distribution and marketing services must cover services sold to outside publishers, but must take account of in-house sales in order to arrive at a full and valid appraisal of the competitive position of the various players in the market.

A.3. MARKETS FOR BOOK SALES

(159) Books are sold at various stages of the book chain. First there is the relationship between the publishers and dealers, based on a price in the form of a discount given to the dealer. Second, there is a quite separate business transaction between the retailer and the final consumer, the reader. Where the dealer is a wholesaler, rather than a retailer, there is an intermediate business relationship between the wholesaler and the smaller dealer. This pattern is similar to the usual model for consumer goods.

A.3.a. THE PUBLISHING MARKET

(1) THE NOTIFYING PARTY

(160) The notifying party puts forward a series of market definitions under which the analysis would have to consider, at one and the same time, markets for book sales by publishers to dealers, by retailers to final consumers and by publishers to final consumers, which it describes as "publishing markets", where supply comes from publishers and demand from final consumers, and which would have to be further subdivided by category of books (123).

(161) It argues that the potential problems flowing from the notified transaction can be considered only in terms of a hypothetical market for sales of books by publishers to the final consumer. This is because under the Lang Act in force in France the publishers determine the retail price (dealers may give a discount of no more than 5%), while the retailer may return unsold books in certain conditions, creating a direct relationship between the publisher and the final consumer, since the publisher bears the sole risk of failure to sell to the final consumer.

(162) In its reply to the statement of objections, the notifying party also claims that recognising the existence of a publishing market would be consistent with the Commission's decision-making practice (124). It argues that in cases involving consumer goods such as pharmaceutical products, foodstuffs and pet products, the Commission analyses relations between producers and the final consumer rather than between producers and dealers, even though on these markets the business risk is borne by the distributor, not the producer. The Commission should therefore apply its standard analysis to the publishing industry, especially as the publisher is exposed to greater risk because of the return mechanism and the price-fixing arrangements.

(2) THE COMMISSION

(163) The notifying party's approach of defining a relevant product market on the basis of potential competitive problems is not methodologically convincing and does not correspond to the reality of the way the market functions.

(164) On the substance, the publisher has no business relationship with the final consumer, except in the very specific and very exceptional case of direct sales (such as mail orders).

(165) The dealer is not an agent for the publisher but an entrepreneur accepting a business risk. In the case of reorders dealers bear the entire business risk inherent in buying books from publishers and in no way act as an intermediary on behalf of the publisher.

(166) Moreover, even in the return mechanism, which applies solely to new issues, the publisher is not alone in bearing the risk of failure to sell to final consumers. Part of the business risk is borne by dealers, who bear the cost of carriage of unsold copies and who cover the costs inherent in advance payments to publishers and above all the opportunity costs engendered by failure to sell (125). The notifying party acknowledges that the publisher bears only part of the cost of returns since it states in its written observations on the decision to open proceedings that "no other business sector has a mechanism comparable to the return scheme, whereby the supplier relieves dealers, in whole or in part, of the business risk involved in sales of its products" (126). The return scheme therefore does not demonstrate the existence of a separate market for sales of books by publishers to the final consumer.

(167) Regarding the fact that the publisher fixes the retail price, it should be pointed out that in France there is legislation allowing producers to fix or influence the retail price of consumer products other than just books. For example, the legislation placing general limits on the possibility for large-scale distributors doing business in predominantly in food products to reduce prices has a decisive influence over the price paid by the final consumer. But this situation has not prompted the competition authorities to speak of a market for sales by producers to the final consumer.

(168) As regards earlier Commission decisions cited by the notifying party, in the first one (127), apart from the fact that the features of the industry are not the same, the Commission left the question of the precise definition of markets open, particularly as regards intermediate markets (128). In the second one (129), the Commission did indeed study the presence of the parties to the transaction in relation to the number of subscribing households, but the study concerned the analysis of the effects of the transaction rather than the definition of the relevant product markets. The notifying party's references to earlier Commission decisions are therefore not relevant.

(169) The decisions cited by the notifying party in its reply to the statement of objections do not define a market between producers and final consumers. Indeed in some of them the Commission even considers the possibility of subdividing the markets for the products in question according to distribution channels (130). Although in these decisions the Commission considers market shares based on the weight of producers in the retail sale of products, this is because no data are generally available to determine the market shares of the various players on the market for sales by producers to dealers, and not because the Commission accepts the existence of such a market. In any event, in the decisions cited by the notifying party, demand is analysed at the level of dealers and final consumers (131). The Commission therefore considers that, contrary to the statements by the notifying party, its previous decision-making practice does not point to the existence of a market for the sale of books by publishers to the final consumer.

(170) As for the multiple definitions of markets put forward by the notifying party, under which the analysis would have to consider at one and the same time markets for sales of books by publishers to the final consumer, markets for sales of books by publishers to dealers and markets for sales of books by dealers to the final consumer, the Commission does not consider that is economically justifiable for a single product (a book) to be the subject of a transaction between a publisher and a dealer, and then between a dealer and a final consumer, while at the same time being the subject of a single transaction between the publisher and the final consumer, bridging the two previous transactions.

(171) In the light of the foregoing, the Commission considers that there is no market for the sale of books by publishers to the final consumer, apart from mail-order markets that are not material to the analysis of this case.

A.3.b. MARKETS FOR THE SALE OF BOOKS BY PUBLISHERS TO DEALERS

(172) A significant feature of the publishing business is that the marketing and distribution of books often involve intermediaries - marketers/distributors - who generally do not own the books they promote or distribute. Marketing and distribution structures are separate entities, most commonly integrated into publishing groups; but some publishers also use independent marketers/distributors.

(173) Marketers/distributors represent different publishers through a common structure consisting of teams of representatives (for marketing) and logistical facilities (for distribution). According to the notifying party (132), representatives hold a catalogue containing titles from in-house and outside publishers. Each publisher can choose its representative or team of representatives, often depending on the catalogues they already manage, which must, according to the notifying party, allow titles to be highlighted through .a subtle combination of complementarity and differentiation.. The marketer/distributor is an intermediary between the outside publisher that it markets and the dealers that it visits. It provides the outside publisher with a commercial platform, in that the publisher has the benefit of its ability to negotiate discounts for the retailer. Business relations are organised on the basis of general terms of business prepared by the marketer/distributor, governing the award of bonus points. But that is not to say that the publisher loses all control over the terms of business offered to dealers. Even if the publisher puts its marketing and distribution in the hands of a third party, it can, according to the notifying party, at any time amend the general terms of business determined by its marketer and therefore the terms on which discounts are given to dealers (133). The Commission's inquiries reveal that different discounts are given for different publishers marketed by the sale entity. Moreover a publisher, in particular when launching a book, can back up its promotion campaign with an extra discount for outlets.

(174) On these markets, consequently, the supply side consists of the publishers and, in certain cases, the wholesalers, and the demand side consists of dealers but not of final consumers.

(1) DISTINCTION BY TYPES OF DEALER

(a) The notifying party

(175) The notifying party argues that are no grounds for defining separate markets for the sale of books to dealers depending on the type of outlet, since there is no clear demarcation line dividing dealers into distinct groups; sales to small dealers are not subject to market considerations but are an additional service rendered to publishers; there is a degree of substitutability on the supply side; and there is intense competition between suppliers to large booksellers, which has repercussions for small outlets (134).

(b) The Commission

(176) The Commission considers that on both the supply and the demand sides a subdivision of markets for the sale of books by publishers to dealers must be made on the basis of the type of dealer. In the same way as for markets for marketing to publishers, there are clear differences between the three types of dealer - bookshops, hypermarkets and wholesalers.

(177) First, as was seen in the analysis of the definition of product markets for marketing services, the classification of dealers by level is standard practice in the publishing business. For the purposes of organising their teams of representatives, analysing sales and managing terms of business, marketers generally use a classification into class 1 (large bookshops), class 2 (local bookshops), hypermarkets, and wholesalers (which in turn serve class 3 - small outlets and supermarkets). Since marketing consists mainly of marketing books to dealers in the different classes for the marketer's own account or for another publisher, the fact that this classification is used for marketing purposes is further evidence that there are different market conditions for the separate segments of sales of books by publishers to dealers, and therefore distinct markets. At the same time, as has been seen, it emerged from the investigation that, while the exact definition of the borderline between classes 1 and 2 may vary widely between publishers/marketers, each of them applies much the same definition of hypermarkets and wholesalers.

(178) Apart from this standard practice in the industry, the separation between classes of customer is based on a number of structural features which distinguish very specific types of customer, regarding matters such as the types of books they buy from publishers and the number of titles they offer to final consumers (135).

(179) Class 1 consists of the largest booksellers, selling all types of books and offering a wide range of titles, and specialised superstores such as FNAC and Virgin, which sell not only books but also other cultural items and equipment. Class 2 outlets are typically local bookshops with a smaller stock but a fairly full range (136). The Commission's investigation has revealed a degree of continuity between the smallest class 1 bookshops and the biggest class 2 bookshops, in terms of sales, number of titles, or surface area given over to books, for example (137). Moreover the fact that certain bookshops are classified in either class 1 or class 2, depending on the publisher/marketer, is evidence of a degree of fluidity between the two categories.

(180) Hypermarkets operate a system of central listing covering a limited number of book titles mainly targeting impulse buyers (138), via a central purchasing facility with which the marketer, acting for the publishers which it represents, must negotiate the general level of discounts. Once titles are listed, managers of individual hypermarkets' book departments place their orders, bearing in mind any promotion campaigns that may be running. Not having specialised staff, they offer little in the way of advice to customers. The range of choice offered to hypermarkets usually covers general literature, children's books, guides and manuals, art books and strip cartoons. The products on offer may be extended temporarily (mainly at the beginning of the school year and in the pre-Christmas period) to include educational supporting materials and reference works, but generally consists of an assortment of works in pocket format.

(181) Wholesalers have a markedly different profile, as they buy books to sell them on to other retailers. Wholesalers are among the publishers' biggest customers in turnover terms. But they are interested only in products that are likely to suit the customers they work with, which are either small outlets not specialised in books (from small newsagents to petrol stations) or supermarkets, so that what they are looking for is a general range containing a limited number of titles that allows these outlets' book sections to be managed from beginning to end. The range offered to wholesalers tends to be quite narrow, consisting primarily of guides and manuals, children's books and educational supporting materials, and an assortment of best-sellers and pocket-format books.

(182) The terms on which publishers sell to these various profiles of bookshops, hypermarkets and wholesalers are very different.

(183) First, there are major differences in terms of the turnover generated by the dealer. If a bookshop generates an average turnover of 100 for Hachette Livre, for instance, a hypermarket generates [...]* and a wholesaler [...]* (139).

(184) Second, terms of business and discounts differ according to the class of dealers. Hachette Livre, VUP and all their competing marketers apply different discounts to the different classes of dealers in their general terms of business, setting different levels of quantity discounts and different categories and levels of quality discounts depending on the class of dealer. (140)

(185) Lastly, there are specific barriers to access to hypermarkets and wholesalers, as distinguished from bookshops. To be listed by a hypermarket's central purchasing facility, a publisher (or its marketer) needs to have a strong reputation as only a limited number of publishers will be listed (141). In the case of wholesalers, the barriers encountered are again related to reputation and catalogue since the products of interest to wholesalers must match the specific needs of the class 3 outlets they serve.

(186) It follows that the markets for the sale of books to dealers must be subdivided according to the type of dealer. A distinction must be made between the markets for the sale of books by publishers to bookshops (sales to class 1 and 2 bookshops being one and the same market on account of their common features and the fluidity of the distinction between them), hypermarkets and wholesalers.

(187) But this distinction does not apply to markets for the sale of school books by publishers to dealers, since publishers do not supply all types of dealers. The Commission's calculations, on the basis of information from the notifying party and others, show that nearly 95% of publishers' turnover on the sale of school books comes from sales to booksellers, the balance being from specialised wholesalers selling school books generally not to retailers but direct to schools.

(2) DISTINCTION BY MAJOR CATEGORIES OF BOOKS

(a) The notifying party

(188) The notifying party considers that no distinction should be made between markets for the sale of books to dealers according to categories of books (142) since it argues that, apart from class 1 bookshops, the discounts given to dealers are the same for all categories of books, both at Hachette Livre and at VUP.

(189) But the notifying party considers that the market which it defines as the market for the sale of books by publishers to the final consumer should indeed be subdivided according to major categories of books - (i) general literature titles, (ii) children's books, (iii) art books, (iv) guides and manuals, (v) strip cartoons, (vi) educational books, (vii) academic and professional books, (viii) reference works and (ix) and books published in instalments (143).

(b) The Commission

(190) The Commission considers that the sale of books by publishers to dealers must be subdivided according to major categories of books.

(191) As the notifying party acknowledges, the argument from the absence of differentiated discounts is not valid for class 1 booksellers, whose discounts are modulated according to categories of books, nor for school books and strip cartoons, to which different scales apply (144). Moreover it emerges from the Commission's investigation that publishers other than Hachette Livre and VUP generally modulate the discounts they negotiate according to categories of books (145).

(192) It further emerges from the Commission's investigation that, on the supply side, publishers distinguish between different categories of books that do not have the same physical properties and are not produced the same way or on same print runs (146). Different publishers (or in the case of a group different publishing houses) may produce different categories of books. (147)

(193) On the demand side, certain dealers do not sell all the categories of books, either because they specialise in a given category or because certain distribution channels do not sell all categories. (148)

(194) It follows from the foregoing that each of the major categories of books constitutes a relevant product market - (i) general literature titles, (ii) children's books, (iii) art books, (iv) guides and manuals, (v) strip cartoons, (vi) educational books, (vii) academic and professional books, (viii) reference works and (ix) and books published in instalments.

(3) SUBDIVISION OF MAJOR CATEGORIES OF BOOKS

(195) Within certain of these categories of books, it may be necessary to further subdivide into smaller segments.

(a) General literature titles

(196) The sale of general literature titles includes novels (classics149, contemporary, romantic, erotic, humorous, crime, science-fiction, horror), drama, collections of poetry, essays, criticism, analytical and current-affairs titles, marketed in different formats, at different prices and possibly in specific series.

(i) Distinction according to genre

* The notifying party

(197) According to the notifying party, these works constitute one market for sales of general literature titles by publishers, there being no difference based on authors, publishers or prices to justify more detailed subdivision, which does not correspond to any form of publishing or business reality (150).

* The Commission

(198) The Commission's investigation broadly confirmed this analysis. The various genres (novels, studies, current affairs, documents, thrillers, science-fiction, etc") are indeed substitutable.

(199) On the supply side, the operators involved in publishing general literature titles publish across the different genres (151) and the borderline between the genres is not a hard and fast one. A thriller or a science fiction book can be regarded as a "novel", and a study might be classified under "current affairs". The same applies to romantic fiction, which, depending on the quality of the writing or the reputation of the author, might be classed under "novels". On the demand side, therefore, while dealers have separate shelving units for certain types of book, this absence of a rigid demarcation between genres can prompt them to vary the classifications, especially of new titles.

(ii) Distinction according to format

* The notifying party

(200) The notifying party considers that there are no grounds for making a distinction between general literature titles depending on their format (large and pocket formats). It puts forward three particular sets of arguments in support (152).

(201) The first argument is that the book chain is identical irrespective of format. In France, all formats come from the same suppliers, sometimes the same machines, the same printshops and the same marketing and distribution circuits.

(202) The second is that there is a degree of continuity in prices between the different formats (153).

(203) The third is that there is a high degree of substitutability on both the supply and the demand sides. On the supply side, substitutability between the different book formats is highlighted by the gap between the date of pocket-format publication of a large- format best-seller and by the coexistence of pocket- and large-format series at the same publishing houses. The notifying party adds that more and more publishing houses have their own paperback series to give a title the second wind that comes with a pocket-format edition. On the demand side, the various formats meet the same need and match a similar type of leisure or culture-oriented consumption pattern. And purchasers of books on the same subject are sociologically comparable, irrespective of format.

* The Commission

(204) Review of the notifying party's arguments and the findings of the Commission's investigation do not bear out the notifying party's conclusions in this respect.

(205) The first obvious criterion for distinguishing a pocket-format book is the immediate visual identity conferred by the specificity of its format (usually 10x18 cm or 11x17.8 cm), which is very different from the format of a large-format book, and there is very little overlap between the two types of product where general literature titles are concerned (154).

<emplacement tableaux>

(206) Moreover, pocket-format books appear in a specific series. There is a unifying graphic character that gives the work and its author a supplementary identity (such as the indication of the number of the work in the series). Large-format series do not have a similar strong and familiar visual identity. And the production of a pocket- format book is very different from that of a large-format book. In particular, for pocket-format books, unlike other books, soft covers are preferred to hard, the signatures are glued rather than stitched, the cover is film-coated, and the printing technique is generally different (rotary presses with paper in reels rather than sheet- paper machines).

(207) Second, publishing a pocket-format book differs from publishing a large-format book in terms of the separate rights that go with it. Publishing a pocket-format book corresponds to a separate form of exploitation of publishing rights, the remuneration being half that of a large-format publication. Likewise, when publishing houses publish a work in the two formats, there are separate accounting entries for the rights deriving from sales of the two formats. And for books originally written in foreign languages, non-French-speaking agents and publishing houses sometimes assign rights in large and pocket-format editions separately (155). Unpublished general literature titles issued for the first time in pocket format are also of limited importance as the percentage is very low, unpublished works in this category generally being published in large format to begin with (156).

(208) Third, the prices and cost structures of the two types of format are very different. The average print run of a paperback is far greater than that of a large-format book (157). Consequently, the cost structure is by no means the same for the two formats. The marginal cost of a typical work published by the notifying party in paperback format is four times less than in large format for an identical number of pages (158). The copyright for the pocket-format edition is lower than the primary publishing rights for the original large-format edition. The price of a pocket-format book is set on a specific scale, which ensures a degree of pricing consistency within the same series and thus facilitates promotion operations relating to more than one title in the series. And the retail price of a pocket-format book is generally determined on the basis of the number of pages and the rhythm of sales (159), whereas the price of a large-format book is based virtually entirely on the cost price plus a margin for publishing expenses. On the general literature side, therefore, the retail price of a pocket-format book is on average three times lower than the price for the large-format (160). In addition, there is a clear gap between the price ranges for pocket-format and large- format books. While 3.8% of pocket-formats cost more than euro 9, and 3.2% of large- formats cost less than euro 10, the fact remains that 91% of pocket-formats are priced between euro 2 and euro 9 and 95% of large-formats are priced between euro 10 and euro 24 (161).

<emplacement tableau>

(209) The average price of a pocket-format book is euro 5.78, as against euro 18.27 for a large- format book, the difference being a factor of three (162). In addition, as has been seen, the technical features of a pocket-format book are very different from those of a large-format book (gluing, trimmed folios, film-coated covers) and the average print run for a pocket-format book is far higher than for a large-format book (163). Consequently, the cost structure is not the same for the two formats. The marginal cost of a typical work published by the notifying party in pocket format is four times lower than for the large format for an identical number of pages (164).

(210) Fourth, books in pocket format tend to differ from large-format books in terms of the movement of demand (165). In 2001, pocket-format sales expanded by 10%, as against 3.2% for publishing overall (including pocket-format books) (166). There also appears to be a strong seasonal fluctuation, in particular in sales of large-format books, which tend to be concentrated at the beginning of the school year and around the award of literary prizes.

(211) Fifth, the Commission has observed that the main publishing houses do not manage their pocket-format series direct but, in most cases, entrust them to specialised subsidiaries (167). The activity of pocket-format publishing houses is specific in that it focuses primarily on the acquisition of specific rights, with no work on content (168) as is usually the case of traditional large-format publishing. In an internal strategy paper [...]* (169). Here it should be added that pocket-format publishing houses generally operate at higher levels of profitability than large-format publishing houses. The specific nature of pocket-format books is underlined further by Dominique Goust, chairman of Le Livre de Poche, the pocket-format books arm of Hachette Livre, who states that "pocket-format publishing is not creative publishing" and that "the question about a new book is always whether 30 000, 40 000 or more readers are going to like it. That is not necessarily the approach of a traditional publisher" (170).

(212) Sixth, the demand-side substitutability invoked by the notifying party refers only to readers, and not to dealers, who are the demand side of the relevant markets. In any event, it emerges from the Commission's investigation that the purpose of a pocket- format book is not the same as that of a large-format book (171) since the way it is made means that it is not designed for long-term conservation or regular or careful reading. The readership profile also seems different. The 15-25 age bracket accounts for 14% of pocket-format book buyers but only 2.5% of large-format book buyers. The over- 65s, on the other hand, account for 14.1% of pocket-format book buyers but 34.4% of large-format book buyers. Young people and students are thus more attached to the pocket format than the older readers (172).

(213) Lastly, pocket-format books are usually published nine months after the first large-format edition (173) and remain on the market thereafter. While a large-format book stays on the shelves for about three months (the duration of a literary season), a pocket-format book will stay there far longer (between five and ten years, and even more for a classic). There may be a degree of substitutability between different periods of time, but it concerns different types of final consumer. Readers who are willing to pay a high price for the larger format will buy it straight away, whereas those whose resources are more limited are more likely to wait for the pocket-format edition and buy later. Substitutability over time thus amounts to a distinction in the relevant population (174). Finally, in its reply to the statement of objections, the notifying party attempts to justify its claim that pocket-format and large-format books belong to the same product market by comparing pocket-format books to generic medicines, which generally belong to the same market as the branded medicines on which they are based. The Commission cannot subscribe to this argument for the following reasons.

(214) First, generic medicines are identical to branded medicines in terms of composition and marketing, which is not the case with pocket-format and large-format books. Also, generic medicines are normally marketed by companies that are in competition with those holding the marketing patents for the branded medicine, as the latter accept the marketing of their product in generic form because they have no choice. In contrast, a pocket-format book is a second edition of a large-format work, published by decision of the first publishing house. The pocket-format edition may be published by a competing publishing house, but it is the publishing house that first published the work that decides whether it wants to assign its rights or exploit them itself, at what point the pocket-format rights are assigned and hence at what point the work is published in pocket format. Finally, and again in contrast to the situation found on the market for pharmaceutical products, any overlap in the marketing periods for the first and second editions of the same work is generally a marginal phenomenon and, most importantly, lasts for a short period of time. This is not the case with medicines, which remain on sale side-by-side for long periods.

(215) Accordingly, the notifying party's argument that there are similarities between the publishing and medicinal product sectors is unconvincing and cannot be endorsed.

(iii) Conclusion

(216) It follows from the foregoing that a subdivision between genres within the general literature category is not relevant. But the sale of large-format general literature titles by publishers to dealers and the sale of pocket-format books by publishers to dealers are distinct relevant markets. By way of conclusion, the notifying party maintains in its reply to the statement of objections that "pocket-format books and large-format books are substitutable", but in fact "it is common knowledge today that a large- format edition and a pocket-format edition are not in competition with each other, but that, on the contrary, they complement one another" (175).

(b) Children's books

(217) Children's books are for a readership aged less than fifteen, and their presentation is specific, as they are commonly illustrated.

(i) The notifying party

(218) According to the notifying party, a finer subdivision of the market, distinguishing children's books and albums or picture books would be excessive and not conducive to a sound understanding of relations between competitors (176).

(219) It argues that the two types of work pursue the same object and are published by the same publishing houses in the same manner and in the same conditions, the authors and illustrators often being specialised in them. The notifying party indicates that Hachette Livre publishes both picture books and reading books. The same applies to the children's publishing houses of VUP, Syros, Gallimard and Flammarion.

(220) The notifying party does not address the question of a possible distinction by format.

(ii) The Commission

(221) The Commission's investigation confirms the notifying party's analysis whereby there is no need to further subdivide the market for the sale of children's books by publishers to dealers, distinguishing between picture books and reading books.

(222) On the question whether a distinction should be made on the basis of format, unlike the market for the sale of general literature titles, the investigation did not provide adequate evidence of the need for such a subdivision of the market for children's books. Although a number of the specific factors described above as regards the publishing of pocket-format general literature titles apply similarly for children's books, the main differentiating factors are no longer found.

(223) First, regarding visual identity, unlike general literature titles, children's books come in a multitude of different formats, and there is extensive overlap between pocket and large-format books. There is a wider variety of formats than there is for general literature titles. There are 22 different formats for general literature titles but 131 for children's books (177).

<emplacement tableau>

(224) There are rarely separate rights for publishing pocket-format and large-format children's books. Children's books published in pocket series are usually in that format originally and rather than being second editions (178).

(225) There is no clear differentiation between the price ranges for pocket and large-format books. The gap between the average price for a children's pocket-format book (euro 4.80) and the average price for a large-format book (euro 8.60) is not as wide as the gap in the case of general literature titles (179). 36% of pocket-format books are priced between euro 4.80 and euro 8.60, and 25% of large-format books are priced between euro 4.80 and euro 8.60. There is also a considerable overlap between the two ranges, as about 25% of large- format children's books are priced between euro 4 and euro 5.

<emplacement tableau>

(226) There are no children's books publishing houses specialising in pocket-format books, as all publishing houses deal in all formats, and the gap between the duration for which pocket and large-format books are stocked that was observed for general literature titles is not mirrored for children's books, since they are rarely re-issued.

(iii) Conclusion

(227) It follows from the foregoing that there are no grounds for making a subdivision between children's books and albums or picture books. Nor is there evidence of a need for a subdivision by type of format. The sale of children's books by publishers to dealers is therefore a single distinct relevant market.

(c) Art books

(228) Art books, albums and illustrated books cover a variety of subjects - painting, sculpture, music, architecture, cinema, photography, the sea, the mountains, sport, cooking, etc. Their most distinctive features are the richness of the illustrations and/or the quality of the photographs, and their presentation, in particular the use of glossy paper. Accordingly, they are much more expensive than any other category of books.

(229) The notifying party states that there are no grounds for a finer subdivision of the market into art books, albums and illustrated books.

(230) The findings of the Commission's investigation broadly confirm this analysis.

(d) Guides and manuals

(231) The purpose of guides and manuals is to provide practical guidance and information that will be useful in everyday life on matters such as decoration, DIY, gardening, nature, cooking, wine, travel, language learning, sports, health etc.

(i) The notifying party

(232) According to the notifying party, there are no grounds for distinguishing by subject-matter. It does not address the question of a possible distinction by format.

(ii) The Commission

(233) The Commission's investigation confirms the notifying party's analysis whereby there is no need to further subdivide the market for the sale of guides and manuals by publishers to dealers on the basis of subject-matter.

(234) The investigation did not provide adequate evidence of the need for a subdivision of the market for guides and manuals by format, such as there is in the market for the sale of general literature titles. Although many of the specific features described above for pocket-format general literature titles also apply to guides and manuals, the main differentiating factors are no longer found.

(235) Regarding visual identity first of all, unlike general literature titles, guides and manuals come in a multitude of different formats and there is extensive overlap between pocket and large-format books. There is a wider variety of formats than there is for general literature titles. There are 22 different formats for general literature titles as against 46 for guides and manuals (180).

<emplacement tableau>

(236) Moreover, there are rarely separate publishing rights for pocket and large-format guides and manuals. Guides and manuals published in pocket-format series usually appear for the first time in that format, rather than being second editions (181). Moreover, as guides and manuals are often the product of collective authorship, there is no real acquisition of rights but rather payment for work done in the conventional sense, whether the work is in pocket or large format.

(237) There is no clear differentiation between the price ranges for pocket-format and large-format books. The gap between the average price for a pocket-format guide or manual (euro 5.80) and the average price for a large-format book (euro 18.50) is comparable to that for general literature titles (182), but there is considerable overlap between the two ranges, as about 25% of large-format guides and manuals and 75% of pocket-format guides and manuals are priced between euro 5 and euro 8.

<emplacement tableau>

(238) There are no publishing houses specialising in pocket-format guides and manuals, as all publishing houses deal in all formats, and the gap between the duration for which pocket and large-format books are stocked that was observed for general literature titles is not mirrored for guides and manuals, since they are rarely re-issued.

(iii) Conclusion

(239) It follows from the foregoing that there are no grounds for making a subdivision between subjects in the guides and manuals category. Nor is there evidence of a need for a subdivision by format. The sale of guides and manuals by publishers to dealers is therefore a single distinct relevant market.

(e) Strip cartoons

(240) A strip cartoon is a sequence of drawings telling the same story or presenting the same character. Strip cartoons differ from other works by their content, the fact that they are often in colour and the fact that they are often published in a series of successive volumes. Furthermore, strip cartoons are often sold in specialised bookshops and - even more often - are published by specialised publishers.

(241) It emerges from the Commission's investigation that there are no grounds for a finer subdivision of this market.

(f) Educational books

(242) Educational publishing includes all educational material below university education level. Adult education material is excluded.

(243) School textbooks include all the books prescribed by teachers for syllabuses defined by government departments or other public education authorities and used by pupils and teachers. This includes textbooks, exercise books and teaching materials used by pupils in the classroom and material specifically identified by teachers for use with the textbook (183).

(244) Educational supporting materials are materials used to support the pupil's passage through school to go into further detail on certain aspects of the syllabus. This category includes chronologies, support and practice books, collections of exercises, holiday workbooks, books containing advice on how to write essays and the like.

(i) Distinction between school textbooks and educational supporting materials

* The notifying party

(245) The notifying party (184) considers that the market for sales of school textbooks is distinct from the market for educational supporting materials, since school textbooks are prescribed whereas educational supporting materials are used simply as adjuncts, and while all publishers of educational supporting materials also publish school textbooks, the converse does not apply. It adds that in addition to prescribed texts there are educational supporting materials that are .highly recommended. by teachers.

(246) School textbooks call for heavy initial investment linked to promotion costs, constraints as to content and drafting times (185), specific formats (large format, four-colour printing), higher average selling prices, different pricing structures, exemption from the resale price maintenance schemes (186), separation between the person prescribing and the buyer, specific promotion by representatives visiting teachers, and seasonal distribution patterns.

(247) Educational supporting materials differ from school textbooks in that they are not prescribed (even if they may be recommended), the cost of buying them is borne entirely by the user, specific formats are used (often in consumable form with less illustration), average prices are lower, price structures differ, dealers are given larger discounts, promotion is mainly at the point of sale or by advertising, and marketing focuses on large-scale retail shops with generalist teams.

* The Commission

(248) The Commission shares the notifying party's analysis, since school textbooks and educational supporting materials have significantly different physical characteristics, contents, forms of use, marketing techniques, discounts and prices.

(249) In particular, school textbooks need to comply with the syllabus determined by the public authorities and are prescribed by teachers, so that, as became clear from the Commission's investigations, the price is a secondary selection factor (187). For the same reason, they have to be marketed through specific channels (specimens sent to teachers, teams of educational representatives), and they are marketed mainly through class 1 booksellers and specialised wholesalers (188).

(250) Publishers, on the other hand, have greater freedom as to the contents and teaching methods of educational supporting materials (189). Although some categories (such as holiday workbooks or final-year preparation books) are also to some extent linked to the syllabus, this is not as true of some others (such as annotated editions of classic authors or grammar books). Parents are commonly quite free to choose their children's educational supporting materials (except where books are strongly recommended by the teacher), and prices can be a major criterion for opting between two books on the same subject (190). Marketing techniques are similar to those for other categories of books (191), and unlike school textbooks they are extensively sold through hypermarkets.

(251) It follows from the foregoing that the sale of school textbooks by publishers to dealers and the sale of educational supporting materials by publishers to dealers are distinct relevant markets.

(ii) Distinction between school textbooks by school level

* The notifying party

(252) The notifying party considers that there are no grounds for subdividing the market for the sale of school textbooks on the basis of the level or type of school (192) (primary, secondary, technical/vocational), since the differences that may exist between these segments are not significant enough to make them distinct markets. In particular, the notifying party argues that there are strong similarities in the drafting, manufacture, physical characteristics, sales circuits, prices and marketing and distribution techniques for books at all levels of schooling.

* The Commission

(253) The Commission's investigation highlights a number of factors, in particular the prices of books, their physical characteristics, their useful life or their buyers, that could be taken as evidence supporting the hypothesis of distinct markets. But all school publishers are active at all levels of schooling, and the investigation did not reveal the existence of barriers to entry by a publisher active at one level of schooling to the market segment of another level. There is thus a high degree of substitutability on the supply side, which means that there is a single market for the sale of school textbooks by publishers to dealers.

(g) Markets for the sale of academic and professional works

(254) According to the notifying party, it would be irrelevant to subdivide the market for the sale of academic and professional works according to the type of reader (193), as the same works are read by professionals and by university students or teachers alike. It also argues against a subdivision according to the type of medium used (CD-ROM or print), as a CD-ROM incorporates the same content as the paper version of a work and the two are most often sold together in inseparable form (194).

(255) However, the notifying party considers that a distinction needs to be made according to discipline (law; science, technology and medicine; economics and social sciences), as publishers specialise in certain disciplines (sometimes even one discipline) and the readers of academic and professional works are themselves specialists (195).

(256) The findings of the Commission's investigation have broadly confirmed the notifying party's analysis. However, no matter what product market definitions are adopted, the transaction is unlikely to create or strengthen a dominant position with the result that competition is significantly impeded, so the question of the exact definition of the markets for sales of academic and professional works by publishers to dealers can remain open.

(h) Reference works

(257) There are two sorts of reference work - smaller works (in one to three volumes) and larger works (more than three volumes).

(258) The notifying party considers that a distinction must be made between the two types by reason of their format, distribution techniques (booksellers or supermarkets as against mail orders and direct selling), price (euro 15 to euro 40 as against euro 1 000) and production costs. It also points out that producers of larger reference works tend to be distinct from producers of smaller reference works (196).

(259) The findings of the Commission's investigation confirm the notifying party's arguments. The markets for smaller reference works and larger reference works are accordingly distinct product markets.

(i) Smaller reference works

(260) Among the smaller reference works a distinction must be made between dictionaries and encyclopaedias, each of the two types of work being capable of further subdivision into several relevant product markets depending on size and manner of use.

* Dictionaries

(261) Regarding dictionaries, the notifying party makes a distinction between single-language dictionaries and bilingual or multilingual dictionaries. But it acknowledges that production costs are similar whatever the type of dictionary (197).

(262) Single-language dictionaries are collections of words in alphabetical order, used by readers to check the meaning of a word, its etymology, its synonyms, and expressions illustrating its use. The notifying party includes in the market for single-language dictionaries both paper and electronic versions (CD-ROMs or Académie Française on-line dictionaries).

(263) Bilingual or multilingual dictionaries are collections of words in alphabetical order accompanied by their translations in one or more languages. As in the case of single-language dictionaries, the notifying party includes in the market both paper and electronic versions (CD-ROMs or translation sites on the Internet). The notifying party also considers that, while to the consumer a French-Italian dictionary is not substitutable for a French-English dictionary, all publishers of bilingual or multilingual dictionaries are able to supply dictionaries in any foreign language: all these dictionaries are worked up from the same list of words; the only changes are in the degree of exhaustiveness and the teams of translators.

(264) The Commission's investigation shows that the publishers of both single-language and bilingual or multilingual dictionaries are the same, and that the two types of work are marketed by the same techniques.

(265) It may be, therefore, that the two types of work do not constitute distinct relevant markets. But the question of the precise definition of the market can be left open since, whatever definition is used, the notified transaction will strengthen a dominant position.

* Encyclopaedias

(266) For encyclopaedias, the notifying party distinguishes general encyclopaedias from specialist encyclopaedias, the former not meeting the same consumer needs and not being published by the same publishers as the latter.

(267) It also distinguishes single-volume encyclopaedias from multi-volume encyclopaedias, the latter being substantially heavier and more expensive. The notifying party considers that multimedia encyclopaedias, sales of which are expanding, should be included in the markets.

(268) The notifying party includes both paper versions and computerised versions of encyclopaedias (198).

(269) The Commission's investigation confirms the notifying party's arguments. The markets for specialist encyclopaedias and general encyclopaedias are accordingly distinct product markets.

(270) On the inclusion of paper versions and computerised versions in the markets for encyclopaedias, the Commission's investigation reveals that computerised versions are often, though not always, sold along with the hard-copy version.

(271) But the question of the precise definition of the market can be left open since, whatever definition is used, the notified transaction will strengthen a dominant position.

(ii) Larger reference works

(272) Multi-volume dictionaries and encyclopaedias (whether general or specialist) are more exhaustive and carry more illustrations than single-volume dictionaries and encyclopaedias. They are also considerably more expensive (at an average price of EUR 1 000 compared with around EUR 30) and are generally sold through special sales channels (direct sales, mainly by sales agents and by mail order).

(273) The notifying party considers that no distinction should be drawn in this market according to medium (paper, pre-installed software, CD-ROM or on-line) (199).

(274) However, it emerges from the Commission's investigation that paper versions of encyclopaedias must be distinguished from multimedia versions, as the former have a number of special features. First, they are marketed in a particular way, via a network of specialised representatives or sales agents, who approach potential buyers direct. Another distinctive feature is the unit price of the works sold (around EUR 1 000) (200). Finally, on the supply side, they cannot be substituted by other types of publication, given their volume, the scope and complexity of the publishing project and the initial investment required, which does not compare with that for other categories of books.

(275) Accordingly the markets for paper and multimedia versions of larger reference works are distinct product markets.

(iii) Books published in instalments

(276) Some specialist encyclopaedias are divided into a number of instalments/booklets, which are published in succession and sold at regular intervals in kiosks or newsagents or by subscription. Originally, instalments of encyclopaedias accounted for the lion's share of this market, but today most works in this category consist of free gifts, supplements to collect (traditional instalments), CDs, CD-ROMs or DVDs which accompany information booklets or explanatory booklets.

(277) It emerged from the Commission's investigation that the market for the sale of instalments by publishers to dealers is a distinct product market and should not be subdivided further.

A.3.c. MARKETS FOR THE SALE OF BOOKS BY WHOLESALERS TO CLASS 3 RETAILERS

(278) As has been seen under the description of publishers' marketing services, since the turnover generated by small outlets, where the sale of books is only a marginal activity, rarely covers the costs of a marketing team set up to cover these outlets and the investment needed to set up a network of salesrooms, the sale of books to this type of retailer goes through wholesalers. Wholesalers are thus the access point for class 3 customers, which includes both small outlets not specialising in books and supermarkets.

(279) The main difference between the mode of business operation of wholesalers and of the publishers and their marketers is that wholesalers stock a range of different categories in what amounts to a one-stop shop. Since class 3 outlets do not specialise in the books business, they rarely have the same ability as booksellers or central buying facilities to select titles from the very wide range available. Class 3 outlets therefore go for a limited number of suppliers who can offer them ready-made solutions comprising the regular supply of a selection of titles matching their clientele, logistics services (orders, restocking, returns) and invoicing. The Commission has been able to ascertain that class 3 outlets sell a limited range of titles which they obtain from a restricted number of wholesalers, that their discount is particularly low (less than 30%) and that their purchases are heavily dependent on automatic supplies or selection by suppliers' representatives.

(280) This activity is exercised by wholesalers mainly through the supply of assortments and/or rack-jobbing in small outlets and supermarkets. Whereas a publisher uses a marketing team that is generally specialised in a major category to sell titles to booksellers and hypermarkets, a wholesaler will select and group the production of several publishers so as to propose a range of supply matching the needs of small outlets and supermarkets, while paring to the bone the costs of marketing to these customers, since it is usually the wholesalers themselves who decide on the assortment to be offered (201).

(281) The size of the standard assortment set up by a wholesaler in these types of outlets depends mainly on the outlets' limited surface area and turnover for books. Not being specialised, small outlets generally offer a limited quantity of books generally ranging from 100 to 200 titles, or 1 000 to 2 000 for supermarkets.

(282) A major proportion of this inevitably eclectic range is accounted for by best-selling general literature titles, enjoying heavy media coverage, and pocket-format general literature titles. The rest of the assortment put on the shelves of these outlets by the wholesalers consists of a selection of guides and manuals, children's books and currently popular strip cartoons (202).

(283) Although certain of these small outlets and supermarkets are supplied by two or even more wholesalers, the size and composition of the average assortment and the practice of having it selected and stacked by the wholesalers do not vary to a point where they might be approached specifically by marketers on the basis of the category of books sold.

(284) While Hachette Livre (operating via LDS) and VUP (operating via La DIL) are generally both represented in these outlets, the main reason is that both VUP and Hachette Livre have opted not to sell each other's products. To fill out their ranges, LDS and to a lesser extent La DIL do obtain supplies from their main competitors, of "must-stock" items meaning very well known titles or series marketed or distributed by Gallimard, Flammarion, Seuil and other smaller marketers/distributors, and the Commission's investigation reveals that combined supplies from VUP and Hachette Livre would cover all the categories of books sold at these outlets and therefore meet their needs.

(285) It follows from the foregoing that the size and composition of the standard assortment and the fact that it is selected and stacked by the wholesalers warrants the conclusion that there is no need to make distinctions within this market on the basis of the categories of titles sold by wholesalers to this type of retailer.

(286) There appear to be good grounds for distinguishing the market for the sale of books by wholesalers to retailers from the market for sales of books by publishers to other types of outlets (general bookshops and hypermarkets) or to wholesalers themselves.

(287) For one thing, sales by wholesalers to class 3 retailers are an extra link completing the book chain, as they form a market downstream of markets for the sale of books by publishers to wholesalers. For another, as has been seen, supply is generally in the form of a basket of titles sold to the retailer rather than a set of individual transactions relating to different types of title. The discounts given by wholesalers to retailers are different from those from publishers to booksellers, hypermarkets (203) and wholesalers themselves. And either the supply side (wholesalers) or the demand side (small outlets and supermarkets) or both at once will be different from the players on other markets for the sale of books by publishers.

(288) It follows from the foregoing that there are good grounds for considering the sale of books by wholesalers to class 3 retailers as a distinct relevant market for the purposes of the competition analysis of the notified transaction.

A.3.d. MARKETS FOR THE SALE OF BOOKS TO THE FINAL CONSUMER

(289) Books are sold in several types of outlet such as bookshops, specialised superstores (204), supermarkets and hypermarkets, newsagents and specialist shops (205). Other forms of retailing include sales through book clubs, distance selling, discount Shops (206) and sales through sales agents.

(290) The notifying party states that (207), with the exception of remainder shops selling end-of-stock titles at low prices, all outlets offer the same books to consumers at identical prices, so that, from the consumer's point of view, they are substitutable. According to the notifying party, this means that they belong to the same product market. It also notes that sales by sales agents and distance selling are distinct product markets because of the specific form of marketing employed.

(291) The findings of the Commission's investigation confirm the existence of a distinct product market for sales by sales agents, the main form of marketing for larger reference works. The market for the sale of larger reference works has a number of specific features, one of which is the form of marketing, via a network of specialised representatives or sales agents, who approach potential buyers direct.

(292) Without stating any opinion on the existence of a distinct market for sales through book clubs, the notifying party points out that the market for retail sales cannot be analysed without taking into account the influence of France Loisirs.

(293) In previous decisions the Commission has indicated that in operations involving the sale of consumer products - a category which covers much of publishing production - it is necessary to identify the traditional market for retail sales in shops which marks the boundary of business relations between dealers and the final consumer. Given the differences in marketing, distance selling and sales by sales agents are a distinct product market. The same applies to remainder shops, in particular because of the differences in selling price.

(294) As regards the existence of a relevant market for sales through book clubs, the Commission has spelt out the characteristics of book clubs in previous decisions, stressing in particular that customers become members after accepting a promotional offer, where membership entails an obligation to make regular purchases (208). Books are bought from book clubs on the basis of a limited catalogue comprising at one and the same time books already published by other publishers - the rights in which have been acquired from those publishers - and unpublished books (which account for 20% in the case of France Loisirs).

(295) Without adopting a final position on the existence of a relevant market for sales through book clubs, the Commission has taken the view that the activities of book clubs belong to the "distance selling" segment, which also includes sales through the Internet and by mail order. Sales through book clubs do not therefore form part of retail sales in shops as defined by the notifying party (209).

B. GEOGRAPHIC MARKETS

(296) The notifying party stresses that all the markets concerned by this transaction are national or smaller markets (in particular the French-speaking part of Belgium). The Commission's investigation, on the other hand, indicates that the relevant geographic market is a single market covering the French-speaking area of the Community (France, the French-speaking part of Belgium and Luxembourg) or the French-speaking area of Europe (France, the French-speaking part of Belgium, Luxembourg and the French-speaking part of Switzerland), with the exception of the market for the sale of school textbooks and the market for the sale of law books by publishers to dealers, which have to be defined as national markets.

B.1. MARKETS FOR PUBLISHING RIGHTS

B.1.a. THE MARKET FOR REPRODUCTION RIGHTS FOR IMAGES AND MAPS

(297) The notifying party claims that reproduction rights for images and maps are acquired from a great many bodies and rights holders throughout the world.

(298) It has been found that such rights are indeed acquired worldwide, an indication that the market for rights for images and maps has a global dimension. However, the exact definition of this market can be left open in that, whatever the geographical dimension, no market is affected by this transaction.

B.1.b. MARKETS FOR PUBLISHING RIGHTS

(1) PRIMARY MARKETS FOR FRENCH-LANGUAGE PUBLISHING RIGHTS ("PRIMARY RIGHTS MARKETS")

(a) Primary markets for French rights

(i) The notifying party

(299) The notifying party argues that a geographic market for first publishing rights (which it claims does not exist) would be on a purely national scale, in that French authors would assign all their rights exclusively to French publishers, while Quebec authors would assign all their rights exclusively to Quebec publishers and most Belgian authors would assign all their rights to Belgian publishers (210). The notifying party none the less admits that some Belgian authors assign their rights to French publishers because they are sufficiently well-known to want to be present on the French national market.

(ii) The Commission

(300) However, the Commission's investigations have not produced any concrete evidence of the existence of a national market.

(301) To begin with, publishers pay no attention whatsoever to an author's nationality when selecting a work (211), the sole essential criterion being that of language. While there are, of course, examples of authors published by publishers of their own nationality, it is clear that many French-speaking foreign authors are published by French publishers. In addition to the examples which the notifying party quotes of Belgian authors published by French publishers (212), many examples can be found of Canadian, Algerian, Moroccan, Czech or Senegalese authors who have contracts with French publishers, provided of course that they write in French (213).

(302) What is more, when publishers acquire publishing rights in French-language works this covers all French-language rights, with no restriction as regards country. In the course of its investigation, the Commission has learnt from publishers and authors alike that "non-translated" authors assign their French-language publishing rights for marketing throughout the world (214).

(303) The organisation of French-language international literary prizes without discrimination by nationality and the nationality of the authors who win these literary prizes also point to a worldwide primary market in French-language publishing rights. Similarly the Goncourt (215) and Renaudot (216) prizes, which the notifying party counts among the most prestigious French literary prizes (217), are not awarded on the basis of the nationality of the author but on the basis of the language in which they write.

(304) Finally, the existence of a cultural and linguistic movement specific to French speakers is another illustration of the supranational dimension of literary and publishing operations which generate French rights in primary markets.

(305) In the light of the above, it can be seen that the primary markets for French-language publishing rights for works originally written in French are worldwide.

(b) Primary markets for foreign rights

(i) The notifying party

(306) The notifying party considers that the acquisition of rights of foreign authors constitutes a national market, in that the rights of foreign authors are always assigned to French publishers, who are therefore the principal players in the market. The notifying party explains this by the fact that Belgian or Luxembourg publishers do not have sufficient financial capacity (218) to compete with French publishers, who are able to offer authors higher royalties and advances.

(ii) The Commission

(307) However, the fact that French publishers are generally chosen by non-French-language publishing houses is not in itself proof that the national market exists, since the non-French-language publisher assigns rights for all aspects of the exploitation of the work in French. The contracts for the acquisition of French-language rights provided by the notifying party (219) and the answers given by literary agents and publishers to the Commission's investigation (220) show that as a rule French-language rights in works are acquired for the entire world. This means that a publisher in a French-speaking country wishing to publish the French-language version of a work by a foreign author in French-speaking countries will be competing with French-language publishers from other countries who would also like to publish the works of the same author in their own countries of origin. The geographic restriction of French-language publishing rights concerns American publishers or literary agents, who in some cases themselves wish to exploit publishing rights for Canada, a bilingual, neighbouring country. But this restriction is still only Marginal (221).

(308) The Commission's investigations have revealed that publishers selected to publish the French-language version of works originally written in a foreign language are not always French (222). Literary agents seeking to sell foreign rights for French-language exploitation operate not only in France but also in other countries where there are French-language publishers, such as Canada, Belgium and Luxembourg. As has already been stated above, competition between publishers for the acquisition of the foreign rights of an author are not therefore confined to the national level, France in particular, but potentially include all French-language publishers in all countries throughout the world.

(309) In the light of the above, it can be seen that the primary markets for foreign-language publishing rights for works originally written in a foreign language are worldwide.

(2) SECONDARY RIGHTS MARKETS

(a) Secondary markets for pocket-format rights

(i) The notifying party

(310) The notifying party considers that the acquisition of pocket-format publishing rights constitutes a national market, claiming that French publishers assign their pocket-format rights solely to French publishers, just as Quebec publishers assign their pocket-format rights solely to Quebec publishers and Belgian publishers generally publish their pocket-format books themselves.

(ii) The Commission

(311) The Commission does not agree with the notifying party's analysis and considers, on the contrary, that the market for pocket-format publishing rights is supranational.

(312) The Commission can see no specific barrier to French-language publishers in different countries competing for the acquisition of pocket-format publishing rights. Cases exist of Canadian publishers assigning pocket-format rights to French Publishers (223). The Commission's investigation also reveals (224) that when publishers assign pocket-format publishing rights this covers all French-language rights, with no restriction as regards country (225), and that French publishers themselves exploit pocket-format publishing rights in Canada.

(313) In the light of the above, the Commission considers the secondary market for pocket-format rights to be a supranational market.

(b) Secondary markets for club rights

(314) The notifying party gives no definition of the geographic dimension of the secondary markets for club rights.

(315) Applying the same arguments as for secondary markets for pocket-format rights, it can be seen that the secondary markets for club rights are supranational, and could be defined as worldwide.

(316) This is further illustrated by the fact that the same operator, Bertelsmann, is one of the leading players in France as well as in Belgium, Switzerland and Canada.

B.2. MARKETS FOR MARKETING AND DISTRIBUTION SERVICES

B.2.a. MARKETS FOR MARKETING SERVICES

(1) THE NOTIFYING PARTY

(317) The notifying party expounds a series of arguments seeking to demonstrate that marketing services constitute national markets.

(318) First of all, it claims that the differences in the regulatory environment between France and other countries, mainly as a result of the Lang Act, give rise to a greater density of retail outlets in France than in other countries.

(319) The notifying party also claims that marketing structures depend on separate, national organisations. It is claimed that in Belgium supermarkets are supplied by wholesalers, whereas in France they are supplied primarily by marketers.

(320) It is also argued that the services offered by marketers in France and Belgium are different. In France rack-jobbing is practised only in supermarkets, whereas in Belgium it is also the practice in hypermarkets, restocking is faster in Belgium and, contrary to what happens in France, some large and medium-sized outlets in Belgium request personalised labelling of works on the shelves. The notifying party also maintains that the work of representatives in Belgium is different and more expensive than that of representatives in France, because, it is claimed, outlets in Belgium are on average smaller than in France.

(321) Finally, the notifying party claims that some publishers do not use the same marketing channels in Belgium as they do in France and that the players on the markets for marketing services to publishers are not the same in France and Belgium. For instance the notifying party points out that La Caravelle operates only in Belgium while Harmonia Mundi does so only in France.

(2) THE COMMISSION

(322) Even though the Commission has found that marketing contracts are exclusive, generally cover the entire world, with France, Belgium, Luxembourg and Switzerland specifically mentioned, and spell out the categories of dealers concerned (226) and even though the Commission's investigation has further shown that some marketers provide an integrated service in France, Belgium, Luxembourg and Switzerland for some of their customers at least, certain aspects of marketing can none the less be subcontracted at a more local or regional level.

(323) On the demand side, there is a high degree of substitutability between the various countries. While some publishers may use different firms for certain geographic areas, in most cases a single firm handles marketing and distribution for the entire French-speaking area of the Community (227). What is more, charges are virtually the same between the different zones; and the operation of the Belgian depots does not appear to differ much from that of French regional depots (228). However, in so far as there are separate markets for the products concerned depending on the level of marketing services as defined above, the geographic dimension of marketing services must be defined for each of the product markets identified.

(324) It can also be seen that a number of criteria pointing to the supranational dimension of marketing-service markets in France, Belgium and Luxembourg also apply to the French-speaking part of Switzerland. Marketing-service charges are in most cases similar for Switzerland and Belgium. Similarly there are significant cross-border flows at the level of book retail sales (and to a lesser extent at the level of bookshops) between French-speaking Switzerland, in particular the Geneva canton, and neighbouring areas of France (229).

(325) There are, however, also a number of features distinguishing the marketing-service markets in French-speaking Switzerland from those in the French-speaking area of the Community. First of all, and contrary to the situation in Belgium and Luxembourg, Swiss bookshops are not served direct by French marketers from France. Swiss bookshops do not themselves import books (230) but go via marketers and distributors, such as the Office du Livre, a subsidiary of the Lagardère group, which imports half the books sold in French-speaking Switzerland. The other players are Diffulivre, owned by Hachette Livre, Servidis in which Seuil has a 50% stake, and Dargaud, a subsidiary of Dargaud in Belgium. The Commission's investigation has found that Diffulivre works almost exclusively for publishing houses in the Hachette group. Publishers outside the group distribute via OLF. In particular, Gallimard, Flammarion and VUP work through OLF. Another difference in relation to the French-speaking countries in the Community is that the existence of the Swiss franc as the national currency, with a rate that varies against the euro, means that price relationships between Switzerland and the Member States can vary, and that for final consumers the price differential is less obvious than between, say, France and Belgium, where the two prices are often shown on the back cover. In addition, even though there are no customs duties between the Community and Switzerland, imports of books to Switzerland involve higher costs in terms of additional carriage and transit costs (231).

(326) The question of whether or not French-speaking Switzerland belongs to the market can, however, be left open in that it does not alter the analysis of the effects the notified transaction would have on the markets in question. It is clear, on the other hand, that the geographic definition of the relevant marketing-services market does not include any other countries. First, French books are marketed only on a very small scale in non-French-speaking countries, and in this case it is very often done direct by publishers or via importers (232). Second, in French-speaking countries such as Canada and the countries of the Maghreb or French-speaking Africa, books are marketed exclusively by local structures, with costs that vary appreciably and on markets where conditions of competition are radically different (books in several languages, different types of customer, differences in prices and services, distance from the language area of the Community, scale of local production, etc").

(a) Marketing to bookshops

(327) The Commission takes the view that marketing services for sales to class 1 and class 2 bookshops each constitute a separate relevant market of supranational geographic dimension corresponding to the French-speaking area of the Community in that, for these levels of marketing, there is a certain degree of integration between the marketing structures of the operators working in this area, the only exception being Hachette Livre and VUP. The Commission's investigation has shown that these two groups will not supply Belgian and Luxembourg bookshops direct from France, and force them instead to use their local structures, whereas other French publishers, such as Seuil, Gallimard and Flammarion, supply Belgium from their marketing structures in France. Quite clearly, even though the effect of these constraints imposed by Hachette Livre and VUP is to partition the Belgian and French markets for their products and in this way to enable the two groups to set prices higher in Belgium than in France, this is not sufficient to show that national markets exist, since the other players do not partition the market in this way.

(328) There are no demonstrable economic grounds for this difference in the net of discount prices applying in France, Belgium and Luxembourg. The marketing structures of Hachette Livre and VUP in Belgium and certain independent Belgian marketers sell books published in France to Belgian bookshops at a price net of discount which is 10 to 15% higher than the price in France. Artificial maintenance of these price differences (the "tabelle") (233) and the obligation to order through local structures cannot reasonably be seen as an argument to conclude that separate national markets exist.

(329) On the argument concerning the impact of the Lang Act on marketing services, the Commission's investigation has not produced any evidence of a higher density of bookshops in France than in Belgium. The comparison of the number of customers of Hachette Livre, which has the highest penetration on the two markets and should therefore cover virtually all outlets in the two countries, shows that densities are equivalent in France and Belgium234.

(330) In its reply to the statement of objections the notifying party challenges this presentation of the facts, mainly on the grounds that it is legitimate for Hachette and VUP to maintain structures in Belgium. However, this argument has nothing to do with the definition of relevant market in Community competition law235.

(b) Marketing to hypermarkets

(331) The Commission considers that the market for marketing services for sales to hypermarkets is supranational and corresponds to the French-speaking area of the Community.

(332) Although French marketers such as Seuil or Gallimard subcontract their marketing services for sales to Belgian hypermarkets to local operators, the Commission's investigation has found that marketing services as such are organised and performed in the same way in France and in Belgium. The only service which differs between France and Belgium is not marketing as such but the additional service of rack-jobbing, which in Belgium is demanded by hypermarkets. But as this service is offered by the same companies providing marketing services, the major marketers, such as the ones mentioned above, have felt that in view of the small amount of their sales at that level and the fact that they do not practise rack-jobbing themselves, it was more effective to use local marketers for marketing to hypermarkets. For instance, Union Distribution (Flammarion) markets strip cartoon albums to Belgian hypermarkets direct from France.

(333) Second, even though it is more common in Belgium than in France for hypermarkets to ask marketers to provide a book-labelling service (in particular because application of the tabelle means that a sticker with the Belgian price must be placed over the French price printed on the book cover), this service is also provided for certain French hypermarkets, in particular on pocket-format books where the price is usually determined via a code. In any case specific labelling is necessary in France in order to place the references required by the central purchasing facility on parcels, and French distributors therefore have the resources to label books intended for the Belgian market.

(334) French marketers are therefore able to supply Belgian hypermarkets direct and there is substitutability on the supply side. For this reason the geographic dimension for marketing services for sales to hypermarkets is the French-speaking area of the Community.

(335) The notifying party, in its reply to the statement of objections, considers this to be no more than a future possibility, but provides nothing to challenge the evidence gathered by the Commission during its investigation (236)

(c) Marketing to wholesalers

(336) For the same reasons as are given concerning marketing services for sales to class 1 and class 2 bookshops, the Commission considers that marketing services for sales to wholesalers constitute a supranational market corresponding to the French-speaking area of the Community. At this level the organisation of marketing by the various operators, with the exception of Hachette Livre and VUP, is integrated. On this point, while these two groups, unlike other marketers, will not supply Belgian and Luxembourg wholesalers direct from France, and force them instead to use local structures, these constraints imposed by Hachette Livre and VUP are not sufficient to show that national markets exist, as this division is not applied by the other market operators.

B.2.b. MARKETS FOR DISTRIBUTION SERVICES

(1) THE NOTIFYING PARTY

(337) In its arguments about the geographic definition of markets, the notifying party does not distinguish between marketing services and distribution services (237). For this reason its analysis of marketing services, i.e. that the market is a national one, also applies to distribution services.

(2) THE COMMISSION

(338) The Commission considers that the market for distribution services is of a supranational and corresponds to the French-speaking area of the Community.

(339) In this connection it should be borne in mind that the French-speaking part of Belgium and Luxembourg are far smaller than France in terms of book sales, and at less than 400 km from Paris, which is the natural centre of gravity of French-language publishing, these territories are much closer than many French regions.

(340) A number of distributors use the Prisme transport platform, located in the Paris region, to supply customers both in France and also in Belgium and Luxembourg. Other findings of the investigation suggest a certain convergence in the computerised ordering and restocking systems used for distribution services in these three countries. Lastly, the distributors questioned by the Commission made it clear that the establishment of the single European market, the removal of tariff and non-tariff barriers and the introduction of the euro had made borders ineffective from the point of view of distribution activities in the French-speaking area.

(341) On the question of the national organisation of distribution structures, the investigation has revealed that the strategy of distributors, principally Hachette Livre and VUP, of maintaining a distribution structure in Belgium, while many Belgian dealers obtain supplies of a number of titles direct from France, cannot reasonably be used as an argument in support of the claim that the market for distribution services should be defined geographically as a national market. It should be added that Hachette Livre also has regional distribution platforms in France (Lyon and Nantes) from which its dealers obtain their supplies.

(342) Here too the question of whether French-speaking Switzerland could be considered to belong to the same geographic market can be left open. Even though the investigation has found some evidence of a unified market, there is other evidence of a Swiss market with specific features, as has been explained for the marketing-services markets. At all events taking Switzerland into account would not change the competition analysis based on the French-speaking area of the Community. On the other hand there can be no question that the geographic definition of the relevant market for distribution services should include other countries, if only for the high transport costs to French-speaking countries such as Canada, the Maghreb and French-speaking countries in Africa and the differences in conditions of competition, as has been explained in the case of marketing.

B.3. MARKETS FOR THE SALE OF BOOKS

B.3.a. MARKETS FOR THE SALE OF BOOKS BY PUBLISHERS TO DEALERS

(1) GENERAL ARGUMENTS

(a) The notifying party

(343) The notifying party has put forward a number of arguments which, it feels, point to markets for the sale of books by publishers to final consumers being defined as national markets (238).

(344) First of all it takes the view that as the legislative and regulatory framework differs from one country to the next (Lang Act on single book prices in France, and also legislation on prior declarations, censorship and protection of minors), it affects the economic features of books differently (239) in these countries. As cultural goods, books are often given special attention by the public authorities, and in some cases are even subject to specific rules intended to maintain the diversity of supply and guarantee distribution throughout the national territory.

(345) Second, the notifying party maintains that there are price differences between France and neighbouring French-speaking countries at the level of retail sales (240). These differences are accounted for by extra distribution costs in Belgium and Luxembourg justifying a price to the public that is [5-15]% higher (241) for Hachette and as much as [10-20]% for VUP (the "tabelle"). As there is no law imposing single book prices in Belgium and Luxembourg, Belgian and Luxembourg dealers can offer discounts, while French dealers cannot. The notifying party also justifies price differences by the different VAT rates (6% in Belgium, 5.5% in France and 3% in Luxembourg). In addition, it is claimed that there are no significant cross-border flows, and in particular no arbitrage, despite the existence of the tabelle, either by dealers or by consumers.

(346) Third, access to the domestic media in each of the countries concerned is required in order to promote books (242).

(347) Fourth, it is claimed that there are differences in the structure of supply and demand between France and Belgium (243). Differences in the structure of supply between France and Belgium are the result of the strength of domestic publishers, differences in the development of market segments and the breakdown of turnover and the unavoidably national nature of certain works. For final consumers, the differences concern the categories of works purchased, the lack of correspondence between best-sellers in Belgium and in France and the existence of different catalogues for France Loisirs and Belgique Loisirs. In the case of dealers, there are also significant differences in their structure.

(348) Fifth, the notifying party relies on the Commission's practice and the case law of Community courts on transnational agreements on fixed book pricing as regards marketing of books in the cross-border homogeneous language areas of Belgium and the Netherlands (244), (Dutch-language books), the United Kingdom and Ireland (245) (English-language books) and Germany and Austria (German-language books) (246). It is claimed that these agreements were designed to bring about an artificial standardisation of cross-border market conditions and their implementation demonstrates that although belonging to the same language area, each Member State constituted a geographic market in itself. It is further claimed that the Commission and the Court of Justice of the European Communities have refused to consider that the language areas in question each constituted one and the same geographic market (247).

(349) Sixth, there are significant variations in market shares from one country to another and significant historical and cultural differences between countries (248).

(b) The Commission

(350) The Commission disagrees with the notifying party and considers that the geographic dimension of markets for the sale of books by publishers to dealers is supranational, corresponding to the entire French-speaking area of the Community, and even of the whole of Europe (on the question of whether the French-speaking part of Switzerland belongs to the French-language market, see above), for the following reasons.

(351) To begin with, it should be pointed out that some of the arguments of the notifying party are only indirectly relevant as they concern final consumers (249).

(352) First, this is the case with the argument relating to the fact that the Lang Act standardises the retail sale price for books throughout France. Although this Act, which is specific to France, prevents price competition between dealers in retail sales prices, it does not in any way rule out competition in sales of books by publishers to dealers. While it lays down rules governing relations between publishers and dealers, this is essentially to ensure that there is no discrimination in terms of sale to dealers (250) and not to set selling prices between these two players. As regards markets for the sale of books by publishers to dealers, the point to be considered is whether the Act influences the level of discounts (and not selling prices to final consumers) between France, Belgium and Luxembourg.

(353) On this, it has been found that discounts are calculated on a comparable basis in these three countries (251), by reference to similar elements such as the quantity criterion of turnover and the quality criterion of participation in promotional operations. The Commission has also had a comparative study conducted into the levels of discounts applied by the notifying party to dealers in France, Belgium and Luxembourg (252), designed to examine whether the method of determining discounts was different between France, Belgium and Luxembourg. The econometric model therefore sought to evaluate the mechanism which is used for determining discounts, on the basis of quantities sold by the dealer and the quality of the service provided. A country variable was included in the regression to see whether the location of the dealer in France, Belgium or Luxembourg affected the determination of the discount. The results of the study show in particular that no "country" effect is observed for general literature titles, whatever the type of dealer (253). This analysis shows that the level of the discount is unrelated to the location of the dealer in the case of discounts for sales of general literature titles.

(354) It is also the case with the argument relating to significant price differences which, it is claimed, exist between France and Belgium. At all events, the analysis of price differences supplied by the notifying party concern a narrow selection of titles and dealers and is not necessarily representative. The Commission's investigation has shown that retail prices in Belgium are strongly influenced by the prices in France. In this connection, the introduction of the euro has given Belgian final consumers a clearer view of prices charged in France (254). As a result, the policy of certain dealers in Belgium, especially hypermarkets and some booksellers such as FNAC, is to align their prices on the list prices charged in France, thus absorbing the "tabelle" effect.

(355) Finally, it applies to the argument on household consumption patterns, which relates to final consumer demand and not demand from dealers.

(356) Second, as regards the notifying party's claim that best-sellers in Belgium are not always best-sellers in France, this is based on a study by the notifying party covering only specific genres, namely novels and essays, and is therefore not conclusive evidence. As the notifying party quite rightly points out (255), it is not relevant to make a distinction, within general literature, between books in the different genres, where they are written by the same category of authors, are produced within the same publishing houses, have similar production costs and meet the same needs for readers. It is therefore even less relevant to highlight within these genres one or other title which has more success in one Member State than in the neighbouring Member State. What it is more important to note is that fundamentally, French and French-speaking Belgian readers tend to read books from the same publishers (most of them French) (256).

(357) Third, on the analysis of the structure of demand, the Commission's past practice (257), as confirmed by the notifying party (258), leads to the conclusion that defining publishing markets from the point of view of demand does not give a satisfactory picture of the conditions of competition in these markets. At all events, it should be pointed out that the difference between France, Belgium and Luxembourg in terms of structures for supplying dealers is essentially the result of they way in which the parties to the transaction are themselves organised, in that Hachette Livre and VUP require dealers in Belgium to obtain supplies of the books they sell exclusively from their Belgian subsidiaries. This is not the case for other French publishers, in particular Gallimard, Seuil and Flammarion, which supply their Belgian customers direct from Paris, without imposing the "tabelle" (259). As a result, it is the internal organisational structures of the parties to the transaction and not trade flows or any economic logic specific to the sector which artificially place a partial and relative partition between France, Belgium and Luxembourg (260). The fact that there is no substantial cross-border supply of dealers, if it was confirmed, is not therefore an indication that separate geographic markets exist.

(358) Fourth, as regards the analysis of the structure of supply in markets for sales of books to dealers, the first point to make is the size of the cross-border flows between France, Belgium and Luxembourg. Nearly 70% of French-language books purchased in Belgium come from France and nearly 60% of French-language books published in Belgium are sold in France and Luxembourg (261). This is clear for general literature, in that Belgian output is, in terms of quantity, very limited and French publishers have a very strong position in this country and publish the best-selling Belgian authors. With certain exceptions (school textbooks, law books, some books on purely national politics) French publishers do not specifically produce books for the Belgian market, while Belgian publishers, with some exceptions, produce books for a worldwide French-speaking market and not specifically for French-speaking Belgians. So while some Belgian publishers specialise in strip cartoons or the social sciences (262), they do not sell exclusively in Belgium but throughout all French-speaking areas. All this indicates that on the supply side, there is a large degree of substitutability between France, Belgium and Luxembourg. A distinction should none the less be made between different product markets, in particular that for the sale of school textbooks by publishers to dealers and that for the sale of law books by publishers to dealers, where supply is national.

(359) Fifth, contrary to what the notifying party claims, access to national media does not appear to be a differentiating factor between France, Belgium and Luxembourg. On the one hand, the vast majority of major French daily newspapers and magazines are available and sold by many newsagents in the French-speaking part of Belgium and in Luxembourg. What is more, the books columns of Belgian newspapers essentially cover works of French publishers and not those of Belgian publishers. And on the other hand, most French television channels, and therefore their literary programmes, are also broadcast in Belgium and Luxembourg. There is clearly, therefore, no need for French publishers to have access to the strictly national media in order to secure effective media coverage for their works. Similarly, the fact that in Belgium, unlike in France, books can be advertised on television does not have a decisive influence on demand from Belgian consumers, in that such television advertising in Belgium does not appear to be very widespread (263).

(360) Sixth, according to the strategic plans of the market players themselves [...]* (264).

(361) Seventh, with reference to the argument based on Commission practice and the case law of the Court of Justice on fixed book pricing, it should first be pointed out that in none of the cases quoted by the notifying party did the Commission or the Community courts demarcate, in a technical sense, the product and geographic markets in the book sector (265). The passages identified in the Court's judgment in VBVB and VBBB concern only the question of whether the price-fixing agreements affect "trade between Member States" within the meaning of Article 81 (1) of the Treaty and do not therefore permit any conclusions on the demarcation of the markets in question (266). Similarly, the reference to the judgment of the Court of First Instance in Publishers Association is not relevant either. The Court's reasoning, on which the notifying party relies, although it quotes it out of context, is not based on a market demarcation analysis but is designed to reject the applicant's erroneous argument that a cross-border fixed-pricing agreement should enjoy exemption under Article 81 (3) of the Treaty because confining its application to a "national market" would cause the established system to collapse (267). For these reasons, the notifying party's arguments based on case law do not bear out the existence of separate national markets. On the contrary, the above-mentioned cases show that the Commission has considered cross-border conditions of competition for marketing books in the same language to be sufficiently homogeneous and effective to conclude that there is a breach of Article 81 (1) of the Treaty.

(362) Eighth, with reference to the notifying party's argument that the differences in the combined market shares of the parties to the transaction in France, Belgium and Luxembourg could be evidence of separate geographic markets, it should be pointed out that this factor is not sufficient by itself to define the geographic market dimension; reference should be made to the Commission's competition analysis of the different markets for the sale of books by publishers to dealers, which reveals that for sales of general literature titles (268), these differences are not great. They are, however, more significant for school textbooks and law books and for works in instalments.

(363) Last, it should be noted that barriers to entry into Belgium, France and Luxembourg are very low, except for school textbooks and law books, which have very specific national characteristics.

(364) Some of the above criteria in support of the argument that markets for the sale of books by publishers to dealers are supranational as regards France, Belgium and Luxembourg also apply to the French-speaking part of Switzerland. In particular, it is clear that even though there is no legislation in Switzerland imposing a single retail price for books, publishers' prices to dealers (i.e. discounts given to dealers) are calculated in the same way as in the French-speaking countries belonging to the Community, namely on the basis of the French list price exclusive of tax. Similarly, there are large cross-border flows in retail sales (and to a lesser extent between booksellers) between the French-speaking part of Switzerland, in particular the Geneva canton, and the neighbouring parts of France (269).

(365) A number of features differentiate the book sales market in the French-speaking part of Switzerland from that of the French-speaking area of the Community. First of all, prices are significantly higher in Switzerland. All the main French-speaking publishers operating in that country apply a "tabelle", the effect of which is that retail prices are between 20% to 25% and 40% to 50% higher than the prices for the same books sold in France (or in Belgium, in the case of certain Belgian publishers). This price difference partly reflects higher costs in Switzerland, one reason being higher salaries, but it also takes account of the greater purchasing power in Switzerland. Another difference in relation to the French-speaking countries in the Community is that the existence of the Swiss franc as the national currency, with a rate that varies against the euro, means that price relationships between Switzerland and the Member States can vary, and that for final consumers the price differential is less obvious than between, say, France and Belgium. In addition, even though there are no customs duties between the Community and Switzerland, imports of books to Switzerland involve higher costs in terms of additional carriage and transit costs (270).

(366) However, the question of whether or not French-speaking Switzerland belongs to the French-speaking area of Europe can be left open, in that it does not alter the analysis of the effects the notified transaction would have on the markets in question.

(367) In the light of the above, it can be seen that in general the markets for the sale of books by publishers to dealers, in particular works of general literature, guides and manuals, children's books and reference works, are supranational, covering at least the French-speaking area of the Community, with the possible inclusion of the French-speaking area of Switzerland. There are, however, certain specific features for school textbooks and educational supporting materials which will be analysed below.

(2) SPECIFIC MARKETS

(a) Markets for the sale of school textbooks by publishers to dealers

(i) The notifying party

(368) The notifying party claims that the geographic market for the sale of school textbooks by publishers to dealers is a national one, mainly because (in addition to the general criteria discussed above) works are prescribed on the basis of national curricula laid down by the ministry of education or equivalent authority in each country, and orders for school textbooks have specific features (timetables, circuits, etc").

(369) Second, in the French market, unlike the Belgian and Luxembourg markets, it is claimed that there are virtually no foreign publishers supplying school textbooks (271).

(ii) The Commission

(370) The Commission, like the notifying party, considers the geographic market for the sale of school textbooks by publishers to dealers to be a national one.

(371) Contrary to the other categories of books, such as works of general literature, it can be seen that school textbooks sold in France are not the same as those sold in Belgium. This is essentially due to the fact that school syllabuses are strictly national and impose a constraint on supply.

(372) It is also found that Belgian publishers of school textbooks are not present in France and French publishers of school textbooks have only a limited presence in Belgium for sales of school textbooks. This results in very significant differences in market shares between France and Belgium.

(373) The operation of the Belgian and Luxembourg markets for school textbooks differs greatly from that of the French market. In the French-speaking part of Belgium, for instance, the use of school textbooks is much less systematic than in France. There are two education networks which each have their own syllabuses and different rules. There are no public invitations to tender and teachers have a free choice as to which textbooks they use. The French practice of sending specimen copies of new textbooks to teachers (272) does not apply there either. In Luxembourg it is the Ministry which sets syllabuses and chooses the textbooks to be used in all classrooms in the country on the basis of existing textbooks offered by Belgian or French publishers (273).

(374) The Commission has also had a study carried out to compare the discounts applied by Hachette and VUP to dealers in France, Belgium and Luxembourg for school Textbooks (274), the purpose being to see whether there was any difference in the way discounts were determined in the three countries. The findings show that a "France" effect can be seen for school textbooks for the various classes of dealers. This analysis therefore shows that for school textbooks the size of the discount depends on the location of the dealer. The economic experts of the Lagardère group did not challenge this finding, but maintained that a snapshot comparison of price mechanisms was of only limited relevance for the definition of geographic markets.

(375) Finally, there are major barriers to entry to the French market in view of the costs involved in drafting a textbook, sending specimen copies to teachers, compiling lists of teachers, acquiring and maintaining a brand name with a long-standing reputation and being able to produce a number of books in the same series in response to the needs of syllabuses.

(b) Markets for the sale of educational supporting materials by publishers to dealers

(i) The notifying party

(376) The notifying party claims that the geographic market for sales of educational supporting materials is a national market mainly because (in addition to the general criteria discussed above) the content of educational supporting materials is closely linked to the syllabuses set by the French Education Ministry and the equivalent authorities in each of the countries concerned. It is claimed that the arguments presented to show that the markets for the sale of school textbooks by publishers are national markets apply to educational supporting materials.

(ii) The Commission

(377) The Commission maintains that the geographic dimension of the market for the sale of educational supporting materials by publishers to dealers is that of a supranational market covering the entire French-speaking area of the Community (with the possible inclusion of the French-speaking part of Switzerland).

(378) The market for the sale of educational supporting materials does have some national features which it shares with the (national) market for the sale of school textbooks. However, compared with the other features of this market for the sale of educational support materials which are shared by the other (supranational) markets for the sale of books by publishers to dealers, these national features do not carry sufficient weight to justify defining the geographic market as a national one. The national features of the market for the sale of educational supporting materials are much less marked than those for the sale of school textbooks.

(379) For instance, while there are differences in market shares between countries, these are far smaller than can be found on the market for the sale of school textbooks and can be compared to those existing on markets for the sale of other categories of books.

(380) These market share also show that French publishers of educational supporting materials have a very strong presence in Belgium and Luxembourg, where they sell the same materials as in France. In addition to French grammar books and language-learning aids, which do not belong to any specific school curriculum, it has been found that holiday workbooks sold in France and Belgium are the same, even though originally they are linked to the French national school curricula. What is more, in order to be able to sell their holiday workbooks more easily in Belgium, it is observed that French publishers have simply added on the dust jackets of all their books a reference to the age of the children concerned in addition to the reference to the French class grade.

(381) In addition, as was explained above, marketing methods for school textbooks differ from those for other categories of books depending on the way in which textbooks are prescribed in each country. This is not the case as regards educational supporting materials, for which marketing methods throughout the French-speaking area are similar to those for books in other categories.

(382) While a link may exist between certain educational supporting materials and the school curricula set by the authorities, this does not prevent the same works being offered by French-speaking publishers of educational supporting materials in Belgium and Luxembourg as well as in France. There is therefore a high degree of substitutability on the supply side between the French market and the Belgian/Luxembourg market for the sale of educational supporting materials. This is confirmed by the fact that all the educational supporting materials published by Hachette Livre and VUP are in principle available and marketed in Belgium (275).

(383) This finding is not contradicted by the analysis of the size of discounts to dealers. The Commission has had a study carried out to compare the discounts applied by the notifying party to dealers in France, Belgium and Luxembourg for educational supporting materials (276), the purpose being to see whether there was any difference in the way discounts were determined in the three countries. The findings show that a "France" effect for educational supporting materials can be seen only on class 1 bookshops, and not on other classes of dealers, unlike the situation concerning school textbooks, where the "France" effect is visible on all classes of dealers.

(384) The other reasons why the Commission considers the markets for the sale of books by editors to dealers to be supranational apply entirely to educational supporting materials, in particular the arguments deriving from the unilateral imposition of the "tabelle" by Hachette Livre and VUP, the existence of large cross-border flows (277) and coverage by the French media.

(385) In the light of the above, it can be seen that the market for the sale of educational supporting materials by publishers to dealers is a supranational market covering at least the French-speaking area of the Community, with the possible inclusion of the French-speaking area of Switzerland.

(c) Markets for the sale of academic and professional works by publishers to dealers

(i) Markets for the sale of law books by publishers to dealers

(386) The notifying party claims that the geographic dimension of the market for the sale of law books by publishers to dealers is national, mainly because (apart from the general reasons set out above) the content of these works varies considerably from one country to the next. The notifying party argues that each country lays down its own laws and regulations and so students and practitioners therefore study and apply principally this national body of law. The notifying party further considers that even on international legal subjects, students and practitioners would naturally prefer the works of authors well-known in their own country, or even in their university of origin (278).

(387) The Commission's investigation has confirmed some of the arguments put forward by the parties and hence the fact that the geographic dimension of the markets for the sale of law books by publishers to dealers is national.

(ii) Markets for the sale of science, economics and social science works by publishers to dealers

(388) The notifying party argues, on the basis of the reasons set out above, that the geographic dimension of the markets for the sale of science, economics and social science works by publishers to dealers is national.

(389) But however the markets for these products are defined, the transaction is not likely to create or strengthen a dominant position which would result in effective competition being significantly hampered, so that the question of the precise definition of the markets for the sale of science, economics and social science works by publishers to dealers can remain open.

B.3.b. MARKETS FOR THE SALE OF BOOKS BY WHOLESALERS TO CLASS 3 DEALERS

(390) The notifying party claims that the geographic markets for the sale of books by wholesalers to retailers are national markets, as (i) wholesalers are different legal entities in the different Member States, (ii) they have limited geographic cover and

(iii) some publishers are distributed by different wholesalers in different Member States.

(391) The first argument concerning the existence of different legal entities is not relevant in that it flows more from internal organisation policy. As regards the limited geographic cover of wholesalers, it should be pointed out that in France only the parties to the merger (279) cover the entire national territory, other wholesalers being regional or even local, while in Belgium the activity of independent wholesalers can be regarded as regional if the French-speaking part of Belgium, as explained above, is considered to operate in the same way as a French region.

(392) The salesrooms of Hachette Livre and of VUP are in practice a network of regional salesrooms enabling the two groups to cover the entire French-speaking area. In this respect the regional coverage of the independent wholesalers is not fundamentally different from that of the merged entity in that DNL, an independent wholesaler specialising in sales to supermarkets, also operates on the basis of two regional platforms. In addition, the rack-jobbing function as performed by a wholesaler does not vary with the geographic location of customers and so can be considered homogeneous over the entire French-speaking area of the Community.

(393) Even if a large proportion of independent wholesalers generally operate on a regional basis, the areas they cover overlap, and are not confined within national borders. On this point the Commission's investigation has shown that a wholesaler established in Belgium could operate in the north of France and Luxembourg. And then the notifying party itself has clearly stated (280) that the areas of action of the various wholesalers exist side-by-side. Regional wholesalers are therefore generally competing with LDS and La DIL in that the wholesale market is made up of overlapping geographic areas where competition is uniform.

(394) As regards prices charged by wholesalers, which are based on the differential between the discount they receive from publishers (around 35%) and the discount they give their class 3 customers (20 to 30%), these too do not vary significantly over the French-speaking area of the Community, but depend more, within a uniform range, on factors such as the size of the sales outlet, participation in promotion campaigns and performance in terms of turnover.

(395) In the light of the above, the market for the sale of books by wholesalers is a supranational market corresponding to the French-speaking area of Europe.

B.3.c. MARKETS FOR THE SALE OF BOOKS TO FINAL CONSUMERS

(1) MARKETS FOR THE SALE OF LARGER REFERENCE WORKS BY SALES AGENTS

(396) The notifying party argues that the market for the sale of larger reference works by sales agents is a national market (281).

(397) As sales through sales agents constitute direct sales to final consumers, and the regulations governing consumer credit differ from one Member State to the next, publishers use different distributors in different countries. This is the case in particular of Hachette Livre and VUP (282).

(398) The market for sales through sales agents is therefore a national one.

(2) MARKETS FOR THE SALE OF BOOKS BY DEALERS TO FINAL CONSUMERS

(399) The notifying party claims that the geographic dimension of the markets for retail sales of books is national (283). The notifying party has not developed the reasons which in its opinion justify this definition of the market.

(400) The Commission has on a number of occasions already commented on the demarcation of the relevant geographic markets in the retail trade (284). It has indicated that for food retailing and DIY stores (285) the geographic markets are local markets or catchment areas in which the parties operate their sales outlets. These catchment areas vary in size depending on the size of the outlets, their location (close to major roads, presence or not of an attractive shopping centre nearby) or the consumption habits of the households concerned. It will be observed that certain sales outlets operated by the parties are to be found in airports (286), areas where the customers are captive. It cannot therefore be ruled out, even from the point of view of the final consumer alone, that the catchment areas are confined to these same geographically demarcated areas. For airports the Commission has already pointed out that the largest airports, in particular Paris, Lyon and Toulouse, constituted substantial parts of the common market (287) given the annual volume of traffic, both passengers and freight.

(401) In addition, as the Commission has already pointed out in connection with specialised distribution or food distribution (288), certain aspects of the competition analysis can be considered in a wider context than the local level. Some major factors of competition between distribution chains can be assessed at national level, as the chains take centralised decisions on the structure of their product listings, conduct promotion campaigns at national level as part of their cooperation with suppliers, and apply customer loyalty policies at national level. Store siting policies are in most cases also determined by each State.

(402) The Commission's investigation has confirmed that these factors must be taken into consideration in the assessment of the position of the various retailers.

(403) In the light of the above, there are a number of factors pointing to the existence of local markets, even though the competition analysis cannot ignore the strategies and positions of the different market players at national level.

SECTION VI: COMPETITION ANALYSIS

I. RELATIONS BETWEEN THE PUBLISHING HOUSES OF A SINGLE GROUP

A. NOTIFYING PARTY

(404) In the competition analysis of the market power of the Hachette and VUP groups, the notifying party argues that a "conventional" analysis of concentrative effects is not appropriate in this case (289). Account should be taken not of the two groups' aggregated market shares on the various relevant markets, but of the individual shares of the publishing houses of which the groups are made up. Given the peculiarities of the publishing sector, the best way of trying to assess the impact of the concentration, according to the notifying party, is therefore to analyse the market shares publisher by publisher, without taking account of whether those publishers belong to a publishing group.

(405) The notifying party claims on a number of grounds that there is internal competition between the various houses within Hachette Livre and VUP, which makes it impossible to aggregate the market power of the publishing houses and thus precludes the existence of group market power. This intra-group competition applies to the purchase of publishing rights, the recruitment of authors from other houses in the group, content (subjects), media products (introduction of competing series), and literary prizes. It is reflected in the lack of a centralised publishing policy, budget and management of financial risk within the group (290). Lastly, price setting is not coordinated between the various publishing houses within Hachette Livre or between those of VUP and clearly cannot be the subject therefore of any coordination whatsoever between Hachette Livre and VUP (291) after the merger.

B. COMMISSION

(406) The approach proposed by the notifying party is tantamount to calling into question the very concepts of group (292) and control (293) and, hence, the Commission's established position on these issues.

(407) It is clear, moreover, that the notifying party's submission does not tally with the way publishing groups actually operate, that it accentuates the sector's peculiarities, and that it refuses to acknowledge the way competition operates in reality in the relevant markets.

(408) The Commission's investigation shows first of all that, even if the different houses of the publishing groups enjoy a degree of autonomy as regards their canvassing and launching activities (294), they do not operate as fully autonomous entities. Thus the subsidiaries of the Hachette Livre, VUP and other publishing groups operate on a budget ratified and controlled by the group's central management, and in practice the strategic decisions affecting these houses, such as mergers between houses or a change of managers, are generally taken by their group's head office (295).

(409) Second, a set of essential functions (marketing, distribution, accounts, management audit, production, IT, human resources, legal department, etc") are usually shared between the publishing houses of a single group, making it possible to reduce costs and increase efficiency within the group. Similarly, the general conditions for selling to dealers are normally laid down for the group as a whole and not for each of the publishing houses individually. The sharing of these functions by different publishing houses facilitates the exchange of information and the coordination of management policies, and thus contributes to the limitation of competition, or lack of competition, between a group's publishing houses.

(410) Lastly, although certain publishing houses operate on the same markets, the works they market are often commercially positioned in different segments. This editorial sensitivity on the part of publishing houses enables the group to offer dealers enough differentiated products and brands to satisfy much of their requirements with group products while maintaining a semblance of choice. The policy of a controlled increase in the range on offer limits the stimulation of competition outside the group, increases the chances of selecting a group product to the detriment of other suppliers and maximises the possibility of generating profits at group level.

(411) In its reply to the statement of objections, the notifying party disputes the facts set out above: it maintains that the autonomy of the Hachette Livre and VUP publishing houses cannot reasonably be challenged, that the centralisation of certain administrative functions is not sufficient to question this conclusion, and that there is no group commercial policy. It considers that this independence is the corollary of the special features of the publishing sector.

(412) The Commission recognises that, to a certain extent, editorial activity, which generates cultural goods, requires appropriate company structures which are capable of stimulating author creativity and maintaining the personality of the brands within the groups. In many respects, however, these particularities are not specific to the publishing sector, and the autonomy in question has very few points in common with that which exists on a competitive market between independent firms. The large majority of groups manufacturing mass consumption products have, to a certain extent, adopted a company structure that operates by brands or product groups. For instance, the L'Oréal group, which markets the shampoo brands Garnier and Elsève, preserves a certain margin of autonomy for these two entities. There is no doubt, however, that while the Garnier and Elsève entities deal with consumers with comparable characteristics and, with their respective products, meet similar needs and could therefore give the impression of being in direct competition, they are still coordinated by L'Oréal, in particular as regards commercial policy (296). The factors mentioned by the notifying party in its reply to the statement of objections follow a similar logic when it is stated that authors recognise the personality and editorial line of the publishing houses within groups. In no way does this prevent the different brands being managed and coordinated by the group, so that the relations between the different houses are quite different from those that apply between publishing houses of competing groups. (413) The Commission's investigation also confirms that the publishing houses are subject to central control. As the La Martinière group explains, for instance, .if there is a problem to sort out, it is for the group chief executive to decide,. which would be hard to imagine if competition were not coordinated. Similarly, in the strategy documents of the Hachette Livre group, the focus is clearly that of a group and not a collection of publishing houses acting without centralised control. The Commission has obtained internal operational documents from Hachette Livre and VUP which clearly show that a group logic applies (297). This situation is confirmed by the dealers, which explained that the general terms of business of the Hachette Livre group applied generally to the different publishing houses in the group (298). (414) For all the above reasons, therefore, the notifying party's argument that the competitive effects of the merger should be analysed publisher by publisher should not be accepted. The analysis of the threat to create or strengthen a dominant position will therefore be conducted, as the French authorities did in their request for a referral, at publishing group level, and thus on the basis of the market power of the Hachette Livre and VUP groups. II. EFFECTS (415) It should be explained by way of introduction that according to the case law of the Court of Justice and the Court of First Instance a dominant position is defined as a position of economic strength enjoyed by an undertaking which enables it to prevent effective competition being maintained on the relevant market by affording it the power to behave to an appreciable extent independently of its competitors, its customers and ultimately of the consumers. The existence of a dominant position may derive from several factors which, taken separately, are not necessarily decisive. Among those factors, the existence of very large market shares is highly important. Nevertheless, a substantial market share as evidence of the existence of a dominant position is not a constant factor. Its importance varies from market to market according to the structure of those markets, especially so far as production, supply and demand are concerned. In addition, the relationship between the market shares of the undertakings involved in the concentration and their competitors, especially those of the next largest, is relevant evidence of the existence of a dominant position. That factor enables the competitive strength of the competitors of the undertaking in question to be assessed (299).

(416) Apart from these horizontal elements, it is the Commission's task to examine the vertical and conglomerate effects of the notified transaction.

(417) It should be explained that the two groups concerned by the transaction are already today the two undisputed leaders of French-language publishing. They are present on a large number of markets and throughout the book chain, in particular as marketers and distributors for a large number of publishers, thus strengthening their power vis- à-vis dealers. The Lagardère group is also a conglomerate that is present on other media markets in particular.

(418) Following the transaction, the merged entity would, in the book field alone, be seven times larger than Gallimard, its nearest rival. It would have a very extensive portfolio of brands and well-known authors and would be present in all book categories, unlike its nearest rivals. Lastly, it would be present at all stages of the book chain, even in retailing. Its position and market power would therefore be unrivalled by its nearest competitors.

A. MARKETS FOR PUBLISHING RIGHTS

A.1. PRIMARY RIGHTS MARKETS

(419) As indicated in the section on the relevant product markets, the primary rights markets should be subdivided by major categories of book. The different primary markets for publishing rights should be subdivided into a market for French rights and a market for foreign rights.

A.1.a. THE PRIMARY MARKET FOR RIGHTS IN GENERAL LITERATURE TITLES

(1) INTRODUCTION: METHOD OF CALCULATING MARKET SHARES

(a) Method supplied by the notifying party

(420) To evaluate the strength of the different publishers as regards purchasing the rights in general literature titles, the notifying party supplied a study of the rankings (in volume terms) of the various publishers in the best-seller lists for general literature titles by recurrent, well-known authors (300). The notifying party uses the concept of "recurrent authors", since it considers that, as well as works by new authors or occasional works, there are authors producing work that is anticipated and often programmed. To purchase the rights to such work publishers compete more directly than in the case of authors of lesser reputation or ones who have had only one literary success. Some well-known authors are recurrent writers who write books regularly and who can even be described as commercial brands in their own right (301); publishers offer them large advances. Although the notifying party's calculations relate only to best-sellers in France and not the world, which is the relevant geographic dimension of the markets analysed, France occupies a very important place in the markets for French-language publishing rights compared with the other French-speaking countries and can therefore be regarded as representative of the market as a whole.

(421) The Commission considers, however, that the importance of the different publishers as regards the primary purchase of the rights in general literature titles cannot be evaluated on the basis of the volume of the sales of these authors' books. The publishing houses do not compete for the acquisition of the publishing rights on the basis of the volume sold but of the price offered to the authors in order to purchase such rights, i.e. on the advances (and the proportional remuneration) which are offered to them by the publishing houses. Competition occurs between publishers interested in acquiring publishing rights when the contract is concluded, and not when the book is marketed. The point at which the rights are purchased is in fact the point when the publisher chooses to take a risk and evaluates the anticipated sales of the book (which will not necessarily be achieved). There is not necessarily a correlation between the volume of sales achieved and the amount of remuneration offered to the author (302).

(422) By using the concept of "recurrent authors", moreover, the notifying party excludes from the market authors who are a significant part of publishers' business but who have less countervailing negotiating power than recurrent authors, since they have not necessarily already enjoyed a literary success.

(423) Further, in order to separate French rights from foreign rights, the study supplied by the notifying party distinguishes authors by nationality, irrespective of the language the book is originally written in. Thus some authors who write their books originally in French and whose works do not have to be translated before being exploited in the French language are classified as foreign authors303. According to the information supplied by the notifying party, 8%304 of French-language "recurrent" authors have a nationality other than French and are part of the French rights market305. The study supplied by the notifying party does not reflect the reality of the market therefore.

(b) Methods used by the Commission in its statement of objections

(424) Thus, to evaluate the importance of the different publishers as regards the primary purchase of the rights in general literature titles appropriately and realistically, the Commission asked the principal players in the market to supply, for each of their publishing houses, the annual amount paid in advances for books. It has proved impossible, however, to aggregate the data supplied by these players, given the different methods of accounting used by the various publishing houses to annualise the spread payments of advances.

(425) To overcome this technical difficulty, the Commission has therefore used another tool for evaluating the importance of the different publishers as regards the primary purchase of the rights in general literature titles. It has collected the amount of advances paid by the different publishing houses to the authors of the 100 best-selling non-translated general literature titles and to the authors of the 100 best-selling translated general literature titles in France in 2002 (306) ("first method"). Although this aggregate relates only to the 100 best-sellers in France and not the world - the relevant geographic dimension of the market analysed - France occupies a very important place in the markets for French-language publishing rights compared with the other French-speaking countries and can therefore be regarded as representative of the market as a whole (307). This method makes it possible to catch all sales of more than 21 000 copies for non-translated titles and more than 12 000 titles for translated titles and is not confined therefore to books which may be regarded as best-sellers (308).

(426) The method assesses market shares for just a single year, 2002, but the Commission considers that this year is representative of the sizes of the players in the market, since the annual amount of advances paid by publishers is broadly stable (309).

(427) Moreover, while this method relates only to advances offered by publishers and not to the amount of proportional remuneration, this simply serves to show the real market power of the players involved, since the proportional remuneration is paid on the basis of the book's actual sales, which publishers cannot know when the rights are purchased. It is true that the percentage (and not the absolute amount actually paid as proportional remuneration) might also be an indicator of the market power of the publishers and the price they are prepared to offer in order to obtain the rights. However, taking this indicator into account would not substantially alter the results of the method adopted by the Commission, since the percentages for the proportional remuneration granted to authors are fairly homogeneous (310).

(428) To use the notifying party's concept of "recurrent authors", the Commission also checked, among the 100 best-sellers, what the books written by such authors (311) were, even though it does not regard this method as appropriate, because it is too simplistic ("second method"). This check shows that 46 books among the 100 best-sellers were written by recurrent, well-known authors.

(c) Reply of the notifying party to the statement of objections

(429) In its reply to the statement of objections, the notifying party considers that the Commission's analysis uses a method of calculating market share which is very doubtful, since it is based only on the amount of advances.

(430) The Commission, it claims, describes the role of advances in the market for the acquisition of primary rights erroneously (312), since (i) advances take no account of the assignment of authors. royalties (they are not laid down in law); (ii) they do not constitute the purchase price, but are a special method of payment of that price; (iii) they are not a factor in authors. decisions and cannot therefore be a relevant consideration in the analysis of competition in rights markets (313) (authors do make the distinction, and are interested in the proportional remuneration but hardly at all in the advance); and (iv) the information contained in the advances cannot provide a statistically reliable indicator because the data are skewed by sources of heterogeneity.

(431) The notifying party also considers that the Commission's theory that publishers are taking a risk when paying an advance is not supported by the facts (314), since the advance is nearly always a straightforward advance which is offset by deducting an equivalent amount from the proportional royalties.

(d) Commission's conclusions

(432) The Commission's investigation shows however that advances play a decisive role in the acquisition of primary rights.

(433) From a legal, contractual and factual point of view, these advances constitute payments, which are made in consideration for the sale of rights. An advance is a payment for the transfer of intellectual property rights, which is a guaranteed minimum and not a method of payment (315). Thus the advance systematically constitutes a guaranteed minimum acquired definitively by the author, even if the revenues from the exploitation of the work should be less than anticipated. From this point of view, the advance is the only payment which the author is sure to receive. It is not therefore just a method of payment.

(434) The advance is a selection criterion for authors, since more than 70% of general literature authors cited remuneration as a criterion for selecting a publishing house (316). In addition, Dominique Goust, Director for Le Livre de Poche in Hachette Livre, himself explains that "publishing is a noble profession, which behind the scenes is like football. Today, authors change publishers like Zidane changes clubs, for money, prestige or reasons of affinity" (317) and therefore considers money to be one of authors' selection criteria. The Commission's investigation has shown that, according to the publishers of general literature, authors choose their publishing house as much for the remuneration proposed as for the personal link between author and publisher (318).

(435) Moreover, contrary to what the notifying party maintains, advances are not intended to cover expenses already incurred by the author (319). Where an author incurs research expenses from writing a work, the coverage or reimbursement of those expenses by the publisher are the subject of a special clause, which is independent of the payment of an advance. Nor are they consideration for a right of first refusal, since advances are provided for each of the works covered by a right of first refusal.

(436) The Commission's investigation also shows that publishers take a risk when paying an advance. As the notifying party itself explains, "since publishing, as already stated, is a risky profession by definition, Hachette Livre acknowledges that a publishing house can make losses following editorial choices that are unfortunate or not immediately recognised by the public" (320) and "the financial risk relates basically to the advance paid to the author", the promotional budget and the number of copies it is decided to print for first publication and, if successful, subsequent printings, and all this without any guarantee of future revenues. The risks associated with each work are taken into account in the provisional operating accounts. The financial risk does not in principle change according to the category of book, type of author or format of the book" (321). All the players who commented on the statement of objections, or on the notifying party's reply to the statement of objections, be they booksellers, publishers of final consumers, said that payment of an advances represented a real risk (322).

(437) Furthermore, the very existence of accounts, within the accounting scheme in force in the publishing sector, intended to show uncovered advances, i.e. advances paid to the author but never offset by sales from a particular work, confirms by itself the risk attached to the notion of an advance. The publishing-house practice of using these accounts for making annual allocations to the reserves for uncovered advances illustrates the reality of the risk incurred by paying an advance, since any uncovered advance will be expressed in these expenditure accounts and will thus hit the profits (or exacerbate the losses) of a particular publishing house. Lastly, the not inconsiderable size of the amounts concerned expresses the degree of risk associated with the practice of paying an advance and with the fact that the payment of advances which are too large can entail the financial collapse of a publishing house. The Commission's investigation has shown that over the past five years allocations to reserves of this kind by the main publishing houses competing with the merged entity averaged close to 10% of the amount they actually paid in advances over the same period. This level may seem relatively low, but it has to be seen in relation to the acceptable limit to which an individual publishing house's incapacity to cover the advances it has paid might be allowed to rise: it is not conceivable that the true theoretical level should be 100%. If year after year a publishing house fails to cover a high proportion of the advances it has paid out, it is unreasonable to suppose that it will be able to go on operating profitably, developing its business and acting as a viable and effective competitor in the market. To take the analysis to the extreme, a rate of failure to cover advances close to 100%, while still acceptable in theory, would leave the firm with no option but to cease trading for want of structural profitability. Thus a publishing house can continue in business only if its rates of failure to cover advances are more reasonable. Against this background it will be seen that a rate of 10% is significant, and reflects the risk taken by publishing houses as they acquire rights year after year.

(438) The arguments put forward by the notifying party in reply to the statement of objections do not, therefore, call into question the method used by the Commission to assess the impact of the notified transaction on competition in the different primary markets for the rights in general literature titles.

(2) PRIMARY MARKET FOR THE FRENCH RIGHTS IN GENERAL LITERATURE TITLES

(439) The Commission's investigation shows that the transaction leads to the creation of dominant positions for the merged entity in the primary market for the acquisition of the French rights in general literature titles through the combination of horizontal, vertical and conglomerate effects. The totality of these effects will enable the merged entity to act independently of its competitors on the primary market for the acquisition of the French rights, since in a structural sense it will have an unrivalled power to attract authors.

(a) Horizontal effects

(440) The notified concentration would lead to the elimination of VUP as a competitor and hence to an immediate reduction of competition in this market, since two of the largest buyers of these rights would now form a single entity.

(441) According to the notifying party's method, the merged entity accounts on average for [25-35]% of the best-seller list for books by recurrent French authors of general literature in France in the last five years (323), ahead of Albin Michel (324) with [10-20]%, Gallimard with [10-20]%, XO with [5-15]%, Seuil with [0-10]%, Flammarion with [0-10]%, and the great many other publishers which have only very small market shares.

(442) According to the methods adopted by the Commission, the market shares of the main players are as follows:

<emplacement tableau>

(443) Thus, irrespective of the method adopted by the Commission, the merged entity will have much greater market shares than its nearest competitors in the market for the purchase of non-translated general literature titles. This is true for the purchase of rights from recurrent, well-known authors (second method), who are generally those who write the works traditionally regarded as best-sellers.

(444) It is also true, in particular, for the rights purchased from authors, whether recurrent or not (first method), who make up a significant part of publishers' business but who have less countervailing negotiating power than recurrent, well-known authors, since they have not yet necessarily enjoyed literary success.

(b) Vertical and conglomerate effects

(445) The vertical integration and the strong presence or dominant position of the merged entity on the different markets in the book chain would help to create a dominant position on the primary market for the purchase of the French rights in general literature titles by increasing its attraction for authors, thus enabling it to act independently of its customers and competitors.

(446) As Hachette Livre has stated, and this is confirmed by the Commission's investigation, authors are sensitive to the human working relationship which they maintain with their publisher, this aspect being, for many of them, an important factor in the selection of a publisher. However, there is no reason to think that the publishing houses in the Hachette Livre and VUP groups would not be as capable as their competitors of meeting authors' expectations. In particular, the fact that, within these groups, the publishing identity of the houses which have gradually been incorporated into them has been preserved helps to provide authors with individualised contact with their publisher "on a human scale". However, with regard to the other criteria whereby authors select publishing houses, the merged entity would enjoy an unequalled competitive advantage over its nearest competitors as a result of the notified transaction (325).

(447) In its reply to the statement of objections, the notifying party considers that, while the criteria stated by the Commission are genuinely important, VUP and Hachette Livre occupied a strong position before the transaction was notified (326). The Commission notes in this respect that, before the transaction was notified, Hachette Livre and VUP were respectively the leader and the third player in the market. Thus, before the notified transaction, the two undertakings already had positions that were not inconsiderable. As a result of the transaction, Hachette Livre would increase its strength in the different publishing sectors and thus create a dominant position in the market for the primary acquisition of the French rights in general literature titles, thereby widening the gap with Gallimard. The other players cited by the notifying party are XO and Albin Michel, which have links with Hachette Livre and VUP (327). The Commission is not arguing that Hachette Livre and VUP constituted a dominant duopoly from before the merger, but rather that the elimination of one of the two main competitors leads to the creation of a dominant position.

(i) Marketing, distribution and sale of pocket-format books

(448) The results of the Commission's investigation reveal that authors are sensitive to their publisher's capacity to market their works as effectively as possible, i.e. by ensuring the widest possible marketing and distribution, with the book being available in all types of outlet (328). Given the power of the merged entity in marketing, distribution and wholesaling and the fact that, for authors producing works aimed at a wide public, only access to the smallest outlets can ensure commercial success, authors will naturally be attracted by the merged entity to the detriment of its competitors which are not integrated or whose strength in these activities is clearly less significant. This is confirmed moreover by the notifying party itself in [...]* (329). In its reply to the statement of objections, the notifying party states that the Commission is being illogical, since coverage would not be affected (330). Even if this were true, it would still be the case, as is described in more detail in the section of this Decision dealing with marketing services, that, following the merger, one of the two channels to the smallest outlets would have disappeared and that the other available channels (the hitherto independent wholesalers) would have less commercial freedom than they have today. The merged entity would thus have greater access to the smallest outlets than its competitors, and its attraction for authors would be increased as a result.

(449) The merged entity's dominant position in the market for the sale of general literature titles in pocket format would enable it to supplement its capacity for attracting authors. Given the many commercial outlets that the "second life" of a work represents for recurrent, well-known authors whose books are systematically published in pocket format, and also for all authors who have enjoyed a literary success and can therefore expect to see their next work published in pocket format, the merged entity's dominant position in the market for the sale of general literature titles in pocket format is an added incentive for an author to be published by the publishing houses in the merged entity. This situation is confirmed moreover by [...]* (331). In its reply to the statement of objections, the notifying party considers that, for an author, being published by a group which has three generalist pocket-format series rather than one has no impact on his "second-life" sales prospects. The Commission considers, however, that for an author the fact of being published by a group which has three generalist pocket format series gives him more chance of seeing his work published in pocket format. Where an author is published by a publisher which has several pocket-format series, his work will more readily suit one of those series than if his publisher had only one series.

(450) The combination of the strength of the merged entity in marketing and distribution and of its dominant position in the markets for the sale of general literature titles in pocket format helps to attract authors, therefore, on the primary market for the purchase of general literature rights. Thus, as Le Soir emphasised in an article on fifty years of Le Livre de Poche (10 February 2003) "[In 1953] Henri Filipacchi, company secretary of the bookseller Hachette had the smart idea, when championing his project [pocket-formats], of building on Hachette's solid marketing and distribution network. A network which he knew was effective". This is confirmed by Mr Dominique Goust, chairman of Hachette's subsidiary, Le Livre de Poche: "The pocket-format book - consolidates success and prepares that of future books. It has a snowball effect. It can help an author reach all his potential customers more quickly". This mutual reinforcement is confirmed by Gallimard, a competing publisher, which explains that "all developments of the firm have synergetic effects, which are reinforced by the size of the business. Thus a firm which has a significant market share, especially in marketing/distribution, will find it that much easier to obtain a high rate of return on its investments in capacity, rights purchasing or external growth" (332).

(451) In addition, the merged entity's strength in marketing/distribution, the sale of books by wholesalers and the sale of general literature titles in pocket format could enable it, if its intrinsic power of attraction is not enough, to offer authors larger financial remuneration, in the form of advances or proportional remuneration, than competing publishers.

(452) Since the amount of the advance offered to an author for publishing his book is calculated as a function of all the anticipated revenues (333) from marketing the work, the unrivalled ability of the merged entity to maximise its profits at the level of pocket-format sales, marketing, distribution and its wholesale activity, where it is in a strong position, would give it a greater ability than its competitors to amortise advances paid out and, hence, strategically and specifically to offer authors higher advances than those offered by publishers competing with it. (334) In its reply to the statement of objections, the notifying party considers that the Commission is making an error of fact concerning the amortisation of advances through the sales of pocket-format books (335).

(453) As a report on Robert Laffont (336) emphasises, "the amortisation of titles, in particular through pocket-formats, is a basic aspect of the business. Robert Laffont's titles are covered at all levels of outlet - The sale and purchase of rights is an essential part of Robert Laffont's activity". It is fairly clear from this description that the more integrated a publisher is in pocket-format sales and in marketing/distribution services, the more its capacity to compete on the primary rights markets is strengthened by its ability to amortise any surplus remuneration offered to authors over a number of copies that is significantly larger and over the resale of the pocket-format rights (in-house or externally). This situation is confirmed by Mr Francis Esmenard, chairman of Albin Michel, who, referring to pocket formats, stresses that "these extra sales are essential if we are to remain competitive in the purchase of foreign authors' rights, which are auctioned systematically, and are also important for publishers' ability to attract French authors" (337). Moreover, as the notifying party itself emphasises, .the amount of an advance that an author may be offered is thus directly determined by the forecast sales of the work (in large format and in secondary publication). (338).

(454) Even before the transaction was notified, competing publishers have been permanently living with the threat of losing their authors, as Odile Jacob, a publisher competing with the merged entity confirms; it states that "a house like Odile Jacob is continually having its authors poached by publishers like Hachette and VUP, which can offer very large marketing capability and advances. Hachette and VUP are also able to amortise excess advances over the whole book chain (distribution, marketing/pocket formats/retailing)" (339). Since the notified transaction would lead to the creation of dominant positions in the markets for marketing and distribution services and in sales of general literature titles in pocket format (340), the situation would be aggravated, and would help to create a dominant position for the merged entity on the primary markets for purchasing the rights in general literature titles.

(ii) Advertising and media

(455) As several parties have pointed out, the considerable presence of the merged entity, unlike its competing publishers, in the advertising (341) and media (342) sectors, such as television, radio and the printed press, augments its power to attract authors. Market players have emphasised the importance of this consideration, which is indeed confirmed by Hachette Livre's own strategic plan. But the Commission takes the view that in the circumstances it is not a decisive factor that would bring about the creation of a dominant position in the market for the acquisition of the primary French rights in general literature titles.

(456) The Lagardère group is by no means a negligible player in the media and advertising sector. As regards the media, the group's radio interests, with nearly six million listeners, have a combined audience of 12% (343). The advertising sales agencies of the Lagardère group account for [20-30]% of the French market for broadcast advertising and [15-25]% of advertising in magazines aimed at the general reader (344). In this context, Lagardère Active Publicité is the exclusive advertising sales agency for all the radio stations controlled by the Lagardère group and for other radio stations not controlled by Lagardère, including inter alia the radio stations covering the French motorway network.

(457) It was found that, even before the notified transaction, Hachette Livre's publishing houses were being charged preferential "group" rates for buying advertising space on the group media (345). "Guaranteed" purchases of advertising space on its media products would increase significantly, because the needs of VUP's publishing houses would now be brought into the group, so the notified transaction would enable the Lagardère group to extend its preferential terms to those publishers (346). For competing publishers access to the Lagardère group's media products would automatically become more difficult, given the greater volume of internal customers who would naturally have priority, and all the publishing houses of the merged entity would also be able to enjoy even more preferential terms, as these would be justified by the larger guaranteed volume of media traffic following the notified transaction.

(458) In order to increase sales of their works or quite simply to make themselves better known to the entire potential readership, authors need coverage in all media, and as a result of its integration and its strength in this type of activity only the merged entity would be able to guarantee them privileged access to its media products. As regards general literature titles, while promotion is not the only factor needed for the successful launch of an unknown author, it does sell books. This is confirmed by VUP, for which a media plan makes it possible "to make a work known to its target audience so as to improve sales" and "to counter the media presence of the competition" (347), and also by Stock, which considers that "the purpose of the media operations is to establish an author permanently and to increase sales" (348). Promotion is particularly important because of works of general literature are produced in substantial numbers, and promotion is therefore necessary to be able to allow promising works to be discerned by the potential readership and sales to be made, as one hypermarket emphasises when it states that "the supply of literature is so big that promotion [here, over the radio] enables certain works to speak to the listener. The latter thus remembers the title of the book and may buy it later as a priority purchase" (349).

(459) The importance of promotion is emphasised by authors, even if it is not their first criterion for selecting a publishing house, and taken into account by publishers. Thus, according to Plon-Perrin, even if "the media plan is not contractual", it "is drawn up with the author's collaboration" (350). In other words, as Grasset states (351), "in literature, paying promotion campaigns are paid for to improve sales of the work and to make the author understand that [the publisher, here Grasset] is ready to spend money for a book in which it believes". This situation is confirmed by the notifying party itself in [...]* (352).

(460) Among the range of incentives at the disposal of the publishing houses in the Lagardère group, the merged entity could make use of its position in the media and advertising sales agencies, in terms of both priority access and advantageous rates, to attract authors.

(iii) Financial capacity

(461) Subsequently, quite apart from any link with its greater ability to amortise advances, the merged entity would have a much greater financial capacity, in terms of the financial flows generated by its activities in the book sector, than its nearest competitors; this would enable it to think strategically in the longer term about acquiring authors and hence attracting them with strategically higher advances and proportional payments when this proves necessary. This analysis is confirmed by the notifying party itself in [...]* (353). The investigation carried out by the Commission has shown that in the large majority of cases authors who have left a publisher have obtained better financial conditions at their new one (354). This importance of the payment to authors is confirmed by the fact that, for authors, it is one of the criteria for selecting a publishing house (355). In its reply to the statement of objections, the notifying party considers that if an author is offered a higher advance by his new publishing house, it is because the sales prospects of the book for which he signs a contract with a new publisher are usually better than those of his previous book (356). However, since publishing is a risky occupation, the Commission considers that the new publisher can never be sure it will be more successful in selling an author. As the notifying party emphasises, "in the field of general literature, publishers seek out books which may be successful in bookshops. As success in this field is highly uncertain, there are no set targets for returns" (357). Since authors, according to the notifying party, are not sensitive to their remuneration, the mere hope of a preferential link with their new publisher should be reason enough for them to leave a publishing house; and a higher advance should not be needed. In addition, if better sales than for his previous books are anticipated, better remuneration for the author will show up in the proportional remuneration, without it being necessary to increase the advance; this will allow the publisher to avoid taking a risk with a non-reimbursable amount.

(462) In view of the importance to a publisher, underlined by VUP, of having substantial financial resources, in particular "to bring recognised authors, who generally ask for high sums, to your company and hence guarantee them large advances" (358), the financial capacity of a publishing house is one of the factors likely to contribute to the attractiveness of a publisher to an author. For the strategic and specific purpose of excluding competitors, this financial capacity could in fact be used to increase authors' remuneration or to guarantee them, as well as preferential access to the media products of the Lagardère group as discussed above, publicity investment in all the opinion-forming media in accordance with their expectations. This is confirmed by Albin Michel, which states that it is important for a publisher of general literature to have substantial financial resources "in order to attract recognised authors, keep recognised authors and ensure works are promoted" (359).

(463) In this respect, the notified concentration results in the merging of the two leading publishers, which are also the most vertically integrated throughout the book chain, i.e. the two groups with the highest revenue and liquidity in the entire French- language publishing sector, with revenues which are about 800% those of its nearest competitor and which generate an operational cash flow (EBITDA) greater than the turnover of the merged entity's nearest competitor (360). As the notifying party itself claims, these financial resources described by Gallimard, a competing publisher as "way in excess of those generated by the book trade" (361) can be mobilised at any time to enable the media activities of Lagardère, of which Hachette Livre (and, potentially, VUP) are part, "to snap up opportunities thanks to its very sound financial resources" (362).

(464) It is evident, therefore, that, on this level as well, the notified transaction would considerably alter the structure of the markets for the primary purchase of rights by creating a player which is much larger than its competitors. This scale effect would enable the merged entity, specifically and strategically, to take marginally more risks than its competitors, since the impact of a failed investment in an author would be more easily absorbed and diluted than in significantly smaller, and hence less resilient, publishing houses or publishing groups. This limitation of risk-taking is intrinsically bound up with the question of size and consequent financial capacity, as emphasised by Albin Michel, when it states that it is important for a publisher of general literature to have sufficient financial resources "to meet the cost of publishing failures, which may be very high" and that "as regards publishing, it can happen that [Albin Michel] foregoes investments, not from a lack of financial resources but because it is faced with a level of risk-taking that the firm's size does not permit" (363). From this point of view, taking such a risk would be easier for a publisher with a financial base many times that of its competitors. This is how the evidence of Éditions du Rocher should be understood: "our policy has always been to be prudent as regards financial commitments - the result partly of our financial stability and partly of a lack of opportunity to purchase authors - and as regards the investment required for launching our products, since we cannot commit investments which exceed our normal financial capacity" (364).

(465) In light of the foregoing, the merged entity's access to less stringent financing terms than those available to its competitors could reinforce, in a specific and strategic manner, the range of factors that make up the entity's unrivalled attractive power, through larger advances or the financing of broader promotion campaigns. Market players have emphasised the importance of this consideration, which is indeed confirmed by Hachette Livre's own strategic plan, but the Commission takes the view that in the circumstances it is not a decisive factor that would bring about the creation of a dominant position in the market for the acquisition of the primary French rights in general literature titles.

(iv) Retail sales

(466) Lastly, the merged entity would have a growing presence on the downstream market for the sale of books by dealers to the final consumer through its different shops (Virgin, Relay and Furet du Nord) (365), whereas competing publishers are not present on that market (or a limited number of them are present to a totally marginal extent); this is not significantly altered by the notified transaction (366), but it nevertheless supplements the merged entity's power to attract authors. The point is confirmed by Odile Jacob, a publisher competing with the merged entity, which emphasises that "at the retail sales level, Hachette has an obvious advantage, [since] publishers which have no significant presence in this activity cannot guarantee that booksellers will buy the books, while publishers/booksellers such as Hachette can more easily persuade authors of their commercial strength" (367). The notifying party itself also confirms this analysis in [...]* (368).

(467) Market players have emphasised the importance of this consideration, which is indeed confirmed by Hachette Livre's own strategic plan. But the Commission takes the view that in the circumstances it is not a decisive factor that would bring about the creation of a dominant position in the market for the acquisition of the primary French rights in general literature titles.

(c) Countervailing power

(i) Existing competition

(468) Some of the existing competitors with the merged entity in the markets for the primary purchase of French rights in general literature titles are publishing houses which are not vertically integrated over the rest of the book chain, or which are so only to a slight extent. Particular examples are Actes Sud, Payot/Rivages, Les Éditions de Minuit, XO, V. Hamy, Desclée de Brouwer, Le Dilettante, Autrement, Le Cherche Midi, Michel Lafon, Les Éditions Du Rocher, De Fallois, Odile Jacob and Anne Carrière. Since these publishers are not vertically integrated (369) and their financial capacity is not comparable to that of the merged entity, they do not have the same ability to attract authors or to compete with integrated publishers as regards authors' remuneration. Thus, even before the notified transaction, while they search for and discover talent, such publishers manage to keep very few recurrent authors within their fold. It is already possible to observe that while these publishers published books that figured among the 100 best-selling non-translated general literature titles in 2002, their names no longer appear, except for XO and, to a lesser extent, V. Hamy and Le Dilettante (370), in the list of publishers of books by recurrent, well-known authors drawn up by the Commission on the basis of the 100 best-selling titles. The same applies to translated books, in respect of which only Payot/Rivages, First Editions, Autrement and Actes Sud still figure, although their sales are small. The representatives of Desclée de Brouwer explain in this respect that "once an author has a first success, the balance of power with the publisher is reversed. Competitors know that the success has occurred - and make specific proposals to the author - This hardening of relations between authors and publishers - weakens small and medium-sized publishers to some extent, since they have fewer means of resisting pressure from authors. Not infrequently, a small publishing house launches a first book and then sees its author move to a larger firm for later works". The representatives of De Vecchi state that "changes occur when an author already has three or four titles published and another - often more powerful - publisher makes a more attractive offer", while those of Fleurus state that "as a general rule, authors use small and medium-sized firms to launch themselves and are hunted by large ones as soon as their success has been secured" (371).

(469) If the notified transaction goes ahead, these publishers would therefore be even less able than before to keep their authors and hence to exert competitive pressure on the merged entity. This is particularly true of those that are dependent on it for marketing and distribution and for selling their books to the smallest outlets. Such is the case in particular with XO, De Fallois and Anne Carrière, which would also have capital links with the merged entity since the latter would have shareholdings in them (372).

(470) It is also true of Albin Michel (373), which has a holding in LGF; this makes LGF the preferential pocket-format publisher for its large-format titles. It is evident, moreover, from Article 4 of the memorandum of agreement signed on 14 June 1999 between Albin Michel, Hachette Livre and LGF that [...]* (374). In addition, Albin Michel is tied to Hachette Livre [...]* (375), (376). The marketing of books to the smallest sales outlets and the effectiveness of distribution are factors in the choice of a publisher by authors (377), and the sale of pocket-format books is a resource necessary in order to amortise the advances paid to authors; it is to be feared, therefore, that Albin Michel's ability to compete with Hachette Livre is already small and is not in any event greater than that of the vertically integrated publishers.

(471) As well as these publishers with only a small measure of integration, there are other larger, vertically integrated publishers (378), but their strength does not compare with that of the merged entity; the latter will generate more than eight times the overall turnover of its nearest rival, Gallimard. Apart from the fact that their positions on the primary markets for the purchase of French rights in general literature titles are clearly smaller than that of the merged entity, these publishers are not vertically integrated to the same degree, nor do they have the same strength on the different markets in the book chain; they would depend on the merged entity, since they do not market their own books to small outlets. They would not be able, therefore, to offer the same quality of services as the merged entity. Nor do they have the merged entity's financial capacity, and they would not be able to compete with it in this respect either. While it cannot be ruled out that publishers like Gallimard or Flammarion might offer competitive advances to an author in certain cases, their size and financial situation oblige them to target such action, and hence to limit the number of authors with high advances in their publishing houses, whereas the merged entity will have more than ten general literature publishing houses in its portfolio, dominant positions at nearly all levels of the book chain, a unique presence in the media and a financial capacity that is many times greater. The competitive constraint exerted by these publishers would not be enough, therefore, to prevent the merged entity from behaving independently of them.

(ii) Potential competition

(472) As far as potential competition is concerned, while it is true in theory that anybody can act as a publisher and canvass unknown authors, entry to the market for the primary purchase of French rights in general literature titles with a view to competing effectively with the merged entity assumes that the new entrant is able to exploit these purchase rights, and therefore that it has the same degree of vertical integration as the merged entity, and especially marketing and distribution facilities will guarantee a sufficient presence for its products in outlets, and that it has a pocket-format series able to give the works a second life; so that the barriers to entry are high.

(473) While a publisher not present in French-speaking countries and, hence, in French-language publishing may have a general knowledge of the market for the purchase of rights, it will have no knowledge of the marketing and distribution of its books, and more generally of how the French-language publishing sector works. These difficulties are important barriers in the way of any potential entrant desiring to penetrate the primary markets for the purchase of French-language publishing rights, for books that may or may not have been originally written in French. The notifying party argues that Bertelsmann in particular ought to be considered a potential entrant (379). It is true that Bertelsmann does acquire rights on primary markets abroad, and exploits them in languages other than French. But, to enter the primary market for French-language publishing rights on a significant scale and not just in a token fashion, i.e. to buy the right to works with a view to publishing them in French, Bertelsmann would have to establish itself as a full-scale publishing house, and would therefore have to present itself to authors as being able to provide creative assistance (working relationship with the author) and to promote their works on the French-language markets, whereas today virtually all the French rights it holds were acquired on the secondary markets from publishers for a club operation and not direct from authors. Furthermore, while Bertelsmann may be considered a conglomerate, it is not, unlike the merged entity, vertically integrated in the publishing of French-language works. To become so, it would have to join the traditional marketing and distribution circuits, when it is currently present only in the sale of books to the final consumer through book clubs and therefore maintains no marketer/distributor relation with booksellers and other dealers. The presence of a book in traditional bookshops is essential if authors are to be sufficiently visible to all consumers and not just to members of a book club. Even if Bertelsmann planned to penetrate these primary markets, its dependence on publishers and in particular on the merged entity, which accounts for more than 50% of its current supplies of French-language publishing rights for club publishing, would be a dissuasive factor difficult to overcome.

(474) The combination of the above factors means that there is no genuine competitive pressure from rival publishers, whether actual or potential.

(iii) Supply

(475) On the primary market for the purchase of French rights in general literature titles, the supply is made up of numerous individual players of a size and economic strength which are far less than those of the merged entity. Given the market shares of the merged entity as a purchaser of these rights, and given that for suppliers the main selection criteria, apart from the relation of mutual trust with one's publishing house that any publisher is potentially capable of providing without difficulty, are the amount of the advances offered and the coverage offered by the marketing and distribution facilities, authors selling rights would naturally be encouraged to contract with the merged entity.

(476) The fact that French-language authors are also attached to the reputation of their publishing house is not a factor that is likely to remove the merged entity's power of attraction, since it is clear that in its portfolio of publishing houses some have a reputation at least equal to that of rival houses.

(477) Similarly, the rights of first refusal (380) that bind authors to their publisher are not a factor that is able to prevent the creation of a dominant position by the merged entity on the primary market for the purchase of publishing rights. The Commission's investigation shows that first-refusal rights do not figure in all publishing contracts, since authors who are fairly well-known often refuse this type of clause. Thus, of all the publishing contracts between Hachette Livre and authors sent to the Commission (381), more than 60% do not have a first-refusal clause. Similarly, for a publisher like Seuil, only 40% of contracts for non-translated books (382) contain a first-refusal right. Furthermore, this type of clause is governed by statute in France and Belgium. Article L 132-4 of the French Intellectual Property Code requires the number of future books concerned to be limited and the types of book concerned to be specified. Article 3-2 of the Belgian Act of 30 June 1994 on copyright and neighbouring rights states that the sale of property rights relating to future works is valid only for a certain period and provided that the types of work covered by the sale are specified. But some authors write books in different genres within the category of general literature, (383) which enables them to get round this potential difficulty. In practice, moreover, publishers have few means of keeping an author who really wants to leave, the cost and energy required to sue the new publisher being generally dissuasive (384). The representatives of Seuil explain in this respect that "It is impossible in practice to force an author to publish a new book with a particular publisher when the complex link between them is broken - It is inconceivable, after all, that compliance with the right of first refusal could be imposed by force when the author warns the publisher in advance of publication by a competitor", and those of Flammarion state that "it cannot be said that the clause makes it possible to keep an author who is bent on leaving" (385). The notifying party itself states that .each new title of an author may give rise to a change of publisher. While in some contracts there may be an option right for the publisher of the previous title, there is of course no commitment to sell on the part of the agent: if another publisher makes a higher or better-quality offer, the rights in the new title will be granted to it. In the field of copyright there is no type of agreement equivalent to the notorious "output deals. concluded between American film studios and television broadcasters" (386).

(478) Given the combination of these factors, the conclusion is that suppliers would have no countervailing power.

(d) Conclusion

(479) The notified transaction would therefore create a dominant position as a result of which effective competition would be significantly impeded in the Common market or a substantial part of it in the primary market for French rights in general literature titles.

(3) PRIMARY MARKET FOR FOREIGN RIGHTS IN GENERAL LITERATURE TITLES

(a) Horizontal effects

(480) According to the method of the notifying party, the combined market shares of the merged entity account on average for [25-35]% of the best-selling list of books by recurrent foreign authors of general literature in France over the last four years (387), behind Albin Michel [40-50]% and ahead of Seuil [0-10]%, First Editions [0-10]%, Autrement [0-10]% and other publishers, which have only very small market shares.

(481) According to the Commission's first method, the merged entity accounts for [30-35]% ([10-15]% for Hachette Livre and [20-25]% for VUP) of the advances paid to the authors of the 100 best-selling translated general literature titles in France in 2002, after Albin Michel ([50-55]%) and ahead of Seuil ([5-10]%).

(482) According to the Commission's second method, the merged entity accounts for [30-35]% ([10-15]% for Hachette Livre and [15-20]% for VUP) of the advances paid to the recurrent authors among the 100 best-selling translated general literature titles in France in 2002, after Albin Michel ([55-60]%) and ahead of Seuil ([5-10]%).

(483) Thus, whichever calculation method is used, the merged entity would have market shares of [30-35]%, smaller than those of Albin Michel.

(b) Vertical and conglomerate effects

(484) The vertical integration and the strong presence or dominant position of the merged entity on the different markets in the book chain would increase its power to attract authors.

(485) However, the Commission's investigation has shown that non French-language authors accord greater importance than French-language ones to the remuneration criterion in selecting their publisher (388). The other selection criteria, such as the marketing, distribution, promotion or sale of their book in pocket format, are less important than for French-language authors. The new entity's power of attraction, linked to the vertical effects, is therefore less.

(486) Since financial capacity is not a sufficient factor in itself for creating a dominant position, the vertical integration and strong presence or dominant position of the merged entity in the different markets of the book chain are not by themselves likely to create a dominant position for the merged entity in the primary market for the foreign rights in general literature titles.

(c) Conclusion

(487) In view of these factors, the horizontal effects, the vertical integration and the merged entity's strong presence or dominant positions in the various markets in the book chain are not such as to create a dominant position held by the merged entity in the primary market for foreign rights in general literature titles.

A.1.b. PRIMARY MARKETS FOR THE RIGHTS IN CHILDREN'S BOOKS

(488) As indicated in the section on the definition of the product markets, it is appropriate to subdivide the primary markets for the publishing rights in children's books into a market for French rights and a market for foreign rights.

(489) For the purposes of the competition analysis, however, since the expected effects of the proposed transaction on the markets for French rights and the markets for foreign rights are relatively similar in nature, they are discussed together below; where appropriate, specific effects on these markets are identified separately.

(1) HORIZONTAL EFFECTS

(a) Method of calculating market shares

(490) The notifying party did not supply an estimate of market shares according to the method used for the primary markets in the rights in general literature titles. Such an estimate would have merited the same criticisms as made above. The concept of recurrent authors is not relevant, either, to children's books and would not therefore constitute a reliable indicator for assessing the competitive impact of the notified transaction. According to the notifying party, recurrent authors of children's books accounted for less than 10% of sales of children's books in 2002 (389).

(491) To assess the strengths of the different publishers in the primary acquisition of rights in children's books, the Commission tried to aggregate the annual amount of advances and proportional remuneration paid for non translated books on which advances were paid, but came up against the same technical impossibilities as for the primary markets in the acquisition of rights in general literature titles. It therefore used the same method as for these markets and recorded the amount of the advances paid by the different publishing houses to the authors of the 100 best-selling translated and non-translated children's books in France in 2002 (390). Among these best-sellers, however, there were collective works, for which there are no advances and which were therefore withdrawn from the list for the purpose of calculating market shares.

(492) Although this collection of data concerns only the 100 best-sellers in France and not in the world - the relevant geographic dimension of the markets analysed - France occupies a very large place in the markets for French-language publishing rights compared with the other French-speaking countries and may therefore be regarded as representative of the market as a whole.

(b) Market shares

(493) According to the method used by the Commission, the market shares of the main players are as follows:

<emplacement tableau>

(494) Thus the merged entity would have higher market shares than those of its nearest competitor on the market for the acquisition of the French rights in children's books, and much lower market shares than those of its nearest competitor on the market for the acquisition of the foreign rights in children's books.

(495) However, while Hachette Livre has a not inconsiderable position in the market for the acquisition of French rights to children's books ([10-20]%), that position has been achieved by buying a single title. It is therefore possible to consider that Hachette Livre does not have a large presence in this market for the acquisition of rights in children's books.

(496) Furthermore, in the children's books sector, the Commission's investigation has shown that the large majority of authors are published by several publishers at the same time (391). Unlike the authors of general literature titles, who generally write first one book before starting another, the authors of children's books usually publish several books at very short intervals.

(2) VERTICAL AND CONGLOMERATE EFFECTS

(497) The vertical integration and the strong presence or dominant position of the merged entity on the different markets in the book, as described for the market for the acquisition of French rights in general literature titles, would increase its power to attract authors of children's books.

(498) However, the Commission's investigation has shown that children's books are often first published in pocket format and that there are no, or very few, second publications. The strength of the parties in pocket format sales has no effect therefore on the merged entity's power of attraction.

(499) In addition, children's titles are much less susceptible to promotion than general literature titles. The strength of the parties in the media therefore has little influence on authors. selection of their publishing house.

(500) The advances offered to authors of children's books are much lower than those offered to authors of general literature titles. The financial capacity which the merged entity will have is therefore of less importance for the primary acquisition of rights in children's books.

(3) CONCLUSION

(501) Given these different factors, the vertical integration and the strong presence or dominant position of the merged entity on the different markets in the book chain are not such as to create a dominant position for the merged entity in the primary markets for rights in children's books.

A.1.c. THE OTHER PRIMARY RIGHTS MARKETS

(502) As stated in the section on the definition of product markets, strip cartoon albums and academic and professional books are also individual works in which publishing rights can be acquired.

(503) Given the weak presence of the parties in these two sectors of publishing activity, the notified transaction is not such as to create or strengthen a dominant position in the markets for the acquisition of rights in strip cartoon albums and academic and professional books.

(504) Hachette Livre is present in the publication of strip cartoon albums only because it is the publisher of 25 old Asterix titles. VUP is present through Hors Collection, which publishes foreign strip cartoons in French. Hachette Livre is therefore not present in the market for the acquisition of foreign rights in strip cartoon albums, while VUP is present in the market for the acquisition of French rights in such albums.

(505) In the case of academic and professional books, Hachette Livre has a marginal presence in this sector and hence on the markets for the acquisition of rights. The transaction does not therefore alter the structure of the market. In addition, for these types of work, the vertical integration of the new entity in the different sectors of the book chain has only a slight impact, since these books are not the subject of a secondary publication, are basically marketed and distributed to large bookshops or specialist bookshops and do not therefore require access to small outlets. The presence of the merged entity in the media and its financial power are not factors which attract the authors of this type of book, since their books are not promoted very much and the advances offered are not very high.

A.2. SECONDARY RIGHTS MARKETS

(506) As stated in the section on the definition of the product markets, it is appropriate to subdivide the secondary rights markets into the secondary market in pocket-format rights and the secondary market in club rights.

A.2.a. SECONDARY MARKET FOR POCKET-FORMAT RIGHTS IN GENERAL LITERATURE TITLES

(507) As indicated in the section on the definition of product markets, the secondary market for pocket-format rights should not be subdivided according to the original language of the work, since the selling publisher has already translated the work the primary publication in large format. The market for secondary rights concerns the purchase of rights only in general literature titles, since the other categories of work very rarely give rise to second publications (392).

(508) The investigation carried out by the Commission shows that the notified transaction results in the creation of a dominant position for the merged entity in the secondary market for the purchase of pocket-format rights in general literature titles through a combination of horizontal, vertical and conglomerate effects. All these effects would lead to the creation of a dominant position for the merged entity on the secondary market for the purchase of rights, since structurally it would have an unrivalled power of attraction for primary publishers and authors, enabling it to behave, to an appreciable extent independently of its competitors and customers, and ultimately of consumers.

(1) HORIZONTAL EFFECTS

(509) The notified concentration will lead to the elimination of VUP as a competitor and, hence, to an immediate reduction in competition on this market, since two of the biggest purchasers of pocket-format rights will now form a single entity.

(510) As a rule, when a book is first published in large format, the pocket-format publisher purchases the rights from the first publisher in return for a payment proportional to sales (393), and generally accompanied by an advance. The suppliers, therefore, are the publishing houses which first published the works (394) and which sell their pocket- format rights (either for reasons of expediency, or because they themselves do not have a pocket-format series).

(511) The Hachette Livre publishers exploiting the pocket-format rights to books are basically LGF ("Le Livre de Poche" series), which makes most of the purchases of pocket-format publishing rights on behalf of the group, and La Librairie des Champs- Elysées. Within VUP, Univers Poche publishes books in pocket format for its "10/18", "Pocket" and "Fleuve Noir" series (395).

(512) The main competitors of the merged entity are Gallimard (the "Folio" series), Seuil (the "Points" series) and Flammarion (the "J'ai Lu", "Garnier-Flammarion" and "Librio" series). While some smaller publishers (Éditions de Minuit, Actes Sud, Phébus, Payot/Rivages, Éditions du Rocher, Odile Jacob, Picquier, etc") exploit pocket-format rights as well, they are not active in the market for the purchase of these rights, since their activity in this field is limited to the exploitation of their own lists. Thus, as the notifying party states (396), some of these publishers run a small pocket-format series, whose only purpose however is "to exploit their list further, which means that they do not require any extensive buying of rights." While the names of some of these publishers figure in the list of the 500 best-selling pocket- format titles in 2002, they no longer appear in that list if books which were the subject of primary exploitation by the same publisher are eliminated (397).

(513) It will be noted, in addition, that certain purchasers of primary publishing rights, some of whom indeed have significant positions on those markets, do not own a pocket-format series. This is the case in particular with publishers such as Albin Michel (which does, however, own a minority holding in LGF), XO or First Editions.

(a) Methods of calculating market shares

(i) Method proposed by the notifying party

(514) While denying the existence of a pocket-format market, the notifying party stated (398) that an assessment of the market shares of the large, generalist series of pocket-format books is sufficient to evaluate the strengths of the various players on the secondary market for the acquisition of pocket-format rights, since a pocket-format book is systematically a new exploitation of a work which has necessarily given rise to the purchase of rights (399).

(515) The Commission considers that this method excludes special series such as crime novels from the scope of the appraisal (400). However, since the Commission's investigation has shown that crime novels are first published basically as pocket-formats, the rights acquired by publishers in such cases are on the primary markets for publishing rights and not on the secondary markets. In this respect, therefore, it is possible to consider the assessment supplied by the notifying party as representative.

(516) Although this examination covers only sales in France and not the world - the relevant geographic dimension of the market examined - France accounts for a very large part of the secondary market in French-language publishing rights compared with the other French-speaking countries and may therefore be regarded as representative of the whole market.

(517) However, since this method concerns only the volume and not the value of purchases, it means that no account can be taken of any gaps between the amounts paid for purchasing the rights and the volume of sales actually generated. To acquire publishing rights, publishing houses do not compete on volumes sold but on the price offered to acquire those rights, i.e. basically on the advances offered (401).

(ii) Methods used by the Commission

(518) Thus, to arrive at a proper and realistic evaluation of the strength of the different publishers in the secondary purchase of pocket-format rights in general literature titles, the Commission asked the main players in the market to state the annual amount of advances paid (402) in order to acquire these rights from outside publishers ("first method") (403).

(519) The Commission also used another way of evaluating the strength of the different publishers in the secondary purchase of rights in general literature titles: collating the amounts of advances paid to outside publishers for the 163 best-selling pocket-formats in 2002 (404) ("second method").

(520) Although the latter data-collecting exercise relates only to sales in France and not the world - the relevant geographic dimension of the market examined - France accounts for a very large part of the secondary market in French-language publishing rights compared with the other French-speaking countries and may therefore be regarded as representative of the whole market.

(521) In its reply to the statement of objections, the notifying party considers that the Commission is wrong to exclude in-house assignments of rights from its analysis, since the conditions of competition are homogeneous as between in-house and external assignments and since the two types of right are substitutable.

(522) Apart from the fact that taking account of internal sales in the analysis of a merger is not part of the Commission's conventional approach to merger control, given Community case law (405), it is evident that in this particular case the conditions of competition are not homogeneous for these two types of right, which are not substitutable. Clearly, in publishing contracts, in-house and external assignments are subject to special rules, in particular as regards redistributing royalties to the author (406). In the initial contracts for first publications, the publisher itself generally acquires from authors, agents or foreign publishers the right to exploit the work in large and pocket format, without a new assignment and a new contract being necessary for in-house exploitation of the work.

(523) In any event, if in-house sales were to be taken into account, the method supplied by the notifying party to the Commission, which includes in-house sales since it covers all pocket-format sales, shows that the Commission's competition analysis would not be substantially altered.

(b) Market shares

(524) According to the notifying party's method (407), the merged entity's combined market share in numbers of copies would amount to [50-60]% ([20-30]% for Hachette Livre and [25-35]% for VUP), ahead of Flammarion with [10-20]% (408), Gallimard with [10-20]% and Seuil with [0-10]%.

(525) According to the methods adopted by the Commission, the market shares of the main players are as follows (409) :

<emplacement tableau>

(526) Thus, whatever the method of calculation chosen (410), after the notified transaction the merged entity will have far greater market shares than its nearest competitors on the secondary market for the purchase of pocket-format rights in general literature titles.

(2) VERTICAL AND CONGLOMERATE EFFECTS

(527) The vertical integration and the strong presence or dominant position of the merged entity on the different markets of the book chain would contribute to the creation of a dominant position on the secondary market for the purchase of the pocket-format rights in general literature titles by increasing the merged entity's attractiveness to primary publishers, authors and literary agents.

(528) The first finding is that publishing houses which publish their own pocket-format books have not normally any great interest in selling their pocket-format rights to others, since this secondary exploitation helps to make the investment in the first publication of the work profitable. It is for this reason, moreover, that many small publishing houses have set up their own pocket-format publishing operation, even though this is limited to exploiting their own lists. Publishers which do not have their own pocket-format series, like Albin Michel, XO or First Editions (and which do not carry out their own distribution or all of their marketing) have no alternative but to sell their secondary publishing rights. Publishing houses sell their pocket-format rights according to the prospective level of profitability which the pocket-format publisher can offer them (or the prospect of greater profitability, for those which have their own pocket-format publishing house). Prospective profitability depends on the advances which the pocket-format publisher can offer and its ability to reach the likely customers for the work in question. Thus, in the case of a work aimed at the general public, a pocket-format publishing house will be best able to sell the work well if it is supported by a marketing/distribution network with a wide coverage. Gallimard explains in this respect that, as far as it is concerned, "sales of pocket-format rights to other publishing houses are rare, unless a better financial offer is made or there is a better chance of reaching the target" (411).

(529) In addition, although the pocket-format rights are most often purchased from the large-format publisher, this does not mean that authors and literary agents have no control over the choice of the publisher that is to publish their book in pocket format. The notifying party states in this respect that literary agents and, increasingly, authors are asking, formally in the contract selling the French or foreign rights, to be able to approve the choice of pocket-format publisher; "the agent may also approach the first publisher informally, but to just as binding an effect, in order to impose a pocket-format publisher on it" (412). This statement is confirmed by the results of the Commission's investigation (413). Thus an author (generally a best-selling one) or the author's literary agent may require the publisher exploiting the primary rights to choose a particular publishing house to exploit the secondary rights. For instance, the representative of Actes Sud states that "Actes Sud has also been forced to sell Paul Auster's pocket-format rights to Le Livre de Poche, given Hachette's commercial image and proposed advances based on a guaranteed minimum - . No other offer compared with Hachette's in the eyes of Paul Auster's agents" (414). As was found in the case of the primary market for the publishing rights in general literature titles, authors and literary agents are very sensitive not only to the reputation of their publishing house and their payment but also to publishers' ability to be present on the shelves of the largest number of outlets of all types.

(530) In its reply to the statement of objections, the notifying party considers that if the criteria stated by the Commission were really important, VUP and Hachette Livre should have a strong position from before the notified transaction (415). In this respect, the Commission finds that, before the notified transaction, Hachette Livre and VUP were the third player (416) and the market leader respectively. It is possible to conclude therefore that, even before the notified transaction, the two undertakings already had not inconsiderable positions. As a result of the notified transaction, Hachette Livre would increase its strength in the different publishing sectors and thus create a dominant position in the secondary market for the acquisition of the pocket-format rights in general literature titles, thereby widening the gap with Flammarion. The Commission is not arguing that Hachette Livre and VUP were a dominant duopoly before the merger but, on the contrary, that the elimination of one of Hachette Livre's main competitors leads to the creation of a dominant position.

(a) Marketing, distribution

(531) The information gathered by the Commission from authors and literary agents shows, as was found in the case of the primary publishing rights in general literature titles, that the merged entity would have an unrivalled advantage over its nearest competitors following the notified transaction, given its attractiveness to authors and literary agents and also to publishing houses selling pocket-format rights.

(532) Thus authors and literary agents, apart from mutual trust and remuneration, are sensitive to the ability of their publisher to market their works as effectively as possible, i.e. by marketing and distributing them as widely as possible and ensuring that they are present in outlets of all types. Given the strength of the merged entity in marketing, distribution and wholesaling activities, and given that, for the holders of the rights to books intended for a wide public and hence most likely to be the subject of secondary exploitation in pocket format, only access to the smallest outlets can confirm the commercial success already encountered by the exploited work in large format, authors and literary agents would naturally be attracted by the merged entity, to the detriment of its non-integrated competitors or competitors supported by a less wide-ranging network for marketing and distribution. In its reply to the statement of objections, the notifying party states that the Commission is not being logical, since coverage would not be altered (417). Even if this were true, it would still be the case, as is described in more detail in the section dealing with marketing services, that, following the merger, one of the two channels to the smallest outlets would have disappeared, and that the other available channels (the hitherto independent wholesalers) would have less commercial freedom than they have today. The merged entity would therefore have greater access to small outlets than its competitors and its power of attraction would be increased.

(533) In this respect, the strength of the merged entity in the markets for marketing and distribution services was a competitive advantage for it vis-à-vis the selling publishers even before the notified transaction. As stated above, publishing houses which do not have their own pocket-format series most often do not have their own distribution structure either, and they do not carry out their own marketing, or in some cases do so only to a limited extent. The purchasers of secondary rights, however, are all present in the marketing and distribution of books on behalf of other parties. There is a link between the distribution/marketing of large-format books and the purchase of the pocket-format rights, since the results of the investigation carried out by the Commission show that a publisher will tend to sell the secondary rights which it does not exploit itself to the parent publishing house of its marketer/distributor. Thus publishers/distributors may have preferential agreements with the publishers they market or distribute. This is the case in particular [...]* (418), and also with other publishers/distributors such as Flammarion (419). Moreover, even where publishers of pocket formats/distributors have no preferential agreement or rights of first refusal, it is possible to find, as Flammarion emphasises, that "in our relations [with the publishers we market or distribute] we try to persuade them to sell us pocket-format titles" (420). Similarly, [...]* (421), all XO's books in the list of the 163 best-selling pocket formats were published in pocket format by VUP, which is its marketer/distributor (422). The same is true of Anne Carrière and De Fallois in the case of Hachette Livre, and of Le Dilettante in the case of Flammarion (423). However, the results of the Commission's investigation show that the ability of a publisher/marketer to persuade the publishers it markets or distributes that they should sell it their pocket-format rights depends on the coverage of its marketing/distribution network (to guarantee optimum sales for the pocket-format edition, which by definition is intended to be marketed widely). Thus the publisher Christian Bourgois, which is marketed and distributed by Seuil, sold the pocket-format rights for all its best-selling books (424) to VUP. The same applies to Le Cherche Midi, marketed and distributed by Gallimard, which sold the pocket-format rights for its best-selling titles (425) to VUP and Hachette Livre.

(534) Lastly, since the size of advances is calculated in the light of all the anticipated revenues (426) from marketing a work, the merged entity's unequalled ability to maximise its profits from the sales of books in pocket format by the publisher, and from marketing, distribution and wholesaling - in which markets it is in a strong position - would enable it to amortise advances more successfully than its competitors. It would thus be in a position, in the specific and strategic situations where its intrinsic attractiveness did not suffice to bring it pocket-format rights, strategically to offer advances higher than those offered by competing publishers of pocket-format books.

(535) In the light of the foregoing, the creation of dominant positions for the merged entity on the markets for marketing and distribution services would strengthen its competitive advantage over competing pocket-format publishers by giving it additional attractiveness to publishers wishing to sell pocket-format rights.

(b) Advertising and media

(536) The fact that only the merged entity, unlike its competing publishers, is present in the advertising (427) and media (428) sectors, such as television, radio and the printed press would be an additional factor in its attractiveness to authors and literary agents. Market players have emphasised the importance of this consideration, which is indeed confirmed by Hachette Livre's own strategic plan; the Commission takes the view that in the circumstances it is not a decisive factor that would bring about the creation of a dominant position in the secondary market for the purchase of pocket-format rights in general literature titles, but that it strengthens the other factors analysed above.

(537) Where general literature titles in pocket format are concerned, what is promoted is usually not just one work but more a collection or set of works in pocket format by a particular author. The impact of promotion is consequently less marked than it is on the primary rights markets, but that impact does exist. In order to increase the sales of their works or simply to make themselves better known to the potential readership as a whole, authors need coverage in all media, and only the merged entity would be able to guarantee them preferential access to its media because of its integration and its strength in this type of activity. Thus Jean-Claude Dubost, Chairman of Univers Poche in VUP, "invests intensively in marketing and promotion, to make both his brand and his authors known - To revitalise their lists and stimulate sales, pocket-format publishers are also developing author policies and backing particular writers" (429).

(538) Among the range of incentives available to publishing houses in the Lagardère group, the merged entity would be able to use its position in the media and advertising sales agencies, in terms of both priority access and special rates, to attract authors and literary agents.

(c) Financial capacity

(539) As was discussed for the primary market for the purchase of rights in general literature titles, apart from any links with its greater ability to amortise advances, the merged entity would have a financial capacity, in terms of financial flows generated by its activity in the book sector, which is clearly greater than that of its nearest competitors and which would allow it strategically to offer much higher advances when necessary. The importance of remuneration for authors and their literary agents is confirmed by the fact that it is an important criterion for selecting a publisher. This is endorsed by the notifying party, which emphasises that "the agent seeks to ensure that the rights granted in France are exploited as well as they can be and, hence, that the pocket-format rights are sold to the highest bidder" (430).

(540) A publishing house's financial capacity is therefore one of the factors likely to contribute to the attractiveness of a publisher to the sellers of pocket-format rights, i.e. publishers, authors and literary agents. Financial capacity could in fact be used specifically and strategically to offer higher advances.

(541) In this respect, the notified transaction results in the merging of the two leading publishers, which are also the most vertically integrated throughout the book chain, i.e. the two groups with the highest revenue and liquidity in the entire French- language publishing sector. As the notifying party itself claims, this financial capacity can be mobilised at any moment to enable the media activities of Lagardère, of which Hachette Livre (and, potentially, VUP) are a part, "to snap up opportunities thanks to very sound financial resources" (431). It is evident, therefore, that the notified transaction would considerably alter the structure of the market for the secondary purchase of pocket-format rights by creating a player which is much larger than its competitors and enabling it to take marginally more risks than they could, since the impact of a failed investment in a book would be more easily absorbed and diluted in a firm the size of the merged entity than in significantly smaller, and hence less resilient, publishing houses or publishing groups.

(542) In light of the foregoing, the merged entity's access to less stringent financing terms than those available to its competitors could reinforce, in a specific and strategic manner, the range of factors that make up the entity's unrivalled attractive power, through larger advances or the financing of broader promotion campaigns. Market players have emphasised the importance of this consideration, which is indeed confirmed by Hachette Livre's own strategic plan; the Commission takes the view that in the circumstances it is not a decisive factor that would bring about the creation of a dominant position in the secondary market for the purchase of pocket-format rights in general literature titles, but that it does strengthen the other factors analysed above.

(d) Retail sales

(543) Lastly, as was discussed in the context of the primary markets for the purchase of rights in general literature titles, the merged entity would have a growing presence on the downstream market for the sale of books by dealers to the final consumer through its different shops, when competing publishers are not present on that market (or a limited number of them are present to a marginal extent); this is not significantly altered by the notified transaction, but it nevertheless supplements the merged entity's power to attract authors. Market players have emphasised the importance of this consideration, which is indeed confirmed by Hachette Livre's own strategic plan, but the Commission takes the view that in the circumstances it is not a decisive factor that would bring about the creation of a dominant position in this market.

(3) COUNTERVAILING POWER

(a) Existing competition

(544) The existing competitors with the merged entity on the secondary market for the purchase of the pocket-format rights in general literature titles are all vertically integrated publishers, but do not compare at all in strength with the merged entity: the latter will generate more than eight times the overall turnover of its nearest rival, Flammarion (432). Apart from the fact that their positions in the market for purchasing the pocket-format rights in general literature titles are clearly smaller than that of the merged entity, these publishers are not vertically integrated to the same degree, nor do they have the same strength on the different sectors in the book chain; they would depend on the merged entity, since they do not market their own books to small outlets. They would not be able therefore to offer the same quality of services as the merged entity to sellers of pocket-format rights. Nor do they have the merged entity's financial capacity, and they would not be able to compete with it in this respect either. While it cannot be ruled out that publishers like Gallimard or Flammarion might on occasion offer competitive advances, their size and financial situation oblige them to target such action, and hence to limit the number of pocket-format books with large advances in their pocket-format publishing house, whereas the merged entity will, by itself, have three generalist pocket-format publishers in its portfolio, dominant positions at nearly all levels of the book chain, a unique presence in the media and a financial capacity that is many times greater. The competitive pressure exerted by competing pocket-format publishers would, therefore, be much reduced.

(b) Potential competition

(545) As to the potential competition, the first group of operators that might be able to enter the secondary market for the purchase of pocket-format rights in general literature consists of publishers which have their own pocket-format series, but use only their own list. Such publishers, however, already have difficulties in publishing in pocket format (or medium format) the works they previously publish in large format, since some authors force them to sell the pocket-format rights to other pocket-format publishers. Thus Actes Sud states that "in the pocket-format field as well, Hachette, through its powerful position, exerts pressure to sell the rights of authors [from other publishing houses] to its subsidiary [LGF]" (433). These potential competitors are small publishing houses which are not vertically integrated over the rest of the book chain, or are only slightly so. They include in particular publishers such as Actes Sud, Payot/Rivages, Les Éditions de Minuit, Le Cherche Midi, and Odile Jacob. As these publishers are vertically integrated only to a marginal extent and are not comparable in financial capacity to the merged entity, they are unable to compete with integrated publishers in terms not only of their intrinsic power to attract authors but also in terms of the amount of advances they would be able to offer if they decided to try and penetrate the secondary market for the purchase of the pocket-format rights in general literature titles; and this would be the case even if the merged entity decided to reduce the prices for acquiring pocket-format rights.

(546) The second group of players who might have an interest in entering the market for the purchase of pocket-format rights are publishers which do not yet have a pocket- format series. It emerges from the Commission's investigation, however, that there are sizeable entry barriers to this market. The majority of the publishers surveyed consider that the two most important barriers to creating one's own pocket-format series are the need to acquire a list so as to make up a consistent series, and access to a marketing and distribution network so as to sell these products in neighbourhood outlets, especially hypermarkets (434). The publisher Casteilla states in this respect that "as long as a publisher which does not already publish pocket-formats has no access to a distribution network, increasing prices [possibly by 5%] will have no effect. It will probably also have difficulty in buying rights!" (435). The other barriers to entry mentioned by publishers are the development of a commercial brand, in particular through commercial and promotional investments, and the ability to renew and develop the established list, which indirectly means that the candidate must have the financial capacity to attract authors, literary agents and publishers of books published in large format by other publishers. Since such publishers are not vertically-integrated, or only to a certain extent, and they do not do indeed run a coverage which includes small outlets, or a financial capacity comparable to the merged entity's, any establishment of a pocket-format series by such players would be limited, as is the case with the other non-vertically integrated publishers which do indeed run their own pocket-format series simply in order to publish in pocket format (or medium format) the works which they already exploit in large format, i.e. in order to exploit their own list further. Editions Fleurus states in this respect that "nobody not present in the pocket format has the slightest chance of successfully launching a new series" (436).

(547) The combination of the factors described above leads to the conclusion that there is no genuine competitive pressure from either existing or potential rival publishers.

(c) Supply

(548) On the secondary market for the purchase of the pocket-format rights in general literature titles, the supply side consists basically of the French-language publishers (437) which are not vertically integrated, or only to a certain extent, into the rest of the book chain, except for Seuil, Gallimard and Flammarion, the first two of which do not offer their books very much in the market for the purchase of pocket-format rights. These are in fact the potential competitors of the merged entity; their ability to exert competitive pressure on that entity, i.e. their countervailing power, has already been analysed above.

(549) In addition, these publishing houses try to satisfy their authors (or literary agents), who want a say in the choice of the pocket-format publisher of their book, and whose main selection criteria are, apart from good relations with their publisher, the amount of advances offered and the widest possible marketing of books. Authors, therefore, will have neither the incentive nor the means to prevent the merged entity from acting independently of them, as has already been explained in regard to the countervailing power of the suppliers on the primary markets for the purchase of the rights in general literature titles.

(550) The combination of these factors leads to the conclusion that the suppliers have no genuine countervailing power.

(4) CONCLUSION

(551) The notified transaction would create a dominant position as a result of which effective competition would be significantly impeded in the Common market or in a substantial part of it in the secondary market for the pocket-format rights in general literature titles.

(552) The merged entity would have a dominant position not only with regard to the purchase of the secondary rights in new works, but also with regard to its competitors' lists. Authors have a definite interest in seeing their books published in a single pocket-format series, even if this does not always happen (438). At first sight this finding might suggest that the merged entity would not be able to capture authors hitherto published by competitors, but the reality is quite different, since secondary rights are regularly renegotiated and reassigned. As the notifying party emphasises, "contracts for the sale of pocket-format publishing rights are always concluded for a limited period (five to ten years), whether the author is foreign or French. Such rights can therefore be regularly renegotiated and reassigned" (439). The fact that authors want some continuity in the series publishing their works in pocket format is not an insurmountable obstacle that would prevent them changing publisher, since they can get their new publisher to buy the rights in their old works, or can be published in pocket format in several series.

A.2.b. THE SECONDARY MARKET FOR CLUB RIGHTS

(1) MARKET SHARES

(553) According to the information supplied by the notifying party (440), the merged entity would have slightly more than 35% of the market for the sale of club rights.

(554) Judging by the purchases of club rights from the various publishers by France Loisirs and GLM in the last three years, it is clear, however, that the market share would in fact be [45-55]%, since Albin Michel is the nearest competitor with about [5-15]% of the market. The entity would therefore have strong positions compared with that of its nearest competitor.

(2) COUNTERVAILING POWER OF THE BOOK CLUBS

(555) In the secondary market for the sale of pocket-format rights, demand consists of only two book clubs.

(556) The largest is France Loisirs with 3.7 million members, which is wholly owned by the Bertelsmann group and which, according to the notifying party, accounts for at least 80% (441) of book club sales.

(557) GLM, jointly controlled by Albin Michel (50%) and Club Français du Livre (50%), has fewer than 1 million members, and a market share of less than 20%.

(558) The relation between publishers and the clubs is one of assymetric power, given the market position of France Loisirs and the dependence of the publishers, which need the exploitation of this derived right. It should be noted that France Loisirs has been censured in France for having imposed on publishers contractual terms that amounted to an abuse of a dominant position (442).

(559) The book clubs therefore have sufficient countervailing power to frustrate any creation or strengthening of a dominant position for the merged entity in the secondary market for club rights.

(3) CONCLUSION

(560) In view of these various factors, the vertical integration and strong presence, or dominant position, of the merged entity in the different markets of the book chain are not such as to create a dominant position for the merged entity in the secondary market for club rights.

B. MARKETS FOR THE SALE OF MARKETING AND DISTRIBUTION SERVICES TO PUBLISHERS

(561) Marketing and distribution are vital links in the book chain: on marketing depends access to customers, and hence sales volumes and prospects of profit, while on distribution depends the proper delivery of orders and the collection of sales. As well as being markets for the benefit of third parties, marketing and distribution are therefore strategically important for the book economy, through the vertical impact which they have on markets for the sale of books. As Seuil explains, "the life of a work is determined mainly by the bookseller's tables. For books of equal quality, it is their position on the tables that determines their sales, and hence their publisher's results" (443). This analysis is shared by all market players. In its investigation, therefore, the Commission analysed both the impact of the transaction on the marketing and distribution markets for third parties and the vertical impact of marketing/distribution on the markets for the sale of books.

(562) The markets for marketing and distribution services are analysed on the basis of the consolidated data for the French-speaking area of the Community (France, Belgium and Luxembourg). It is not necessary, for the purposes of analysing the notified transaction, to include the Swiss market for French books in the area considered. First, with 1.5 million potential consumers, the market is small compared with the whole French-speaking area of the Community. Second, about 80% of the books sold in French-speaking Switzerland come from France, so that the market shares of the different publishers would not change significantly as between Switzerland and the French-speaking area of the Community on account of the Swiss publishers. Adding Switzerland to the market for the marketing of books in the European French-speaking area would not have an impact on the positions of the various players in the market. It seems appropriate, therefore, to carry out the competition analysis on the basis of the market that combines France, Belgium and Luxembourg.

(563) Since the notifying party did not provide figures for each class of dealer, the Commission has collected these data from the largest marketers/distributors in the market in order to build a picture of the relevant market shares (444).

B.1. INTRODUCTION

(564) As regards the market for marketing and distribution services provided to outsiders, it is necessary to explain how competition operates in these markets, in particular since marketing and distribution are closely linked by their sequence and their interaction within the book chain. First, as has already been explained above, the competitive position of the different players in the markets for marketing and distribution services provided to outsiders is linked to their overall position, including their own sales.

(565) Second, since a publisher generally uses a single provider for marketing and distributing its production to all classes of dealer, although some publishers do their own marketing to some outlets445, marketing and distribution services are to some extent complementary. The complementarity between them is clearly manifested in the fact that the contract that the publisher concludes with the service provider for both marketing and distribution is a single contract, and an exclusive contract at that.

(566) The Commission's investigation shows that publishers using these services do not choose the marketer and distributor for their works separately (446) but, apart from publishers who themselves market to certain outlets, choose as their distributor the publisher who markets their books (447). In addition, the marketer and the distributor, apart from the fact that they are one and the same in practice, are exclusive providers of these services to their publisher customers.

(567) The very tight link between the provision of marketing services and the provision of distribution services in practice means that it is not possible to offer marketing services without the associated distribution services, and that the two kinds of service are provided under one and the same contract: the result is that a strong position in marketing services will have an impact on the competitive position of a marketer in the market for distribution services provided to publishers.

(568) Marketing services for sales to bookshops, wholesalers and hypermarkets constitute, as indicated in the section on market definitions, separate relevant markets; these services may be sold jointly to a publisher under an overall contract, but account has to be taken of the respective positions of a marketer on each of these different markets in order fully to assess each of these competitive positions individually.

B.2. MARKETS FOR MARKETING SERVICES

(569) The notified transaction would lead to the creation of a dominant position for the merged entity on all the markets for marketing services on behalf of outsiders, through a combination of horizontal, vertical and conglomerate effects.

B.2.a. HORIZONTAL EFFECTS

(a) Players

(570) The providers of marketing services are distinguished principally by the extent of their vertical integration in the sale of books to outlets, and in particular class 3 outlets, and by the specialisation of their catalogues.

(571) Thus, unlike the other marketers, Hachette Livre and VUP are vertically integrated into the sale of books by wholesalers to class 3 retailers, through their respective wholesale structures. They have large and diversified catalogues, which would become more so following the proposed transaction.

(572) Alongside these two leaders, there is a second group with medium-sized marketing structures (Gallimard, Flammarion and Seuil) which is focused on general literature but is also active in other types of work. However, unlike the merged entity, these marketers do not have an integrated wholesale structure and are therefore obliged to use external wholesalers, including those of Hachette Livre and VUP, so that their works can reach class 3 outlets.

(573) Lastly, specialist marketing structures (Diff-edit, Sofédis, Etai, and Dilisco) generally focus on a limited number of categories of work and related channels.

(574) By way of illustration, the following chart shows that, compared with total sales of books on all markets together, most marketers specialise in a few categories of book, with the notable exception of Hachette Livre (448) and VUP, which are present in all categories.

Breakdown of sales by category of book (group + outsiders marketed)

<emplacement tableau>

(575) Conversely, publishers using marketing and distribution services are also distinguished by certain features. To begin with, they do not have sufficient critical mass to have their own teams of representative for each class (bookshops, hypermarkets or wholesalers). Second, they tend to be niche players, specialising in quite specific types of book: general literature, guides and manuals, children's books, strip cartoon albums, etc.

(2) MARKETING SERVICES, ALL LEVELS

(576) The merger between Hachette Livre and VUP eliminates competition between the two largest marketers and upsets the market structures by creating an entity which is very much larger than its direct competitors. This overall negative effect can be seen market by market to varying degrees, depending on whether the services are provided for sales to bookshops, where the horizontal effect is significant, or to hypermarkets and wholesalers, where it is very great.

(577) According to the notifying party's calculations, the combined share in 2001 of all marketing services provided to outsiders, all retail classes combined, held by Hachette Livre and VUP, in the French-speaking area of the Community, was about [35-45]% (449) ([25-35]% for Hachette Livre and [5-15]% for VUP). The Commission calculates the position of the merged entity to be about [40-50]% ([25-35]% for Hachette Livre and [5-15]% for VUP), i.e. at a similar level to that identified by the notifying party. The nearest competitors are Seuil ([5-15]% of the French market, according to the notifying party, [10-15]% in the French-speaking area, according to the Commission), the Gallimard group ([5-15]% of the French market, according to the notifying party, [5-10]% in the French-speaking area, according to the Commission), and Flammarion ([0-10]% of the French market, according to the notifying party, [5-10]% in the French-speaking area, according to the Commission).

(578) The merged entity would therefore have a total strength in marketing services provided to outsiders about three times as large as its nearest competitor. The Herfindahl-Hirschmann Index ("HHI") for all marketing services to outsiders, all classes combined (450), would be more than 2 400 after the transaction, according to the Commission's calculations, which is an increase of more than 800.

(579) Including all in-house and external sales, and on the basis of sales expressed in terms of the list price exclusive of tax in France, Belgium and Luxembourg, the combined share of the merged entity in 2001 is, according to the notifying party, and likewise according to the Commission, [40-50]% ([15-25]% for Hachette Livre and [15-25]% for VUP). The Commission estimates the shares of Gallimard, Flammarion and Seuil, on the same basis, to be [5-10]%, [0-5]% and [0-5]% respectively. No other competitor exceeds 3% of all marketing services, all classes combined. These findings indicate high levels of concentration.

(580) In its reply to the statement of objections, the notifying party considers that the combination of market shares is not a relevant indicator for the analysis of competition in marketing, because the market shares aggregate independent marketing forces; to this effect, it cites the example of Robert Laffont's marketing forces, which operate separately with regard to class 1 bookshops. However, the Commission notes that the marketing forces of the Hachette and VUP groups have pooled their efforts (451), and that even Robert Laffont uses teams common to other VUP publishers for marketing to hypermarkets and wholesalers; it is not correct, therefore, to say that the marketing forces are independent and cannot be aggregated.

(581) Another argument put forward by the notifying party is that the market for marketing services operates in the manner of a system of tendering, and that market shares are not appropriate for assessing the competitive position of the players. Apart from the fact that this argument was not included in the notification form, it should be pointed out that the Commission's investigation did not identify any formal invitations to tender for marketing/distribution contracts; instead, publishers changed their providers in response to approaches made to them by other marketers/distributors. Marketing/distribution contracts generally included a clause providing for tacit renewal. In addition, it is the same teams that will present the works of all of a marketer's customers, which would conflict with the idea of a specified service offered via a tendering system. Lastly, the market shares used by the Commission represent the flows that transit through the marketing system of the different players on the market (452). By taking account of sales for in-house publishers, these market shares thus reflect a marketer's capacity to provide its customer with access to the retailer's shelves and its sales negotiating power vis-à-vis dealers in a given class. By taking account of sales to outsiders on their own, the market shares indicate the capacity of a marketer to offer services to outsiders and its power of attraction.

(582) In this respect, the market shares of the parties to the merger and of the other marketers make it possible to give an opinion on "the existence of a sufficient number of marketers that can offer attractive services", which indication according to the notifying party (453) should be of use in examining the impact of the transaction. The Commission nevertheless agrees with the notifying party that the main criteria of publishers with regard to the choice of marketers should also be included. This is what it has done during the investigation and, as the notifying party observes, the first criterion for selecting a marketer is its capacity to cover the different classes of customer. This criterion is examined below in the analysis of the vertical and conglomerate effects of the transaction on the markets for marketing on behalf of outsiders. As regards the horizontal effects of the merger, however, the Commission maintains that adding market shares - for outsiders and including own-account services - is a relevant factor in assessing the market power of the new entity.

(583) As well as radically transforming marketing structures, as stated above, the transaction would eliminate competition between the two largest marketers and, what is more, between the only competitors with integrated wholesaling structures active in the downstream market for the sale of books by wholesalers to class 3 outlets. Before the merger, as will be illustrated below, Hachette Livre and VUP are the two players providing the best access for their customers to hypermarkets and wholesalers. The Commission's investigation has also shown that Hachette Livre and VUP have offered each other stiff competition. One bookseller states with regard to these activities, "the Hachette group has dominated the situation for a long time, and the recent creation of the Vivendi group weakened this position, which to our mind explains the great advantage of this operation to Hachette, which had a competitor of its own stature" (454). Thus of the 24 publisher customers won by Hachette Livre between 1999 and 2002, 18 come from VUP (455), one from the Gallimard group and one from Médias Participations, while four are recently established publishers. VUP, for its part, has won Michel Laffon from Hachette. Conversely, over the same period, only eight marketed customers left, six of them because they became insolvent (456). The case of Albin Michel, which is the largest publisher that is a customer for marketing and distribution services provided to outsiders, amply illustrates the way the groups could be played off against each other in order to obtain better conditions; Albin Michel changed its service provider, switching from VUP to Hachette Livre.

(584) In its reply to the statement of objections, the notifying party disputes the importance of the rivalry between Hachette and VUP, on the grounds that there are many transfers between small and medium-sized structures. But this submission is manifestly flawed in regard to the analysis of the competitive impact of the notified transaction: the task here is to examine how far the transaction would enable Hachette-VUP to behave independently of its competitors and its customers, i.e. to examine what competition Hachette-VUP would be subjected to, and not to examine the remaining competition between the other players on the market. As was explained above, the characteristics of the market for marketing services are differentiated; transfers between Hachette and VUP are incontrovertible proof that there is competition between these marketers, which are very close substitutes, because of their large size and their integration into a wholesaling structure. This competition would disappear following the merger, and the fact that medium-sized structures would compete with each other is not relevant to this analysis. The only relevant factor is the competitive pressure which the latter would exercise on Hachette and VUP. It will be shown below that this is not sufficient to prevent the creation of a dominant position.

(585) The notifying party maintains, moreover, that the merger would have no impact on markets for marketing on behalf of outsiders. The transaction, it claims, would involve no significant increase in marketing coverage equivalent to Hachette Livre's before the transaction and, consequently, would not be very likely to make competing marketers less attractive. It should be pointed out here that the notifying party is distorting the arguments of the Commission when it states that the latter is basing its reasoning on an increase in Hachette's and VUP's coverage. As mentioned for example in paragraph 451 of the statement of objections, the Commission's investigation has shown that it is not coverage as such, but rather access to the different levels which, for a publisher, is a criterion for selecting a marketer.

(586) The merger leads to a radical strengthening of the new entity, compared with other marketers, in its capacity to access outlets. The new entity would have more than [...]* representatives ([...]* for Hachette and [...]* for VUP), who made more than [...]* visits to outlets in 2002 ([...]* for Hachette and [...]* for VUP) (457). This change has nothing to do with the coverage of the two entities, which were competitors and therefore visited the same outlets. The merged entity would also have an integrated wholesaling structure, consisting of representatives and sales offices, giving it easy access to class 3 outlets, whereas its competing marketers would have to use independent wholesalers and the wholesaling structures of Hachette and VUP. After the merger, the new entity would have very large resources with which to increase commercial pressure on outlets and would account in particular for a considerable share of their turnover, which would make it all the more attractive for outside publishers wanting to use its marketing services. In addition, the merger, by integrating the two main wholesaling structures in the market, would eliminate one channel to class 3 outlets for other marketers and make access to the remaining independent wholesalers more difficult.

(587) Lastly, the notifying party maintains that the Commission contradicts itself with regard to the segmentation of marketing services by class, since marketing rates are uniform. But it is making an error of economic analysis here, since the total price of a package of services may quite easily comprise differences in the values placed on its components by customers. Further, the Commission's investigation shows that, in the case of marketing/distribution contracts of the "dealer" type, rates are differentiated by class (458).

(3) MARKETING TO BOOKSHOPS (CLASS 1 AND 2)

(588) The combined market shares of the merged entity on the market for marketing services for sales to class 1 and 2 bookshops are [25-35]% ([15-25]% for Hachette Livre and [5-15]% for VUP) as against [15-20]% for Seuil, [10-15]% for Gallimard and [5-10]% for Flammarion. The HHI after the transaction is just below 1 800, an increase of 465.

(589) Taking in-house sales into consideration, the merged entity accounts for [35-45]% of marketing at this level ([15-25]% for Hachette Livre and [15-25]% for VUP), as against [5-10]% for Gallimard, Seuil and Flammarion and less than 3% for the other marketers, i.e. more than four times the share of its nearest competitor (459).

(590) The notifying party emphasises that from market shares alone the Commission cannot deduce the existence of a dominant position in marketing to class 1 and 2 dealers. It should be noted here, however, that the total strength of the new entity with regard to bookshops would be very great on this market, and the gap with its direct competitors obvious. On this basis, the Union of French Booksellers (Syndicat de la Librairie Française, SLF), which represents independent booksellers (460), thinks that "the creation of a dominant position in the market for the sale of marketing services supplied to class 1 and 2 bookshops can already be discerned".

(591) The Commission would point out in this respect that a marketer's power to attract outside publishers is also affected by the overall position with regard to retailers.461 As is explained below in the analysis of the vertical effects of marketing on sales markets, the marketer establishes the general terms of business for the works which it markets (in-house and outside publishers) to retailers. These general terms of business determine the bookseller's remuneration in accordance with a set of parameters, and contain incentives. In Hachette's general terms of business for booksellers, an additional discount of between [...]%* is granted, if the bookseller follows the recommendations (concerning titles and quantities) of the Hachette group. This has an effect on priority access for works, as the SLF explains very clearly: "a bookseller who intends to "maximise" the discount from the Hachette group (and all the more so from Hachette-VUP after the merger) will have an interest in following the recommendations of the Hachette group in terms of titles and quantities and also in keeping works on the shelves for as long as possible so as to try and limit returns. This will inevitably tend to keep the products of other publishers/marketers away, since the bookseller's shelves will be almost entirely occupied by the works of the merged entity" (462).

(592) In conclusion, the Commission considers that the strong position of the merged entity as a marketer vis-à-vis bookshops helps to create a dominant position on this market for marketing services. This strong position, moreover, produces vertical effects on the markets for the sale of books, which will be analysed below.

(4) MARKETING TO HYPERMARKETS

(593) According the Commission's calculations, the merged entity's share of the market for marketing services for outside publishers selling to hypermarkets is [55-65]%, as against [5-10]% for Gallimard, [0-5]% for Flammarion and Seuil and less than 2% for the other marketers. The notified transaction would therefore mean that the merged entity would be about ten times as strong as its nearest competitor. On this market, the transaction has an HHI of 3 976, with an increase of 1 408, which is a very high level of concentration.

(594) Taking in-house sales into account, the merged entity's market share is [45-55]%*, as against [5-10]% for Gallimard, [0-5]% for Flammarion and Seuil, and less than 3% for the other marketers.

(5) MARKETING TO WHOLESALERS

(595) The merged entity has [55-65]% of marketing services for outside publishers selling to wholesalers ([40-50]% for Hachette Livre and [10-20]% for VUP). By contrast, Seuil has a market share of [5-10]%, Gallimard and Flammarion both have [0-5]% and all the other players are below 2%. On this market, therefore, the merged entity would be seven times as strong as its direct competitors. Here the transaction generates an HHI of 4 077, an increase of 1 600. The transaction radically transforms the structures of the market, therefore, by eliminating a direct competitor, which is also the only competitor which has an integrated wholesale structure on the downstream market for the sale of books by wholesalers to class 3 outlets.

(596) Taking in-house sales into account, the merged entity's market share is [45-55]% ([25-35]% for Hachette Livre and [15-25]% for VUP) as against less than 5% for all the other competitors. It is remarkable that its share of the market in marketing services for outside publishers selling to wholesalers is higher than its share when in-house publishers are included. This indicates that its power to attract is much greater than that of its rival marketers.

B.2.b. VERTICAL AND CONGLOMERATE EFFECTS

(597) Since marketing services for sales to the different classes of dealer are the subject of a single, exclusive contract when they are sold, even selectively (463), the competitive position of a marketer in one of these markets will have an impact on its position in the other markets for marketing services.

(598) Since the first criterion for selecting a marketer mentioned by publishers is its capacity to penetrate all the different networks of dealers, including class 3 outlets (464), vertical integration in a wholesale structure is a discriminating factor in the choice of marketer and helps to strengthen a marketer's competitive position in the market for the sale of marketing services, in particular to publishers selling to wholesalers.

(1) MARKETING TO BOOKSHOPS (CLASS 1 AND 2)

(599) Since marketing to hypermarkets and wholesalers is essential to a publisher's ability to make sales in those places, which, in addition to bookshop sales, will enable it to ensure profitability and hence viability, and since it is established practice to combine all marketing to the different classes of dealer in a single contract, and an exclusive contract at that, the merged entity's dominant positions in the markets for marketing services for sales to hypermarkets and wholesalers would contribute to the creation of a dominant position in the market for marketing services for sales to bookshops.

(600) With regard first of all to the importance of marketing to hypermarkets and wholesalers, the Commission's investigation reveals that, although books are sold mainly through class 1 and class 2 bookshops (which account for about 60% of sales on average, apart from clubs and direct sales), sales in hypermarkets and to wholesalers account for a considerable share of a publisher's total turnover and, therefore, a publisher will not be able to achieve its full sales potential if it is not marketed on all these levels.

(601) The notifying party, in its reply to the statement of objections, considers for its part that the Commission overestimates the importance of access to wholesalers, for two reasons: (i) generally speaking, small outlets and wholesalers are not more important for best-selling titles than for other titles, and (ii) reduced access to wholesalers would not mean loss of turnover.

(602) The Commission considers nevertheless that access to wholesalers has considerable importance for publishers, and that its investigation clearly shows this. Although, as the following chart shows465, the volume of sales in hypermarkets and to wholesalers differs according to the category of work in question, such sales are significant for a substantial proportion of the works sold in this market. While in the case of academic and professional books and school textbooks a publisher may possibly be content to be marketed in bookshops, since this is where it will find nearly all the demand for these categories (466), in the case of large-format general literature or guides and manuals sales in hypermarkets and to wholesalers may account for over 30 percent of the revenues generated by such works over all classes of dealer.

Breakdown of book sales by class of dealer

<emplacement tableau>

(603) Sales to wholesalers appear all the more essential when related to the number of titles which generate them. Although the outlets supplied by wholesalers generally achieve a small turnover, that turnover is concentrated on a limited number of titles sold by many outlets. Thus, LDS.s customers, which are about [...]* in number, sold an average of some 1 500 books in 2001. Of the more than 200 000 Seuil books sold by LDS in 2002, 50% were accounted for by 10 titles, and 80% by 35 titles (467).

(604) The importance of class 3 for a very large number of titles is undeniable. One could give many more examples of works which achieve between 20% and 40% of their sales through this channel: the Grasset title Les Ombres Errantes achieved 22% of its sales through LDS; Tcho la planète, Titeuf, 39%; Le Parfum 36% in Le Livre de Poche; while the Guide Hachette des vins de France achieved 25%. In its reply to the statement of objections, the notifying party observed that it is not only the 20 best-sellers which achieve a considerable proportion of their sales through wholesalers. This does not mean, however, that the wholesale channel is not important, but rather that it accounts for more than the 20 best-sellers of certain publishers. The works in question are, as VUP explains, the 20% of publishing production which achieves 80% of sales and which wholesalers select in accordance with their own criteria (468).

(605) This observation about the importance of sales to wholesalers also applies to sales in hypermarkets, which constitute one of the largest growth points in terms of retail sales (469). Books sold in hypermarkets are also among the works that sell well (470). An analysis of the 1 500 best-selling general literature titles according to Ipsos shows that nearly 40% of sales of these 1 500 titles (471) are achieved outside class 1 and 2 bookshops (472) and that, among these best-sellers, certain individual titles achieve more than 50% of their total sales (473) through hypermarkets and wholesalers (474).

(606) Thus mere access to bookshops is not enough for certain books to achieve their full sales potential. Consequently, the sales made in hypermarkets and to wholesalers may be regarded as essential for a publisher's profitability and viability: without access to these classes of customer, a publisher will suffer losses in sales and hence in turnover, with a detrimental effect on its general profitability.

(607) If one considers the titles that sell well in wholesalers and hypermarkets, one is bound to acknowledge that they are of considerable importance for the general economy of several publishers. For many publishing houses, a limited number of successful titles generates a large share of their turnover (475). In addition, however, the additional margin obtained through sales in wholesalers and hypermarkets can be very significant. To give only one example, Patricia Cornwell's book Dossier Benton (Le Livre de Poche) had a gross margin on sales in wholesalers and supermarkets of euro [...]* and more than euro [...]* in hypermarkets, compared with less than euro [...]* in bookshops. Taking all sales of the publishers marketed and/or distributed by certain competitors of Hachette and VUP, one finds that for a great many of them sales to wholesalers and hypermarkets are considerable, especially when compared with the average operating margin (the average operating margin of Hachette Livre for example from the sale of literature, children's books or guides and manuals varies between [0-10]% and [0-10]% of the corresponding turnover (476)).

<emplacement tableau>

(608) Consequently, a loss of sales resulting from the loss of access to hypermarkets and wholesalers can bring a publisher below break-even point and seriously threaten its economic viability. However, sales forces that give sufficient access to small outlets served by wholesalers can generate additional turnover, at that level alone, of 5% to 10%, which for the same investment can make a work profitable rather than unprofitable (477).

(609) The notifying party produces a second argument against the importance of sales to wholesalers and hypermarkets, based on the porosity of the sales channels: it is not important for a publisher to have its books in hypermarkets and class 3 outlets, since consumers can find them in the bookshops anyway.

(610) It should be noted at the outset that this argument clashes with Hachette's entire strategy on this market. The firm maintains LDS, when that organisation is plainly not profitable, and must therefore find an economic reason for doing so: why go to the lengths of investing in a costly structure so as to be present in all small outlets, if readers are going to find the works they want in class 1 bookshops anyway? In reality, this argument about the porosity between the different sales channels tends erroneously to present books in economic terms as "experience goods" (478). This is what the SLF explains in its comments on Lagardère's reply: "losses in one channel are definitive losses, since it is fairly rare for consumers to buy a particular title in advance - Thus, apart from a few titles which are promoted intensively, the customer in hypermarkets, supermarkets and small outlets will only buy what is present".

(611) Purchases in hypermarkets and class 3 outlets are, for the most part, not "considered" purchases but more impulse buys, meaning that the consumer does not decide in advance which book to buy. In such a situation the argument used by the notifying party is purely theoretical: it consists in saying that a consumer about to buy an Hachette book in a hypermarket will abandon that purchase in favour of a book from an Hachette competitor which is not in the hypermarket but somewhere in another shop, when it is very probable that the consumer (479) is not aware of the competitor's book, because he or she cannot see it, and therefore is not entertaining the possibility of buying elsewhere at all. Similarly, Lagardère obscures the importance of the retailer, who can promote works by making them visible. The increased power of the new group compared with its competitors would automatically give it preferential access to retailers, not only at wholesaler and hypermarket level but also with bookshops.

(612) In light of the foregoing, the Commission maintains its position that a marketer operating for an outside publisher cannot be content to be present just in bookshops, as the Flammarion group in its capacity as a marketer emphasises ("only publishers that are absolutely convinced they will never have a major success with their books could accept that they should be marketed in bookshops only" (480)), since a marketer's customers have every incentive to sell (and hence to be marketed) at all levels (481).

(613) Lastly, the notifying party, in its reply to the statement of objections, contends that the Commission has not tried to show that the new entity's competitors would be less efficient in marketing to hypermarkets and/or wholesalers after the merger (482). In order to be more specific, the Commission will here set out the reasons why the operation would make access to the different outlets, in particular hypermarkets and wholesalers, more difficult for other marketers.

(614) As was explained above, there is a link between a marketer's position in regard to retailers and its position as a supplier of services to publishers. As is explained in the analysis of the markets for the sale of books, the notified transaction radically changes the balance of power between the new entity and the retailers it canvasses, since after the merger the new entity would account for a considerable share of their turnover: to the combined sales of the two merged entities would be added the combined sales of the firms which they respectively market. The general terms of business, which are determined by the marketer, contain incentives aimed at developing exclusive links with retailers.

(615) Such is the case in particular with hypermarkets. The merged entity's strong position vis-à-vis hypermarkets would create incentives, already mentioned above in the case of bookshops, to maximise the discount obtained from Hachette-VUP by following the group's recommendations as regards taking part in promotions, book displays, following advice and increasing turnover, all of which are means of giving the new entity more space or a better position on hypermarkets' shelves, to the detriment of other marketers (483). The same phenomenon is to be expected in the case of wholesalers which generate a substantial share of their turnover with the new entity and/or which depend on it for their supply. Over and above this, however, as explained in the analysis of the market for sales by wholesalers, the new entity itself owns the dominant wholesaling structure.

(616) Consequently, the capacity of other marketers to gain satisfactory access to hypermarkets and wholesalers for the publishers which they market would be reduced by the notified transaction.

(617) It is in this way that the dominant positions of the merged entity in the marketing-service markets for sales to hypermarkets and wholesalers would contribute, over and above the horizontal effects described earlier, to the creation of a dominant position in the marketing-service market for sales to bookshops.

(2) MARKETING TO HYPERMARKETS

(618) Since, as was discussed in the context of the creation of a dominant position in the marketing-service market for sales to bookshops, marketing to wholesalers is essential to a publisher's profitability, and since it is the established practice to combine all marketing to the different classes of dealer in a single, exclusive contract, the merged entity's dominant position in the markets for marketing services for sales to wholesalers would help to create a dominant position in the market for marketing services for sales to hypermarkets.

(3) MARKETING TO WHOLESALERS

(619) The dominant position of the merged entity in the market for the sale of books by wholesalers to class 3 outlets (supermarkets and small outlets) would help significantly to create a dominant position for it in the marketing-services market for sales to wholesalers.

(620) The merged entity's dominant position in the sale of books by wholesalers to class 3 dealers would enable it to facilitate access to wholesalers by its integrated marketing teams. Other marketers, which unlike the merged entity are not integrated into a wholesaling structure, would not have this ability. The transaction therefore threatens to make access to the dominant wholesalers, i.e. the structures integrated with the merged entity, very difficult for marketers competing with it, thus rendering competing marketers significantly less attractive than the merged entity's own marketers. Its own marketers would be vertically integrated into the wholesalers' activity and, hence, in a position to control the marketing of a customer publisher's books among wholesalers.

B.2.c. COUNTERVAILING POWER

(1) EXISTING COMPETITION

(621) Apart from the fact that their positions are clearly inferior to that of the merged entity on the different markets for marketing services, in particular for sales to hypermarkets and wholesalers, competing marketers cannot claim to exert real countervailing power and thus try to compete with the merged entity's dominant position, given the size of its captive marketing activities, and given that competitors are not vertically integrated into a wholesaling structure.

(622) In its reply to the statement of objections the notifying party maintains that the captive activities are not relevant to the analysis of competition in marketing and that there is no barrier to expansion in marketing.

(623) There is an obvious inconsistency in the position of the notifying party: on the one hand it claims that there is no market for services to outside publishers and that shares in the market for marketing services must include sales of books published by in- house publishers (484), and on the other that such sales should not be taken into account for the competition analysis. The Commission's view is that sales for in-house publishers have an impact on the competitive position in markets for marketing services for outside publishers and that there are barriers to expansion in marketing activities linked principally to opportunity cost and the time required for investment in additional capacity.

(624) The market survey shows that marketing costs increase in steps (485): there is a minimum size for a team to be viable, because of the number of customers and their geographical distribution, and representatives can increase the number of titles they market only marginally, with the result that more representatives must be taken on if additional titles are to be marketed (and with the problems of coordination, at a certain point it becomes necessary to form new teams, etc"). This means that the competitors of Hachette-VUP, which are small in size (sales of their own titles account for less than 7% of the total market) and hence have restricted absorption capacity with the present number of representatives (486), would have difficulties in taking on new customers immediately and would first have to invest before being able to offer additional services. Besides, the number of representatives cannot be increased overnight (487).

(625) The Commission's investigation then identified barriers to change (488) (mainly in connection with distribution, but having an impact on combined marketing/distribution services). One barrier to change mentioned by publishers and marketers alike concerns the management of returns after termination of a marketing/distribution contract: retailers must be recredited by the marketer/distributor for the returns of publishers who change supplier, which can impose a sizeable burden on cash-flow that has to be financed by the publisher if the marketer/distributor is not prepared to make a gesture (489). Furthermore, as has already been explained, marketing and distribution are associated services that are provided under a single exclusive contract. To take on a new customer for its marketing service, therefore, a marketer needs not just additional marketing capacity but also additional distribution capacity; and it will be shown below that here too there are appreciable barriers to expansion.

(626) What is more, the customers of Hachette and VUP are even more dependent than those of the competitors of Hachette and VUP in relation to hypermarket and wholesaler channels, as shown by the chart below. (490) This dependence explains why they choose to be marketed by Hachette or VUP, who guarantee them best possible access to small sales outlets and to hypermarkets, and would make it difficult for them to turn to another marketer unless it was one which could offer them comparable access.

<emplacement tableau>

(627) As has been shown in the analysis of the merged entity's dominant position in the market for the sale of books by wholesalers to class 3 outlets, it is clearly not reasonable to think that at this stage marketers competing with the merged entity could begin selling books to class 3 outlets by setting up their own vertically integrated wholesaling structures. The barrier to access to wholesalers, in addition to the points already identified in terms of appropriate supply and reputation, would consist, after the merger, in the ability of Hachette-VUP to handicap its marketing competitors in their access to its own wholesale structures (in a dominant position, it will be recalled) or make it more difficult for them to secure access to outside wholesalers, who will have become dependent on Hachette/VUP for their supplies. To get around this, competing marketers would have to set up their own structure for class 3 outlets.

(628) The establishment of an efficient wholesaling structure requires the generation of enough turnover to make the network that is needed for effective coverage of class 3 outlets profitable. To do this, a wholesaler must have a body of works to sell, either in-house through the publishers of the group to which it belongs or through outside publishers, which is not only sufficiently large but which also, and above all, matches the categories of work traditionally marketed to this class of customer (i.e. mainly best-sellers, pocket-format books, strip cartoon albums, etc")

(629) Although the main marketers competing with the merged entity are all vertically integrated into the sale of books in their capacity as publishers, it is clear that the size of the investment in representatives and an appropriate network of salesrooms, and the inadequacy of their respective ranges of works and sales volumes have not so far allowed them to contemplate setting up their own wholesaling structures. The notifying party itself states that this activity is a loss-maker today; the merger, by creating a dominant position for the new entity as a wholesaler, would make such an undertaking even more risky.

(630) Lastly, concerning the possibility of setting up a grouping of independent marketers as a joint alternative wholesaling structure, on the lines of the existing Livre Diffusion economic interest grouping, this is an unrealistic proposal for several reasons.

(631) To begin with, Livre Diffusion itself is an association between Gallimard, Flammarion and Seuil which manages three logistical structures located only in France (Paris, Nantes and Lyon), which are of the depository type and thus not similar to a salesroom such as those owned by Hachette Livre and VUP (491). Furthermore, as experience has shown, even before the notified transaction, the marketers competing with Hachette Livre and VUP were not capable of operating such a wholesaling structure. Even if it is considered that Livre Diffusion could be swiftly converted to a salesroom system, it could not constitute an operational wholesaling structure, since - and this is confirmed by the way the merged entity is organised - to cover the French-speaking area of the Community calls for a minimum number of salesrooms492, which is a sizeable investment on top of the requisite sales teams.

(632) Moreover, it is difficult for the Commission to regard as a valid alternative, and hence an effective obstacle to the merged entity's dominant position, any speculation about the feasibility of a reaction of this type to the transaction, amounting to a radical alteration of the marketing structures of the merged entity's main competitors. And even if it were conceivable, this additional concentration of the market for the sale of books by wholesalers would still not help to ensure any countervailing power for the marketers competing in the markets for the provision of marketing services. To do that, as has been explained above, the new alternative concentrated wholesaler would also have to be concentrated in the sale of books by publishers to wholesalers, so as to have a sufficiently wide and appropriate range of published books to cover the costs of marketing. Gallimard, Seuil and Flammarion do not together have a complete supply, since they are not present on the markets for the sale of educational supporting materials and reference books and do not have such large, diversified catalogues as Hachette Livre and VUP combined. For this reason, a wholesaling structure set up on the initiative of these participants would need to buy large quantities from other publishers, especially Hachette Livre and VUP. Moreover, the three marketers have highlighted the practical difficulties that such a structure would entail and the clashes between their respective works which would result: small outlets can absorb only a tiny proportion of new titles coming out each week; a wholesaler is therefore obliged to choose constantly between these new titles. The choices would become particularly awkward - and a constant source of strain - in a structure shared by a number of publishers. In order to overcome all the potential problems of discrimination between the books marketed by this joint structure, which as a joint structure would be economically inefficient, it would ultimately be necessary to envisage merging the marketing teams, and therefore, in practice, the publishing houses. The Commission cannot base its competition analysis on a hypothesis which is so difficult to imagine and so lacking in credibility. The parties directly concerned have confirmed the Commission's analysis on this point by voicing doubts about the permanence of a financial structure combining independent publishers of very diverse sizes and structures, which therefore, according to them, would not have an economic rationale (493).

(633) All these factors prompt the Commission to conclude that it is not reasonable to consider this possibility, which indeed has so far never materialised, as a practicable exertion of countervailing power in regard to the merged entity by its existing marketing competitors.

(2) POTENTIAL COMPETITION

(634) As regards potential competition, the need to recoup high fixed costs from a significant mass of operations, and the size of the captive sales already accounted for by the merged entity and its main competing marketers, mean that operators likely to have an interest in entering the market for marketing services must reasonably regard this alternative as a difficult one, on the ground that it would not be economically profitable.

(635) It is hard to see what publishing house, other than those already present on the market, would be able, within a reasonable period, to set up a marketing structure comparable to that of the merged entity, since the barriers to entry involve establishing a sales force for the whole French-speaking area of the Community. This requires a minimum turnover to cover the costs of representatives for each class of dealer (494). In addition, to be able to exercise any competitive pressure at all on the merged entity, mainly on the market for the provision of marketing services for sales to wholesalers and, indirectly, on the market for the provision of marketing services for sales to bookshops and hypermarkets, any potential competitor will have to be able to access an integrated wholesaling structure, the need for which was explained in the above discussion of the pressure exerted by existing competition.

(636) The importance of barriers to entry and to expansion on each of the markets for marketing services is confirmed by the fact that there has been no entry or major expansion by a supplier of marketing services in France in recent years. In Belgium, new entries have essentially been due to publishers acquiring shareholdings in small local distributors.

(3) DEMAND

(637) In these markets, the demand side consists essentially of small-scale publishers, except for Albin Michel. The latter's incentives for opposing the merged entity are limited by virtue in particular of its shareholding in the Hachette Livre's pocket-format publishing house and its use of Hachette Livre's marketing structure. Whereas, in the past, Albin Michel has been able to profit from the rivalry between the two major marketing centres of VUP and Hachette Livre by switching from one to the other, the other marketers could not offer it identical access to hypermarkets and small outlets. Albin Michel's representatives confirmed this analysis ("as to the alternative options that [Albin Michel] would have after any merger between Hachette and VUP, if [Albin Michel] were to be thrown out in the street it would have no other possibility that would provide equivalent penetration of outlets" (495)). So even if Albin Michel would probably have the volume of sales and would be well enough known to do its own marketing to hypermarkets and wholesalers, this option would be even less advantageous than finding an outside marketer with an extensive catalogue, given that Albin Michel specialises essentially in general literature and these dealers prefer to work with a limited number of suppliers. What is more, it is clear that Albin Michel with its buying power alone would not be able to exert competitive pressure on the merged entity's dominant position (496).

(638) Lastly, concerning the possibility of publishers which are customers of the merged entity to exert competitive pressure on its dominant position by deciding to do their own marketing rather than use its services, the Commission's investigation shows that a publisher must overcome several obstacles in order to have its own marketing force, the chief one being the minimum turnover necessary for financing a team of representatives for each of the different classes of dealer (497). Thus, as long as a publisher has not reached the critical minimum thresholds enabling it to contemplate doing its own marketing, it will be impossible for it to exert any competitive pressure on the merged entity's dominant positions on the respective markets for the sale of marketing services for sales to the different classes of dealer. But the Commission's investigation has shown that the present customers of Hachette and VUP which do not do their own marketing do not have the size needed to do so themselves (498). They would therefore be dependent on the new entity after the merger.

(639) The combination of these factors prompts the conclusion that there is no real countervailing power among existing or potential competitors or customers.

B.2.d. CONCLUSION

(640) The notified transaction would thus create dominant positions as a result of which effective competition would be significantly impeded in the Common market or in a substantial part of it in the markets for the provision to publishers of marketing services for sales to bookshops, hypermarkets and wholesalers.

B.3. MARKETS FOR DISTRIBUTION SERVICES

(641) The notified transaction would lead to the creation of a dominant position for the merged entity in the markets for the sale of distribution services to outside publishers, through a combination and interaction of horizontal and conglomerate effects. The combination of the merged entity's substantial market shares and its strong vertical integration in the markets for the provision of marketing services for sales to the different classes of dealer would enable the merged entity to behave to an appreciable extent independently of its competitors and its customers, and ultimately of consumers.

B.3.a. HORIZONTAL EFFECTS

(1) MARKET SHARES

(642) The first effect of the transaction would be to deprive outside publishers of the principal alternative to the notifying party on this market. As well as the radical change the transaction would make in the market's structures, it would create a wide gap between the merged entity and its immediate competitors.

(643) In its reply to the statement of objections, the notifying party considers here too that market shares are not a relevant criterion for assessing the intensity of competition in the distribution business. As already explained in connection with marketing, the Commission argues that market shares, which represent the flows through the distribution channels of the various operators, both for outside publishers and with the inclusion of their own sales, give an adequate measure of the market power and the power of attraction in the distribution market.

(644) The combined market shares of Hachette Livre and VUP in distribution services provided to outside publishers in France, Belgium and Luxembourg are [35-45]% according to the notifying party (499) and also according to the Commission's calculations ([25-35]% for Hachette and [5-15]% for VUP) in 2001, as against [15-20]% for Gallimard, [10-15]% for Seuil and [5-10]% for Flammarion; the other competitors have less than 3% of the market. The merged entity will therefore be in a particularly strong position compared with its direct competitors. According to the Commission's calculations, the HHI after the merger will be over 2 500, with an increase of 800 as a result of the merger, i.e. the levels of concentration will be high.

(645) Including in-house sales, and on the basis of a market that includes France, French-speaking Belgium and Luxembourg, the merged entity's combined market shares for 2001 would, according to the notifying party, be [45-55]% ([20-30]% for Hachette and [15-25]% for VUP) and, according to the Commission, [45-55]% (([20-30]% for Hachette and [15-25]% for VUP) as against [10-15]% for Gallimard, [5-10]% for Flammarion and Seuil and less than 3% for the other competitors. The merged entity would thus be more than four times larger than its nearest competitor.

(2) OTHER HORIZONTAL EFFECTS

(646) As well as the classic horizontal effects identified above, the results of the Commission's investigation show that, given the importance of the merged entity as Prisme's biggest customer, with over one third of the volumes handled, the combination of its distribution activities on behalf of its own group's publishers and outsiders threatens, at this level, to contribute to the creation of a dominant position for the merged entity in the market for distribution services.

(647) Given that the merged entity would have the capacity, in terms of volumes and flows handled, to set up its own dispatching platform, it cannot be ruled out that it will leave Prisme, thereby helping to increase the costs of rival distributors, reduce their ability to compete, and thus create a dominant position for the merged entity in the market for distribution services. It should be emphasised in this respect that, for 2003, 75% of the costs of the Prisme platform generated by "outward" flows (i.e. consignments as such) are borne by the publishers/distributors (500). Any decision by the merged entity to leave Prisme would have the immediate consequence of reducing the volume of transactions handled and hence the economies of scale created by centralising parcels and consolidating consignments through Prisme (501).

(648) The notifying party maintains that Hachette and VUP would have no incentive to leave the Prisme platform. But in this it seems to confuse the role of Prisme and that of a carrier, when it states that .withdrawing from Prisme would increase the transport costs for books of the new entity much more than those of its competitors". It should be borne in mind that Prisme is a platform for breaking up consignments, where parcels from the different publishers are opened and then regrouped for dispatch to the addressees; it is then independent carriers who deliver the parcels. Prisme is thus an intermediate stage in the delivery of parcels. While use of the platform does involve some cost, it is undoubtedly of benefit to booksellers and small publishers: instead of paying transport costs per parcel for each consignment from a publisher, booksellers are able, by grouping consignments in a smaller number of parcels and deliveries, to cut transport costs; the benefit to small publishers is that by grouping consignments they secure better rates from carriers than they would have obtained solely on the basis of the volumes carried (502).

(649) It would be hard to imagine that Hachette and VUP, which together account for [45-55]% of total deliveries in terms of value and even more in terms of volume (503), would not be large enough to secure more advantageous transport rates than their competitors. At the same time, Hachette and VUP today finance a large proportion of Prisme's running costs and might consider it pointless financing a tool from which they derive little benefit. For all these reasons and because that would enable it push up its rivals. costs, the merged entity would have both the possibility and the incentive to leave Prisme. As the Prisme economic interest grouping itself explains, withdrawal would mean the end of the grouping (504). Many booksellers have also pointed out that even if Hachette-VUP promised to remain a member as a gesture of solidarity, the power of the merged entity within Prisme would give it de facto control over the use of the facility (505).

B.3.b. CONGLOMERATE EFFECTS

(650) As explained above, distribution services are generally covered together with marketing services by a joint, simultaneous contract. For this reason, the dominant positions of the merged entity in the markets for marketing services for sales to the different classes of dealer will make a decisive contribution to the creation of a dominant position for the merged entity in the market for distribution services provided to outside publishers.

(651) The Commission's investigation has shown that, for contractual and practical reasons, such as the difficulty for a marketer of dealing with several different distributors, an outside publisher choosing a marketer to handle its books will always have recourse to the integrated distribution services of that marketer. Therefore, the market position of the provider of marketing services, whatever it may be, will have a direct impact on its position as a supplier of distribution services (506).

B.3.c. COUNTERVAILING POWER

(1) EXISTING COMPETITION

(652) Apart from the fact that their positions are clearly inferior to that of the merged entity, competing distributors cannot claim to rival its position on account in particular of the size of its captive distribution activities. As regards distribution activities performed for outside publishers, competing distributors do not have the same position as the merged entity on the markets for the sale of marketing services to be able to claim that they exert real countervailing power.

(653) The notifying party once again contests the link between the size of captive activities and market power in sale of distribution services for outside publishers. Here too there is an obvious inconsistency in the position of the notifying party: on the one hand it claims that there is no market for services to outside publishers and that shares in the market for distribution services must include sales of books published by in- house publishers (507), and on the other that such sales should not be taken into account for the competition analysis.

(654) The Commission also considers that there are sizeable obstacles to expansion for distribution activities (508). Firstly, as in the case of marketing, there are opportunity costs linked to the need to invest in additional capacity for the competitors of Hachette and VUP, who, in view of their comparatively small size, have only limited capacity to absorb new customers. More specifically, the biggest distributors have provided the Commission with detailed information on what an increase in their capacity would entail. It is clear that even a limited extension of capacity would require an extension of reception, storage and preparation areas, changes in the organisation of work and the recruitment of new staff; and the process would have to be spread over time, both because of the time required to alter the structures and the sunk costs that would be involved (509). What is more, an oversized distribution platform would lead to diseconomies of scale, with the result that there would be no economic rationale for such investments if they were based on anticipated but uncertain gains in customers.

(655) Finally, the Commission's investigation identified barriers to change, as was explained in connection with marketing services (510). They mainly concern the conditions for terminating marketing/distribution contracts and the liquidity requirement associated with managing returns. It must also be pointed out that the links between marketing and distribution are a barrier to change for customers of Hachette and VUP, which are very dependent on access to hypermarkets and wholesalers for selling their books and which would have great difficulty, as has already been explained in the analysis of the markets for the sale of marketing services, in finding a supplier offering equivalent access.

(656) Nor is there a serious prospect that distributors competing with the merged entity might be able to restrain the merged entity and become a credible alternative after the merger by reducing the rates they offer to customer publishers. Even without the merger Hachette and VUP each have costs lower than those of their other main competitors (511).

(2) POTENTIAL COMPETITION

(657) As regards the potential competition, the Commission's investigation reveals that entry barriers in the market for distribution services consist mainly in the considerable investments in terms of IT and logistical equipment, estimated at between euro 10 million and euro 15 million which are needed for the quick, efficient handling of the many different flows. Moreover, given the need to recoup substantial fixed costs from a significant mass of activity, and the size of the captive sales already accounted for by the merged entity and its main competing distributors, it is to be feared that operators likely to have an interest in entering the market for distribution services would regard this alternative as a difficult one, on the grounds that it would not be economically profitable.

(658) The Commission's investigation shows that distribution is characterised by high fixed costs that increase in steps (512). The Commission has found that, seen in a historical context, distribution infrastructures have become increasingly computerised, since, according to some distributors, the facilities needed for properly carrying out this function are so considerable that small and medium-sized structures are increasingly less competitive, as is attested by the exit of certain medium-sized firms (Géodif, Dervy, Le Cerf, Iotadis, or Desclée de Brouwer) (513). All these factors indicate that an entity's size is of decisive importance for its production costs and, hence, its ability to compete.

(659) Moreover, if a potential entrant is to have any hope of making its decision to invest in distribution services profitable and hence penetrating the market, it will, as has been explained above, have to base itself on a sufficiently strong position in the distribution services markets to enable it to capture a sufficiently large volume of marketed works for distribution. Apart from the merged entity, which is the dominant player on these distribution markets, and its nearest competitors, which all have integrated distribution services, the genuinely available share of these services is limited, and would therefore not enable any potential entrant to exert real competitive pressure on the merged entity.

(660) The importance of barriers to entry and expansion in distribution services is confirmed by the fact that the only recent major entry into the distribution services market was that of MDS (514).

(661) Lastly, the prospect of bringing the competing distributors together in a single organisation is not realistic or credible, as is attested by the history of these same distributors. First, from a purely technical viewpoint, the Commission finds it difficult to regard as a valid alternative, exerting effective competitive pressure on the merged entity, any speculation as to the feasibility of a reaction of this kind to the transaction. Even if it were conceivable, this additional concentration of the market for distribution services would still not help to ensure the expected countervailing power. To do that, as has been explained above, the new alternative concentrated distributor would also have to be concentrated in the marketing of the works to be distributed. Lastly, it would also be necessary to envisage merging not only the distribution facilities of the current independent competitors but also, and above all, their marketing teams and, hence, in practice, their publishing houses. Apart from the simple fact that this scenario is not easily achievable for evident practical and economic reasons, the Commission cannot base its competition analysis on a hypothesis which is so unrealistic and lacking in credibility. As history has shown, even before the notified transaction, the distributors competing with Hachette Livre and VUP have not been able to create an alternative structure of this kind, even though incentives are already available.

(662) In light of the foregoing, neither existing nor potential competitors would be able to prevent the merged entity from acting independently of them.

(3) DEMAND

(663) In these markets, the demand side consists essentially of small publishers, except for Albin Michel, whose motivation for opposing the merged entity, in particular as a result of its marketing/distribution contract with Hachette, has already been discussed.

(664) These publishers, therefore, would not have sufficient buying power to act as a counterweight to the merged entity, particularly since in-house sales would account for the bulk of the books distributed. The notifying party's argument that the smallest publishers would have greater bargaining power than the bigger publishers is contradicted by the operation of any market, as well as by the rates applied by Hachette and VUP (515). In addition, the current customers of Hachette and VUP would be confronted with obstacles to any change of provider, as has already been said.

(665) The combination of these factors leads to the conclusion that no real countervailing power would be exerted by competitors or by customers.

B.3.d. CONCLUSION

(666) The notified transaction would therefore create a dominant position as a result of which effective competition would be significantly impeded in the Common market or in a substantial part of it in the markets for the sale of distribution services to publishers.

C. MARKETS FOR THE SALE OF BOOKS

C.1. MARKETS FOR THE SALE OF BOOKS BY PUBLISHERS TO DEALERS

(667) As indicated in the section on the definition of product markets, markets for the sale of books by publishers to dealers should be subdivided into markets for the main categories of books concerned.

(668) With the exception of school textbooks and reference works, which will be considered later, the Commission takes the view that in all of the markets defined the notified transaction would lead to the creation of dominant positions, as a result of a combination of horizontal, vertical, and conglomerate effects (516). Together these effects would enable the merged entity to behave to an appreciable extent independently of its competitors and customers, and ultimately of final consumers.

(669) The merged entity's dominant position might in particular be reflected in an increase in prices to the final consumer, or to dealers (in the form of lower discounts), or both. The Commission's investigation has shown that many booksellers fear a reduction in their discounts, or an alignment of VUP's discounts on those of Hachette Livre (517). Some of them cite the example of the takeover of the marketing of Albin Michel's books by Hachette Livre, when they say the discount given to dealers on Albin Michel's books fell by an average of 1% (518).

(670) The merged entity's dominant position might also be reflected in an ability to reduce supply, in terms of numbers of new titles and reprints, or in terms of the originality of new publications, or both; this would impoverish creativity, quality and diversity in publishing (519). According to one author, "artistic quality has never been the result of monopoly; it springs from the diversity of choice" (520); a bookseller says that "twelve years or so ago the market for French books was somewhat lethargic. The appearance of VUP, a new pole competing with the dominant group, Hachette, allowed the book trade, contrary to forecasts at that time, to pull itself together and develop. A merger between VUP and Hachette would create a super-dominant group, which would make for sclerosis - and tend to smother any attempt at renewal in the sector" (521). Fabrice Piault points out that "the changes in the book market during the super-concentration of the 1980s show that it did not have the effect of encouraging competitors to enter the business - After cutting staff and closing surplus production capacity, they have ended up by eliminating most forms of internal competition. And the cost of entering the market has become so high - that the spaces they have left free cannot easily be occupied by competitors" (522).

(671) Lastly, the dominant position might also be reflected in a reduction in the openings available to publishers that the merged entity does not itself market, as the merged entity could monopolise shelf space and the areas devoted to book promotion (523). With the marginalisation of publishers it did not market, it would have a growing incentive to reduce the market shares of the publishers it did market.

(672) This is because, as the investigation carried out by the Commission has shown, the limited availability of shelf space at sales outlets means that competition in the markets for the sale of books to dealers operates mainly in terms of access to shelves, through discounts negotiated between publishers and dealers, but also through the system of automatic distribution and promotion campaigns: in the biggest bookshops, access to promotion campaigns and prominent displays is even more important than access to shelf space, which is less constricted there (524). More specifically, in order to maximise sales of certain titles, especially new titles, a bookshop will usually choose to display them in special sales areas, perhaps on tables near the entrance. Access to prominent display of this kind is also a scarce resource, and is of great commercial and financial value to publishers in a context where over 40 000 new titles are published every year.

(673) Against this background, given its size in the relevant markets as compared with its nearest competitors, and its importance in the successive markets in the book chain, the merged entity would be in a position to marginalise its competitors while maintaining the illusion of choice for the final consumer, as a result of the number of publishing houses and consequently of separate imprints that would belong to it. This marginalisation would be more pronounced in the case of best-sellers and well-known authors; other publishers would be confined to prospecting for new talent, and, as has been explained in the analysis of the markets for the acquisition of rights, once these new writers had achieved success they would very likely be approached in their turn by the merged entity.

(674) In its reply to the Commission's statement of objections, the notifying party puts forward the argument that concentration among publishing houses is not a factor that tends to limit supply; the number of titles, it says, has not ceased to increase, despite the links established between some publishers.

(675) The results of the Commission's investigation do not cast any doubt on the growth in the number of titles, and it will be useful, therefore, to go beyond the explanation just given regarding the reduction in supply to put the notifying party's argument in perspective, by clarifying the concept of supply and the limitation of supply, and explaining why there is no contradiction between the facts adduced by the notifying party and the Commission's reasoning regarding damage to the consumer.

(676) Contrary to what the notifying party seeks to imply, the concept of supply in the publishing industry embraces more than just the number of titles offered for sale in a year. It has a qualitative dimension that is a great deal subtler than the number of books sold (525). A diversified supply will not include products that are physically different but whose content is the same. In the book trade, especially in general literature, there has been a proliferation of uniform and similar books. This clearly does not mean that there are fewer and fewer books being published. As the notifying party says, the number of new titles has never been bigger. But what the notifying party omits to point out is that the growing number of titles is not necessarily an indication of greater diversity, but rather of a proliferation of books that resemble one another and which are sold in just the same way as mass consumption products. The positive link between concentration and supply that the notifying party hopes to establish appears very doubtful when it is observed that an expansion in the number of titles is not the same as a diversification of supply (526).

(677) One of the strategies available to the merged entity would be to offer booksellers an incentive to increase the volume of their purchases from its publishing houses. In the course of the investigation it was claimed that even before the transaction the strategy followed by Hachette Livre and VUP had been to "occupy the ground and [dealers'] time", by bringing out new titles at a considerably faster rate than their competitors, and taking up a large part of booksellers' time especially with visits from representatives (527). After the transaction the merged entity could intensify this strategy, but it could also give booksellers an incentive to increase the volume of their purchases, especially of books distributed automatically (528), and to promote its products (529), given its importance in dealers' turnover (530). An analysis of Hachette Livre's commercial practice shows that it is in fact already pursuing these strategies. Hachette Livre offers quantity discounts if a dealer increases turnover with the publishing houses it markets by comparison with the previous year (531). Hachette Livre gives a further discount [...]* if the shelf is occupied entirely by titles marketed by LDS. There are also partnership contracts between Hachette Livre and dealers [...]*, which give an extra discount on condition that the dealer undertake to promote the books of the publishing houses in the group on a preferential basis (532). The merged entity would be even stronger in its dealings with these sales outlets, and this would thus increase their incentive to buy from its publishing houses.

(678) If any dealers were to resist increasing the volume of their purchases from the merged entity, the merged entity would be in a position to force them to do so, for example by recasting the structure of discounts, or by threatening to lower their total discounts, to review the time allowed for payment533, or to delay or cut off deliveries.

C.1.a. MARKETS FOR THE SALE OF BOOKS OF GENERAL LITERATURE BY PUBLISHERS TO DEALERS

(679) As indicated in the section on the definition of product markets, markets for the sale of books of general literature should be subdivided according to the format of the books, that is to say whether they are in large format or pocket format. Markets for the sale of books of general literature should also be subdivided by type of dealer. For purposes of the competition analysis, however, the anticipated effects of the notified transaction in the various markets for the sale of books of general literature are relatively similar in nature, and they will be considered together here, with any effects specific to particular markets being identified separately where necessary.

(680) According to information supplied by the notifying party and others, sales of general literature titles by publishers amounted to euro 931 million in 2001 (including euro 330 million for books in pocket format); this was almost one third of the total market for the sale of books. To distinguish between sales channels, bookshops accounted for euro 602 million of publishers' general literature sales (including euro 192 million for pocket-format books), hypermarkets accounted for euro 154 million (including euro 63 million for pocket-formats), and wholesalers accounted for euro 175 million (including euro 74 million for pocket-formats).

(681) The Commission's investigation has shown that the notified transaction would lead to the creation of dominant positions which would be held by the merged entity in the French-language markets for the sale of general literature in pocket format, and in the French-language market for the sale of general literature in large format, by publishers to dealers of all types, as a result of a combination of horizontal, vertical and conglomerate effects. Together these effects would enable the merged entity to behave to an appreciable extent independently of its competitors, of its customers, and ultimately of final consumers; and this could result in an increase in retail prices, or a reduction in the discounts given to bookshops, or a reduction in the quality and variety of books published, or all of these.

(1) HORIZONTAL EFFECTS

(682) The notified transaction would eliminate VUP as a supplier of books of general literature competing with the Lagardère group, and would thus immediately reduce competition in those markets, as what had hitherto been two of the biggest suppliers of such books would now form a single entity.

(a) Players

(683) Hachette Livre operates in the French-language market for general literature in large format through Calmann-Lévy, Editions n° 1, Fayard (Mazarine, Mille et Une Nuits and Pauvert), Grasset et Fasquelle, Hachette Littératures, Harlequin, Jean-Claude Lattès, Le Masque/Champs Elysées, and Stock. VUP is present through Robert Laffont (Bouquins, Julliard, Nil and Seghers), Plon (Plon, Perrin and Orban), Presses Solar-Belfond (Presses de la Cité, Solar, Belfond, Acropole, Omnibus, Le Pré aux Clercs and Hors-Collection), La Découverte et Syros, and Presses de la Renaissance.

(684) The merged entity's main competitors are Gallimard (Gallimard, Denoël, P.O.L., Mercure de France, La Table ronde, Quai Voltaire and La Pléiade), Flammarion (Flammarion and Pygmalion), Seuil (Seuil, L'Olivier and Baleine), Albin Michel, and a great many small or medium-sized publishers (XO, Éditions de Minuit, Actes Sud, Odile Jacob, Le Cherche Midi, etc). It should be borne in mind, however, that for the marketing and distribution of their books some of these publishers depend wholly or partly on Hachette Livre or VUP - this is the case of Albin Michel and XO in particular - and their ability to conduct their own commercial policy, and thus to exercise competitive pressure on the merged entity, is thereby reduced. The fact that Hachette Livre and Albin Michel are partners in LGF also helps to give them interests in common.

(685) The notifying party takes the view that its competitors also include France Loisirs, a wholly owned subsidiary of the Bertelsmann group; it argues that France Loisirs has recently established itself as a full-scale publisher, bringing out new works, and that it has a buyer profile comparable to that of conventional publishers.

(686) But France Loisirs is a book club which buys almost all its publishing rights from the publishers already mentioned here, and operates almost entirely in the secondary market for the purchase of club rights534 and in the retail sales market. As for the buyer profile, the consumer, to buy from a book club, joins the club by accepting a promotional offer, and membership carries an obligation to make regular purchases, in the case of France Loisirs one book every three months for two years (535). Members buy books from a limited catalogue, most of them second editions, though with a few new titles. France Loisirs does not sell to traditional booksellers; the bulk of its outlets are integrated into the company, and open to members only. As things stand, therefore, it would not be justified to consider France Loisirs a competitor with the merged entity in the market for the sale of books by publishers to dealers.

(687) In its reply to the statement of objections, the notifying party insists once again on the importance of Bertelsmann in the market for the sale of books of general literature, though it does not make it clear whether it is speaking of the market for sales by publishers to dealers or the market for retail sales. It says that France Loisirs "has had first publication contracts concluded with [the author] Yann Queffélec". The results of the Commission's investigation, however, confirm that France Loisirs is absent from the market for the sale of general literature by publishers to dealers, and that its role as a publisher is less than marginal, since contrary to what the notifying party maintains France Loisirs has not concluded first publication "contracts" with Yann Queffélec, but only one contract, for one book which is not expected before 2004, and it has done so in full agreement with Mr Queffélec's usual publisher, Fayard. Lastly, France Loisirs concludes over 600 rights transfer contracts with publishers every year in France, but has concluded only three with authors of general literature, including Yann Queffélec.

(688) Hachette Livre operates in the French-language market for the sale of general literature in pocket format through LGF (in which Albin Michel has a minority holding), with its "Le Livre de Poche" series; Harlequin (in which Hachette has a half share); and La Librairie des Champs-Elysées. VUP is present through Univers Poche, which publishes the "10/18", "Pocket" and "Fleuve Noir" pocket-format series.

(689) Here the main competitors with the merged entity are Gallimard (the "Folio" series), Flammarion (the "J'ai Lu" series, in which Hachette has a 35% stake), and Seuil (the "Points" series). Some smaller publishers also publish their own pocket-format series simply in order to exploit their own lists further, and consequently do not need to engage in extensive buying of rights for the purpose: this is the case of Actes Sud (the "Babel" series), Odile Jacob ("Poches Odile Jacob") or Payot Rivages (the "Rivage" series).

(b) Market shares

(690) In order to estimate the market shares of the merged entity in the French-speaking area of the Community, the Commission bases itself on the estimates provided by the notifying party and on its own calculations, which are themselves based on data supplied by the notifying party and by others.

(691) Neither the notifying party nor the Commission has information on the players. market shares in Switzerland, or, consequently, in a French-language market comprising France, Belgium, Luxembourg and Switzerland. The notifying party has estimated the total volume of publishers. sales in French-speaking Switzerland at 3 to 4% of sales in France (536). It is probable that the ratio of size between the French-speaking areas in the Community and Switzerland is comparable whatever product market is considered, and this ratio may therefore be deemed to be representative of the ratio for sales of books of general literature. It is fair to conclude that market shares for the French-speaking areas in the Community provide a sufficient basis for a valid assessment of the market strength of the merged entity in the French-language markets for the sale of books of general literature (537).

(i) Markets for the sale of large-format books

(692) According to the information provided by the notifying party, the different players. market shares, by value, were as follows in 2001 (538):

<emplacement tableau>

(693) According to the findings of the Commission's investigation, the different players. market shares, by value, were as follows in 2001 (539):

<emplacement tableau>

(694) Whatever the source, these figures indicate that in the markets for the sale to wholesalers, hypermarkets and bookshops of general literature in large format, the position of the merged entity would be far ahead of those of its main competitors, and that a sizeable part of the market (over 40%) is scattered in small shares.

(695) In the market for the sale of general literature in large format to wholesalers, the HHI indicates that after the transaction the degree of concentration would be relatively high, at 1 714 points, and that the increase would be significant, at 782 points.

(ii) Markets for the sale of pocket-format books

(696) According to the information provided by the notifying party, the different players. market shares, expressed in value, were as follows in 2001:

<emplacement tableau>

(697) According to the findings of the Commission's investigation540, the different players. market shares, expressed in value, were as follows in 2001:

<emplacement tableau>

(698) Whatever the source, these figures show that even before the notified transaction Hachette Livre and VUP were the two main players in the markets for the sale to dealers of general literature in pocket format; they both have shares of the market far ahead of those of their competitors. Irrespective of the level of dealers considered, the merged entity would have more than twice the market share of its nearest competitor; the gap reaches 70% in the case of the market for sales of such books to hypermarkets.

(699) The HHI for all sales channels together was already at 2 212 points before the transaction, indicating a very high level of concentration, and would increase after the transaction by 1 890 points to 4 102. In each of the relevant markets the index after the transaction would easily exceed the 1 800 threshold and an increase of 1 000 points: in the market for sales to booksellers it is 3 255 points, up 1 457; in the market for sales to hypermarkets it is 5 726 points, up 2 495; and in the market for sales to wholesalers it is 5 325 points, up 2 464.

(c) Non-coordinated (or unilateral) effects

(700) The non-coordinated (or unilateral) effects have been calculated in an econometric study carried out for the Commission into the market for the sale of books of general literature by dealers to final consumers (541). The non-coordinated effects considered in the study measure the impact of a concentration on the list price exclusive of tax (542), the consumers. surplus, and the profits of the firms concerned. Before the merger, if Hachette Livre were to decide to increase list prices unilaterally, some final consumers would turn to competing publishers. As a result of the merger with VUP, Hachette Livre would have absorbed part of the competition, and would thus be able to recover part of this clientele. In addition, competing publishers may also take advantage of the weakening of competitive pressure, themselves recover part of the demand, and in their turn find that the optimal course is to raise their prices. After estimating final demand for general literature in large format and pocket format, a study of this type simulates an increase in prices which may be brought about by a merger between two rival firms as a result of the new configuration of the market. The analysis of the non-coordinated effects of the merger does not incorporate the vertical and conglomerate effects. In the present case, for want of data and an available econometric model, the study concentrates on the increase in prices to the final consumer, and does not cover the level of discounts given to dealers, who constitute the demand side in the relevant markets. The model assumes a market of a size that allows the final consumer to buy a book of another type or no book at all.

(701) But although the study looks at the increase in the price of books of general literature to the final consumer, it is also indicative of the turnover of dealers, as the structure of the discounts given to dealers is expressed in percentages of the list price.

(702) In the markets for the sale of all general literature books in pocket format and large format, by dealers of whatever type, the study indicates that after the notified transaction the list prices exclusive of tax for books published by the merged entity would increase by an average of 4.84%, no account being taken here of the vertical and conglomerate effects of the transaction. According to the study there is only a 5% chance that the increase in prices resulting from the transaction would be outside a range from 3.74% to 5.54% (543). This average increase in prices would reduce consumer's surplus by 6.04%, equivalent to 1.5% of the industry's turnover in general literature.

(703) In the markets for the sale of general literature in pocket format by dealers of all types to final consumers, the study indicates that following the notified transaction the list prices exclusive of tax of the books published by the merged entity would rise by an average of 5.51%, no account being taken of the vertical and conglomerate effects of the transaction. In the markets for the sale of general literature in large format by dealers of all types to final consumers, the list prices exclusive of tax of books would increase by an average of 1.59%, no account being taken here either of the vertical and conglomerate effects of the transaction. The study also indicates that the bulk of the average price increase would be due to the increase observed in pocket-format books.

(704) In its reply to the statement of objections, the notifying party argues that the non-coordinated effects scenario put forward by the Commission is contradictory in several respects. First, the analysis of non-coordinated effects does not relate to any of the markets identified by the Commission. Second, the non-coordinated effects contradict the definition of the relevant markets, in which the market for the sale of general literature is divided between pocket-format books and large-format books. And third, the non-coordinated effects take no account of the way publishing works in reality and of the way book prices are determined.

(705) The Commission takes the view that these criticisms are unfounded. The econometric study does examine the increase in the list price exclusive of tax to the final consumer, but it is also indicative of the market for the sale of general literature titles to dealers, because discounts to dealers are calculated as a percentage of the list price exclusive of tax. As explained earlier, as a result of the transaction, no account being taken of the vertical and conglomerate effects, the list prices exclusive of tax of the notifying party's books would increase proportionately more than those of the competing publishers. All other things being equal, if discount rates remain the same, the transaction would give the notifying party a wider margin than those of the parties separately, so that it could take advantage of this extra leverage in its dealings with competing publishers.

(706) The econometric study does not contradict the existence of a market for pocket-format books distinct from that for large-format books, and it is not correct to say, therefore, that substitution between large-format books is far weaker than substitution between pocket-format and large-format books. The measurement of the non-coordinated effects, combining the two formats, does not diverge widely from a separate measurement of the two formats, since the merged entity is well represented in all categories and all formats. The study offers an objective method of encompassing the two dimensions and testing the level of substitutability between books in the different formats, with no predetermined view of the pre-merger level.

(707) Nor does the study contradict the working of the publishing industry observed by the Commission. It considers that the price of a book is set at publishing group level, but that in setting the price the group has to acknowledge that it is itself the expression of the sum of the different publishing houses within it and the sharing of their costs.

(2) VERTICAL AND CONGLOMERATE EFFECTS

(708) Vertical integration and the strong or dominant position of the merged entity in various markets in the book chain would contribute to the creation of dominant positions in the relevant markets for the sale of general literature by publishers to dealers, enabling it to behave to an appreciable extent independently of its competitors and customers.

(a) Acquisition of publishing rights for general literature titles in French

(709) First, the dominant positions held by the merged entity in the primary and secondary markets for publishing rights for general literature titles (544) would enable it to become dominant in the markets for the sale of books of general literature, all formats and dealers together. Once publishers have acquired the rights in a book, they exploit them by publishing the book. The more authors the merged entity could attract, and hence the more publishing rights it acquired, the more titles it would offer for sale, and the fewer titles could be offered by its competitors, who would have lost authors, and especially their best-selling authors. The dominant positions held by the merged entity in the markets for rights would then help to create dominant positions for the merged entity in the markets for the sale of books of general literature by publishers to dealers.

(710) As regards general literature in large format, the creation of a dominant position in the primary market for French-language publishing rights would enable it structurally to increase its strength in the market for the sale by publishers to dealers of general literature in large format, as a straightforward result of the gradual extension of its ownership of the primary publishing rights, and would thus contribute to the creation of a dominant position in the markets for the sale by publishers of general literature in large format, whatever the level of dealers considered.

(711) As regards general literature in pocket format, the creation of a dominant position in the secondary market for pocket-format publishing rights, explained above, would enable it structurally to increase its strength in the market for the sale by publishers to dealers of general literature in pocket format, as a straightforward result of the gradual extension of its ownership of the secondary pocket-format publishing rights, and thus would thus contribute to the creation of a dominant position in the markets for the sale by publishers of general literature in pocket format, whatever the level of dealers considered.

(b) Marketing and distribution

(712) The Commission's investigation also shows that the merged entity's strong position in marketing and distribution, for its own account and on behalf of others, would help to create dominant positions in the markets for the sale of general literature titles irrespective of format or type of dealer.

(i) Marketing

(713) As has been explained in the analysis of the markets for marketing services, Hachette Livre and VUP both provide substantial marketing services for the publishing houses in their own groups and for the account of other publishers.

(714) In the marketing of books, the dealer in practice does business mainly with the marketer rather than with the publisher, so that the strength of the merged entity's position in its relations with dealers has to be evaluated not on the basis merely of its own share of book sales but on the basis of all the books it markets to those dealers. The shares of the various marketers in the volume of books bought by dealers are an important indicator of the strength of the merged entity in its relations with dealers, and hence in the competition analysis of the markets for the sale of books by publishers to dealers (545).

(715) The Commission's investigation shows very clearly that the marketer plays an important role in the purchase of books by dealers (and is often their only contact), notably because the quantity discounts given to dealers are based on their turnover on all general literature titles marketed by that marketer, and not on turnovers in the products of individual publishers546.

(716) The merged entity has substantial positions in the marketing of general literature in large format to dealers (547):

<emplacement tableau>

(717) As a marketer, therefore, the merged entity would have a position in its business with dealers which was substantially stronger, both in absolute terms and in relation to its immediate competitors, than it had in its role as a publisher.

(718) The creation of dominant positions held by the merged entity in the markets for marketing services, which has been analysed above, would enable it structurally to increase its strength in its dealings with booksellers, and thus to augment the contribution made by its marketing activities to the creation of dominant positions in the markets for the sale of general literature in large format by publishers, whatever the level of dealers considered.

(719) These positions in marketing services, whatever the level of dealers considered, would help to create dominant positions in the various markets for the sale of general literature in large format, because they would give the merged entity the ability substantially to increase its sales at the expense of its most immediate competitors. The ability to control the marketing of a large number of books, including books published by the publishers for whom it handled marketing or distribution, would enable it to behave independently of its competitors.

(720) As already pointed out, marketing is a stable resource by comparison with the other branches of the book trade, and is consequently an essential source of financing for vertically integrated publishing houses, which enables them to invest in the acquisition of publishing rights, a business where risks are much higher, and thus to maintain the supply of general literature in both large format and pocket format.

(ii) Distribution

(721) As has been explained in the analysis of the markets for distribution services, Hachette Livre and VUP both provide substantial distribution services for the publishing houses in their own groups and for the account of other publishers.

(722) In the payment mechanism for books (invoicing, management of financial flows for returns, etc"), the dealer in practice does business mainly with the distributor rather than with the publisher, so that the merged entity's position as a distributor would have a powerful influence on its position in its dealings with booksellers. The shares of the various distributors in the volume of books bought by dealers are an important indicator of the strength of the merged entity in its relations with dealers, and hence in the competition analysis of the markets for the sale of books by publishers to dealers.

(723) The Commission's investigation shows very clearly that the distributor plays an important role in the purchase of books by dealers (and is often their only contact), because it is the distributor who directly manages invoicing, terms of payment, returns and deliveries.

(724) The merged entity has substantial positions in the distribution of general literature in large format to dealers (548):

<emplacement tableau>

(725) It is clear, therefore, that in its role as a distributor the merged entity would have a position in its business with dealers which was stronger, both in absolute terms and in relation to its immediate competitors, than it had in its role as a publisher.

(726) These positions in distribution, whatever the level of dealers considered, would contribute to the creation of dominant positions in the various markets for the sale of general literature in large format, as a result of the merged entity's capacity to monitor booksellers. invoicing, and thus their financial resources and their stocks of new books (549). As regards liquidity in particular, the Commission's investigation shows that for a great majority of dealers the merged entity would be their biggest creditor, and would account for a large proportion of the credit linked to returns. If the merged entity were to credit returns late, these booksellers would find it more difficult to buy new books. The Commission's investigation also shows that if the merged entity were to penalise them by stopping or interrupting supplies, they, or at least the smallest of them, would not be able to continue as dealers (550).

(iii) Conclusion

(727) The strong positions held by the merged entity in the markets for marketing and distribution services would therefore combine to contribute to the creation of dominant positions in the markets for the sale of general literature, irrespective of format and type of dealer considered.

(c) Other markets for the sale of books by publishers to dealers

(728) Furthermore, given the strategic interest of the sale of general literature in pocket format and of school textbooks, the merged entity's particularly large shares of the markets for the sale of such books, irrespective of the type of dealer, would contribute to the creation of a dominant position for the merged entity in the markets for the sale of general literature in large format. According to the findings of the Commission's investigation the sale of books in pocket format is highly profitable, and along with marketing and distribution and the sale of school textbooks constitutes a major source of financing for publishers which enables them to invest in the acquisition of publishing rights, a business where risks are much higher, and thus to maintain the supply of general literature in large format (551).

(729) The majority of booksellers selling school textbooks say that the loss of this business would be catastrophic for their operating balance, and confirm that after the transaction the merged entity could use its strength in the market for the sale of school textbooks to encourage them to buy other categories of books, such as general literature titles (552).

(d) Advertising and media

(730) The fact that Lagardère, unlike the competing publishers, also has a presence in advertising (553) and in media such as the press and broadcasting (554) would give the merged entity an advantage as a seller of books of general literature, essentially in large format, since any promotion of a book helps its sales and encourages dealers to display it prominently at their sales outlets (555). This is especially so in the case of class 1 bookshops, who do not generally subscribe to a system of automatic distribution but determine the number of books they order on the basis of forecasts of future sales, and consequently of planned promotion campaigns.

(731) The Lagardère group's conglomerate structure puts it in a position to give preferential treatment to its own books in its own media, because it would be in a position to give its own books preferential coverage, and to choose the scale and timing of publicity, and also enables it to promote books in the media more heavily than its competitors can, because of the preferential rates available to the publishing houses in the group; these advantages would enable the merged entity to improve its position by comparison with those of it competitors, who have no presence in the media.

(732) The merged entity's unique position in advertising and the media gives it a competitive advantage in the various markets for the sale of books of general literature by publishers to dealers. Market players have emphasised the importance of this consideration, but the Commission takes the view that in the circumstances it is not a decisive factor that would bring about the creation of a dominant position in these markets.

(e) Retail book sales

(733) Lastly, as explained with regard to the markets for the acquisition of rights in general literature titles, the merged entity, with its own shops, has a growing presence in the downstream market for the sale of books by dealers to final consumers, whereas competing publishers do not operate at that level (or, in the case of a limited number of them, do so only to a marginal extent); this situation would not be changed appreciably by the notified transaction, but the merged entity's substantial presence in retail sales could enable it to promote the sale of books published by its own group's publishing houses more heavily, or to give them preferential treatment by comparison with books published by its competitors. The Commission's investigation reveals that even before the notified transaction independent publishers say they have had difficulty in gaining access to shelf space in sales outlets belonging to the Hachette Livre group, and more especially the Relay shops (556). After the merger the merged entity would be in position to extend any practices favouring the group's own publications to include those of VUP (557). Market players have emphasised the importance of this consideration, but the Commission takes the view that in the circumstances it is not a decisive factor that would bring about the creation of a dominant position in this market.

(3) COUNTERVAILING POWER

(a) Existing competition

(734) Some of the merged entity's present competitors in the markets for the sale of books of general literature are publishers who are not vertically integrated over the rest of the book chain, or who are so only to a small extent. This is the case of large-format publishers such as Actes Sud, Payot/Rivages, Les Éditions de Minuit, XO, V. Hamy, Desclée de Brouwer, Le Dilettante, Autrement, Le Cherche Midi, Michel Lafon, Les Éditions du Rocher, De Fallois, Odile Jacob or Anne Carrière. Some of these - such as De Minuit, Actes Sud, Phébus, Payot/Rivages, Du Rocher, Odile Jacob and Picquier - also run pocket-format collections, but their pocket-format business is confined to the exploitation of their own lists (558). They are not vertically integrated outside the primary markets for the acquisition of publishing rights, and for the sale of their books they depend on marketers and distributors.

(735) The relationship between a publisher and the firm providing it with marketing and distribution services means that in these markets a publisher can be a customer of another publisher, dependent on its service provider in a way that curbs its competitive potential in the markets for the sale of books by publishers to dealers; for several reasons, the dependent publisher becomes a satellite of the provider of marketing and distribution services. First, marketers possess a large volume of information regarding the publishers whose books they market. In order to be able to do their job, marketers always have the publisher's programme of publications for the coming months, presentations of the books, and prices, and also possess sales statistics for each of these publishers (559). Direct contact with sales outlets on behalf of these publishers enables marketers to form an idea of the reception of a book in the market (560), and to keep close track of changes in demand; in addition to acting as service providers on behalf of the publishers they market, therefore, they have considerable capacity to influence those publishers' commercial policy (561). This reduction in direct competition between a publisher and a marketer which is a publisher in its own right is also favoured by the fact that it is difficult for a marketed publisher directly to control the quality of the service being provided by the marketer (562). A number of publishers have spoken of the possibility of a marketer giving some of the books it markets preferential treatment over others, with positive or negative repercussions on these books' sales. One publisher, for example, explains that "where there is head-on competition between different publishers in the same marketer's portfolio, there is a danger of a reduction in the scale and quality of the service provided, to the detriment of non-dominant publishers. The marketer may give preferential treatment to its main supplier, by reason of targets or quotas to be met" (563). The notifying party itself confirms that marketing has a decisive influence on the potential success of a title, independently of its intrinsic quality (564).

(736) In reality such publishers. commercial policy is closely bound up with that of their marketer and distributor, because the price at which they sell their products, that is to say the amount of the discount given to dealers, is laid down in the marketer's general conditions of sale, and the appearance of new books has to be planned in consultation with the marketer, who is also the distributor and delivers books, handles invoicing, and manages financial relations with dealers, as already explained in this Decision. These publishers, then, do not exercise any real competitive pressure on the publishing houses controlled by their marketer and distributor in the markets for the sale of books, irrespective of the type of dealer considered. Quite the reverse, in relations with dealers they help to increase the competitive strength of the publishing houses belonging to the marketer and distributor with which they are themselves associated (565).

(737) Although Albin Michel is a substantial publisher, and performs some of its own marketing to bookshops, it is tied to Hachette Livre by an exclusive marketing and distribution contract under which Hachette Livre handles marketing to other dealers (566) and all distribution, including distribution to class 1 bookshops, on Albin Michel's behalf. Albin Michel also has capital links with Hachette, via the LGF pocket-format series. The marketing of books to the smallest sales outlets and the effectiveness of distribution are determining factors in the choice of a publisher by authors (567), and the sale of pocket-format books is a resource necessary in order to amortise the advances paid to authors; it is to be feared, therefore, that Albin Michel's ability to compete with Hachette Livre is already small. It would in any event become so following the notified transaction, because with the disappearance of VUP, the only other firm able to provide marketing and distribution services as extensive as those of Hachette Livre, Albin Michel would lose any real possibility of turning to another provider of marketing and distribution services as an alternative to Hachette Livre.

(738) Alongside the non-integrated publishers, there are a number of vertically integrated publishers who operate in the markets for the sale of books of general literature in both large format and pocket format, though on a scale nothing like that of the merged entity, which would have a turnover eight times that of its nearest competitor, Gallimard (568). Not only do these publishers occupy positions far behind that of the merged entity in all markets for the sale of books of general literature, they are not vertically integrated to the same extent, as they do no business in the media and in the markets for the sale of books by wholesalers to supermarkets and small retailers, in the various markets in the book chain where they are present they have not got the same importance.

(739) In the specific market for the sale of books of general literature to wholesalers, the countervailing power of Gallimard, Seuil and Flammarion would be even weaker, because they would be dependent on the merged entity for a substantial part of their sales of best-sellers (569). The merged entity would be the only integrated unit in the markets for sale by wholesalers to small retailers and supermarkets, so that Gallimard, Seuil and Flammarion would have no means of inciting the merged entity to buy their books.

(b) Potential competition

(740) In the markets for the sale of general literature in large format, the only operators who might have an interest in mounting a commercially credible and significant entry into the markets for the sale of books to dealers would be operators who were capable of becoming significant players in the primary markets for the purchase of publishing rights in French-language books, that is to say operators who were able to acquire and exploit those rights throughout the whole book chain. Given the barriers to entry described in the analysis of the primary markets for rights in works of general literature, no significant entry can be envisaged on the part of a new player able to counter the creation of dominant positions by the merged entity in the markets for the sale of general literature in large format. In addition, as indicated earlier, in order to exercise a competitive constraint on the merged entity a new entrant in the markets for publishing rights would have to be independent in terms of marketing and distribution. The barriers to entry to these activities make such a development improbable, at least on any significant level.

(741) In the markets for the sale of general literature in pocket format, given the additional barriers to entry described in the analysis of the secondary markets for rights in books of general literature, the entry of a new player able to counter the creation of dominant positions by the merged entity is even less credible. The observation made in the preceding paragraph applies once again to the entry to the market of a new competitor independent in terms of marketing and distribution, and consequently to potential competitors such as Albin Michel.

(742) Taking these factors together, it has to be concluded that there would be no real competitive pressure from competing publishers, either existing or potential.

(c) Demand

(743) In the markets for the sale of books of general literature to bookshops, the demand side is made up essentially of players of small size and little economic weight, with the exception of the large specialised shops, primarily FNAC and Virgin, which like Hachette Livre is controlled by Lagardère.

(744) The investigation carried out by the Commission clearly shows that these small players would not be able to stand in the way of any independent behaviour engaged in by the merged entity (570). In any event the merged entity would be able to overcome any opposition, given its importance in their turnover and liquidity, and their financial fragility. Almost all the booksellers surveyed by the Commission said they had no way of withstanding a lowering of discounts (571) on the part of the merged entity (572); as the Kléber bookshop put it: "radical measures, such as an "automatic distribution strike" or a boycott of certain books that are not being recommended or that are not strategic (cookery, guides and manuals, children's etc") could very quickly backfire, with falling discounts and a questioning of the discount points linked to quality, and hence to the breadth of the stock of the publisher's books the bookseller holds" (573). If the merged entity were to set out to marginalise competing publishers, dealers would have no incentive to try to prevent it, because a marginalisation of this kind would increase the volume of sales by Hachette Livre to booksellers, and hence their quantity discount. Their quality discount might also be increased quite mechanically if they joined more campaigns to promote the merged entity's books or to display them prominently. In any event the merged entity could also force booksellers to go along with such a marginalisation strategy, for example by threatening recalcitrants with a general reduction in their discounts, by refusing to allow time for payment574, or by interrupting supplies.

(745) The larger dealers would likewise be unable to stand up to the merged entity if it were to threaten them with a general reduction in their discounts or an interruption of supply, or indeed if it were to carry out such threats (575), by reason of its importance in their turnover. FNAC is the leader in the French markets for the retail sale of books: the merged entity would account for 45% of its turnover in books, and would consequently be an essential supplier. FNAC states that "there are only a few areas of publishing in which there is frontal competition between products, so that one can be substituted for another without damage to the bookseller's turnover. Otherwise the network of sales outlets is sufficiently dense to allow the marketer to rely on a transfer of activity in the event of a local boycott" (576), against a background in which Hachette Livre, through the Virgin stores, is FNAC.s main competitor in the French markets for the retail sale of books (577).

(746) In the markets for the sale of books of general literature to hypermarkets, the demand side is made up of the big food distribution chains; books are not their main business. But even these players take the view that they would not be able to resist the strategy of the merged entity, because they could not do without the books the merged entity distributes, notably pocket-format books and best-sellers (578).

(747) In the markets for the sale of books of general literature to wholesalers, the demand side is made up mainly of players of a size and economic weight nothing like those of the merged entity. The investigation carried out by the Commission shows that they would not exert any competitive pressure on the merged entity. If they attempted to oppose any independent behaviour on the part of the merged entity, the merged entity would have the means to constrain them, by threatening a general reduction in their discounts or an interruption in supplies, by reason of the horizontal effects of the transaction in the markets for the sale of books of general literature and also of the merged entity's importance in the various markets in the book chain. Their countervailing power would be further reduced by the fact that the merged entity's customers in the markets for the sale of books to dealers are also its competitors in the downstream market for the sale of books by wholesalers to class 3 dealers. In its reply to the statement of objections, the notifying party states that "such behaviour is already - theoretically - possible as things stand", that is to say before the transaction, and that consequently "the transaction will change nothing". To confine itself to the first of these points (579), the Commission can only agree with the analysis put forward, with the qualification, however, that the behaviour the notifying party describes as "theoretically" possible has in fact taken place even before the notified transaction, in dealings with a wholesaler customer, DNL (580).

(748) Taking these factors together, it must be concluded that buyers would not have any real countervailing power.

(4) CONCLUSION

(749) The notified transaction would therefore create dominant positions as a result of which effective competition would be significantly impeded in the Common market or a substantial part of it in the markets for the sale of books of general literature, irrespective of the format and the type of dealer.

C.1.b. MARKETS FOR THE SALE OF CHILDREN'S BOOKS BY PUBLISHERS TO DEALERS

(750) The Commission's investigation has shown that the notified transaction would lead to the creation of dominant positions which would be held by the merged entity in the French-language markets for the sale of children's books to hypermarkets and wholesalers, as a result of a combination of horizontal and vertical effects. Together these effects would enable the merged entity to behave to an appreciable extent independently of its competitors, of its customers, and ultimately of final consumers; and this could result in an increase in retail prices, or a reduction in the discounts given to bookshops, or a reduction in the quality and variety of books published, or all of these.

(1) HORIZONTAL EFFECTS

(751) Hachette Livre operates in the French-language market for children's books through Deux Coqs d'Or, Didier Jeunesse, Disney Hachette Édition, Gautier-Languereau, Grasset-Jeunesse, Hachette (Bibliothèque Rose, Bibliothèque Verte and Le Livre de Poche Jeunesse), Hatier and Rageot. VUP is present through Nathan Jeunesse, Hemma, Larousse Jeunesse, Syros et Univers Poche (Pocket Jeunesse).

(752) The merged entity's main competitors are Gallimard (Gallimard Jeunesse and Folio Junior), Flammarion (Flammarion jeunesse, Castor poche, Casterman jeunesse, and J'ai Lu), L'École des Loisirs (École des loisirs, Pastel, Kaléidoscope, Neuf, and Mouche), Albin Michel and Bayard.

(753) According to the information provided by the notifying party, the different players. market shares, by value, were as follows in 2001 (581):

<emplacement tableau>

(754) According to the findings of the Commission's investigation, the different players. market shares, by value, were as follows in 2001 (582):

<emplacement tableau>

(755) Whatever the source, these figures indicate that in the markets for the sale of children's books to hypermarkets and wholesalers, the market shares of the merged entity would be close to or above 40%, twice that of its closest competitor, Gallimard (583), while a sizeable part of the market (over 30%) is scattered in small shares.

(756) The HHI indicates that after the transaction the degree of concentration would be relatively high: in the case of hypermarkets it would be 1 977 points, with a significant increase of 815 points, and in the case of wholesalers it would be 2 243 points, with a significant increase of 870. In its reply to the statement of objections, the notifying party contests the existence of horizontal effects in the markets in the sale of children's books to hypermarkets and to wholesalers; it argues that the merged entity's market shares would be lower in the market for the sale of children's books to bookshops, and in the market for the sale of such books in general irrespective of the class of dealer, and says that the bulk of the merged entity's sales of children's books would go to bookshops.

(757) This argument cannot be accepted, for a number of reasons. In its reply to the statement of objections the notifying party does not deny that the markets for the sale of books by publishers to dealers should be segmented by category of dealer, and consequently accepts that the conditions of competition in the markets for the sale of children's books to hypermarkets and wholesalers may be different from those observed in the market for the sale of such books to bookshops. Furthermore, the existence of horizontal effects in the markets for sales to hypermarkets and wholesalers cannot be made to depend on the relative sizes of these markets as compared with that of the neighbouring market for sales to bookshops.

(2) VERTICAL EFFECTS

(758) Vertical integration and the strong or dominant position of the merged entity in various markets in the book chain would contribute to the creation of dominant positions in the relevant markets for the sale of children's books by publishers to dealers, enabling the merged entity to behave to an appreciable extent independently of its competitors and customers.

(759) As has been explained in the analysis of the markets for marketing services, Hachette Livre and VUP both provide substantial marketing services for their own publishing houses and for the account of other publishers. In the marketing of books, as already explained with regard to books of general literature, the dealer in practice does business mainly with the marketer rather than with the publisher. The strength of the merged entity's position in its relations with dealers consequently has to be evaluated not on the basis merely of its own share of book sales but on the basis of all the books it markets to dealers. The shares of the various marketers in the volume of books bought by dealers are an important indicator of the strength of the merged entity in its relations with dealers, and hence in the competition analysis of the markets for the sale of books by publishers to dealers.

(760) The merged entity has substantial positions in the marketing and distribution of children's books to dealers, and particularly hypermarkets and wholesalers (584):

<emplacement tableau>

(761) It would appear, therefore, that in its role as a marketer and distributor the merged entity would have a position in its business with hypermarkets and wholesalers which was substantially stronger, both in absolute terms and in relation to its immediate competitors, than it had in its role as a publisher of children's books. The ability to control the marketing of a large number of books, including books published by other publishers for whom it handled marketing or distribution, would contribute to its ability to behave independently of its competitors and customers in these markets.

(762) The merged entity's importance as a distributor, responsible among other things for invoicing, would also have implications for its strength in its relations with dealers, and would thus likewise contribute to the creation of dominant positions in the markets in the sale of children's books by publishers to dealers.

(763) The strengthening of the vertical integration of the merged entity over the different levels of the book chain would therefore help to create dominant positions for the merged entity in the markets in the sale of children's books to hypermarkets and wholesalers, enabling it to behave independently of its competitors, its customers and ultimately of consumers.

(3) COUNTERVAILING POWER

(a) Supply

(764) In addition to the fact that the merged entity's competitors have weaker positions in the markets for the sale of children's books to hypermarkets and wholesalers, they do not possess the same degree of vertical integration over all levels of the book chain.

(765) In its reply to the statement of objections, the notifying party contends that competition, existing and potential, is more vigorous in the markets for children's books, because of the dynamic character of these markets: new publishers have entered recently, and new series have been launched.

(766) These special features of children's book publishing go some way towards explaining the absence of a dominant position significantly impeding competition in the market for the sale of children's books to booksellers, but they have a smaller impact on competition in the markets for the sale of children's books to hypermarkets and wholesalers. The markets for sales to hypermarkets and wholesalers have special barriers to entry and expansion, deriving in particular from the publishers' capacity to market books to these classes of dealer; these barriers are not fundamentally different from the barriers observed in the markets for the sale of general literature titles to the same classes of dealer. Given the special barriers, and the high market share of the merged entity and the dispersion of its competitors and the new entrants in particular, with the sole exception of Gallimard, whose success is largely due to the Harry Potter books, the countervailing power of existing and potential competitors cannot credibly be seen as a factor that would enable them to counter the creation of dominant positions by the merged entity in the markets for the sale of children's books to hypermarkets and wholesalers.

(767) Neither existing competitors nor potential competitors, then, would be able to counter the merged entity's ability and incentive to become dominant in the markets for the sale of children's books.

(b) Demand

(768) As for countervailing power exercised by buyers, it has already been explained that the demand side is made up essentially of players who would have no means of individually opposing a strategy of dominance pursued by the merged entity, or no motive to do so. This also applies to hypermarkets, who likewise consider that they would not be able to resist the strategy of the merged entity, because they could not do without the books the merged entity would distribute, especially pocket-format books and best-sellers (585).

(769) Taken together, these factors lead to the conclusion that neither competitors nor customers would wield any real countervailing power.

(4) CONCLUSION

(770) The notified transaction would therefore create a dominant position for the merged entity as a result of which effective competition would be significantly impeded in the markets for the sale of children's books to hypermarkets and wholesalers.

C.1.c. MARKETS IN THE SALE OF ART BOOKS BY PUBLISHERS TO DEALERS

(1) HORIZONTAL EFFECTS

(771) Hachette Livre operates in the French-language market for the sale of art books to dealers through E/P/A, Les Éditions du Chêne, Les Éditions Hazan and Les Éditions Filipacchi; some of its publishers of general literature or guides and manuals, such Fayard, Stock and Calmann-Lévy, also publish a few art books every year. VUP is present through Bordas, Dessain & Tolra, Robert Laffont, Larousse, Plon and Presses Solar-Belfond.

(772) The parties. main competitors in the publication of art books are La Martinière and Sélection du Reader's Digest, followed by Gallimard, Flammarion, Seuil, Albin Michel and a very large number of small or medium-sized publishers.

(773) According to the information supplied by the notifying party, the merged entity would have a share of the market for art books equal to [10-20]% by value in 2001.

(774) According to the findings of the Commission's investigation, the various players. market shares by value in 2001 were as follows (586):

<emplacement tableau>

(775) Thus the merged entity's shares of the markets considered would be small.

(2) VERTICAL EFFECTS

(776) As has been explained in the analysis of the markets for marketing services, Hachette Livre and VUP both provide substantial marketing services for their own publishing houses and for the account of other publishers. In the marketing of books, as already explained with regard to books of general literature, the dealer in practice does business mainly with the marketer rather than with the publisher. The strength of the merged entity's position in its relations with booksellers has to be evaluated not on the basis merely of its own share of book sales but on the basis of all the books it markets to dealers. The shares of the various marketers in the volume of books bought by dealers are an important indicator of the strength of the merged entity in its relations with dealers, and hence in the competition analysis of the markets for the sale of books by publishers to dealers.

(777) According to the findings of the Commission's investigation, the shares held by the merged entity and by other publishers in the marketing and distribution of art books to bookshops, hypermarkets and wholesalers are as follows (587):

<emplacement tableau>

(778) As a marketer and distributor, therefore, the merged entity might have a relatively strong position in the sale of art books to hypermarkets. But this position would largely be accounted for by VUP, and the overlap would be small, and the merged entity's position would be open to challenge by La Martinière, a competitor very active in this area.

(779) Thus the strengthening of the vertical integration of the merged entity over the different levels of the book chain would not create a dominant position for the merged entity in the market for the sale of art books to hypermarkets.

(3) CONCLUSION

(780) The notified transaction is therefore not such as to create or strengthen a dominant position as a result of which effective competition would be significantly impeded in the market for the sale of art books to dealers, be they bookshops, hypermarkets or wholesalers.

C.1.d. MARKETS FOR THE SALE OF GUIDES AND MANUALS BY PUBLISHERS TO DEALERS

(781) The Commission's investigation has shown that the notified transaction would lead to the creation of dominant positions for the merged entity in the French-language markets for the sale of guides and manuals to wholesalers and hypermarkets, as a result of a combination of horizontal and vertical effects. Together these effects would enable the merged entity to behave to an appreciable extent independently of its competitors, of its customers, and ultimately of final consumers; and this could result in an increase in retail prices, or a reduction in the discounts given to bookshops, or a reduction in the quality and variety of books published, or all of these.

(1) HORIZONTAL EFFECTS

(782) Hachette Livre operates in the sale of guides and manuals in France, Belgium and Luxembourg through Hachette Pratique (le Guide Hachette des Vins Millésimés, Les Petits Pratiques Hachette, Hachette Tourisme, La Librairie Générale Française, Marabout and Les Éditions Filipacchi). VUP is present through Presses Solar- Belfond, Hors-Collection, Larousse, Bordas, Dessain & Tolra, Nathan, Laffont and Pocket.

(783) The merged entity's main competitors would be Sélection du Reader's Digest, La Martinière, Albin Michel, Odile Jacob, followed by a very large number of small or medium-sized publishers.

(784) The notifying party takes the view that its competitors also include France Loisirs, a wholly owned subsidiary of the Bertelsmann group. As has already been explained with reference to the market for the sale of books of general literature, the Commission considers that as things stand it would not seem justified to consider France Loisirs a competitor with the merged entity in the market for the sale of books by publishers to dealers.

(785) According to the information provided by the notifying party, the different players. market shares, by value, were as follows in 2001 (588):

<emplacement tableau>

(786) According to the findings of the Commission's investigation, the different players. market shares, by value, were as follows in 2001 (589):

<emplacement tableau>

(787) Whatever the source, these figures indicate that in the markets for the sale of guides and manuals to wholesalers the merged entity's share is close to 40%, whereas none of its main competitors. has a share of over 10%, and a sizeable part of the market (over 50%) is scattered in small shares. The HHI shows an appreciable measure of concentration post-merger, at 1 568 points, which is an increase of 736 points.

(788) The merged entity's share would be lower in the market for the sale of guides and manuals to hypermarkets, but there the share held by its nearest competitor does not exceed 5%. In its reply to the statement of objections, the notifying party contests the existence of horizontal effects in the markets in the sale of guides and manuals to hypermarkets and to wholesalers; it argues that the merged entity's market shares would be lower in the market for the sale of guides and manuals to bookshops, and in the market for the sale of such books in general irrespective of the class of dealer, and says that the bulk of the merged entity's sales of guides and manuals would go to bookshops.

(789) This argument cannot be accepted, for a number of reasons. In its reply to the statement of objections the notifying party does not deny that the markets for the sale of books by publishers to dealers should be segmented by category of dealer, and consequently accepts that the conditions of competition in the markets for the sale of guides and manuals to hypermarkets and wholesalers may be different from those observed in the market for the sale of such books to bookshops. Furthermore, the existence of horizontal effects in the markets for sales to hypermarkets and wholesalers cannot be made to depend on the relative sizes of these markets as compared with that of the neighbouring market for sales to bookshops.

(2) VERTICAL EFFECTS

(790) Vertical integration and the strong or dominant position of the merged entity in various markets in the book chain would contribute, as in the markets for the sale of general literature and children's books, to the creation of dominant positions in the markets for the sale of guides and manuals to wholesalers and hypermarkets, enabling it to behave to an appreciable extent independently of its competitors and customers.

(791) As explained with regard to books of general literature, the merged entity's importance in marketing and distribution services is an important indicator of its strength in its relations with dealers, and hence in the competition analysis of the markets for the sale of books by publishers to dealers.

(792) The merged entity has substantial positions in the marketing and distribution of guides and manuals to dealers, and especially wholesalers and hypermarkets590:

<emplacement tableau>

(793) As a marketer and distributor of guides and manuals, therefore, the merged entity would have a position in its business with hypermarkets and wholesalers which was substantially stronger, both in absolute terms and in relation to its immediate competitors, than it had in its role as a publisher. The ability to control the marketing of a large number of books, including books published by other publishers for whom it handled marketing or distribution, would contribute to its ability to behave independently of its competitors and its customers in those markets.

(794) The merged entity's importance as a distributor, responsible among other things for invoicing, would also have implications for its strength in its relations with dealers, and would thus also contribute to the creation of dominant positions in the markets in the sale of guides and manuals by publishers to dealers.

(795) The strengthening of the vertical integration of the merged entity over the different levels of the book chain would therefore contribute to the creation of dominant positions for the merged entity in the markets in the sale of guides and manuals to hypermarkets and wholesalers, enabling it to behave independently of its competitors and its customers.

(3) COUNTERVAILING POWWER

(a) Supply

(796) In addition to the fact that the merged entity's competitors have weaker positions in the markets for the sale of guides and manuals to hypermarkets and wholesalers, they do not possess the same degree of vertical integration over all levels of the book chain.

(797) In its reply to the statement of objections, the notifying party contends that competition, existing and potential, is more vigorous in the markets for guides and manuals because of the dynamic character of these markets: new publishers have entered recently, and entry would be easy, especially for foreign publishers who could publish translations.

(798) These special features of the publication of guides and manuals go some way towards explaining the absence of a dominant position significantly impeding competition in the market for the sale of guides and manuals to booksellers, but they have a smaller impact on competition in the markets for the sale of guides and manuals to hypermarkets and wholesalers. The markets for sales to hypermarkets and wholesalers have special barriers to entry and expansion, deriving in particular from the publishers' capacity to market books to these classes of dealer; these barriers are not fundamentally different from the barriers observed in the markets for the sale of general literature titles to the same classes of dealer. Given the special barriers, the high market shares of the merged entity, and the dispersion of its competitors, the countervailing power of existing and potential competitors cannot credibly be seen as a factor countering the creation of dominant positions by the merged entity in the markets for the sale of guides and manuals to hypermarkets and wholesalers.

(799) Neither existing competitors nor potential competitors, then, would be able to counter the merged entity's ability and incentive to become dominant in the markets for the sale of guides and manuals to wholesalers and hypermarkets.

(b) Demand

(800) As for countervailing power exercised by buyers, it has already been explained that the demand side is made up essentially of players who would have no means of individually and preventively resisting the dominance of the merged entity, or no motive for doing so. This also applies to hypermarkets, who likewise consider that they would not be able to resist the dominance of the merged entity, because they could not do without the books the merged entity would distribute, especially pocket-format books and best-sellers (591).

(801) Taken together, these factors lead to the conclusion that neither competitors nor customers would wield any real countervailing power.

(4) CONCLUSION

(802) The notified transaction would therefore create a dominant position for the merged entity as a result of which effective competition would be significantly impeded in the markets for the sale of guides and manuals to hypermarkets and wholesalers.

C.1.e. MARKETS IN THE SALE OF STRIP CARTOON ALBUMS BY PUBLISHERS TO DEALERS

(1) HORIZONTAL EFFECTS

(803) Hachette Livre is present in the French-language market for the sale of strip cartoon albums to dealers only because it publishes 25 old Asterix titles. VUP operates in this market through Hors Collection, which publishes French-language editions of foreign strip cartoons.

(804) The parties. main competitors in the sale of strip cartoon albums are Dupuis, Dargaud/Lombard, Glénat, Albert René, Seuil and Les Éditions du Soleil.

(805) According to the information supplied by the notifying party, the merged entity would have a share of the market for strip cartoon albums not exceeding [0-10]% by value in 2001; this figure is estimated on the basis of market shares in France, Belgium and Luxembourg, respectively [0-10]%, [0-10]% and [0-10]%.

(806) According to the findings of the Commission's investigation, the various players. market shares by value in 2001 were as follows (592):

<emplacement tableau>

(807) The merged entity's market shares are very small, and there is almost no overlap in the markets considered.

(2) VERTICAL EFFECTS

(808) As has been explained in the analysis of the markets for marketing services, Hachette Livre and VUP both provide substantial marketing services for their own publishing houses and for the account of other publishers. In the marketing of books, as already explained with regard to books of general literature, the dealer in practice does business mainly with the marketer rather than with the publisher. The strength of the merged entity's position in its relations with booksellers has to be evaluated not on the basis merely of its own share of book sales but on the basis of all the books it markets to dealers. The shares of the various marketers in the volume of books bought by dealers are an important indicator of the strength of the merged entity in its relations with dealers, and hence in the competition analysis of the markets for the sale of books by publishers to dealers.

(809) According to the findings of the Commission's investigation, the shares held by the merged entity and by other publishers in the marketing and distribution of strip cartoon albums to bookshops, hypermarkets and wholesalers are as follows (593):

<emplacement tableau>

(810) It is clear that as a marketer and distributor the merged entity would have a large share of the market for the sale of strip cartoon albums to bookshops, hypermarkets and wholesalers. But the position would not have been altered by the transaction, because in these markets VUP's contribution is negligible. In addition, the merged entity's position would be open to challenge by MDS (through its publishing house Dargaud) and Flammarion, two competitors who are very active in these markets.

(811) Thus the strengthening of the vertical integration of the merged entity over the different levels of the book chain would not create a dominant position for the merged entity in the market for the sale of strip cartoons to dealers.

(3) CONCLUSION

(812) The notified transaction is therefore not such as to create or strengthen a dominant position as a result of which effective competition would be significantly impeded in the market for the sale of strip cartoons to dealers, be they bookshops, hypermarkets or wholesalers.

C.1.f. MARKETS FOR THE SALE OF SCHOOL TEXTBOOKS BY PUBLISHERS TO DEALERS

(1) FRENCH MARKET FOR THE SALE OF SCHOOL TEXTBOOKS BY PUBLISHERS TO DEALERS

(a) Introduction

(813) Before the market for the sale of school textbooks is analysed, it should be pointed out that it possesses special features which become apparent when its functioning is compared with that of the other markets for the sale of books (594). The school textbook is thus a quite specific product, and the school textbook chain consists of operators peculiar to this sector.

(814) The production of a school textbook cannot be compared with that of other categories of book, in view of its complexity and specificity. As the notifying party states (595), school textbooks generally require input from a large number of people, calling for a whole team of writers (from three to ten), documentalists and illustrators. School textbooks require a layout suitable for teaching purposes, with illustrations, reproductions, photographs, maps and tables. The production process is therefore more complex and calls for heavier investment than is the case with any other category of book (apart from reference works), and what is more school textbooks have to be produced in a very short space of time once new syllabuses are released. The investment by the parties to the transaction in school textbooks and educational supporting materials is thus extremely high as a proportion of their total investment in publishing ([35-45]% for Hachette Livre and [35-45]% for VUP (596)). This helps to explain why there are so few operators in this market, despite the fact that it is highly profitable, for two main reasons: (i) the print run for textbooks, which corresponds to a demand determined by the number of pupils per age group for each level of education, is large and known in advance; and (ii) the risk to the publisher of books being returned is very low, as the purchase and use of textbooks is compulsory.

(815) School textbooks have to follow syllabuses that are national and compulsory and whose content is decided on by the Minister for Education. While school textbook publishers have to follow these syllabuses, and any revisions thereof, they have a free hand when it comes to interpreting the syllabus and to designing and marketing textbooks.

(816) When a syllabus is changed, the Ministry of Education informs publishers accordingly. By law, such a change must be made public 12 months before it is implemented (save in exceptional circumstances) to give publishers sufficient time to design and produce textbooks and make them available to teachers and pupils.

(817) A revision of the syllabus is usually (597) what triggers a decision to publish a new school textbook. The largest generalist publishers publish almost systematically a new textbook at each revision. It has also become commonplace for some of the largest publishers to publish two competing works under the same brand, differentiating between them by means of the teaching method proposed.

(818) The choice of textbooks is a matter for the teachers responsible for implementing the syllabus, under the watchful eye of the various inspectorates. Neither the central administration nor the local one participates directly or indirectly in this selection process.

(819) At all levels, the selection is made by the teachers concerned by the change of syllabus after an individual and/or collective examination of the specimen copies sent by publishing houses. Thus, a key requirement for those wishing to operate in the market for the sale of school textbooks in France is the ability to send teachers specimens free of charge.

(820) The sending of specimens presupposes the maintenance of up-to-date indexes containing teachers. details. As a rule specimens are sent at certain times of the year (e.g. in May in the case of a textbook due for release when the schools go back in September). Promotional material for other textbooks and educational supporting materials is often enclosed. It is essential to have enough distribution capacity to dispatch specimens on the appointed dates and also to ship textbooks when the new school year begins, bearing in mind the need to deliver large quantities at very short notice in response to numerous orders placed simultaneously with thousands of bookshops.

(821) Another feature of the school textbook market is the specific nature of the marketing teams, made up as they are of educational sales representatives (délégués pédagogiques) who are often trained specifically for the school textbook market and who develop a rapport with the teachers who prescribe textbooks. Alongside their sales forces, some publishers also have specialised infrastructures, such as salesrooms, for school textbooks.

(822) According to the findings of the Commission's investigation, a textbook's price is not the prime consideration when teachers choose and prescribe a textbook. The choice is based on the teachers' assessment of the extent to which the textbooks proposed by publishers are suited to attaining the syllabus.s aims, the scientific rigour of their content, the quality of the source material and the teaching methods the teachers use. In this context, the investigation has shown that teachers adopt a fairly conservative approach in that they may stay with a particular brand for a long time. The brand image of textbooks and the reputation of the publisher are important factors here which take into account the standing among teachers and, in some cases, the particular speciality of each publishing house.

(823) The role of the bookseller in the school textbook chain must also be discussed. As the notifying party indicates, booksellers perform a logistical function and therefore the discount granted by publishers to booksellers on school textbooks is significantly lower than that on other categories of book. Moreover, school textbooks are excluded from the application of the Lang Act on uniform book prices, and it is common practice for booksellers to grant a not insubstantial further discount in turn to their customers, especially where they are local authorities or purchaser groups.

(824) As regards the financing of the purchase of school textbooks, a distinction must be made according to level of education:

- At primary school level (Order (décret) of 1890), it is the municipalities that finance and administer the purchase of textbooks. The teachers. council (conseil des maîtres) decides freely on the spending of municipal resources and therefore of the portion devoted to the purchase and renewal of textbooks.

- At lower secondary school (collège) level, the purchase of textbooks is the responsibility of the State (Act of 1977). The textbooks chosen by teachers are ordered by schools within the limits of the funds available. The renewal of worn textbooks is theoretically possible every four years.

- At upper secondary school (lycée) level (whether general, technical or vocational in nature), no law governs the arrangements to be made. Pupils. families have to pay for their school equipment. For a number of years now, some Regions have taken the initiative of contributing, in various ways, towards its financing.

(825) The transaction would lead to the creation of a dominant position for the merged entity in the French market for the sale of school textbooks to dealers through a combination of horizontal, vertical and conglomerate effects. These effects together would enable the merged entity to act, to an appreciable extent, independently of its competitors, its customers and, ultimately, the final consumer, and this might lead to an increase in retail prices, or a reduction in the discounts granted to booksellers, or a decline in the quality and diversity of publications, or all of these.

(b) Horizontal effects

(826) Hachette Livre is present in France in the sale of school textbooks to dealers through Hachette Éducation, Hatier, Didier, Foucher and Istra. VUP is present through Nathan, Bordas and Retz.

(827) The merged entity's main competitors can be divided into two categories, the first consisting of generalist publishers present across the publishing of school textbooks for all levels of education and all subjects, and the second consisting of publishers who are generally smaller and more often than not specialised. Apart from the merged entity, only two school textbook publishers belong to the first category, namely Belin, an independent publisher whose main activity is publishing school textbooks, and Albin Michel, which is present through its subsidiary, Magnard. The leading publishers in the second category are Flammarion (Delagrave and Lanore), Bréal, Bernard Lacoste and Fontaine Picard.

(828) According to information furnished by the notifying party, the merged entity had a share of the French market for the sale of school textbooks to dealers of [70-80]% in value terms in 2001 ([40-50]% for Hachette Livre: Hachette [15-25]%, Hatier [10-20]%, Didier [0-10]% and Foucher [0-10]%; and [25-35]% for VUP: Nathan [15-25]%, Bordas [0-10]% and Retz [0-10]%), ahead of Belin ([0-10]%) and Albin Michel ([0-10]%) (598).

(829) The investigation has confirmed these data, with combined market shares of [70-80]% ([40-50]% for Hachette and [25-35]% for VUP), as against [5-10]% for Belin and [0-5]% for Albin Michel (Magnard), Flammarion (Delagrave) and the other competitors. According to these estimates, the merged entity would therefore occupy a position almost 10 times stronger than that of its closest competitor. What is more, the HHI indicates an extremely high degree of concentration, at 5 965 points, and a very significant increase of 2 811 points. These indicators are strong signs of a dominant position held by the merged entity in the French market for the sale of school textbooks to dealers.

(830) In its reply to the statement of objections, the notifying party states that market shares are not an indicator of market power as each revision of the syllabus leads to a major redistribution of market shares among school textbook publishers. This statement is backed up by examples of market shares in the sale of books in certain subject areas and for specific education levels. However, if the trend in the market shares of Hachette and VUP in the markets for the sale of school textbooks is analysed, it becomes apparent that there is no significant variation from year to year. Moreover, the high level of the merged entity's combined market shares, at between 75% and 85%, removes any doubt about its market strength.

(831) This strong position of the merged entity would be enhanced by the fact that the price of textbooks is not a decisive factor in the choice made by the prescribing teachers, as long as it falls somewhere within an average range that is acceptable in the light of the budgets allocated by authorities (599), and by the fact that, on the contrary, there is a fair amount of brand loyalty on the part of teachers, which translates into a degree of stability and foreseeability of market shares. Hence, although syllabus revisions can pit publishers once more against one another for the sale of new textbooks, a certain recurrent pattern in market shares is to be observed. To this must be added the fact that publishers propose more often than not series spanning several levels of education, which enables them to win the loyalty of teachers who prefer to work with an already familiar tool used for another class studying the same subject (600).

(832) The marketing of school textbooks is a case apart in so far as it makes use of specific means aimed at the teachers who prescribe the textbooks. In this respect, there is no marketing proper directed at bookshops: booksellers place their orders on the basis of those they receive from teaching establishments601. The school textbook marketing teams are made up with teachers in mind, being composed of specialists in education known as pedagogical advisers (conseillers pédagogiques).

(833) The role of these advisers is to visit teaching establishments and present their new titles. They take part in inter-publisher exhibitions organised for example by the Institut Universitaire de Formation des Maîtres and the Centre Régional de Documentation Pédagogique. They meet primary school inspectors and teaching advisers and help conduct market surveys into how works are perceived and into new layouts. And they may organise meetings with authors. In short, the pedagogical adviser's role is essential not only in order to promote books and sales but also in order to obtain a clear perception of market expectations.

(834) The investigation has shown that, owing to the size of their network, the parties to the transaction are currently the most active in the area of promotion by pedagogical advisers. The merged entity would have a much bigger pedagogical adviser network than its competitors, which have networks more limited in scale whose functions are often shared with the marketing staff for other works, such as educational supporting materials (602). For this reason, the competitors of Hachette Livre and VUP generally select a limited number of establishments for their representatives to visit, whereas the merged entity would have a large number of representatives (Diffusion Hatier, Diffusion Hachette Éducation, Diffusion Nathan and Diffusion Bordas), enabling it to canvass the entire teaching profession. The merged entity would therefore have easier, more direct access to the teachers who prescribe school textbooks, which would enable it to "push" its works at the expense of those of its competitors. The competitors of Hachette Livre and VUP would be less well placed, moreover, to promote their books and to gather marketing information from the users of books, essential as this is to developing suitable products.

(835) Besides pedagogical advisers, the usual means of marketing and promotion is the sending of free specimens. Specimens are sent to teachers in February in the case of primary schools and May in the case of secondary schools. The cost of sending specimens is extremely high (between 10 and 20% on average of the list price exclusive of tax depending on the education level (603)) and in principle is not dependent on the number of sales. Were it not for such specimens, textbooks would not sell, as teachers would be unable to gain an idea of a book's worth (604). The investigation has shown that only school textbook publishers falling into the first category defined above, in the introduction to the analysis of the horizontal effects in the French market for the sale of school textbooks, are able to finance the sending of samples to the entire teacher population concerned, smaller publishers having to limit their target audience to part of the prescribing teachers only (605), which automatically limits their sales proportionately.

(836) The merged entity would be able to rely on its existing marketing tools, which it alone has the resources to employ. Foremost among these are the "teaching rooms" (salles pédagogiques) of Hachette Livre, which are used to transmit information to teachers and to publicise books. Teaching rooms are centres for providing advice, selling and, where appropriate, giving individual samples to visiting teachers. Hatier and Hachette Éducation have one teaching room each. The Hachette Livre group also owns the Librairie Pédagogique du Centre, which sells books by the unit and, where appropriate, sends specimens by the unit on a mail-order basis. During term time, VUP's salesrooms are also used to propose school textbooks to bookshops in their area and therefore act as additional go-betweens.

(837) The merged entity would have at its disposal a portfolio of brands unequalled in the market for the sale of school textbooks: Hachette, Hatier, Didier, Nathan, Bordas and Retz. This could confer a decisive advantage on it in a market the functioning of which is often based on the "call for tenders" system whereby, upon each revision of the syllabus, teachers must choose between the new school textbooks proposed by publishers. The merged entity might thus be able to create the impression of a choice between the brands belonging to the group by offering a range of textbooks for a given class (606). The main textbook publishing houses of Hachette Livre and VUP already have the capacity to publish several textbooks under the same brand but offering different teaching methods (607). Thus, not only would the merged entity have at its disposal a larger number of brands, but it would also be the only publisher able to offer a choice between several textbooks under the same brand. This was confirmed by the head of Hachette Éducation during an interview reported in an article in Livres Hebdo, transmitted to the Commission by the notifying party (608).

(838) This superabundance of falsely competing products - combined with huge promotional and sales canvassing resources - would limit the exposure of genuinely competing textbooks to teachers, who would have no opportunity to familiarise themselves with every work (609). By monopolising teachers' attention, the merged entity would increase its chances of being chosen by prescribers.

(839) In its reply to the statement of objections, the notifying party questions this strategy, maintaining that it would be neither economically rational nor effective. It acknowledges that the costs of creating and promoting a school textbook are high, and considers therefore that the marketing by the merged entity of a larger number of works would be risky as far as the publishing houses' financial equilibrium is concerned. However, it does not put forward any arguments which contradict the Commission's analysis and omits in particular to mention the possibility for the merged entity to share the costs both of creation and of promotion.

(840) By implementing such a strategy, the merged entity would be able to raise prices (more so in the case of upper secondary schools, where it is families that pay, than in that of primary schools and lower secondary schools, where an excessive price rise would be limited by the budgets allocated by local authorities); where a price rise would not be possible (for the reasons just given), the merged entity would be able to increase its profitability by cutting production costs (which, as explained above, are very substantial in the school textbook field) by reducing the quality of works: in a market in which calls for tenders are the norm, in so far as all the potential major tenderers were controlled by one and the same group there would be nothing to prevent them from dividing the contracts among themselves; nor would there be anything to prevent each of the publishing houses concerned from specialising in a particular area. Those publishing houses would therefore no longer be forced to make as big an effort in terms of creativity and investment as they are today by the competition in order to gain the favour of textbook prescribers. The disappearance of this constraint would lead to a drop in the quality of publications without, however, resulting in any loss of market share for the merged entity (610).

(c) Vertical and conglomerate effects

(841) The vertical integration and the strong presence or dominant position of the merged entity in the various markets in the book chain, notably at the distribution level, would help to create a dominant position in the French market for the sale of school textbooks to dealers.

(842) The merged entity's position in distribution would enable it to strengthen further its position in the sale of school textbooks. School textbooks must be shipped at a precise moment in time. Orders are taken at the beginning of the new school year within a very short period and any delay can lead to a textbook order being cancelled. The greater reliability of the merged entity's school textbook delivery service would therefore induce teachers to opt for it (611).

(843) The leading competing school textbook publishers all have their own distribution networks. The findings of the investigation have shown that smooth distribution is a decisive factor in this sector, where specimens and textbooks have to be dispatched on very precise dates when teachers make their choice of textbook, failing which exclusion from the market ensues (612). These distribution networks specialising in school textbooks are, however, much smaller than those of the merged entity, which is present in the other markets for the sale of books. Whereas Hachette Livre and VUP account for more than 50% of total distribution in France, Flammarion accounts for less than 8%, and the other school textbook publishers for less than 3%, of distribution. Moreover, apart from Flammarion, the other distributors are specialised in textbook publishing. Because of the seasonal nature of school textbook sales, there is a degree of complementarity between the flows of such books and those of other books: the distribution facilities, less used for the other categories of book during the summer holiday period, can be put to good use to distribute school textbooks (613). Textbook publishers which have a distribution facility devoted solely to school textbooks do not enjoy such flexibility.

(d) Countervailing power

(i) Supply

(844) With regard to potential competition, it must be pointed out first of all that there are high barriers to entry to the markets for the sale of school textbooks. Inasmuch as the activity of selling school textbooks to dealers is characterised by the prescribing of those books by teachers, the promotion of textbooks involves sending specimens to teachers - expenditure on which (614) constitutes a very substantial initial cost - and calls on teachers by educational sales representatives.

(845) Inasmuch as the activity of selling school textbooks to dealers responds to the prescribing of those books by teachers, publishers must first of all produce or rent indexes of teachers (more than 300 000 contacts in France), the cost of which is very high (between euro 50 000 and euro 150 000 for the rental of indexes, according to the notifying party (615)). The leading school textbook publishers belong to one of two economic interest groups, GIDEC (Groupement d'Information d'Éditeurs Classiques) and GECRI (Groupement d'Éditeurs Classiques pour le Recueil d'Informations), the function of which is to gather information on publishers' behalf on educational establishments ranging from nursery schools to universities. Hachette and VUP form part of these two groups, with Hatier presiding over GIDEC in place of Magnard following its takeover by Albin Michel, and Nathan presiding over GECRI. With [75-85]% of the market for the sale of school textbooks and a presence at every educational level, the merged entity would be able to create its own structure dedicated to the production of teacher indexes and thereby increase the cost to its competitors, which would have to bear the high cost alone.

(846) Publishers must then send out advertising material, information on school textbooks and samples (10 000-15 000 for each work to upper secondary schools and 35 000 to lower secondary schools) and have at their disposal pedagogical advisers to promote their textbooks.

(847) Publishers must be able to produce several books in the same series (616) in order to satisfy syllabuses and teacher preferences. School textbook publishing is characterised by the high cost of creating a new textbook, which may be estimated at between euro 75 000 and euro 230 000 (617) and which it takes several years to recoup (618).

(848) The results of the Commission's investigation show that teachers become fairly attached to the books they are in the habit of using and are sensitive to the image of teaching quality linked to the brands they already use. As a result, in a context where textbooks are renewed, on average, only once every four years, a new entrant to the market for the sale of school textbooks has to invest for many a year before it can hope to enjoy a reputation equal to that of the incumbent firms and establish itself in the market on a lasting basis (619).

(849) The notifying party maintains that digitisation is a competitive alternative to school textbooks and that powerful operators might enter the market for the sale of school textbooks by that means.

(850) The results of the Commission's investigation show, however, that serious obstacles (training of teachers, development of "content", lack of computer equipment in households) render the introduction of these new teaching techniques still problematic, at least as a substitute for conventional textbooks.

(851) Thus, according to a study carried out by Boston Consulting Group entitled Analyse de l'importance du numérique dans le secteur de l'éducation en France (analysis of the importance of digitisation in the education sector in France) (620), the necessary investment in data processing equipment by pupils and teachers in France is of the order of euro 6.8 billion. If this amount is compared with that of the budget earmarked by the Ministry of Youth, National Education and Research for information and communication technology in teaching, which is put by the notifying party at approximately euro 50 million for 2002 (621), it is unlikely that the resources allocated will allow school textbooks to be replaced by new technology in the near future.

(852) This analysis is shared by school textbook publishers, who state in response to the Commission's investigation that paper-based products are not about to disappear in the next few years to be replaced by new technologies, but that they should survive, with digital tools acting more as a complement to than a substitute for the textbook (622). The investigation has shown, moreover, that the textbook publisher's know-how would be irreplaceable when it comes to developing digital tools in the education sector and makes entry by an operator not present in textbook publishing, but present in, say, the data processing market, unlikely (623).

(853) The Commission's investigation has thus shown that the sums invested both by Hachette Livre and VUP and by their competitors in producing digital education modules, as well as sales of such products, remain marginal compared with the turnover achieved by these publishing houses from the sale of school textbooks (624).

(854) Although it is still too early at this stage in the development of new technology use in the education sector to decide whether digital products are to be regarded as forming part of the school textbook market, it is nevertheless possible to conclude that the merged entity would be best placed owing to its importance on this market and its capacity to invest in developing its digital market shares. The results of the investigation show that only large groups have the resources with which to mobilise staff and capital in order to operate in this sector and propose technical platforms and digital content on a large scale (625). One textbook publisher thus states that developing digital products is very burdensome for his company and that as a result it will not make a move unless forced to do so by the competition (626). Moreover, the schemes to introduce new technologies in teaching are being financed largely by the Ministry of Education and regional councils. The strength of the merged entity would be such that, from the authorities' point of view, it would be an unavoidable and privileged interlocutor when it came to drawing up and implementing any future schemes in this area (627). Not only would digital products not make it possible to counterbalance the merged entity's strong position in school textbooks, but its strong position in these new developing products constitutes an aggravating factor. There is a risk that the transaction might lead to the monopolisation of an emerging market by a single enterprise, which would secure every new outlet in the market and perhaps impose its own standards, thereby preventing other competitors from developing their businesses in this segment (628).

(855) In its reply to the statement of objections, the notifying party contests the Commission's analysis of the place occupied by digital products in the school textbook market. First, it questions the size of the investment that would be needed to introduce digitisation into the school textbook sector, stating that the sums involved are smaller than those to which the Commission refers - based though these are on information received from the notifying party - and that they are also financed by the Regions and not only by the Education Ministry - something which the Commission also pointed out. However, the amounts of investment mentioned by the notifying party, the calculation of which is, moreover, open to challenge, are not sufficiently smaller to call into question the Commission's analysis following the in-depth investigation.

(856) The notifying party maintains further that Hachette and VUP are not the best placed to cope with the emergence of digitisation, and that others - either large foreign groups present in book publishing or operators active in the IT sector - might out-distance the merged entity. On this point, the investigation has not revealed the emergence to date of such players in the school textbook field. On the contrary, it has shown that it is more important to be present in school textbook publishing than in the IT sector when it comes to developing a presence in this market in France.

(857) Neither existing competitors nor potential competitors, then, would be able to counter the merged entity's ability and incentive to become dominant in the French market for the sale of school textbooks.

(ii) Demand

(858) With regard to the countervailing power of the demand side, the notifying party considers that the merged entity might be restrained by the countervailing power exerted by the French State in its capacity as prescriber of and payer for school textbooks. The investigation has shown, however, that it is not the French State that is the prescriber, but teachers. In the case of primary education, it is the teachers that choose school textbooks, which are then paid for by the municipalities. In the case of upper secondary schools, it is the staff meetings within each school that are the prescribers, with the cost still being borne in most instances by parents. There are nowadays a number of regional initiatives aimed at taking over responsibility for financing school textbooks. That leaves lower secondary schools as the only establishments for which it can be said that the State intervenes directly in the financing of school textbooks through the regional education authorities (rectorats). But here, too, it is the schools. staff meetings that decide on the choice of textbooks.

(859) As for the possible countervailing power of booksellers, the demand side is made up essentially of operators whose size and economic strength cannot be compared to those of the merged entity. Although purchases are often made by bookseller/tenderers performing a role similar to that of wholesalers, their size is nothing like that of the merged entity. The results of the Commission's investigation show, moreover, that school textbook sales generally account for a substantial portion of the turnover of bookshops active in this segment and that they take place during a short period of the year. This seasonality and the need to build up sizeable stocks of books weaken considerably the financial position of bookshops, which as a rule do not have enough working capital to pay in advance and therefore have to negotiate a credit period and/or facilities with publishers and/or distributors (629). Even if they have the incentive to do so, therefore, dealers do not have the means to oppose the creation of a dominant position for the merged entity in these markets.

(860) Taken together, these factors lead to the conclusion that neither competitors nor customers would wield any real countervailing power.

(e) Conclusion

(861) The notified transaction would therefore create a dominant position for the merged entity as a result of which effective competition would be significantly impeded in the Common market or a substantial part of it in the French market for the sale of school textbooks.

(2) THE BELGIAN MARKET FOR THE SALE OF SCHOOL TEXTBOOKS BY PUBLISHERS TO DEALERS

(862) Hachette Livre is present in Belgium in the sale of school textbooks to dealers through Hachette Éducation, Hatier, Didier, Foucher and Istra. VUP is present there through Nathan, Bordas and Retz.

(863) The parties. main competitors are, besides the French publishers, Van In,

belonging to the Standard Book group, Labor and De Boeck.

(864) According to information furnished by the notifying party, in Belgium in 2001 the merged entity had a share of the market for the sale of French-language school textbooks to dealers of [10-20]% in value terms, behind De Boeck, with [45-55]% (630).

(865) The merged entity's small market shares and the presence of a significant competitor which holds a market share more than three times the size of that of the merged entity lead to the conclusion that the notified transaction is not likely to create or strengthen a dominant position as a result of which effective competition would be significantly impeded in the market for the sale of school textbooks in Belgium.

(3) THE LUXEMBOURG MARKET FOR THE SALE OF SCHOOL TEXTBOOKS BY PUBLISHERS TO DEALERS

(866) Hachette Livre is present in Luxembourg in the sale of school textbooks to dealers through Hachette Éducation, Hatier, Didier, Foucher and Istra. VUP is present there through Nathan, Bordas and Retz.

(867) The parties. main competitors in Luxembourg are De Boeck, Albin Michel (Magnard), Belin, Flammarion (Delagrave and Lanore), Bréal and Labor.

(868) According to information furnished by the notifying party, in Luxembourg in 2001 the merged entity had a share of the market for the sale of school textbooks to dealers of approximately [25-35]% in value terms, ahead of De Boeck, with 20-30%, and numerous other operators, each with less than 5% of the market (631). The notifying party has not furnished any information on a possible distinction by level. It does, however, point out that the Luxembourg educational model is based on a system whereby classes are taught in German at primary and secondary technical level and in French at general secondary level. It would appear, therefore, that, if a distinction were to be made by level of education, the merged entity's market shares would not be noticeably different.

(869) The notifying party states that the operation of the school textbook market in Luxembourg displays the following features. First, the textbooks used by pupils are prescribed by the Luxembourg Ministry of Education and Vocational Training, which each year draws up a list of textbooks by subject and for each class. Secondly, teaching is provided predominantly in German and there are numerous operators in the market. The prescriber, who can react to any anticompetitive behaviour by the merged entity, therefore has real countervailing power.

(870) The prescriber's strong countervailing power and the presence of a significant competitor with a market share close to that of the merged entity lead to the conclusion that the notified transaction is not likely to create or strengthen a dominant position as a result of which effective competition would be significantly impeded in the market for the sale of school textbooks in Luxembourg.

(4) THE SPANISH MARKET FOR THE SALE OF SCHOOL TEXTBOOKS BY PUBLISHERS TO DEALERS

(871) Hachette Livre is present in Spain in the sale of school textbooks to dealers through its subsidiary Bruño and VUP is present there through Anaya.

(872) The parties. main competitors are Santillana and Santa Maria.

(873) According to information furnished by the notifying party, the merged entity had in Spain in 2001 a share of the market for the sale of Spanish-language school textbooks to dealers of [15-25]% (including at least [0-10]% for Hachette Livre) in value terms, ahead of Santillana, with [15-25]%, and Santa Maria, with [10-20]% (632).

(874) The significant presence of competitors and the lack of any significant overlap lead to the conclusion that the notified operation is not likely to create or strengthen a dominant position as a result of which effective competition would be significantly impeded in the market for the sale of school textbooks in Spain.

C.1.g. MARKETS FOR THE SALE OF EDUCATIONAL SUPPORTING MATERIALS BY PUBLISHERS TO DEALERS

(875) The French-language market for the sale of educational supporting materials to dealers accounted in 2001, according to the Commission's calculations based on information furnished by the notifying party and others, for a turnover of euro 119 million. Distinguishing by sales channels, the market for sales to class 1 and class 2 bookshops is worth euro 62 million, that for sales to hypermarkets euro 35 million and that for sales via wholesalers euro 21 million. There are strong links between the market for the sale of school textbooks and that for the sale of educational supporting materials in so far as educational supporting materials are often complementary to school textbooks and are often sold under the same brand names.

(876) The results of the Commission's investigation show that the transaction would lead to the creation of a dominant position for the merged entity in the French-language market for the sale of educational supporting materials by publishers, irrespective of the type of dealer, through a combination of horizontal, vertical and conglomerate effects. These effects together would enable the merged entity to act, to an appreciable extent, independently of its competitors (whose presence in the markets might be reduced), its customers and, ultimately, the final consumer, and this might lead to an increase in retail prices, or a reduction in the discounts granted to booksellers, or a decline in the quality and diversity of publications, or all of these.

(877) In its reply to the statement of objections, the notifying party contests the likelihood of the scenario of a reduction in discounts and an increase in retail prices. It bases its case on the example of the takeover of Hatier by Hachette, and says that discounts and retail prices remained virtually unchanged after the operation. It should be noted, however, that the scenario depicted by the Commission would come about following the merger between the educational supporting materials businesses of Hachette and VUP when the main source of competition would have disappeared. The example chosen by the notifying party therefore has no probative force, the operation having taken place at a time when VUP exercised competitive pressure on the Hachette group, pressure which would disappear after the merger.

(1) HORIZONTAL EFFECTS

(878) Hachette Livre is present in France, Belgium and Luxembourg in the sale of educational supporting materials to dealers through Hachette Éducation, Hatier and Foucher. VUP is present through Nathan, Bordas and Retz.

(879) The merged entity's main competitors are Albin Michel (Vuibert and Magnard), Playbac, Belin, Flammarion, Bréal and Labor (in Belgium and Luxembourg only). De Boeck, which used to be present in the Belgian and Luxembourg markets, recently withdrew from the educational supporting materials market.

(880) According to information furnished by the notifying party, the different players. market shares in value terms in 2001 were as follows:

<emplacement tableau>

(881) According to the Commission's calculations based on information furnished by the notifying party and third parties, the market shares in value terms in 2001 of the merged entity and its main competitors in the French-language market for the sale of educational supporting materials to dealers may be estimated as shown in the following table:

<emplacement tableau>

(882) It follows from these data, whatever their source, that after the transaction the merged entity would hold extremely high market shares both in absolute terms and in comparison with those of its closest competitors. The finding of a dominant position flowing from these market shares is borne out by the degree of concentration as measured by the HHI, which comes post-merger to 6 426 points, with an increase of 3 072 points for the market as a whole (all classes combined).

(883) In its reply to the statement of objections, the notifying party maintains that the supply of digital products should be included in the market for educational supporting materials. In that event, and if the sale of multimedia encyclopaedias is also included, it considers that its market share would be [50-60]%. The Commission's investigation has not confirmed the importance of digitisation in educational supporting materials, and it has even been said that it is a segment in sharp decline in which publishers have reduced their investment. It is therefore doubtful whether the parties. market strength would be weakened, all the more so since the [50-60]% market share is already in itself a sign of a certain market strength.

(2) VERTICAL AND CONGLOMERATE EFFECTS

(884) Including internal sales and the sales of publishers whose works it markets (notably Magnard, through hypermarkets and wholesalers), the merged entity would account for nearly [75-85]% of the turnover of bookshops, [80-90]% of that of hypermarkets and [75-85]% of that of wholesalers. It would be able on its own to furnish a sufficient supply of products to cover the entire needs of its customer dealers. It would possess an unequalled portfolio of brands which would benefit from their brand image, linked as it is to other sectors such as school textbooks and reference works.

(885) The merged entity would have other major advantages over other publishers, including a strong sales network for distributing works among the general public via the largest number of sales outlets of all classes, and considerable financial capacity in the areas of marketing and advertising (633). Market players have emphasised the importance of this consideration, but the Commission takes the view that in the circumstances it is not a decisive factor that would bring about the creation of a dominant position in this market.

(886) The merged entity's significant presence in almost every market for the sale of books, all categories combined, and in particular in the sale of school textbooks, together with its substantial activities in markets related to book publishing such as magazines, would further strengthen its dominant position in the markets for the sale of educational supporting materials (634), thereby enabling it to act independently of its competitors, its customers and, ultimately, consumers.

(3) COUNTERVAILING POWER

(a) Supply

(887) Besides occupying positions markedly inferior to that of the merged entity in the market for the sale of educational supporting materials, the merged entity's competitors are not characterised by the same degree of vertical integration as the merged entity, and in particular have a less extensive marketing and distribution network.

(888) As to potential competition, the Commission's investigation has shown that, while it does not have entry barriers as high as those of the market for the sale of school textbooks in particular in relation to marketing costs, the market for the sale of educational supporting materials is characterised by the need to hold a brand that is already well known in the school textbook field (635). Since all significant school textbook publishers are already active in the sale of educational supporting materials, any new entrant would therefore first have to establish itself in the market for the sale of school textbooks, where there are high barriers to entry (636). The Commission acknowledges that some educational supporting material publishers are absent from school textbook publishing and that therefore a presence in school textbook publishing is not an obligation and does not constitute a barrier to entry; however, a brand recognised in the school textbook field remains a factor conducive to successful establishment in the educational supporting materials field.

(889) The notifying party maintains that digitisation is a competitive alternative to educational supporting materials and that powerful operators might enter the market for the sale of educational supporting materials by this means. The Commission's investigation has, however, not revealed conclusive evidence of any foreseeable competition from new technologies in this market in the long term. It would appear rather that digital products are complementary to paper-based products owing to their additional functionalities and their interactivity. Inasmuch as digital products are still at the development stage, it is impossible to predict whether they will constitute a separate market in future or whether all educational supporting materials will become integrated into a digital module. At all events, Hachette and VUP today hold a very strong position in digital product development, compared with the so far fairly limited activities of other educational supporting material publishers in this domain (637). It is therefore unlikely that, in the short or medium term, the arrival of digital products on the market would be able to counterbalance the strong position of the merged entity. Moreover, the investigation has shown that the know-how of the school textbook/educational supporting material publisher is irreplaceable when it comes to developing digital tools in the educational supporting material sector and makes the probability of entry by an operator absent from the school textbook publishing market, but present in, for example, the data processing market, unlikely. Rather than counterbalancing the merged entity's position, the development of products in digital format could be used by the merged entity to exclude its competitors, which at present do not have sufficient financial resources to develop this kind of medium and therefore risk becoming dependent on the standards and modules developed by Hachette Livre and VUP.

(890) In its reply to the statement of objections, the notifying party contests the Commission's analysis of the importance of digitisation in the educational supporting materials market but does not adduce any evidence calling into question the above analysis.

(b) Demand

(891) As far as the countervailing power of buyers is concerned, it has already been shown (638) that the demand side is made up essentially of operators whose size and economic strength cannot be compared to those of the merged entity and which would not be able individually and prospectively to counter the merged entity's strategy of dominance. Even if hypermarkets - despite the merged entity being able to provide them with a sufficient supply of educational supporting materials - were to seek another supplier, at a concentration level of more than 80% they would not always have a competitive alternative. For instance, in the holiday workbook segment, only Magnard (Albin Michel) proposes a genuine range of products rivalling that of the merged entity, but the products are marketed to hypermarkets by Hachette Livre and are therefore dependent on the merged entity.

(892) Taking these factors together, it must be concluded that neither competitors nor customers would have any real countervailing power.

(4) CONCLUSION

(893) The notified transaction would therefore create a dominant position for the merged entity as a result of which effective competition would be significantly impeded in the Common market or a substantial part of it in the French-language market for the sale of educational supporting materials by publishers to dealers.

C.1.h. MARKETS FOR THE SALE OF REFERENCE WORKS BY PUBLISHERS TO DEALERS

(1) SMALLER REFERENCE WORKS

(a) Dictionaries

(894) The transaction would lead to the strengthening of dominant positions held by the merged entity in the sale of dictionaries to dealers. The creation of a near-monopoly would enable the merged entity to act independently of its competitors and its customers, and this might result in an increase in retail prices, or a reduction in the discounts granted to booksellers, or a decline in the quality and diversity of publications, or all of these.

(i) Horizontal effects

(895) The principal effect of the transaction in these markets would be to strengthen the presence of VUP however the markets are segmented. The elimination of the Lagardère group as a supplier of dictionaries to dealers, in competition with VUP, the unchallenged leader, would be reflected in an immediate reduction in competition in these markets, with two of the largest potential suppliers of such works then forming a single entity. Before the transaction, VUP was already in a particularly strong - not to say dominant - position in the markets for the sale of dictionaries, with Hachette Livre as its main competitor.

(896) Hachette Livre is present in France, Belgium and Luxembourg in the sale of dictionaries to dealers through Hachette Éducation, Marabout and LGF. VUP is present there through Larousse, Dictionnaires Le Robert, Bordas, Nathan and Harrap. The merged entity's main competitors are Micro Application in France and Le Cri in Belgium.

(897) In the French-language market for single-language dictionaries, the merged entity would, according to Commission estimates based on information furnished by the notifying party and its competitors, have a market share of [90-100]% (with Hachette Livre accounting for [5-15]%). It would therefore be in a near-monopoly position.

(898) In the French-language market for bilingual and multilingual dictionaries, the merged entity would, according to Commission estimates, have a market share of [90-100]% (with Hachette Livre accounting for [0-10]%). It would therefore be in a near-monopoly position.

(899) In the French-language market for dictionaries (both single-language and bilingual/multilingual), the merged entity would, according to Commission estimates, have a market share of more than [90-100]%. It would therefore be in a near-monopoly position.

(900) The Commission's investigation has also confirmed that the merged entity would be in a very strong position whatever the class of dealer (639).

(901) In the light of the above considerations, the merged entity's large market shares, constituting as they do a near-monopoly, would lead to a strengthening of dominant positions in these markets whatever the sales channel or type of dictionary. Even if Hachette Livre's contribution is fairly small, especially in the case of bilingual and multilingual dictionaries, at least it performed a role of potential competitor in so far as it was the only player in the market to be of a size globally comparable to that of VUP vis-à-vis the various dealers and to be in the business of selling reference works.

(902) In its reply to the statement of objections, the notifying party maintains that adding Hachette Livre's market share to that of VUP would have a negligible impact. The findings of the Commission's investigation have shown, on the contrary, however, that the competition between Hachette Livre and VUP in the market for the sale of dictionaries is real and that it is clearly identified by VUP, as can be seen from internal documents (640). The disappearance of Hachette Livre as a competitor to VUP cannot therefore be passed off as marginal, all the more so since the notified transaction would lead to the creation of a near-monopoly.

(ii) Countervailing power

* Supply

(903) Not only are there hardly any competitors in the various markets for the sale of dictionaries, but what competitors there are are not characterised by the same degree of vertical integration as the merged entity at all levels of the book chain and they do not have privileged access to various kinds of media services as is the case with the merged entity.

(904) As to potential competition, in view of the barriers to entry and expansion existing in this sector it is not credible to consider this argument to be a factor countering the creation by the merged entity of dominant positions in the markets for the sale of dictionaries. The Commission's investigation has shown that the costs of producing, promoting (641) and establishing a brand are high. The notifying party maintains that the arrival on the French market of Micro Application, with a dictionary on a CD-ROM, illustrates the possibility of penetrating the market and therefore of re-establishing effective competition. However, although Micro Application did launch a dictionary in 1998, it still has a market share of less than 1%, which illustrates its relative lack of success in penetrating the market.

(905) The notifying party also maintains in its reply to the statement of objections that software producers, such as Microsoft, would exert competitive pressure on the merged entity, citing as proof the case of the Microsoft .Word. software dictionary. The Commission considers, however, that word processing software such as "Word" is not a substitute for the dictionaries published by VUP and Hachette Livre and is not about to become so in the near future.

(906) Competitors, both existing and potential, would therefore be unable to counter the strengthening of dominant positions held by the merged entity in the various markets for the sale of dictionaries to dealers.

* Demand

(907) As far as the countervailing power of buyers is concerned, it has already been shown (642) that the demand side is made up essentially of operators whose size and economic strength cannot be compared to those of the merged entity and which would not have any alternative supplier. Those operators would therefore have no means of countering the merged entity's dominance.

(908) Taking these factors together, it must be concluded that neither competitors - both existing and potential - nor customers would have any real countervailing power.

(iii) Conclusion

(909) The notified transaction would therefore strengthen a dominant position as a result of which effective competition would be significantly impeded in the Common market or a substantial part of it in the French-language markets for the sale of dictionaries by publishers to dealers.

(b) Smaller general encyclopaedias

(910) The transaction would lead to the strengthening of a dominant position held by the merged entity in the markets for the sale of smaller general encyclopaedias to dealers. This dominant position would enable the merged entity to act independently of its competitors and its customers, and this might result in an increase in retail prices, or a reduction in the discounts granted to booksellers, or a decline in the quality and diversity of publications, or all of these.

(i) Horizontal effects

(911) Hachette Livre does not publish any smaller general encyclopaedias. VUP is present in France, Belgium and Luxembourg in the sale of single-volume general encyclopaedias through Larousse and Robert Laffont. The merged entity's main competitors in France are Chanteclerc, Dargaud, Gallimard and Magnard.

(912) According to information furnished by the notifying party, VUP has a share of the French-language market for the sale of smaller general encyclopaedias of [50-60]%.

(913) If the market for smaller general encyclopaedias were to be segmented according to the various types of dealer (643), the Commission's investigation has confirmed that the merged entity would be in a strong position in every class, especially in France.

(914) In view of the foregoing, the merged entity's large market shares are a sufficient ground for presuming that VUP currently has a dominant position in the sale of smaller general encyclopaedias in the French-speaking countries, whatever the sales channel.

(915) Even if Hachette Livre is not present in the markets for the sale of smaller general encyclopaedias, it at least played the role of a potential competitor. Since it is present in the market for the sale of larger general encyclopaedias and in view of its financial strength, Hachette Livre would have been able easily and quickly to gather together the content needed to produce single-volume general encyclopaedias. And since it is present in the markets for the sale of smaller dictionaries, it could just as easily and quickly have gained access to the sales channels for single-volume general encyclopaedias, these being the same as for dictionaries. Consequently, since Hachette Livre was one of the most credible potential entrants to these markets, the transaction would lead to the strengthening of a dominant position by the elimination of potential competition.

(ii) Vertical and conglomerate effects

(916) The vertical integration and strong presence or dominant position of the merged entity in the various markets of the book chain, notably in marketing/distribution services for sales to hypermarkets, which are a major sales channel for smaller reference works, would help strengthen dominant positions in the markets for the sale of smaller general encyclopaedias to dealers.

(iii) Countervailing power

* Supply

(917) As well as occupying weaker positions than the merged entity in the various markets for the sale of smaller general encyclopaedias to dealers, the merged entity's competitors are not characterised by the same degree of vertical integration as the merged entity at all levels of the book chain and do not have privileged access to various types of media service as is the case with the merged entity.

(918) As far as potential competition is concerned, in view of the barriers to entry and to expansion existing in this sector (644) it is not credible to consider this to be a factor countering the creation of dominant positions by the merged entity in the markets for the sale of smaller general encyclopaedias. Competitors, both existing and potential, would therefore be unable to prevent the strengthening of dominant positions held by the merged entity in the markets for the sale of smaller general encyclopaedias.

* Demand

(919) With regard to the countervailing power of buyers, it has already been established (645) that the demand side is made up essentially of operators whose size and economic strength are not comparable to those of the merged entity and who would have no means of countering the strengthening of the merged entity's dominant position, which could lead to an increase in the price payable by the final consumer.

(920) Taking these factors together, it must be concluded that neither competitors, both existing and potential, nor customers would have any real countervailing power.

(iv) Conclusion

(921) The transaction would therefore strengthen a dominant position held by the merged entity as a result of which effective competition would be significantly impeded in the Common market or a substantial part of it in the markets for the sale of smaller general encyclopaedias to dealers.

(c) Smaller specialist encyclopaedias

(922) Hachette Livre is present in France, Belgium and Luxembourg in the sale of smaller specialist encyclopaedias to dealers through LGF. VUP is present there through Larousse, Le Robert, Robert Laffont, Nathan and Bordas.

(923) The merged entity's main competitors in the sale of smaller specialist encyclopaedias to dealers are Citadelles, Mazenod, Gallimard (Découverte) and, on a joint publishing basis, Albin Michel and Encyclopaedia Universalis.

(924) According to information furnished by the notifying party, the merged entity had in France in 2001 a share of the market for the sale of smaller specialist encyclopaedias to dealers of [0-10]% in value terms (646). In Belgium in 2001 the merged entity had a market share of [15-25]% in value terms (647). The notifying party estimates the merged entity's market share in Luxembourg in 2001 at [0-10]% (648).

(925) In a French-language market comprising France, Belgium and Luxembourg, the merged entity's share in 2001 of the market for the sale of smaller specialist encyclopaedias to dealers can be put at [0-10]%.

(926) The merged entity's small market shares in the various relevant geographic markets lead to the conclusion that the notified transaction would not create or strengthen a dominant position as a result of which effective competition would be significantly impeded in the various markets for the sale of smaller specialist encyclopaedias to dealers.

(d) Larger reference works in multimedia form

(927) Hachette Livre is present in France, Belgium and Luxembourg in the sale to dealers of larger reference works in multimedia form through Hachette Multimédia. VUP is present there through Larousse, Le Robert and Bordas.

(928) The merged entity's main competitors in the sale to dealers of larger reference works in multimedia form are Microsoft, Micro Application, Encyclopaedia Universalis and Planeta.

(929) According to information furnished by the notifying party, the merged entity had in France in 2001 a share of the market for the sale of larger reference works in multimedia form of [15-25]% in value terms, behind Microsoft, with [35-45]%, and Encyclopaedia Universalis [25-35]% (649). The notifying party estimates Microsoft's market share in Belgium at [75-85]% (650). It puts the merged entity's market share in Luxembourg at less than [0-10]% (651).

(930) In a French-language market comprising France, Belgium and Luxembourg, the merged entity's share of the market for the sale to dealers of larger reference works in multimedia form should not be larger than that which it holds in France.

(931) The merged entity's small market shares in the various relevant geographic markets and the presence of significant competitors lead to the conclusion that the notified transaction would not create or strengthen a dominant position as a result of which effective competition would be significantly impeded in the various markets for the sale to dealers of larger reference works in multimedia form.

C.1.i. MARKETS FOR THE SALE OF BOOKS PUBLISHED IN INSTALMENTS

(932) Hachette Livre is present in France and Belgium in the sale to dealers of books published in instalments through Hachette Collections. VUP is present there through an indirect 50% stake in Del Prado.

(933) The parties. main competitors in the sale of books published in instalments are De Agostini, the European leader, RCS Libris, Les Éditions Atlas and Fabbri Orbis.

(934) According to information furnished by the notifying party, the merged entity had in France in 2001 a share of the market for the sale of books published in instalments of [25-35]% in value terms, on a par with Les Éditions Atlas, likewise with [35-45]%, and ahead of Fabbri Orbis, with [5-15]% (652).

(935) In Belgium in 2001 the merged entity had a share of the market for the sale of books published in instalments of [20-30]% in value terms, behind Les Éditions Atlas, with [35-45]%, and ahead of Fabbri Orbis, with [5-15]% (653).

(936) In a French-language market comprising France, Belgium and Luxembourg, in 2001 the merged entity had a share of the market for the sale of books published in instalments of [25-35]% (Hachette Livre, [15-25]%; VUP, [5-15]%) in value terms, behind Les Éditions Atlas, with [25-35]%, and ahead of Fabbri Orbis, with [5-15]%.

(937) The presence of significant competitors and the absence of fear of eviction among the other market operators lead to the conclusion that the notified transaction would not create or strengthen a dominant position as a result of which effective competition would be significantly impeded in the various markets for the sale to dealers in books published in instalments.

C.2. MARKET FOR THE SALE OF BOOKS BY WHOLESALERS TO CLASS 3 RETAILERS

(938) As indicated in the section on the definition of product markets, the market for the sale of books by wholesalers to retailers should not be subdivided by category of work concerned.

(939) In order to be able to serve the customers of class 3 retailers, any supplier must be capable of proposing a sufficiently representative, comprehensive supply of the best-selling works at any given time. To this end, it must either hold a sufficient stock of its own, and hence be vertically integrated, and purchase some books from publishers to supplement its own stock, or purchase all books from other publishers for subsequent resale to retailers. It must also have suitable sales forces and a salesroom network making it possible to resupply retailers.

(940) In order to be able to sell their works to class 3 sales outlets, publishers competing with Hachette Livre and VUP have no choice but to go through these wholesalers whoever they may be. Inasmuch as a comprehensive, representative supply of best-sellers is necessary to satisfy the demand emanating from this type of sales outlet, independent wholesalers must obtain supplies from all publishers. The same goes, therefore, for the integrated wholesaling structures of Hachette Livre and VUP, which obtain part of their supplies from publishers competing with their own publishing houses (654).

(941) According to the findings of the Commission's investigation, the notified transaction would lead to the creation of a dominant position for the merged entity in the market for the sale of books by wholesalers to class 3 retailers through a combination of horizontal and vertical effects which would enable the merged entity to impede effective competition to a significant extent in this market.

C.2.a. HORIZONTAL EFFECTS

(1) PLAYERS

(942) Hachette Livre and VUP are present in the market for the sale of books to class 3 retailers through LDS and La DIL (655), their wholesaling structures. They are the only publishers to possess integrated wholesaling structures. LDS achieves a turnover which may be estimated at approximately euro [...]* million. La DIL achieves a turnover which may be estimated at approximately euro [...] million. The customers of LDS and La DIL are local sales outlets, supermarkets, and those retailers who frequent salesrooms. Their profile is therefore typically that of a class 3 sales outlet, in so far as their main business is not that of selling books and they do not usually employ assistants whose sole purpose is to sell books, as a result of which they need a wholesaling structure to serve them. As the notifying party states, "LDS also serves wholesalers who work with the same type of retailer" (656). LDS and La DIL have the same features as the other wholesalers operating in this market in so far as they provide the same specific rack-jobbing services (657) to the same types of sales outlet and have sales counters to which purchasers can come to obtain their own supplies.

(943) The notifying party is also present in this market through its Relay structure. Relay shops, which are class 3 sales outlets, effect a large part of their book purchases through an in-house central purchasing facility (658) to which all publishers have to turn if they wish to sell their titles in most Relay shops (659), the remainder of Relay's purchases being made directly from the principal publishers' marketers (660). It can therefore be considered that this part of sales by Relay shops is made via a structure equivalent to the in-house wholesaling structures of Hachette Livre and VUP (661). As a result, the merged entity's market shares in the market for the sale of books by wholesalers to class 3 retailers would include the turnover achieved by the Relay central purchasing facility from its (internal) sales to Relay shops (662).

(944) The notifying party has put forward the argument that the wholesaling activities of the parties to the transaction are more complementary than substitutable in so far as they do not sell their works to each other directly, which, in the notifying party's opinion, justifies the parallel existence of the two structures despite the small size of the sales outlets they serve. The transaction would therefore have a very limited impact at this level.

(945) In this connection, the notifying party fails to take into account the fact that there is strong substitutability between the works sold at this level by the wholesaling structures of Hachette Livre and VUP in so far as their respective publications cover practically every book category and these wholesalers buy, albeit in different proportions, the "must-have" works of outside publishers. The findings of the Commission's investigation have thus shown that, before the transaction, a class 3 retailer can still carry on business using the titles of either of these wholesalers as its main source of supply, supplemented by any additional works proposed by an independent wholesaler or wholesalers.

(946) The competitors of LDS, La DIL and the Relay central purchasing facility (though this supplies only Relay shops) are either smaller or much smaller than the merged entity's structures. Unlike the merged entity's wholesaling structures, moreover, competitors are generally limited to a clearly defined geographic area or specialise in a particular type of retailer in a larger geographic area, or, in the case of the smallest competitors, in a given range of works in a clearly defined area.

(947) DNL is thus a supraregional wholesaler which supplies only to supermarkets. It is active in mainland France and Corsica (and, formerly, part of Belgium), and achieves a turnover of euro [45.55] million.

(948) The vast majority of other wholesalers (663) are regional or even local in dimension (664) and achieve for the most part a turnover of less than euro 1 million, with a few notable exceptions, namely Générale du Livre (Paris area, less than euro 15 million), Les 3 Épis (western France, less than euro 15 million), SFL (Paris area, less than euro 10 million), La Caravelle (Belgium and, formerly, northern France, approximately euro 10 million), Distri Corse Livre (Corsica, approximately euro 5 million), Daphné (Belgium and, formerly, northern France, approximately euro 5 million) and Comptoir du Livre (Toulouse area, less than euro 5 million).

(2) MARKET SHARES

(949) In this market, following the notified transaction the merged entity would hold a minimum combined market share of around [50-60]% ([30-40]% for Hachette Livre and [10-20]% for VUP) (665) of sales to class 3 sales outlets made through wholesalers, ahead of DNL, which has [10-15]%, while other operators, such as Comptoir du Livre or Générale du Livre, each account for not more than 5% of the market.

(950) A distinction should be drawn between the different ways in which effective competition might be significantly impeded as a result of the merged entity's dominant position in this market, in the light of the various possible positions occupied by the merged entity in small sales outlets and supermarkets.

(951) In those sales outlets where, before the transaction, LDS and La DIL were both present, the merged entity would be in a position of considerable strength vis-à-vis dealers (666) and hence would be able to impede effective competition to a significant extent by inter alia providing them with a lower quality of service and/or reducing the discount granted to them so as to maximise the economic profitability of its joint presence in these sales outlets. Since retailers would no longer be in a position, as a result of the notified transaction, to benefit from competition between Hachette Livre and VUP in this market, the merged entity would probably find it even easier to engage in such conduct because the unilateral effects on the list price exclusive of tax (667) would mean that the relative reduction in the discount granted to retailers could be offset, though in absolute terms and in part only. It is therefore not surprising that, in view of the reduction in the discount, the sales outlets concerned, i.e. those whose book-generated turnover is mostly earned from the merged entity's titles, wonder whether "there would still remain any possibility of competition in the long term" (668) and state that "for us, [the merger between Hachette and VUP] means the loss of 5% off our discount and longer distribution times (five days)" (669). This risk is confirmed by Générale du Livre, a wholesaler competing with the merged entity, which states that "the discounts granted by these two groups [to class 3 retailers] are already among the lowest in the market. It is to be feared that the general terms of business might make the level of discounts subject to a sizeable stake in their shops and to acceptance of automatic consignments". (670)

(952) In the case of those sales outlets where, before the notified transaction, only one of the parties thereto was present with its whole range of books, or neither of them was, so that they did not account for the totality of the dealer's supplies, the significant impeding of effective competition resulting from the merged entity's dominant position might take the form of reducing the discounts granted to retailers or of using those discounts strategically (671) in order to impose itself as the new lead wholesaler holding most of the works needed to supply those retailers in-house (672) and thus to contribute to the marginalisation, or even elimination, of competing wholesalers. The existence of this state of affairs is confirmed by Générale du Livre in its analysis of the effects of the transaction, about which it states that "a transaction - grouping together the resources of these two structures and accounting for 60% of French production would pose a major threat to the survival of other independent wholesaling structures" (673).

(953) Because of its dominant vertical integration in the sale of books to wholesalers (especially general literature in large format and pocket format, guides and manuals etc"), the merged entity would have the capacity to increase the input prices paid by its rivals in the market for the sale of books by wholesalers to class 3 retailers; this would enable it to take the place of existing wholesalers as supplier to such outlets, and would thus help to strengthen its dominant position in this market.

(954) Dealers in this market generate relatively little turnover from books, so that in order for the coverage of a sales outlet by a wholesaler to be profitable the wholesaler must account for a high proportion of the outlet's total book purchases (674). Before the notified transaction, Hachette Livre and VUP separately covered a proportion of supplies to class 3 outlets which, though sizeable, was limited by the fact that they did not purchase each other's books (675), and they did not therefore individually consider it profitable to supply certain sales outlets that were supplied by independent local or regional wholesalers who could supply a range including books published by both Hachette Livre and VUP. Since the notified transaction would enable the merged entity to increase significantly, if not double, its potential turnover from a sales outlet, such previously unprofitable sales outlets would become profitable simply by dint of the Hachette Livre and VUP catalogues being combined. Because of its dominant vertical integration in the sale of books to wholesalers (especially general literature in large format and pocket format, guides and manuals etc"), the merged entity would have the capacity to increase the input prices paid by its rivals in the market for the sale of books by wholesalers to class 3 retailers; this would enable it to take the place of existing wholesalers as supplier to such outlets, and would thus help to strengthen its dominant position in this market.

C.2.b. VERTICAL EFFECTS

(955) The creation of a dominant position for the merged entity in the market for the sale of books by wholesalers to class 3 retailers would be enhanced by its vertical integration into the upstream markets for the sale of books by publishers to wholesalers, and in particular by its strong positions in the markets for the sale of works of general literature to those same wholesalers, in both large and pocket format.

(956) Of all the works sold by publishers to independent wholesalers, about one quarter are marketed by the publishing houses of the merged entity or by the outside publishers which it markets (676). Although this proportion may not correspond immediately to the levels generally associated with the generation of competition problems stemming from vertical economic integration, it is nevertheless significant, for a number of reasons.

(957) First, it represents an average, meaning that of the independent wholesalers competing with the merged entity some are much more dependent than others on the works sold to them by the merged entity's publishers. There is thus a fairly wide dispersion of independent wholesalers. dependency around the mean. For example, 60% and 55% respectively of the turnover of Générale du Livre and Roussillon-Diffusion is earned from books published by the merged entity, while only 13% of the turnover of Comptoir du Livre is earned from such books. The scale of the vertical effects on this market varies, therefore, from case to case and is such that they cannot be excluded from the analysis.

(958) Second, the average level of vertical integration also reflects the fact that, before the transaction, some independent wholesalers have already tried to reduce as far as possible their dependence on suppliers involved in the same activities as themselves, and on Hachette Livre in particular (677), by seeking to limit purchases of their products to "essential works which they cannot do without (approximately 20% of supplies to class 3 dealers), and to buy from other suppliers the most substitutable books, such as guides and manuals or children's books", as attested to by Comptoir du Livre (678). This attempt at protection made even prior to the notified transaction is in fact due to the way Hachette Livre, acting as an independent entity, has behaved toward certain wholesalers. Hachette Livre has already been investigated by the competent French authorities, and has been found guilty of anticompetitive practices against a rival independent wholesaler (679). In this connection, the French authorities concluded that the fact that "DNL is treated unfavourably - has the effect of impeding DNL.s activity, or even, ultimately, of ousting it from the market for the supplying of supermarkets" and that "this risk is brought about by the paradoxical situation whereby Hachette Livre is, at one and the same time, both supplier and rival to DNL and in competition with itself where supermarkets are concerned" (680). Another independent wholesaler, Les 3 Épis, states that "Hachette's commercial policy strikes them as abusive, as they receive no explanation about the terms granted and because they are not supplied with goods in accordance therewith" (681). (959) Thus, whereas before the transaction it was still possible for some wholesalers - primarily those acting essentially as a complement to LDS and La DIL and those specialised particularly in a specific range of books such as specialist works of a regional dimension - to reduce their economic dependence on Hachette Livre and VUP, in their capacity as independent entities, now the structural changes resulting from the notified transaction would reduce their past efforts to naught, owing to the immediate increase in their dependence on the merged entity.

(960) It cannot therefore be ruled out that anticompetitive practices such as those engaged in by Hachette Livre vis-à-vis DNL might also be applied to wholesalers which, individually, were, before the notified transaction, less dependent on LDS and La DIL taken separately. In this connection, Générale du Livre, a wholesaler competing with the merged entity, states that "while the good relations we enjoy with independent publishers enable us to talk with them and negotiate conditions of sale on good terms and in a spirit of partnership, the same cannot be said of the two major groups, Hachette and Vivendi, which impose draconian terms markedly worse than those to which a medium-sized general bookstore might aspire. These general terms of business have been tailor-made for wholesalers and are designed to stifle the competition which in their view [independent] wholesalers engage in vis-à-vis their own distribution and wholesaling structure (La DIL and LDS)" (682). Générale du Livre concludes that "it is to be feared that increased concentration in this sector may have the effect of further reducing the discounts granted to wholesalers with the aim of expanding and making profitable the DIL/LDS structure. The first to lose out will be the wholesalers, whose disappearance would be a severe blow to class 2 and 3 bookshops" (683).

C.2.c. COUNTERVAILING POWER

(1) EXISTING COMPETITION

(961) Besides occupying positions markedly inferior to those of the merged entity in the market for the sale of books by wholesalers to class 3 retailers, the wholesalers competing with the merged entity are, unlike the latter, not present as suppliers in the upstream market for the sale of books by publishers to wholesalers. This rules out any attempt by competitors of LDS and La DIL to exercise any countervailing power. Any decision by a competing wholesaler no longer to purchase works from the merged entity's publishers, in an attempt to exert countervailing power, would only cause difficulties for itself, given the importance of the merged entity's products to the average assortment of a wholesaler (684).

(962) As regards the potential expansion of existing competitors, the principal barrier is the need for any wholesaler seeking to thwart the merged entity's market strength to be vertically integrated into the sale of books by publishers to wholesalers. In other words, wholesalers would have to act as publishers to a substantial extent, ideally matching in this respect the merged entity, if they were to stand any chance of countering the latter's power. In view of the dominance of the merged entity's wholesaling structure, such vertical integration would in practice have to be on such a scale that common sense dictates that no wholesaler could envisage such a strategy.

(2) POTENTIAL COMPETITION

(963) As regards potential competition, operators likely to have an interest in entering the markets for the sale of books by wholesalers to class 3 retailers in a credible manner and to a significant extent are essentially publishers that might want their own wholesaling structure in order to ensure access to such sales outlets in keeping with their representativeness throughout the markets for the sale of books by publishers to retailers. In view of the dominance of the merged entity in this market, it is scarcely conceivable that any other individual operator might decide to establish itself as an independent wholesaler so as to try to compete with the merged entity. The only possible eventuality would be the establishment of a niche wholesaler with a local or regional dimension, which would therefore not really constitute a full-blooded competitor with the merged entity.

(964) Would it be feasible for an established publisher, competing with the merged entity in the upstream market for the sale of books by publishers to wholesalers, to penetrate the market for the sale of books by wholesalers to retailers, and hence to establish itself as an integrated wholesaler? A number of entry barriers make this possibility scarcely credible and highly unrealistic.

(965) As has already been explained elsewhere, the establishment of an efficient wholesaling structure has to be based on a sufficiently large turnover to make profitable the network needed to cover class 3 sales outlets effectively. To this end, a wholesaler must have at its disposal not only a sufficiently large quantity of books for sale, either in-house via publishers forming part of the group to which it belongs or via outside, client publishers, but also and above all a range of books corresponding to the categories of work traditionally marketed at this customer level (i.e., mainly best-sellers, pocket-format books, strip cartoon albums, etc").

(966) Unless it is already successful in the markets for the sale of books in large format, a publisher desirous of establishing itself as a wholesaler with a view to countering the merged entity's market strength must be able to market pocket-format books, which are one of the product categories that sell best in class 3 sales outlets. The example of Albin Michel is revealing in this respect. Although it is one of the publishers richest in talent, and is highly successful, it has never, at least not directly, had its own pocket-format book series. This alone would suffice to explain why Albin Michel is not integrated into a wholesaler-type marketing activity despite the fact that the kind of works it publishes meets the demand from class 3 sales outlets. As explained elsewhere, for these two activities Albin Michel depends economically on the merged entity, the dominant operator in both areas.

(967) The likelihood of competing wholesalers grouping together in one and the same organisation is unrealistic and lacking in credibility, because what is involved here is a large number of relatively fragmented, independent entities which are therefore difficult to concentrate from a practical, legal and economic point of view. In addition, the Commission finds it difficult to regard as a valid alternative, exerting effective competitive pressure on the merged entity, any speculation as to the feasibility of a reaction of this kind to the transaction.

(968) As regards the possibility of independent publishers grouping together to set up an alternative common wholesaling structure, past experience has shown that, even before the notified transaction, these publishers competing with Hachette Livre and VUP - the two largest publishing groups in the French-speaking world - were incapable of creating such a wholesaling structure among themselves, despite the fact that the market leader, LDS, was barely half the size of the merged entity and therefore potentially more vulnerable, and at all events less integrated at all levels of the book chain than the merged entity would be (685). Although in its reply to the statement of objections (686) the notifying party considers that implementation of the transaction would constitute a very strong incentive to induce competing publishers to group together, and that the benefits of such a structure to the publishers and marketers forming part of it would far exceed the direct profit from selling books, it does not show how, in practice, such a structure could see the light of day and be exploited competitively, except as a grouping "not [of] the combined range of Gallimard, Seuil and Flammarion, but rather of [the] combined range of the competitors of the new entity not marketed or distributed by it". In other words, the solution proposed by the notifying party is none other than the banding together of all publishing houses other than those belonging to the Hachette Livre and VUP groups. Even if this - to say the least - bold proposal is regarded as practically seductive, it has never been realised even on a much more limited scale, and it is not reasonable to regard it as conceivable or credible (687).

(969) Moreover, as a matter of principle, the Commission finds it difficult to regard as a valid alternative, exerting effective competitive pressure on the merged entity, any speculation as to the feasibility of a reaction of this kind to the transaction. And even if it were conceivable, this additional concentration of the market for the sale of books by wholesalers would still not exert the desired countervailing power. To do that, as has been explained above, the new, alternative, concentrated wholesaler would also have to be concentrated in the sale of books by publishers to. In order to overcome all the potential problems of trade-offs (688) between these publishers' books, and hence of economic inefficiency, it would be necessary ultimately to consider merging the marketing teams, and thus in practice also the publishing houses. Beyond the mere fact that this is hardly feasible, for obvious practical and economic reasons, the Commission cannot base its competition analysis on a hypothesis that is scarcely conceivable and therefore unrealistic and lacking in credibility.

(970) In view of the foregoing, competitors, both existing and potential, would be unable to prevent the merged entity from acting independently of its competitors in the market for the sale of books by wholesalers to class 3 retailers.

(3) DEMAND

(971) As explained above, in this market the demand side is made up essentially of operators whose size and economic strength cannot be compared to those of the merged entity, even allowing for the fact that, as far as the sale of books in supermarkets is concerned, there are a number of operators, such as major distribution chains, which have some countervailing power in their capacity as buyers. According to the findings of the Commission's investigation, the vast majority of sales outlets unquestionably have no means of countering the effects of a dominant position held by the merged entity; immediately after the implementation of the notified transaction, the merged entity would in some of these sales outlets be the sole supplier of books. Super U Unival states in this connection that "if the new group tries to reduce the discount level we have no way of preventing it" (689). Even shops as large as the E. Leclerc establishments point out that "owing to the position of strength which the new Hachette/VUPS group would occupy vis-à-vis the bookselling activities of E. Leclerc, it is to be feared that we would have no means of countering such a discount reduction" (690) and that "LDS and DIL, the main marketer-wholesalers serving small sales outlets, including supermarkets - would henceforth be controlled by the Hachette/VUPS group alone, thereby increasing our dependence on that supplier" (691).

(972) Whatever the strategy developed by the dominant entity, such as a policy of marginalising competitors or evicting them from the shelves of class 3 sales outlets or of reducing the discounts granted to such retailers, it is therefore difficult to see, on the basis of the findings of the Commission's investigation, how most of these dealers would be able to resist the competitive pressure to which they would be subjected (692).

(973) Taking these factors together, it must be concluded that neither competitors - both existing and potential - nor customers would have any real countervailing power.

C.2.d. CONCLUSION

(974) The notified transaction would create a dominant position held by the merged entity as a result of which effective competition would be significantly impeded in the market for the sale of books by wholesalers to class 3 outlets, through a combination of its market share, which is very large and out of proportion to that of its nearest competitors, and the capacity resulting from its close vertical integration in the sale of books to wholesalers, as a consequence of which it would be able to act to a significant extent independently of its competitors and customers. This conclusion is borne out by Roussillon-Diffusion, an independent wholesaler more than half of whose turnover is achieved through the sale of the merged entity's works: "as a small wholesaler, it seems to me that the Hachette/VUP merger might be harmful to the wholesaler, as there would be no more competition in supply, whence the possibility of the formation of a monopoly - discounts, negotiations, carriage, grace periods, payments. might be called into question, and automatically the wholesaler could no longer serve the retailer in the same way - in the long run, would [independent] wholesalers and therefore small retailers still have any reason for existing?" (693).

C.3. MARKETS FOR THE SALE OF BOOKS TO FINAL CONSUMERS

C.3.a. MARKETS FOR THE SALE OF LARGER REFERENCE WORKS BY SALES AGENTS

(1) SPAIN

(975) In the market for the sale of larger reference works to final consumers in Spain, Hachette Livre operates through Salvat. VUP operates through Spes.

(976) In 2001, according to the information supplied by the notifying party, the merged entity's share of the Spanish market for the sale of larger reference works in paper form to the final consumer was [20-30]% ([0-10]% for Hachette Livre and [10-20]% for VUP) (694), behind Planeta, which had [30-40]%, and in front of CIL, which had [0-10]%.

(977) The presence of a substantial competitor in the Spanish market allows it to be concluded that no dominant position would be created or strengthened in the Spanish market for the sale of larger reference works in paper form to the final consumer.

(2) BELGIUM AND LUXEMBOURG

(978) In 2001, according to the information supplied by the notifying party, the merged entity's share of the Belgian market for the sale of larger reference works in paper form to the final consumer was [20-30]% by value (695), behind Encyclopaedia Universalis, which had 40%.

(979) In 2001, according to the information supplied by the notifying party, the merged entity's share of the Luxembourg market for the sale of larger reference works in paper form to the final consumer was less than [0-10]% by value (696); Hachette Livre had no presence in the market.

(980) The presence of a substantial competitor in the Belgian market and the smallness of the merged entity's share of the Luxembourg market allow it to be concluded that no dominant position would be created or strengthened in the Belgian and Luxembourg markets for the sale of larger reference works in paper form to the final consumer.

(3) FRANCE

(981) In 2001, according to the information supplied by the notifying party, the merged entity's share of the French market for the sale of larger reference works in paper form to the final consumer was [30-40]% by value and [30-40]% by volume (697), in front of Encyclopaedia Universalis, which had [0-10]% by value and [0-10]% by volume.

(982) VUP has been withdrawing from this market for several years. It sold its sales subsidiary, SGED, in 2002, despite the fact that it holds such reputable brands as Le Grand Robert and Le Grand Dictionnaire Universel de Pierre Larousse.

(983) In addition, the French market for the sale of larger reference works in paper form to the final consumer has fallen considerably especially as a result of the progress of less expensive and more voluminous reference works in multimedia form.

(984) VUP's withdrawal and the steady shrinkage of the market allow it to be concluded that no dominant position would be created or strengthened in the French market for the sale of larger reference works in paper form to the final consumer.

C.3.b. MARKETS FOR THE SALE OF BOOKS BY RETAILERS TO FINAL CONSUMERS

(985) According to the notifying party, the merged entity would have a presence in the sale of books by retailers to final consumers in France through the bookshop chains Le Furet du Nord, Virgin/Extrapole and Relay, which have a combined share of [0-10]% of the French market. The notifying party estimates the merged entity's share of the Belgian market at between [0-10]% and [0-10]%. VUP is almost absent from this market.

(986) Hachette Livre has no significant shares of local or regional markets, with the exception of airports and hospitals.

(987) Given the absence of addition of market shares, it may be concluded that no dominant position would be created or strengthened in the French markets for the sale of books by retailers to final consumers.

SECTION VII: REMEDIES

(988) On 2 December 2003 the notifying party offered a number of commitments with a view to overcoming the competition concerns the Commission had identified in its statement of objections. On 23 December 2003 the notifying party put forward a number of amendments to do with the way in which these commitments were to be implemented. The wording of the commitments is set out in Annex II to this Decision, of which it forms an integral part.

I. Summary of commitments

(989) The notifying party undertakes to sell all the assets of VUP with the exception of the following:

- the Larousse publishing house and all of its business and its publisher's lists;

- the Anaya group and all of its business and its publisher's lists;

- the Dalloz publishing house and all of its business and its publisher's lists;

- the Dunod publishing house and all of its business and its publisher's lists;

- the academic lists made up of the publishers' lists of Armand Colin, Sedes and Nathan Université and the academic journals;

- the Ivry distribution centre.

(990) The notifying party undertakes to conclude irrevocable contracts of sale within [...]* months of the date of reception of this Decision. The notifying party states that if it has not itself satisfied the commitments within [...]* months, the trustee will have an irrevocable mandate to sell the assets on behalf and for the account of the notifying party. The contracts of sale must then be concluded within [...]* months.

(991) The notifying party undertakes to use its best efforts to sell all the assets to a single buyer, but in order to secure the best possible return on the sale reserves the right to sell to two or more buyers.

(992) Pending completion of the sale, a trustee appointed by the notifying party and approved by the Commission will ensure that the assets to be sold are held and managed in a structure separate and independent from the Lagardère group, under the responsibility of an independent hold-separate manager, and that their viability and competitive capacity are preserved.

II. Assessment of the commitments

A. CAPACITY TO OVERCOME THE COMPETITION CONCERNS

A.1. ELIMINATION OF HORIZONTAL OVERLAPS

(993) The commitments offered by the notifying party would eliminate almost all the horizontal overlaps between the parties. businesses in all of the relevant French-language markets in which the transaction might create or strengthen a dominant position.

(994) With the exception of the Larousse publishing house, the assets which are not to be sold, and are to be retained by the notifying party, are used in markets in respect of which the Commission raised no objection in the statement of objections it addressed to the notifying party. Anaya.s main business is the publication and sale of school textbooks and general literature titles in Spain and Latin America; it does no business in the French-language book trade. Dalloz, Dunod and the academic lists operate in the market for academic and professional books, in respect of which the Commission raised no objection in the statement of objections it addressed to the notifying party. The Ivry distribution centre has in the past been used for the distribution of school textbooks and works of reference, but it is now being used for the distribution of mobile telephones, decoders and video games for the account of Vivendi Universal among others. It is not currently in use for the distribution of books.

(995) Larousse operates mainly in the markets for the sale of works of reference by publishers to dealers. The assessment that the Commission has set out here shows that in all the markets for the sale of dictionaries the transaction in the form notified would strengthen VUP's existing dominant position by adding Hachette Livre's market share, and would create a near-monopoly with a market share of over [90-100]% in each of these markets. But even without Larousse, VUP's shares of these markets would still be greater than those of Hachette Livre, so that the sale of VUP's assets other than Larousse would mean that the merged entity would have a market share lower than that of VUP in the past. The commitment offered would therefore allow the emergence of a competitor whose market shares would be greater than those of Hachette Livre before the transaction.

A.2. VERTICAL AND CONGLOMERATE ASPECTS

A.2.a. IF ALL THE ASSETS ARE SOLD TO A SINGLE BUYER

(996) The assets that the notifying party undertakes to sell account for about [60-70]% of VUP's worldwide turnover, and about [70-80]% of its turnover in French-speaking markets. If the sale is to a single buyer, the commitments offered would eliminate the bulk of the vertical and conglomerate effects assessed in this Decision which result from the overall size of the merged entity in the French-language book trade and which contribute to the creation or strengthening of dominant positions in the relevant markets.

(997) The sale of VUP's marketing and distribution assets would make it possible to eliminate the horizontal competition concerns identified by the Commission in the markets in marketing and distribution services, and also to eliminate some vertical effects of the notified transaction due to the size of the merged entity in marketing and distribution.

(998) If the assets are sold to a single buyer the unit sold would own all of VUP's marketing and distribution assets, and in particular the Malesherbes distribution centre, with the exception of the assets dedicated to the publishing houses retained by the notifying party. The unit sold would consequently be in a position to market and distribute for its own publishing houses, thus ensuring their viability.

(999) The sale of the general literature publishing houses would eliminate the horizontal competition concerns identified in the markets for the acquisition of primary rights and for the sale of books of general literature; if the sale were to be to a single buyer, and the publishing houses were to keep their publishers' lists, and hence their authors and the key staff in contact with their authors, it would also ensure that the pocket-format publishing houses sold were viable, by providing them with publishers' lists and thus with a supply of secondary rights.

(1000) Lastly, the sale of a large part of VUP's publishing houses would ensure the viability of the distribution facilities sold, which would suffer as a result of the loss of the distribution of the publishing houses retained by the notifying party.

(1001) If the sale is to a single buyer, then, the commitments offered would eliminate the bulk of the vertical and conglomerate effects assessed in this Decision which result mainly from the overall size of the merged entity in the French-language book trade, and especially in the marketing and distribution of books, and which contribute to the creation or strengthening of dominant positions in the relevant markets.

A.2.b. IF SOME ASSETS ARE SOLD TO DIFFERENT BUYERS

(1002) In the light of what has been said, if the notifying party were to propose to sell these assets to two or more buyers, in order to overcome the vertical and conglomerate problems caused by the notified transaction the following requirements would have to be met in full:

- the capacity of each of the units sold to secure a supply of copyrights, and to ensure its marketing and distribution, so as to constitute a viable business able to exert real competitive pressure on the merged entity;

- the capacity of the units sold, taken together, to exert sufficient competitive restraint on the merged entity in each of the relevant markets, notably in terms of the relative size and the marketing and distribution capacities of each of the players in those markets;

- the continued viability of the pocket-format publishing houses as a competitive force in the markets in the sale of general literature in pocket format;

- the continued viability of VUP's distribution facility in Malesherbes as a competitive force in the markets for the marketing and distribution of books.

(1003) If the notifying party were to propose a sale to two or more buyers, therefore, the Commission could approve the buyers only if the tests just set out were fully satisfied.

B. FINAL CONSIDERATIONS

(1004) Before 2 December 2003, when it offered the commitments it subsequently amended on 23 December 2003, the notifying party had entered into talks with Commission staff on the basis of a smaller sell-off, which would not for example have included any distribution facility currently in use in the market. An initial assessment of commitments contemplated at that time, following a market survey, showed that there were some doubts as to the viability of the plan and its capacity to reestablish the conditions for effective competition in the relevant markets on a lasting basis. The notifying party was informed of these doubts; it did not formally propose commitments along the lines previously discussed, but instead, on 2 December 2003, offered the new package of commitments just described, which clearly overcomes the competition problems identified and thus renders the transaction compatible with the Common market.

(1005) By entering into commitments to be met within a precise time of the authorising Decision the notifying party is modifying the notified concentration in line with the second subparagraph of Article 8 (2) of the Regulation (698). In a Decision under Article 8 (2) or (3) of the Regulation, therefore, the Commission must pronounce on the transaction as thereby modified. This interpretation is in line with the principle that it is the notifying party that freely determines the transaction and any modifications to it, always subject to the obligation to preserve effective competition.

(1006) In order to meet the objections raised by the Commission in the statement of objections, and the doubts expressed regarding commitments of the kind initially suggested by the notifying party, the notifying party has not proposed separate divestitures in the markets referred to in the statement of objections but has instead undertaken to sell the VUP group as a whole, with the exception of certain specified assets.

(1007) As already explained, the Commission must pronounce on the transaction as thereby modified. The Commission finds that the commitment to sell the bulk of VUP's French-language assets as a coherent whole would preserve the integration of the publishing houses and other structures in the VUP group, including the marketing and distribution structures. This commitment therefore clearly meets the need to reestablish a player with sufficient vertical integration over the whole French-language book chain to be able to exert genuine competitive restraint on the merged entity, especially as regards the acquisition of copyrights and the marketing and distribution of the books that that player publishes.

C. CONCLUSION

(1008) The Commission accordingly considers that the commitments entered into by the notifying party, once they have been put into practice, are enough to render the notified concentration compatible with the Common market.

III. Conditions and obligations

(1009) The second subparagraph of Article 8 (2) of the Regulation empowers the Commission to attach conditions and obligations to its decision which are intended to ensure that the undertakings concerned comply with the commitments they have entered into vis-à-vis the Commission with a view to rendering the concentration compatible with the Common market. The requirement for achievement of each measure that gives rise to a structural change of the market, for example that a business is to be divested, is a condition; the implementing steps which are necessary to achieve this result are obligations on the parties (699).

(1010) The Decision to declare the transaction compatible with the Common market is accordingly subject to the condition that the notifying party comply in full with the commitments to sell assets which are set out in paragraphs 1 to 3 and 10 of Annex II. Full compliance with the other commitments set out in Annex II is an obligation imposed on the notifying party.

SECTION VIII: OVERALL CONCLUSION

(1011) For the reasons set out above the transaction contemplated will not create or strengthen a dominant position as a result of which effective competition would be significantly impeded in the Common market or in a substantial part of it, provided Lagardère complies with the commitments it has entered into. The transaction must therefore be declared compatible with the Common market and with the EEA Agreement provided the commitments set out in Annex II are complied with in full, in accordance with Article 2 (2) and Article 8 (2) of the Regulation and with Article 57 of the EEA Agreement,

HAS ADOPTED THIS DECISION:

Article 1

The notified transaction, as amended by the package of commitments of 23 December 2003, by which Lagardère acquires sole control of the assets of Vivendi Universal Publishing, now called Editis, is declared compatible with the Common market and with the functioning of the European Economic Area.

Article 2

Article 1 shall apply on condition that Lagardère complies in full with the commitments referred to in paragraphs 1 to 3 and 10 of Annex II.

Article 3

This Decision is subject to an obligation on Lagardère to comply in full with the other commitments set out in Annex II.

Article 4

This Decision is addressed to Lagardère SCA 4, rue de Presbourg F - 75116 Paris.

ANNEX I

METHODOLOGY: CROSS-TABULATIONS

* In the notification, the Lagardère group did not provide any information on the market shares of the players by category of book and by sales channel, taking the view that there was no need to distinguish markets for the marketing and sale of books by sales channel. The Commission considered that the information was necessary for its inquiries; it had to gather the information from market players, therefore, in order to evaluate the relative sizes of the various sales channels by category of book, in standardised cross-tabulations, and to be able calculate market shares for a given category of books in a given channel.

Compilation of the cross-tabulations

* The first step was to draw up cross-tabulations that would show the information needed for the investigation. The point of departure was to ask for figures for sales expressed in list prices exclusive of tax (French PPHT), in line with ordinary practice in the industry. These prices form the basis of the discounts given to dealers and the fees charged by marketers and distributors. Using this price as the basis of calculation avoids conflicts in the figures that could result from differences in the levels of discounts between publishers and between categories of book. On the basis of the list price exclusive of tax, it was possible to assemble information relevant to an assessment of the competitive position of the various players in the market for marketing and distribution (by showing the size of the flows through marketing and distribution mechanisms) and in the markets for the sale of books (by showing sales by channel and by category of book).

* As a common frame of reference, the categories of book making up the relevant markets were identified using the categories employed by the CLIL, the French Book Industry Liaison Committee, which are commonly referred to in book orders in the trade. Where a party's information system did not use those categories, the Commission contacted that party in order to standardise the categories employed. The sales channels identified were class 1 bookshops, class 2 bookshops, hypermarkets, supermarkets and wholesalers. Direct sales and sales through book clubs were expressly excluded. The investigation has shown that the figures gathered include sales through salesrooms. During the first stage of the inquiry the Commission considered the possibility that class 1 and class 2 might form separate markets, and drew up a standardised list of a little over 800 class 1 sales outlets which were common to the main marketers; class 2 was obtained by subtracting the other channels. The cross-tabulations drawn up consequently use standardised categories corresponding to the categories usually referred to in the trade.

* It will be useful to explain the calculations made to determine the shares of Hachette Livre and VUP in the wholesaler channel. The wholesalers LDS and La DIL are structures integrated into Hachette Livre and VUP respectively, but Hachette Livre and VUP did not initially provide specific figures for LDS and La DIL; in several replies Hachette stated that such figures could not be obtained. After repeated requests the Commission did obtain further information regarding the customer profile and sales of LDS and La DIL by category of book. It became clear that a proportion of the clientele of LDS and La DIL did not match the profile of the class 3 customers usually served by wholesalers. To avoid penalising the Lagardère group in the assessment of its position and VUP's position in the wholesaler channel, the Commission deducted part of the sales by LDS and La DIL from the wholesaler channel, and allocated it to class 2 bookshops (700). This allowed sales via wholesalers to be calculated on the basis of the totals for wholesalers (including LDS and La DIL) and supermarkets, in line with the definitions of the relevant markets.

Assembly of data

* The Commission observed that the distributors of French-language books had very precise and powerful information systems that allowed sales to be monitored book by book at each level of dealers. It therefore asked the main marketers and distributors operating in the market in France, Belgium and Luxembourg to provide the figures necessary to complete the cross-tabulations. More specifically, in order to distinguish between the marketing, distribution and sale of books, the following information was requested: total sales for in-house and outside publishers, sales for in-house publishers only, sales for outside publishers marketed but not distributed, and sales for outside publishers distributed only.

* Answers were obtained from the main distribution groups. These data were supplemented by similar requests to the main players in each category of book, and through the other questionnaires sent by the Commission in order to arrive as near as possible to a total for the market. Data were obtained for sales by the following publishers and those they market or distribute: Hachette Livre, VUP, Gallimard, Seuil, Flammarion, Diff-Edit, MDS, Sofedis, Tech & Doc, Etai, Vilo, Casteilla, Les Belles Lettres, Lavoisier, Belin, Albin Michel, Éditions Atlas, Encyclopédia Universalis, Van In, De Boeck, Labor, Éditions Lefebvre, Bayard, Elsevier, Kluwer, Magnard, Pearson, Lamy, Juriste Classeur, De Vechi, Reader's Digest, Michelin, Gründ, and École des Loisirs.

* The Commission checked most of the figures for consistency on the basis of the declarations made by the firms to the SNE, the Union of French Booksellers; they appeared to be correct. The figures assembled account for almost 80% of the total market estimated by the notifying party.

Evaluation of market shares

* In order to calculate market shares, the main difficulty now was to clarify the size of the various markets, by category of book and by sales channel. The Commission had figures for sales by the main players, but there was still some doubt as to the size of the total market. For example, the notifying party estimated the French market at euro 3.2 billion, while some other publishers estimated it at euro 2.5 billion.

The Lagardère group's estimates

* After repeated requests, the Lagardère group finally supplied estimates of total sales by category and by sales channel, though it indicated that it had reservations regarding the results obtained. It had first determined total sales by category of book, extrapolating from the SNE totals for France, which give publishers' turnovers rather than list prices. Average discounts were then used to convert SNE turnovers into list prices. A breakdown by sales channel was carried out on the basis of percentages using the Sofrès data base, which gives sales by channel. The wholesalers category does not appear in these totals, because the Sofrès data base covers final sales, and does not indicate whether or not they passed through the hands of a wholesaler.

Estimate of sizes of markets by channel

* The Commission has used a method slightly different from that of the Lagardère group: it first determined the size of the markets by channel (class 1 bookshops, class 2 bookshops, hypermarkets and wholesalers), with a view to determining the market shares for marketing and distribution, and then worked out cross-tabulations for channel and category of book. Since it had numerous answers from other publishers and from public sources, such as the SNE statistics for the 13 main distributors, which diverged from the Lagardère group's estimates, it first determined the size of the total market in France. For Belgium and Luxembourg, for want of more precise information, it used the parties. market totals. In general, the Commission sought to adopt a conservative approach that would not penalise the notifying party. The size of the market comprising France, Belgium and Luxembourg was ultimately calculated to be very slightly less than the parties' own estimates, that is to say euro 3 244 million in terms of list prices exclusive of tax, as compared with the euro 3 404 million estimated by the Lagardère group. For the size of the markets in marketing and distribution for outside publishers, the Commission used the answers to its questionnaires and the estimates of the parties. There too the estimates of the total size of the market by the Commission and by the Lagardère group are close (701).

* On the basis of the samples collected, answers to questionnaires and the analysis of sales by wholesalers, including LDS and La DIL, the Commission determined the size of the markets for marketing and distribution in the various channels. The breakdown by channel was worked out in percentage terms on the basis of the actual standardised-basis figures that the Commission possessed for the 14 main marketers and distributors in France, Belgium and Luxembourg and a certain number of publishers and wholesalers who proved to represent a very high percentage of the total market. All of these factors allowed a size to be determined for the markets for the marketing and distribution, by channel, of sales for outsiders and total sales, in list prices exclusive of tax. These figures also provided an estimate of the size of the different channels in France, Belgium and Luxembourg.

Estimates of sizes of markets by channel and by category of book

* Alongside these estimates of the overall size of the market by channel, the Commission sought to estimate the total sales and breakdown by channel for each category of book. The point of departure was the actual figures obtained from the main publishers for each category of book. Some additions were nevertheless made to the figures supplied by Lagardère, which were incomplete especially in respect of pocket-format books, where percentages supplied for one channel had been arbitrarily applied to all channels, and in respect of travel, where Lagardère's estimate of market share of travel and food books had been applied uniformly across the different channels. Apart from these minor additions the figures used by the Commission came direct from the players in the market, who had completed the Commission's standardised tables.

* The Commission then compared these actual figures with the estimates of the Lagardère group, and with the SNE figures that had formed the basis of Lagardère's calculations. For school textbooks, educational supporting materials and reference books, the Commission's figures were practically identical to Lagardère's, except of course that the Commission's were more precise with regard to the breakdown by channels, and with regard to wholesalers in particular. At a practical level it emerged that Lagardère had not excluded the road maps included in the SNE figures, which meant that it had overestimated the size of the market in guides and manuals. Apart from this correction, it became clear that here too the Commission's figures were more precise because they were based on actual figures. For children's books, strip cartoon albums and general literature, the Commission's estimates arrived at a total market bigger than that estimated by Lagardère. For the academic market, on the other hand, which Lagardère had estimated in a different fashion owing to inconsistency in the SNE figures, the Commission arrived at a far smaller size for the market, on the basis of actual figures for the biggest players present.

* In general, the estimates of the size of the different channels made use of percentages drawn from actual figures obtained by the Commission, while seeking to ensure that they were consistent with Lagardère's own estimates. But a final adjustment was made to ensure that the figures thus arrived at were in line with the estimates of total market by channel. The following table shows the detail of the sizes of markets calculated by the Commission for the market comprising France, Belgium and Luxembourg in 2001, in list prices exclusive of tax:

Estimates of the French-language publishing market (in France, Belgium and Luxembourg) by value (list prices exclusive of tax, euro .000) in 2001

<emplacement tableau>

* list of sales outlets drawn up by Commission

Calculation of market shares

* The calculation of market shares was carried out on the basis of actual figures which the Commission obtained from the main players. These calculations were carried out for the relevant markets, that is to say for bookshops (the sum of the class 1 and class 2 columns in the table above), hypermarkets and wholesalers (the sum of the wholesalers and supermarkets columns in the table above).

* In general, the market shares arrived at proved very close to those submitted by the Lagardère group.

ANNEXE II : COMMITMENTS OFFERED BY LAGARDÈRE

In connection with the acquisition by the Lagardère group (hereinafter "the notifying party" or "Lagardère") of Editis SA (formerly Investima 10, hereinafter Editis) and in accordance with Article 8 (2) of Council Regulation No 4064-89 on the control of concentrations between undertakings, the notifying party is prepared to enter into the following commitments. These commitments are offered subject to the adoption by the Commission of a Decision declaring the transaction compatible with the Common market in accordance with Article 8 (2) of Regulation No 4064-89 already referred to, and will take effect immediately upon reception of such a Decision.

TITLE I. SCOPE OF THE COMMITMENTS

1. The notifying party undertakes to sell all the assets of Editis with the exception of the following:

- the Larousse publishing house and all of its business and its publisher's lists,

- the Anaya group and all of its business and its publisher's lists,

- the Dalloz publishing house and all of its business and its publisher's lists,

- the Dunod publishing house and all of its business and its publisher's lists,

- the academic lists made up of the publishers' lists of Nathan Université, Armand Colin and Sedes and the academic journals, and

- the Ivry distribution centre (hereinafter collectively referred to as the "Assets Retained"), on the terms described below under the common mechanisms for meeting the commitments. The assets covered by the commitment to sell are hereinafter referred to as the "Assets Sold".

2. The Assets retained are defined in detail in Annex 1.

TITLE II. COMMON MECHANISMS FOR FULFILLING THE COMMITMENTS

(1) Time-limit

3. The notifying party undertakes to conclude irrevocable contracts of sale of the Assets Sold within [...]* of the date of reception of the Decision authorising the concentration in accordance with Article 8 (2) of Regulation No 4064-89 (hereinafter the "First Time-limit"). The transfer of the Assets Sold will take place within [...]* of the conclusion of the contract of sale (hereinafter the "Second Time-limit").

4. If exceptional circumstances prevent the conclusion of a contract of sale or the transfer of an asset within the First or the Second time-limit as the case may be, that time-limit may be extended by the Commission, in response to a properly substantiated application on the part of the notifying party.

5. Any application for the extension of the first Time-limit must be submitted to the Commission no later than [...]* the First Time-limit. Any application for the extension of the Second Time-limit must be submitted no later than [...]* the Second Time-limit. The Commission will reply to the application for extension within no more than [...]* from the date on which it has received all the information it needs in order to be able to reach a decision on the application. If the Commission does not reply within that time, the time-limit to which the application for extension refers will be suspended until the Commission replies to the application.

(2) Preservation of the conditions of competition and of the value of the Assets Sold until the transfer takes place

6. The notifying party undertakes to preserve the full economic and competitive value of the Assets Sold until the date of the transfer of the Assets Sold, in accordance with sound commercial practice and to the extent permitted by the resources at its disposal in accordance with these commitments. The notifying party undertakes:

(a) until the date of transfer of the Assets Sold, to do nothing on its own authority that might significantly affect the economic value, the management or the competitiveness of the Assets Sold;

(b) until the date of transfer of the Assets Sold, to do nothing on its own authority that might significantly change the nature or scope of the business of the Assets Sold or industrial or commercial strategy or investment policy relating to the Assets Sold;

(c) not to contract with the outside publishers currently distributed or marketed by the Assets Sold, not to contract with authors whose sales of at least one book published by one of the publishing houses of the Assets Sold in 2002 exceeded 10 000 copies (a list of these authors is attached as Annex 2), and not actively to solicit authors whose sales of at least one book published by one of the publishing houses of the Assets Sold in 2000, 2001 or 2002 exceeded 14 000 copies (in the case of the commitment not to solicit actively, with the exception of authors contracted to an Hachette Livre publishing house); the commitment in this paragraph 6 (c) will apply until the transfer of the Assets Sold;

(d) to take the steps necessary to prevent the divulgation of confidential information concerning the Assets Sold within Lagardère or to outsiders, with the exception of the information necessary for the sale of the Assets on the best possible terms in accordance with these commitments;

(e) to provide the trustee referred to in paragraph 15 with all the necessary resources and all the information the trustee may consider relevant so as to enable the trustee to familiarise itself with the day-to-day management of the Assets Sold. Subject to the preceding, Lagardère may manage the Assets Retained as it chooses.

(3) Obligation to hold the Assets Sold separate and independent

7. Until the date of transfer, the notifying party undertakes that the Assets Sold will be held and managed in a distinct structure, separately from and independently of the Assets Retained and Lagardère's other businesses.

8. Until the date of transfer, the notifying party will give the trustee all the assistance it needs to satisfy itself that the Assets Sold are being held and managed separately in a distinct structure.

9. The notifying party will appoint a hold-separate manager within the time-limits and in accordance with the mechanisms for the appointment of the trustee laid down in paragraph 16 to 19. The hold-separate manager will be responsible for the management of the Assets Sold, under the supervision of the trustee. The hold-separate manager must manage the Assets Sold independently and in the ordinary course of business with a view to ensuring the preservation of their economic viability, their saleability, their competitiveness and their independence of the Assets Retained and Lagardère's other businesses. If an officer of a subsidiary of Editis within the scope of the commitment to sell terminates his or her duties, the hold-separate manager will have authority to appoint a successor, under the supervision of the trustee.

(4) Quality of the buyer

10. In order to preserve effective competition on the relevant markets the notifying party undertakes to sell the Assets Sold to one or more buyers independent of the notifying party and satisfying the following tests:

(a) Lagardère may have no significant interest, direct or indirect, in the buyer.

(b) The buyer or buyers must be viable operators capable of maintaining or developing effective competition, and having the economic incentive to do so; this wording is not to be interpreted as excluding any category of industrial or financial buyers in advance.

(c) The acquisition of one or more of the Assets Sold by a potential buyer must not be such as to create fresh competition difficulties or to threaten to delay the fulfilment of these commitments. The notifying party must be able to show the Commission that the buyer satisfies the tests laid down in these commitments and that the Asset or Assets Sold are being transferred in accordance with these commitments.

(d) The buyer or buyers must have obtained or be reasonably supposed to be able to obtain all the approvals necessary for the acquisition and operation of the Assets Sold.

11. The notifying party will use its best efforts to sell all the Assets Sold to a single buyer, while nevertheless maintaining the objective of securing the best possible return on the sale of all the Assets Sold.

12. To preserve the structural effect of these commitments, the notifying party undertakes:

(a) not to acquire control of all or part of any of the Assets Sold within [...]* of the transfer of the Assets Sold, unless the Commission has determined that the structures of the market have developed in such a way that in order to render the concentration compatible with the Common market it is no longer necessary that there should be no influence over the Assets Sold;

(b) not to recruit, directly or indirectly, any of the members of the executive board of Editis or the main senior publishing staff of the Assets Sold who maintain essential relationships with authors, a list of whom is attached as Annex 3 (hereinafter .Key Staff") before the transfer of the Assets Sold and for a period of [...]* after that transfer, except that no provision in these commitments is to prevent the notifying party or any of its subsidiaries from (i) asking the hold-separate manager, in order to complete the sale of the Assets Sold as quickly as possible and on the best possible terms, to introduce incentive payments for all or some of the Key Staff, payable only if the Key Staff remain with Editis at least until the date of the transfer, and (ii) recruiting Key Staff who are not kept on by the buyer or buyers.

13. The notifying party will as quickly as possible submit to the Commission, which undertakes to comply absolutely with its duty of confidentiality:

(a) the draft or drafts of information documents relating to the sale of each category of Assets Sold, to be given to potential buyers;

(b) the list of the potential buyers that the notifying party intends to contact (though this list need not necessarily be exhaustive); and

(c) the draft of the contract of sale.

14. The choice of the buyer or buyers will be subject to the Commission's approval. The application for the approval of the buyer or buyers will include the information necessary to enable the Commission to establish that the proposed buyer or buyers satisfy the tests set out in paragraph 10. The Commission will inform the notifying party of its approval or rejection of the proposed buyer or buyers within [...]* from the date on which it has received all the information referred to. If the Commission does not reply within that time, the First Time-limit or the Second Time-limit, as the case may be, will be suspended until the Commission has replied to the application for approval.

(5) Trustee

15. The notifying party will appoint a trustee to perform the duties set out below. The trustee will be independent of Lagardère and of Editis, will possess the qualifications needed to perform its mandate, being for example a consultant bank, consultant or auditor, and will not be exposed to any conflict of interests. The trustee will be remunerated by Lagardère in a manner that does not compromise the proper performance of the trustee's duties or the trustee's independence.

16. No later than [...]* from the date of reception of the Decision authorising the concentration in accordance with Article 8 (2) of Regulation No 4064-89, the notifying party will propose the name or names of one or more trustees to the Commission, and will submit to it a draft mandate, complying with these commitments, setting out in detail the scope of the mandate, the responsibilities of the trustee and the manner of the trustee's remuneration.

17. The Commission will state its position on the identity of the trustee and the draft mandate within [...]* of the date on which it has received all the information it needs in order to do so. If the Commission so requests, the notifying party will amend the draft mandate.

18. If the Commission does not reply to the notifying party's proposals within [...]* of the date on which it has received all the information it needs in order to do so, the First Time-limit will be suspended until the Commission replies to the proposals.

19. The notifying party will appoint the trustee within [...]* of the Commission's approval. A copy of the mandate given to the trustee will be supplied to the Commission. Once the mandate has been signed the notifying party may not amend them without the Commission's agreement. At the trustee's request the Commission may require that the mandate be amended if it is shown that it does not allow the trustee properly to perform the duties entrusted.

20. The object of the trustee's actions is to ensure that these commitments are met. On its own initiative or upon application by the trustee or the notifying party, the Commission may address any instruction to the trustee which is designed to ensure that these commitments are met.

21. The trustee's duties will consist of the following:

(a) to ensure that the Assets Sold are held and managed in a distinct structure, separately from and independently of the Assets Retained and Lagardère's other businesses, until the date of the transfer of the Assets Sold;

(b) to ensure that the hold-separate manager maintains the viability and saleability of the Assets Sold and the management and operation of the Assets Sold in the ordinary course of business and in accordance with previous practice until the date of the transfer of the Assets Sold;

(c) to ensure that effective steps are taken to prevent any information that is sensitive with regard to competition form being communicated to the notifying party, with the exception of information necessary for the sale of the Assets Sold on the best possible terms in accordance with these commitments;

(d) to ensure that the restructuring measures are taken in accordance with these commitments, to ensure that it is informed of discussions in progress between Lagardère and Editis regarding demarcation, and if necessary to attend such discussions;

(e) if the notifying party fails to fulfil the commitments within [...]*, to negotiate with interested outside parties and to conclude a contract for the sale of the Assets Sold with one or more buyers approved by the Commission in accordance with paragraph 25;

(f) generally to ensure that the economic and competitive value of the Assets Sold is preserved, and to take any steps conducive to that end;

(g) generally to ensure that the notifying party satisfactorily fulfils these commitments.

22. The trustee will regularly report to the Commission on the progress of the fulfilment of the commitments referred to below (including progress on demarcation) and more generally the performance of the trustee's duties. To this end the trustee will draw up and transmit to the Commission a confidential report [...]* at the end of each period or whenever requested. The report will in particular:

(a) confirm that the Assets Sold are being kept and managed in a distinct structure, separately from and independently of the Assets Retained and Lagardère's other businesses, in accordance with paragraph 21 (a);

(b) confirm that the Assets Sold are being managed in such a way as to preserve their full economic and competitive value in accordance with paragraph 21 (b);

(c) indicate the steps taken to fulfil the commitments, the reactions of outside parties contacted (potential buyers, outsiders holding a right of approval or first refusal, social partners and administrative authorities) and progress with the formal conclusion of contracts of sale; and

(d) identify any aspects of the mandate which the trustee has been unable to carry out and the reasons for that failure to carry out the mandate.

23. A non-confidential version of the report from the trustee to the Commission will be transmitted to the notifying party.

24. In the event of disagreement between Lagardère and Editis on the restructuring measures necessary for the fulfilment of these commitments, either party may inform the trustee accordingly by registered letter, sending a copy to the other party. The trustee, having heard both sides, will then make a recommendation, as quickly as possible, regarding the scope of the restructuring measures needed. The trustee will report to the Commission informing it of the recommendation. If the disagreement between Lagardère and Editis persists, either party may request the Commission, after hearing both sides, to determine the scope of the restructuring measures needed.

25. If the notifying party fails to fulfil the commitments within [...]*, the trustee will have the duty to conduct negotiations with interested outside parties on the trustee's own initiative with a view to selling in good faith the Assets Sold [...]* to one or more buyers approved by the Commission. The contracts of sale must in that event be concluded within [...]*, in accordance with paragraph 3, possibly extended in accordance with paragraphs 4 and 5.

26. In the event of a substantial failure on the part of the notifying party to fulfil its commitments, the Commission may amend the duties of the trustee so as to enable the trustee to ensure that the commitments are fulfilled.

27. The hold-separate manager will give the trustee all reasonable assistance and all information the trustee may consider necessary for the performance of the trustee's duties as described above. The hold-separate manager will place one or more offices at the disposal of the trustee on the premises of the Assets Sold.

28. The notifying party will hold regular meetings with the trustee, at intervals agreed between them, in order to provide all information, verbal or written, which is necessary for the performance of the trustee's duties. At the trustee's request, the notifying party will give the trustee access to the premises being sold.

29. Once the trustee's duties have been performed, the notifying party will ask the Commission to discharge the trustee from its duties. But the Commission may require the appointment of a new trustee if it should subsequently appear that the commitments have not been fulfilled in their entirety.

(6) Alteration of the legal form of Editis

30. After Editis.s new constitution has been approved by the Commission, the notifying party will convert Editis into the form of a simplified limited company (société par actions simplifiée, SAS). Thereafter the governing bodies of the company will be (i) a chief executive officer, who will perform the duties of the hold-separate manager, and (ii) a shareholders' committee, made up of three representatives of the trustee referred to in paragraph 15 and two representatives of Lagardère.

31. The simplified limited company will be organised as follows: [...]*.

32. During the period between the adoption by the Commission of a decision authorising the notified transaction and the conversion of Editis [...]*. During that time, Lagardère, in its capacity as shareholder in Editis, will be entitled to access to all information concerning the Assets Retained. In the case of the Assets Sold the trustee will ensure that the information referred to in paragraph 31 (c) is transmitted to Lagardère.

33. Lagardère will submit the draft of Editis's new constitution to the Commission for its approval within [...]* of the adoption by the Commission of a Decision authorising the notified transaction.

(7) Review clause

34. In response to a properly substantiated application on the part of the notifying party, and after consulting the trustee, the Commission may:

(i) grant an extension of the time-limits for the divestiture;

(ii) authorise the sale or transfer of the Assets Sold without certain assets, contracts, rights or obligations;

(iii) lift one or more of the conditions and obligations which are the subject of these commitments.

The notifying party must seek any extension of the time-limit at least [...]* before the expiry of that time-limit. The notifying party may seek an extension of [...]* time-limit only in exceptional circumstances.

(8) No waiver of rights of redress open to parties

35. These commitments do not imply any waiver of any rights of redress open to the parties against decisions taken by the Commission in proceedings relating to Case No COMP/M.2978, and in particular against any authorising Decision under Article 8 (2) of Regulation No 4064-89.

Paris, 23 December 2003

For Lagardère SCA

Dominique d'Hinnin and Norbert Giaoui

ANNEX 1: ASSETS NOT INCLUDED IN THE COMMITMENT TO SELL ("ASSETS RETAINED")*

The following are not included in the commitment to sell.

1. The Larousse publishing house and all of its business and its publisher's lists

1. The main business of the Larousse publishing house is the publication and sale of works of reference; its assets comprise the following:

(a) The following "Larousse" series:

- Encyclopaedias: [...]*

- Dictionaries: [...]*

- Educational supporting materials: [...]*

- Foreign-language dictionaries: [...]*

- Guides and manuals, leisure and the arts in general: [...]* and various dictionaries, albums and art books not published as part of a series.

- Children's books: [...]*

(b) The [...]* list, comprising: Guides and manuals

* Fine arts: [...]*

* Crafts: [...]*

* Decoration: [...]*

* Creative leisure: [...]* Children's

* [...]*

(c) The [...]* guides and manuals list, comprising: In the gardening section:

- [...]* In the astronomy section: - [...]*

In the nature section:

- [...]*

(d) [...]*

(e) the following subsidiaries and holdings abroad: [...]*

2. The following will be taken over by the notifying party, subject to the demarcation exercise needed to ensure that the Assets Sold have the resources necessary in order to fulfil these commitments:

(a) all the tangible and intangible assets linked to the business of the Larousse publishing house, and in particular the outright ownership of the intellectual property rights associated with the Larousse publishing house, subject to Section 8 of this Annex;

(b) all the staff necessary to the operation of the Larousse publishing house;

(c) all the existing contracts linked to the business of the Larousse publishing house.

2. The Anaya group and all of its business and its publisher's lists

3. The main business of the Anaya publishing house is the publication and sale of textbooks (from nursery school to university), general literature, and books published in instalments. Anaya also carries on a multimedia business and a distribution business. The assets of the Anaya publishing house comprise the following:

- Textbooks for the 17 regional curricula from nursery school to school leaving certificate, in the four official languages and two regional languages;

- The regional imprints [...]*

- The academic publishing houses [...]*

- The [...]* publishing house and the general literature series [...]*

- The publishing houses [...]*

- The [...]* publishing house, whose main business is the publication and sale of dictionaries and encyclopaedias.

- The [...]* publishing house, whose main business is the publication and sale of books for primary and secondary education, dictionaries, encyclopaedias and children's books.

These assets are operated by the following companies (foreign subsidiaries): [...]*

4. The following will be taken over by the notifying party, subject to the demarcation exercise needed to ensure that the Assets Sold have the resources necessary in order to fulfil these commitments:

(a) all the tangible and intangible assets linked to the business of the Anaya publishing house, and in particular the outright ownership of the intellectual property rights associated with the Anaya publishing house, subject to Section 8 of this Annex;

(b) all the staff necessary to the operation of the Anaya publishing assets;

(c) all the existing contracts linked to the business of the Anaya group.

3. The Dalloz publishing house and all of its business and its publisher's lists

5. The main business of the Dalloz publishing house is the publication and sale of books on law and economics.

Dalloz's assets comprise the following series and subsidiaries:

- Law codes: [...]*

- Academic: [...]*

- Economics and management: [...]*

- Professional: [...]*

- Law encyclopaedia: [...]*

- [...]*

- [...]*

- [...]*

- [...]*

- Other: [...]*

- [...]* and all of its business, in particular:

* Periodicals: [...]*

* Guides and manuals: [...]*

6. The following will be taken over by the notifying party, subject to the demarcation exercise needed to ensure that the Assets Sold have the resources necessary in order to fulfil these commitments:

(a) all the tangible and intangible assets linked to the business of the Dalloz publishing house, and in particular the outright ownership of the intellectual property rights associated with the Dalloz publishing house, subject to Section 8 of this Annex;

(b) all the staff necessary to the operation of the Dalloz publishing house;

(c) all the existing contracts linked to the business of the Dalloz publishing house.

4. The Dunod publishing house and all of its business and its publisher's lists

7. The main business of the Dunod publishing house is the publication and sale of books in science, technology, medicine and the social sciences. Dunod's assets comprise the following series:

- higher education department, social sciences sector: [...]*

- higher education department, fundamental and applied sciences sector: [...]*

- working environment department, social sciences and management sector: [...]*

- working environment department, science and technology sector: [...]*

- [...]*

- reference

- [...]* and all of its business (provision of services).

8. The following will be taken over by the notifying party, subject to the demarcation exercise needed to ensure that the Assets Sold have the resources necessary in order to fulfil these commitments:

(a) all the tangible and intangible assets linked to the business of the Dunod publishing house, and in particular the outright ownership of the intellectual property rights associated with the Dunod publishing house, subject to Section 8 of this Annex; (b) all the staff necessary to the operation of the Dunod publishing house;

(c) all the existing contracts linked to the business of the Dunod publishing

house.

5. Other academic lists

9. The academic lists of Librairie Fernand Nathan SA, comprising the following publisher's lists:

- [...]*

- [...]*

- [...]*

- The academic journals [...]*

- Titles not published as part of a series (see attached addendum).

10. The following will be taken over by the notifying party, subject to the demarcation exercise needed to ensure that the Assets Sold have the resources necessary in order to fulfil these commitments:

(a) all the tangible and intangible assets linked to the business of the other academic lists, and in particular the outright ownership of the intellectual property rights associated with the other academic lists, subject to Section 8 of this Annex;

(b) all the staff necessary to the operation of the other academic lists;

(c) all the existing contracts linked to the business of the other academic lists.

6. The Ivry distribution centre

11. The Ivry centre is a centre with a total surface area of about 24 000 m², including offices, on a site with a total surface area of about 30 000 m². It is located at 11/15, Rue Rigaud, Ivry (94200), and is distinct from the head office and main place of business of Interforum, located at 3, Allée de la Seine (Immeuble Paryseine), Ivry (94200), which is among the Assets Sold.

12. The following will be taken over by the notifying party, subject to the demarcation exercise needed to ensure that the Assets Sold have the resources necessary in order to fulfil these commitments, and in particular the demarcation of the assets and resources of Interforum housed in the Ivry logistics centre:

(a) all the tangible and intangible assets linked to the logistics business of the Ivry centre;

(b) all the staff necessary to the operation of the Ivry centre;

(c) all the existing contracts linked to the business of the Ivry centre.

7. Transitional agreements

13. Lagardère's takeover of the Larousse publishing house, the academic lists and the Ivry distribution centre will necessitate legal reorganisation and the conclusion of transitional agreements in order to allow the publisher's lists to be operated independently.

14. The marketing and distribution currently performed in France and abroad for the Assets Sold by Interforum and its subsidiaries, including La DIL, will continue to be performed by Interforum for a period of [...]* after the Commission's authorising Decision. During that period marketing to class 3 bookshops of books published by Larousse and by Bordas Pratique will be performed exclusively by La DIL.

15. [...]*

8. Principles for the transitional use of imprints between the Assets Sold and the

Assets Retained

16. In all cases of co-copyright between Assets Retained and Assets Sold [...]*

17. The following mechanisms will apply:

(a) Each of the publishing Assets Sold (hereinafter .the Company") will retain title to the intellectual property rights corresponding to the operation of its publisher's lists [...]* and will grant a non-exclusive right of use to the Asset Retained (hereinafter "the User") [...]* in order to clear its stocks, in accordance with the mechanisms described below.

(b) [...]*

(c) [...]*

(d) [...]*

(e) [...]*

(f) [...]*

ANNEX 2

AUTHORS REFERRED TO IN PARAGRAPH 6 (c) OF THE COMMITMENTS (702) [...]*

ANNEX 3

PERSONS REFERRED TO IN PARAGRAPH 12(b) OF THE COMMITMENTS [...]*

Notes :

1 OJ L 395, 30.12.1989, p. 1; corrected version OJ L 257, 21.9.1990, p. 13; last amended by Council Regulation (EC) No 1310-97 of 30 June 1997 (OJ L 180, 9.7.1997, p. 1).

2 OJ C ..., ... 200., p. ....

3 OJ C ..., ... 200., p. ....

4 Since the date of the notification, Investima 10 has become Editis SA.

5 Vivendi Universal also divested itself of its publishing assets in the United States (Houghton Mifflin), acquired by a third party.

6 Turnover calculated in accordance with Article 5 (1) of the Regulation and the Commission communication on the calculation of turnover (OJ C 66, 2.3.1999, p. 25).

* Parts of this document have been omitted to ensure that no confidential information is disclosed; these are contained in square brackets and marked with an asterisk.

7 In France, consumer prices are governed by Act No 81-766 of 10 August 1981 (the "Lang Act") on uniform pricing, which provides that the same book may not be sold at different prices (subject to a 5% margin of fluctuation). In Belgium, a bill to the same effect is before the federal parliament.

8 In France, the effect of the Galland Act of 1996 is that prices are virtually identical for all distributors. "Back margins" (marges arrières) are then based on specific services rendered by the distributor to the producer, such as gondola-end display or prominence in the distributor's advertising.

9 Members generally receive a catalogue at home, setting out the titles that have been selected "Clubs" clientele consists entirely of members who have signed a contract imposing obligations such as a commitment to buy a specified minimum number of books over a given period or to accept a selected book if they do not respond.

10 Source: French Ministry of Culture and Communication (key figures for publishing, 2001).

11 Source: ADEB.

12 Source: Supplement to Livres Hebdo, 22.3.2002.

13 Though growth rates were slightly lower for school textbooks.

14 The source of the turnover figures in this section is the supplement to Livres Hebdo, 22.3.2002.

15 [...]*.

16 LGF has [...]* in the supplementary contract to the joint holding agreement of 20 July 1971, signed on 14 June 1999 between Albin Michel and Hachette Livres, providing that [...]* (notification, annex 69).

17 Notification, annex 69.

18 Apart from publishers of school textbooks and academic and professional books, who commonly handle their own marketing and distribution, since their retailer clients are large and specialised booksellers, of which there are relatively few.

19 As will be seen, some of the largest of these smaller operators have developed their own pocket-format series publishing only their own titles. But the marketing of these series is far less extensive than that of the major publishing houses, and inadequate to guarantee a second life for a best-seller (e.g. Actes Sud, publisher of the 2002 Nobel Prize winner, Imre Kertecz, who until then was not widely sold in France, had to sell the paperback rights to the VUP group (the "10/18" series) in order to give his works a second life).

20 Unlike moral rights, the exploitation rights conferred on authors are limited in time. When they expire the work falls into the public domain and can then be used freely, provided that the author's moral rights are respected.

21 However, if the book is not reprinted after the first edition has gone out of print, the publisher may ask to recover the rights. In this case it will take back the rights to illustrations, but the translation rights will remain the property of the French-language publisher which acquired them.

22 Alongside individual and collective works, there are also "collaborative" works, which may be defined as works to which a number of natural persons have contributed (Article L. 113-2 (1) of the French Code de la propriété intellectuelle). Such persons enjoy the status of (co-)authors and hold copyright (economic and moral rights).

23 Notification, p. 34.

24 The notifying party estimates that commissions account for [75-85]% of general literature titles, [90-100]% of school textbooks and dictionaries, [85-95]% of educational supporting materials, [85-95]% of academic works and [45-55]% of children's books published by VUP.

25 These rights are covered by a contract not between an author and a publisher, but between two publishers, or between a publisher and a literary agent who basically represents the original publisher.

26 Answer to questions 8 and 20 in request for information 14.685.

27 Or they may be well-known personalities from outside the literary world, e.g. journalists, artists, sports personalities, politicians, etc.

28 An advance is a firm and final payment by the publisher to the author before a work is published. The amount is calculated on the basis of an estimate of expected sales, but is guaranteed by the publisher whatever level of sales is subsequently achieved. In practice, the advance is often paid in separate instalments - on signature of the contract between author and publisher, on submission of the manuscript, and on publication of the work.

29 Notification, p. 297.

30 As part of its investigation, the Commission asked various publishers to supply the contracts of authors who had joined or left their publishing houses in the last five years (answers to question 14 in request for information 14.685 and question 9 in the "Publishers Phase II" request for information sent on 17 July 2003). Analysis of these contracts shows that in over 70% of cases where the Commission obtained both the old and new publishing contracts the advance paid to the author by the new publisher (the proportional payment being equal or higher) was greater than that paid by the old publishing house. The Commission has obtained the old and new contracts in 70 cases. The publishing houses which benefited most from these transfers were Hachette Livre (37%), VUP (27%), Seuil (14%) and Gallimard (11%). While it cannot be ruled out that authors receive higher advances even without changing publishing house, such an increase would reflect the pressure exerted by competing publishing houses on the initial publisher, which is forced to increase authors' advances in order to hold on to them even if their most recent book has sold no better than the previous one.

31 Notification, p. 299.

32 Notification, p. 62.

33 Answers to question 6 in the "Authors" request for information sent on 30 July 2003 and question 7 in the "Literary Agents" request for information sent on 30 April 2003.

34 Answers to question 3 in the "Authors" request for information sent on 30 July 2003.

35 "The concept of a "commission" applies to a contract concluded when the text of the work to be published has not been completed. This covers three possible situations: (i) a project defined by the publisher and entrusted to an author, (ii) a project submitted by an author and accepted by the publisher or (iii) a work expected under the terms of a right of first refusal previously granted by the author to the publisher", notification, p. 32.

36 Under a right of first refusal, the author of a work grants the publisher preference for the acquisition of a number of works to be written in future. This right binds the author to the publisher.

37 Notification, p. 32.

38 Answers to questions 8 and 20 in request for information 14.685 and question 15 in the "Publishers Phase II" request for information sent on 17 July 2003.

39 Answer to question 8 in request for information 14.685.

40 Notification, p. 106.

41 Intellectual property protection lasts for 70 years after the author's death in France and Belgium (50 years in Canada).

42 Answers to question 2 of the "Publishers Phase II" request for information sent on 17 July 2003.

43 However, English-language authors or their agents retain rights of approval over the choice of the pocket-format publisher acquiring these rights; the best-known authors even assign pocket-format publishing rights by separate contract (answers to questions 29 and 30 in request for information 14.685 and questions 23 and 24 in the "Publishers Phase II" request for information sent on 17 July 2003).

44 Answers to question 6 in the "Literary Agents" request for information sent on 30 April 2003. Seven of the 23 leading literary agents cited by the notifying party replied to the Commission's questionnaire, of which only one stated that it did not operate according to the auction system.

45 Answers to question 6 in the "Authors" request for information sent on 30 July 2003 and question 7 in the "Literary Agents" request for information sent on 30 April 2003.

46 Notification, p. 107.

47 This emerged from the answers to questions 18 and 24 in request for information 14.685 and questions 13 and 18 in the "Publishers Phase II" request for information sent on 17 July 2003: the advances paid to the authors of the 100 best-selling general literature titles that are not translations were compared with the advances paid to the authors of the 100 best-selling general literature titles that are translations. The same proportions are found for children's books.

48 XO is a publishing house whose books are marketed and distributed by VUP; VUP owns 25% of its capital.

49 The annual advances paid by this publishing house to acquire French rights were [...]* than those paid to acquire foreign rights in 2001 and 2002 (answer to questions 16 and 19 in request for information 14.685).

50 Notification, p. 30.

51 Notification, p. 30.

52 Sometimes the authors of school textbooks are clearly identified, and indeed may be known to teachers, who prescribe them for their pupils (e.g. the textbook on French literature by Lagarde and Michard), but such cases of identified authors remain marginal.

53 In its decision to initiate the detailed examination under Article 6 (1) (c) of the Regulation, the Commission envisaged the existence of a market for primary publishing rights for guides and manuals. However, the investigation carried out during the detailed examination showed that most such works were collectively authored, as claimed by the notifying party (notification, p. 30). In responses to the statement of objections third parties have argued that guides and manuals and art books come under the category of collaborative works for which there are royalties. However, since Ipsos classifies 58% of guides and manuals as collective works and the publishers canvassed provided no figure for advances for the guides and manuals they themselves classify as collective works, guides and manuals and art books will be considered as collective works for the purposes of this analysis

54 In the (rare) cases where individual works are produced in other categories, the advances paid by publishers to authors are very low, if indeed any advance is paid at all.

55 To obtain this result, the Commission referred to the amount of the advances which publishers said they had paid to authors on the best-seller lists for general literature titles and children's books according to the Ipsos study submitted by the notifying party (answers to questions 18 and 24 in request for information 14.685 and questions 13 and 18 in the "Publishers Phase II" request for information sent on 17 July 2003).

56 Notification, p. 111.

57 Notification, p. 106.

58 In the case of books by translated authors, literary agents sometimes negotiate the second life of a book independently of the large-format publishing rights (answers to questions 29 and 30 in request for information 14.685 and questions 23 and 24 in the "Publishers Phase II" request for information sent on 17 July 2003).

59 In general, primary publishing contracts stipulate that 50% of the royalties paid by the publishing house publishing the work in pocket format (advance and proportional payment) go to the author, the remainder going to the primary publisher.

60 To obtain this result, the Commission referred to the amount of the advances which publishers said they had paid to authors on the best-seller lists for general literature titles, according to the Ipsos study submitted by the notifying party (answers to questions 18 and 24 in request for information 14.685 and questions 13 and 18 in the "Publishers Phase II" request for information sent on 17 July 2003) and the amount of the advances which publishers said they had paid for pocket-format books on the best-seller list for general literature titles, according to the Ipsos study supplied by the notifying party (answers to question 3 in the "Paperbacks" request for information sent on 7 October 2003).

61 The proportional payment for a general literature title that is not a translation is generally 10% for the first 10 000 copies sold, 12% for sales of between 10 000 and 20 000 copies and 14% beyond that. For a translated book the author is generally paid 8%, 10% and 12% respectively.

62 Hachette Livre holds 35% of its capital.

63 Answers to questions 25, 26 and 27 in request for information 14.685 and questions 19, 20 and 21 in the "Publishers Phase II" request for information sent on 17 July 2003.

64 Nor for guides and manuals, which are essentially collective works and therefore rarely the subject of the acquisition of rights.

65 Notification, p. 113.

66 Reply to the statement of objections, pp. 71 et seq.

67 Notification, p. 113.

68 Answer to question 12 in the - France Loisirs request for information sent on 23 April 2003.

69 Club rights for general literature titles and children's books are generally higher than for guides and manuals and art books.

70 As regards class 1 bookshops, the minimum turnover must be euro 20 million to cover the expenditure of a team of representatives (answers to question 3 in the "Marketing/Distribution" request for information sent on 16 April 2003).

71 See pp. 268 and 269 of the notification for a breakdown of investments and sunk costs. The Commission's investigation found that the minimum sunk industrial and IT investments for a small organisation are around euro 12 to 15 million; the size of the minimum investment will of course depend on the volumes handled and may therefore be considerably greater if volumes are higher.

72 In some cases, the publisher may decide to do the marketing itself for a given class of customer, in general class 1 bookshops.

73 See the answers to question 43 in the "Publishers" request for information sent on 17 July 2003.

74 Notification, pages 247 and 248.

75 Distribution also comprises other secondary functions, such as book restoration. This may easily be outsourced.

76 Hachette Livre's main site in France is in the Paris region. It runs two purely logistical sites at Nantes and Lyon, which also serve as salesrooms together with seven other salesrooms at Vanves, Toulouse, Marseille, Nice, Bordeaux, Rennes and Lille. VUP's main distribution centre in France is in the Paris region; it also has nine regional salesrooms.

77 The book trade is a unique creative industry made up of a large number of different flows. A total of 45 787 new titles and new editions were launched in France in 2002. On average the flow of new titles accounts for 40% of publishers' turnover and requires an automatic distribution function combined with a right for retailers to return unsold copies. There is also a "pull" effect caused by orders from retailers for restocking, together with promotion campaigns. Finally, the system must be flexible enough to absorb uneven flows at the same time: there may be a sudden surge in demand for fashionable titles following a promotion campaign or television exposure. Such titles co-exist alongside stable titles, for which demand is much more regular, and with school textbooks, for which demand tends to be seasonal.

78 Prisme is an economic interest grouping (EIG) of which equal shares are owned by ABX Logistics, Danzas and Geodis. The participating parties are the Commission de Liaison Interprofessionnelle du Livre ("CLIL" - representing publishers/marketers and bookshops) and the above-mentioned members of the EIG.

79 In its reply the notifying party states that Prisme is used mainly by class 1 bookshops in the provinces. It emerged from the Commission's investigation that class 2 bookshops also use Prisme, while other dealers (hypermarkets and wholesalers) are supplied direct, by-passing Prisme.

80 Information supplied by Prisme.

81 Answer to question 40 in request for information 14.685.

82 Answer to question 76 in the "Marketing/Distribution" request for information sent on 20 June 2003.

83 See the answers to question 34 in the "Publishers Phase II" request for information sent on 17 July 2003 and question 45 in the "Marketing/Distribution" request for information sent on 20 June 2003.

84 See Annex 108 to the notification for a description of independent wholesalers.

85 For a full description of how wholesalers operate, see the section on the market for the sale of books by wholesalers to class 3 dealers.

86 Answer to question 49 in request for information 10.069.

87 Notification, p. 260.

88 Albin Michel in particular does its own marketing to class 1 bookshops.

89 For example, in the case of the merged entity, Hachette Livre's class 1 sales force in France is divided into [...]* sales teams, the latter also serving hypermarkets. The Hachette Livre sales force has [...]* in all. In Belgium, Hachette Livre exercises its activities via Dilibel for the various classes mentioned above and via AMP for small outlets and supermarkets. VUP's class 1 sales force in France is divided into [...]* sales teams ([...]* in all). The other sales forces are [...]* class 2 [...]* and [...]* for hypermarkets [...]*. In Belgium, VUP exercises its activities via VUPS Bénélux for the various classes mentioned above.

90 In its reply to the statement of objections, the notifying party points out that it is normal practice to make a distinction between specialised superstores and class 1 bookshops. However, this is not how the situation was presented in the notification (see pp. 235-242), which shows that the same marketing teams are used for large bookshops and specialised superstores. It emerged from the Commission's investigation that all marketers deal with specialised superstores such as FNAC or Virgin via class 1 marketing teams.

91 Hachette Livre considers that there are [...]* class 1 bookshops, and divides this class into two sub-categories, with the [...]* largest customer marking the end of the first sub-category. For VUP the number of class 1 bookshops is [...]*. Depending on the structure of the lists marketed, class 1 contains between 750 and 1 250 retailers. Class 2 begins where class 1 ends and may contain different numbers of outlets depending on the penetration capacity of marketers/distributors. For Hachette Livre class 2 comprises [...]* outlets, for VUP it comprises [...]*. For Seuil and Gallimard class 2 comprises between 2 000 and 3 000 outlets. For the other marketers, the borderline is between 500 and 1 000 outlets.

92 In some cases, the representatives assigned to wholesalers and to the largest bookshops may be the same, as they are experienced representatives who also deal with the biggest booksellers.

93 See for example pages 282 and 283 of the notification and the answers to question 29 of the "Publishers" request for information sent on 30 April 2003.

94 Answer to question 93 in request for information 14.685.

95 Class 2, however, also includes bookshops specialising in, for example, strip cartoons or esoteric literature.

96 Answers to questions 2 and 3 in the "Dealers" request for information sent on 14 March 2003 and analysis of the file supplied by Hachette Livre in response to question 3 in request for information 6.509.

97 In its reply to the statement of objections, the notifying party points out that hypermarket chains, specialised superstores and bookshop chains are bigger customers, if all their individual outlets are added together. However, the Commission's investigation showed that each shop in these chains was canvassed individually.

98 Where a hypermarket chain also includes supermarkets, the central listing is generally common to both. However, product listing is only the first stage. The range on offer in a supermarket, being so much narrower, will be very different from the range in a hypermarket. Moreover, the choice of books is made by the department manager in a hypermarket, whereas in a supermarket, it is usually left to the wholesaler.

99 The products on offer may be extended temporarily (mainly at the beginning of the school year) to include educational supporting materials and reference works, but generally consist of an assortment of works in pocket format.

100 Notification, p. 284, and answers to request for information 9.755.

101 See Annexes 106 and 107 to the notification.

102 See in particular the answers to questions 81 to 84 in request for information 14.685.

103 Answers to question 6 in the "Marketing/Distribution" request for information sent on 16 April 2003.

104 Answers to question 26 in the "Marketing/Distribution" request for information sent on 20 June 2003.

105 Answer to question 85 in request for information 14.685.

106 Taking 15 wholesalers served as an arbitrary basis.

107 See the marketing/distribution margins on works selected by Hachette Livre and VUP in answers to the additional questions sent on 2 October 2003.

108 See Annex 79 to the notification, and the 1992 and 2001 memoranda of agreement.

109 Answers to questions 82 to 84 in request for information 14.685.

110 Answers to question 38 in the "Marketing/Distribution" request for information sent on 20 June 2003.

111 Answer to question 10 in the .Hypermarkets/Supermarkets. request for information sent on 11 August 2003.

112 For example, Vilo, whose main activity is the marketing of art books.

113 A publisher signs a single contract rather than separate contracts for distribution to large bookshops, to hypermarkets and to small outlets.

114 Case T-221-95 [1999] ECR II-1229.

115 Answer to questions 31 and 32 in the "Marketing/Distribution" request for information sent on 20 June 2003.

116 See in particular pp. 417 et seq. of the notification.

117 Answers to questions 24, 25, 27, 28, 58 and 59 in the "Marketing/Distribution" request for information sent on 20 June 2003.

118 For example, when Albin Michel, a publisher with substantial marketing and distribution needs, switched from VUP to Hachette Livre, [...]* (answers to questions 43 to 53 in request for information 14.685).

119 Contrary to what the notifying party states in its reply, marketing and distribution costs are by nature fixed, as explained by Hachette in reply to question 82 in request for information 14.685.

120 See also the judgment of the Court of First Instance in Case T-310/01 Schneider v Commission [2002] ECR II-4071, and the guidelines on vertical restraints (OJ C 291, 13.10.2000, p. 1), paragraph 98.

121 For example, most bookshops reported that they tend to take more books from representatives who visit them more often (question 13 in the .Dealers Phase II. request for information sent on 11 July 2003).

122 For example, the platforms of Hachette and VUP in Belgium are smaller than their French counterparts and have higher average costs (see answer to questions 97-99 in request for information 14.685).

123 Notification, pp. 67 et seq.

124 Reply to the statement to the objections, pp. 122 et seq.

125 In the book industry there is a very large number of titles available (including new issues) and limited space available for selling them (booksellers. shelves). Booksellers who opt for a new issue that they fail to sell bear not only carriage costs when returning the title and the loss of the premium (publishers give booksellers quantity premiums and penalise them on the basis of the average frequency of returns) but also lose the margin (difference between the price paid to the publisher and the price charged to the customer) that they would have earned if they had chosen a more saleable new issue.

126 Written observations by the notifying party on the Commission decision to open proceedings, 26 June 2003, p. 2.

127 Commission Decision of 22 August 2000 relating to the referral of the case No COMP/M.2044 Interbrew/Bass to the United Kingdom Competition Authorities, pursuant to Article 9 of Regulation 4064-89.

128 .While it cannot be excluded that the different routes to market represent different relevant markets, for the purposes of this decision it is not necessary to define narrower markets since even on the widest possible basis, i.e. the supply of beer to the on-trade and the supply of beer to the off-trade, competition problems arise and the markets fulfil the criteria in Article 9. For the same reason it is not necessary to decide whether distribution, and in particular secondary distribution, represents a separate service market. (paragraph 31).

129 Commission Decision of 22 June 2000 in Case COMP/JV.40 - Canal+/Lagardère/Canal Satellite.

130 See in particular Commission Decision of 15 February 2002 in Case COMP/M.2544 . Masterfoods/Royal Canin (OJ C 79, 3.4.2002, p. 10).

131 See in particular Commission Decision of 17 May 1999 in Case COMP/M.1397 - Sanofi/Synthélabo (OJ C 23, 27.1.2000, p. 4).

132 Notification, p. 241.

133 Even if it is the marketers who initially determine general terms of business (Notification, pp. 277 et seq").

134 Notification, pp. 279 et seq.

135 Cf. pages 282 and 283 of the notification and the answers to question 29 in the request for information on publishers sent on 30 April 2003.

136 Class 2, however, also includes bookshops specialising in, for example, strip cartoons or esoteric literature.

137 Answers to questions 2 and 3 in the "Dealers" request for information sent on 14 March 2003, and the analysis of the file supplied by Hachette Livre in response to question 3 in request for information 6.509.

138 Where a hypermarket chain also includes supermarkets, the central listing is generally common to both. But product listing is only the first stage. The range on offer in a supermarket, being so much smaller, will be very different from the range in a hypermarket. And the choice of books is made by the department manager in a hypermarket, whereas in a supermarket it is usually left to the wholesaler.

139 Answer to question 84 in request for information 14.685.

140 Notification, annex 76, and replies to question 21 in the "Marketing/Distribution" request for information sent on 26 June 2003.

141 Answer to question 38 in the "Marketing/Distribution" request for information sent on 26 June 2003.

142 Notification, pp. 69 and 71.

143 Notification, p. 105 and pp. 114 et seq.

144 Notification, p. 71.

145 Answers to question 21 in the .Publishers B. request for information sent on 16 April 2003 and to question 27 in the .Publishers A. request for information sent on 16 April 2003.

146 Answers to question 21 in the .Publishers B. request for information sent on 16 April 2003 and to question 27 in the .Publishers A. request for information sent on 16 April 2003.

147 Robert is active in reference books but not in general literature. Gallimard and Seuil, reputed publishers of general literature titles, are not active in the sale of educational books. Hatier, publisher of educational books, is not active in the sale of general literature titles. And La Martinière, publisher of art books, is not active in the sale of general literature titles or guides and manuals.

148 Notification, p. 284, and answers to question 9 in the "Dealers" request for information sent on 16 April 2003.

149 Mainly works written before the Second World War.

150 Notification, pp. 114 et seq.

151 With the possible exception of romantic fiction.

152 Notification, pages 119 et seq.

153 In particular, it is argued that there are wide ranges of prices for each of the formats, and a multiplicity of formats, so that there is a continuum between categories, with porous demarcation lines. This analysis is based in particular on an analysis of the price ranges for the various pocket-format series and concludes the prices for most series correspond to prices equivalent to the bottom end of the price range for large- format books. This continuity in pricing is further reinforced by qualitative factors such as the emergence of "intermediate" - "medium" formats - which have the smart appearance of a large-format and can take the place of either a pocket or a large-format book.

154 The Ipsos study of the 5 000 best-selling pocket-format titles supplied by the notifying party shows that 89% of the titles indexed have a standard pocket format ranging from 10x17 cm to 13x19 cm (54% of total sales being accounted for by the 11x18 cm format alone). By way of comparison, the Ipsos studies of the 1 500 best-selling large-format books records only 2.1% of the titles having this format. 155 Though this is a relatively rare situation.

156 Except for romantic fiction, crime novels and science-fiction, often published only in paperback format in specific series (Harlequin, SAS or Fleuve Noir, for example). Answers to question 30 in the .Publishers A. request for information sent on 16 April 2003.

157 In the period from 1997 to 2001, the average print run of a pocket-format book was 11 822, as against 7 363 for other formats, i.e. 61% higher. The average print run for a new issue in pocket format in 2001 was 14 208, as against 9 571 for large formats overall (notification, annex 43). In the specific case of general literature, the SNE reports in its Repères Statistiques 2002 that the average print run of a pocket- format book is 11 898, as against 7 949 for large formats, i.e. 50% higher.

158 Notification, p. 117.

159 Answers to question 30 in the .Publishers A. request for information sent on 16 April 2003. This is the principle for titles not yet in the public domain. Where a title is in the public domain, the price depends on market forces (prices for other pocket-format series).

160 On the basis of Ipsos data supplied by the notifying party for 2002.

161 Although there are "medium-format" series, the Commission's investigation reveals that this type of intermediate series is not widely represented on the market and really belongs in the large format category (smart appearance, prices).

162 Calculations on the basis of Ipsos data supplied by the notifying party for 2002.

163 In the period from 1997 to 2001, the average print run of a pocket-format book was 11 822, as against 7 363 for other formats. The average print run for a new issue in pocket format in 2001 was 14 208, as against 9 571 for large formats overall (notification, annex 43).

164 Notification, p. 117.

165 Notification, annex 43, p. 7.

166 Statistiques de l.édition en France en 2001, published by SNE, p. 7.

167 For Hachette Livre, LGF (for general literature - Hachette Livre also has a 35% holding in J'ai Lu). For VUP, Univers Poche (including "Pocket" and "10/18"). For Flammarion, J'ai Lu. Seuil manages its pocket-format titles through an independent department.

168 Answers to question 30 in the .Publishers A. request for information sent on 16 April 2003.

169 See [notifying party internal paper]*.

170 " Dominique Goust - Le poche offre une sorte de carte sous-marine de la lecture en France ", in Livre Hebdo 499, 31.1.2003, p. 8.

171 See in particular the answer given by the Syndicat de la Librairie Française to the "Dealers" request for information sent on 14 April 2003.

172 See annex 55 to the notification and the Sofres study.

173 Though it seems that certain best-sellers come out earlier.

174 One of the main results presented by the notifying party (notification, page 123) is an analysis of the (simple and not time-based) correlation between purchases by "major book buyers". Major buyers are defined by the notifying party as being the top 20% book buyers. The coefficient of correlation between "the fact of buying at least one pocket-format general literature title and the fact of buying at least one large-format general literature titles" is, according to the notifying party, 0.109. This analysis concerns demand substitutability among final consumers rather than among the dealers who are the demand side of the relevant markets. In any event, this low degree of correlation reflects the fact that buyers are not in the habit of buying the two categories of books (pocket and large format) together. This confirms the low level of substitution between the two types of format for the consumer.

175 See the article entitled "Livres: A fond les poches" in Le Vif/L.Express, 14 February 2003.

176 Notification, p. 143.

177 For legibility reasons, only the first 65 formats are represented in the charts (beyond the 23x19 cm format, only large-format books are present).

178 Answers to questions 25, 26 and 27 in request for information 14.685 and to questions 19, 20 and 21 in the "Publishers Phase II" request for information sent on 17 July 2003.

179 Calculations on the basis of Ipsos data supplied by the notifying party for 2002. For general literature, the average price of a pocket-format book is euro 5.78 whereas the average for a large- format book is euro 18.27.

180 For legibility reasons, only the first 28 formats are represented in the charts (beyond the 24x30cm format, only large-format books are present).

181 Answers to questions 25, 26 and 27 in request for information 14.685 and to questions 19, 20 and 21 in the "Publishers Phase II" request for information sent on 17 July 2003.

182 Calculations on the basis of Ipsos data supplied by the notifying party for 2002.

183 In France, school books are defined by an Order (décret) of 8 August 1985 implementing the Books (Uniform Pricing) Act (the Lang Act) as "textbooks and the exercise books and practical workbooks that accompany them, regularly used in teaching irrespective of level and designed to correspond to a syllabus defined or approved in advance by the Ministry of National Education or the authority with powers to supervise education".

184 Notification, pp. 166 et seq.

185 The publisher must produce a work that corresponds to the official syllabus imposed by the national education authorities and to teachers' expectations. The time allowed between publication of the syllabus and publication of the work is usually no more than 18 months.

186 In France, the Lang Act excludes school textbooks from the provision confining variations in resale prices to 5%.

187 See answers given by Albin Michel and Flammarion to questions 49, 57 and 60 in the .Publishers A. request for information and the records of telephone conversations between the Commission and Magnard (Albin Michel) and Delagrave (Flammarion) on 23 September 2003.

188 Answer given by Bréal to question 28 in the .Publishers B. request for information.

189 Answer given by Bréal to question 28 in the .Publishers B. request for information.

190 Record of telephone conversations between the Commission and Magnard (Albin Michel) on 23 September 2003.

191 Reply given by Flammarion to question 42 in the .Publishers A. request for information and record of telephone conversations between the Commission and Delagrave (Flammarion) on 23 September 2003.

192 Notification, pp. 164 and 165.

193 Notification, p. 188.

194 Notification, p. 190.

195 Notification, p. 189.

196 Notification, p. 203.

197 Notification, p. 205.

198 Notification, pp. 204 and 209.

199 Notification, p. 222.

200 The notifying party points out in its notification that computerised versions are sold at the price of single-volume reference works, see notification, p. 222.

201 In practice, wholesalers can either visit and deliver at outlets direct or give retailers the possibility of collecting books direct from salesrooms. In the latter case, class 3 outlets are not routinely visited but contacted by the wholesalers by telephone or any other convenient means.

202 This is the standard assortment, but it is quite possible that the actual assortment in certain small outlets might diverge from it as some outlets, depending on their geographical location and clientele profile, might specialise in one or more non-typical ranges of titles (such as regional tourist guides or art books dealing with cooking, culture, history or geography).

203 The Commission's investigation reveals that the level of discounts given to small outlets and supermarkets by wholesalers lies around 20%-30% whereas those given by publishers to wholesalers, hypermarkets and booksellers are around 35%, 30%-35% and 30%-40% respectively of the list price exclusive of tax.

204 For example FNAC or Virgin.

205 Shops whose main activity is to sell products relating to a leisure activity (gardening, pet care, DIY, etc") and which generally offer books concerning these activities.

206 In France books are sold in remainder shops (solderies) at a discount that exceeds the statutory 5% applicable in traditional outlets. Under Article 5 of the Lang Act, only books which have been published or imported for over two years and which were last supplied more than six months ago may be sold at a discount of more than 5%.

207 Notification, p. 289.

208 Commission Decision of 22 April 1999 in Case COMP/M.1407 - Bertelsmann/Mondadori (OJ C 145, 26.5.1999, p. 4).

209 In Decision 89-D-41 the French Competition Council held that the sale of books through book clubs formed a specific market because of the restricted nature of the catalogue, the membership requirement for purchasing a book and the difference in the presentation of works. This view was upheld by the Paris Court of Appeal in its judgment of 11 March 1993 in France v Grand Livre du Mois.

210 Notification, p. 106.

211 The notifying party does not, in fact, include this among the criteria determining the choices of publishers and authors (see notification, p. 28 et seq").

212 Such as Jacqueline Harpmann, Raymond Devos or Amélie Nothomb (see notification, p. 106).

213 See, for example, participants in the Salon International du Livre de Tanger 1999 and their publishers http://www.ambafrance-ma.org/archives/espacult/silt/ind_part.html.

214 Answers to question 5 in the "Publishers Phase II" request for information sent on 17 July 2003 and to question 12 in the "Authors" request for information sent on 30 July 2003.

215 The Goncourt was awarded to the Lebanese writer Amin Maalouf in 1993, and in 1995 to Andrei Makin, born in the Soviet Union. It should be added that it is sufficient to be a French-language author to be entitled to belong to the Académie Goncourt, hence the election of the Spanish writer Jorge Semprun in 1996.

216 The Renaudot prize was awarded in 2000 to the Ivorian writer Ahmadou Kourouma.

217 Notification, p. 63.

218 The notifying party claims that French publishers, .who have a domestic market that is larger than that of Belgian, Swiss or Luxembourg publishers, can offer foreign publishers and agents the best financial terms: as the volumes of sales in the French market are larger than what can be sold in the other French- speaking countries, proportional royalties are naturally higher and the amounts of advances offered can be higher as it is easier for the publisher to cover them. (Notification, p. 108).

219 Notification, annex 15.

220 Answers to question 5 in the "Publishers Phase II" request for information sent on 17 July 2003 and question 20 in the "Literary Agents" request for information sent on 30 April 2003.

221 Answers to question 5 in the "Publishers Phase II" request for information sent on 17 July 2003 and question 20 in the "Literary Agents" request for information sent on 30 April 2003.

222 It should be pointed out that Belgian publishers such as Éditions Complexe, les Éperonniers, La Longue Vue and Luce Wilquin publish French translations of works in foreign languages.

223 For instance, the Canadian publisher Leméac, which has assigned pocket-format rights to Flammarion for publication in the "J'ai Lu" series or Boréal, which has assigned such rights to Seuil for publication in the "Points" series. The Belgian publisher De Boeck has also assigned pocket-format rights to Seuil for publication in the "Points" series. Casterman apparently did likewise before being taken over by Flammarion.

224 Answers to question 6 in the "Publishers Phase II" request for information sent on 17 July 2003.

225 In this connection see in particular contracts for assignment of pocket-format rights supplied by the notifying party (notification, annex 15).

226 Notification, annex 107.

227 Answers to question 36 in request for information 14.685; to remedy the .lack of econometric rigour. claimed here by the notifying party in its reply to the statement of objections, it can be stated in the case of Hachette that of the 72 publishers marketed and distributed by Dilibel in Belgium, only 12 are not marketed and distributed by Hachette Livre in France, which means that in around 83% of cases, a single service-provider is used by the publisher for the entire French-speaking area.

228 Answers to questions 67 to 69 in request for information 14.685 and questions 66 to 69 in request for information 10.069.

229 Answer given by Hachette Livre to question 18 in request for information 17.204 and answers to questions 65 to 73 in the "Publishers Phase II" request for information given by Eyrolles, Flammarion, Gallimard, Gründ, Lavoisier, Payot/Rivages, Sélection du Reader's Digest and Seuil.

230 FNAC, however, may on occasion supply its Swiss shops direct from its French warehouses.

231 See notification, p. 84 and the answer given by the notifying party to question 131 in request for information 14.685 and the answers to questions 65 to 73 in the "Publishers Phase II" request for information given by Albin Michel, De Boeck, La Documentation Française, Eyrolles, Flammarion, Gallimard, Gründ, Lavoisier, Pearson Education France, Payot/Rivages, Sélection du Reader's Digest and Seuil. See also the article "Prix du livre : un marché de funambules" dated 20 March 2001 published on the Internet site of Télévision Suisse Romande at http://www.tsr.ch/emission/abe/archive/01/010320.html.

232 In particular export commission agents who receive orders from purchasers (in particular foreign bookshops), group them by publishing house and send them to the publishing house concerned. Once they have received the books, they send them to the bookshops. In addition to the generalised operators SIDE and CELF, the other operators in this sector specialise in highly specific segments (Amateur de livres International in supplying libraries, SFL in supplying bookshops in the overseas departments and territories, Touzot in scholarly works, and Lavoisier and ESKA in scientific and technical works).

233 Originally the tabelle was designed to offset the additional import costs incurred by French publishers/marketers selling in Belgium as a result of customs duties, exchange rate fluctuations, additional transport costs, etc. There are now indications, however, that following the removal of barriers to cross-border trade and in particular the introduction of the euro, this practice is no longer justified.

234 Around 15 000 customers in France for 60 million inhabitants, a ratio of 0.25 per thousand as against some 1 000 customers in the French-speaking part of Belgium for some 4 million (French-speaking) inhabitants, giving the same ratio of 0.25 per thousand.

235 See Commission communication on the definition of relevant market for the purposes of Community competition law (OJ C 372, 9.12.1997, p. 5).

236 Such as the documents sent on 18 July 2003 by the representatives of Seuil explaining the order circuit for hypermarkets.

237 Notification, p. 285.

238 It should, however, be borne in mind that such markets, which put publishers in direct contact with final consumers, could not be identified, except in the case of direct sales. In virtually all cases publishers sell to retailers (class 1, class 2 and class 3 bookshops and hypermarkets) which then sell to final consumers. For this reason the notifying party's arguments relating to final demand have to be viewed with caution.

239 Notification, p. 77 et seq.

240 Notification, p. 82 et seq.

241 This is the rate practised by Dilibel, a subsidiary of Hachette Livre, in Belgium and Luxembourg. The increase is, however, smaller for certain works such as strip cartoon albums, which Dilibel increases by only [0-10]%.

242 Notification, p. 87.

243 Notification, p. 88 et seq.

244 Commission Decision of 25 November 1981 in Case IV/428 - VBBB/VBVB (OJ L 54, 25.2.1982, p. 36); Court of Justice in Joined Cases 43/82 and 63/82 VBVB and VBBB v Commission [1984] ECR 19.

245 Commission Decision 89-44-EEC of 12 December 1988 in Case IV/27.393 and Case IV/27.394 - Publishers Association - Net Book Agreements (OJ L 22, 26.1.1989, p. 12); Court of Justice in Case C-360-92 Publishers Association v Commission [1995] ECR I.23.

246 Notification, pp. 79 and 80.

247 See Court of First Instance in Case T-66-89 Publishers Association v Commission [1992] ECR II-1995, paragraph 84; Court of Justice in Joined Cases 43-82 and 63-82 VBVB and VBBB v Commission [1984] ECR 19, paragraphs 44 and 47.

248 Notification, p. 96.

249 See the section of this Decision on the definition of markets for the sale of books by publishers to dealers.

250 The Lang Act does in fact cover conditions for negotiating discounts, stipulating for instance that quality aspects must be taken into account, and laying down certain compulsory services, such as orders for single items.

251 Percentage of the list price (fixed in France) or the recommended price (in Belgium and Luxembourg).

252 Marc Ivaldi, "Report No 1 on Case COMP/M.2978 - Lagardère/Natexis/Vivendi Universal Publishing", 24 June 2003, study conducted for the purposes of this case on the basis of the information supplied by the notifying party.

253 The study covers general literature titles, school textbooks and educational supporting materials. The results of this study for the latter two categories of books are described in the part of this Decision relating to the definition of markets.

254 The French list price, if not deliberately covered by a sticker applied by the distributor showing the "tabelle" price, is as a rule printed on the book and can therefore be seen by the final consumer. Transparency is further enhanced by the advertisements and promotion campaigns concerning books sold in France and Belgium which appear in the French newspapers widely available from newsagents in Brussels and the French-speaking part of Belgium.

255 Notification, p. 114 et seq.

256 The ten best-selling works shown each week by the newspaper Le Soir are, with some rare exceptions, published by French publishers, as are the ten titles ("Nos choix étoilés") recommended by the newspaper (see in particular Le Soir of 24 October 2003).

257 Commission Decision of 15 February 1999 in Case COMP/M.1377 - Bertelsmann/Wissenschaftsverlag Springer (OJ C 122, 4.5.1999, p. 19).

258 Notification, p. 27.

259 The reference price for determining the discount is then the French list price as printed on the book cover.

260 On this point, see the analysis of market definition in the media sector in "Market definition in the Media sector, Economic issues" (2002) report by Europe Economics for the European Commission, DG Competition", p. 97 et seq.

261 See "Le marché du livre de langue française en Belgique: Evolution 1996 - 2000", a study conducted for the literary promotion department of the Directorate-General for Culture in the Ministry for the French-speaking Community in Belgium, page 6. The notifying party estimates that imported books account for 77% of all books sold in Belgium (answer to question 14 in request for information 10.069).

262 Notification, p. 89.

263 The notifying party has given only one example of television advertising for books in Belgium, that of the Larousse campaigns for the Petit Larousse illustré to coincide with the start of the new school year (answer to question 15 in request for information 10.069 of 16 May 2003).

264 In the internal paper [...]*. By contrast, nowhere in this document from the notifying party is there a reference to separate national markets within the same language area.

265 No such analysis was necessary because there were agreements which by nature restricted competition (fixed-price clauses) and were caught by Article 85 (1) (now Article 81 (1)) of the Treaty. See also Article 4 (a) of Commission Regulation (EC) No 2790-1999 of 22 December 1999 on the application of Article 81 (3) of the Treaty to categories of vertical agreements and concerted practices (OJ L 336, 29.12.1999, p. 21).

266 Court of Justice in Joined Cases 43-82 and 63-82 VBVB and VBBB v Commission [1984] ECR 19, paragraphs 44 and 47. In paragraph 47 the Court rejected the applicants' argument that trade between Member States would not be affected because the homogeneous language area between the Flemish part of Belgium and the Netherlands was the relevant geographic unit and not the national territories.

267 Court of First Instance in Case T-66-89 Publishers Association v Commission [1992] ECR II-1995, paragraph 84.

268 In particular where France Loisirs is not included in these calculations and only books published in French are taken into consideration.

269 See answers given by Hachette Livre to question 18 in request for information 17.204 and the answers to questions 65 to 73 in the "Publishers Phase II" request for information given by Eyrolles, Flammarion, Gallimard, Gründ, Lavoisier, Payot/Rivages, Sélection du Reader's Digest and Seuil.

270 See page 84 of the notification and the answer given by the notifying party to question 131 in request for information 14.685 and the answers to questions 65 to 73 in the "Publishers Phase II" request for information given by Albin Michel, De Boeck, La Documentation Française, Eyrolles, Flammarion, Gallimard, Gründ, Lavoisier, Pearson Education France, Payot/Rivages, Sélection du Reader's Digest and Seuil. See also the article "Prix du livre: un marché de funambules" dated 20 March 2001 published on the Internet site of Télévision Suisse Romande at http://www.tsr.ch/emission/abe/archive/01/010320.html.

271 Notification, p. 175.

272 See notification, pp. 175 and 176, and the answers given by De Boeck and Labor to questions 72 and 76 to 78 in the "Publishers Phase II" request for information.

273 See notification, p. 358, and the record of the telephone conversation between Commission staff and De Boeck on 3 October 2003.

274 See the note on VUP discounts to retailers in Benelux and France by Marc Ivaldi in the study conducted for the purposes of this case on the basis of information supplied by the notifying party, 5 July 2003.

275 Answers to questions 135 and 184 in request for information 14.685.

276 See the note on VUP discounts to retailers in Benelux and France by Marc Ivaldi in the study conducted for the purposes of this case on the basis of information supplied by the notifying party, 5 July 2003.

277 It has been found that French publishers sell exactly the same educational supporting materials in Belgium and Luxembourg as are sold in France.

278 Notification, p. 192.

279 With the exception of their immediate rival, DNL, which specialises in supermarkets in France and, previously, in Belgium.

280 Notification, annex 108.

281 Notification, p. 224.

282 VUP uses Encyclopaedia Britannica for direct sales of its books in France, and uses Eurolivres in Belgium. Hachette users various Belgian distributors (SDS, Sidec and Organi-Team) and has its own network for France.

283 Notification, p. 289.

284 See, for example, Commission Decision 1999-674-EC of 3 February 1999 in Case IV/M.1221 - Rewe/Meinl (OJ L 274, 23.10.1999, p. 1), or the Commission Decision of 25 January 2000 in Case COMP/M.1684 . Carrefour/Promodès (OJ C 164, 14.6.2000, p. 5).

285 See, for example, the Commission Decision of 13 December 2002 in Case COMP/M.2898 - Leroy Merlin/Brico (OJ C 187, 7.8.2003, p. 11).

286 And also stations and hospitals.

287 Commission Decision of 9 February 2000 in Case COMP/M.1628 - TotalFina/ELF (OJ L 143, 29.5.2001, p. 1).

288 See, for example, the cases already cited, Case IV/M.1221 - Rewe/Meinl, Case COMP/M.1684 - Carrefour/Promodès, or Case COMP/M.2898 - Leroy Merlin/Brico.

289 Notification, p. 313.

290 Notification, pp. 41 to 60.

291 Notification, p. 313.

292 In this respect see, for instance, point 36 et seq. of the Commission notice on calculation of turnover under Council Regulation (EEC) No 4064-89 on the control of concentrations between undertakings (OJ C 66, 2.3.1998, p. 25).

293 In this respect see, for instance, the Commission notice on the concept of concentration under Council Regulation (EEC) No 4064-89 on the control of concentrations between undertakings (OJ C 66, 2.3.1998, p. 5).

294 This feature is not specific to publishing but is found in other areas of business, even though the fields in which such autonomy is exercised may vary from one sector to another depending on the peculiarities of the sector.

295 Notification, annex 22. 1 500 best-selling books

296 See, for example, http://www.loreal-finance.com.

297 For instance, the annexes to question 6 in request for information 14.685 show presentations and tables of figures from Hachette Livre's commercial reporting, in which several houses in the same segment are systematically compared and results expressed for the group as a whole. These documents confirm therefore that the different houses are controlled and coordinated and that a common focus applies, since each house's results are aggregated by type of work and uniform ratios are applied to the group as a whole.

298 Answers to question 17 in the "Dealers" request for information sent on 14 April 2003.

299 See, in particular, Court of Justice in Case 85-76 Hoffmann-La Roche v Commission [1979] ECR 461, paragraph 39, and Court of First Instance in Case T-102/96 Gencor v Commission [1999] II-753, paragraphs 201 and 202.

300 Notification, p. 300.

301 For the purposes of this analysis, "well-known authors" are those who figure in the 1 500 best selling books according to the Ipsos survey supplied by the notifying party.

302 According to the method used by the Commission and explained below, while the merged entity alone accounts for [50-60]% of the advances paid for the 100 best-selling "non-translated" books in 2002, its market share of the sales volumes for these same titles is [30-40]%. On the other hand, while the merged entity accounts for only [30-40]% of the advances paid for the 100 best-selling "translated" books in 2002, its market share of the sales volumes for these same titles is [35-45]%. The same lack of correlation is found with other players, such as Albin Michel or Gallimard.

303 E.g. in particular Amélie Nothomb or Didier Van Cauwelaert.

304 The notifying party states, however, in its answer to question 22 in request for information 14.685, that, since there are few titles concerned and given the difficulty of identifying the language of authorship, the figures must be interpreted with considerable caution.

305 This situation highlights the fact that it is not an author's nationality which determines whether he or she belongs to the French rather than the foreign rights market, but the language in which he or she originally produced the work.

306 The Commission has extracted from among the 1 500 best-selling books in the Ipsos survey supplied by the notifying party the 100 "translated" best-sellers and the 100 "non-translated" best-sellers.

307 Contrary to the study supplied by the notifying party, which proposes that French and foreign rights should be distinguished by the author's nationality, the Commission has resorted to a breakdown by original language of the work ("translated" and "non-translated" works), in accordance with the definition of the relevant product markets.

308 Such titles account for 32% of sales of general literature titles in France.

309 In Hachette Livre's case, for example, the amount paid in advances to non-translated authors was euro [...]* million in 2001 and euro [...]* million in 2002 (answer to question 16 in request for information 14.685).

310 It was found that about 50% of the authors on the list of 100 "non-translated" best-sellers received advances in excess of euro 50 000. It was also found that above the euro 50 000 figure, the large majority of French-speaking authors (more than 70%) received a proportional remuneration of 15% or above. Below euro 50 000, however, the proportional remuneration was usually between 10% and 12% (answers to question 14 in request for information 14.685 and to question 9 in the .Publishers Phase II. request for information sent on 17 July 2003).

311 To do this, the Commission took the list of recurrent, well-known authors supplied by the notifying party in annex 14 to the notification.

312 Reply to the statement of objections, p. 32 et seq.

313 The notifying party explains that, while 48% of authors mention the level of authors. royalties/the proportional remuneration among the main criteria for choosing a publisher, only 14% of them mention the "moment of payment (amount of advances)" among the three main selection criteria, and only 7% mention the advance as the decisive criterion.

314 Examination of the titles published by an Hachette Livre publishing house figuring in the list of the 100 best-selling general literature titles in 2002.

315 See the Jurisclasseur littéraire et artistique, which shows the advance as a payment.

316 It will be noted in this respect that the notifying party assimilates the moment of remuneration to the criterion of the advance, thereby minimising in its reply to the statement of objections the criterion of the amount of the remuneration comprising both the advance and the proportional remuneration.

317 .Livres à fond les poches", Le Vif-L'Express, 14 February 2003.

318 42% of general literature publishers gave as the first criterion for selection of a publishing house by an author the personal link and the remuneration (answers to question 10 in the "Publishers B" request for information and question 11 in the "Publishers A" request for information sent on 30 April 2003). VUP also stated that "recognised authors generally demand high amounts" and that therefore "they must be guaranteed large advances" (answer to question 1 in request for information 18.544).

319 Answer to question 14 in request for information 14.685.

320 Notification, p. 58.

321 Notification, p. 59.

322 Comments on the statement of objections or the notifying party's reply to the statement of objections were received from five publishers, from the Syndicat de la Librairie Française, from the European Consumers. Organisation (BEUC), and from the members of Editis's representative consultation body (IRC).

323 Notification, p. 298.

324 Albin Michel is a publisher all of whose works are distributed, and some of which are marketed, by Hachette Livre, with which it has capital links through its holding in LGF.

325 In their answers, 83% of the French-speaking general-literature authors cited the human relation as one of the three principal criteria in selecting their publisher. The marketing criterion was cited as one of the first three criteria by 52% of authors, the amount of remuneration by 48%, promotion by 31%, distribution and the moment of remuneration by 14% and, lastly, vertical integration in the sales of the pocket format by 7% of authors (answer to question 6 in the "Authors" request for information sent on 30 July 2003). It may reasonably be considered that the marketing criterion (authors attracted by a marketer/distributor likely to market a book well) is, in part, linked to the remuneration criterion: a better marketed book provides its author with more remuneration.

326 Reply to the statement of objections, p. 39.

327 VUP has a 25% stake in XO, is its distributor and publishes its pocket-format books; Albin Michel has a stake in Le Livre de Poche, owned by Hachette Livre, and is distributed for all classes of customer, and marketed for all classes of customer except class 1, by Hachette Livre.

328 For instance, the author Max Gallo summarises the situation very well when talking about his publisher XO, marketed, distributed and 25% owned by VUP: "I ask only two things of a publisher", comments Max Gallo, who has been working with Bernard Fixot for seven years" "First, that he respects the author and, second, that he makes the novel available to the maximum number of readers". (see http://etudiant.lefigaro.fr/formation/cas_ecole/xo1.php).

329 [Notifying party internal paper]*.

330 Reply to the statement of objections, p. 42.

331 [Notifying party internal paper]*.

332 Answer to question 154 in the "Publishers" request for information sent on 30 April 2003.

333 "All the anticipated revenues" covers not only the income generated by sales of large formats and pocket-formats and commissions (as a percentage of the list price exclusive of tax) received from marketing and distribution, but also any profits generated by spreading the fixed costs of marketing and distribution over a greater volume of works.

334 On the link between publishing houses and distribution, the notifying party states that .since publishing, as has already been stated, is a risky occupation by definition, Hachette Livre acknowledges that a publishing house can make losses as a result of editorial choices that are unfortunate or not immediately recognised by the public. In addition, if a publishing house were to close, its contribution to shared services such as distribution would be lost without a corresponding reduction in the latter's fixed costs. (notification, p. 58).

335 Reply to the statement of objections, p. 44.

336 Report drawn up on 5 December 2002 by Secafi Alpha, chartered accountants and financial and strategic analysis consultants, on the proposal to sell the publishing division of VUP to Lagardère, as part of a commission assigned by the conciliation body representing the works councils affected by the notified transaction, p. 10.

337 See record of the meeting between Commission staff and the representatives of Albin Michel, 25 June 2003.

338 Reply to the statement of objections, p. 46.

339 See record of the meeting between Commission staff and the representative of Odile Jacob, 16 July 2003.

340 See the sections on the competition analysis of the notified transaction with regard to the markets for the sale of general literature titles in pocket format and the markets for marketing and distribution services.

341 The Lagardère group operates in the market for the purchase of advertising space through its advertising sales agencies Lagardère Active Publicité, for the broadcast media, and Interdeco, for printed media.

342 The Lagardère group has 100% control of the radio stations Europe 1, Europe 2 and RFM, and indirect holdings in MFM and Voltage FM. It has extensive holdings in the French-language press: (100% of La Provence, Corse Matin, Nice Matin, TV Hebdo, Var Matin, Elle, Isa, Paris Match, Parents, Jeune & Jolie, Pariscope, Photo, Première, L'Echo des Savanes, France Dimanche, Ici Paris and Télé 7 Jours; 80% of Entrevue, 60% of Le Journal du Dimanche, 42% of Marie Claire, Cosmopolitan, etc").

343 According to the Lagardère group's letter to shareholders of April 2003, available on the Internet at http://www.lagardere.com/info_financieres/lettres/Lettre-avril-2003.pdf.

344 According to the letter from the notifying party, dated 1 October 2003 and registered under the reference 15.519.

345 The information supplied by the notifying party on 1 and 10 October 2003 shows that, among the various publishers of general literature, the publishing houses in the Hachette Livre group receive a preferential discount [...]* from the radio stations in the Lagardère Group (Europe 1/Europe 2/RFM) [...]*, whereas for a campaign promoting a best-seller-type work on the Lagardère group radio stations an independent publisher receives an average reduction of some [...]*. The information supplied by the notifying party in its answer to question 8 in request for information 6.405 shows that the publishing houses in the Hachette Livre group also receive the highest discount [...]* from the Hachette Filipacchi Média publications (Paris Match, Elle, Le Journal du Dimanche, Parents, Télé 7 Jours, etc").

346 This possibility is confirmed by VUP's publishing houses in their answer to question 23 in request for information 18.544 [...]*.

347 See VUP's answer to question 6 in request for information 18.544.

348 See VUP's answer to question 6 in request for information 18.544.

349 See Coop-Alsace answer to question 6 in the "Conglomerate Effects" request for information sent on 25 August 2003.

350 See VUP's answer to question 6 in request for information 18.544. Éditions Robert Laffont go further, stating in this same answer that "a budget may be planned with the author but this remains extremely rare".

351 See VUP's answer to question 8 in request for information 18.544.

352 [Notifying party internal paper]*.

353 [Notifying party internal paper]*.

354 As part of its investigation, the Commission asked various publishers to supply the contracts of authors who had joined or left their publishing houses in the last five years. Analysis of these contracts shows that in more than 70% of cases where the Commission obtained the old and the new publishing contract the advance paid to the author by the new publisher (the proportional payment being equal or higher) was greater than that paid by the old.

355 Remuneration is considered to be a criterion for selecting a publishing house by more than 70% of authors, even if, as the Commission has already explained, it is not the leading criterion. For publishers, remuneration is cited on the same level as the personal link as the most important selection criterion for authors.

356 Reply to the statement of objections, p. 51.

357 Notification, p. 62.

358 See VUP's answer to question 1 in request for information 18.544.

359 Answer to question 1 in the "Conglomerate Effects" request for information sent on 20 August 2003.

360 [...]*. Gallimard's turnover (2001) = euro 234.7 million.

361 Answer to question 154 in the "Publishers A" request for information sent on 30 April 2003.

362 See page 7 of the brochure presenting the Lagardère group Repères 2002, included in annex 6 to the notification.

363 Answers to questions 2 and 3 in the "Conglomerate Effects" request for information sent on 20 August 2003.

364 Answer to question 3 in the "Conglomerate Effects" request for information sent on 20 August 2003.

365 Including the largest bookshop in Europe (Le Furet du Nord, Lille), according to the brochure presenting the Lagardère group Repères 2002.

366 VUP increases the presence of the merged entity in the sale of books to the final consumer by contributing the retail sales business it carries on through the Robert Laffont shop.

367 See record of the meeting between Commission staff and the representative of Odile Jacob, 16 July 2003.

368 [Notifying party internal paper]*.

369 As stated by the notifying party (see notification, p. 323), some of these publishers run a small pocket-format series, whose only purpose however is "to exploit their lists further, which means that they do not require any extensive buying of rights".

370 The latter two publishers, however, have only one book cited in the list of 47 books by recurrent, well-known authors drawn up by the Commission. XO, which publishes only 20 titles a year, is marketed and distributed by VUP, which has a 25% holding in it.

371 Answers to the "Publishers B" request for information sent on 16 April 2003.

372 VUP owns 25% of XO, and Hachette Livre owns 34% of De Fallois and 10% of Anne Carrière.

373 It will be noted that Albin Michel's situation is quite special in the French publishing world. While the other publishers of comparable size, such as Gallimard, Seuil and Flammarion, are all owners of marketing and distribution facilities and a pocket-format series so that they can give their successful works a "second life", this is not the case with Albin Michel. See in particular the description of the players involved.

374 Notification, annex 69.

375 In 2002 the outlets benefiting from Hachette Livre's marketing services on behalf of Albin Michel generated about 60% of the copies sold of the 20 best-selling Albin Michel works.

376 As indicated elsewhere (see description of the competitors) the nearest competitor of LDS is La DIL, a subsidiary of VUP. Only Hachette Livre and VUP, through LDS and La DIL, have networks serving small outlets throughout the French-speaking area, in which Albin Michel, a publisher of best-sellers, has to be present. When La DIL has been absorbed by Hachette Livre, Albin Michel will lose the possibility of getting LDS and La DIL to compete. Its dependence on Hachette Livre will clearly be greater therefore.

377 This is particularly true for authors of best-sellers, in whose publication Albin Michel specialises.

378 All of which generate between euro 100 million and euro 250 million in turnover, except for the merged entity, which in 2001 had a (hypothetical) combined turnover of euro [...] million.

379 It is the only player cited by the notifying party.

380 The right of first refusal is a right taken by the publisher in order to keep an author it has "launched" in its own "stable"; the principle is that if an author's first work is hugely successful, the publisher has every interest in ensuring that the author does not leave for another publisher that makes more attractive offers.

381 Contracts between the Hachette Livre publishing house and authors who had left or joined the group in the last five years, supplied in answer to question 14 in request for information 14.685. There were 59 such contracts; 38 of them contained no first-refusal clause.

382 And 10% of contracts for "translated" books.

383 Such as novels, essays on current affairs, historical essays, etc.

384 Even where a publisher does sue another publisher or an author for failing to comply with a first-refusal clause, this does not always make it possible to keep the author. In a recent example, Grasset won a case against Albin Michel complaining that the book L'Impératrice had been published by the latter in breach of a first-refusal clause linking the author to Grasset, but the author was still able to have her book published by Albin Michel.

385 Answer to question 4 in the "Pocket-format Rights" request for information sent on 7 October 2003.

386 Notification, p. 301.

387 Notification, pp. 300 and 301.

388 Most of the literary agents who replied to the Commission's questionnaire say that they operate on an auction basis (answers to question 6 in the "Literary Agents" request for information sent on 30 April 2003) and give as much place to remuneration as to the personal link with the publisher (answers to question 7 in the "Literary Agents" request for information sent on 30 April 2003).

389 Answer to question 22 in request for information 14.685.

390 The Commission extracted the 100 best-selling children's books from the 1 500 best-selling titles in the Ipsos survey supplied by the notifying party.

391 Answers to question 8 in the "Authors" request for information sent on 30 July 2003.

392 For the other categories of book, in the majority of cases, publication in pocket format means first publication in pocket format.

393 In practice this is 5% of the list price exclusive of tax, irrespective of the volume of sales.

394 In the case of books by "translated" authors, literary agents sometimes negotiate the primary and secondary publishing rights separately (answers to questions 29 and 30 in request for information 14.685 and to questions 23 and 24 in the "Publishers Phase II" request for information sent on 17 July 2003). In such cases, the seller is not the publisher of the primary publication but the literary agent.

395 Fleuve Noir is a special crime series, which consists almost entirely of first editions in pocket format and in which therefore pocket-format rights are seldom acquired.

396 See notification, p. 323.

397 To obtain this result, the Commission extracted from among the 5 000 best-selling pocket formats in the Ipsos survey supplied by the notifying party the 500 best-selling general literature titles (the Harlequin and Barbara Cartland series were disregarded, since these series are obviously pocket-format first editions).

398 Notification, p. 302.

399 However, the notifying party excludes from its calculations works which are in the public domain, i.e. for which the publisher does not have to pay copyright, and also "in-house" pocket-format series, which are designed to extend the exploitation of publishers' own lists, which means that they do not require extensive buying of rights, and "popular" pocket series, which mainly publish previously unpublished material.

400 The purchase of pocket-format rights to general literature titles does not include purchases of pocket-format rights to crime novels published in special series.

401 See the discussion of the methods for calculating the primary market for the purchase of rights to general literature titles. While it is true that risk-taking is less important for the acquisition of secondary rights than for the acquisition of primary rights, this activity is not without risk for a publisher, which cannot know in advance what its sales will be.

402 The proportional remuneration has even less impact on the purchase of secondary publishing rights, since it is fixed at 5% for virtually all contracts.

403 Given the difficulties encountered, only 2002 is representative, since the Commission did not obtain exact amounts for all players for 2000 and 2001.

404 From the 5 000 best-selling pocket formats in the Ipsos survey supplied by the notifying party, the Commission extracted the 500 best-selling books (the Harlequin and Barbara Cartland series were omitted, since they are obviously first published in pocket format). It then checked to see which of these 500 titles was first published by a publisher in the group and which by an outside publisher. Of the 500, 163 were first published by different publishers (outside the group) in pocket format, which means that in their case the pocket-format rights were purchased.

405 Court of First Instance in Case T-221-95 Endemol v Commission [1999] ECR II-1299, paragraph 109.

406 In the case of an external assignment, primary publishing contracts provide that 50% of the royalties paid by the pocket-format publishing house (allowance plus proportional remuneration) go to the author, the balance going to the primary publisher. In the case of an in-house assignment, publishing contracts provide that the publisher will pay to the author 5% of the author's royalties on the pocket format sales.

407 Notification, p. 324.

408 It should be noted for the purposes of this analysis that Hachette owns 35% of the "J'ai Lu" series, but that all the market shares of this series ([5-15]%) have been attributed to Flammarion.

409 The Commission has here departed from the method it followed in the statement of objections by incorporating into its calculation of market shares the fact that in the future Hachette Livre's acquisitions of secondary rights from VUP, and VUP's from Hachette Livre, will be regarded as in-house sales and need not therefore be taken into account.

410 Whether the notifying party's or those of the Commission.

411 Answer to question 25 in the "Publishers A" request for information sent on 16 April 2003.

412 Notification, p. 304.

413 Answers to question 13 in the "Authors" request for information sent on 30 July 2003 and to question 15 in the "Literary Agents" request for information sent on 30 April 2003. It is also confirmed by the publishers, who consider that, even if there is not always a formal (contractual) agreement, the author's or literary agent's permission is always requested (answers to questions 23 and 24 in the "Publishers Phase II" request for information sent on 17 July 2003).

414 Note on the contractual relations between authors and publishers drawn up by Actes Sud and received by the Commission on 4 August 2003.

415 Reply to the statement of objections, p. 61.

416 And even the second market player, according to certain calculation methods.

417 Reply to the statement of objections, p. 63.

418 LGF has [...]* in the supplementary contract to the joint holding agreement of 20 July 1971, signed on 14 June 1999 between Albin Michel and Hachette Livres, providing that [...]*.

419 Flammarion has a preferential agreement with Actes Sud (answer to question 1 in the "Pocket-format Rights" request for information sent on 7 October 2003).

420 Answer to question 1 in the "Pocket-format Rights" request for information sent on 7 October 2003.

421 Answer to question 1 in the "Pocket-format Rights" request for information sent on 7 October 2003.

422 And in which VUP has a 25% stake.

423 Le Dilettante is a publisher which is distributed by Flammarion but does its own marketing.

424 In the list of the 163 best-selling general literature pocket formats.

425 In the list of the 163 best-selling general literature pocket formats.

426 "All the anticipated revenues" covers not only the income generated by sales of large-formats and pocket-formats, plus commissions (as a percentage of the list price exclusive of tax) received from marketing and distribution, but also any profits generated by spreading the fixed costs of marketing and distribution over a greater volume of works.

427 The Lagardère group operates in the market for the purchase of advertising space through its advertising sales agencies Lagardère Active Publicité, for the broadcast media, and Interdeco, for the print media.

428 The Lagardère group has 100% control of the radio stations Europe 1, Europe 2 and RFM, and has indirect holdings in MFM and Voltage FM. It has extensive holdings in the French-language press: 100% of La Provence, Corse Matin, Nice Matin, TV Hebdo, Var Matin, Elle, Isa, Paris Match, Parent, Jeune & Jolie, Pariscope, Photo, Première, L'Echo des Savanes, France Dimanche, Ici Paris, and Télé 7 Jours; 80% of Entrevue; 60% of Le Journal du Dimanche; 42% of Marie Claire, Cosmopolitan, etc.

429 " Livres à fond les poche ", Le Vif-L'Express, 14 February 2003.

430 Notification, p. 304.

431 See page 7 of the brochure presenting the Lagardère group Repères 2002, included in annex 6 to the notification.

432 Hachette Livre owns 35% of J'ai Lu, a 65% subsidiary of Flammarion active in the markets for the sale of pocket-formats.

433 Note on the contractual relations between authors and publishers drawn up by Actes Sud and received by the Commission on 4 August 2003.

434 Answer to question 50 in the "Publishers B" request for information sent on 16 April 2003 and to question 79 in the "Publishers A" request for information sent on 16 April 2003.

435 Answer to question 51 in the "Publishers B" request for information sent on 16 April 2003.

436 Answer to question 51 in the "Publishers B" request for information sent on 16 April 2003.

437 Literary agents are also suppliers in the secondary market for the purchase of pocket-format rights in general literature titles, but since they are natural persons their countervailing power is even less than that of a publisher.

438 Thus Amélie Nothomb.s first book came out in a Seuil series, but her other works have been published by LGF. Similarly, Bernard Werber and Stephen King have books published in pocket format by J'ai Lu and LGF, and J.R.R. Tolkien by LGF and Pocket.

439 Notification, p. 305.

440 Notification, p. 308.

441 Notification, p. 309.

442 Paris Court of Appeal, 11 March 1993, France Loisirs v Grand Livre du Mois and Others, BOCCRF, 26.12.1993, p. 104.

443 Answers to question 37 in the .Marketing/Distribution. request for information sent on 16 April 2003.

444 Certain competitors of the merged entity did also supply estimates of market shares per class (see the .Methodology. paper attached as Annex I).

445 Class 1 bookshops mainly.

446 Answers to question 78 in the "Publishers B" request for information sent on 14 April 2003.

447 The Commission's investigation shows that the choice of distributor is influenced by that of marketer (answers to the "Publishers B" request for information sent on 14 April 2003).

448 Hachette Livre's sales of academic and professional works are not shown, as they are classified under the school segment.

449 Answer to question 43 in request for information 14.685.

450 Contrary to what the notifying party maintains in its reply to the statement of objections, MDS is included in this calculation; more specifically, the calculation comprises the following marketers: Hachette, VUP, Gallimard, Flammarion, Seuil, MDS, Diff-edit, Dilisco, Sofédis, Tech & Doch, Etai, Vilo, Casteilla, Les Belles Lettres and Lavoisier.

451 See notification, p. 235.

452 Annex 1.

453 Reply to the statement of objections, p. 77.

454 Answer from Librairies Decitre to question 22 in the "Dealers" request for information sent on 14 April 2003.

455 Answers to questions 37 to 40 in request for information 10.069; the fact that several customers switched form VUP to Hachette at the same time, as the notifying party points out, does not change anything with respect to the scale of the rivalry illustrated by these transfers.

456 Answers to questions 37 to 40 in request for information 10.069.

457 Answers to question 81 in request for information 14.685.

458 See, for example, VUP's answer to question 35 in request for information 14.685.

459 On the basis of the amounts marketed for own account and on behalf of outsiders, the HHI is more than 1 850, which represents a large increase of 850.

460 According to the SLF, 31% of its member bookshops generate more than 50% of their turnover with Hachette and VUP (answer to the "Dealers" request for information sent on 14 April 2003).

461 This was clearly explained in the statement of objections, e.g. at paragraph 136.

462 Answer to the "Dealers Phase II" request for information sent on 11 July 2003.

463 I.e. where a publisher does not want the marketing service to cover a particular class (e.g. bookshops), but only the others.

464 Answers to questions 66 et seq. in the "Publishers B" request for information sent on 14 April 2003. In the question "Please indicate the basis on which publishers choose their marketers. Please place the following criteria for the choice of a marketer by a publisher in order of importance" the criterion "Levels of customer covered, extensiveness of the network" was on average ranked highest (56% of the publishers that replied put extensiveness first or second). The Lagardère group had also explained (see notification, p. 434) that although it was in deficit, its LDS activity was being maintained in order precisely to offer a complete service to its publisher customers.

465 LDS's and La DIL's sales to class 3 outlets are included in "Wholesalers".

466 Virtually all academic, professional and school publishers have their own vertically integrated operations for marketing to bookshops.

467 Answer to question 3 of the supplementary questions sent on 2 October 2003; of the more than 120 000 Seuil books sold by DNL in 2002, 33% were accounted for by 10 titles, and 60% by 35 titles.

468 Answer to question 4 in request for information 19.897.

469 Answers to question 2 in the "Bookshops" and "Hypermarkets" requests for information sent on 11 July and 11 August 2003 respectively. See also the Xerfi 2002 study on bookshops and book distribution.

470 Regarding the profile of works in hypermarkets, François Rouet said in Le livre, mutations d'une industrie en crise, la Documentation Française, Paris, 2000, p. 209: "The rule is to promote the desire to buy by systematically putting forward works and titles with a strong media impact, even if one doesn't always expect large turnovers from this preferential treatment. To be displayed on hypermarket shelves, books do not just need to be the subject of intensive advertising or to be hardwearing like certain children's series, or again interchangeable, like certain romantic or crime series; they must also be aimed at a sufficiently large readership and priced acceptably".

471 For instance, Éditions du Rocher has 7 books in the Ipsos top 1 500 general literature best-sellers in 2002, 33% of whose sales on average were achieved through Ipsos's mass distribution channel; Le Cherche Midi has 7 books in the Ipsos top 1 500, 32% of whose sales were achieved through the mass distribution channel; Lonely Planet has 10 books in the Ipsos top 1 500, 30% of whose sales were achieved through the mass distribution channel; Odile Jacob has 11 books in the Ipsos top 1 500, 25% of whose sales were achieved through the mass distribution channel; Minuit has 4 books in the Ipsos top 1 500, 40% of whose sales were achieved through the mass distribution channel, etc.

472 These are calculated on the basis of Ipsos's "specialist independent bookshops and large-scale retail outlet" channels (e.g. FNAC).

473 In the Ipsos top 1 500 large-format books, 399 titles achieved more than 50% of their sales through the Ipsos "mass distribution" channel. In the Ipsos top 5 000 pocket-format works, 1 546 titles achieved more than 50% of their sales through the same channel.

474 Answers to question 48 in the "Marketing/Distribution" request for information sent on 20 June 2003 and to question 74 in the "Publishers" request for information sent on 17 July 2003. For the 15 large-format titles most sold by Gallimard, Albin Michel, Flammarion and Seuil, about 30% of sales on average were made in hypermarkets and wholesalers.

475 Among the publishers marketed by Gallimard, Éditions du Rocher has 7 books in the Ipsos top 1 500 general literature best-sellers in 2002, and the 7 books accounted for 12% of turnover; Le Cherche Midi has 7 books in the top 1 500, and the 7 books accounted for 12% of total turnover; Lonely Planet has 10 books in the Ipsos top 1 500, and the 10 books accounted for 16% of total turnover. Among the publishers marketed by Seuil, Odile Jacob has 11 books in the top 1 500, and the 11 books account for 41% of its total turnover; Minuit has 5 books in the top 1 500, and the 5 books account for 55% of its total turnover; Milan has 9 books in the top 1 500, and the 9 books account for 26% of its total turnover.

476 Notification, Annex 29.

477 See the record of the meeting between the representative of Odile Jacob and Commission staff on 16 July 2003.

478 "A book is one of those goods whose quality (in the general sense of the term) is known through consumption, i.e. after it has been "experienced" or read. This is not the only characteristic of a book, but it shows that information plays an important role here. All individuals are different in their consumption of books, through the subjectivity of their appraisals and the ways they use the product. There is both moral hazard and adverse selection in the demand for books". Report No 1 by Marc Ivaldi, 24.6.2003, p. 1.

479 It will be noted that French-language publishers produce more than 40 000 new works a year and that there are more than 350 000 titles available. At the start of the new literary year in September 2002, more than 660 novels were placed on the market. It is therefore not credible to assume that the average consumer could be aware of more than a minute fraction of this production.

480 Answer to question 2 of the questions sent on 21 October 2003 to certain marketers.

481 To the question sent out on 21 October 2003 which asked "To what extent do publishers which are your customers for marketing services require that you market them to hypermarkets and wholesalers?" the marketers consulted replied unanimously that "It's essential for them: I could not keep those customers if I did not supply marketing services for sales to hypermarkets and wholesalers".

482 The statement of objections clearly indicates, however, for example in paragraphs 487 and 533, that the merged entity would be able to reduce the shelf-space of publishers which are not marketed by it and therefore make it more difficult for marketers to access sales channels.

483 This is what those most affected anticipate: "We think the new group, Hachette-VUP, will try to increase its sales at E. Leclerc to the detriment of other publishers, through extra discounts and greater participation in advertising where the works on their list are concerned". (Leclerc's answer to question 17 in the "Hypermarkets/Supermarkets" request for information sent on 11 August 2003).

484 See, for example, notification pp. 275 and 417-420.

485 Answer to question 35 in the "Marketing/Distribution" request for information sent on 20 June 2003.

486 Flammarion, for instance, explains that structures are optimised in terms of number of visits and number of sales outlets visited and that their prospection capacity is limited: "the principal factor of saturation is the number of new titles to be sold by representative, weighted by the type of works." (answers to question 29 in the "Marketing/Distribution" request for information sent on 20 June 2003).

487 A number of marketers explain that there is no reserve of skills available on the employment market and so some time is needed for training before it is possible to increase the size of a marketing team (see answers to question 35 in the "Marketing/Distribution" request for information sent on 20 June 2003).

488 The statement of objections refers to these barriers in paragraph 526.

489 Answers to question 53 in the "Marketing/Distribution" request for information sent on 20 June 2003, and answers to question 69 in the "Publishers B" request for information sent on 14 April 2003; the other points mentioned by publishers concern the clauses in contracts which set time limits for changing service provider, and the complexity of changing supplier.

490 To preserve the confidentiality of information supplied to the Commission the main marketers/distributors competing with Hachette and VUP are here shown as "marketer 1" to "marketer 4"; the graph shows the marketers' customers' sales, in list prices exclusive of tax, by class of dealer. It will be seen that Hachette's and VUP's customers achieve a large proportion of their sales through hypermarkets and wholesalers.

491 Salesrooms (comptoirs de vente) are primarily shops reserved for people in the trade, with an attractive presentation of selected works, where open days can be held.

492 9 salesrooms, according to the answers to question 41 in the "Marketing/Distribution" request for information sent on 20 June 2003; Hachette Livre and VUP have 7 and 10 respectively.

493 See the answers to questions 41, 45, 46 and 80 in the "Marketing/Distribution" request for information sent on 20 June 2003.

494 Thus, according to the results of the Commission's investigation, distribution to class 1 and 2 bookshops throughout the French-speaking area of the Community requires two teams of at least 25 representatives each and a minimum turnover of about euro 50 million, i.e. an estimated cost per representative of roughly 4%-5% of the list price exclusive of tax and net of returns. As regards marketing services for sales to hypermarkets, the prerequisites are a turnover of around euro 30 million and sufficient commercial reputation to be listed by central purchasing facilities.

495 See record of the meeting between the representatives of Albin Michel and Commission staff of 25 June 2003.

496 Albin Michel accounts for only [0-10]% of the merged entity's total marketing business (in the French-speaking area of the Community) and [5-15]% of the distribution services performed by the merged entity for outside publishers (these percentages rise to [0-10]% and [10-20]% respectively in the case of distribution services provided by the merged entity for Albin Michel).

497 The Commission's investigation shows in this respect that the current customers of the main independent marketers (Gallimard, Seuil, Flammarion and Diff-Edit) do not achieve sufficient turnover to be able to do their own marketing. According to the notifying party, these publishers' sales are less than euro 49 million, which, according to the answers to question 26 in the "Marketing/Distribution" request for information sent on 20 June 2003, is the minimum critical size required for financing class 1 (euro 21 million) and class 2 (euro 28 million) marketing teams.

498 Answers to question 4 in request for information 23.229.

499 Answer to question 46 in request for information 14.685.

500 Information supplied by Prisme.

501 It should be mentioned that a decision by the merged entity to leave Prisme would automatically increase the transport costs invoiced to booksellers, since the latter incur 25% of the cost of dispatches and 100% of the costs of returns. On average, these costs are currently 1.8% of the list price exclusive of tax for a customer bookseller and, because the price per kilo transported falls with increasing weight, are inversely proportional to the size of the outlets.

502 Transport costs are generally paid by retailers. For a publisher, having high transport costs therefore means less possibility of offering a high trade discount.

503 MDS puts the figure at over [...]% of the volume handled (answer to question 74 in the .Marketing/Distribution. request for information sent on 16 April 2003).

504 Answer given by Prisme to question 22 in the request for information sent to it on 24 April 2003.

505 As the SLF states in its answer to the "Dealers" request for information sent on 14 April 2003, "The fact that Hachette undertakes to be a full member of this platform would not resolve the problem caused by concentration in the transport field: such an undertaking would inevitably be for a period that by definition was too short in relation to what was at stake; even if the undertaking was given and honoured, it would in no way change the fact that a single distributor made up of VUP and Hachette would be in a position to impose its own distribution conditions (even if use of the Prisme platform was guaranteed) in such a way that they would upset the entire distribution system to the disadvantage of booksellers. The constraints applied in this way would completely distort the rules of competition."

506 The opposite does not necessarily apply.

507 See, for example, notification pp. 275 and 417-420.

508 The notifying party implicitly recognises them in its analysis of the economic outcome of accepting a new customer (reply to the statement of objections, p. 120). In addition, there are forms of subcontracting that might be possible, as the notifying party points out; but they relate only to specific and limited distribution functions, such as long-term storage, or else involve changes so fundamental that they call the entire architecture of the distribution mechanism into question, and are consequently improbable, an example being [...]* (answer to question 54 in request for information 14.685, and in particular the feasibility study concerning Albin Michel).

509 Answers to questions 5 and 6 in the additional requests for information sent to distributors on 2 October 2003, and to request for information 19.897. Costs are estimated by all the players at around euro 5 million for an increase in capacity of only 20% that would take at least twelve months to introduce.

510 Reference was made to these barriers in paragraph 526 of the statement of objections.

511 Answers to question 22 in the .Marketing/Distribution. request for information sent on 20 June 2003 and to question 83 in request for information 14.685. The answers show that Hachette's and especially VUP's separate average costs before the merger are lower than those of their main competitors. The Commission also collected figures from most of the main distributors (question 70 in the "Marketing/Distribution" request for information sent on 20 June 2003 and question 107 in request for information 14.685), and this enabled it to estimate a production function (memo of 9 September 2003 from the independent expert Marc Ivaldi). On average over the sample the level of returns to scale is 2.35. In other words, if production increases by 1%, costs increase by only 0.43%, which is characteristic of a technology with rising returns to scale, and indeed very high returns to scale. According to these calculations, Hachette and VUP before the merger each have marginal costs lower than those of their main competitors.

512 Answers to questions 58 to 59 and 67 to 69 in the "Marketing/Distribution" request for information sent on 20 June 2003, and answers to questions 97 and 98 in request for information 14.685.

513 Answers to questions 69 and 74 in the "Marketing/Distribution" request for information sent on 20 June 2003, and answers to questions 37 to 39 in request for information 14.685.

514 New entries into the market also occurred as a result of publishers buying shareholdings in small local distributors.

515 Answers to questions 4 and 5 of request for information 23.229. It will be seen quite unequivocally that between two publisher customers of Hachette Livre and VUP, the one with the higher turnover for a given service is offered the lower rate.

516 The conglomerate effects, which derive from the merged entity's media business, are confined to the markets for the sale of general literature.

517 Answers to questions 10, 34 and 35 in the "Dealers Phase I" request for information.

518 This allegation is contested by Albin Michel.

519 Answers to questions 56 and 82 in the "Publishers B" request for information given by Actes Sud, Juris-Classeur and Play Bac, and the answers to questions 34 and 35 in the "Dealers Phase I" request for information given by the bookshops Alinéa, Luxembourg; À Livre Ouvert and Alsatia, Mulhouse; Auchan, Villeneuve d.Ascq; La B.L., Saint-Jean d'Angély; Bisey, Mulhouse; Bouquin Bouquines, Schirmeck; Brioson, Saint-Lô; Broglie, Strasbourg; Camponovo and Castéla, Toulouse; Colobri, Montreuil; Delbert and Le Divan, Paris; Ducher, Verdun; Edit et Moi, Ploeruel; Ernster, Luxembourg; Frimaudeau, La Ribotière; E. Guy, Lons-le-Saunier; Hartmann, Colmar; Gibert Jeune, Paris; Kléber, Strasbourg; Lecuru, Liancourt; Loddé, Orléans; du Lycée, Beziers; du Lycée, Perpignan; Maison de la Presse, Saint-Dié-des-Vosges; MDP, Perpignan; Mollat, Bordeaux; Mot à Mot, Rodez; Naud, Les Andelys; Ombres Blanches, Toulouse; Page et Plume, Limoges; Le Passage, Alençon; Pleine Page, Sélestat; Privat, Nantes; Sarion, Lyon; de l.Université, Poitiers; Vincenti, Haguenau; Zunino, Saverne; de Casino, FNAC, FNAC Belgique, GALEC (Leclerq group) and Monoprix. See also the answers to the "Authors" request for information sent on 30 July 2003 given by the authors Tahar Ben Jelloun, Françoise Chandernagor, Michael Cunningham, Didier Debord, Hans Magnus Enzensberger, Hubert Reeves, Jean Lacouture, Jean-Christophe Rufin and Michel Salzedo, and by Catherine Camus, who holds the rights of the author Albert Camus. In its answer to question 35 in the "Dealers Phase I" request for information, the Syndicat de la Librairie Française says there is a danger that a large part of the range of titles on offer in pocket format may disappear, because many titles in the various series published by Hachette Livre may not be reprinted.

520 Answer given by Claire Lubac to the .Authors. request for information sent on 30 July 2003.

521 Answer given by the Vincenti bookshop to question 35 in the "Dealers Phase I" request for information.

522 L'Édition depuis 1945 (La Documentation française), page 638.

523 Answers to question 100 in the .Publishers A. request for information given by Gallimard, La Martinière and Seuil, and answers to question 82 in the "Publishers B" request for information given by Brepols, Casteilla, Dargaud, De Boeck, Desclée de Brouwer, Ellipses, Gründ, Imprimerie Nationale, Lavoisier, Masson, Michelin, Micro Application, Milan, Odile Jacob, Payot & Rivages and Le Rocher. See also the minutes of the meeting with Albin Michel on 25 June 2003, and the answer given by the Kléber bookshop to question 34 in the "Dealers Phase I" request for information.

524 Answers given by Gallimard, La Martinière and Seuil to question 100 in the .Publishers A. request for information sent on 16 April 2003.

525 This analysis is in line with the model of the "discriminating monopoly" developed by economists (e.g. Pigou [1910], "Producers" and consumers' surplus, Economic Journal, 20, pp. 358-70), in which a monopolist creates false distinctions between identical products in order to exploit differences in demand among consumers. The asymmetry of information that characterises the quality of books makes it easier for a monopolist to offer products that are only seemingly different.

526 See in particular the article http://acrimed'samizdat.net/article.php3?id_article=861.

527 Minutes of the meeting between Commission staff and the Syndicat de la Librairie Française on 15 July 2003.

528 Of the booksellers surveyed by the Commission, 83% said that it had happened more than once that a publisher had sent them parcels containing more copies of a book than they had ordered. Of the booksellers surveyed 41% said they returned any unordered books immediately, while 46% said they kept them and stacked them on their shelves (answers to questions 6 and 7 in the .Dealers Phase II. request for information).

529 Of the booksellers surveyed, 68% said it was already impossible for them to take part in marketing operations mounted by publishers other than Hachette Livre and VUP (59% were not able to take part even in promotion campaigns organised by Hachette and VUP). 16% said that after the merger they would be able to take part in all operations organised by Hachette and VUP, but that this would limit their ability to take part in operations organised by competing publishers. Only 6% said they had sufficient space and funds to take part in all marketing operations mounted by all publishers.

530 The merged entity would account for [55-65]% of sales of general literature, all formats and all categories of dealer combined.

531 Under Hachette Livre's general conditions of sale, [...]*.

532 Answer to question 1 in request for information 17.604. [...]*.

533 Some bookshops have a very low overall margin, of the order of 2% of their sales, and have a large amount of money tied up in stocks. Reducing discounts or tightening payment facilities would have a negative impact on these outlets' liquidity, and would leave them financially dependent on their distributor, who is responsible for collecting payment.

534 See the analysis of secondary markets for club rights set out above.

535 If they do not themselves choose books to buy, members must buy the book selected for them every three months by France Loisirs, which may explain the impressive performance of books published by France Loisirs in the best-seller list: the notifying party states that four of the five best selling general literature titles in large format in 2002 were marketed by France Loisirs.

536 According to the notifying party the total publishing market in French-speaking Switzerland amounts to euro 113 million, which compares with a figure of euro 3 220 million for the French market (answer to question 2 in request for information 10.069).

537 This is so especially as 80% of French-language books sold in Switzerland are published in France (see the article "Prix du livre : un marché de funambules").

538 Based on estimates of the size of the Lagardère group's market provided in the answers to request for information 9.755.

539 The Commission arrived at these figures by asking market players to estimate the size of the French-language market and to provide their turnover figures for the sale of books to dealers at all levels in 2001.

540 The Commission arrived at these figures by asking market players to estimate the size of the French-language market and to provide their turnover figures for the sale of books to dealers at all levels in 2001.

541 Jérôme Foncel and Marc Ivaldi, " Evaluation Econométrique des Effets de la Concentration Lagardère- VUP sur le Marché du Livre de Littérature Générale ", final version, revised and expanded, September 2003. The study was carried out for use in the present case, on the basis of data provided by the notifying party, namely figures for sales and book prices assembled by the Ipsos survey institute supplemented by data from Sofres. The entire sample contains a total of 9 566 observations. The Ipsos data concern the 5 000 best sellers in pocket format in 2002, and the 1 500 best sellers in large format. The sample covers a big part of the market in 2002, since these books accounted for over 96% of pocket-format sales and 44% of large-format sales. The econometric study is based on a Bertrand competition model in which consumer preferences are represented by a nested logit model. The estimation used the non-linear three-stage least squares method.

542 For want of reliable data on the discounts given to dealers.

543 To increase the robustness of the results Foncel and Ivaldi calculated a confidence interval by the bootstrap method. This allows a possible range of price increases to be worked out by simulating the merger a large number of times. A 95% confidence interval means that there is a 5% possibility of error in the conclusion that the merger will result in a price increase in the range arrived at. The study is particularly robust by reason of the very large number of observations, the stability of the various parameters estimated, the high degree of statistical power of the tests provided, and the simulation of a confidence interval for the calculation of a price increase.

544 See the competition analysis of the effects on the primary and secondary markets for rights.

545 "Control of its marketing and distribution puts a publisher in a better position to ensure that its products are available and visible in bookshops", answer given by FNAC to question 22 in the "Dealers Phase I" request for information.

546 On this point, [notifying party internal paper]*.

547 Shares of the French, Belgian and Luxembourg markets, by value, in 2001.

548 Shares of the French, Belgian and Luxembourg markets, by value, in 2001.

549 Answers to question 100 in the .Publishers A. request for information given by Gallimard, La Martinière and Seuil, and answers to question 82 in the .Publishers B. request for information given by Brepols, Casteilla, Dargaud, De Boeck, Desclée de Brouwer, Ellipses, Gründ, Imprimerie Nationale, Lavoisier, Masson, Michelin, Micro Application, Milan, Odile Jacob, Payot & Rivages and Le Rocher. See also the report of the meeting between Commission staff and representatives of Albin Michel on 25 June 2003, and the answer to question 34 in the "Dealers Phase I" request for information given by the Kléber bookshop in Strasbourg.

550 Over half of booksellers believe that the merged entity could use its strength in schoolbooks and reference books to encourage them to buy its books in other product categories (answers to question 15 in the "Dealers Phase II" request for information sent on 11 July 2003).

551 Notification, Annex 29. Answer to question 50 in request for information 14.685 and answers to question 12 in the "Marketing and Distribution" request for information sent on 26 June 2003. See also the answer given by Actes Sud to the "Publishers" request for information sent on 17 July 2003.

552 Answers to question 19 in the .Dealers Phase II. request for information.

553 The Lagardère group operates in the market for the purchase of advertising space through its advertising sales agencies Lagardère Active Publicité, for the broadcast media, and Interdeco, for the print media.

554 The Lagardère group has 100% control of the radio stations Europe 1, Europe 2 and RFM, and has indirect holdings in MFM and Voltage FM. It has extensive holdings in the French-language press: 100% of La Provence, Corse Matin, Nice Matin, TV Hebdo, Var Matin, Elle, Isa, Paris Match, Parent, Jeune & Jolie, Pariscope, Photo, Première, L'Echo des Savanes, France Dimanche, Ici Paris, and Télé 7 Jours; 80% of Entrevue; 60% of Le Journal du Dimanche; 42% of Marie Claire, Cosmopolitan, etc.

555 It should be pointed out that all the promotion activities planned when a book appears are listed and described in the notes sent by publishers to dealers when initial orders are being taken.

556 Answers given by Gründ and Séléction du Reader's Digest to question 82 in the .Publishers B. request for information. For example, books with large print runs published by outside publishers were ordered by Relay only much later than books published by the Hachette group, and the orders were a great deal smaller; an outside publisher was told the terms of purchase of promotional space in Relay shops only after repeated enquiries (see the report of the meeting between Commission staff and representatives of Actes Sud on 15 July 2003, and the memorandum on contractual relations between authors and publishers drawn up by Actes Sud and sent to the Commission on 4 August 2003).

557 See the answer to question 152 in the .Publishers A. request for information given by Flammarion, and the answers to question 82 in the .Publishers B. request for information given by Actes Sud, Bayard Presse, Belin, Bréal, Desclée de Brouwer, Dupuis, Ellipses, Gründ, Juris-Classeur, Payot & Rivages and Sélection du Reader's Digest.

558 Notification, p. 323.

559 See Annexes 106 and 107 of the notification, and the marketing and distribution contracts assembled during the Commission's investigation.

560 Answers to questions 5 and 8 in the "Dealers" request for information sent on 14 April 2003.

561 Answers to question 43 in the "Publishers" request for information sent on 17 July 2003.

562 At the meeting between Commission staff and representatives of Éditions Albin Michel on 25 June 2003, the Albin Michel representatives told the Commission of the difficulty Albin Michel had had in having its school workbooks marketed "properly" by its marketer, Hachette Livre, given that Albin Michel was in direct competition here with books published by Hachette ("Hachette would be in a position to adopt behaviour designed to evict its competitors, for example by using quality discounts to exclude other publishers from bookshops; in the same way, it is easy to oust competitors from supermarkets (an example is provided by the eviction of Magnard for school workbooks)").

563 Answer given by Juris-Classeur to question 70 in the .Publishers B. request for information sent on 14 April 2003.

564 Notification, Annex 16 (Philippe Moatti, " La filière du roman: de la passion à la rationalité marchande? ", in Cahiers de l'économie du livre, No 7, March 1992).

565 As already indicated, marketed and distributed publishers are most often also tied to their marketer and distributor by other cooperation agreements or practices, such as the sale of secondary rights for pocket-format publication. Financial mechanisms, and notably the management of returns, increase this dependence.

566 In 2002, sales outlets covered by Hachette Livre marketing services on behalf of Albin Michel accounted for about 60% of sales of copies of Albin Michel's 20 best-selling titles; Albin Michel states that it lives on its 10 best sellers (see the report of the meeting between Commission staff and representatives of Albin Michel on 25 June 2003).

567 This is especially true of best-selling authors, and Albin Michel has specialised in the publication of best-sellers.

568 These publishers all have a turnover of between euro 100 million and euro 250 million, with the exception of the merged entity, whose hypothetical combined turnover in 2001 amounted to euro [...]* million.

569 The merged entity would represent an average of about 25% of the sales of each independent publisher in the sale of books (of all categories) to wholesalers.

570 More than 50% of the booksellers surveyed by the Commission said that after the merger the merged entity would be in a better postion to incite them to buy the books it markets in preference to those of competing publishers or marketers (answers to question 9 in the "Dealers Phase I" request for information).

571 Many booksellers say they fear a reduction in their discounts, or an alignment of VUP's discounts on those of Hachette Livre (answers to questions 10, 34 and 35 in the "Dealers Phase I" request for information). Some of them cite the example of the takeover of the marketing of Albin Michel's books by Hachette Livre, when the discount given to dealers on Albin Michel's books fell by an average of 1%. But this allegation is contested by Albin Michel.

572 Answers to question 11 in the "Dealers Phase I" request for information.

573 Answers to question 11 in the "Dealers Phase I" request for information.

574 Some bookshops have a very low overall margin, of the order of 2% of their sales, and have a large amount of money tied up in stocks. Reducing discounts or tightening payment facilities would have a negative impact on these outlets. liquidity, and would leave them financially dependent on their distributor, who is responsible for collecting payment. Bookshops that sell school textbooks would be even more sensitive to any tightening of conditions of payment, because school books bought in mid-June are sold only when schools resume at the end of summer. The operating balance of many of them depends on the continuation of this business, something which Hachette Livre, with its easily dominant position in school textbooks, would be in a position to threaten.

575 Some publishers surveyed by the Commission fear that the merger would result in the eviction of a number of bookshops, especially large specialised bookshops (answers given by Ellipses, Lavoisier and Odile Jacob to question 82 in the .Publishers B. request for information).

576 FNAC states that its capacity to withstand any reduction in discounts imposed on it by the merged entity would in reality be fairly small (answer to question 11 in the "Dealers Phase I" request for information).

577 According to its own estimates, the merged entity, through the Le Furet du Nord, Virgin/Extrapole and Relay network of bookshops, would have a market share [0-10]% behind FNAC.s [10-20]%.

578 Answer to question 11 in the "Dealers Phase I" request for information.

579 As for the second point, it will be obvious that the Commission does not share the notifying party's analysis.

580 See the judgment of the Paris Court of Appeal (25th division, section B), 24 November 2000.

581 On the basis of the estimates of size of market provided by the Lagardère group in its answers to request for information 9.755.

582 The Commission arrived at these figures by asking market players to estimate the size of the French-language market and to provide their turnover figures for the sale of books to dealers at all levels in 2001.

583 Gallimard's market share is largely due to the Harry Potter books by J.K. Rowling, which it publishes in French.

584 Market shares by value for France, Belgium and Luxembourg in 2001.

585 See the competition analysis of sales of books of general literature.

586 The Commission arrived at these figures by asking market players to estimate the size of the French-language market and to provide their turnover figures for the sale of books to dealers at all levels in 2001.

587 Market shares by value for France, Belgium and Luxembourg in 2001'.

588 Based on estimates of the size of the Lagardère group's market provided in the answers to request for information 9.755.

589 The Commission arrived at these figures by asking market players to estimate the size of the French-language market and to provide their turnover figures for the sale of books to dealers at all levels in 2001.

590 See the competition analysis of sales of books of general literature. Shares of the French, Belgian and Luxembourg markets, by value, in 2001.

591 See the competition analysis of sales of books of general literature.

592 The Commission arrived at these figures by asking market players to estimate the size of the French-language market and to provide their turnover figures for the sale of books to dealers at all levels in 2001.

593 Market shares by value for France, Belgium and Luxembourg in 2001.

594 See pp. 164 et seq. and 352 et seq. of the notification and the answers given by Flammarion to questions 42-63 in the .Publishers A. request for information. See also the answers given by Albin Michel, Belin, Bréal and Flammarion to questions 76-78 in the .Publishers B. request for information and the records of the telephone conversations Commission staff had with Delagrave (Flammarion) and Magnard (Albin Michel) on 23 September 2003. See also the answer given by the French Ministry of Youth, National Education and Research to the request for information that was sent to it.

595 Notification, p. 167.

596 Answer to question 50 in request for information 14.685.

597 It sometimes happens as well that a publisher decides to launch a new textbook without a change of syllabus, e.g. where an existing textbook has become obsolescent or can be improved or where educational sales representatives pass on information about the existence of an expectation on the part of teachers.

598 Notification, pp. 351 and 352, supplemented by the answer to question 23 in request for information 10.069.

599 See the answers given by Albin Michel and Flammarion to questions 49, 57 and 60 in the "Publishers A" request for information and the records of the telephone conversations Commission staff had with Magnard (Albin Michel) and Delagrave (Flammarion) on 23 September 2003.

600 See the answers given by Albin Michel and Flammarion to questions 80-82 in the "Publishers A" request for information and the answer given by Belin to question 52 in the "Publishers B" request for information. See also the records of the telephone conversations Commission staff had with Magnard (Albin Michel) and Delagrave (Flammarion) on 23 September 2003. For example, in the event of a revision of the French syllabus in lower secondary schools, the first-form textbook will be published the first year, the second-form one in the same series the following year, and so on, so that the success of a series established by the first-form textbook has a knock-on effect.

601 Answer to question 24 in request for information 10.069.

602 See the records of the telephone conversations Commission staff had with Magnard (Albin Michel) and Delagrave (Flammarion) on 23 September 2003 and the answers given by Belin and Bréal to question 34 in the "Publishers B" request for information.

603 Answers given by school textbook publishers to question 76 in the "Publishers Phase II" request for information sent on 17 July 2003.

604 See the answers given by Albin Michel and Flammarion to question 57 in the "Publishers A" request for information and the records of the telephone conversations Commission staff had with Magnard (Albin Michel) and Delagrave (Flammarion) on 23 September 2003.

605 Record of the telephone conversation between Commission staff and Delagrave (Flammarion) on 23 September 2003.

606 Notification, pp. 353 and 354.

607 Record of the telephone conversation between Commission staff and Delagrave (Flammarion) on 23 September 2003.

608 Notification, Annex 20.

609 Record of the telephone conversation between Commission staff and Delagrave (Flammarion) on 23 September 2003.

610 See the memo sent by Norton Rose on 15 January 2003.

611 Record of the telephone conversation between Commission staff and Delagrave (Flammarion) on 23 September 2003.

612 Records of the telephone conversations between Commission staff and Magnard (Albin Michel) and Delagrave (Flammarion) on 23 September 2003.

613 Answers given by the notifying party to request for information 10.069.

614 The costs consist mainly in the production or rental of an index of teachers, the sending of a free specimen to the prescriber and the associated carriage costs.

615 Answer given by the notifying party to question 115 in request for information 14.685.

616 It is rare for a publisher to publish only one book for a particular subject and/or class. Generally speaking, it must offer all the textbooks on a subject for the entire educational cycle concerned.

617 Answers given by school textbook publishers to question 52 in the "Publishers" request for information sent on 16 April 2003.

618 Annex 71 to the notification shows in this respect that the impact of producing a school textbook on cashflow is very weak or even negative the first year and increases only gradually over the textbook's lifetime.

619 See the answer given by Flammarion to question 81 in the .Publishers A. request for information and the records of telephone conversations Commission staff had with Delagrave (Flammarion) and Magnard (Albin Michel) on 23 September 2003. New syllabuses are drawn up in France every four years or so, depending on the subject and educational level.

620 Notification, Annex 73.

621 Answer given by the notifying party on 22 July 2003.

622 See the answer given by Bréal to question 39 in the "Publishers B" request for information and the records of the talks with De Boeck on 29 April 2003 and Delagrave on 23 September 2003.

623 See the answers given by Albin Michel and Flammarion to question 62 in the "Publishers A" request for information and the answer given by Bréal to the "Publishers Phase II" request for information. See also the record of the telephone conversation between Commission staff and Delagrave (Flammarion) on 23 September 2003.

624 Answers to question 39 in the "Publishers B" request for information and to question 80 in the "Publishers Phase II" request for information.

625 See the answer given by Flammarion to question 62 in the "Publishers A" request for information and the answer given by Bréal to question 39 in the "Publishers B" request for information. See also the record of the telephone conversation between Commission staff and Delagrave (Flammarion) on 23 September 2003.

626 Answer given by Belin to question 79 in the "Publishers Phase II" request for information.

627 See the record of the telephone conversation Commission staff had with Delagrave (Flammarion) on 23 September 2003 and the answer given by Bréal to question 39 in the "Publishers B" request for information.

628 Hachette Livre and VUP are already leading members of the Savoir-Livre association, which was formed in 1985 by Belin, Bordas, Hachette, Hatier, Magnard and Nathan for the purposes of encouraging studies and meetings on education and educational means and taking part in discussions on access to reading and culture. The association has the ear of the authorities, in particular the Ministry of Education, but is not open to all school textbook publishers (see the answer given by Editions Bréal to question 52 in the "Publishers B" request for information sent on 16 April 2003).

629 Answers to question 20 in the "Dealers Phase I" request for information. This picture is all the more realistic as the level of discounts granted by school textbook publishers to retailers, irrespective of their size, is, at 31.5%, already well below the average rates for other book categories, which is a reflection of the fact that, for this book category, the bookseller is not a prescriber.

630 Notification, p. 357.

631 Notification, p. 359.

632 Notification, p. 361.

633 See the answers given by Albin Michel, Flammarion and Gallimard to questions 42 and 66 in the "Publishers A" request for information and the answer given by Belin to questions 43 and 53 in the "Publishers B" request for information.

634 See the analysis of the conglomerate effects of the transaction on the markets for the sale of works of general literature by publishers to dealers.

635 See the answers given by Flammarion and Gallimard to questions 50, 64 and 65 in the "Publishers A" request for information and the answers given by Belin and Bréal to questions 33, 41 and 42 in the "Publishers B" request for information. Most of the merged entity's competitors are publishers active in school textbook publishing, apart from Gallimard, Vuibert (Albin Michel) and Playbac. According to Gallimard, its entry into this market was limited to an area close to its core business and was only possible thanks to its notoriety as a general literature publisher.

636 In this connection, see the competition analysis of the markets for the sale of school textbooks.

637 Answers given by school textbook publishers to question 85 in the "Publishers Phase II" request for information.

638 See, inter alia, the competition analysis of the markets for the sale of works of general literature.

639 Market shares in the market for dictionaries: (i) bookshops: Hachette Livre, [0-10]%; VUP, [85-95]%; (ii) hypermarkets: Hachette Livre, [5-15]%; VUP, [80-90]%; and (iii) wholesalers: Hachette Livre, [15-25]%; VUP, [70-80]%.

640 [VUP internal paper]*.

641 Notification, p. 386.

642 See, inter alia, the competition analysis of the markets for the sale of works of general literature.

643 Smaller reference works are sold mainly in (class 1) bookshops and hypermarkets.

644 See the section on the markets for the sale of dictionaries to retailers.

645 See, inter alia, the competition analysis of the markets for the sale of works of general literature.

646 Notification, p. 390. Mention may be made here of the notifying party's confusion about the merged entity's share of this market in France. On p. 390 of the notification, the market share is given as [0-10]%, whereas on pp. 218 and 219 the merged entity is described as having [10-20]% of the market.

647 Notification, p. 219.

648 Notification, p. 220.

649 Notification, p. 394.

650 Notification, p. 403.

651 Notification, p. 225.

652 Notification, p. 410.

653 Notification, p. 412.

654 Notification, p. 263, footnote on p. 602.

655 LDS is a separate entity within Hachette Livre's marketing/distribution set-up. La DIL is a former subsidiary of VUPS which, though absorbed by the latter, has retained its name to designate the teams that canvass a particular customer base. For convenience, LDS and La DIL are to be understood as referring to the wholesaling structures of Hachette Livre and VUP respectively, including their respective twin structures in Belgium covering Belgium and Luxembourg (namely AMP - and Dilibel for some sales outlets - in the case of Hachette Livre and VUPS Bénélux in the case of VUP).

656 Answer to question 6 in request for information 13.495.

657 This service to supermarkets, which is provided both in France and in Belgium, consists in the marketer or wholesaler managing every aspect of the book department of a store which considers it lacks the necessary expertise.

658 A minimum of [55-65]% according to the notifying party's reply to the statement of objections, p. 192.

659 The Commission's investigation has thus confirmed that publishers' marketers treat the Relay central purchasing facility in their statistics as a wholesaler (see inter alia the answers to question 8 in the "Marketing/Distribution" request for information sent on 20 June 2003.

660 For example, in France, sales outlets such as the Librairie H in Saint Lazare and that in Roissy Airport's Terminal F obtain supplies almost exclusively from publishers' marketers without going through the central purchasing facility (equivalent to [0-10]% of Relay retailers. total turnover). Moreover, [25-35]% of total turnover is achieved by Relay retailers which combine purchases via the central purchasing facility with direct purchases from publishers' marketers.

661 Even if other retail outlets (e.g. FNAC) also carry out a centralised listing of works, unlike the Relay central purchasing facility's manner of proceeding which involves dealing with all marketers in a centralised manner, specialised superstores are the object of an individualised marketing effort per point of sale and hence cannot be considered wholesalers.

662 Although this impacts directly on the market for the sale of books by publishers to wholesalers rather than on the market for the sale of books by wholesalers to class 3 dealers, Relay's purchasing policy is considered by publishers competing with the merged entity which also supply Relay as discriminating against them. According to them, when buying, Relay favours the works of the Hachette Livre group. Sales by the shops also reflect, so they say, discriminatory conduct consisting in .pushing. works in so far as, in the general literature field, the titles of Hachette Livre and its marketed publishers accounted for 27% of the 15 Relay best sellers in 2002 whereas, according to the Ipsos panel, they should have accounted for only 20% of the 15 best sellers overall. It cannot therefore be ruled out that this discriminatory policy may be extended to VUP's titles and may thus strengthen the vertical effect of the integration of the merged entity in the sale of books by publishers to wholesalers in the market for the sale of books by wholesalers to class 3 dealers as discussed below.

663 Notification, Annex 108.

664 The number of sales outlets they cover ranges from a few dozen in the case of the smallest of them to about 2 500 (supermarkets) in the case of the largest, DNL.

665 According to the findings of the Commission's investigation and for 2001 (see the "Methodology" paper attached as Annex I).

666 The share of these dealers' turnover from book sales accounted for by the merged entity would vary between 50% and 100%.

667 As indicated in the section on markets for the sale of works of general literature (in both large and pocket format), the Commission, through its independent economic consultant, carried out an econometric study which finds that the notified transaction would make possible unilateral effects which would result in an increase in the list prices exclusive of tax of these types of work. Knowing that the average class 3 sales outlet supplies mostly works of general literature in both large and pocket format, it is reasonable to consider that these unilateral effects might contribute to the implementation of the practice described.

668 Answers to question 12 in the "Class 3 Retailers" request for information sent on 19 September 2003.

669 Answers to question 12 in the "Class 3 Retailers" request for information sent on 19 September 2003.

670 Answer to question 50 in the "Wholesalers" request for information sent on 16 April 2003.

671 Hachette and VUP would be able to step up a commercial and financial policy aimed at developing exclusive links with a number of retailers through, for example, an additional consideration based on a percentage of turnover and end-of-year discounts or bonuses where titles marketed by LDS are stocked exclusively (see notification, Annex 76). Were they to be extended to VUP - on the basis, moreover, of a catalogue that could satisfy almost all of sales outlets' needs - such terms would further enhance exclusive links with sales outlets, such links being, according to the investigation's findings, not only possible but already in the course of being established with certain supermarkets. GALEC, the central listing facility of the Leclerc supermarket chain, confirms, for example, that "we have in effect concluded exclusivity agreements with certain suppliers (LDS and DIL, which are both wholesalers) because they are able to offer our supermarkets all of the works that are likely to be of interest to them, whatever the category of work" (see answer to question 9 in the "Hyper/Super Phase II" request for information sent on 11 August 2003).

672 Analysis of the works available in the salesrooms of Hachette and VUP (annex to question 66 in request for information 14.685) already shows that only those titles of rival publishers that are considered "must haves" or necessary additions to the range supplied are put on display by Hachette Livre and VUP (for instance, VUP, which does not produce any travel guides, puts Michelin guides on display in its salesrooms, together with best-sellers such as Seuil's La Vie Sexuelle de Catherine M. Similarly, Hachette puts on display such must-haves as the Harry Potter books, published by Gallimard).

673 Answer to question 48 in the "Wholesalers" request for information sent on 16 April 2003.

674 Answer to questions 26, 45 and 50 in the "Marketing/Distribution" request for information sent on 20 June 2003.

675 Apart from VUP's sales to the Relay central purchasing facility.

676 The [15-25]% in question can be broken down as follows: [10-20]% sold by the merged entity's publishing houses, and [0-10]% sold by publishers marketed by the merged entity.

677 Of all books sold to independent wholesalers by the publishing houses of Hachette Livre and VUP, those of Hachette Livre accounted for only [15-25]% (the remainder being accounted for by VUP), or [20-30]% if one includes sales by publishers marketed by Hachette Livre (the remainder being accounted for by VUP).

678 See the record of the telephone conversation between the Commission and a representative of Comptoir du Livre on 2 July 2003.

679 See the judgment of the Paris Court of Appeal (25th division, section B), 24 November 2000.

680 See p. 68 of the report on the investigation of practices by Hachette Livre towards DNL of the Directorate for Competition, Consumption and the Suppression of Fraud of the Île de France Region, Paris, 18 May 2000.

681 See the record of the telephone conversation between the Commission and a representative of 3 Épis on 24 June 2003.

682 Answer to question 16 in the "Wholesalers" request for information sent on 16 April 2003.

683 Answer to question 16 in the "Wholesalers" request for information sent on 16 April 2003.

684 See in this connection what was stated in the vertical effects analysis.

685 Answers to questions 26, 41, 45, 46 and 80 in the "Marketing/Distribution" request for information sent on 20 June 2003. By way of example, Gallimard's representatives state that "the presence at class 3 level of a domestic publisher necessitates a wholesaler, sales networks and sales outlets throughout France. Our group's output is nowhere near large enough to justify this type of action and investment. The investment cost alone would be enormous. But other, even more insurmountable, obstacles prevent us once and for all from considering such a possibility: the wholesaling skills, the access to Hachette's and VUPS's catalogues and the large market share of the two groups, which would constitute a strong barrier".

686 pp. 197 and 198.

687 The economic interest grouping set up by a number of independent publishers (Gallimard, Flammarion and Seuil) cannot be equated with an integrated wholesaling structure in so far as its main task is not to sell a "basket" of books suited to class 3 dealers but simply to serve as an advance logistical depot for the totality of those publishers' offerings. Nor must it be forgotten that this type of structure generates all the traditional economic inefficiencies of an incompletely integrated cooperation structure (such as higher transaction costs).

688 Unlike in the case of newspaper distribution (where all dailies are distributed irrespective of their ownership), the necessary trade-offs resulting from the way in which a wholesaling structure is organised correspond to choices between the works of the publishers involved and hence to commercial decisions having a direct impact on those publishers' profitability and their interest in participating in and contributing to such a structure.

689 See the answer to question 11 in the "Dealers" request for information sent on 14 April 2003.

690 See the answer to question 11 in the "Dealers" request for information sent on 14 April 2003.

691 See the answer to question 34 in the "Dealers" request for information sent on 14 April 2003.

692 Nor can it be excluded that the merged entity might turn to its advantage the position of strength - notably financial - which it would occupy vis-à-vis certain small sales outlets through its presence not only as a book supplier but also as a magazine distributor, so as to intensify its links with these outlets and exclude them to competing wholesalers.

693 Answer to question 50 in the "Wholesalers" request for information sent on 16 April 2003.

694 Notification, p. 407.

695 Notification, p. 402.

696 Notification, p. 225.

697 Notification, p. 393.

698 Commission notice on remedies acceptable under Council Regulation (EEC) No 4064-89 and under Commission Regulation (EC) No 447-98 (the .Remedies Notice"), paragraph 5.

699 Remedies Notice, paragraph 12.

700 The equivalent of the sales of euro [...]* million by LDS to customers not normally served by wholesalers, or about [15-25]% of LDS.s total sales less supermarkets, and the equivalent of the sales of euro [...]* million made by La DIL to customers not normally served by wholesalers (channel B), or about [15-25]% of La DIL.s total sales less supermarkets.

701 The totals coincide almost exactly in the case of figures for distribution for outsiders (E1 026 million, as compared with euro 1 055 million by Lagardère), but the Commission reckons the market for distribution for outsiders is slightly smaller than the Lagardère group's estimate, at euro 634 million as compared with euro 712 million for Lagardère, on the basis of actual figures for 14marketers and distributors.

* The description of the Assets Retained is given to the best of the notifying party's knowledge to date, and may be adjusted, but not in any significant respect.

702 Authors whose sales of at least one book in 2002 exceeded 10 000 copies