Commission, April 5, 2013, No M.6789

Dear Sir/Madam,

Subject: Case   No   COMP/M.6789   -   BERTELSMANN/  PEARSON/  PENGUIN RANDOM HOUSE

Commission  decision  pursuant  to  Article  6(1)(b)  of  Council Regulation No 139/2004

1.   On 26 February 2013, the European Commission received notification of a proposed concentration pursuant to Article 4 of the Merger Regulation by which the undertakings Bertelsmann SE & Co. KGaA ("Bertelsmann") of Germany and Pearson Plc ("Pearson") of the UK acquire within the meaning of Article 3(1)(b) of the Merger Regulation joint control of the newly created joint venture Penguin Random House (the "JV") by way of purchase of shares.2 Bertelsmann and Pearson are designated hereinafter as the "parties".

I.              THE PARTIES AND THE JV

2.   Bertelsmann is an international media company whose core divisions encompass television and television production (RTL Group), trade publishing (Random  House), magazine publishing (Gruner + Jahr), music rights management (BMG) and services (Arvato) in some 50 countries.

3.   Pearson is active in publishing educational materials (Pearson), business  information (Financial Times) and trade publishing (Penguin; which includes  Dorling Kindersley Books and Rough Guides).

4.     Penguin Random House will contain all of the English language trade publishing divisions of Bertelsmann's Random House division in the United States, Canada, the United Kingdom, Australia, New Zealand, India and South Africa as well as the Spanish language trade publishing divisions of Bertelsmann's Random House division in Spain and Latin America, but will exclude Bertelsmann's German language trade publishing division Verlagsgruppe Random House. It will also contain all of Penguin's business and assets including its US, European, Australasian and Indian trade publishing divisions, its trade publishing company in China and its 45% share in a Brazilian Portuguese language publishing house. According to the initial strategic business plan for the years 2013-2016, the joint venture is expected  to have a turnover (net external sales) of USD […] in 2014.

II.                THE CONCENTRATION

5.     Joint control. The JV will comprise a Delaware entity ("USCO") that will operate directly and through subsidiaries in the United States, and an existing UK corporation or such other entity as will be agreed by Bertelsmann and Pearson ("ROWCO") (together "the Ventures"). The ROWCO will operate directly and through subsidiaries in the EU and other countries outside the United States.

6.     The USCO and ROWCO will each:

a) have a board of representatives ("Board"), Chief Executive Officer  ("CEO"), Chief Operating Officer ("COO"), Chief Financial Officer ("CFO"), and senior management; and

b) be owned, directly or indirectly, by Bertelsmann and Pearson, with shareholdings of respectively 53% and 47%.

7.     According to the "Framework Agreement" between Bertelsmann and Pearson, which sets out the governance and voting arrangements for the Ventures, Bertelsmann will have the power to appoint the majority of the members of the Board with voting rights. While most decisions relating to the Ventures require only a simple majority vote of the Board, certain "Reserved Matters" require the approval of Pearson as minority shareholder.

8.     The Framework Agreement confers a veto right upon Pearson for decisions regarding:

a) […]; and b) […]

9.     Pearson will also have the right, […].

10.  Finally, the Framework Agreement affords Pearson […]: a) […];

b) […];

c) […];

d) […];

e) […];

f) […];

11.    According to the Commission's Consolidated Jurisdictional Notice, a minority shareholder will be deemed to enjoy joint control over a JV if it is able to block the adoption of strategic decisions by the JV without having the power, on its own, to impose such decisions.3 Veto rights which confer joint control typically include decisions on issues such as the budget, the business plan, major investments or the appointment of senior management.4 By contrast, veto rights that are normally accorded to minority shareholders in order to protect their financial interests as investors are insufficient to establish joint control.5

12.    On the one hand, the following rights afforded to Pearson under the Framework Agreement are in themselves insufficient to lead to Pearson acquiring joint control over the JV:

a)              Pearson's veto rights in relation to […]. During the period between 2009 and  2012, Bertelsmann […]. During that same period, Pearson […]. It therefore seems unlikely that these thresholds will be attained for many strategic decisions concerning the JV and that they are rather designed to protect Pearson's financial interests as a minority shareholder

b)             Pearson's right […].

c)              Pearson's rights in relation to […].

13.    In addition, while in the Bertelsmann/KKR/JV,6 the minority shareholder, Bertelsmann, was an industrial shareholder that contributed its business to a joint venture with a majority shareholder, KKR, that was a mere financial investor, in this case, both Bertelsmann and Pearson will be industrial shareholders.

14.    On the other hand, when assessing Pearson's rights in the JV as a whole, and in particular in view of […], Pearson's influence over the JV will, on balance, lead to Pearson acquiring joint control over the JV:

a) […];

b) […]; and

c) […].

15.    Consequently, it can be concluded that Bertelsmann and Pearson will exercise joint control over the newly created entity.

16.    Full function joint venture. The JV qualifies as a full-function joint venture as it will operate on the market autonomously from its parent companies, performing all the functions of undertakings normally active on the same market. In more detail:

a)              The JV will sell trade books on the market and not depend on sales to its parent companies.7

b)             The JV will have its own management.8

c)              The JV will have a dedicated team of employees.9

d)             The JV will have adequate funding to perform its activities.10

e)              The JV will have access to sufficient assets and infrastructure to allow it to operate on a lasting basis.11

f)                 The JV will be formed for an indefinite period.12

17.    For the above reasons, the proposed transaction leads to the creation of a joint venture performing on a lasting basis all the functions of an autonomous economic entity within the meaning of Article 3(4) of the Merger Regulation.

18.    The transaction constitutes a concentration within the meaning of Article 3(1)(b) of the Merger Regulation.

III.             UNION DIMENSION

19.    The undertakings concerned have a combined aggregate worldwide turnover for the year 2011 of more than EUR 5 000 million¹³ (Bertelsmann: EUR 15 253 million,Pearson: EUR 6 754 million). In the same year, the parties each achieved more than EUR 250 million EU-wide turnover (Bertelsmann: […], Pearson: EUR […]), and did not achieve two thirds of their EU turnover in one and the same single EU Member State.

20.    The notified operation therefore has a Union dimension within the meaning  of Article 1(2) of the Merger Regulation.

IV.              COMPETITIVE ASSESSMENT

1.                  Introduction

21. The trade book publishing value chain for any given title comprises the acquisition of publishing rights, editorial input and design, production of the print and/or e-book version of the title and its sale to dealers and/or consumers as well as marketing and promotion activities. Some publishers also engage in book production and distribution activities for third party publishers.

22. In the past, publishers acquired, edited and marketed authors' works almost exclusively. More recently, to a large extent due to technical development, vertical integration can be observed on several levels of the value chain: authors started to engage in self- publishing while both literary agents and retailers such as Amazon started to provide publishing services and thereby compete with traditional publishers.

Table 1: Schematic overview of the trade publishing industry today

Source: Parties

23. Random House and Penguin are both active in the upstream acquisition of  publishing rights for English language titles, as well as on the downstream sale of English books to dealers. The Commission therefore assessed the likelihood of horizontal non-coordinated and coordinated effects on these markets taking also into account the vertical link between acquisition of publishing rights for English language books and sale of those books to dealers (sections 2 and 3). In addition, the Commission assessed the proposed transaction's impact on third party book distribution, in which both Random House and Penguin are present (section 4). Finally, the Commission assessed the proposed transaction's impact given the  vertical relationship between Bertelsmann's activities in book production (through its Arvato division and BePrinters) and the JV's activities in the sale of English language books to dealers (section 5).

2.              Acquisition of Publishing Rights and Supply of Services to Authors

2.1.          Product market definition

24.   The JV will be active upstream in the acquisition of authors' rights and offer of services, such as the editing of manuscripts, to authors.

2.1.1.       View of the parties

25.  The parties submit that the relevant upstream market encompasses the acquisition of authors' rights for the publishing of English language books without further differentiation based on different categories of books (for example general literature, children's books, and guides and manual); the formats (hardback versus paperback) or the original language of the work (whether the title was originally written in English or a foreign language).

26.  According to the parties, there is no reason to segment the upstream market by book categories, since the vast majority of works involve the licensing of authors' rights in a similar manner. Furthermore, from a supply perspective it would be easy for publishers to expand into the various book categories if they are not yet present in those.

27.  The parties further argue that publishers typically acquire rights to publish a title in all formats including as a hardback and a paperback and that unlike in the French publishing market, English language paperback publishing rights are not sold on to other publishers for exploitation, but exploited by the same publisher who publishes the hardback.

28.  As regards the publishing rights for a title originally written in a foreign language, the parties explain that these are generally the same as publishing rights for an original English title and that they are acquired in the same way as publishing rights for titles written in English language.

29.    Moreover, according to the parties, the acquisition of publishing rights for English language print books, e-books and audiobooks all belong to the same product  market, since publishers typically acquire them together.

30. In addition, the parties explain that the acquisition of publishing rights in bestsellers does not constitute a separate product market, as publishers cannot predict with certainty whether a title will develop in a bestseller or not, when they acquire the respective rights. Moreover, there would not be an industry-wide definition of a bestseller in any event.

31.   They further submit that on the upstream market, book publishers would provide a number of services to authors such as the payment of advances, editing, pre- publication legal advice, proofreading and design. For this reason, the parties consider the acquisition of publishing rights from authors to be closely analogous to its equivalent market in the context of music publishing, which the Commission has previously defined as a market for the provision of publishing services to authors.14

2.1.2.       Commission's assessment

32.   In its decision in the Lagardère/Natexis/VUP case ("Lagardère") concerning the French language publishing sector,15 the Commission defined separate product markets for the acquisition of publishing rights according to the original language of the work, that is to say (i) French-language publishing rights for an original work in French, and (ii) French-language publishing rights for an original work in a foreign language. In that case, the Commission further found that the categories of books  that were affected by publishing rights were all individual works, since it rarely happened that rights were acquired for collective works. Instead, in the case of collective works, typically, a fee would be paid for the individual contributions to the overall work. The Commission therefore considered that only general  literature titles, children's books, strip cartoon albums and academic and professional works constituted individual works for which publishing rights were acquired. It found that the markets for French-language publishing rights had to be subdivided according to the categories of books in question. In addition, the Commission defined separate secondary markets for French-language publishing rights for publication in (i)  pocket format and (ii) book clubs.

33.    While the market investigation validated several of the arguments of the parties, it also showed that several of the Commission's findings in Lagardère are also applicable to the English language publishing sector.

Services typically provided by publishers

34. The market investigation conducted for the purpose of the present transaction confirmed that publishers typically offer the following services to authors: editorial service, physical book production, e-book production (digitisation), audio-book production, sales, design (jacket and book interior), rights sales, and distribution.16 These services are provided by using the publishers' own capacities, or smaller publishers would source some of these services (for example distribution) from third parties.

Distinction between publishing rights based on the original language of the work

35. The results of the market investigation were mixed as regards a distinction between the acquisition of publishing rights for a title in English language and a title in another language to be translated into, and published in English language. While several publishers and literary agents indicated that the respective contracts do not differ, others indicated that there are differences with respect to (i) the person/entity from whom these rights are acquired (author/another publishing house/etc.), (ii) the scope of publishing rights (coverage of print version, e-book version, hardback, paperback), (iii) the contractual duration of publishing rights, and (iv) the level of author advances and royalties.17 In this context, it was pointed out that translation rights are often acquired from other publishers and not directly from the author.Advances and royalties would tend to be lower for a work that firstly needs to be translated into English language, among other reasons due to the cost of translation. Furthermore, the respective rights might be acquired for a limited period in time only.

Distinction between publishing rights for English language titles according to the category of books

36.The market investigation provided mixed results as to whether the acquisition of publishing rights for English language titles varies for the different categories of books, as defined in Lagardère. The majority of publishers who replied in the market investigation indicated that the typical publishing contracts do not vary significantly for (i) general literature, (ii) children's books, (iii); guides and manuals, (iv) strip cartoon albums, (v) art books and (vi) academic and professional books.18 However, the majority of literary agents who participated in the market investigation explained that publishing contracts in the above-mentioned categories differ from each other, for example as regards the level of royalties.19

Distinction of separate secondary markets for English language publishing rights for the publication of a title as paperback or via a book club

37. The market investigation confirmed that publishers typically acquire the rights to publish an English language title as hardback and paperback.20 Furthermore, the vast majority of respondents confirmed that the same publishers publish the hardback and paperback versions of a title.21

38. Moreover it was confirmed that the sale of a title via a book club does not require any specific publishing rights in the UK and Ireland.22

Distinction between the acquisition of publishing rights for print books, for e-books and audiobooks

39. In its market investigation, the Commission further tested whether the acquisition publishing rights for print books, e-books and audiobooks belongs to the same relevant product market. In this context, it found that today the vast majority of publishers acquire publishing rights for the e-book version of a title together with the publishing rights for the print version in over 90% of the cases.23 However, several respondents indicated that contracts that were concluded prior to the market launch of e-books may not necessarily cover publishing rights for e-books.

40.  As regards publishing rights for audiobooks, the results of the market investigation revealed that these are less frequently acquired together with the publishing rights for the print title than publishing rights for e-books. Answers varied from 25% to 80% of the cases.24

Publishing rights for English language potentially bestselling titles

41.  The Commission found that the level of advances paid to authors vary to an outstanding degree. While a […] number of authors do not receive any advances at all for their works, a small fraction of top selling authors may receive advances which are higher than GBP one million. In the period from 2009 to 2011, Penguin and Random House paid advances of GBP 500,000 or more only for respectively [0- 5]% and [0-5]% of the titles for which they acquired publishing rights.

42.  Taking into account the difference in the level of advances paid for a title as described above, the Commission considered whether the acquisition of publishing rights for English language titles written by well-reputed authors (which are more likely to be paid higher advances) could constitute a separate product market or a specific market segment with particular competitive dynamics. For this purpose the Commission defined an author as well-reputed if the author has written a title that was recorded in the weekly top 100 bestselling titles according to the Nielsen weekly 5000 database in any of the five years preceding the contract for new title(s), and/or if the author can be characterized as a celebrity in an area different from book publishing.

43.   The level of advances that the parties paid in 2011 for titles written by well-reputed authors was […] higher than the level of advances paid for titles written by any type of author. While the median advance paid by Random House and Penguin in the years 2009 to 2011 was GBP […] and GBP […] respectively,25 in 2011, the median advance for titles by well-reputed authors amounted to GBP […] in the case of Random House and GBP […] in the case of Penguin.26 This difference in median advances suggests that titles written by well-reputed authors could belong to a possible separate segment of the market.

44.  However, the results of the market investigation did not allow the Commission to define the acquisition of publishing rights by well-reputed authors as a separate market.27

45.  Nevertheless, taking into account that the competitive dynamics in the segment of  the acquisition of authors' rights for a consideration of more than GBP 500,000 may be different from the overall market for the acquisition of publishing rights, the Commission treats the acquisition of authors' rights for a consideration above GBP 500,000 as a possible separate market segment for the purpose of the competitive assessment.

Conclusion

46. The results of the market investigation were mixed as to the question whether the acquisition of English-language publishing rights for titles originally written in English and those originally written in a foreign language belong to the same product market. Furthermore, the results of the market investigation were mixed as to the question whether the acquisition of publishing rights for English language titles needs to be further divided according to the category of titles concerned ((i) general literature; (ii) children's books; (iii); guides and manuals; (iv) strip comics, (v) art books and (vi) academic and professional books. While the market investigation revealed strong indications that publishing rights for English language print books and e-books belong to the same product market, its results were mixed as to the question whether the acquisition of publishing rights for audiobooks belong as well to that product market. Furthermore, the market investigation revealed strong indications that the acquisition of authors' rights to publish an English language title as paperback or via a book club does not constitute separate secondary rights markets. Finally, the market investigation did not allow the Commission to conclude that the acquisition of publishing rights for titles written by well-reputed authors constitutes a separate product market.

47.   For the purpose of the present decision, the Commission considers that the acquisition of authors' rights to publish an English language title as paperback or via a book club does not constitute separate secondary rights markets. Apart from this, the exact product market definition as regards the acquisition of publishing rights for English language titles can be left open, given that the proposed transaction does not raise competition concerns under any possible market delineation.

2.2.             Geographic market definition

2.2.1.            View of the parties

48. The parties submit that the relevant geographic market for the acquisition of primary publishing rights is at least EEA wide and possibly worldwide. They add that the Commission can leave the definition open as, regardless of the geographic scope of the market, the transaction will not lead to a significant impediment to effective competition.

49. The parties explain that publishers active in the UK and Ireland will typically acquire exclusive English language publishing rights for the UK and Ireland, as well as, to the extent possible, the EEA and the Commonwealth.28 They underline that publishers in the UK and Ireland will invariably seek to obtain exclusive rights for the EEA, which are particularly important because of the implications of the exhaustion principle.29 However, in instances where exclusive rights for the EEA  may not be made available, publishers would opt to acquire non-exclusive rights.  The parties further submit that publishers in the UK and Ireland sometimes try to acquire "World English Language" rights, including US and Canadian rights, if they believe that they will be able to exploit these either directly or through sub-licensing arrangements with local publishers.

50.  In addition, the parties submit that they face essentially the same competitive dynamics in all countries where they are active. They face competition for the acquisition of publishing rights from at least four other large international publishers, Hachette, Harper Collins, Pan MacMillan and Simon & Schuster – as well as a large number of medium sized and smaller publishers. This similarity of competitive conditions globally would be reflected in the fact that compared to the EEA, Penguin and Random House's market share is not […] higher in other major English language countries in which they operate outside the EEA. To substantiate this statement the parties provide their market shares on the respective downstream markets for the sale of English language books to dealers as a proxy for their position on the upstream markets for the acquisition of publishing rights in those countries, as illustrated by table 2 below.

Table 2: Combined market share of physical books by value in major English language countries – 2012

2.2.2.            Commission's assessment

51.  In Lagardère, the Commission found the relevant geographic market for the acquisition of French-language publishing rights to be worldwide.30

52.  The market investigation conducted for the purpose of the present transaction confirmed that likewise the relevant geographic market for the acquisition of English-language publishing rights is at least EEA wide, if not worldwide. Furthermore, the geographic scope of publishing rights does not vary for whether for print books, e-books or audiobooks or according to the categories of books in question.

53. The majority of publishers who participated in the Commission's market investigation indicated that they aim at acquiring publishing rights for English language print books which are worldwide in scope. A smaller number of publishers indicated that they aim at acquiring publishing rights for English language titles for a territory that covers the EEA and the rest of the world without the United States and Canada. In addition, several publishers explained that they do not always succeed in obtaining the geographic scope of the publishing rights that they aim for, but sometimes have to settle for a narrower geographic scope, for example exclusive rights for the UK and Ireland and non-exclusive rights for the rest of the EEA or world.31 This is true for publishers of (i) general literature, (ii) children's books, (iii) guides and manuals, and (iv) reference works.32

54.   As regards publishing rights for e-books, the majority of publishers who participated in the Commission's market investigation confirmed that they aim at acquiring publishing rights which are either worldwide in scope or cover at least a territory comprising the EEA and the rest of the world without the United States and Canada.33 This is true for publishers of e-book titles falling into the following categories: (i) general literature, (ii) children's books, (iii) guides and manuals, and (iv) reference works, (v) strip cartoon albums, (vi) art books, and (vii) academic and professional books.

55.  As regards publishing rights for audiobooks, all publishers who participated in the Commission's market investigation indicated that they aim at acquiring publishing rights which are either worldwide in scope or cover at least a territory comprising the EEA and the rest of the world without the United States and Canada.34 Overall, the responses by publishers do not vary between general literature audiobooks and children's audiobooks.

Conclusion

56.  For the purposes of the present decision, the exact geographic market definition as regards the acquisition of publishing rights for English language titles can be left open since the proposed transaction will not raise competition concerns under any possible geographic market delineation.

2.3.             Competitive assessment

2.3.1.            Non-coordinated effects

Introduction

57.   In this section, the Commission assesses whether the proposed transaction will cause horizontal non-coordinated effects on the upstream market for the acquisition of publishing rights for English language books and its various segments, which possibly constitute separate product markets. The Commission considered the following possible separate markets for (i) the various book categories (that is to say general literature, children's books, guides and manuals, strip cartoon albums, art books and academic and professional books), (ii) the different book formats (that is to say print book, e-book and audio-book), and (iii) works originally written in English and works to be translated into English. Furthermore, within the competitive assessment, the Commission analyzed whether the proposed transaction will lead to particularly pronounced effects as regards the acquisition of publishing rights for English  language  titles  in  a  high  price  segment  of  the  market. The Commission carries out the competitive assessment of non-coordinated effects at EEA-wide and  at worldwide levels.

58.    In particular, the Commission assessed whether the merged entity would have  market power such as to be able to unilaterally raise the level of remuneration (advances and royalties) paid to authors on the upstream market(s) or specific segments of the upstream market(s) and thereby foreclose competitors from  accessing publishing rights for specific categories of titles as essential input for their downstream sales. If in the long run such foreclosure leads to the exit of other publishers from the downstream market and thereby reduces competition on this market, this might lead to consumer harm in the form of increased prices or reduced choice and innovation.

Views of the parties

59.  According to the parties, the transaction will not significantly impede effective competition in respect of the acquisition of publishing rights, for the following reasons:

60. First, they submit that as demonstrated in table 3 below, the merged entity's market shares in the upstream market for the acquisition of publishing rights would be modest. They propose to use the market shares in the downstream market for the sale of English language print books as a proxy for assessing their position in the acquisition of publishing rights. The parties submit that the method used in Lagardère to calculate market shares in the upstream market for the acquisition of  publishing rights by using advances paid to authors for bestsellers35 is not appropriate.

Table 3: Publishing rights for English language works (based on sales of print books to dealers as a proxy – UK and Ireland, by value, 2009-2012)

Sources: Nielsen bookscan data;

*As parent of Penguin Random House, Pearson will continue its educational business, including very limited trade book publishing.

61.  Post-transaction, Penguin Random House will have a share of [20-30]% by value in the market for the acquisition of all English language publishing rights. Even if   this were further sub-divided by category into general literature and children’s books, Penguin Random House would have a share of only [30-40]% and [20-30]% respectively for English language publishing rights in 2011.

62.  Second, the parties submit that Random House and Penguin are not each other’s closest competitors in relation to the acquisition of publishing rights. According to them, it follows from the use of the same market shares that publishers have in relation to the sale of books that there is no segment of the market where they do not face strong competition from the four other large international publishers, Hachette, HarperCollins, Pan MacMillan and Simon & Schuster, as well as from medium-sized and smaller publishers.

63. Third, according to the parties, authors choose a publisher based on a range of considerations, not just advances. Key factors would include the identity of particular editors, the marketing plan, the publisher’s perceived experience with similar titles, the culture of the publisher and other intangibles. They explain that the acquisition of publishing rights from both debut authors and established authors is generally conducted through a bidding process - managed by literary agents on behalf of authors, whether through a formal auction or one or more bilateral negotiations.

64.In this context the parties submit that in 2011, more than [80-90]% of publishing rights were purchased for less than GBP 50,000, an affordable amount even for smaller publishers. They explain that a significant proportion of the market is accounted for by medium-sized competitors such as Bloomsbury, Quercus, Canongate, Faber and Faber, Profile, Usborne, Walker, Egmont and Scholastic who will also continue to exert a competitive constraint in the market because they are frequent competitors on all bids, and are often willing to bid up to GBP 500,000 (EUR 616,622) for an advance manuscript. Medium-sized publishers may also bid in excess of GBP 500,000 for instance, where they place a higher value on a book matching their expertise or if they believe in a book more than other publishers and are therefore willing to pay a higher advance. Smaller publishers are also prepared to bid in excess of GBP 100,000 (EUR 123,000).

65.  In addition, the parties submit that even if the scale of the advance were to be the prime motivator behind an author selecting a particular publisher, there are many publishers that could easily afford the necessary expenditure. For example, all of Hachette, HarperCollins, Simon & Schuster and Pan Macmillan, and AudioGo in the context of audiobooks, will continue to have similar financial capacity as Penguin Random House post-transaction. All of these suppliers secured rights to established brand authors in fiction, as well as more expensive celebrity autobiography titles.

66.   Fourth, the parties explain that bidding dynamics will be unchanged by the transaction. According to them, there are several powerful agents active in the English language trade publishing industry which represents important players in the publishing value chain. These agents are well resourced, in many cases active across several industries in addition to books (for example film, theatre), represent a number of established authors, and their reputations, built-up over many years, mean that the titles they offer to publishing houses are likely to be seriously considered for publication.

67. According to the parties, the transaction will not affect the negotiating position of the literary agents vis-à-vis the parties especially as there will remain several other competitors that would have similar financial capacity to Penguin Random House (in particular, larger international competitors such as Hachette, Harper Collins, Pan Macmillan and Simon & Schuster).

68.  Fifth, the growth of self-publishing would have allowed authors to bypass the market and increased authors’ bargaining position, therefore placing additional constraint on publishers.

69.  According to the parties, the market for the acquisition of publishing rights would be currently undergoing significant change as a result of the explosion in popularity of e- books and the corresponding emergence of self-publishing companies that offer authors an alternative route to market. Self-publishing providers generally offer core publishing services (for example distribution via online sales platforms) for free. In addition, many also offer add-on services that approximate the packages offered by the traditional publishing houses for a fee (for example editing, cover design, promotional materials etc.). Furthermore, the option of print-on-demand books, where physical books are only manufactured when they are ordered, would have considerably reduced the cost and  risk involved in self-publishing and has opened up a low cost and low risk route to consumers for authors.

70. Finally, according to the parties, barriers to entry would be low and there would be significant potential for new market entry (for example as a provider of self-publishing services, or by non-English language publishers).

71.  This would be demonstrated by the growing number of self-publishers and Amazon’s own recent vertical move into publishing.

72. Also, the growth of Internet retailers would have leveled the playing field and provided easy access to consumers; and the growth in the self-publishing industry more generally would have led to the increasing commoditisation of publishing services, which may now be purchased separately from a range of suppliers. There would also be a large freelance marketplace in which the skills and expertise associated with publishing services may be purchased as needed.

73. The parties submit that it would therefore be relatively easy for companies with no previous publishing activities (for example retailers - as demonstrated by the example of Amazon - or Internet companies) to quickly enter the market as self-publishing service providers. Similarly overseas publishers with a focus in non-English language books could expand into English language countries with relative ease by purchasing the English language specific services they might lack (for example copy-editing and UK retailer distribution networks).

74. For all the above reasons, the parties submit that the transaction is unlikely to significantly impede effective competition in the market for the acquisition of publishing rights in the UK and Ireland.

Commission's assessment

75.   In the absence of any third party market share data for the acquisition of publishing rights for English language titles, in order to accurately determine the parties' market position, it would be necessary to fully reconstruct the upstream market(s) by identifying all amounts that are being paid as consideration for publishing rights, that is to say all advances and royalty rates negotiated between authors and publishers. However, for practical reasons such approach is not feasible. For this reason, the Commission first had to assess how best to approximate the parties' market position on the market(s) for the acquisition of publishing rights for English language titles. In a next step, the Commission applied the chosen method in order to assess the parties' market position within the overall market structure, before considering additional elements that impact upon competition in the upstream market(s).

a)   Methods to measure the parties' market position in the upstream market for the acquisition of authors' rights

76. The Commission considered and market-tested two different methods in order to approximately measure the parties' position in the upstream market(s) for the acquisition of authors' rights. Firstly, based on the methodology used in Lagardère, the Commission considered using the advances paid by publishers for bestselling titles as proxy to measure the parties' market position on the upstream market(s). Secondly, as suggested by the parties, the Commission considered using their market shares on the downstream market(s) for the sale of English language books to dealers as a proxy for their position on the upstream market(s) for the acquisition of authors' rights.

77.  In Lagardère, the Commission analysed the market position of the then merging parties by looking at their share in the payment of author's advances paid by publishers for the top 100 annual bestsellers. Following this approach, the Commission also investigated, for the purpose of the present transaction, whether measuring the authors' advances paid by publishers for the top 100 annual bestsellers in general or in the relevant categories (general literature, children's books, guides and manuals) may be a relevant proxy for determining the parties' competitive position in the upstream market(s) for the acquisition of authors' rights.

78.   The market investigation provided mixed results on this question. On the one hand, two publishers explained that measuring the level of advances was a good measure of competitive strength since advance payments are a key priority for agents and authors, and as such, are therefore important in assessing a publisher's market power in the upstream market. However, on the other hand, a literary agent explained that the level of advances is so varied that they cannot be compared to take a view on the upstream market power of different publishers. Another publisher also explained that bestsellers do not represent the entire publishing market and that "there are many other types of copyrights that can be successfully acquired beyond the top 100".36

79.   Furthermore, the Commission was not able to collect the advances being paid for all titles that figured among the 100 annual bestsellers in the above-mentioned categories. For this reason, the Commission was only able to partially reconstruct the sample. While this partial reconstruction does not allow the Commission to determine the parties' position on the upstream market(s) based on the level of advances they paid, it does however allow to confirm that in the present case the second method to measure the parties' market power, that is to say using their market shares on the downstream market(s) for the sale of English language books  to dealers, constitutes a good proxy for their position upstream, as explained in greater detail below.

80.   The market investigation also yielded mixed results on whether the downstream market shares for the sale of English language books to dealers constitute a relevant proxy for the calculation of the upstream market shares for the acquisition of authors' rights. On the one hand, a number of literary agents and publishers explained that the downstream market shares for the sale of English language books to dealers could be indicative to a certain extent of the market power in the upstream market for the acquisition of authors' rights, since, as one publisher summarized: "a publisher's strength in the markets for the sale of books to distributors is a factor that some authors may take into account in selecting among publishers and […] may find that success in the downstream market is indicative of a publisher's market prowess, and favour certain publishers accordingly".37

81.  On the other hand, several respondents indicated that since authors take into account a variety of factors when choosing between publishers, some of which being highly subjective, a publisher's position in the author's rights acquisition market cannot be said to generally reflect its position in the market for the sale of book to dealers. In addition, one respondent suggested that an extrapolation from the market position in the downstream market to the upstream market could offer misleading conclusions as for instance, in order to gain a large market share, "a publisher would need to acquire considerably more rights than other publishers, as a good proportion of what is published will prove unsuccessful in the market place. Equally, a publisher could gain an increased market share [downstream] boosted by a particularly strong selling author, series or book", which would not necessarily reflect its position in the upstream market for the acquisition of author's rights.38

82.  Based on the advances collected from the parties and several of their main competitors, given the mixed views of participants in the market investigation, the Commission carried out a comparison of the parties' scaled market share on a sample upstream market and on a sample downstream market. The Commission used the information submitted by the parties on the advances that they paid in 2011 and 2012 for top 100 bestselling titles in the categories of general literature, children's books and guides and manuals, as well as these respective amounts collected from their main competitors. The Commission then calculated the parties' advance shares in the sample upstream market, which obviously only represents a share of all advances paid by publishers to authors in 2011 and 2012. In a next step, the Commission calculated the sales values of the downstream sales of books that corresponded to the titles included in the sample upstream market. The Commission subsequently proceeded to the calculation of the shares that the parties and their competitors that were included in this exercise achieved with the sale of titles that were classified among the top 100 bestselling general literature titles, children's books or guide and manuals on the downstream market. As shown by table 4 below, the parties' scaled shares on the sample upstream and the sample downstream markets are in the same order of magnitude.

Table 4: The parties' scaled market shares for a sample markets comprising the top 100 bestselling general literature titles, children's books and guides and manuals in 2011 and 2012.

83.   For this reason, the Commission is satisfied that despite certain doubts raised by participants in the market investigation, in the present case, the parties' market shares on the downstream market(s) for the sale of books to dealers constitute a good proxy for their position on the upstream market for the acquisition of the respective publishing rights.

b)  The parties' position on the upstream market(s) for the acquisition of publishing rights for English language titles

84. The market data available for the parties' sales on the downstream market for the sales of English language books to dealers is limited to the UK and Ireland. As explained in more detail in paragraph 184 below, the Commission considers that the parties' market share in UK and Ireland reflect accurately their position on a wider geographic market covering the entire EEA. For this reason, the Commission considers that an assessment of the parties' position on a possibly EEA-wide upstream market for the acquisition of publishing rights for English language titles can be based on the available market data for the downstream sale of English language books in the UK and Ireland.

85.   Furthermore, the Commission considers that the parties' position on the upstream market on the level of the EEA would be the same or stronger than on the worldwide level. The parties explained that they face the same internationally active  competitors on a worldwide level, and that the United States, where the majority of English language rights originate is lower than in the EEA, see table 2 above. Hence, their combined market share on the worldwide level would be roughly the same if not lower on the worldwide level. For this reason, the Commission can limit the competitive assessment of the upstream market to the EEA. Since the transaction does not raise competition concerns on the more narrow geographic level of the EEA, the Commission can also exclude competition concerns on the worldwide level.

86. Based on the market shares of the merging parties on the downstream market(s) for the sale of English language books to dealers as a proxy for their position on the upstream market(s) for the acquisition of author's rights, their combined market position post-transaction will remain limited, and there will remain a number of significant competitors with relevant market shares that will continue to compete with the merged entity, overall but also in separate categories of books such as general literature, children's books or guides and manuals, where the parties are mostly active.

87.   As outlined above in table 2, the parties combined market share on the overall  market for the sale of English language books to dealers in the UK and Ireland, amounted to [20-30]% by value in 2012. Hence, their combined market share on an overall market for the acquisition of publishing rights for English language titles would remain in the same order. As shown by table 2, the merged entity would continue  competing  against  several  larger  publishers,  including  Hachette  with  a market share of [10-20]% by value in 2012, HarperCollins with a market share of [5- 10]% by value in 2012 and Pan Macmillan with a market share of [0-5]% by value in 2012, as well as against a large number of smaller publishers, which in 2012 made  up [40-50]% of the market. Furthermore, according to the majority of literary agents who participated in the market investigation, the merged entity is unlikely to unilaterally increase or decrease the level of advances paid to authors. None of the respondents esteemed that the merged entity will offer higher advances to authors and only two out of seven respondents anticipated a decrease in advances paid by the merged entity. The majority of respondents either believe that the transaction will have no impact on the level of advances or find it impossible to predict the possible effect of the transaction.39

88. Even if the upstream market for the acquisition of publishing rights for English language titles were to be divided by category of titles, the picture would not change significantly. As shown by table 3 below, the parties combined market share by  value in the possible markets for the acquisition of publishing rights for (i) children's books; (ii) guides and manuals; (iii) strip cartoon albums, (iv) art books and (v) academic and professional books remained below 30% from 2009 to 2012. In the category of general literature, the combined market share was slightly higher in the period from 2009 to 2011, during which it ranged between [30-40]% and [30-40]%. Only in 2012, the parties' combined market share amounted to [30-40]%, which is explained with the exceptional success of a particular trilogy published by Random House in that year ("Fifty shades of grey", "Fifty shades darker", and "Fifty shades freed" by EL James, hereafter "Fifty Shades trilogy"). The Commission considers that the unprecedented success of this particular trilogy needs to be taken  into account when assessing the merged entity's market position. Hence, the 2012 market shares alone do not accurately reflect the parties' market position on the downstream market for the sale of English general literature titles, which is better depicted by their market shares over the period from 2009 to 2012. By analogy, the parties' position on the upstream market for the acquisition or publishing rights for general literature titles is more accurately reflected by the level of their downstream market shares in the period from 2009 to 2012 than by their downstream share in 2012 only.

Table 5: Parties' combined share in the markets for the acquisition of publishing rights divided by book category (based on sales of physical books to dealers as a proxy – UK and Ireland, by value, 2009-2012)

Source: Nielsen Bookscan

89.   On this basis, the Commission concludes that the parties' combined positions on possible separate upstream markets for the acquisition of publishing rights for English language titles divided by category of title are not of such magnitude that would allow the merged entity to act independently of its competitors and cause non- coordinated effects.

90.   If the upstream market was to be divided into the acquisition of publishing rights for English language print books, e-books and audiobooks, the assessment would not change as shown by the respective market share levels on the downstream markets, see paragraphs 193 to 200.

91.  Finally, the market investigation did not reveal any indications for a stronger  position of the parties on a possible market for the acquisition of publishing rights of a foreign-language title to be translated into English and released in English language.

92.  In summary, the Commission considers that based on the merged entity's combined market share on the downstream market(s) for the sale of English language books to dealers as a proxy for its position on the upstream market(s) for the acquisition of the respective publishing rights, the merged entity would not hold a sufficiently strong market position on the overall upstream market and its possible segmentation in order to cause the risk of non-coordinated effects.

93.  This finding is strengthened by the following additional factors, which have a  bearing on the level of competition in the upstream markets for the acquisition of publishing rights for English language titles.

c)  Existence of a variety of factors on which publishers compete vis-à-vis authors

94.  First, the market investigation did not fully confirm the parties' submission that it is not uncommon for a publisher that has not offered the most lucrative advance or royalties' proposal to be nonetheless the successful bidder because the author has a personal preference for that publisher due to non-price reasons. On the one hand, the Commission analysis of the 2011-2012 authors’ win and loss data provided by the parties shows that, […]. However, this analysis could not be considered as complete as the level of advances offered by competing publishers was not available in a significant number of instances to the parties. In addition, a large number of market participants confirmed that "right contracts depend on a multitude of factors" and "it is difficult to predict since negotiations with each author are treated on a case by case basis and authors have different preferences and affinities".

d)  Level of advances generally affordable to medium-sized and smaller publishers

95. Taking the worst case scenario hypothesis that the level of advances was the sole or primary motivator behind an author's selection of a particular publisher, the Commission assessed the likely impact of the merger on the level of advances overall. Furthermore, as one publisher explained that "the top end of celebrity and genre fiction publishing is expensive, as the books have major appeal and are financially highly lucrative. To buy these rights and to publish them successfully requires strong financial resources"40, the Commission also assessed the impact    of the merger on a possible market segment within the overall market of the acquisition of authors' rights, namely the acquisition of authors' rights for a consideration above GBP 500,000 (which corresponds to the higher advances paid to authors).

96.   As regards the overall market, the Commission was able to collect partial information from a number of literary agents and publishers on the average amount of advances they received for the authors they represented in 2011 and 2012. These amounts varied greatly for different literary agents and publishers. The respective average advances ranged from around EUR 30,000 to not more than EUR 145,000.41

97.    The Commission acknowledges the comments from a number of market participants that some of the high amount of advances paid to bestselling or celebrity authors  may distort the picture given by the average amount of advances paid by  publishers.42

98.  Nevertheless, the parties also submitted figures showing that the median advances paid to authors by Random House and Penguin between 2009 and 2011 were GBP […] and GBP […] respectively. In addition, according to the parties' figures, [80- 90]% or more of their authors' rights were purchased for GBP 50,000 (EUR 57,325) or less, and for more than [90-100]% of all titles published, authors received an advance below GBP 100,000 (EUR 114,650).

99.  The Commission considers that the amounts of advances paid by the parties may be taken as a proxy for the entire publishing industry in the UK and Ireland, given the relatively low level of these amounts even for the first and third largest publishers in the UK and Ireland that Random House and Penguin are. Furthermore, the Commission, also analysing the 2011-2012 authors’ win and loss data provided by the parties, confirms that in practice, competing mid-sized or smaller publishers have bid for such amounts or larger amounts of advances in the past two years. As a  result, the Commission considers that amounts of up to GBP 50,000 or even GBP 100,000 must be affordable also for a number of larger and even smaller publishers competing with the merging parties in the acquisition of author's rights.

100.   As regards the possible separate market segment for the acquisition of publishing rights for a consideration above GBP 500,000, the Commission, analysing the  figures of the parties first recognises that a number of authors have been paid advances which ranged between GBP 500,000 to GBP 2 million. However, according to the parties' own computations, only [0-5]% for Random House and only [0-5]% for Penguin of all books titles published between 2009 and 2011 received an advance in excess of GBP 500,000 (EUR 573,250). Therefore, even if, taking the worst case scenario, the merging parties were the only publishers to be able to afford to pay such high advances post-transaction, they would only be able to foreclose a very small minority of the market (between [0-5] and [0-5]%), leaving other publishers to be able to compete for at least [80-90]% if not [90-100]% of the market for the acquisition of authors rights.

101.  Furthermore, analysing the 2011-2012 authors’ win and loss data provided by the parties,  competing  publishers  have  bid  during  that  period  for  authors  in  a large number of instances for amounts ranging over GBP 100,000 and going up to between GBP 1 million, and up to GBP 3 million in […] instances.

102.  The Commission considers that this demonstrates that other publishers are currently able and do currently compete with the merging parties, also on the high bids for  well reputed or bestseller/ brand authors.

103.   The Commission assessed whether the proposed transaction would change this situation.

104.   First, the results of the market investigation were mixed as to the question whether the merged entity would be in a position to offer more attractive publishing contracts (including a higher level of advances) to authors compared to their offerings today.

105.   On one hand, two publishers believed that the marketing power of the merged entity will be "more attractive to many authors" as it will "have more distribution power due to its size and therefore [will be] able to command more power at retail and therefore probably reach a wider audience". Another publisher also believed that the merged entity would be able to pay higher advances to authors.43

106.   On the other hand, a publisher explained that advances are already generous today and the proposed transaction should not change that situation, while two other publishers explained that it was difficult to predict whether the merged entity would be more attractive to authors post-transaction as "right contracts depend on a multitude of factors" and "it is difficult to predict since negotiations with  each authors are treated on a case by case basis and authors have different preferences and affinities". Likewise, another publisher explained that "it does not seem likely […] that the contracts will be greatly different in ways other than industry norms."  In addition, a publisher also explained that even if "the merger will generate synergies that the merged entity may be able to pass on to authors in the form of more attractive terms […] the merged entity will still face fierce competition from other publishers who may be able to offer equally attractive or more attractive terms".44

107. Furthermore, none of the literary agents which replied to the market investigation believed either that the merged entity would be in a position to offer more attractive publishing contracts to authors compared to the contracts currently offered by Random House and Penguin or by their competitors.45

108.   Secondly, the market investigation did not fully confirm either that the merged entity would be in a position to successfully sign more promising, well-reputed or best- selling authors than Random House and Penguin do today. Even if one literary agent believed that "their market share and financial strength would enable them  to outbid, in advance and resources terms, other publishers, more often that they do now", two explained that already prior to the transaction, the merging parties did  not have difficulties to attract such authors so the situation will not change post- transaction.46

109.Thirdly, when asked whether the combined Penguin Random House would become so strong for the acquisition of publishing rights for English language print and e- books and audio books that other publishers would find it difficult to compete with the merged entity, one literary agent explained that the merged entity's "position in the upstream market will be strengthened relative to their market share of the downstream market to the extent that in a significant number of cases we would expect them to prevail over competitors". However, another literary agent explained that "there is plenty of competition – both from other publishers and because authors can also self-publish. Amazon's self-publishing service will continue to provide serious competition – and it is much larger than either Random house or Penguin".47

110. In addition, the vast majority of publishers did not consider that the combined Penguin Random House would become so strong for the acquisition of publishing rights that other publishers would find it difficult to compete with the merged entity. As one publisher explained: "book publishing is a fiercely competitive industry comprising numerous players of widely varying sizes, ages and specialties […] authors take into account a wide array of factors in deciding between publishers,  and as such, while the merged entity may be more attractive to certain [authors] in certain respects, it will not be capable of making it appreciably more difficult for other publishers to compete in the rights acquisition market". Similarly, according to another publisher, "the merged business may be stronger than the separate companies, but not to a very great degree in our views. As far as marketing and editorial strengths are concerned, they won't become "so strong" through merger".48

111. In light of the results of the market investigation, the Commission considers it more likely than not that post-transaction, larger as well as medium-sized and smaller publishers will continue to compete with the merged entity on the level of advances in the overall market for publishing rights, as well as in the segment for higher  priced publishing rights.

Conclusion on non-coordinated effects

112. In light of the above, the Commission concludes that the proposed transaction does not raise serious doubts as to its compatibility with the internal market in the upstream market(s) for the acquisition of author's rights as regards non-coordinated effects.

2.3.2.            Coordinated effects

113. The Commission assessed whether the proposed transaction may create any risk of coordinated effects in relation to the acquisition of authors rights by publishers. In particular, the Commission assessed whether the publishers (all of them or only the larger ones) may tacitly coordinate their behaviour in order to reduce the amount of advances paid or the level of royalties paid to authors.

114.  To assess coordinated effects, well-established case law49 and Commission guidelines on the assessment of horizontal mergers50 require proof that the merger will make coordination more likely, more effective and more sustainable. The analysis needs to focus in particular on: (i) the ability to reach terms of coordination; (ii) the ability to monitor deviations; (iii) the existence of a credible deterrent mechanism if deviation is detected; and (iv) the reactions of outsiders such as potential competitors and  customers.

115. The Commission first investigated whether, and if yes to what extent, the amount of the advances and the level of royalties that was agreed among the winning publisher and an author in an auction or bidding procedure were generally known in the market, so that competing publishers would be able to detect any deviation from the amount of advances or level or royalty tacitly agreed between competing publishers.

116.   As regards first the amount of advances, even though a number of author's agents  and publishers which replied to the market investigation explained that "on occasion, the specific amount paid becomes known [and that] more commonly there may be a range of speculation about the advance, [...]", a vast majority of author's agents and publishers confirmed that this information was not publicly available, or made known to other publishers.51

117. As regards second the level of royalties, the vast majority of literary agents and publishers confirmed that coordination on royalty rates was unlikely given that the rates are negotiated individually and are not normally disclosed publicly. As one literary agent explained: "there is not as much variability in royalty rates as there is in advances. To the extent that agreed royalty rates resulting from an auction may vary from customary industry parameters, the royalty rates may sometimes become known to other publishers participating in the auction, but it is our sense that the rates are generally not made known to other publishers".52

118. The Commission therefore considers that the market for the acquisition of authors' rights is unlikely to be sufficiently transparent to allow monitoring deviation from any hypothetical tacitly agreed amount of advance or level of royalty and that therefore, the transaction is unlikely create any risk of coordinated effects on that market.

119.  The Commission also investigated whether it would be likely that only large publishers - that is to say, excluding small and medium-sized publishers - would coordinate their behaviour on the market for the acquisition of authors' rights.

120.  A large majority of publishers, including smaller publishers, explained that it was unlikely, given the competitive nature of the publishing industry, the number of competing publishers, and the continual changes within the industry, that large publishers  would  be  able  to  coordinate  their  behaviour  on  the  market  for    the acquisition of author’s rights.53 Only one publisher and one literary agent raised the concern that there may be coordination between the large publishers, specifically on the level of royalty rates for e-books.

121.  On this latter point, the Commission notes however that the e-book royalty rate is just one of a range of key parameters on which publishers compete for authors’ rights. These parameters include non-price parameters such as editorial skill, the reputation of publishers, and their marketing and promotional skills, as well as other price parameters such as the level of advances paid and royalty rates, not only for e- books but also for all other formats or sales.

122.   Furthermore, as was confirmed in paragraph 117, e-books royalty rates negotiated with authors/authors’ agents (and the other key pricing parameters for acquiring authors’ rights such as advances and royalty rates on print books) are not generally made public and are not tracked by third party data providers such as Nielsen Bookscan.

123. Moreover, […], different royalty rates apply for other formats such as physical books; for sales in home markets and export markets; […]. Against this backdrop,  the disclosure of e-book royalty rates cannot bring any meaningful transparency to the entirety of the price component of the way in which large publishers compete for authors’ rights.

124.   Based on the above, the Commission takes the view that it is unlikely that there is sufficient market transparency to facilitate coordination on e-book royalty rates.

Conclusion

125.    The Commission concludes that the proposed transaction does not raise serious doubts as to its compatibility with the internal market in the upstream market(s) for the acquisition of author's rights arising from a risk of coordinated effects.

3.             Sales of English Language Books to Dealers

3.1.          Product market definition

3.1.1.            View of the parties

126. The parties submit that the relevant market is the sale of English language consumer books to dealers in the EEA (primarily in the UK and Ireland).

127.  According to the parties, print books, e-books and audio books are part of the same product market. Moreover, they do not consider it necessary to segment the market by sales channel, categories of titles (that is to say general literature, children's books, art books, guides and manuals, etc.) or, as regards print books in the general literature category, by format (that is to say hardback and paperback). However, the parties provide the relevant market data and submit that the exact product  market definition can be left open as the competitive analysis remains unchanged.

128. First, as regards a possible segmentation between print books, e-books  and  audio books, the parties submit that all formats are part of the same relevant product market.

129. E-books are considered to be a competitive constraint on print books and vice-versa  as (i)   publishers are generally active across e-books and physical books; (ii) publishing rights typically are acquired jointly for physical and digital sales; (iii) the content of the title is essentially identical whether purchased in digital or physical form; (iv) the same factors will influence consumer purchases of both e-books and physical books and consumers tend to purchase a given title in only one format; (v) the e-book and hardback versions of a title are published at the same time; (vi) the decline in print sales coincides with the growth in e-book sales; and (vii) the wholesale pricing of e- books is linked to the recommended retail price ("RRP") of physical books.

130.The parties submit that it is neither necessary nor meaningful to sub-divide the supply  of e-books into separate markets for the supply of e-books to dealers under a resale (wholesale) model and the supply of e-books through dealers under an agency model. The choice of distribution model is simply a commercial issue to be negotiated between the publisher and the dealer within the confines of the market for the sale of consumer books. The choice between agency and resale distribution is thus not a choice between two separate markets, but simply between the commercial terms in which the parties distribute within the market via intermediaries (that is to say dealers).

131.  In relation to audio books the parties submit that there is supply- and demand side substitution between audio books and other book formats because (i) with a few exceptions54, there is no discernible difference in customer taste between physical, e- book and audio book formats; (ii) the same publishers are largely active across all three formats, although certain audio book publishers (notably the BBC) are substantially less active in print book and e-book publishing; (iii) publishers seek to and generally do obtain audio book rights together with print book and e-book rights; (iv) the content is substantially the same as for the print book and e-book format; (v) if possible, audio books are published at the same time as the e-book or the hardback print book55; (vi) marketing is directed at print, e-book and audio book format of a given title; (vii) many print book and e-book retailers also sell audio books (for example Waterstones, WH Smith, Amazon); and (viii) pricing to retail customers is determined in essentially the same way as print books/e-books; that is to say, a discount off the RRP for physical audiobooks and a further discount off that RRP for digital audiobooks. The parties do not consider a further distinction between digital and physical audio books.

132. Second, the parties do not consider it necessary to subdivide the market by sales channel. First, the conditions of competition are the same across all dealers, as the same dealers will purchase books across a large range of publishers. Second, different publishers organise their sales teams across many different lines and not necessarily along the different types of dealers.56 Finally, in relation to e-books the parties submit that any further segmentation is unnecessary as, according to their estimates, Amazon accounts for approximately [90-100]% of all e-book sales in the UK and Ireland.

133.  Third, the parties do not consider there to be a need for a subdivision of the market by categories of titles. They submit that there is potential for both substantial supply side and demand side substitutability across the categories. In particular, the book acquisition, publication, and sales process are very similar across different categories and publishers will tend to be active across a large range of categories.

134.Also, due to strong supply side and demand side substitutability the parties do not consider a further subdivision within any possible categories according to genre (that is to say romance, crime thriller, science fiction, travel literature, children's fantasy etc.) to be meaningful or appropriate.

135.  Fourth, as regards print books in a possible general literature category, the parties submit that hardback and paperback formats are part of the same product market.  The large majority of titles will be published both as hardback and paperback title, with the edition reflecting the life cycle of a book and the price differential reflecting the cost differences in production as well as consumer preferences. From a supply- side perspective, as set out in paragraph 27, the parties do not consider it appropriate to distinguish between the acquisition of rights for both formats. From a demand- side perspective, customers are said to be generally prepared to purchase both formats. Moreover, there is competition between both formats as is evidenced by the negative effect the "Fifty shades" trilogy published by Random House in 2012 had  on other book sales even though very few hardback copies of this trilogy have been sold.

136. Fifth, the parties do not consider that bestsellers constitute a separate market. They submit that there is no industry standard definition for bestsellers and no step-change in the volumes of books sold according to their chart positions thereby not supporting a market definition according to chart positions. Moreover, sales rankings vary substantially over time so that a given title may be in the top 200 titles in a given week but only in the top 1000 a few weeks later. Apart from demand, seasonal factors are also claimed to affect sales, which suggests that sales of a given title may vary depending on the date of its launch. Furthermore, consumers and dealers make their choices across the full range of bestsellers and non-bestsellers. In addition, the publishers' objective is to increase the volumes of each title sold but it is uncertain if a title is going to be a bestseller or not when the rights are acquired and when a title is launched. Finally, the parties note that high-selling generally titles tend to be discounted more and enjoy greater promotional spend than less successful titles.

3.1.2.            Commission's assessment

137.  The Commission concluded in Lagardère that the market for the sale of French language (print) books to dealers should be subdivided according to the types of dealer.57 The Commission also subdivided the market by categories of titles and, for some categories defined further subcategories58 but it did not further subdivide these categories by genre59. Within the general literature category, the Commission further defined separate markets according to the format (hardback vs. paperback).60

138.  Preceding the assessment of the relevant product markets, the Commission notes in relation to the publishers' sales of e-books that, for the purposes of the present decision, no distinction is made as to whether e-books are sold to dealers under the wholesale model or whether they are sold through dealers under the agency model.

Possible distinction between print books, e-books and audio books

139.  In Lagardère, the Commission did not consider a distinction between print books and audio books.61

140.   The market investigation revealed, first, that print books and e-books may constitute separate product markets.

141.  The majority of responding publishers and customers consider that print books and e-books differ in terms of (i) sales channel; (ii) pricing at wholesale level; (iii) pricing at retail level; and (iv) promotion of specific titles. Unlike the responding customers, the majority of publishers did not see any differences in terms of (i) the typical end-customer; and (ii) the mode of consumption, that is to say if a title is read at home, when travelling, etc.62

142.  The results of the market investigation are more mixed as regards substitutability of the two formats. Nevertheless, the majority of responding publishers and customers do not consider that wholesalers or retailers would switch from print books to e-books and vice-versa in case of 5-10% increase in price.63 Moreover, the majority of responding customers consider that the vast majority of consumers would not switch from print books to e-books and vice-versa in case of a 5-10% increase in the retail price Only two customers consider, without providing further details, that 10-25% of end-customers would switch while a minority of customers consider that most consumers would switch in the event of a price increase of this magnitude.64

143. Second, the market investigation showed that audio books are distinct from print books.65

144.The majority of responding publishers and customers consider audio books to differ from print books in terms of (i) pricing at wholesale level; (ii) pricing at retail level (audio books tend to be more expensive both at wholesale and retail level); and (iii) promotion of specific titles. The majority of customers also see differences in terms of the typical end customer and the mode of consumption while results are mixed in this respect for the responding publishers. Overall, results are mixed as regards the sales channels through which both products are sold.

145.  Moreover, the vast majority of publishers and customers do not consider that wholesalers or retailers would switch from audio books to print books (and  as regards the responding publishers also vice-versa) in case of a 5-10% increase in price.66 In addition, the majority of customers do not think that consumers would switch from audio books to print books in case of a 5-10% increase in the retail price or consider that only a "low percentage" of end-customers would switch. However, one customer states that "the introduction of digital audio books is more likely to have an impact".67

146. Third, the market investigation showed that audio books are also distinct from e- books.

147.  The majority of responding publishers and customers consider audio books to differ from e-books in terms of (i) pricing at wholesale level; and (ii) pricing at retail level (again, audio books are said to be more expensive at both levels). The majority of customers also see differences in terms of the typical end customer, sales channel  and the mode of consumption while results are mixed in this respect for the responding publishers. Results are mixed for both publishers and customers as regards differences in promotion of e-books and audio books.68

148.   Furthermore, none of the responding publishers and customers considers that wholesalers or retailers would switch from print books to audio books (and  as regards the responding publishers also vice-versa) in case of a 5-10% increase in price for print books.69 A vast majority of customers do not think that consumers would switch from audio books to e-books in case of a 5-10% increase in the retail price of audio books. One customer considers that only 25% of end-customers would switch, explaining that "potentially people would switch to the e books and listen via text to speech technology instead".70

149.   For the purposes of the present decision, the exact product market definition as regards the sale of English language print books, e-books and audio books to dealers can be left open since the proposed transaction will not raise competition concerns under any possible product market delineation.

Possible distinction by sales channel

150.  In Lagardère, the Commission concluded that the market for the sale of books to dealers needs to be subdivided according to the following sales channels: (i) bookshops (including both large bookshops/specialised superstores and local bookshops); (ii) hypermarkets; and (iii) wholesalers.71

151. The Commission investigated a possible subdivision of the market based on sales channels such as chain booksellers, independent booksellers/small bookshops, supermarkets/hypermarkets and e-tailers (which had not been considered in Lagardère).

152. The market investigation does not confirm the parties' claims that a distinction by sales channel is not necessary. The vast majority of respondents consider that the above sales channels vary in terms of (i) final customers; (ii) service offered to final customers; (iii) level of discounts granted; (iv) the person in charge of discount negotiations on the side of the reseller (for example specialised purchaser, librarian, owner of the business) as well as on the side of the publisher; and (v) the categories of books offered. The respondents also consider the level of specialisation of a retailer to differ between the different sales channels.72

153.  In any event, for the purposes of the present decision, the exact product market definition as regards the different sales channels for the sale for English language titles can be left open since the proposed transaction will not raise competition concerns under any possible product market delineation.

Possible distinction by categories of titles

154.   As indicated in paragraph 137, the Commission, in Lagardère, found certain categories of books to constitute separate relevant product markets. Those categories comprise (i) general literature; (ii) children's books; (iii) art books; (iv) guides and manuals; (v) strip cartoon albums; (vi) educational books (further subdivided into school text books and educational supporting materials); (vii) academic and professional books (further subdivided into law books, books on science, technology and medicine and books on economics and social sciences); and (viii) reference works (further subdivided into smaller reference works (dictionaries and encyclopedias), larger reference works and books published in installments). The Commission did not further subdivide any of these categories by genre.

155.  The market investigation confirmed a possible distinction by category of books. In particular, the following criteria vary between categories: (i) the typical end customer; (ii) the RRP; (iii) the display of the category of books at the different sales points; (iv) cover design and format; and (v) the graphical presentation of the book (for example photos, text, pictures, tables etc.). Results are mixed as regards differences in the final use of the book by the consumer and the publishing house  that would typically publish the respective category of books.73

156. Moreover, the market investigation showed that the different categories of books are not interchangeable from the demand-side if wholesale prices in any one category would increase by 5-10%.74

157. The Commission notes that some respondents consider retailers' choices not only to be driven by price: "It is the product mix rather than the price mix that wholesale customers are used to working with. Each category has a different set of price bands and even those these move from year to year, they have not in the past shifted their demands of products based on price levels. The more likely drivers for each category are age bands of populations, education levels and the economy as a whole."75

158. Finally, the Commission, in light of its findings in Lagardère, considers there to be no need for a further subdivision of any of the above categories by genre.

159.  For the purposes of the present decision, the exact product market definition as regards the different categories of English language titles for the sale to dealers can be left open since the proposed transaction will not raise competition concerns under any possible product market delineation.

No need to define separate markets for hardback and paperback print editions  within a possible general literature category

160. The majority of responding publishers explain that they generally publish both formats. The paperback would usually be published between six months and two years after the hardback. Some publishers indicate that certain types of books falling under a potential general literature category are only published in one of the two formats. Overall, the main business rationale behind such choices is said to relate to expected sales of a given format.76

161. The vast majority of responding customers, to the extent that they sell print titles, submit that they typically purchase both the hardback and the paperback edition of a print book, and that they generally would do so in line with the respective  publication dates.77  One customer submits that they "do not make buying    decisions based on format – [they] buy based on customer demand".78 Nevertheless, the Commission notes that "there are certain types of books that would be more price- sensitive, such a[s] first-time fiction, which would be an issue for our market in hardback. Sometimes, depending on the publisher's investment in a title, we would only buy it in the cheaper, smaller paperback format."79

162.  However, it appears from the market investigation that publishers' choice to only publish one edition and customers' choice to only purchase one edition is rather more exceptional. The market investigation has not revealed any further differences between hardback and paperback format such as those found by the Commission in Lagardère in relation to the sale of French language titles. Therefore, the Commission considers that, for the purposes of the present decision, there is no need to define separate markets for hardback and paperback English language print editions of titles within a possible general literature category.

No need to define a separate market for bestsellers

163. The market investigation confirmed the parties' argument that there is no industry standard definition for bestsellers.80 Some respondents base themselves on sales data and rankings provided by Nielsen. In this respect, the Commission notes that, as explained by the parties, Nielsen data does not cover e-books. One publisher adds that Nielsen data is predominantly UK and Ireland based and that they "typically utilise international publication data and specific sale route data to define bestsellers". Other publishers merely assess the number of copies sold of a given title.  Thresholds  for when a title is considered to be a bestseller vary (for example "100,000 [copies] in one season" or "an exceptional number of copies"). In the same vein, a number of retailers, including Amazon, base themselves on their own sales rankings. Other respondents submit not to have any definition for bestsellers.81

164. Respondents do not generally consider that bestsellers, apart from selling well, have unique inherent characteristics that would unambiguously distinguish them from non-bestselling titles. However, the greater historic sales of the author (by brand authors) and/or his or her notoriety, the amount of marketing spend allocated to promoting a title, the quality of the content, the media coverage of the title and whether its subject matter concerns a "subject area in vogue" are possible factors common to bestselling titles.82

165.  The market investigation revealed mixed results as to the substitutability of bestsellers with non-bestsellers or "deep range" titles from a publisher, dealer and consumer's perspective. A majority of customers consider bestsellers and other titles to be interchangeable while the majority of publishers do not consider this to be the case. One argument put forward by several respondents against substitutability is the price difference.  Bestsellers  are  considered  to  be  more  price  sensitive  and  tend  to  be discounted more heavily than deep range titles. Without further specifying, some respondents submit that substitutability depends on the category of titles or genre in question. On the other hand, more generic titles by authors with no brand name are considered to be more interchangeable than others. The majority of respondents do not consider their assessment to differ when considering print books, e-books and audio books separately.83

166.  The market investigation confirmed the parties' submission that most (type of) dealers make their choices across the full range of bestsellers and non-bestsellers, often aiming at offering the broadest possible range of titles. Some customers submit that they carry out their own assessment of the titles they acquire. The vast majority of respondents do not consider their assessment to differ when considering print books, e-books and audio books separately.84

167.  The market investigation did not unequivocally confirm the parties' submission that  it is uncertain whether a title is going to be a bestseller when the rights are acquired and when a title is launched. While this statement is generally sanctioned with  respect to new or unknown authors, a majority of respondents consider that it is  likely for a publisher to correctly predict whether an individual title will be a bestseller for established brand authors/ top authors with a bestseller track record (although also in this scenario an element of uncertainty would remain). 85

168.  In addition, for the vast majority of customers, bestsellers are not a "must have" for a publisher in order to win a retailer as a customer.86 The result appears to be slightly more mixed in relation to certain sales channels such as supermarkets or multimedia chains carrying a limited range of titles as they do not strive to offer a wide range of titles but concentrate on lower-priced titles (due to heavy discounts and/or promotional spend) and/or popular titles driving sales volumes.87 In the words of one competitor: "Supermarkets and WHSmith [a multimedia and stationary retail chain] drive the bestseller lists. It is possible to have a book on the bestseller list without these 2 stocking the title but it would be rare."88 A majority of competitors confirmed that a publisher must have bestsellers in his catalogue in order to win supermarkets as wholesale customers, albeit some publishers also pointed out that they do sell to supermarkets without having any bestsellers, that supermarkets are now starting to take  a wider range of books and that they sell titles which may not be bestsellers elsewhere.89 However, the majority of customers, including a large supermarket chain, submit that they try to source across a broad range of titles - not necessarily only bestsellers - and across a number of publishers. In particular, a responding supermarket emphasised that they also purchase a number of educational books, general reference and cookery books that would not necessarily be considered to be bestsellers.90

169.  In light of the above, the Commission considers that, for the purpose of the present decision, the sale of English language bestsellers is unlikely to constitute a separate market but is best considered, in the competitive assessment, as a specific sub-segment of the sale of English language books.

3.2.          Geographic market definition

3.2.1.            View of the parties

170. The parties submit that the relevant geographic market for the sale of English language books to dealers should be defined along linguistic boundaries. It should therefore comprise all English language books sold to dealers in the EEA, recognising that most sales will be recorded in the UK and Ireland. The geographic scope of the market would not change if print books, e-books and audio books are considered separately. Furthermore, the parties submit that the success of a title in  the UK and Ireland is likely to be reflected in demand for the same title in the EEA.

171.  The parties consider that neither the US nor other countries outside the EEA should form part of the relevant geographic market as authors' rights are almost always sold separately for the US market, or other regions, titles benefit from being published locally, format and cover design may differ and consumer tastes are markedly different so that for example certain UK or Irish authors do not sell well in the US and the popularity of certain genres may vary depending on the country. In addition, the parties submit that there are slight linguistic differences between the US and the UK/Irish versions of any given title that is published in both regions.

3.2.2.            Commission's assessment

172.  The Commission has previously defined the relevant geographic market for the sale of French language titles to be the relevant territory of the EU/EEA in which the language of the titles is primarily spoken, with the exception of the market for the sale of school textbooks and the market for the sale of law books, which were defined as national markets.91

173. The Commission notes that the parties, although basing themselves on Lagardère, propose a slightly different approach to the definition of the relevant geographic market as regards the sale of English language titles to dealers by not limiting the scope of the market to those countries where the language of the title is primarily spoken. However, the market investigation broadly confirms that the relevant geographic market should also cover areas outside the UK and Ireland where the parties sell English language titles to dealers.

174. Contrary to the parties' views, a majority of responding publishers consider the geographic scope to go beyond the EEA so as to also cover the rest of the world except the United States and Canada or even to be worldwide.92 This is consistently claimed for print books, e-books and audio books and does not vary according to the categories of titles sold (that is to say general literature, children's books, guides and manuals and reference works). Overall, the responding publishers do, however, not substantiate their claim for such a broader market. A significant number of responding publishers stresses that the geographic scope of the downstream market may be limited by the scope of the underlying rights acquired upstream93 and/or the geographic reach of their respective dealers.94

175. The Commission acknowledges that a significant number of both publishers and dealers of English language titles are active beyond the EEA and do not necessarily organise their sales/sourcing on a regional or national basis for either print books, e- books or audio books.95 Notwithstanding this, the Commission considers, in line  with the arguments put forward by the parties regarding differences in the sale of English language titles to dealers, between the EEA and other countries outside the EEA, both on the demand and on the supply side between different countries and regions that the scope of the geographic market does not extend beyond the EEA. The market investigation has not yielded any substantial argument casting doubt on the parties' assessment in this respect.

176.  The Commission also investigated whether, within the EEA, the UK and Ireland on the one hand and the rest of the EEA on the other hand constitute two separate relevant product markets for the sale of English language titles to dealers. The results of the market investigation are mixed. A majority of responding publishers and customers96 consider there are differences between the two geographic areas as regards (i) the level of discounts granted to dealers – a certain number of respondents claims that such discounts tend to be higher outside the UK and Ireland; (ii) end- consumer preferences, notably due to the primary language spoken; and (iii) publishers present in the respective area. However, the majority of the responding publishers and customers consider there are no differences as regards the RRP and the categories of dealers present on the market, although some respondents note that there are some dealers operating only locally.97

177.  In light of the above and taking into consideration the parties' argument that demand for the respective English language titles is similar across both regions, there is a strong suggestion that the UK and Ireland on the one hand and the rest of the EEA  on the other hand belong to the same geographic market.

3.2.3.            Conclusion

178.   However, for the purpose of the present decision, it can be left open whether the exact definition of the geographic market as regards the sale of English language books to dealers comprises the entire EEA or whether the UK and Ireland together and the remaining countries of the EEA constitute separate geographic markets given that the proposed transaction does not raise competition concerns under any of these geographic market delineations.

3.3.          Competitive assessment

3.3.1.            Market structure

Overall market for the sale of English language books

179.  Five major publishers98 and numerous mid-sized and smaller publishing houses currently compete in the market(s) for the sale of English language books to dealers in the UK and Ireland and in other EEA countries.

a)  UK and Ireland

180. Random House and Penguin are currently among the three largest publishing houses, together with Hachette. The proposed transaction would create the market leader in the publishing industry in the UK and Ireland with Hachette remaining the second largest publisher in the UK and Ireland. The structure of the markets/ segments constituting the overall market for the sale of English language books, namely the print books, e-books and audiobooks markets/ segments, is presented in detail in the sections below.

b) EEA-wide market excluding UK and Ireland

181.  Both Random House and Penguin also sell their own English language books in  other EEA countries. As regards the EEA sales excluding the UK and Ireland, the parties submit that the market shares in the UK and Ireland are a reasonable proxy for the EEA as a whole as they capture the vast majority of original English language books sales. Moreover, Random House’s and Penguin’s export sales would be a derivative form of exploitation - success in the home market will lead to demand in export markets. Conversely, if a book fails in the home market, it would be highly unlikely to succeed in export markets.

182. The parties submit that while highly reliable data for English language book sales in other EEA countries does not appear to be available, they estimate, on the basis of publicly available export data from the UK and the US and a number of assumptions,99    that  the  total  size  of  the  EEA  English  language  book      market excluding UK/Ireland in 2011 was […] by value and […] by volume.100 On this  basis, the parties would have a combined share of the sales of English language books in the EEA excluding UK/Ireland of [10-20]% by value and [10-20]% by volume in 2011. The parties note that they would not expect that Random House’s and Penguin’s respective share of supply of English language books in the rest of the EEA to be […] lower than their share in the UK and Ireland, and therefore still consider that the best proxy for market shares in English language book publishing  in the rest of the EEA are the competitive conditions observed in the UK and Ireland (which in any event account for the […] majority of each publisher’s English language book sales in the EEA).

183. The market investigation provided mixed results as to whether successful sales in the UK and Ireland determine the success of an English language book in other EEA countries.101 Success in the UK and Ireland of any given title does not necessarily correlate with successful sales in other EEA countries depending on such factors as, for example, subject matter, publicity and whether a local language translation is available. Moreover, success in the US may also help sales in the rest of the EEA. In the words of one publisher, "in broad terms, though, as long as the subject matter (if non-fiction) is not UK-specific, it normally helps the sales in the EEA and for fiction, good selling UK titles stand more chance of selling more successfully in the EEA."¹⁰²

184.  Given that reliable data for sales in the rest of the EEA is not available, the Commission considers that market shares in the UK and Ireland are a reasonable proxy for the EEA as a whole. First, 85% by value of sales in the EEA-wide market for the sale of English language books in 2011 occurred in the UK and Ireland. Second, the views expressed in the market investigation support that, on a more aggregated basis (rather than on a per-title basis), success of sales in the UK/ Ireland may also reflect successful sales in the rest of the EEA.

185. In any event, on the basis of the parties' best estimates of the size of the non- UK/Ireland segment, the Commission notes that the merged entities' share in the rest of the EEA is significantly lower ([10-20]% as compared to [20-30]% in 2011). The market investigation does not contradict this finding and in particular it does not suggest higher market shares of the parties for this segment. Therefore, it can be reasonably assumed that the merged entity's share is unlikely to be higher in the rest of the EEA than in the UK and Ireland. On this basis, the competitive assessment on the basis of market shares in the UK and Ireland would also be applicable to the rest of the EEA even if the rest of the EEA were to constitute a separate relevant geographic market.

Market for the sale of English language print books

186. For the market for the sale of English language print books, the parties submitted market shares based on "bookscan" data compiled by industry analyst Nielsen that, according to the parties and Nielsen, covers over 90% of print book sales in the   UK and approximately 65% of print book sales in Ireland. As a result, the parties submit that Nielsen bookscan data will not fully reflect sales across the entire spectrum of retail outlets. However, given that Nielsen does not monitor the "long tail" of smaller retailers, in which the share of the parties' sales is likely to be lower relative to their overall share as these outlets tend to sell a higher proportion of niche and specialist books than the larger chains, the parties submit that Nielsen data may provide an upper bound to the parties' shares across the entire consumer market.

187.  The market investigation confirmed that the Nielsen bookscan data is a reasonable proxy for market shares in print book sales in the UK and Ireland. The majority of both publishers and wholesalers/retailers confirmed that the Nielsen bookscan data is the most accurate data available in this respect for both UK and Ireland.103 As  regards Ireland (for which Nielsen's coverage of sales is lower), a responding publisher described the Nielsen bookscan data to "give a broad indication of market shares in Ireland and true market shares are unlikely to be too different". According to the major publisher HarperCollins, the Nielsen "data cover the major retailers in Ireland, but does [sic] not cover the large independent retail sector which exists in Ireland". While pointing out that the “65%” figure likely overstates the coverage of the Nielsen bookscan data in Ireland, HarperCollins still "considers them to be the best available". According to one retailer, the Nielsen bookscan data provides "a good representation of what is being sold in the market even if every independent bookseller is not reporting their data in".104

188.  On the basis of the market investigation and the parties' arguments, the Commission considers that the Nielsen bookscan data are a reasonable proxy for market shares in print book sales in the UK and Ireland.

189. The combined entity accounted for [20-30]% in value ([20-30]% in volume) of print books sold to dealers in UK and Ireland in 2011 and [20-30]% in value in 2012. As for the figures for 2012, the parties submit that these shares are affected by the unprecedented success of the bestselling "Fifty Shades" trilogy.

Table 6: Sales of print books in the UK and Ireland

Sources: Nielsen bookscan data;

* As parent of Penguin Random House, Pearson will continue its educational business, including very limited trade book publishing.

190. As for English language print book sales in the entire EEA, the Commission considers, on the basis of the conclusion in paragraph 184 above, that in the absence of more reliable data market shares in the UK and Ireland are a reasonable proxy for the EEA as a whole (including the various sub-segments/ markets). In any event, the parties submit that on the basis of the methodology discussed in paragraph 182 above, the total segment size of EEA sales excluding UK/Ireland in 2011 was EUR […] by value ([10-20]% of the total market value in Ireland and UK) and […] by volume. On this basis, the merged entity's volume share in 2011 would be [10-20]%.

191.  As regards the various Lagardère categories of titles, in the UK and Ireland, the merged entity would have a market share (in value) exceeding 15% in the following categories in 2012 and 2011, respectively: general literature ([30-40]% and [30- 40]%), children's books ([20-30]% and [20-30]%), guides and manuals ([20-30]% and [20-30]%) and reference works ([10-20]% and [10-20]%). The table below lists the combined market shares of the merged entity in the various markets/segments, including those that are not affected by the proposed transaction due to market shares lower than 15%, notably art books, strip cartoon albums, educational books and academic and professional books.

Table 7: The merged entity's combined share by value in the various content markets for physical books to dealers in the UK and Ireland, by value, 2009-2012

Source: Nielsen Bookscan

192. When sub-dividing the English language print book market by sales channel, the merged entity's 2011 value shares are [20-30]% of sales to chain booksellers (such as Waterstones), [30-40]% to mass market retailers (such as supermarkets Tesco or Asda) ([20-30]% Random House and [10-20]% Penguin), [10-20]% to independent bookstores, [10-20]% to internet retailers such as Amazon and [10-20]% to other dealers.105

Table 8: The merged entity's combined share in the various sales channels for the sale of print books to dealers in the UK (*), by value

Source: Bowker Survey data

*The parties submit that there is no data by sales channel available for Ireland.

Market for the sale of English language e-books

193.  As regards the possible e-books market, the parties submit that the only available data to them - the Bowker Books and Consumers Survey 2012 which is based on online interviews of 3 000 book buyers from January to September 2012 - has a number of shortcomings.106 The parties further submit that sales from Amazon's online platform selling into the UK and Ireland would represent a more accurate estimation of market shares given that they account for approximately [90-100]% of all e-books sales in the UK and Ireland.

194. The Commission considers that Amazon sales through its .co.uk domain is a good proxy for the parties' shares for English language e-books in the UK and Ireland as well as for the rest of the EEA and requested from Amazon e-books sales data to compute market shares. According to this data, the parties' combined share in e- books sales amounted to [20-30]% in value in 2011 ([10-20]% in volume), and [20- 30]% in value in 2012 ([10-20]% in volume).

195. As for English language e-book sales in the entire EEA, the Commission considers, on the basis of the conclusion in paragraphs 184 above and 186 above, that market shares in the UK and Ireland are a reasonable proxy for the EEA as a whole as well as for the EEA excluding Ireland and the UK. In any event, the parties submit that on the basis of the methodology discussed in paragraph 182 above, the total segment size of EEA sales excluding the UK and Ireland in 2011 is EUR […] by value and […] by volume in 2011. On this basis, the merged entity's volume share would be [20-30]% (according to their own sales data).

196.  As regards the various Lagardère categories applied to Amazon's e-books sales through its .co.uk domain in the years 2012 and 2011 respectively, the merged entity had market shares exceeding its overall share in e-books only in the categories general literature ([20-30]% and [20-30]%) and in children's books ([20-30]% and [30-40]%).

Table 9: The merged entity's combined share in the various content markets for e-books sales to dealers in the UK and Ireland, by value

Source: Amazon's sales data; Commission's calculations

Market for the sale of English language audiobooks a) UK and Ireland

197. The parties submit that according to Bowker Books and Consumer Survey data that is only available for January to September 2012, the merged entity's combined volume share of audiobooks amounted to [10-20]% (Random House at [10-20]% and Penguin at [0-5]%) and [10-20]% by value (Random House at [10-20]% and  Penguin at [0-5]%).107 The parties further submit that it is not possible to present market shares split by physical audiobooks and digital audiobooks for the Bowker data as the data are supplied on a combined physical and digital audiobooks basis only.

198.  The parties further submit that Bowker data for audiobooks is more reliable than alternative data sources, notably a combination of data from Nielsen bookscan and the Publishers' Association, since its coverage is more exhaustive. In particular, the digital audiobooks sales information provided by the Publishers' Association would focus on traditional publishers and, therefore, include the sales of only a subset of UK digital audiobook publishers which were active in the audiobooks segment. In particular, it would exclude the BBC Group (AudioGo), Audible publishers Brilliance (part of the Amazon group) and WF Hawes (which trades as Whole Story Audiobooks), all of whom the parties consider to be key competing publishers of audiobooks and which represented at least 10% of the overall audiobook market by value. Therefore, the parties submit that the merged entity's combined market share based upon a combination of data from Nielsen bookscan and the Publishers' Association amounting to [10-20]% in 2011 in the UK and Ireland, overstates their market share and is not an accurate reflection of their market position. The merged entity's 2011 share in physical audiobooks in the UK and Ireland would be [10-20]% whereas its share in digital audiobooks – which would be overstated according to above considerations – would amount to [20-30]%.

199.  The Commission considers that on the basis of the Bowker data, which appears to be more reliable, the overall audiobooks market in the UK and Ireland would not be affected. While based on this data source, market shares for physical and digital audiobooks are not available, the Commission considers that the alternative data based on Nielsen and the Publisher Association data shows that physical audiobooks would remain unaffected for the purposes of assessing the proposed transaction, and that the merged entity's market share for digital audiobooks – according to imperfect data – is [20-30]% or likely lower.

200. Given that only the potential market for digital audiobooks would be affected, that,  as noted in footnote 65 to paragraph 144 above, only a small minority of the respondents consider physical and digital audio books to constitute separate markets, that the market share for digital audiobooks is in any case below 25% and that the market investigation did not reveal any concern with respect to audiobooks in  general or digital audiobooks in particular, the Commission considers that the proposed transaction is not likely to result in an impediment to effective  competition in the possible audiobooks market and its sub-segments in the UK and Ireland and therefore does not need to be further assessed in this decision.

b) EEA-wide market excluding UK and Ireland

201.  The parties submit furthermore that there are no reliable data sources for English language audiobook sales in the rest of the EEA excluding the UK and Ireland. Notwithstanding the lack of reliable data, the parties' best estimates of the total size of the EEA English language audiobook market (excluding the UK and Ireland) in 2011 are […] by volume. On this basis, the parties would have a combined share of the sales of English language audiobooks in the EEA (excluding the UK and Ireland) of [5-10]% by volume (Random House [0-5]%, Penguin [0-5]%) in 2011.  The parties' combined shares by value are estimated to be [5-10]% (Random House [5- 10]%, Penguin [0-5]%). However, the parties consider the estimate of the total market size by value to be less reliable than the estimate of the total market size by volume.

202.  The Commission considers that on the basis of the parties' best estimates, the overall audiobooks market in the EEA excluding the UK and Ireland would not be affected. While no data is available as the sub-segments of physical and digital audiobook sales, the volume of sales of English language audiobooks in the EEA excluding the UK and Ireland represented only [10-20]% of the entire volume of sales in the EEA in 2011. Moreover, there is no reason to believe that the market shares for physical and digital audiobook sales are significantly different in the rest of the EEA than in the UK and Ireland. Therefore, by analogy of the Commission's considerations in paragraph 199 above, the Commission considers that the proposed transaction is not likely to result in an impediment to effective competition in the possible audiobooks market and its sub-segments in the EEA excluding UK and Ireland and therefore does not need to be further assessed in this decision.

3.3.2.            Assessment of non-coordinated effects

203. The Commission investigated whether the proposed transaction would remove an important source of competition in the relevant market(s) and sub-segments for the sale of English language books and may make it more likely that the merged entity could profitably increase wholesale prices for its titles on these markets/segments or foreclose its rivals from accessing certain markets.

204.  The Commission carried out its assessment of the non-coordinated effects of the proposed transaction on the print books and e-books markets/segments which are addressed in turn by the following sections.

Market for the sale of English language print books a) Views of the parties

205.  According to the parties, the proposed transaction will not significantly impede effective competition in respect of the sale of English language print books, for the following reasons:

206.  First, the merged entity's market shares would be modest and several large competitors as well as medium-sized and smaller publishers would continue to exercise significant competitive constraints on the merged entity. Furthermore, the parties are not uniquely close competitors.

207. As regards the market segment of bestsellers, while the six larger publishers might have a higher number of “brand authors” and therefore may generate more bestsellers, medium-sized and smaller publishers will nonetheless also be able to generate bestselling titles and constrain the parties in the segment for bestsellers. In support of this argument, the parties submit that in every week, from 2009 to 2012ytd, at least nine different publishers (including, by necessity, smaller publishers) obtained titles in the Top 100 lists. Additionally, an analysis of the weekly top 100 bestseller lists would also confirm that bestselling titles achieved by the smaller publishers are not ‘one hit wonders’ and that in fact many smaller publishers (and not only the six largest publishers) achieve bestselling titles consistently over time. For example, in 2012, 26 publishers would attain at least two different bestselling titles in the weekly Top 100 lists and 12 publishers would have at least five different bestselling titles in the weekly Top 100 lists. In light of these arguments, the parties believe that post-transaction, the merged entity will continue  to face a large number of competitors (beyond the large publishers) also within the bestsellers segment.

208.  As regards in particular the mass retailing channel, the parties submit that the only retail channel in which access to shelf space is subject to material limitations is in the supermarket channel, where books compete against other product categories for shelf space. The parties explain that all publishers face this competitive pressure equally. Small publishers would not be disadvantaged relative to larger publishers when it comes to securing prominent shelf space, although smaller publishers would, naturally, have fewer titles that they wish to promote in-store.

209.   The parties submit further that the ability of even smaller publishers to compete against larger and medium-sized publishers including in the mass retailing channel is also facilitated by the presence of services such as the Independent Alliance  initiative through which the publisher Faber and Faber provides the sales force and administrative team to sell physical books for other (smaller) publishers.108 Smaller publishers would be able to negotiate with retailers including supermarkets through the Independent Alliance enabling them jointly to market titles of potential interest  to the mass market retailers from a range of publishers.

210.   Second, the parties submit that the book retailing business is increasingly concentrated given that the increased competition from supermarkets and e-tailers over the past decade has substantially affected the viability of independent book retailers. Therefore, the parties submit that the merged entity will also face considerable constraint from the strong position enjoyed by retailers. Random House’s five largest physical book retailers account for [60-70]% of its sales and [50-60]% of Penguin’s sales in 2011. Across all publishers, the parties submit that Bowker estimates the top 5 retailers to account for [60-70]% of volumes sold.

211. The parties further submit that the strength of the bargaining position of major book dealers is not purely a function of their relative importance as customers but also a product of their role as gatekeepers to consumers, with brick and mortar retailers serving as showrooms for books. Publishers would rely […] on the visibility of their titles in-store or on online retail platforms to showcase their books to consumers and thus drive sales. That visibility would be a product of the way in which the retailer chooses to promote the books (which may occur within the framework of an agreement with the publisher, but also will be within the retailer’s discretion). For instance, the largest chain bookseller in the UK, Waterstones, […].

212.  The parties submit that shelf space (particularly promoted shelf space) is considered to be a premium opportunity for publishers and is scarce and fiercely contested by publishers. According to the parties, [description of commercial negotiations with customers].

213.  Third, the parties submit that barriers to entry and expansion across different book categories / geographies are low and declining. As explained in more detail in paragraph 71 above, new technologies such as print-on-demand and digital printing have reduced costs for small-scale publishing so that authors can more easily self- publish their titles and access the market(s) for the sale of print books.

b)  Commission's assessment

Market shares and competitive constraint exercised by competing publishers

214.  The Commission considers that while the proposed transaction would create the market leader, the merged entity's combined market share of [20-30]% in the overall market for the sale of English language print books remains relatively limited. According to the Commission's Horizontal Guidelines, an indication that a concentration is not liable to impede effective competition exists in particular where the market share of the undertaking concerned does not exceed 25%.109 Moreover, three larger and internationally established competitors (Hachette, HarperCollins and Pan Macmillan) will remain in the market. These are followed by numerous mid- sized and smaller publishers that together account for [40-50]% of the overall print publishing market value in 2011 in the UK and Ireland.

215. The Commission investigated whether post-transaction, a sufficient number of competitors would remain active in the market(s) for the sale of English language print books so that it is unlikely that the merged entity could profitably increase prices, for instance, by increasing the recommended retail price ("RRP") for its titles or by reducing the level of base discounts and/or promotional discounts granted to dealers.

216.  Overall, the market investigation revealed that the majority of respondents consider that the merged entity will not be in a position to successfully raise prices in the market(s) for the sale of English language print books.110

Categories of print books

217.  On the basis of the Commission’s Lagardère segmentation of the print books markets, the merged entity would have an aggregated market share   post-transaction in excess of its overall print book market share in the categories of general literature ([30-40]% and [30-40]% in 2011 and 2012 respectively) and children's books ([20-30]% and [20-30]% in 2011 and 2012 respectively). These markets/ segments respectively account for approximately [40-50]% and [10-20]% of the overall value of print book sales in 2012.

218. The parties submit that the relatively higher 2012 market share in the general literature category is largely due to the unprecedented success of the "Fifty Shades" trilogy published by Random House. By comparison, the parties' combined value market share in general literature would vary between [30-40]% and [30-40]% in 2009 to 2011. Finally, the effect of the Fifty Shades trilogy in 2012 would highlight that market shares may be volatile and can be transitorily affected by a few titles  with high levels of sales.

219. The Commission considers that a number of larger competitors remain active in the general literature market, notably Hachette (with a [20-30]% market share in 2011), HarperCollins ([10-20]%) and Pan Macmillan ([5-10]%). Moreover, as below table demonstrates, a number of mid-sized and smaller publishers are also present in this market, including Simon& Schuster, Bloomsbury, Harlequin Mills & Boon, Faber and Faber as well as Quercus.

Table 10: Sale of print general literature books to dealers in the UK and Ireland, by value 2011

Source: Nielsen Bookscan

*As parent of Penguin Random House, Pearson will continue its educational business, including very limited trade book publishing.

220.  As regards the children's books market/ segment, the merged entity's market shares remain below or equal to 30%. Moreover, as shown in below table, a variety of  larger and mid-sized competitors – some of them (such as Egmont, Usborne, and Scholastic) specialised on children's and young adult books, have sizeable shares in the market and are likely to continue to constrain the merged entity.

Table 11: Sale of print children’s books to dealers in the UK and Ireland, by value, 2011

Source: Nielsen Bookscan

221. With respect to the general literature category, a large majority of responding publishers and half of the customers who expressed a view said that the merged  entity would not be in a position to successfully raise prices. The same answer was given by a large majority of each of publishers and customers expressing a view for the other affected possible product markets, notably children's books, guides and manuals, and reference works. 111

222. On the basis of the above, the Commission considers that post-transaction, the merged entity's market shares are not of such magnitude that they would allow the merged entity to act independently of its competitors and cause non-coordinated effects in any of the affected sub-markets. It would likely be constrained by a significant number of competitors, including larger, international publishers. All of these publishers are active across the full range of sub-markets/ segments, operating under a range of well recognised imprints.

223. Furthermore, the Commission considers that there are also a large number of successful medium and small-sized publishers constraining the merged entity. These medium and small-sized publishers accounted for over [40-50]% by volume and [40- 50]% by value of the market in 2011. Some medium and small-sized publishers also have international publishing divisions, including in the USA (for instance, Egmont, Bloomsbury, Scholastic), Australia (Egmont and Bloomsbury) and India (Bloomberg).

224.  While medium and smaller firms will typically publish fewer titles and across fewer categories of titles in any one year than larger publishers, the Commission considers that they may nonetheless act as a competitive constraint on the larger publishing groups in particular in their categories of specialisation (such as children's books). Moreover, while taken individually, these publishers will not necessarily have the range of publications that the larger publishers may have in any one category, they collectively account for a substantial proportion of titles sold.

Sales channels

225. With respect to the sub-division of the market by sales channels, the Commission considers that the merged entity would have limited market shares in the affected possible markets of sales to chain booksellers and Internet dealers (see the table in paragraph 192 above).

226. The only market where the merged entity's combined share significantly exceeds its overall market share in print books of [20-30]% concerns the possible market for the sale to mass market dealers such as supermarkets, with market shares ranging from [30-40]% to [30-40]% between 2009 and 2011. Based on Bowker Survey data, this segment accounts for approximately [5-10]% of the overall value of print book sales in 2011.

227. As shown by the table below, Hachette with market shares ranging from [20-30]% to [20-30]% in the same period, followed by HarperCollins and Pan Macmillan, are other large competing publishers active in that market/ segment. Moreover, mid- sized and smaller publishers account for a quarter of that market.

Table 12: Sale of print books to mass market dealers in the UK (*), by value

Source: Bowker Survey data

*The parties submit that there is no data by sales channel available for Ireland.

228. The Commission investigated, first, whether the merged entity could successfully raise prices to mass market dealers by increasing the RRP for its titles or by reducing the level of discounts to dealers. Second, the Commission investigated whether the merged entity would likely be able to market its titles more effectively through, for instance, a combination of privileged sales contacts and higher discounts (in the form of base discounts, promotional discounts, "range space" discounts, and/or retrospective discounts) to mass retailers in order to foreclose its rivals from effectively accessing supermarkets' "shelf space", thereby increasing its position on the possible market for the sale of English language books to mass retailers.

229. As regards the first concern, a large majority of responding publishers who  expressed  a  view  said  that  the  merged  entity  would  not  be  in  a  position  to successfully raise prices in any of the sales channels discussed in paragraph 192 above, including with respect to supermarkets.112 A supermarket operator in the UK who responded to the market investigation did not raise any concerns with respect to the proposed transaction.113

230.   As regards the second concern, the market investigation confirmed the parties claim with respect to the Independent Alliance run by Faber and Faber. It has a sales team dedicated to selling, on a commission basis, the books of currently 7 small UK publishers including to supermarkets.114 The market investigation also showed that other smaller publishers employ sales personnel specifically dedicated to contacts with supermarkets.115

231.  Moreover, the Commission considers that unlike in the Lagardère case, where the merging parties undertook sales services on behalf of a large part of publishers and therefore had the ability (and incentive) to foreclose these competitors, post- transaction, from accessing sales channels (or gaining a privileged position on shelves),116 the parties are not active in providing sales services for third party publishers. Any privileged position with respect to shelf space would need to be obtained through additional discounts paid to retailers, which would, if anything, lead to a pro-competitive outcome (that is, potentially cheaper books). Moreover,  any such hypothetical foreclosure strategy from the mass retailing channel would be costly for the merging parties in the short to medium term as they would need to  fund payments to retailers. This implies that, financially speaking, the strategy would only make sense if the initial investment in foreclosure could be recouped over time. Therefore, a hypothetical foreclosure strategy of competitors would appear risky from a commercial perspective, in particular as a significant marginalisation over time of rival publishers on the possible market for sales to supermarkets appears unlikely in the light of a market shares of between [30-40]%-[30-40]% (2009-2011).

232. On the basis of the above, the Commission considers that post-transaction, the merged entity is unlikely to be in a position to successfully and profitably raise wholesale prices across the various retailing markets for print books. There would remain a significant number of competitors, including large publishers in all sub- markets/ segments who are likely to provide a competitive constraint on the merged entity. Furthermore, the merged entity would be unlikely to have the ability to foreclose its rivals from effectively accessing the mass retailing sales channel.

Bestsellers

233. As noted in paragraph 163 above, the market investigation confirmed the parties' submission that there is no generally accepted and used definition of the bestseller segment of the market(s) for print books. The parties submitted their combined share for those titles that (according to the volume of sales) fell into the Nielsen  Bookscan categories of (i) weekly top 100 bestselling titles, (ii) weekly top 200, (iii) weekly top 500, and (iv) weekly top 1000, (v) annual top 100, (vi) annual top 200, (vii) annual top 500, and (viii) annual top 1000.

234. Compared to the overall market for English language print books, the merged  entity's combined share is relatively higher in the various bestseller segments,  ranging for 2011 and 2012ytd (January –October) respectively, from [50-60]% and [60-70]% in value in the annual top 100 bestselling books (the smallest sample) to [40-50]% and [50-60]% in value in the weekly top 100 (a medium-sized sample of bestselling titles) and [30-40]% and [40-50]% in value in the weekly top 1 000 bestselling books (the largest of the seven samples mentioned above), as shown in the following tables.

Table 13: Print bestseller shares - Annual top 100, by value

Source: Nielsen Bookscan; the parties' calculations

*2012ytd corresponds to the period from January to October 2012.

Table 14: Print bestseller shares - Weekly top 100, by value

Source: Nielsen Bookscan; the parties' calculations

*2012ytd corresponds to the period from January to October 2012.

Table 15: Print bestseller shares - Weekly top 1000, by value

Source: Nielsen Bookscan; the parties' calculations

*2012ytd corresponds to the period from January to October 2012.

235. The Parties submit that the figures for 2012ytd are affected by the unprecedented success of the bestselling Fifty Shades trilogy as is demonstrated by significantly lower bestselling shares in the years 2009-2011 under all three definitions. The Commission could confirm this claim for the annual top 100 bestseller data. The unprecedented success of the Fifty Shades trilogy alone accounted for [40-50]% of the 2012 annual top 100 sales value. This would suggest that the success of one or several titles can significantly alter yearly shares in bestsellers. Without the Fifty Shades trilogy, the 2012 annual top 100 bestseller share of the merged entity would be [20-30]%. In addition, it should be noted that the bestseller segments constitute only a small part of the print books market. Notably, the annual top 100 accounted for [5-10]% and [5-10]% of the sales value in 2011 and 2012 respectively.

236. Moreover, the market investigation revealed that a majority of publishers and customers expressing a view considered that the merged entity will not be in a position to successfully raise prices for bestsellers post-merger.117

237. In light of the above, the Commission is of the view that  post-transaction,  the merged entity is unlikely to be able to successfully and profitably raise wholesale prices for print bestsellers. Other major publishers are likely to constrain the merged entity. Moreover, smaller publishers are also a constraint given that publishers other than the five largest ones also generated bestsellers in the past, accounting for between at least [10-20]% (for the annual top 100 in 2011) and [20-30]% (for the weekly top 1000 in 2011 and 2012) of the bestseller sales in the period 2009 to 2012ytd. In any event, the Commission recalls its conclusion in paragraph 169 above that the market investigation did not confirm that bestsellers are a separate market  but that most dealers make their choices across the full range of bestsellers and non- bestsellers.118

Closeness of competition

238. The market investigation revealed that while the merging parties are close competitors in the print books market(s), respondents to the market investigations do not consider them uniquely close competitors to each other in the print books market(s). Notably, Hachette is ranked by at least as many publishers expressing a view and a majority of customers as the closest competitor to Random House and Penguin, respectively, in both, the UK and Ireland. Furthermore, all responding publishers and almost all responding customers rank HarperCollins and Pan Macmillan in the top 5 of close competitors of each of Random House and Penguin.119 There therefore appears to be a high degree of similarity between the portfolio of books offered by the parties and those supplied by these other major publishers.

Categories of print books

239.  As regards the categories of titles, the market investigation provided mixed results as to whether Penguin and Random House are each other's closest competitors in any of the sub-markets. A majority of responding customers perceive the parties as closest competitors in general literature, although some responding customers relativised their answer. For instance, the wholesaler Bertram pointed out: "While not  the closest competitor in many of these categories, they would be within [the top] 2-3 in the vast majority of cases." The Irish book retailing chain and wholesaler Eason and Son noted: "Both publishers have very strong and varied lists and, again, would  have no great advantage or disadvantage over any other UK-based publisher in the market." The retailing chain WHSmith pointed out that in "Fiction and Non Fiction [which is part of the general literature market as defined by the Commission], Random House and Penguin are number 1 and 3 in the market." Similarly, a publisher pointed out that based on Nielsen bookscan 2012 market share figures, Random House ranked first and Penguin fourth in general fiction (which is part of the general literature market as defined by the Commission) and that in children's books, Penguin ranked first and Random House fifth. Another publisher responded that "Penguin and Random House are close competitors across a number of categories but they are equally competitive with Hachette, HarperCollins and […] other publishing companies". As regards children's books, Usborne (a publisher specialised on children's books) noted that "we do not consider Random House or Pearson to be each other's closest competition".120

240.   On the basis of the above, the Commission takes the view that while the parties are certainly well described as close competitors, they are not each other's closest competitors in the overall print book market or any of the possible sub-markets delineated by the Lagardère content categories and there remain a number of similarly strong competitors. As regards general literature, Hachette is a  considerably stronger competitor to Random House than Penguin, and HarperCollins is a close runner-up to Penguin in terms of 2011 market shares. As regards children's books, while Penguin is the clear market leader, Hachette is the second largest publisher in this category, followed by HarperCollins, Random House, Egmont, Pan Macmillan and Usborne which all have enjoyed similar levels of sales in 2011.

Bestsellers

241. The Commission also investigated the parties' position with respect to bestsellers and their relative closeness of competition in order to determine whether the proposed transaction would remove an important source of competition in this market  segment.

242.   The market investigation provided mixed results as to whether Penguin and Random House produce close substitutes in the market segment of bestsellers.121 Half of the dealers expressing a view considered that Penguin and Random House are each other's closest competitors in bestsellers.122 Competing publishers did not, however, consider the parties to be the closest competitors in bestsellers.123

243.   In light of the above, the Commission considers that whilst the parties can be described as close competitors in bestsellers, they are not uniquely close competitors. While Hachette's share in bestsellers has been declining over the past four years, it retains an important share in the bestsellers segments, followed by HarperCollins. Furthermore, large jumps in shares in bestsellers can be seen over the years and for the majority of publishers (most notably for Random House in 2012ytd, with its share [increasing] in the annual top 100 largely due to the success of the Fifty Shades trilogy). This would suggest that the success of one or several titles can significantly alter yearly shares in bestsellers. For these reasons, the relatively higher shares in the bestseller segment would not, per se, mean that the parties are uniquely close competitors and that the removal of competition amongst them would give raise to competition concerns.

244. In any event, the Commission recalls its conclusion in paragraph 169 above that the market investigation did not confirm that bestsellers are a separate market but that most dealers make their choices across the full range of bestsellers and non- bestsellers.124 Indeed, almost all customers responded that a publisher does not have to offer bestsellers in his catalogue in order to win them as customers.125 Therefore, even  if the parties were uniquely close competitors in this segment in which they hold significant shares, this would not mean that they could raise prices or otherwise  leverage their position (as discussed in paragraph 236).

Countervailing buyer power

245. The market investigation suggests that larger print books customers enjoy a certain degree of countervailing buyer power that they do exercise. Many responding publishers report that their top 5 customers account for well beyond 50% of their sales in the UK, and beyond 75% in Ireland.126 A majority of competitors consider that their print book customers typically exercise their bargaining power during commercial negotiations by threatening to switch to other publishers, in other words: delisting (certain of) a publisher's titles; threatening to stop purchasing other products provided by them and, to a lesser extent, threatening to delay purchases from them.127 A large majority of responding competitors find these threats credible.128

246. In the light of the above, the Commission considers that the major resellers of print books appear to be commercially significant to publishers. Given the results of the market investigation revealing high shares of the top 5 customers and that the shares of the parties' own top 5 customers are high, the Bowker estimate submitted by the parties according to which the top 5 retailers account for [60-70]% of volumes sold appears broadly realistic. Moreover, as concluded in paragraphs 220, 232, and 238, there are numerous credible alternative publishers in all markets/ segments, to which customers can easily switch (part of their) orders. Finally, publishers indicated in the market investigation that important customers are able, and regularly do, use certain credible bargaining techniques, notably with respect to delisting of (certain) titles of a publisher. Such techniques would appear to be even more credible for mass retailers (such as supermarkets) who are not specialised on selling print books only. For all these reasons, the Commission considers that large and sophisticated dealers, in particular supermarkets, have a certain degree of countervailing buyer power that is likely to remain effective post-transaction.

Barriers to entry

247.  The market investigation indicates that there are appreciable barriers to enter the print books market relating, for example to scale in order to access major retailers as well reputation and brand recognition of imprints.129 At the same time, the investigation confirms that a number of publishers did successfully enter the market in the more recent past, including Quercus (entering in 2004 and now among the top 10 publishers) but also Head of Zeus, Igloo, and Nosy Crow in children's books.130

248.  The parties' claim regarding the increased ease for authors to access the market through self-publishing was not confirmed by the market investigation. The vast majority of responding publishers did not consider that digital printing-on-demand (for smaller print runs) and self-publishing are an effective alternative to access the market of the sale of English language print books.131

c)  Conclusion

249.  In light of the above, the Commission concludes that the proposed transaction does not raise serious doubts as to its compatibility with the internal market in each of the possible relevant English language print book markets as a number of credible alternative competitors, including numerous smaller competitors, remain post- transaction.

Market for the sale of English language e-books a) Views of the parties

250.  The parties submit that their combined market share in the sale of English language e-books is modest. Based on third party survey data,132 their combined share amounted to [30-40]% in the period from January to October 2012. In that period, Random House would hold a market share of [20-30]% and Penguin of [5-10]%. Furthermore, the merged entity would continue facing competition by numerous competitors who are active across print and e-books, including the four larger international publishers Hachette, HarperCollins, Pan MacMillan and Simon & Schuster, as well as medium and smaller publishers.

251. In addition to competitive constraints exercised by traditional publishers, the merged entity would face further competition by self-published titles and free titles. The parties explain that in particular self-published e-books have had significant commercial success in recent years. They illustrate this by the success of self- published e-books in the United States, where the number of self-published e-book titles on the US Daily Kindle Top 100 list from January to November 2012 exceeded that of any individual publisher, accounting for [10-20]% of the total unique title share. In the UK and Ireland, [5-10]% of e-books were self-published in 2011. The parties add that Amazon recently revealed that 15% of its bestselling e-books in  2012 were written by self-published authors. They further explain that free books which are typically new books that are given away for free by authors in order to raise their popularity play a particular role on the market for e-books. According to the parties, there are currently over one million free e-book titles available on the Internet. They put forward that according to consumer research carried out by Bowker, consumers substitute between paid for books (both physical and e-books) and free e-book downloads.

252.     In particular as regards bestselling e-books titles, the parties submit that an analysis of the Kindle Top 100 charts published in real time by Amazon shows that smaller publishers maintained a strong presence in the e-book top selling lists and  particularly in the free e-book lists: for paid e-books the six larger publishers accounted for between [40-50]% and [60-70]% of the titles in the top 100 in 2012 (with data only being available for 2012); and for free e-books the six larger publishers only accounted for between [0-10]% and [0-10]% of titles in the top 100  in 2012 (with data only being available for 2012). Additionally, the number of publishers with titles in the paid e-book top 100 list ranged from […] to […] over the period surveyed and the number of publishers with titles in the free e-book top 100 ranged from […] to […]. As with physical books, smaller publishers’ successes would not be limited to one hit wonders as […] publishers had at least two  bestselling titles in the paid e-book top 100 lists and […] publishers had at least five bestselling titles in the paid e-book top 100 lists. In the free e-book top 100 lists in 2012, […] publishers had at least two bestselling titles and […] publishers had at least five bestselling titles.

253.    The parties submit moreover that the barriers to enter the publishing of e-books would be low given the recent emergence of self-publishing and that, as in the  market for the sale of print books to dealers in the UK and Ireland, the parties would not be uniquely close competitors as regards the sale of e-books.

254.   Finally, the parties put forward that buyer power is particularly pronounced in the e- books market given that the distribution business is virtually controlled by a single platform (Amazon). They further explain that dealers for e-books in the UK and Ireland form an even more concentrated customer base. Four dealers account for [greater than 75%] of Random House’s e-book sales in the UK and Ireland in 2012. In particular, Amazon alone accounted for [greater than 75%] of Random House’s sales and [greater than 75%] of Penguin’s sales, and an estimated [greater than 75%] of all e-books sales to consumers in the UK and Ireland in 2012 and would therefore have clear and unambiguous buyer power. The parties believe that Amazon would maintain its strong position in the e-book segment of the book market, which they consider represents more than just a first-mover advantage.133

b)  Commission's assessment

Market shares and competitive constraint exercised by competing publishers

255.  As outlined in more detail in paragraphs 193 to 196 above, the parties' market share in the overall market for the sale of English language e-books in the UK and Ireland (measured as a proxy by their sales through Amazon's .co.uk domain) amounted to [20-30]% by value in 2012 and [20-30] by value in 2011.

Categories of e-books

256. As regards the various Lagardère categories applied to Amazon's e-books sales through its .co.uk domain in the years 2011 and 2012 respectively, the parties' combined market share by value in general literature ([20-30]% and [20-30]%) are hardly higher than its overall market share in e-books. In children's books, the merged entity accounted for ([30-40]% and [20-30]%).

257. As such, the merged entity's market shares on the overall market for the sale of English language e-books as well as in the various Lagardère categories will remain at levels, which by themselves do not raise a risk of non-coordinated effects on competition.

258.  Moreover, in the market/ segment of general literature (see the following table), the merged entity will continue facing competition by the several other  large international publishers, a number of medium-sized publishers and a very long tail of small publishers who also successfully entered the emerging e-books market.

Table 16: Annual market shares for e-books sales (General Literature) sold through amazon.co.uk, by value

Source: Amazon's sales data; Commission's calculations

259.  In the market/ segment of children's e-books (see the following table), the merged entity will continue facing competition by several larger, number of medium-sized and a long tail of smaller publishers. [20-30]%

Table 17: Annual market shares for e-books sales (Children's Books) sold through amazon.co.uk, by value

Source: Amazon's sales data; Commission's calculations

260. The vast majority of competitors who participated in the market investigation considered that the merged entity will not be in a position to successfully raise e- book prices in the various Lagardère categories.134 The evaluation of the risk of possible post-transaction price increases given by customers were more nuanced. While the majority of customers who replied in the market investigation did not foresee a risk of post-transaction price increases in the categories of guides and manuals, reference works, art books, strip cartoon albums, and academic and professional books, a slight majority of respondents expressed the view that post- transaction price increases were possible for general literature e-books. However, the vast majority of respondents who feared a price increase in general literature e-books did not provide any further substantiation and only two respondents made a general remark as to the loss of a strong competitor and possibly high combined market shares in the field of general literature. As regards children's e-books, the same number of respondents indicated that the merged entity could successfully increase prices, as the number of respondents who answered that this was not the case.135

261. The Commission further notes that e-books contract terms are generally negotiated across all categories of titles offered by a publisher and discounts off the electronic RRP do not differ among categories of books.

262.  In the light of the above, it appears unlikely that the proposed transaction would result in the merged entity's ability to successfully raise prices within a specific category of titles.

Bestsellers

263. As to e-book bestsellers, the parties' combined market share is relatively higher in  the various bestseller segments compared to the overall market for English language e-books, but lower than for print book bestsellers. In 2012, the parties accounted for [40-50]*% of the sales value of the annual top 100 bestselling e-books and for [30- 40]*% of the weekly top 100 bestselling e-books.136

Table 18: Annual market shares for annual TOP 100 eBooks (*) sold through amazon.co.uk, by value

Source: Amazon's sales data; Commission's calculations

(*) TOP100 defined as titles which were in the top100 by year i.e. titles are ranked by annual sales and subsequently the top 100 list is extracted by year.

Table 19: Annual market shares for weekly TOP 100 eBooks (*) sold through amazon.co.uk, by value

Source: Amazon's sales data; Commission's calculations

(*) TOP100 defined as titles that were in the top 100 in any week, and are also treated as a bestseller in all other weeks of that year.

264.  However, a closer analysis of the bestseller data reveals that the unprecedented success of the Fifty Shades trilogy alone accounted for [20-30]% and [10-20]% of the 2012 annual and weekly top 100 shares, respectively. This would suggest that the success of one or several titles can significantly alter yearly shares in bestsellers. The 2011 bestseller shares for the annual and weekly top 100 were significantly lower ([20-30]% and [20-30]% respectively). Without the Fifty Shades trilogy, the 2012 bestseller shares would be equal or below the 2011 bestseller shares, which in turn hardly exceed the parties' overall shares in the overall e-books market in 2011 [20- 30]% and the general literature segment [20-30]%.

265.Moreover, the Commission considers that whilst the e-books market/ segment is still emerging and data is not available over several years, the three major international publishers in 2012 achieved very similar shares, taken together, in bestsellers as in the overall e-books market and in the general literature category and their 2011 bestseller shares significantly exceeded their overall and general literature shares. This would suggest that the major publishers are at least as capable of generating e- books bestsellers as the merged entity. Finally, between a quarter and more than a third of e-books bestseller sales are generated by publishers other than the five major publishers.

266. For these reasons, the parties are not uniquely close competitors in the e-books bestseller segments and the removal of competition amongst them would not give rise to competition concerns. In any event, the Commission recalls its conclusion in paragraph 169 above that the market investigation did not confirm that bestsellers  are a separate market but that most dealers make their choices across the full range of bestsellers and non-bestsellers.137

Closeness of competition

267. Similarly to the results for print books, the market investigation revealed that while the merging parties are close competitors in the e-books market, respondents to the market investigations do not consider them uniquely close competitors to each other. Notably, Hachette is ranked by a majority of customers as the closest competitor to Random House and Penguin, respectively, in both, the UK and Ireland.138 The answers of competing publishers are more nuanced.139 However, taking into account that all responding publishers and almost all responding customers rank HarperCollins and Pan Macmillan in the top 5 of close competitors of each of Random House and Penguin, there appears to be a high degree of similarity between the portfolio of titles offered by the parties and those supplied by these other major publishers. For these reasons, the Commission does not consider the parties to be uniquely close competitors.

Barriers to entry

268. The market investigation confirmed that barriers to entry for the sale of English language e-books are lower than for the sale of print books.140 As explained by one publisher, "the market conditions for e-book publishing are even more favorable towards entry than those for print books. This is due to the fact that e-book publishing does not involve any of the significant printing and storage costs associated with print book publishing." Several competitors mentioned that the main barrier to entry in e-book publishing relates to the discoverability of a new entrant and its access to e-tailers. Amazon and Pottermore were mentioned as recent entrants in the context of the publishing of e-books. However, the parties' argument that self- publishing is an alternative route for authors to access the market was not unambiguously confirmed by the market investigation as a majority of competitors expressing a view did not consider so.141 The main explanation given for this relates to the issue of discoverability and the need for professional marketing.

Countervailing buyer power

269.     The market investigation confirmed that retailers of English language e-books have buyer power. All publishers who responded to the respective question indicated that Amazon was their most important sales platform by far for English language e- books.142

270.    Asked about the way in which customers of the supply of English language e-books typically exercise their bargaining power during commercial negotiations, several competitors indicated that they experienced threats to switch to other publishers; threat to stop purchasing other products; threat to sponsor expansion of other publishers and other forms of threats such as making titles undiscoverable or withdrawing the "buy button" from the webpage for a publisher's e-books. The majority of respondents consider these threats to be credible.143

271.  Given the concentrated retail market structure and that publishers believe that  various threats that customers can make are credible, the Commission considers that large and sophisticated e-book retailers are likely to enjoy a certain degree of countervailing buyer power that is likely to remain effective post-transaction. However, as opposed to print books where shelf space limits retailers in stocking books, e-book retailers do not face such constraints. E-book stores aim to offer the widest possible range of e-books across all categories144 as one of the parameters likely determining competition for consumers is the breadth of the available content.

c)  Conclusion

272.  In light of the above analysis, the Commission concludes that the proposed transaction is unlikely to result in an impediment of effective competition as regards non-coordinated effects in each of the possible relevant English language e-book markets as a number of credible alternative competitors including numerous smaller competitors remain post-transaction.

3.3.3.            Coordinated effects

Views of the parties

273.  The parties submit that the transaction would not create any risk of coordinated effects in relation to the sale of print books, e-books or audio books to dealers.

274.  First, according to the parties, the large number of competitors, together with the asymmetry of respective publishers’ market shares would undermine any incentive and ability to coordinate publishers’ activities tacitly. Achieving or sustaining any coordinated outcome would be difficult in light of the large number of smaller, medium-sized and large publishers active on the market. Furthermore, market shares  of the major publishers are asymmetric, which increases the difficulty of reaching a coordinated agreement. The asymmetry will increase post-transaction in both the physical and the digital segments. Finally, each publisher has a very large stock of  titles. Agreeing on and monitoring a coordinated strategy across such a large number of titles would be implausible.

275.  Second, the parties submit that the potential for all publishers to (unpredictably) release bestseller titles which fundamentally affect the demand across all books would render tacit coordination more difficult to sustain. Publishers whose sales are negatively affected by a blockbuster title would have a much stronger incentive to deviate from coordination to increase their sales.

276. Third, monitoring of a coordinated agreement on print books would be extremely difficult due to a lack of price transparency. For print books, the relevant net wholesale price paid by a retailer to a publisher is comprised of a number of elements, including promotional discounts, volume rebates and co-op spend, and it varies by retailer, by  title and over time. Discounts and co-op spend are negotiated between retailers and publishers individually and are not publicly observable. Thus, coordination on the RRP of a title would be ineffective as any agreement to set high RRPs could be offset by offering steeper discounts and increased promotional spend.  Because  these discounts can vary by retailer, by title and are not publicly observable, it would be impossible for publishers to detect deviation from a coordinated agreement.

277.   Furthermore, it would be implausible for publishers to infer the level of wholesale pricing from the retail pricing (i) due to the retailers' ability to flex their own margins and (ii) due to the fact that co-op spend cannot be deduced from retail prices and monitoring its effects would be unfeasibly costly.

278. The parties also put forward that there is no risk of coordination within different retail channels, and in particular supermarkets where the parties have higher market shares. Despite the parties' higher market shares, the number of publishers selling through the mass market channel remains the same. Also, all publishers have a strong incentive to introduce their suitable titles to the mass market to drive volume and customer awareness while at the same time the number of titles supermarkets sell is relatively limited. Furthermore, pricing discussions with retailers are particularly complex in the mass market channel, prices are not transparent, there are strong incentives to deviate due to mass market retailers' ability to offer a platform for a book across a very large number of stores and deviation cannot be monitored and, finally, the mass market retailers' significant countervailing buyer power would undermine any attempt at coordination.

279. Fourth, monitoring of a coordinated agreement on e-books would be very difficult due to a range of pricing models used in the industry, the lack of price transparency, and a multiplicity of products and prices that would need to be monitored. Even under the agency model where the price set by the publisher is directly observable at the retail level, e-books, in comparison to RRPs of physical books are sold at a large number of price points. Moreover, in digital publishing there exists a variety of distribution  models. Different publishers may be setting (unobservable) wholesale prices through some distribution channels and retail prices on others.

280. Furthermore, the parties submit that the settlement agreed with the Commission and the US Department of Justice ("DOJ") by various publishers […].

281. Fifth, publishers face strong external constraints in the form of very powerful dealers which could defeat any attempt at coordination.

Commission's assessment

282.  To assess coordinated effects, well-established case law145 and Commission guidelines on the assessment of horizontal mergers146 require proof that the merger will make coordination more likely, more effective and more sustainable. The analysis needs to focus in particular on: (i) the ability to reach terms of coordination; (ii) the ability to monitor deviations; (iii) the existence of a credible deterrent mechanism if deviation is detected; and (iv) the reactions of outsiders such as potential competitors and  customers.

283.  The Commission takes into account the parties' argument that the large number of competitors, together with the asymmetry of respective publishers’ market shares undermines any incentive and ability to coordinate publishers’ activities tacitly. It considers this a valid argument to the extent that coordination among all publishers active in the market seems rather unlikely given the asymmetry of market shares.

284.    Although the Commission acknowledges that market shares would become more asymmetric as regards print books, e-books and physical books post-transaction, thus possibly decreasing the likelihood of coordination, it investigated whether large publishers, independently from small and medium-sized publishers would be able to reach terms of coordination.

285.  The vast majority of market players consider that coordination between large publishers in relation to the sale of English language books to dealers is unlikely and the risks for it to happen would not increase as a result of the transaction given the level of competition in the market, the differentiated nature of the products in question, the high asymmetry of market shares and volatile market positions.147

286.   The Commission furthermore investigated whether the on-going change in the the publishing supply chain, notably as regards the growth of self-publishing, the upstream integration of certain retailers and the emergence of providers of dedicated technical services such as on-demand printing has any impact on the ability or incentives of large publishers to coordinate their behaviour.

287.The majority of market participants believe that the on-going changes do not alter  the ability or incentives of large publishers to coordinate their behaviour on any segment of the market. Among the few respondents seeing an impact, some consider that the incentives for large publishers to coordinate are increasing, while their  ability to actually do so is decreasing. One respondent limits this statement to the  sale of e-books to dealers.

288.  Furthermore, as coordination could also concern aspects other than price, the Commission investigated whether publishers could tacitly agree on delaying the release of the e-book version of a given title as compared to the release of the print version ("windowing") in order to reduce the pressure of e-book sales on the sales and prices of print books. However, publishers do not consider windowing a commercially viable behavior and the overall majority of respondents does not consider coordination on windowing possible and furthermore does not consider coordination on windowing to become easier post-transaction.148

289. With a view to assessing the ability of market players to monitor deviations of a potential coordination, the Commission investigated transparency in the market, in particular as regards retail prices of English language titles.

290. The majority of competitors and customers are aware of and/or use software to monitor retail prices of e-books, online sales of print books and/or physical and/or digital audio books. Also, the majority of competitors and customers consider that the data made available on the publishing industry creates a level of transparency allowing (large) publishers to monitor each other’s behaviour as regards notably the pricing of their respective titles. Such transparency is said to exist in particular regarding retail prices. The market investigation yielded mixed results as regards pricing transparency at the wholesale level for both print and e-books, supporting the parties' argument in relation to print books, that due to the discounts negotiated individually between publishers and dealers, the actual wholesale price can hardly be deduced from the publicly available data.149

291.  As regards the sale of English language e-books to dealers the Commission takes into account the commitments that were made binding in its decision pursuant to Article 9  of Regulation 1/2003 on the implementation of the rules on competition laid down in Articles 81 and 82 of the Treaty150 of 12 December 2012 in in case COMP/AT.39847 - E-Books on Apple and four publishers – Simon & Schuster, HarperCollins, Hachette and Holtzbrinck (owner of, inter alia, Macmillan). These commitments aim at resetting the market following the alleged concerted practice between these five parties in  relation to the sale of trade e-books to consumers in the EEA. In the Commission's preliminary view, as set out in the above decision, the four publishers and  Apple engaged in a concerted practice with the object of either raising retail prices of e-books in the EEA or preventing the emergence of lower prices for e-books in the EEA by jointly converting the sale of e-books from a wholesale model to an agency model with the same key pricing terms on a global basis.

292. The case is on-going against the fifth implicated publisher, Pearson. […]. As the outcome of this procedure is yet unclear, the Commission is not in a position to take any possible future commitments into account for the purposes of assessing the likelihood of coordinated effects in the present case.

293. However, the commitments made binding on four publishers and Apple required these parties, among others, to terminate certain agency agreements for the sale of e-books to consumers in the EEA and to grant a degree of pricing discretion to retailers even in the event of the conclusion of a new agency agreement for the sale of trade e-books to consumers in the EEA. Moreover, the commitments ban for a period of 5 years the use of various types of Most Favoured Customer clauses between publishers and retailers. This removes the key coordination mechanism that was employed as part of the alleged concerted practice (that is to say the use of a retail price Most Favoured Customer clause).

294. In its Article 9 decision the Commission took the following view "[T]he […] [c]ommitments […] will create, over a sufficient period of time, conditions for a competitive reset in the EEA. This would result in sufficient uncertainty regarding the future intentions of publishers and retailers regarding the choice of business models (that is to say, wholesale, agency or a novel model) and the pricing terms used therein. The Final Commitments offered by each of the Four Publishers and Apple will also decrease incentives for each of the Four Publishers to renegotiate agreements for e-books with the same key terms."151

295.   The Commission acknowledges that in the coming years, these commitments will have an impact on the sale of e-books by publishers to dealers. However, this only relates to some or part(s) of the relevant market(s) and particular aspects of competition between publishers. To that extent, these commitments are only partially relevant for the likelihood of coordination post-merger.

296.   Finally, as described in paragraphs 245 and 269-271, the market investigation revealed that customers have, and do exercise, their buyer power, thereby rendering any possible coordination less stable and sustainable.

Conclusion

297.  In light of the foregoing, the Commission considers that the proposed transaction does not create any risk of coordinated effects on any of the possible markets for the sale of English language books to dealers.

4.     Third Party Book Distribution

4.1.      Product market definition

4.1.1.            Views of the parties

298.  In line with the distinction between book distribution services and marketing services carried out for third parties, which the Commission used in Lagardère, the parties submit that the market for third party book distribution would only comprise the provision of logistic services for the distribution of books.

299.  The parties further explain that Random House participates in the third party book distribution market in the UK and Ireland through two wholly owned subsidiaries, TBS td ("TBS") and Grantham Book Services Ltd ("GBS"). With the exception of […] cases, Pearson and Penguin do not engage in third party book distribution. As regards marketing services for other publishers, the parties' activities would be de minimis   and strictly ancillary to their distribution services. In 2011, TBS and GBS combined […] generated an income of GBP […] from marketing services carried out for third parties.

300.  For this reason the parties submit that only the market for third party book distribution constitutes a relevant product market in the present case.

4.1.2.            Commission's assessment

301.  In Lagardère, the Commission observed that "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."152At the time, the Commission differentiated between marketers (diffuseurs) and distributors (distributeurs) and considered that their respective business activities constituted separate product markets. The respective activities were described as follows: "Marketers (diffuseurs) approach dealers to promote the books (except for school textbooks) in the catalogue of the publishers for who 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 exisiting titles and engage in other promotion operations. The marketer takes orders from dealers and passes them on to distributors. 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."153

4.1.3.            Conclusion

302.   Given that the marketing services provided by the parties to other publishers are clearly de minimis in nature, for the purposes of the present decision, the Commission limits its competitive assessment to a relevant market for third party book distribution.

4.2.    Geographic market definition

4.2.1.            Views of the parties

303.   Based on the geographic scope of their standard distribution contracts, which comprise the UK and Ireland, the parties submit that the market for third party book distribution would comprise the UK and Ireland, but that the precise definition could ultimately be left open.

4.2.2.            Commission's assessment

304.   In Lagardère, the Commission considered that the market for third party book distribution was supranational and corresponded to the French-speaking area of the EU.154

305.  The results of the market investigation suggest that the geographic scope of third party book distribution services may even be larger than the UK and Ireland. All five competing publishers replied in the market investigation and who offer third party book distribution indicated that they distribute worldwide.155 The answers of customers of third party distribution services (competing publishers and book dealers) as to the question of the geographic scope of distribution services that they purchase were very mixed covering the UK only, the UK and Ireland, the EEA and worldwide.

4.2.3.            Conclusion

306. For the purposes of the present decision, the exact geographic market definition as regards third party book distribution can be left open, since the proposed transaction does not raise competition concerns under all alternative geographic market definitions.

4.3.          Competitive assessment

4.3.1.            Views of the parties

307.  The parties submit that the proposed transaction will not result in a significant impediment to effective competition on the basis of non-coordinated effects given   that (i)  Penguin Random House's market share will remain modest and there will be an increment of less than [0-5]%; (ii) there are numerous other competitors and significant excess capacity, which is growing, across the market; and (iii) barriers to entry or expansion are low. In addition, Penguin Random House would not have the ability or the incentive to foreclose either inputs (book distribution) or customers.

4.3.2.            Commission's assessment

308. For the following reasons, the proposed transaction is unlikely to raise any horizontal competition concerns.

309. First, the parties' combined market share in the UK and Ireland will amount to approximately [10-20]% by value with an increment of less than [0-5]%, which is below the threshold of 25% which is an indication for the likely absence of competition concerns.156 The parties' combined market share at the level of the EEA would be the same or lower.

310.  Second, the merged entity will continue to face several competitors including publishers  who  engage  in  third  party  book  distribution,  as  well  as      independent distributors and book producers. Indeed, four competing publishers confirmed in the market investigation that they also offer third party book distribution services157 and customers of third party book distribution services mentioned several further service providers.158

311.  Third, the market investigation confirmed that post-merger price increases by the merged entity are unlikely. Only three out of twenty-one respondents (competitors and customers) took the view that the merged entity would be in a position to profitably raise prices for third-party book distribution services in the UK and Ireland. Nine out of twenty-one respondents stated that the merged entity would not be in a position to do so, while further nine out of twenty-one did not express a firm view on this issue.159

312. The market investigation did not, however, clearly confirm that the market for third party book distribution is currently characterised by overcapacity due to the decline of sales of physical books as a result of the rise of e-books. Only half of the respondents to this question (competitors and customers) shared this view (ten out of twenty), while the other half considered that there is currently no overcapacity (six out of twenty) or did not give an assessment (four out of twenty). 160

313.  Furthermore, vertical competition concerns can be excluded on the following grounds.

314.  First, given the limited market power of the merged entity on the market for third party book distribution (combined market share of approximately [10-20]%), it will not be in a position to foreclose its competitors from the access to book distribution services. Indeed, only one out of six respondents who currently use the parties' book distribution services indicated that it would face any difficulties accessing alternative distribution services post-merger if the merged-entity was to discontinue offering third party book distribution services on the market.161

315.  Second, Random House and Penguin […]. Hence, any concerns of possible customer foreclosure can be excluded on this basis.

5.       Book production

5.1.          Product market definition

5.1.1.            Views of the parties

316. The parties submit that book manufacturing comprises both printing and binding for all types of books whatever language is used. They explain that book producers typically offer a range of printing options (print on demand, short run and long run) and formats (hardback, paperback, colour and monochrome) to suit a publisher's need over the lifecycle of a book title.

5.1.2.            Commission's assessment

317.   In the previous decision Electra/CVC/CPI,162 it was submitted that owing to the  different printing processes used and the different customers involved, the book production market was different from the printing of other products such as newspapers, magazines and commercial brochures. However, in that case, the Commission was able to leave the precise product market definition for book production open.

318.   The market investigation confirmed that the production of print books requires distinct production processes from other print products such as newspapers and magazines. As regards specifically the production of e-books, the market investigation revealed that  for new titles e-books can be created as a by-product from the print file, which needs to be converted into an e-pub file or proprietary e-book formats. In the case of older titles where a PDF print file is not/no longer available, the e-book production requires scanning of the print title with optical character recognition (OCR) software and conversion into the e-pub format. Conversion of PDF files into e-pub files is typically carried out by specialised departments of book manufacturers or specialised companies. However, conversion is also undertaken in-house by some publishers.163

5.1.3.            Conclusion

319. In any event, the Commission can limit the competitive assessment to the production of print books given that the parties do not engage in the production of e-books for third parties.

5.2.          Geographic market definition

5.2.1.            Views of the parties

320. The parties suggest leaving the geographic market definition open as the transaction will neither lead to competition concerns on the level of the EEA, nor on a regional level comprising the UK and Ireland.

5.2.2.        Commission's assessment

321. In the decision Electra/CVC/CPI, the then parties submitted that the geographic market for book production was at least EEA-wide given (i) the existence of cross- border sales, (ii) the absence of barriers to trade and the homogenous conditions for competition within the EEA, (iii) the low transportation costs, (iv) the EEA wide activities of competitors, and (v) the fact that customers do no source their procurements on a purely national basis.164 The Commission however ultimately left open the exact geographic market definition.

322. The market investigation in the present case shows that publishers of English language books source manufacturing services from manufacturers located in the  UK and Ireland, as well as in other countries of the EEA and outside the EEA.165    In  choosing the location of a book manufacturer, publishers take into account timing of delivery, quality and pricing. While for certain front list titles, in particular bestsellers, timing of delivery is crucial and therefore, a closer location of the book manufacturer to the destination of delivery is important, for heavily illustrated books it may be commercially advantageous to be produced in countries with low labour costs. As regards specifically the production of e-books, producers mentioned in the market investigation are located in the UK, Germany, India and the US.166

5.2.3.            Conclusion

323. For the purposes of the present decision, the exact geographic market definition as regards book production can be left open, since the proposed transaction does not raise any competition concerns under all alternative geographic market definitions.

5.3.          Competitive assessment

5.3.1.            Views of the parties

324.  The parties submit that the merged entity will not have the possibility or incentive to foreclose either inputs (book production) or book producers from accessing Random House or Penguin as customers. First, Bertelsmann, who is active in book production through its Arvato division and BePrinters, has only limited market shares on the upstream market for book production. The parties estimate that Bertelsmann's share on the market for book production will be around [5-10]% at the level of the EEA and around [0-5]% at the level of the UK and Ireland. Second, the merged entity will only hold a modest market share of around [20-30]% of the overall market for the sale of English language books to dealers in the UK and Ireland (2011), which would leave book producers with sufficient alternative customers if the merged entity was  to discontinue sourcing book production from third parties.

5.3.2.            Commission's assessment

325.  According to the Non-Horizontal Merger Guidelines, a merger is likely to result in foreclosure where actual or potential rivals’ access to supplies or markets is hampered or eliminated as a result of the merger, thereby reducing these companies' ability and incentive to compete.167 Such foreclosure is regarded as anticompetitive where, as a result of the merger, the merged entity, and possibly also some of its competitors, are able to profitably increase the price charged to consumers.

326. When assessing the likelihood of such an anticompetitive foreclosure scenario, it needs to be examined "first, whether the merged entity would have, post-merger, the ability to substantially foreclose access to inputs, second, whether it would have the incentive to do so, and third, whether a foreclosure strategy would have a significant detrimental effect in the competition downstream".168

327.  The market investigation confirmed that Bertelsmann does not enjoy any market power on the market for book production and that switching is possible given the presence of credible  alternative  book  manufacturers.  Only  a  limited  number  of  publishers who participated in the market investigation have their books produced by Bertelsmann and none of those relies exclusively on Bertelsmann as supplier.169 Furthermore, all of the market participants answered that switching from Bertelsmann to another book manufacturer would be possible and none pointed out that they would face significant challenges in terms of associated timing and cost.170 For this reason, the merged entity will not have the ability to foreclose competitors from accessing book production services post-merger.

328.  In addition, the parties' combined market share on the overall market for the sale of English language books to dealers in the UK and Ireland of [20-30]% in 2011 and around [20-30]% in 2012 is not of such magnitude that it would enable the merged entity to successfully foreclose competing book producers from accessing downstream customers.

V.        CONCLUSION

329.   For the above reasons, the Commission has decided not to oppose the notified operation and to declare it compatible with the internal market and with the EEA Agreement. This decision is adopted in application of Article 6(1)(b) of the Merger Regulation.

1          OJ L 24, 29.1.2004, p. 1 ("the Merger Regulation"). With effect from 1 December 2009, the Treaty on the Functioning of the European Union ("TFEU") has introduced certain changes, such as the replacement of "Community" by "Union" and "common market" by "internal market". The terminology of the TFEU will be used throughout this decision.

2          Publication in the Official Journal of the European Union No C 064, 5.3.2013. p. 14.

3        Commission Consolidated Jurisdictional Notice under Council Regulation (EC) No 139/2004 on the control of concentrations between undertakings (OJ C 95, 16.4.2008, p. 1).

4        Consolidated Jurisdictional Notice, paragraph 67.

5       Consolidated Jurisdictional Notice, paragraph 66.

6        Commission decision of 8 September 2009 in Case COMP/M.5533 Bertelsmann/KKR/JV, paragraphs 22 and 23.

7        Section 1.01 of the Contribution Agreement.

8        Articles 4 and 6 of the Framework Agreement.

9        Section 1 03. and Article 5 of the Contribution Agreement.

10     Section 103. of the Contribution Agreement.

11     Section 103. of the Contribution Agreement..

12     Article 7 of the Contribution Agreement and Article 3 of the Framework Agreement.

13     Turnover calculated in accordance with Article 5(1) of the Merger Regulation and the Commission Consolidated Jurisdictional Notice.

14     Commission decision of 22 May 2007 in Case COMP/M.4404 – Universal/BMG Music Publishing; Commission decision of 19 April 2012 in Case COMP/M. 6459 Sony/Mubadala Development/EMI Music Publishing.

15     See Commission decision of 7 January 2004 in Case COMP/M.2978 Lagardère/Natexis/VUP, paragraph 69 to 114.

16     See reply to questionnaire Q1 – Publishers / competitors of 26 February 2013, question 3; and replies to questionnaire Q2 – Authors/literary agents of 26 February 2013, question 2.

17     See reply to questionnaire Q 1 – Publishers / competitors of 26 February 2013, question 7; and replies to questionnaire Q2 – Authors/literary agents of 26 February 2013, question 6.

18     See reply to questionnaire Q1 – Publishers / competitors of 26 February 2013, question 6.

19     See reply to questionnaire Q2 – Authors/literary agents of 26 February 2013, question 5.

20     See reply to questionnaire Q1 – Publishers / competitors of 26 February 2013, question 8.

21     See reply to questionnaire Q1- Publishers / competitors of 26 February 2013, question 9, and reply to questionnaire Q2 – Authors/literary agents of 26 February 2013, question 8.

22     See reply to the questionnaire Q1 – Publishers / competitors of 26 February 2013, question 10; and reply  to questionnaire Q2 – Authors/literary agents of 26 February 2013, question 9.

23     See reply to questionnaire Q1 – Publishers / competitors of 26 February 2013, question 5; and reply to questionnaire Q 3 – Authors/literary agents of 26 February 2013, question 4.1.

24     See reply to questionnaire Q1 – Publishers / competitors of 26 February 2013, question 5; and reply to questionnaire Q2 – Authors/literary agents of 26 February 2013, question 4.2.

25     See Form CO, paragraph 6.334.

26     See the parties' response of 27 February 2013 to the second request for information of 11 February 2013, paragraph 32.4.

27     In their replies to questions 3 to 31 of Questionnaire Q1 – Publishers / competitors of 26 February 2013, none of the competing publishers made any comment relating to the possible existence of a separate product market for the acquisition of publishing rights by well-reputed authors.

28     The parties submit that Commonwealth rights would typically include at least Australia, New Zealand, South Africa and India.

29     According to the exhaustion principle, which is an expression of the free movement of goods, once goods that are subject to intellectual property rights ("IPR") have been released for sale in the territory of the EU with the consent of the holder of the IPR, this consent has "exhausted" any remedies for later breach of the IPR by onward sale within the territory of the EU. This means that a title legitimately sold into, for example, Germany could lawfully be parallel-imported into the UK and Ireland.

30     Commission decision of 7 January 2004 in Case COMP/M.2978 Lagardère/Natexis/VUP, paragraph 305 and 309.

31     See reply to questionnaire Q1- Publishers / competitors of 26 January 2013, question 11.1.

32     See reply to questionnaire Q1 – Publishers / competitors of 26 January 2013, question 11.

33     See reply to questionnaire Q 1 – Publishers / competitors of 26 January 2013, questions 12 to 14.

34     See reply to questionnaire Q 1 – Publishers /competitors of 26 January 2013, question 15.

35     Commission decision of 7 January 2004 in Case COMP/M.2978 Lagardère/Natexis/VUP, paragraph 338.

36      See reply to questionnaire Q2 - authors / literary agents of 26 February 2013, question 23 and reply to questionnaire Q1 - Publishers / competitors of 26 February 2013, question 27.

37     See reply to questionnaire Q2 - authors / literary agents of 26 February 2013, question 22 and reply to questionnaire Q1 - Publishers / competitors of 26 February 2013, question 26.

38      See reply to questionnaire Q2 - Authors/literary agents of 26 February 2013, question 22 and reply to questionnaire Q1 - Publishers / competitors of 26 February 2013, question 26.

39      See replies to questionnaire Q2 – Authors/literary agents of 26 February 2013, question 28.

40      See reply to questionnaire Q1 - Publishers / competitors of 26 February 2013, question 27.

41      See reply to questionnaire Q2 - Authors / literary agents of 26 February 2013, question 13.

42      See reply to questionnaire Q2 - Authors / literary agents of 26 February 2013, question 23.

43    See reply to questionnaire Q1 - Publishers / competitors of 26 February 2013, question 22.

44    See reply to questionnaire Q1 - Publishers / competitors of 26 February 2013, question 22.

45      See reply to questionnaire Q2 - Authors / literary agents of 26 February 2013, question 18.

46    See reply to questionnaire Q2 - Authors / literary agents of 26 February 2013, question 30.

47    See reply to questionnaire Q2 - Authors / literary agents of 26 February 2013, question 27.

48    See reply to questionnaire Q1 - Publishers / competitors of 26 February 2013, question 25.

49    Case T-342/99, Airtours plc v Commission, Case T-464/04, Impala v Commission.

50    Guidelines on the assessment of horizontal mergers under the Council Regulation on the control of concentrations between undertakings, OJ C 31, 5.2.2004, p.5, paragraphs 39-57.

51    See reply to questionnaire Q1 - Publishers / competitors of 26 February 2013, question 28, and to questionnaire Q2 - Authors / literary agents of 26 February 2013, question 31.

52    See reply to questionnaire Q1 - Publishers / competitors of 26 February 2013, question 29, and to questionnaire Q2 - Authors / literary agents of 26 February 2013, question 32.

53    See reply to questionnaire Q1 - Publishers / competitors of 26 February 2013, question 30, and questionnaire Q2 - Authors / literary agents of 26 February 2013, question 33.

54    For example, cookery books are not appropriate for audio books and generally not transformed into audio book format.

55    Audiobooks are not published at the same time as an e-book or physical book if (i) the audiobook rights were not acquired with the physical and e-book rights; for example, if the literary agent sold these separately to another publisher; or (ii) they were acquired but it is not possible (because, for example, the narrator that the author has requested is not available until a later date) or commercially viable (the author is a debut author and there is a risk that the title will not sell enough copies to make an audiobook recording worth the cost) to publish the audiobook at the same time as the other formats.

56    For instance, Random House's sales team is, […]. Penguin's sales teams […].

57    Commission decision of 7 January 2004 in Case COMP/M.2978 Lagardère/Natexis/VUP, paragraphs 186, 187.

58    Commission decision of 7 January 2004 in Case COMP/M.2978 Lagardère/Natexis/VUP, paragraph 194.

59    Commission decision of 7 January 2004 in Case COMP/M.2978 Lagardère/Natexis/VUP. For the general literature category see paragraphs 198, 199. Further subcategories were found in relation to educational books (school textbooks and educational supporting materials were found to be separate product markets), see paragraph 251 and reference works (smaller reference works with two further subcategories – dictionaries and encyclopaedias, larger reference works and books published in instalments), see paragraphs 257 – 277.

60    Commission decision of 7 January 2004 in Case COMP/M.2978 Lagardère/Natexis/VUP, paragraph 216.

61    As e-books were hardly developed at the time of the decision, the Commission did not consider any distinction between e-books on the one hand and print and/or audio books on the other hand in Lagardère. However, as regards the various markets for the sale of larger reference works to the different types of dealers, the Commission found separate markets for paper and so-called multimedia versions, comprising pre-installed software, CD-ROM and online versions (paragraph 275), while leaving open the decision whether within the category of encyclopedias, a sub-category of smaller reference works, computerised and print versions constitute separate product markets (paragraphs 270, 271). Furthermore, in its decision in Candover/Cinven/Bertelsmann-Springer (Commission decision of 29 July 2003 in Case COMP/M.3197 Candover/Cinven/Bertelsmann-Springer) the Commission considered whether print and online versions of academic journals on the one hand and print and CD-ROM versions of academic books on the other hand form part of, respectively, one relevant product market. In the absence of competition concerns the decision on this point was left open (paragraph 21).

62      See reply to questionnaire Q1 – Publishers / competitors of 26 February 2013, question 33; reply to questionnaire Q3 – Wholesalers, retailers / customers of 26 February 2013, question 9.

63    See reply to questionnaire Q1 – Publishers / competitors of 26 February 2013, questions 34, 35; reply to questionnaire Q3 – Wholesalers, retailers / customers of 26 February 2013, questions 11, 13.

64    See reply to questionnaire Q3 – Wholesalers, retailers / customers of 26 February 2013, questions 10, 12.

65    The Commission notes that only a small minority of the respondents considers physical audio books and digital audio books to constitute separate markets.

66    See reply to questionnaire Q1 – Publishers / competitors of 26 February 2013, questions 37, 38; see  replies to questionnaire Q3 – Wholesalers, retailers / customers of 26 February 2013, question 16.

67    See reply to questionnaire Q3 – Wholesalers, retailers / customers of 26 February 2013, question 15.

68      See reply to questionnaire Q1 – Publishers / competitors of 26 February 2013, question 39; reply to questionnaire Q3 – Wholesalers, retailers / customers of 26 February 2013, question 17.

69    See reply to questionnaire Q1 – Publishers / competitors of 26 February 2013, questions 40, 41; see reply to questionnaire Q3 – Wholesalers, retailers / customers of 26 February 2013, question 19.

70    See reply to questionnaire Q3 – Wholesalers, retailers / customers of 26 February 2013, question 18.

71    Commission decision of 7 January 2004 in Case COMP/M.2978 Lagardère/Natexis/VUP, paragraphs 179, 186, 187.

72      See reply to questionnaire Q1 – Publishers / competitors of 26 February 2013, question 53.

73    See reply to questionnaire Q3 – Wholesalers, retailers / customers of 26 February 2013, question 28.

74    See reply to questionnaire Q3 – Wholesalers, retailers / customers of 26 February 2013, question 29.

75    See reply to questionnaire Q1 – Publishers / competitors of 26 February 2013, question 51.

76    See reply to questionnaire Q1 – Publishers / competitors of 26 February 2013, question 52.

77    See reply to questionnaire Q1 – Publishers / competitors of 26 February 2013, question 51.

78    See reply to questionnaire Q1 – Publishers / competitors of 26 February 2013, question 51.

79    See reply to questionnaire Q3 – Wholesalers, retailers / customers of 26 February 2013, question 30.

80    See reply to questionnaire Q1 – Publishers / competitors of 26 February 2013, question 42 and questionnaire Q3 – Wholesalers, retailers / customers of 26 February 2013, question 20.

81    See reply to questionnaire Q1 – Publishers / competitors of 26 February 2013, question 42 and questionnaire Q3 – Wholesalers, retailers / customers of 26 February 2013, question 20.

82    See reply to questionnaire Q1 – Publishers / competitors of 26 February 2013, question 44; reply to questionnaire Q3 – Wholesalers, retailers / customers of 26 February 2013, question 23.

83    See reply to questionnaire Q1 – Publishers / competitors of 26 February 2013, question 43 and questionnaire Q3 – Wholesalers, retailers / customers of 26 February 2013, question 21.

84    See reply to questionnaire Q3 – Wholesalers, retailers / customers of 26 February 2013, question 22.

85    See reply to questionnaire Q1 – Publishers / competitors of 26 February 2013, question 45 and questionnaire Q3 – Wholesalers, retailers / customers of 26 February 2013, question 24.

86    See reply to questionnaire Q3 – Wholesalers, retailers / customers of 26 February 2013, question 27.

87    See reply to questionnaire Q3 – Wholesalers, retailers / customers of 26 February 2013, question 22.

88    See reply to questionnaire Q1 – Publishers / competitors of 26 February 2013, question 47.

89    See reply to questionnaire Q1 – Publishers / competitors of 26 February 2013, question 49.

90    See reply to questionnaire Q3 – Wholesalers, retailers / customers of 26 February 2013, question 27. See also the similar reply by a multimedia and stationary retailer.

91    Commission decision of 7 January 2004 in Case COMP/M.2978 Lagardère/Natexis/VUP, paragraph 296.

92    See reply to questionnaire Q1 – Publishers / competitors of 26 January 2013, question 54.

93    One publisher clarifies that a limitation in geographic scope due to copyright would only affect 5% of the titles in his catalogue.

94    See reply to questionnaire Q1 – Publishers / competitors of 26 January 2013, question 54.

95    See reply to questionnaire Q1 – Publishers / competitors of 26 January 2013, questions 54-56; reply to questionnaire Q3 – Wholesalers, retailers / customers of 26 January 2013, questions 31, 32.

96    The Commission notes that eight out of sixteen dealers did not respond to this question.

97    See reply to questionnaire Q1 – Publishers / competitors of 26 January 2013, question 62; reply to questionnaire Q3 – Wholesalers, retailers / customers of 26 January 2013, question 34.

98    While the parties describe US-based publisher Simon & Schuster as a sixth major publisher on the English language book markets, the Commission considers that, given the latter's limited market share, it is best described as a mid-sized publisher in the UK/Ireland. That said, Simon & Schuster is an internationally established publisher and active across all major content categories.

99    See Annex 12 to the Form CO for a full explanation of the methodology. The parties note that audiobook sales have not been included due to the paucity of available data on these sales. However, sales of physical audiobooks relative to the total market would suggest that the inclusion/exclusion of audiobook sales is unlikely to affect the market share calculations. To illustrate, in the UK and Ireland, physical audiobook sales would comprise only [0-5]% of total physical book and audiobook sales in 2011. Consequently, assuming that audiobook sales will comprise a similarly small proportion of total sales in the EEA excluding the UK and Ireland, the inclusion/exclusion of audiobook sales would be unlikely to affect the estimation of shares of English language sales in this region.

100 Pearson’s share of sales of English language books in the EEA (excluding UK and Ireland) would be [0- 5]% by volume and [0-5]% by value in 2011.

101See reply to questionnaire Q1 – Publishers / competitors of 26 February 2013, question 73 and reply to questionnaire Q3 – Wholesalers, retailers / customers of 26 February 2013, question 39.

102See reply to questionnaire Q1 to Publishers / competitors of 26 February 2013, question 73.

103See reply to questionnaire Q1 – Publishers / competitors of 26 February 2013, question 72 and reply to questionnaire Q3 – Wholesalers, retailers / customers of 26 February 2013, question 38. However, it was noted that the Nielsen bookscan data does not cover print book sales through book clubs.

104Ibid.

105The share of the business remaining with Pearson is minimal, with shares of [0-5]% of sales to chain booksellers, [0-5]% to mass market dealers, [0-5]% to independent bookstores, [0-5]% to Internet retailers and [0-5]% to other dealers.

106The parties submit that, first, the Bowker market share are based on a small sample size of which an even smaller proportion purchase e-books. Therefore, the results are likely to be highly sensitive to the preferences of the particular consumers selected to complete the survey. Second, the Bowker data does not include free books which may result in a distortion of the volume share.

107The equivalent share for Pearson Education in the Bowker data is [0-5]% by volume and less than [0-5]% by value.

108  The parties submit that these publishers are Atlantic Books, Canongate, Icon Books, Portobello Books, Profile Books (including Serpent’s Tail), Short Books, Quercus and Granta Books.

109Horizontal Guidelines, paragraph 18.

110See reply to questionnaire Q1 – Publishers / competitors of 26 February 2013, question 85 and reply to questionnaire Q3 – Wholesalers, retailers / customers of 26 February 2013, question 49.

111See reply to questionnaire Q1 – Publishers / competitors of 26 February 2013, question 85 and reply to questionnaire Q3 – Wholesalers, retailers / customers of 26 February 2013, question 49. See the sub- questions to these questions for the replies concerning the various Lagardère content categories.

112112  See reply to questionnaire Q1 – Publishers / competitors of 26 February 2013, question 89.

113113  See reply to questionnaire Q3 – Wholesalers, retailers / customers of 26 February 2013, questions 66 and 67. 

114114  See reply by Faber and Faber to questionnaire Q1 – Publishers / competitors of 26 February 2013, question 71.

115115 Ibid.

116116  Commission decision of 7 January 2004 in Case COMP/M.2978 Lagardère/Natexis/VUP, paragraphs 593-594.

117See reply to questionnaire Q1 – Publishers / competitors of 26 February 2013, question 88 and reply to questionnaire Q3 – Wholesalers, retailers / customers of 26 February 2013, question 52.

118See reply to questionnaire Q3 – Wholesalers, retailers / customers of 26 February 2013, question 22.

119See reply to questionnaire Q1 – Publishers / competitors of 26 February 2013, questions 75 and 76, and reply to questionnaire Q3 – Wholesalers, retailers / customers of 26 February 2013, questions 40 and 41.

120See reply to questionnaire Q1 – Publishers / competitors of 26 February 2013, question 76.1.1 and 76.1.2, and reply to questionnaire Q3 – Wholesalers, retailers / customers of 26 February 2013, question 42.

121For the purpose of the questionnaires, the Commission defined bestsellers as "any title which achieved a ranking in the top 100 titles in TCM 5000 Nielsen bookscan database in any one week in the course of a year" (weekly top 100).

122See reply to questionnaire Q3 – Wholesalers, retailers / customers of 26 February 2013, question 42.

123See reply to questionnaire Q1 – Publishers / competitors of 26 February 2013, question 76.

124See reply to questionnaire Q3 – Wholesalers, retailers / customers of 26 February 2013, question 22.

125See reply to questionnaire Q3 – Wholesalers, retailers / customers of 26 February 2013, question 27.

126See reply to questionnaire Q1 – Publishers / competitors of 26 February 2013, question 90.

127See reply to questionnaire Q1 – Publishers / competitors of 26 February 2013, question 91.

128See reply to questionnaire Q1 – Publishers / competitors of 26 February 2013, question 92.

129See reply to questionnaire Q1 – Publishers / competitors of 26 February 2013, question 96.

130See reply to questionnaire Q1 – Publishers / competitors of 26 February 2013, question 97.

131See reply to questionnaire Q1 – Publishers / competitors of 26 February 2013, question 100.

132The parties used data of the Bowker Books and Consumers Survey, which is based on surveys composed of individual online interviews with 7 800 consumers per month, representative of the UK population aged 13 to 84 by sex, age, social grade and region.

133  The parties submit, for instance, that consumer options for alternative dedicated e-book readers  (as opposed to smartphones and tablets) are limited (for example Kobo, Nook) and that Amazon’s distribution agreement with Waterstones has had the effect of […] reducing the potential for new entry into e-book retailing in the UK and Ireland.

134See reply to questionnaire Q 1 – Publishers / competitors of 26 February 2013, question 86.

135See reply to questionnaire Q 3 – Wholesalers, retailers / customers of 26 February 2012, question 50.

136For the purposes of these calculations, the Commission defined the annual top 100 as a title which was in the top 100 by year, that is titles are ranked by annual sales and subsequently the top 100 list is extracted for that year. The weekly top 100 are defined as a title that was in the top 100 in any week, and this title has also been treated as a bestseller in all other weeks of that year.

137See reply to questionnaire Q3 – Wholesalers, retailers / customers of 26 February 2013, question 22.

138See reply to questionnaire Q 3 – Wholesalers, retailers / customers of 26 February 2012, questions 43 and 44.

139See reply to questionnaire Q1 – Publishers / competitors of 26 February 2013, questions 78 and 79.

140See reply to questionnaire Q1 – Publishers / competitors of 26 February 2013, question 101.

141See reply to questionnaire Q1 – Publishers / competitors of 26 February 2013, question 105.

142See reply to questionnaire Q1 – Publishers / competitors of 26 February 2013, question 93.

143See reply to questionnaire Q1 – Publishers / competitors of 26 February 2013, questions 94 and 95.

144  See reply to questionnaire Q 3 – Wholesalers, retailers / customers of 26 February 2012, question 3.

145Case T-342/99, Airtours plc v Commission, Case T-464/04, Impala v Commission.

146Guidelines on the assessment of horizontal mergers under the Council Regulation on the control of concentrations between undertakings, OJ C 31, 5.2.2004, p.5, paragraphs 39-57.

147See reply to questionnaire Q1 – Publishers / competitors of 26 February 2013, question 108; reply to questionnaire Q3 – Wholesalers, retailers / customers of 26 February 2013, question 53.

148See reply to questionnaire Q1 – Publishers / competitors of 26 February 2013, question 109; reply to questionnaire Q3 – Wholesalers, retailers / customers of 26 February 2013, question 54.

149See reply to questionnaire Q1 – Publishers / competitors of 26 February 2013, question 111; reply to questionnaire Q3 – Wholesalers, retailers / customers of 26 February 2013, question 53.

150 OJ L 1, 4.1.2003, p. 1.

151 Commission decision of 12 December 2012 in Case COMP/AT.39847 – E-Books, paragraph 134.

152 Commission decision of 7 January 2004 in Case COMP/M.2978 Lagardère/Natexis/VUP, paragraph 115.

153 Commission decision of 7 January 2004 in Case COMP/M.2978 Lagardère/Natexis/VUP, paragraphs 24 and 25.

154 Commission decision of 7 January 2004 in Case COMP/M.2978 Lagardère/Natexis/VUP, paragraph 338.

155See reply to questionnaire Q1 – Publishers / competitors of 26 February 2013, question 125.

156Guidelines on the assessment of horizontal mergers under the Council Regulation on the control of concentrations between undertakings, paragraph 18.

157See reply to questionnaire Q1 – Publishers / competitors of 26 February 2013, question 122.

158  See reply to questionnaire Q3 – Wholesalers, retailers / customers of 26 February 2013, question 58.

159  See reply to questionnaire Q1 – Publishers / competitors of 26 February 2013, questions 132; reply to questionnaire Q3 – Wholesalers, retailers / customers of 26 February 2013, question 65.

160  See reply to questionnaire Q1 – Publishers / competitors of 26 February 2013, question 128; reply to questionnaire Q3 – Wholesalers, retailers / customers of 26 February 2013, question 63.

161  See reply to questionnaire Q1 – Publishers / competitors of 26 February 2013, question 131.

162Commission decision of 27 September 2005 in Case COMP/M.3939 Eectra/CVC/CPI.

163See reply to questionnaire Q1 – Publishers / competitors of 26 February 2013, question 116.

164Commission decision of 27 September 2005 in Case COMP/M.3939 Eectra/CVC/CPI, paragraph 14.

165  See reply to questionnaire Q1 – Publishers / competitors of 26 February 2013, question 117.

166   See reply to questionnaire Q1 – Publishers / competitors of 26 February 2013, question 118.

167   OJ C265, 5.2.2008, p. 6, paragraph 31.

168   OJ C265, 5.2.2008, p. 6, paragraph 32.

169See reply to questionnaire Q1 – Publishers / competitors of 26 February 2013, question 119.

170See reply to questionnaire Q1 – Publishers / competitors of 26 February 2013, question 121.