EC, June 21, 1994, No M.430

THE COMMISSION OF THE EUROPEAN COMMUNITIES,

Having regard to the Treaty establishing the European Community, Having regard to Council Regulation (EEC) No 4064-89 of 21 December 1989 on the control of concentrations between undertakings (1), and in particular Article 8 (2) thereof, Having regard to the EEA Agreement and in particular Article 57 (1) thereof, Having regard to the Commission Decision of 17 February 1994 to initiate proceedings in this case, Having given the undertakings concerned the opportunity to make known their views on the objections raised by the Commission, Having regard to the opinion of the Advisory Committee on Concentrations, Whereas:

(1) On 17 January 1994 Procter & Gamble GmbH notified the Commission of a proposed concentration by which it intends to acquire the whole of Vereinigte Papierwerke Schickedanz AG. This notification follows the withdrawal on 17 January 1994 of its initial notification of 9 December 1993.

The second notification relates to a modified transaction in so far as it includes an offered divestment of a part of VPS's catamenials business (see recital 8 below).

(2) After examination of the notification the Commission has concluded that the notified concentration falls within the scope of Regulation (EEC) No 4064-89.

On 17 February 1994, the Commission decided to initiate proceedings in application of Article 6 (1) c of the Merger Regulation.

I. THE PARTIES

(3) Procter & Gamble GmbH (P& G) is a wholly-owned subsidiary of the Procter & Gamble Company which is active in consumer laundry, cleaning, health and beauty products, paper products including sanitary protection products (except tampons), and food and beverages.

(4) Vereinigte Papierwerke Schickedanz AG (VPS) is a wholly-owned subsidiary of Gustav und Grete Schickedanz Holding KG (GGS), a partnership organized under the laws of Germany. VPS produces household paper and sanitary protection products.

II. THE OPERATION

(5) Procter & Gamble GmbH will acquire the entire share capital of VPS, together with that of other GGS subsidiaries active in related businesses. Under the terms of the share transfer agreement between P& G and GGS and a side agreement between GGS, VPS and P& G, VPS will separate its baby nappy business from its other activities and hold it in a separate company prior to completion of the transaction. The agreements provide for P& G to acquire the VPS shares and then, immediately at closing, to transfer the shares of the separated company holding VPS's baby hygiene business to a third party, i.e., a trustee designated by Procter & Gamble on 22 December 1993.

(6) This third party shall be assigned the mandate of finding final buyers within an initial period of [...] (2) months following closing, to be extended if necessary by a further [...] months under certain conditions.

(7) This commitment not to acquire control of VPS's baby diaper business forms an integral part of the notification and for this reason, despite the objections the Commission would have to any such acquisition, this Decision does not address this market.

(8) Futhermore, with regard to concerns as to possible effects of the notified concentration as to feminine hygiene products, P& G included in its notification a commitment, offered unilaterally to the Commission and supplemented by an additional agreement between the parties, not to acquire control of the 'non-Camelia business' (i.e. the tangible and intangible assets related to three VPS brands - Bluemia, Femina and Tampona - and VPS's private label business) of VPS's catamenial activities. Subsequently, and in the light of the objections raised by the Commission, P& G substantially altered both the brands to be divested and the terms of such a divestment (see Part VI of this Decision). It substituted the Camelia-branded feminine hygiene products for the non-Camelia products of VPS.

III. THE CONCENTRATION

(9) P& G will gain sole control over VPS (subject to the exceptions of the baby diaper business). The operation is thus a concentration within the meaning of Article 3.

IV. COMMUNITY DIMENSION

(10) The notified operation has a Community dimension since total worldwide turnover of the undertaking concerned (P& G ECU 23,626 billion and VPS ECU 681 million in 1992/93 respectively) exceeds the ECU 5 billion threshold; the aggregate Community-wide turnover of both undertakings concerned (P& G ECU [...] million and VPS ECU [...] million) is more than ECU 250 million; and the parties do not achieve more than two thirds of their EC turnover in one and the same Member State.

V. COMPATIBILITY WITH THE COMMON MARKET

(11) The concentration concerns the following products manufactured by VPS: household hygiene paper products, feminine products (catamenials), adult incontinence products, cotton products and certain personal body care products (cosmetics).

(12) Proceedings were initiated with respect to sanitary towels.

Household hygiene paper products

(13) P& G is not active in Europe in these products which comprise, inter alia, toilet paper, paper handkerchiefs, facial tissues and kitchen towels. Sales of these products account for [...] (3) of the total sales of VPS (about [...] (4)). P& G stated that the strategic aim of the proposed merger is to enter the European 'tissue market'. This is thought to be as large as the nappy and catamenial markets together (about ECU 5 billion). P& G is already a market leader in this sector in the US and Canada and it hopes to bring to bear its superior technological and product expertise in Europe through this transaction. The overall market shares of VPS are modest ([...] (5) in the EU and [...] (6) in Germany by volume; value figures not available). As for the individual products EU-wide market shares lie between [...] (7) (kitchen towels) and [...] (8) (handkerchiefs) and in Germany VPS will have [...] (9) in handkerchiefs and a [...] (10) share in toilet paper. Scott and JA/Mont are the leading suppliers and private labels play an important role on a market which is expected to grow in the coming years.

In the absence of any overlap between P& G and VPS in this sector and in the light of VPS's limited market shares, the operation does not give rise to any competition concerns for these products.

Adult incontinence products Product market

(14) Adult incontinence protection products include fitted briefs, shields and pads, insert pads and undergarments. The different products are used for different degrees of loss of bladder and bowel control. Although adult incontinence products are similar to baby diaper and external sanitary napkins, suppliers and customers do not regard them as substitutable because they perform completely separate functions. The absorbency of incontinence products ranges from 150 ml of urine for light incontinence to 800-1 000 ml for heavy cases whereas feminine protection products are designed to absorb between 5 and 20 ml of fluid and the most absorbent baby diaper is designed to absorb about 160-190 ml of urine.

(15) Adult incontinence products differ from diaper and fempro products not only in size and product design but also in raw material and to some extent in technology. The technology relating to super-absorbency is different in respect of incontinence products, baby diapers and sanitary napkins in order to meet the different absorbency requirements in each case. It is true that machines for napkins or baby diapers could theoretically be adapted in order to produce incontinence pads or briefs, but this, as other suppliers confirmed, does not seem to be commercially feasible. P& G agrees that notwithstanding a certain level of supply-side substitutability, adult incontinence products are a separate market.

(16) As to the relevant geographic market, the investigation of the Commission produced contradictory results. VPS manufactures incontinence products in Germany and sells them in the Benelux, France, Italy, Spain, Germany, Austria and Switzerland (mostly under the Certina brand). P& G produces and sells the products under Attends in Western Europe and Lindor or Linidor in Italy, Spain and Portugal.

(17) Similarly, the other big suppliers in Europe such as Moelnlycke ('Tena') and Hartmann ('Molicare') market their products in most of the EEA countries. They compete apparently with the same range of products under European-wide brands, a trend that could be further strengthened by proposed EU legislation harmonizing standards for medical devices, including adult incontinence products. This could suggest a Western European market.

(18) Indication to the contrary, suggesting national markets, are, however, to be found in the form of specific supply and distribution channels which clearly distinguish these products from all other hygiene products. Adult diapers are basically considered a health care item. Products are distributed through hospitals, pharmacies and institutional suppliers even if some suppliers (e.g., Moelnlycke) currently try to distribute through retail and drug stores. In the United Kingdom, for example, approximately 80 % of the purchase are made by the National Health Service through six regional buying groups. In Germany pharmaceutical distributors tend to purchase from German suppliers.

(19) However, the definition of the relevant geographic market for adult incontinence products can be left open, since even in the case of a national market, e.g. in Germany, the conclusion of the competitive analysis would not lead to serious doubts as to the compatibility of the concentration with the common market.

(20) According to the notification, the new entity will achieve a market share in Western Europe of between [...] and [...] (11) in the Community (throughout this paragraph volume figures have been taken when value figures were unavailable). Moelnlycke is the largest supplier with more than [...] (12) while Hartmanns has [...] (13) of Western European sales. In Germany, P& G would account for [...] (14) of the volume of the market and would have a [...] (15) value share. VPS's share in both volume and value would be [...] (16). This would amount to a combined market share of [...] (17) in value terms in Germany. Moelnlycke has [...] (18) of the German market (by volume) and Hartmann [...] (19) (also by volume).

(21) Given this market structure and the fact that Moelnlycke and not P& G is the market leader on a European level, the Commission considers that the concentration will not give rise to the creation of a dominant position in Germany in respect of adult incontinence.

Cotton products and cosmetics

(22) P& G does not sell cotton products (such as cotton pads, cotton balls and so-called basic cotton) in Europe and there is thus no increase in VPS's market shares. VPS has a leading cotton pad brand, Demak'up, which has a [...] (20) market share in the seven Member States where it is sold. VPS estimates its EC market share for all cotton products at [...] (21). The Commission considers that, in the absence of any indications to the contrary, the mere fact of these market shares being acquired by P& G would not be sufficient to give rise to competition concerns.

(23) Personal body care products include a large number of product lines (body/skin/hair care products, oral hygiene, scents, etc.), of which VPS only produces a limited range and sells most of it in Germany with an estimated market share of [...] (22). P& G's market shares also fall below 10 % in the EU and below [...] (23) in single Member States. For both, cosmetics represent a business unit of minor importance. Furthermore, numerous new products are launched in Western Europe each year following volatile consumer habits. The market seems to be intensely competitive. The concentration thus does not raise serious doubts as to its compatibility with the common market in this sector.

Baby nappies

(24) As described above (paragraph 5), P& G would carry out a divestment of VPS's nappy business simultaneously with the closing of the acquisition.

(25) P& G has, according to the parties' calculations, an EU volume-based market share of [...] (24), (value-based shares are not available), excluding the nappy business of Finaf, which is to be divested in the P& G/Finaf case while VPS has a share of [...] (25). While the increment is perhaps not large, the Commission has already made clear in the P& G/Finaf case that P& G should not increase its market share in this sector.

(26) Given the market shares, the financial resources, the advanced technologies and the strong position of P& G in relation to retailers, the Commission considers that, in the absence of the commitment contained in P& G's notification not to acquire control of the nappy business, the operation would create a dominant position for P& G even on the assumption of a Western European market for baby nappies. It accepts, however, that the proposed divestment will meet its concerns in this area.

Feminine hygiene products

A. The relevant product market

(27) Feminine protection ('fempro') products comprise pant-liners, tampons and sanitary towels (pads). VPS manufactures all three while P& G makes only towels and pantliners, not tampons. Pantliners and towels are both external forms for protection relying on an adhesive pad being placed in underwear, whilst a tampon is worn internally. Collectively these products are also known as catamenials.

(28) P& G claimed in its notification that all three feminine hygiene products constitute a single market. It argued, in respect of tampons and towels, that the two products share the same use; that women switch readily between them, especially in response to new products; and that manufacturers of towels compete vigorously with tampon producers. The Commission, however, considers that pantliners, tampons and sanitary towels each constitute separate product markets.

(29) Pantliners are mostly worn as protection for inter-menstrual vaginal secretions, i.e. outside the menstrual period. They are also used menstrually, whether on their own as an insurance immediately before and after the actual period, during the so-called 'lighter days' (normally the last two days of the period when the menstrual flow is much lighter), or in conjunction with a tampon at any time during menstruation. They are not suitable for use alone as a primary method of menstrual protection. This usage pattern is confirmed by the absorbent capacity of a pantliner which is around one quarter that of a towel. Furthermore, the marketing of pantliners is geared to addressing a need for 'feminine freshness' rather than protection.

(30) In terms of the supply side, it should be noted that while there may be some limited supply-side substitutability between the production of pantliners and towels because of similarities in the technology and materials used, in practice that substitutability is extremely limited. All suppliers other than Procter & Gamble stated that the substantial retooling required to alter the product produced on the dedicated machines used by the major producers makes such retooling uneconomic. Even if there were some limited supply side substitutability, it would be difficult for any increased capacity to be converted into market share by a company not already producing towels, given the non-technological barriers to entry to the towel market which are described later in the Decision.

(31) For the above reasons the Commission does not consider pantliners to be substitutable for tampons or towels as a primary method of menstrual protection.

(32) Towels are designed for menstruation (although there may be some marginal use for light incontinence). They consist of an absorbent pad in an envelope consisting of an impermeable back sheet and a permeable cover sheet. Within this basic design there are a large number of variations which manufacturers have developed in order to differentiate their product:

- a stay-dry versus fabric surface,

- with or without wings to hold the pad in place and protect underwear,

- shaped (e.g. butterfly shape) or straight-sided,

- flat or curved to fit the body,

- packaging - folded and individually wrapped or packed flat and unwrapped,

- central protection strip to wick moisture away from the centre of the towel and improve efficiency,

- thickness.

Perhaps the most important innovation in recent years has been the launch of new ultra-thin high-absorbency towels, such as the P& G product 'Always Ultra' and Johnson & Johnson's 'Silhouettes'.

(33) Tampons are used solely for menstruation. They are composed of a carded fleece of staple fibres compressed into a tampon. Tampons are internal devices and are either put in position manually (digital tampons) or with the aid of a plastic or cardboard applicator (applicator tampons).

(34) Both tampons and towels are sold in a range of absorbency capacities to cope with the variation in the rate of flow between women and between different days of a period.

(35) In Germany, approximately one third of women use only tampons, one third only towels and one third use both products. In this decision these groups are referred to as solus tampon users, solus towel users and dual users. The exact proportions of the groups depend upon the source: a Link study supplied by P& G shows these groups in ratio 34:22:34 while figures derived by the Gesellschaft fuer Konsummarkt- und Absatzforschung (GfK) and also supplied by P& G show 28:35:33. Studies for other Member States show still different proportions.

(36) Among dual users, some purchases are for use of a tampon and towel in combination while others are for separate use. Where a woman buys both products but uses each at a different time, this does not mean that she is indifferent as to which method she uses. Dual use is based on the differences between tampons and towels which make the products non-substitutable in the individual consumer's eyes for a particular use.

(37) While both internal an external methods meet the basic need for some form of menstrual protection in the broadest sense, women appear to use each for distinct purposes, and to develop usage patterns according to strong personal preferences. The Commission's investigation has shown that this is not true merely of solus users but also of dual users. A survey of dual users prepared for Moelnlycke in Germany showed that they had specific usage patterns according to the situation. Although the sample of this study is small, it is sufficient to establish the existence of these preferences, whilst it is true that a larger sample would be needed to quantify these attitudes more reliably.

<emplacement tableau>

Source: Study on dual users in Germany prepared by Infratest Burke March 1994.>

As can be seen, not all the women surveyed coincided in their choice of method for a particular situation, but what is of note is that such high proportions of women had a habitual preference for one or the other method for each situation. That the method preferred is not the same between women points to the intensely personal nature of the product and the vital role played by individual preferences.

(38) This is not to ignore certain objective constraints on the choice of method, e.g. only tampons can be used for swimming, towels are less discreet than tampons, and women may be unable to use tampons immediately following childbirth, but these constraints are very limited in impact. Of much greater influence are consumer preferences. The importance of such consumer preferences was recognized in the earlier case 92-553-EEC (Nestlé/Perrier) (26), where the Commission concluded that there are separate markets for soft drinks and bottled water because, while both quenched thirst, they were not recognized as substitutable by the consumer.

(39) The differences between countries in the ratio of solus to dual use is also a function of the relative popularity of towels and tampons between women in different countries. This can be seen in the different 'penetration ratios' or ratio of usage of tampons to towels:

<emplacement tableau>

Source: (A) Nielsen volume figures for September 1992 - August 1993 compiled by a competitor.>

(40) While the choice of protection method is very subjective and personal, there are certain general tendencies. A young woman's initial choice of both menstrual protection method and brand of product is often strongly influenced by the mother, elder sisters, close friends and health education programmes. Frequently the initial choice appears to be for a towel. Some young women then switch to tampons, particularly after leaving home. Some tampon users may then revert to towel use either permanently or temporarily after childbirth and in middle age. Medical or physiological reasons may oblige a woman to change her method of protection (e.g. heavy post-natal periods, choice of contraceptive method), and certain lifestyle and attitudinal changes may also be reflected in her choice of method.

(41) In selecting a method of menstrual protection, non-price issues such as comfort, security and discretion are central. Towels traditionally had the disadvantage of being thick and indiscret, although the new ultra-thin products overcome this problem to a large extent. Tampons still have some advantage in discretion and can be worn under tight-fitting clothing, while some women consider a tampon to be more comfortable, particularly for sports. Other non-performance considerations also play a role: towels may be seen by some women as more 'natural' because they allow the menstrual blood to flow and leave the body; for others there are perceived issues of morality in tampon use and for some women health concerns may play a role, particularly for teenagers considering whether to use tampons for the first time and who are afraid of damaging themselves by inserting a foreign body internally. Other health fears may arise out of toxic shock syndrome (TSS). The link between tampon use and TSS was first identified in the United States of America in 1981 and whilst in the USA and the United Kingdom almost all women are aware of the issue of TSS, elsewhere in Europe awareness of TSS is far lower.

(42) While it is true that both tampons and towels broadly perform the same function, they do so in such a different way that they are not regarded as substitutable by the consumer once she has established a preference or pattern of usage. Consumers appear to have very personal preferences for a method, or particular combination of methods in the case of dual users, and these are dictated as much by the physiological and physical considerations discussed above as by the technical characteristics of the product. Changes in a woman's choice of protection method or combination of methods is not proof of the substitutability of towels or tampons for these women, however. The fact that they have a preferred method which they change as a result of changes in their personal circumstances does not mean that the women are indifferent as to the method they use; quite the contrary: it is because of the differences in characteristics between internal and external methods that women change, the better to meet their new circumstances.

Own label

(43) Own-label towels account for about [...] (27) by value and approximately [...] (28) by volume of towel sales in Germany. The same figures for Spain are [...] (29) and [...] (30) respectively. The very size of this difference between value and volume market shares raises the question of whether branded and own-label towels are in the same market. The own-label and store brand towels are a lower quality, lower price alternative to the branded products and are sold at a discount to branded towels (for example branded towels produced by VPS retail at prices 100 % higher than the own-label towels they produce for distribution chains). Retailers stated that whilst there was a price-conscious segment of consumers who bought own-label towels, considerable price increase would be necessary to persuade purchasers of branded towels to trade down to own label. Furthermore, it is necessary to recognize that own-label products in Germany, in terms of quality and reputation, are arguably not as developed as they are in the United Kingdom, for example.

(44) The Commission considers that because of the sizeable discount at which own-label products are sold, there is only limited competition between branded and unbranded towels. Most products purchased involve for the consumer a trade-off between the best quality available and price. Sanitary towels would appear to be an eara where many women are less prepared to compromise on quality for the sake of a low price, but faced with a branded towel price increase, some women may reassess this trade-off and decide that the new price is simply too high and buy a cheaper towel, accepting lower performance.

Prices

(45) P& G argue that tampons and towels are priced in such a way as to compete head on with each other. In arguing this, P& G compare the average price of tampons (the tampon market leader in Germany) with that of Camelia pads (the market leader in pads in Germany in 1993) and show that they are the same. Whilst this is correct, it should be noted that Camelia's actual retail prices range from DM 0,136 to DM 0,377 per unit (excluding VAT) and cannot be said realistically to be priced against ob prices. The Commission considers also that it is appropriate to look at the prices of P& G's own brand Always, which has become the market leader since the beginning of 1994: its average unit price is approximately 150 % that of ob. The Commission considers that in such a mixed situation it is not possible to draw even provisional conclusions from the relative price levels of tampons and towels.

Third party views

(46) P& G's main competitors in Europe, Johnson & Johnson, Kimberly-Clark and Moelnlycke, all argued that there are separate markets for towels and tampons. Smaller competitors such as Rauscher and Hedwigsthal were of the same point of view. While the tampon manufacturer, Tambrands, at first stated that there is only one product market for both tampons and towels, subsequently it argued that there is in fact a separate product market for towels and that the Commission should give consideration to the availability of tampons as an alternative form of menstrual protection.

(47) The Commission also contacted a large number of retailers in Germany, the UK, France, Spain and Austria. All but one of these retailers considered that tampons and towels were in different markets. Of the 21 retailers who expressed a view on the effect of towel promotions, 15 said that such promotions would have no effect on tampon sales, three said that the effect would be minimal, while one felt that such a promotion would increase sales for all catamenials.

Stability of usage patterns

(48) As part of its evidence that women switch readily between tampons and towels, P& G commissioned a survey from GfK which asked women, inter alia, whether they would 'be prepared to switch to tampons/sanitary pads occasionally?' if a better pad/tampon were available or if the price of one method were increased by 10 %.

(49) The results of this survey showed that 44 to 65 % of women would consider occasionally purchasing a different form of protection, or using it more frequently on the basis of an improved product, and 31 to 50 % on the basis of a 10 % price rise in the other form. In evaluating what conclusions it was legitimate to draw from this survey, the Commission considered first that there is a considerable difference between a willingness perhaps to try a product or to use it on an occasional basis and a lasting change in consumer habits. Secondly, it is necessary in interpreting data from surveys such as this to correct for overstatement on the respondent's part. A reputable model developed by a market research company used by several companies in this market describes how it is necessary to adjust for the natural desire of the interviewee to please; the fact that it is easier to say yes than no, and that the interest generated in the issue at the time of interview may not be lasting and may not be reflected in actual purchase behaviour in the future. This model assumes that it is necessary to adjust such scores downwards by 50 to 80 %.

(50) A different consumer study commissioned in early 1993 by a Member State competition authority as part of an informal investigation of a national tampon market produced a very different result on the question of whether a woman would change her method of protection in the event of a price increase in one method. If prices for all brands of internal protection products rose by 10 %, then 95 % of solus tampon users would continue to use the same number of tampons (while 3 % would use fewer) and 80 % of dual users (tampons and towels) would continue to use the same number (12 % would use fewer). The same question asked for external products showed that the buying habits of 89 % of solus towel users and 78 % of dual users respectively would remain unchanged by a 10 % price increase while 11 % and 8 % respectively would use fewer. The degree of continued loyalty to a particular method is very high, particularly given the nature of this sort of study, which draws a woman's attention to a hypothetical price increase and emphasizes the possibility of switching. Especially significant is the large percentage of dual users who would be reluctant to change the intensity of their usage of a product, even though they are familiar with and actually use the alternative product.

(51) P& G has provided data on the purchasing behaviour of individual women from the GfK purchase panel in support of its claim that women do not have a stable usage pattern of menstrual protection products. The purchase panel data shows that of those women on the panel who had purchased both a tampon and a towel in the first half of 1993, only 57 % had also purchased both a tampon and a towel in the second half of the year. This would seem to imply that nearly one half of all dual users become solus users every six months. Similar panel data from A. C. Nielsen's household purchase panel of households with only one woman comparing nine-month periods show a more rigid purchase pattern, however. Of exclusive towel or tampon users in the first period, 87 % and 82 % respectively remained exclusive users in the second period. When the Nielsen panel data for the same periods are examined in terms of the number of units sold, which is a more important measure when assessing the price-constraining impact of any marginal substitutability between the two forms, it can be seen that for those women who only bought tampons in April to December 1992, only 6,3 % of their purchases in January to September 1993 were of pads. Similarly, 8 % of the purchases in the second period of those women who were exclusive pad purchasers in the first period were of tampons.

(52) However, the fundamental problem in using purchase panel data in this context is that they reflect purchase patterns and not usage patterns. Average purchase cycles of towels are, according to Nielsen data, around 1,5 to two months and close to three months for tampons for all households in the panel. Women who are predominantly towel users, for example, may only buy tampons once a year or even less frequently and vice versa for predominantly tampon users. This problem is particularly relevant if comparisons are made between relatively short time periods such as six months. Furthermore, the panel data cannot explain why a purchase is made. In order to understand whether usage patterns are stable or continuously changing it would be necessary to know how many purchasee took place owing to such factors as seasonal variations, child birth, women approaching menopause, and trial purchases.

(53) P& G's arguments contradict the views of both competitors and retailers that women are particularly conservative and stable in their buying of feminine protection products, and are in fact highly loyal not only to a particular method or combination of methods but also to individual brands.

Price and cross-price elasticity

(54) P& G claims that sales of towels are influenced by the price level of tampons and that they are therefore in the same product market. GfK scanner data provided by P& G purports to show a high own price elasticity for Always as well as the existence of cross price elasticity between towels and tampons. The scanner data from GfK of weekly sales for 52 weeks from a sample of 120 stores with a surface of more than 800 m² thus shows that a [...] reduction in the price of Always produced on average a [...] increase in Always sales (own price elasticity of [...] (31)) and a decrease in sales of ob tampons of [...] (cross price elasticity of [...] (32)), while a [...] increase in Always' price resulted in a [...] drop in Always sales (own price elasticity of [...] (33)) and an [...] increase in sales of ob (cross price elasticity [...] (34)). These data therefore seem to indicate a rather high own price as well as cross price elasticity to ob, in case of a price decrease. On the other hand the own price and cross price elasticities are much lower in case of a price increase. P& G has provided scanner data from other sources as well. The type of analysis is the same as for the GfK data and shows similar results.

(55) This type of analysis suffers, however, from several weaknesses. First, there are technical problems in calculating elasticities in this way. An own price elasticity calculated as [...] is for example an average of the effect of price reductions in the range 0 to 20 %. A 10 % price increase could for example exhibit a much lower price elasticity than a 20 % increase. Second, the calculation is only based on weeks when the price of Always was reduced. The analysis is therefore based on a rather limited number of observations. Thirdly, and more fundamentally, this type of analysis explains all variations in sales simply in relation to price changes on Always. This is obviously a very simple approach, which confuses variations in sales caused by such different factors as seasonal variations or promotions in other stores not included in the scanner sample with general price changes.

(56) A better way to estimate elasticities is to estimate a model including as many relevant variables as possible, i.e. to use a multivariate rather than a univariate approach. In this way it is possible to distinguish the various effects of the explanatory variables from each other.Therefore, a model based on a multivariate approach explains market behaviour in a more comprehensive way than would just price variations, and therefore reduces the risk of confusing one type of variation with another.

(57) P& G has not provided an econometric analysis based on the GfK scanner data. However, A. C. Nielsen has developed an econometric price-promotion model ('Scanpro') based on scanner data similar to the GfK data, by means of which it is possible to estimate the effects of price promotions and general price changes whilst eliminating the effect of seasonal and other variations. The analysis is a standard product sold by A. C. Nielsen in many different countries for many different categories of consumer products and was supplied to the Commission in this case by Johnson & Johnson. The purpose is to enable companies to plan their promotional efforts better. An analysis was carried out using this model to compare Always Super Thin with ob. Such ultra towels are the towel type which could be expected to compete most closely with tampons. This analysis showed that neither ob promotions nor its long-term price development had had any significant impact on the sales of Always Super Thin when measured on a 95 % confidence level.

(58) The multivariate analysis by A. C. Nielsen provides additional useful insights and is a better way than the univariate analysis from GfK to understand the complex behaviour in the market place. It is, for example, likely that the price elasticities from the GfK analysis to a large extent reflect promotion effects. If they reflected long term rather than promotional elasticities, then even small price changes would lead to dramatic shifts in total market demand for ob tampons and Always Super Thin towels from one year to the next. P& G has also recognized that the GfK data could not be projected to the total market.

(59) These results are consistent with attitudinal survey data indicating that price is relatively unimportant for the consumer for this type of product. Furthermore, the analysis of scanner data provided by P& G contradicts the views of almost all retailers questioned, who said that promotions on towels resulted in reductions in sales of other towel brands but not of tampons. Where there was a slight effect, it was felt that this was due to stocking up by regular purchases, particularly dual users familiar with both products, of a brand when that brand was on promotion. However, as has been seen, a more elaborate econometric analysis makes it possible to reconcile the results from GfK's analysis with the views of retailers and competitors as well as providing elasticity estimates which can be projected to the market level.

(60) Whilst the A. C. Nielsen model is still, in some respects, crude, its results were confirmed by a separate multivariate analysis made by the American RLS company which was submitted to the Commission by a competitor. This study gave a price elasticity of 0,5 for pads, a cross-price elasticity of 0,3 for pads with respect to non-promoted price changes, and non-significant promotional elasticities. The study thus confirmed a low level of cross-price elasticity between tampons and towels found by the Nielsen study. Whilst this study relates to the USA and not Germany, the Commission notes that key market characteristics are similar: dual users constitute 39 % in Germany and 32 % in the USA while the penetration ratio is 40:60 in the USA and 39:61 in Germany.

(61) The conclusions of the econometric analyses above are confirmed by observations at the market level.

Impact of launch of Always

(62) Were tampons and towels in the same market, it could be expected that events such as the entry of Always into the towel market and price increases in either market would have repercussions on the other market.

(63) Always was launched in Germany in July/August 1991 and the following table shows the response within six months and two years of average tampon and towel prices.

<emplacement tableau>

Source: Nielsen figures collated by a competitor.>

As can be seen, average prices of towels other than Always at first fell slightly and then rose by only 2,3 % in two years. Tampon prices on the other hand rose slightly in the half year following the introduction of Always and then rose 18,2 % over the next two years.

(64) P& G has argued that tampons are substitutable for towels and that its Always towels have taken market share from tampons since their launch. The table below shows the relative volume share of towels and tampons of the combined towel and tampon market in Germany (referred to as the 'penetration ratio'), where Always was launched in July 1991.

Evolution in the penetration ratio (pads/tampons) showing P& G's share of towel usage <emplacement tableau>

Source: Commission calculation on the basis of Nielsen volume figures supplied by P& G.>

(65) It should be noted for this table that the market growth in 1991/92 is thought to result from the inclusion of the new German Laender since reunification. It is necessary to use volume figures here in order to assess whether switching has taken place. As can be seen, in Germany the entry of Always in July 1991 did not result in any reduction in tampons' share of the combined tampon and towel markets.

(66) P& G argue that this picture is misleading and that while the penetration ratio is stable in Germany, Always has in fact obtained [...] of its market share (and [...] of its gains from other brands) from tampons, and that tampons have gained a similar volume from towels, particularly the thicker and also less premium variants. This analysis was prepared for P& G by GfK, and is again on the basis of a purchase panel with no explanation of the results obtained. Furthermore, the shifts shown in the graph in the next paragraph for the old Laender of Germany are repeated in virtually all EEA states following the launch of Always, i.e. tampon volumes were unaffected. It seems very improbable that in all these cases exactly compensating shifts in volume from tampons to Always and from other towels to tampons occurred.

(67) The impact of the launch of Always in the old Laender of Germany is shown in the following graph:

Sales in mio. units in West Germany

(68) As can be seen, tampon sales in 1993 are slightly higher than in the previous year and there appears to be little evidence of Always taking from tampon volume.

(69) What can be seen in the next graph, supplied by a competitor but based on Nielsen data, is that Always has taken large market shares from other towel brands since its launch in the old Laender of Germany, particularly from VPS's Camelia brand:

NAPKIN MARKET SHARE DEVELOPMENT (WEST; VALUE)

(70) Another example of the low levl of interplay between the two markets can be seen in the price rise of ob tampons in May/June 1993. Ob packs were reduced from 80 to 64 units while price was held constant (an effective 25 % price increase) and ob, with a 75 % share, is a good proxy for the German tampon market as a whole. Always prices remained constant following the ob move and its rate of growth in market share not only did not increase but slowed through the last six months of 1993. Similarly, volume of all towels did not show any marked response to the tampon price increase.

(71) Figures supplied by Moelnlycke show that in the Nordic countries, where Moelnlycke sells both tampons and towels, it dramatically increased its advertising and promotion spend on towels in response to the launch of Always in mid-1992 (mid-1993 in Norway) but there was little increase in the level of spend on tampons. Between 1991 and 1993 towel A& P expenditure was increased by over 200 %, whilst that on tampons by only 34 %.

Targeting of advertising

(72) The parties stated that manufacturers of both tampons and towels launched publicity campaigns aimed at encouraging users of external products to switch to internal ones and vice versa, arguing that manufacturers would not make such expenditure unless the products were in the same market. In the Commission's view, such advertising is not incompatible with the existence of separate markets because there is indeed a degree of switching in the long term and over the course of a woman's life as described earlier. The point of such advertising campaigns is to encourage such switching and to familiarize women with the arguments for switching in advance of major changes in their lives. Furthermore, much of this advertising is directed at adolescent women still experimenting with the two methods.

Supply-side substitutability

(73) Tampons are manufactured in an entirely different fashion, from a compacted fleece of staple fibres, while towels are made from 'cellulose fluff' which is the same wood-based raw material used in the production of nappies and household paper products. The technologies, while relatively unsophisticated and inexpensive, are unrelated.

Conclusion on relevant product market

(74) P& G argues that there is only one feminine hygiene market for towels, tampons and pantliners, and based itself upon a variety of household panel, scanner data and price studies. Whilst the Commission recognizes the seriousness of the research presented, it considers that the data is too narrowly specific to support the broad market conclusions which P& G draws from them. Furthermore, the Commission has examined market research studies provided by third parties which, while also presenting certain methodological difficulties, are broader-based and appear strongly indicative of separate markets. The Commission has been assisted in its analysis of the conclusions which it is legitimate to draw from the market research materials at its disposal by two independent experts in market research analysis.

(75) While none of the market research studies presented to the Commission is decisive, it considers that several are strongly indicative of the existence of separate markets for pads and tampons. If this indicative evidence is placed alongside the Commission's own analysis of the market and the views not only of other suppliers but also of retailers, with all of which it is consistent, the Commission concludes that there are indeed clearly separate markets not only for pantliners but also for pads and tampons.

B. Relevant geographic market

(76) P& G has submitted that the relevant geographic market is Western Europe, arguing that the major suppliers of feminine hygiene products are active in several or even most of the EEA countries and that European brand names used by these manufacturers are predominant in each country. Furthermore, they refer to the absence of legal of tariff barriers, to a high degree of standardization in advertising and packaging, low transport costs, European sourcing and an increasing pan-European purchase policy of retailers.

(77) In view of the structural elements set out below, and after examination of the competitive environment in the sanitary towel market in the EEA, the Commission has, however, concluded that at least in respect of Germany and Spain the relevant geographic market within which the market power of Procter & Gamble post-merger has to be assessed is national. Several factors indicate that the parties to the concentration are and will be able to determine their competitive behaviour in Germany and Spain without suffering significant competitive constraints from outside these countries. It can be left open whether Austria also constitutes a separate geographic market for sanitary towels as the concentration does not lead to the creation of a dominant position in Austria as set out below.

1. Consumer preferences

(78) Consumer demand for sanitary towels presents different characteristics in each Member State. This can be seen not only in the variation of the relative popularity of towels as opposed to tampons between Member States but also, and more importantly, within the category of sanitary towels. Whilst traditional thick and wingless towels are still the most commonly used sanitary towels (about 70 % in volume in the large Member States), innovations such as ultra-thin and winged-towels have led to varying consumer preferences. Ultra-thin, winged towels, for example, are used by French and British consumers more than by Spanish, Italian and German consumers. Thick, winged towels show the opposite tendency. A greater proportion of women use ultra-thin wingless towels in Germany than in France and the UK. While these usage patterns may not be sufficient on their own to identify national markets, they must be seen in the context of other points set out below.

2. Prices for sanitary towels

(79) The Commission investigated the level of ex-factory prices and prices paid by distributors net of all discounts and rebates (net net prices) for towels in different Member States. It then indexed the prices, taking the average as 100. In the case of Procter & Gamble, the difference between the highest and the lowest indexed ex-factory prices in France, Germany, UK, Belgium and the Netherlands was [...] (36) points. The same figure for net net prices was [...] (37) points. There were significantly higher differences in competitors' prices, however, from country to country. For each competitor, the difference between the highest and the lowest prices in these countries was up to 34 points for ultra-thin winged towels (in terms of net net prices) and up to 52 points in ex-factory prices. Maximum differences in thick, wingless towels range between eight and 34 points for different suppliers (net net prices) and 19 to 57 points (ex-factory prices) (38).

(80) In Spain and Italy, ex-factory prices and net net prices range significantly below the prices in all other Member States (notwithstanding the rapid growth in prices over recent years from a low base). This is particularly true for thick towels, where products by P& G (Arbora) or by Johnson & Johnson are more than 50 points cheaper in Spain than in Germany and other Member States. Even for ultra thin towels, prices differ by more than 25 points between Spain and other countries. This is equally true for ex-factory prices and for net net prices. Spanish retailers contacted by the Commission confirmed that the ultra-thin products Evax and Ausonia supplied by the two joint ventures by P& G and other companies are equivalent to the ultra-thin towel of Always.

(81) Prices for towels have evolved at markedly different rates in Germany and other Member States. According to Nielsen figures supplied by a competitor, between January to April 1991 and January to April 1993 average towel prices in Germany rose by 10,6 %, whilst between May to August 1991 and May to August 1993, average towel prices rose by 24 % in France, 43 % in the UK and 45,5 % in the Netherlands.

3. Supply side

(82) While it is true that the big suppliers of sanitary towels operate in several Member States and EFTA countries, it cannot be disregarded that there are 'home markets' for each supplier and that market shares vary significantly even between neighbouring countries.

(83) Companies which have strong market positions in countries bordering on Germany (e.g. Kimberly-Clark's [...] (39) share in the Netherlands market, Moelnlycke's [...] (40) in Denmark, [...] (41) in France and [...] (42) in the Netherlands) are either not present or only to an insignificant extent in Germany itself. Even Johnson & Johnson, the strongest competitor of P& G in Germany post-merger with [...] (43), holds a significantly higher market share ([...] (44)) in Austria. Suppliers with strong positions in France (Kaysersberg, Moelnlycke and Kimberly-Clark) are similarly absent from the Spanish market.

(84) P& G itself markets sanitary towels and pantliners in 13 EEA countries and by means of joint ventures in a further three: Italy, Spain and Portugal. VPS sells the full range of feminine hygiene products in Germany, Austria, Switzerland, Spain and Italy. VPS production facilities, however, are located in Germany with a plant in Spain. P& G sources [...] of its towel products for Italy and Spain from its plants in these countries. P& G's German plant in Crailsheim supplies Germany ([...]), France ([...]), UK/Ireland ([...]) and other EEA countries (around [...]).

(85) The other main suppliers all sell their products in a range of EEA states, but they also have several production sites. Johnson & Johnson, which is the most widely active of Procter & Gamble's competitors, has six production sites in Western Europe. Moelnlycke, which operates primarily in France, the Netherlands, Belgium, UK, Denmark and the Nordic countries, has plants in the UK, the Netherlands, France and Norway, although in January 1994 they also acquired the Italian company Sodalco ('Nuvenia'). Kimberly-Clark is mainly active in the UK, Ireland and the Netherlands and supplies other countries such as France from its plants in the UK and the Netherlands. Thus none of Procter & Gamble's competitors sources its supplies from a single site and while they do not have separate production facilities in every EEA country, there appears to be a regional supply correlation. In any event, the mere centralization of plants would not, in itself, be incompatible with separate national markets.

(86) The investigation made by the Commission as to the destination of towels produced in each plant showed that most of the suppliers and in particular Procter & Gamble situate their plants in a way which avoids long distances of transport, at least for large volumes. In particular, the principal markets such as Germany, the UK, France, Italy and Spain, are thus supplied over relatively short distances. One competitor explained that in order to enter an EEA country's market it purchased a local company because it was uneconomic to supply from its geographically remote production sites. Procter & Gamble has confirmed that a supplier needs a manufacturing location which allows affordable distribution costs. It gave figures for transport costs of [...] (45) for deliveries from Germany to France and [...] (46) from Germany to the UK, while a competitor provided costings that show that for journeys to the UK, Sweden and Italy, transport costs represent [...] (47), [...] (48) and [...] (49) respectively of their inter-company transfer price. Most of the suppliers provide their destinations from neighbouring countries or at least over relatively short distances.

(87) From the limited number of production facilities of the big suppliers it can be concluded that economies of scale are more important to some producers than the saving of transport costs. On the other hand, the impact of transport costs is, as was shown above, far from insignificant, and the big suppliers apparently tend to supply their major markets over relatively short distances.

(88) Procter & Gamble themselves submitted that 90 % of suppliers to the German market for sanitary towels, for example, are provided by German suppliers (VPS, Pelz, Hedwigsthal) or by German plants of international suppliers (Procter & Gamble, Johnson & Johnson). The remaining share falls to private label supplies, which are partly imported. Other than this, imports are minimal although there are no legal or tariff barriers. In spain local producers with local plants (including VPS) supply over 80 % of the market, and again most of the remaining share falls to private label.

4. Brands

(89) Procter & Gamble argues that European brand names are predominant in Western Europe and that, even where names vary, manufacturers are using transnational technologies and multilingual packaging. The enquiries of the Commission did not confirm this, particularly in Germany and Spain. The existence of some 'Eurobrands' does, in any case, not automatically prevent the market from being essentially national.

(90) None of the suppliers of sanitary towels uses the same brand name across Western Europe, not even P& G, although it is present in 10 out of 12 EC Member States. The following survey of Member States reveals a substantial variety of brands which shows that the competitive structures in the different countries still differ.

<emplacement tableau>

(91) As can be seen, brands can be classed either as European (Always and Silhouette) or essentially national. Some of the latter category may extend over two or three neighbouring countries with close ties (e.g. Vania in France and Belgium, Ausonia and Evax in Spain and Portugal). A large proportion of sales in Europe are actually made by important, essentially national brands (eg, Lines and Nuvenia in Italy, Camelia in Germany, Evax and Ausonia in Spain). Since brands like Serena and Vespré (Johnson & Johnson), Kotex and Freedom (Kimberly-Clark) and Libresse and Nana (Moelnlycke) cannot, or cannot yet, be qualified as Pan-European brands, it can be said that sales of current European brands - essentially Always and Silhouette - achieved [...] (51) in value and [...] (52) in volume in 1993 in Western Europe (based on Nielsen market share data from P& G and sales figures from Johnson & Johnson).

5. Demand side

(92) The German retail trade remains substantially national in character. Commercial terms for German retailers differ significantly from those for French or Italian retailers for example. Trade margins are much lower in Germany than in the United Kingdom, for example, and the number of retail outlets differs significantly from some other Member States (only 70 000 in Germany compared to 100 000 or even more in Italy and Spain, despite Germany's far greater population). Own-label products in Germany have a lower quality image and are less important in competitive terms than they are in other markets, e.g. the United Kingdom, where own-label is a much more developed concept and is priced closely to premium branded products. Accordingly, price differences between premium brands and private label products are significantly larger in Germany. In the case of VPS, the average retail price excluding VAT of its branded Camelia products in Germany is DM [...] (53) as compared to DM [...] (54)-DM [...] (55), for its distributor brands.

(93) Purchasing is conducted on a national basis. German, United Kingdom and French retailers stated that their contracts with multinational manufacturers of feminine hygiene products such as P& G were negotiated through the national sales organizations of those manufacturers and at a national level. The existence of national sales organizations is not only, as P& G submitted, due to language differences among EEA States. As was confirmed by retailers contacted by the Commission, retailers claimed to be unaware of, and indifferent to, prices in other countries, claiming that an attempt by a French distributor to purchase from Procter & Gamble (Germany) would result only in their being referred to Procter & Gamble (France). If the major distribution groups of France, Germany, Spain and the United Kingdom are unable to do this, it cannot be expected that the smaller and more fragmented retail groups in Spain and some other EEA states should do so. Only private label products might be purchased abroad to a limited extent by retail chains anxious to ensure the lowest prices. The retail chains then of course impose their own branding on the goods to conform to local tastes.

(94) Transnational buying groups, the growing importance of which has been stressed by P& G, appear not to play a significant role in the purchase of, or negotiation of terms for, feminine hygiene products. Furthermore, it would appear that large suppliers such as Procter & Gamble are in many cases able to rebuff attempts by such groupings to impose European conditions. Many such groupings are in any case simply groupings of large retail groups which occasionally operate common European promotional actions.

6. Market entry barriers

(95) In addition to the varying competitive conditions prevailing in different EEA States, there are significant barriers to entry into the national markets for sanitary towels. These barriers (which are discussed in detail in the Assessment) in themselves indicate that the geographical market is national.

High level of concentration

(96) The high degree of concentration in the German towel market, with three companies - Procter & Gamble, Johnson & Johnson and VP Schickedanz - holding [...] (56) (in value) and [...] (57) (in volume), constitutes a considerable barrier to entry. It increases the risks associated with new entry in the sense of a reaction by the established suppliers against newcomers with a view to defending the acquired market positions and profitability. The same is true of the Spanish market, where Procter & Gamble alone holds [...] (58) by value and [...] (59) by volume.

Brand loyalty

(97) The existence of a relatively high level of brand loyalty in the sector of sanitary towels makes it difficult to persuade users to switch to a new product and, for suppliers, to enter the market. In countries such as Spain and Italy, P& G entered the sanitary towel market by acquiring strong national brands which were retained afterwards, because of the value represented by these brands, i.e. the loyalty of customers to them.

Access to distribution

(98) The towel market is a mature market in terms of the number of brands and range of products. In Germany and Austria, access to retail chains with a new brand is not impossible, it was shown by the introduction of Always, but requires an innovative and premium product and substantial efforts in advertising and promotion (sampling, etc.).

Advertising sunk costs

(99) The establishment of a new brand would require heavy investment in advertising and promotion in order to persuade brand loyal customers to switch away from their usual brand. Such expenditure is a sunk cost and adds to the risk of entry.

Past attempts of market entry

(100) The barriers to entry to the German market in particular can be illustrated with reference to the failure of several attempts in recent years to enter the sanitary towel market in Germany, notwithstanding the success of the launch of 'Always' due to P& G's high level of commitment and investment in its launch (see Assessment).

Conclusion

(101) Factors such as the possibility to price differently in Germany and Spain as opposed to other EEA States, the brand loyalty of customers, the high level of existing concentration and the investment risk associated with entry, form a conglomeration of factors creating an important barrier to entry to the German and Spanish markets. This has been confirmed almost unanimously by the retailers and industrial suppliers consulted by the Commission. The price differentials between Germany and other countries, the importance of the Camelia brand, national purchasing by retailers (even by large chains) and the failed attempts at market entry, all demonstrate Germany to be a national market. Similarly, the market structure in Spain is characterized by the very high combined positions of Ausonia and Evax, produced by Arbora and Ausonia, which are jointly controlled by Procter & Gamble. Other suppliers do not play a significant role. Spain, thus, presents the highest level of concentration within the EEA in sanitary towels. The existing brands, especially those mentioned above, are purely national; Always has not been introduced in Spain. In addition, the Spanish retail sector is still markedly fragmented in comparison with Germany.

(102) P& G subsequently argued that the relevant geographic markets were regional. They identified three such regional markets: the nordic countries; other north European countries (Benelux, Germany, France, the United Kingdom and Ireland); and southern Europe. Whilst it is true that these groupings of countries may share certain common characteristics (such as penetration ratio), there are significant differences in the supply structure even within these groupings. In particular, major players may be absent or hold radically different market shares between neighbouring countries. Kimberly-Clark and Moelnlycke are not present on the German market but have considerable market shares in neighbouring countries. VPS, which was the market leader in Germany until 1993, has no presence in France, Benelux or the United Kingdom while the Belgian and French markets are characterized by important brands which are not supplied to neighbouring countries (Vania, Nana). Similarly the major brands in Spain, Ausonia and Evax, are not present in Italy while major Italian brands such as Lines and Novenia are not present in Spain. While it cannot be excluded that there might be a France-Benelux market, the differences in market structure and brands would exclude Germany and Spain from such a market.

Prospective of the relevant geographic market

(103) P& G submits that the relevant geographic market is already Western European. In particular, they refer to the presence of large transnational suppliers in the sanitary towel sector across Western Europe and to the increasing importance of European brand names. P& G itself appears to follow a European strategy with Always and in the light of the results of its investigation the Commission recognizes that there are some indications of a development towards a European market. At least one supplier other than P& G is attempting to harmonize national or regional brand names while others are using multilingually-labelled packages. Furthermore, developments in the retail trade, such as cross-border acquisition, and formal and informal transnational groupings of retailers, suggest the first tentative signs of an integration and Europeanization of the retail trade in the EEA. Relatively low transport costs and centralization of plants may support this view.

(104) This does not mean, however, that the Spanish and German markets are anything other than national at present. Nor will the completion of a common market in this sector be achieved within the foreseeable future (e.g. within three years). In a sensitive, brand-loyal market such as this one, suppliers cannot switch to a European brand name too quickly without losing sales. The timescale for a possible Europeanization of the market exceeds the relevant perspective to be established by the Commission under the Merger Regulation and the assumption that markets will indeed become European at the end of this period must be regarded in the light of the evidence as by no means certain. The fact that, in the longer term, the sanitary towel market may, in fact, become European cannot therefore be taken into account at present.

C. Assessment Overview of the towel market

(105) The market for sanitary towels has been growing at less than 5 % in volume in recent years in Europe. With population growth relatively stagnant, increased demand is mainly due to use of a greater number of product units per period by women. In value terms the market has grown enormously since 1991/92. According to adjusted Nielsen figures supplied by P& G, the West European towel market has grown by 9 % and 16 % over the last two years while the same figures for Germany show increases of 32 % and 21 % in the value of the market. The West European towel market was worth over ECU 1 500 million per annum as of June 1993, while the German market was worth ECU 269 million per annum. This growth in the value of the market is due to the introduction in the early 1990s of new, sophisticated products which, for example, were thinner, considerably more absorbent and had 'wings' (adhesive tabs on the side of the towel to hold it in place) or were body-shaped, etc. These innovative products were and are able to command a considerable premium over more traditional products.

Suppliers

(106) The major players in the towel market in the EEA are P& G, which makes towels and pantliners, and Johnson & Johnson (J& J), Moelnlycke and Kimberly-Clark, which all make towels, pantliners and tampons. P& G is present in all EEA countries and Switzerland, although its presence in Spain, Portugal and Italy is through joint ventures with local companies. J& J is also present in all EEA states except Denmark, although not for all three products. Moelnlycke's business is centred on the Nordic countries, France, the UK, Benelux and Greece, while that of Kimberly-Clark on France, the UK, Benelux and Ireland, with a marginal presence in Germany.

(107) Johnson & Johnson is part of a worldwide personal care and toiletries company. It is the next largest towel producer after P& G in Western Europe, with [...] (60). It is the largest supplier of pantliners in Western Europe with [...] (61), and the second largest tampon producer behind Tambrands, with [...] (62). In Germany, Johnson & Johnson is the third largest towel producer, behind P& G and VPS, as well as being the largest tampon supplier in Germany, with about 80 %. In Austria it is the largest towel producer, with a market share of [...] (63) by value, but the merger of VPS and P& G will result in P& G becoming the new market leader, with [...] (64). In Spain, Johnson & Johnson has only minimal (1 %) share of the towel market, although it is stronger in other catamenial products in Spain. Its towel product range includes ultra-thin high-absorbency towels similar to the 'Always' ultra towels, which were launched at a similar time and are also premium priced.

(108) Moelnlycke, which owns Peaudouce of France, is the second largest towel manufacturer and also makes tampons and pantliners. It is wholly owned by Svenska Cellulosa AB, a Swedish corporation. Geographically its strength lies in Scandinavia, with a regional market share in excess of 50 % for pads (over 90 % in Norway), and it also has significant shares in the UK, France, Netherlands and Belgium. It is not present in Germany, Ireland, Spain, Portugal or Austria and has just bought a company in Italy which markets the Nuvenia brand.

(109) Kimberly-Clark is one of the leaders in the US market for towels but has significant market shares only in the UK, Ireland, France and the Netherlands. It is not present in Austria and Spain and has only a 0,9 % share of the German market. It also makes tampons and pantliners. Kimberly-Clark does not market an ultra-thin product in Europe.

(110) Further competitors in Germany are Pelz and Hedwigsthal, both of which have market shares clearly well under 5 %. Aldi, the retail discounter, sells an own-label brand which has [...] (65) in value and [...] (66) in volume of the German towel market according to a P& G estimate.

(111) Kaysersberg is active in the household hygiene paper and sanitary protection sectors. The only catamenial products it produces are sanitary towels, sold under the brand name 'Vania', primarily in France and Belgium. Since February 1990, Kaysersberg has been the wholly-owned subsidiary of a Dutch holding company named Jamont NV, jointly owned by James River Corporation and Cragnotti and Partners. Kaysersberg is active in France in the feminine hygiene and adult incontinence protection sector through two companies: Vania Expansion for retailer chains and Polive for pharmacies. Both are owned 50/50 by Kaysersberg SA and Johnson & Johnson but are controlled by Kaysersberg alone.

Market shares

(112) In assessing the market shares of the parties, the Commission considers the most appropriate measure to be that of market share measured by value. The parties, however, have argued that volume market shares are more appropriate because the difference in price between different towels does not arise because they can be used for longer or because fewer of them can be used, but simply because the more expensive ones are more comfortable or perform better when used. Thus the number of units used by a consumer is a good indication of the hold of the manufacturer over that consumer in terms of future purchases. Whilst this might be true if higher prices did not reflect higher quality, in the case of the towel market they generally do.

(113) Feminine hygiene products, and more particularly towels, are heterogeneous products differentiated according to size, absorbency (mini, normal, super or overnight) and the special features described earlier. These differences, which determine the attractiveness of the product to a consumer, are reflected in its price and thus in value market shares.

(114) In the case of sanitary towels, premium branded products cost between 50 and 100 % more than secondary brands or private label products. The most expensive towels are the technically sophisticated ultra type, which constitute the growth sector in the market as a whole. A company's strength in this growth sector, which is reflected by its value market share, is an important indication of its competitive positioning.

(115) The ability of a manufacturer to command a higher price for its products than for competing products, whether as a result of product innovation, advertising, branding or marketing, is an important indication of the relative market power of that company compared with its competitors. This ability is reflected in value market shares but not in volume.

(116) Furthermore, in a sector where heavily promoted branded goods predominate, value shares are an approximate indication of the financial resources available to a manufacturer for reinvestment in its brand, particularly through advertising.

(117) For these reasons, the Commission considers that market shares in value better reflect the real market strength of companies in this market than market shares in volume (see also Commission Decisions 92-553-EEC of 22 July 1992 'Nestlé/Perrier', recital 40 and 93-252-EEC of 10 November 1992 'Warner Lambert/Gillette' (67), recital 22).

(118) The following market shares were supplied by P& G and are for both volume and value. The figures are based on Nielsen data and have been adjusted by P& G to account for Nielsen's imperfect coverage (68). The data given in the table below is for those three national markets where the operation results in a significant increase in market share for P& G.

(119)

National markets, market shares of sanitary towels - 1993

<emplacement tableau>

Source: A.C. Nielsen adjusted by P& G. Source: A.C. Nielsen adjusted by P& G.>

Germany

(120) The transaction will give rise to an addition of market shares on national markets in the four markets where VPS is present. Even after taking into account the divestment of VPS's non-Camelia business offered by Procter & Gamble, its post-merger market share would be [...] (70) in value ([...] (71) in volume) of the German market. Without any divestment of the non-Camelia business the shares would be [...] (72) (value) and [...] (73) (volume). The next largest competitor after the merged entity would be Johnson & Johnson with [...] (74). Aldi's share of the market through its sales of own-label towels is estimated by P& G to be [...] (75), while small producers like Pelz and Hedwigsthal have together [...] (76) (their individual shares are not recorded by Nielsen). Kimberly-Clark has only [...] (77) of the German towel market.

(121) Procter & Gamble's market share continues to grow, although the rate of increase has recently slowed somewhat in Germany. In the year to June 1992 following the launch of Always in Germany, P& G gained a [...] (78) (in value) which increased to [...] (79) in the year to June 1993 and to [...] (80) in the year to December 1993. VPS's Camelia brand, like other brands, suffered considerable losses in market share in this period, so the combined market share of P& G and Camelia rose more slowly, although there has been a steady increase (81).

<emplacement tableau>

Spain and Italy (122) In Spain, P& G's already strong position via its joint ventures Arbora and Fater of [...] (82) by value ([...] (83) in volume) will be increased by [...] (84). Johnson & Johnson's share is only [...] (85) value ([...] (86) volume), with the bulk of the remaining sales in the market being own-label sales and other small producers.

(123) In Italy, the operation will give P& G, through its joint venture with Finaf which currently has a [...] (87) ([...] (88)) share of the Italian market, an increase of only [...] (89) in market share.

Austria

(124) In Austria, VP Camelia will add [...] (90) to P& G existing [...] (91) market share, giving the merged entity a total share of [...] (92) (value) and [...] (93) (in volume). The nearest competitor is Johnson & Johnson with [...] (94) ([...] (95) in volume) followed by a local producer Rauscher with [...] (96) ([...] (97)).

Barriers to entry Brand loyalty

(125) Should the chosen method of menstrual protection fail, the consequences are both physically unpleasant and socially embarrassing for the woman. Women appear to place great importance on the security and reliability of the product. They are therefore reluctant to experiment with an untried product. Most women display a very high level of loyalty not only to their chosen method of protection but also to a particular brand of product. They have a brand in mind when they shop. As one United Kingdom retailer phrased it: 'Customers do not browse in a store'.

(126) The Commission has examined a number of consumer research studies which show a high level of brand loyalty:

- data supplied by a competitor and prepared by GfK shows that in 1993, of total Always volume in Germany, 49 % is sold to women who are 100 % loyal to Always for their purchases of towels throughout an entire year,

- a study prepared by Dymparez in 1993 on the Spanish market concluded that the importance of brands and brand loyalty was growing and that 54 % of consumers always buy the same brand,

- a study conducted by Infratest Burke in Italy for a competitor of P& G in March 1994 shows that 89 % of women questioned concurred with the statement 'I have been using the same brand of protection for a long time'.

(127) P& G's own data for Germany shows that [...] of Always towel customers over 100 % loyal to the brand over a year. A further [...] bought Always towels in 67-99 % of their purchases. Thus the effective level of brand loyalty to Always for towel users was that over half used Always alone or almost alone.

(128) Such levels of brand loyalty reflects the cost of failure, force of habit and desire for familiarity with a trusted product in such an intimate area.

(129) Furthermore, some women distinguish between different types of towel for different purposes and buy one brand of ultra towel and another brand of maxi/overnight towel. Such women may be totally loyal to the two brands but will be shown up in statistics such as these as disloyal.

(130) Such brand loyalty acts as a barrier to entry since, with a market which is growing at under 5 % by volume per annum, any entrant must take market share away from an incumbent. To do this, the entrant must persuade women not only to try out their product, but also to buy it on a repeated basis and to stop buying their habitual brand.

Price insensitivity

(131) When Always Ultra and the other innovative towels, such as J& J Silhouettes, were launched, they were priced at a considerable premium compared with ordinary towels. While they were innovative and offered superior performance, the size of the premium can be seen as indicative of the low level of price sensitivity for this product, as can the size of the price range for pads in Germany: from ECU 0,052 to ECU 0,181 each, i.e. the most expensive is over three times the price of the least expensive. Consumer surveys show price not to be a decisive factor for the majority of women in their choice of towel:

- 48 % of dual users surveyed in Germany for a competitor of P& G said that they would still buy their preferred brand of towel if it increased in price by 10 %, even if other brands were still available in the store,

- 64 % of women surveyed in a study made by Infratest Burke in Italy said they purchased a brand regardless of price. Far more important are factors such as comfort, absorbency and reliability,

- the retailers contacted almost all stated that most women did not change brands because of price. More precisely, the retailers contacted felt that a price difference of at least 10-20 % would be necessary for a woman to consider changing brand and even then it was unlikely.

Innovation

(132) The market is already well developed with a number of well-known and well-established brands and a wide range of product variants. Given the high method loyalty and significant brand loyalty in sanitary towels as set out above, market entry appears to be extremely difficult without a product which is perceived by the consumer as genuinely innovative in a number of ways. Only marginal changes or slight superiority of a new brand may not be sufficient to overcome the attachment consumers have developed to established brands. P& G's Always presented a substantially innovative package of improvements which was able to penetrate the market rapidly. It is difficult to forecast new product development of similar importance by a competitor in the near future. In particular, new entrants who are not already manufacturers of towels in other geographic markets would have difficulty developing a leading-edge product and so would be handicapped in any entry attempt.

(133) Furthermore, there appears to be a considerable 'first mover advantage' in that the first company to launch a new generation of products can establish a market position which is extremely difficult to attack, even with an improved product. The first company with something new to offer may be able to persuade women to switch to its new product because it offers improved performance, but it will be more difficult for companies introducing a similar product subsequently to gain custom. Those women who were originally attracted to the new product will already have switched to the first company's product, while those who have not switched are perhaps less likely in any case to do so. In Denmark, the Netherlands and Sweden, where Moelnlycke was the first to launch an ultra product, Procter & Gamble, even with a superior product, has found it difficult to repeat the gains in market share it has made elsewhere.

Difficulty of low-price entry strategy

(134) Competition in this market is largely conducted on advertising, performance and image. Since consumers are, as was explained above, relatively price-insensitive, a low-price low-promotional spend entry strategy based solely on undercutting the incumbent is inherently less likely to succeed. To the extent that advertising creates demand and determines sales, it is an important factor in the willingness of distributors to stock a new product. A low-price strategy, however, would not generate the returns necessary to fund such advertising, or to pay high listing fees to distributors to encourage them to stock a poorly-promoted brand.

Access to retail

(135) With the increasing concentration of the retail industry, especially in Northern Europe, entry for new brands is becoming more difficult. Instead of having ten or more important distributors from which to gain a presence on the supermarket shelf, suppliers have to gain distribution from only five or six key accounts if they are to achieve a viable position. Furthermore, there are some indications of a sporadic trend for distributors to rationalize the number of brands they stock in order to concentrate their negotiating power and gain larger volume discounts. While retail concentration in southern Europe is considerably lower currently, such a trend is beginning there as well.

Advertising and promotion

(136) P& G's launch of Always in Europe was accompanied by what other industry participants have viewed as an unprecedented promotion of the new product, including both consumer sampling and media promotion.

(137) Advertising is a key factor in generating demand for branded consumer goods. It is a means for manufacturers to attempt to reduce the demand elasticity, or price sensitivity of consumers, for their product. If successful, it gives rise to a 'virtuous circle' from a supplier's point of view, since the enhanced turnover generated allows the supplier to fund yet more sales-generating advertising. Since the reverse is also true, a 'vicious circle' of low sales reducing the ability of a supplier to fund advertising, it strengthens the market position of the strong and weakens that of the weak. Competition in the branded towel market is thus based as much on advertising as on price. As in more price sensitive markets, the ability to be competitive depends on operational efficiency and economies of scale which help keep the cost base low, although this is in order not to reduce prices but to leave as much cash-flow free for promotion as possible. A distinction should also be made between advertising and promotion expenditure aimed at encouraging women to test a new entrant's product and on-going brand support.

(138) In terms of market entry, it is clear that such entry must be sufficiently decisive and successful in order to break into the virtuous, self-reinforcing levels of market share and advertising expenditure, and that this must be achieved quickly. If it is not, then the sunk costs of entry are further and dramatically increased. Kimberly-Clark's attempted entry into the German market would appear illustrative: after the initial failure to win sufficient distribution and market share, Kimberly-Clark appears to have stopped funding promotion of its brand which resulted in sales rapidly dwindling to a negligible level. A new entrant needs not only to spend on advertising in order to encourage women to try out its new product, but it must spend sufficiently to stand out above the background of on-going brand support advertising.

(139) P& G also seemed aware of these factors and in their plans for the launch of Always in Belgium spoke of achieving '. . . . . . high brand awareness during the introduction. . . . . . [by spending] heavily in the media for the first [...] (98) at a level of [...] (99) the total yearly pad segment's media spending', that is at a level of [...] (100) times the expenditure of all other suppliers put together for those [...]. Furthermore, [...] key sizes of the new towel were to be sampled in [...] of households.

(140) P& G spent the following sums on advertising, sampling, trade promotion and listing fees for retailers when launching Always ultra thin in individual EEA countries:

<emplacement tableau>

(141) The Commission accepts that these figures represent the launch of an innovative product in national markets where P& G was not present (although it was already present in Europe through its Iberian and Italian joint ventures). However, the size of these expenditures is markedly higher than those of competitors on launches of new ultra-thin products. For example, while Procter & Gamble spent DM [...] million in launching ultra towels in Germany, Johnson & Johnson spent less than [...] of this amount. Moelnlycke's launch expenditure in France and the United Kingdom was less than [...] that of Procter & Gamble.

(142) P& G has also continued to spend markedly more than its competitors on all types of promotion in the years following the launch of Always. Its competitors have responded by increasing their own on-going brand support in order to defend their market share, as is shown in the following table:

(143)

<emplacement tableau>

(144) Procter & Gamble quote figures from DMB& B showing that, between 1989/90 and 1992/93, advertising as a percentage of total feminine hygiene sales in Germany rose from [...] (101) to [...] (102). This compares with [...] (103) to [...] (104) for diapers and [...] (105) to [...] (106) for detergents over the same period. Whilst the overall ratio of market sales to advertising expenditure may not therefore be out of line with some other consumer goods, the barrier to entry caused by the high levels of advertising is now significantly higher than it was at the time of P& G's entry.

(145) In terms of barriers to entry, the marked increase in the level of on-going promotional support has significantly raised the level of expenditure necessary for a new entrant wishing to stand out above the 'background noise' of this on-going brand support, so that the absolute amount of money required to overcome the promotional barrier to entry has been significantly increased. Furthermore, the effect of this increase in the underlying level of advertising expenditure is also to increase the minimum viable market share on entry since the potential entrant knows that in order to fund competitive levels of advertising, it must achieve sufficient sales. Incumbents such as P& G, with large market shares, can spread their advertising costs over much larger sales volumes. Given that a campaign in a particular magazine or media channel has a fixed cost irrespective of the turnover of the product being advertised, there are important economies of scale in advertising. Retaliation by incumbents with significant market shares against new entrants is thus cheap and easy, and until the new entrant achieves a similar market share, incumbents have a cost advantage on one of the key competitive parameters.

(146) This increase in the market level of advertising thus increases risk not only for potential entrants but also for competitors, particularly those who may not have the resources or, as importantly, the will to spend such large sums.

(147) Advertising and promotion are sunk costs which are not recoverable in the event of failure in the market, so that the two effects described above will dramatically increase the financial risk of market entry, even if they are not in and of themselves insuperable barriers and even if there are potential competitors with the financial resources necessary: entry must be not only possible but a sound business proposition.

High level of concentration

(148) Prior to this transaction, the top three companies in Germany held over 80 % of the market by value while in Spain, P& G's two joint ventures alone held 80 %. Post-merger, the two largest companies in Germany, P& G and J& J, will hold [...] (107) by value between them of the market (108), while in Spain, P& G's two joint ventures will hold [...] (109) more of the market as a result of the merger. These high levels of concentration constitute in themselves a barrier to entry for newcomers. The higher the share of the market leader, the harder it will be for a retailer not to stock the leader's brand and prefer a newcomer's product. It furthermore facilitates reactions against newcomers by incumbents who can defend shelf-space through special rebates for retailers or additional in-store promotion and media-spending. Furthermore, the larger the market share of the incumbents, the larger their advantage in terms of economies of scale in advertising and other areas. Overheads such as sales and marketing expense, as well as advertising itself, can all be spread over larger sales to release more cash flow for advertising and other retaliatory action.

(149) The barriers to entry enumerated above are not necessarily insuperable, particularly when taken individually, but taken together they are a very strong disincentive for any company contemplating entry into this market. P& G was able to overcome these barriers with the combination of an innovative product and massive sampling and publicity spend but, precisely as a consequence of this entry, some of the barriers, most notably advertising, access to retail and the high level of concentration, have increased significantly since the entry of P& G.

(150) The reality of these barriers can be seen in the history of entry in the German market. Kimberly-Clark launched the Freedom brand of towels in Germany in 1988/89, but the brand never achieved higher than a 2 % share and in 1993 had only 0,9 %. In 1979 Moelnlycke entered the German market with an innovative body-shaped product, Libresse Formé, via a distribution agreement with Henkel Kosmetik, part of Henkel AG. Notwithstanding the distribution agreement with Henkel, Moelnlycke was not able to achieve more than 35 to 40 % weighted distribution and finally had to withdraw with a loss. In the early 1980s Unilever tried to launch a towel under the brand Cosmea. This product was innovative (ultra thin and individually wrapped) but still did not succeed, and Unilever sold the brand to Pelz, which holds 3 % of the market with it. Between 1970 and 1985, there have been several unsuccessful attempts to establish the brand Vania in Germany by Kaysersberg.

Dominance and contestability

(151) As has been shown above, P& G will hold high market shares post-operation in Germany, Spain, Austria and Italy on a towel market characterized by high barriers to entry. In establishing whether or not P& G will be dominant on these markets, the question remains as to whether P& G's position will be constrained by actual or potential competitors or by retailers.

(152) In the case of Austria the investigation of the Commission has led it to conclude that, in view of the structure of the Austrian sanitary towel market both before and after P& G's acquisition of VPS, P& G is unlikely to hold a dominant position as a result of this transaction. P& G launched Always in Austria in the summer of 1992. It gained considerable market share from Johnson & Johnson and Rauscher, which together with Camelia would afford P& G a [...] (110) share of the Austrian market. P& G currently faces two competitors, however, with significant market shares: Johnson & Johnson, the previous market leader in Austria and a well-resourced multinational, with [...] (111) and Rauscher with [...] (112) through its two established national brands (Senta, Cresta). The structure of the Austrian sanitary towel market is thus more balanced than in Germany. Whilst P& G's position will be strengthened as a result of the concentration, the Commission considers that the existence of the two other suppliers with significant shares in this market will constrain P& G's competitive behaviour. Whilst it is possible that P& G will further increase its market share through organic growth in the future, this is not certain and in any case the market situation would not have been caused by this concentration but by organic growth.

(153) In the case of Italy, P& G will increase its [...] (113) share by only [...] (114) (by value). While the Commission considers that market share is prima facie evidence of dominance, it does not believe that a [...] (115) increase will constitute a strengthening of a dominant position.

(154) Given P& G's strength in Germany and the implications of its purchase of VPS for the Spanish market, it must be assessed whether P& G will, as a result of the increased share of the German and Spanish towel markets which it will obtain by acquiring VPS, be in a position to act independently of its competitors and customers in these markets on a lasting basis.

P& G's market position

(155) The speed and success of Procter & Gamble's entry into Europe has been remarkable. It has wrested market share not only from traditional national brands such as Camelia, but also from its global competitors, Johnson & Johnson, Kimberly-Clark and Moelnlycke. P& G's competitors have developed their own innovative products and, in response to Always' challenge, have dramatically increased their promotional expenditure in defence of their market shares (see above). However, the combination of the Always product and Procter & Gamble's other competitive advantages has resulted in extremely fast organic growth.

(156) Sanitary towels are products where suppliers enjoy a relatively high level of brand loyalty and reduced price sensitivity. Women must purchase them on a regular basis but they constitute only a small percentage of a woman's total monthly expenditure. As a product, therefore, the sanitary towel sector is one in which price increases beyond competitive levels would not necessarily result in reduced sales.

(157) P& G is particularly strong in the ultra-thin segment of the towel market. Its product is a recent innovation and its share of the total market is growing rapidly. P& G's strength in the fastest growth area of the market is an additional factor strengthening its market power.

(158) P& G is one of the largest suppliers of branded grocery goods to the retail trade. Many of its brands are 'must-haves' for the retailer, such as Always for most of the retailers consulted by the Commission. So P& G is in a much stronger position in commercial relations with grocery distributors than many of its competitors. By supplying [...] (116) of the German retailers' feminine hygiene sales, P& G is in a strong position when it comes to obtaining access to all-important shelf-space in major supermarket chains for its new products.

(159) P& G has considerable expertise in developing and marketing non-food consumer brands and understands well that the very heavy up-front investment required to create a brand can nonetheless be extremely profitable when evaluated over a relatively long time period.

(160) P& G's total turnover is over twice that of Johnson & Johnson, its next largest competitor in the sanitary towel business. More importantly, the sanitary towel turnover of Johnson & Johnson and Moelnlycke is around one third of that of P& G in Europe. This is important in the context of the advertising economies of scale explained above. For the same advertising/sales ratio as its competitors, Procter & Gamble can fund up to three times as much promotional spend in absolute terms. P& G's financial position is further strengthened by its positions on other markets such as washing powders and detergents, since even if resources are not transferred between different business segments, the ability of each to incur risk is determined by the strength of the whole.

Competitive pressure from tampons

(161) P& G has argued that, even if there were two separate product markets for tampons and towels, tampons would exercise a competitive pressure on the towel market which should be taken into account in the assessment of the merger. The Commission does not consider such a competitive pressure from tampons as significant in respect of price competition.

(162) The first aspect of this is the degree to which women change their usage for non-price reasons, which can be several. A change in contraceptive methods can necessitate a change in the method of menstrual protection since the IUD can cause heavy bleeding, which is best dealt with by a towel, while the contraceptive pill reduces bleeding, facilitating tampon use. Women cannot use tampons immediately after childbirth and are obliged to use towels for a certain period of time, while other women approaching menopause may also only be able to use towels. In such cases there is clearly little competition between methods either on price or performance (given that one method has a clear functional advantage over the other). Other women, however, may change their method of protection, again not for price reasons, but because of the performance characteristics of the two methods. Particularly important here are adolescent women who wish to experiment with both methods to find which suits them best. Similarly, it is likely that there are some women who change method later in life because of the non-price characteristics of the two products. This might be because of perceptions of towels as, for example, the more 'natural' method, or of tampons as more 'modern'.

(163) There is thus a range of women who do change their usage habits for a variety of reasons other than price. These changes would be made regardless of the relative prices of the products and are essentially a given in the market which would occur whether P& G raised its towel prices or not. The level of usage change these women account for is difficult to estimate but one competitor estimated that, over the nine-month period taken in the A.C. Nielsen household purchase study, about 4 % will have had babies while 3 % will have become menopausal, which may require a change to towels. Changes in contraceptive method and adolescent experimentation would also need to be taken into account. Given the size of the switching found by the A.C. Nielsen household panel study (6 to 8 % products purchased or 13 to 18 % of women involved), the fact that part of this 'switching' is more apparent than real due to long purchase cycles and stocking, and the number of women switching for non-price reasons, the amount of price-sensitive switching is clearly very small.

(164) Nevertheless, even if price-sensitive switching were not minimal, given the low levels of cross-price elasticity found in the studies described earlier, it is clear that it would still be profitable for P& G to raise prices since the revenue generated by the consumers which they risk losing would be less than the increased revenue they would gain from all other consumers.

(165) Tampons not only form a separate market from towels, but also do not exercise any noticeable competitive pressure on them since most movement from one to the other is on the basis of non-price factors.

Purchasing power by food retailers

(166) P& G has argued that retailers exert strong competitive pressure in the sector of feminine hygiene products owing to the significant purchasing power exercised by large retailers and transnational buying groups.

(167) As was set out above, transnational buying groups do not currently and will not for the foreseeable future play a significant role in the field of sanitary towels. As to the structure of the retail sector in Germany, it is true that it is highly concentrated. However, the existence of purchasing power which would be able to constrain the competitive behaviour of P& G can currently not be assumed for several reasons. Compared to other consumer food products, which are exclusively sold in grocery outlets, the demand side is less concentrated in sanitary towels because they are also sold through pharmacies and drugstores. Less than [...] (117) of P& G's sales (VPS: [...] (118)) go to the three biggest retailers in Germany, which in turn account for less than 40 % of total retail sales.

(168) An expert report ('Sondergutachten') published on 18 February 1994 under the German Act against Restrictions of Competition by the Monopolkommission, a group of independent competition experts appointed by the President of the Federal Republic of Germany, on 'Market structure and competition in retail' stated that the purchasing markets of retail in Germany are not characterized by purchasing power since suppliers of grocery goods dispose of alternatives for selling their goods and the demand side is marked by competition. The high intensity of competition between German retailers with respect to consumers has a further competitive impact on competition on the demand side. The retail sector in Spain is far more fragmented than in Germany and so has much less power than the German trade.

(169) Even if it were assumed that retailers were in a strong position vis-à-vis suppliers of grocery goods, this would have to be considered differently in the field of sanitary towels because of the significant loyalty and brand method loyalty of consumers. Whilst retailers might try to limit P& G's negotiating power by reducing the number of put-ups stocked per manufacturer, they are caught between the supplier and the consumer demand generated by that supplier's advertising. This advertising creates consumer pressure on a distributor to stock a brand which is advertised in order to meet the demands of its customers. Given the advertising capacity of P& G, the leading brands Always and Camelia are therefore almost indispensable for retailers, as confirmed by most of the retailers contacted by the Commission.

(170) Even if it were accepted that retailers did have some countervailing power to use against P& G, it is also important to examine what the motivation of the stores would be to do so. P& G has, in launching Always, argued in promotional literature for distributors that Always was a value-added product and that part of the increased sale price it could command would be shared with distributors which stocked it. Higher prices might thus simply mean higher margins for distributors. Furthermore, given the relative price insensitivity of consumers in this sector and their brand loyalty, and the effect of advertising, which is to reduce the price elasticity of the advertised good, the distributor will be able to pass on the increase to the customer without losing volume. There is thus little motivation for a retailer to resist a price increase by P& G.

(171) It could be argued that post-concentration, retailers will wish to diversify their supplier base since P& G will supply the top two brands in Germany. Retailers are, however, generally reluctant to delist premium brands like Always and Camelia, which present a high level of product support, high margins and, thus, significant turnover, in favour of an untried new product.

Current competitors

(172) P& G's closest competitor in Germany is Johnson & Johnson (J& J) with a market share of [...] (119). Its product line is not markedly inferior to that of P& G (it launched its ultra-thin variant in August 1992) but its market share has been halved from [...] (120) in the two-and-a-half years since the launch of Always. Between 1990 and 1991 J& J increased its advertising support for its brand Silhouette by 90 % to respond to the launch of Always in July 1991 but appears to have had only limited success gainst P& G. In Spain, J& J, despite being the only other multinational present, has only a share of [...] (121) which does not pose a significant competitive threat to P& G's joint ventures.

(173) There are two other small competitors in Germany, Pelz and Hedwigsthal. Both of these are, in competition terms, marginal with market shares of under 5 % and with only secondary brands. They do not have the resources or products to pose any serious threat to P& G's market position or constrain its behaviour.

(174) As mentioned earlier, own-label products constitute [...] (122) by value and [...] (123) by volume of the German towel market and sell at around half the price of premium branded products. As was also explained earlier, the competition offered by own-label products in Germany and Spain is extremely limited given the price and quality gap. What own-label products do provide is an alternative for the consumer in the face of extreme increases in the prices of branded goods. There is a point at which a consumer, even in this market, would conclude that the improved performance of the branded product over the own label did not merit the increased price. Given the relative price insensitivity of this market, however, it is not likely that own-label products would constrain price increases in branded products unless they were highly exaggerated.

Potential competition

(175) For market entry or the potential for such entry to constrain P& G's behaviour in the German and Spanish markets, it would be necessary to examine whether such entry is probable, whether it would be competitively meaningful and effective and whether it could take place within a time frame short enough to deter P& G from exploiting its market power.

(176) The first group of possible entrants are those feminine hygiene suppliers who are present in other States but not in a particular country or group of countries. In the case of Germany and Spain, Moelnlycke, Kaysersberg and even Kimberly-Clark, to the extent that its share of the German market is minimal and declining, could be considered as potential entrants. Outside the existing players in Europe there are two groups of companies which might be considered potential entrants: sanitary protection companies such as the Japanese firms Unicharm and Kao and own-label specialists such as Confab (US) or Disposable Soft Goods (Hong Kong) and also other branded consumer goods companies.

(177) Kao and Unicharm are the only major producers of feminine hygiene products who are not present on the EEA market at present (although Unicharm is in a joint venture with Moelnlycke on the baby nappy market). However, they lack marketing experience and distribution organizations in Europe. One Japanese company stated that it is 'almost impossible' for them to enter an EEA towel market given the position of P& G in Europe and also that, with only absorbent products to sell, it would be too expensive to establish its own distribution network. Manufacturers of consumer goods such as Unilever, Nestlé or Philip Morris would have the financial resources but no experience in the feminine hygiene sector or related markets and would also have to create a product and brand de novo.

(178) While there are thus several conceivable potential entrants, the question is whether any of them would be likely to enter the German or Spanish market within the next two to three years or as a timely response to excessive pricing in the market. As was discussed earlier with regard to Germany, there have been several failed attempts at entry in the last 10 to 15 years while illustrate the difficulty of penetrating this market on any scale less than that undertaken by P& G with Always.

(179) As explained earlier in the discussion of the barriers to entry in the towel market, successful entry is all or nothing and must combine a product which is perceived as innovative by the consumer with a huge advertising and promotion effort. This last point is necessary to build market share and distribution quickly. Competitors of P& G have estimated that the minimum viable market share is between 15 and 20 % if a supplier is to be able to generate the resources to fund the advertising necessary to obtain adequate weighted retail distribution (estimated by competitors for Germany to be at about 70 %) and retain market share once won. A minimum market share is also necessary to begin to be able to reap economies of scale, particularly in advertising. This requirement of a minimum viable market share limits further the potential for entry. A low-cost entry strategy based on undercutting the existing brands rather than out-promoting them would be unlikely to succeed, given not only the relatively low level of price sensitivity among consumers but also the need to fund the advertising necessary to persuade brand loyal consumers to switch brands and thus obtain retail distribution and any sales at all.

(180) Even if a company were to decide to enter the market, such a decision would be a long-term one and could not provide an immediate constraint on excessive pricing. Sanitary towels are not commodity products and new entry would entail the development, in some cases, of a new brand and probably an innovative product which would then have to be market tested, sampled, and heavily promoted before being launched. Furthermore, the sort of investment required to enter the German and Spanish market, particularly with P& G as an incumbent, and the associated investment risk, means that entry, it if comes, will not be 'hit-and-run' entry taking opportunistic advantage of short-term price levels but will be a long-term strategic decision by the entrant that the German and Spanish markets offer a profitable business opportunity despite the barriers and difficulties and that attacking P& G's dominant position in this market through such entry is a corporate priority. It is therefore unlikely that P& G will feel constrained by the threat of such entry, aware as it is both that the costs and risks of entry are too high for all but the most committed of entrants, and that the level of commitment necessary for entry into the market is such that any decision to enter will not be taken quickly, nor a product introduced to the market precipitously.

(181) While P& G's increment on the Spanish market is admittedly not large at [...] (124) by value, it is significant given both P& G's very high existing market share ([...] (125) by value) and the fact that, by this transaction, P& G denies any potential entrant the possibility of entering by buying the Camelia brand in Spain. Apart from Camelia, there is on the premium brand level only Johnson & Johnson with marginal market shares (Silhouette [...] (126), Serena and Vespré [...] (127)). Other manufacturers' brands together hold [...] (128) and private label products [...] (129). Such small market positions will not enable these competitors to constrain the two brands of P& G's joint ventures, which have [...] (130) and [...] (131) respectively. Given the significant brand loyalty in this sector and the well-established position of Ausonia and Evax, the acquisition of one of the remaining premium brands reinforces the dominant position of P& G's joint ventures on the Spanish market.

(182) Such dominance is, in the view of the Commission, likely to cause harm to the consumer in several ways, in particular on the German market:

- price - P& G will be able to increase prices for both Camelia and Always independently of the competitive reaction of other suppliers. Its freedom to do so will spring both from its market dominance in Germany and from the relatively low price sensitivity of consumers in this market. P& G will be particularly free to do so in the ultra-thin segment where the only competitor to the Camelia and Always versions are Johnson & Johnson's Silhouettes brand. Furthermore, it is probable that Johnson & Johnson's strategy as the closest competitor would be to follow any price increases introduced by P& G in order to rebuild profitability in its loss-making towel business in Germany,

- innovation and quality - the acquisition will eliminate competition on innovation between VPS and P& G. Both Camelia and Always will use globally developed P& G technology and any innovations which might have been developed by VPS, or by VPS under different ownership, will be lost to the consumer. Once again the only serious competitor on innovation will be Johnson & Johnson,

- choice - P& G's acquisition is unlikely to result in any immediate reduction in consumer choice. It is highly probable, however, that P& G will subsequently rationalize the broad range of product variants of Camelia and Always in order to maximize the efficiency of its towel business in Germany, thus reducing consumer choice.

Conclusion

(183) In view of the above results of the investigation and particularly the discussion of market shares, barriers to entry and potential competition, the Commission considers that there are a number of factors relevant to the German and Spanish markets for sanitary towels whereby the acquisition by P& G of VPS, even after the divestment of VPS's baby nappy business, will enable P& G post-merger to act independently of its customers and competitors in these markets.

(184) In Germany, P& G will hold a post-merger market share of [...] (132) by value and [...] (133) by volume on a market where its closest competitor will have only [...] (134) by value and [...] (135) by volume, the balance being fragmented between other suppliers and own-label sales. These figures take into account P& G's initial offer to divest the non-Camelia feminine hygiene business of VPS (this offer has subsequently been replaced by one to divest Camelia - see paragraph 8 of this Decision). Given these figures, the barriers to entry to the German market and the history of attempts at market entry, the Commission considers that P& G's market power will not be constrained by either actual or potential competitors. Indeed, the very acquisition of VPS and its important German brand, Camelia, which is the last major national independent brand, would render entry into the German market for other entrants more difficult by obliging them to enter de novo rather than through the acquisition of an existing player.

(185) As was already set out above, the increment of another [...] (136) to the market position of P& G on the Spanish market would strengthen P& G's dominant position on a market which is isolated by the high level of concentration. Not only would there be an increment in an already dominant position but the Camelia brand, which might constitute an entry vehicle for a new entrant, would be lost.

VI. COMMITMENTS PROPOSED BY PROCTER & GAMBLE

(186) P& G has offered to modify the original concentration plan as notified by entering into the following commitments:

'P& G hereby gives the following undertakings to the Commission with respect to VP's Camelia-branded feminine hygiene products business, which comprises: (i) the Forchheim plant and the production lines dedicated to the manufacture of feminine hygiene products; (ii) the Camelia brand name; (iii) all other assets and liabilities that form part of or are necessary for the operation of VP's Camelia-branded feminine hygiene products business (hereafter referred to as the "Business").

1. P& G undertakes that, as soon as practicable after the Commission has adopted a favourable decision under Regulation (EEC) No 4064-89 and in any event no later than at closing of its acquisition of the shares of VP, it shall appoint an independent trustee to be approved by the Commission, to act on its behalf in overseeing the ongoing management of the Camelia Business to ensure its continued viability and market value and its rapid and effective divestiture from the rest of P& G's activities (hereafter referred to as "the Trustee"). The Trustee shall simultaneously appoint Goldman Sachs International Ltd ("Goldman Sachs") to act on its behalf in conducting good faith negotiations with interested third parties with a view to selling the Business. P& G shall agree in turn with the Trustee and Goldman Sachs on their respective remuneration, it being understood that part of the remuneration of Goldman Sachs shall consist of a fee related to the consideration of the sale.

2. P& G undertakes that it shall give the Trustee an irrevocable mandate to find a valid purchaser for the Business within [...], it being understood that such purchaser shall be a viable existing or prospective competitor independent of and unconnected to P& G and, possessing the financial resources and proven expertise in consumer product markets, enabling it to maintain and develop the Business as an active competitive force in competition to P& G's catamenials business on the various markets concerned. [...] P& G shall take all reasonable steps to encourage the relevant personnel currently employed in the Business, including sales and administrative personnel, to take up employment with such independent third party. P& G shall be deemed to have complied with this undertaking if, within this [...] period, it has entered into a binding letter of intent for the sale of the Business, provided that such sale is completed within a time limit agreed to by the Commission. P& G undertakes to give, on an arm's length basis, all assistance requested by the Trustee and Goldman Sachs prior to the sale of the Business to a third party.

3. In the reports referred to in paragraph 10 below, the Trustee shall indicate to the Commission whether it believes that a purchaser with which it is proposed to sign a letter of intent fulfils the description of a valid purchaser set out in paragraph 2 above, and it considers that negotiations with such a purchaser should continue. If within one week of receipt of the relevant advice from the Trustee, the Commission does not formally indicate its disagreement with the Trustee's assessment of a purchaser, negotiations with such party as a valid purchaser shall be free to proceed.

4. Providing the offers concerned have been received from purchasers recognized as valid according to the procedure laid down in paragraph 3, P& G alone shall be free to accept any offer or to select the offer it considers best in case of a plurality of offers. The value of any such offers shall be determined by the price offered plus other obligations affecting the value of such offers.

5. Where a binding agreement for the sale of the Business has been signed, the purchaser shall be associated forthwith to any ongoing contractual negotiations for supply of catamenial products with Germany retail distributors in order to ensure that the viability of the Business is preserved. Until such a binding sales agreement exists, the Trustee shall be associated with these negotiations.

6. P& G undertakes that, within the [...] period referred to in point 2 above and in any event before the completion of the sale of the business to a third party, the Forchheim plant shall be rendered capable of being transferred to an independent third party and, most particularly, that the Forchheim plant is capable of being managed separately from P& G.

7. Prior to the completion of the sale of the Business to a third party, P& G shall ensure that the Business is managed as a distinct and saleable entity with its own management accounts and a sales and distribution effort for the Business that is separate from P& G's catamenials business and with its own research and development facilities as presently exist under VP's management. P& G further undertakes that the Business shall have its own management composed of ex-VP or other currently non-P& G personnel that shall, under the guidance and control of the Trustee be under instructions to manage it on an independent basis in order to ensure its continued viability, market value and independence from P& G. On the request of the Trustee, P& G shall provide sufficient financial resources to this end in the ordinary course of business. Prior to the completion of the sale of the Business to a third party, P& G shall not integrate the Business into any P& G business unit, nor shall it appoint or second any P& G employee to the Business. P& G also undertakes that it shall make no structural changes to the Business without prior Commission approval.

8. P& G shall not integrate VP's secondary and own-label catamenial business into its own commercial and production structures for catamenials until the sale of the Camelia Business is completed.

9. P& G shall not obtain from the Business management any business secrets, know-how, commercial information, or any other industrial information or property rights of a confidential or proprietary nature relating to the Business.

10. P& G undertakes that it shall cause the Trustee to provide a written report either before a letter of intent is to be signed and in any event on a bi-monthly basis on relevant developments in its negotiations with third parties interested in purchasing the Business, and that such reports, together with supporting documentation, shall be furnished to the Commission. Such supporting documentation shall include a report by the management of the Business on its on-going commercial operations.

11. Without prejudice to the powers of the Commission to ensure enforcement of the above undertakings as conditions and obligations under Article 8 (2) of Regulation (EEC) No 4064-89, any dispute between P& G and the third party purchasing the Business arising out of or in connection with the implementation of these undertakings shall be submitted to independent arbitration to be mutually agreed between P& G and such third party, it being understood that the time needed for such arbitration shall not affect the deadline laid down for the completion of the sale of the Business set out in paragraph 2.'

(187) The Commission is satisfied that P& G's offer to divest a business including the Camelia towel brand will prevent P& G from acquiring a dominant position in Germany and from reinforcing its dominant position in Spain. Post-concentration and post-divestment of Camelia the market structure in Germany and Spain will be as follows, taking into account that P& G will not now divest the non-Camelia business of VPS (137):

<emplacement tableau>

As can be seen, P& G will increase its share of the German market by [...] to a total share of [...] by value with Camelia holding a [...] and J& J a [...] share. The increase in P& G's market share will be solely attributable to its acquisition of VPS's secondary and store brand business (i.e. non-premium brands) while P& G's existing Always business will be subject to competition from two significant suppliers of branded premium towels. In Spain, P& G's share will increase by less than [...]. The Commission has therefore concluded that the commitments offered by P& G in respect of the Camelia-branded feminine hygiene business of VPS are sufficient to prevent the creation or reinforcement of a dominant position on the German and Spanish markets, or indeed elsewhere in the EEA.

(188) If the sale of assets to be divested has not taken place by the end of the time period set out in P& G's commitment, the Commission reserves the right to require that P& G divest all assets and interests of VP Schickedanz, and thereby that P& G and VP Schickedanz be fully separated in order to restore conditions of effective competition, as provided by Article 8 (4) of the Merger Regulation. Furthermore, if any of the obligations accepted by P& G are breached, the Commission reserves the right pursuant to Article 8 (5) to revoke its authorisation decision.

(189) These actions will be taken without prejudice to the Commission's right to impose fines pursuant to Article 14 (2),

HAS ADOPTED THIS DECISION:

Article 1

Subject to the full compliance will all conditions and obligations contained in Procter & Gamble's commitment vis-à-vis the Commission in respect of the Camelia-branded feminine hygiene business of VPS, as set out in recital 186 of this Decision, the concentration notified by Procter & Gamble GmbH on 17 January 1994 relating to the acquisition of VP Schickedanz AG is declared compatible with the common market and the functioning of the EEA Agreement.

Article 2

This Decision is addressed to:

Procter & Gamble GmbH, Procter & Gamble European Technical Center, Temselaan 100,

B-1820, Strombeek-Bever.

(1) OJ No L 395, 30. 12. 1989, p. 1. (Corrigendum: OJ No L 257, 21. 9. 1990, p. 13).

(2) [...]: In the published version of this Decision, some information has hereinafter been omitted, pursuant to Article 17 (2) of Regulation (EEC) No 4064-89 concerning non-disclosure of business secrets.

(3) Between 40 and 45 %.

(4) Between ECU 250 and 300 million.

(5) Between 1 and 5 %.

(6) Between 15 and 20 %.

(7) Less than 1 %.

(8) Between 15 and 20 %.

(9) Between 35 and 40 %.

(10) Between 15 and 20 %.

(11) Between 15 and 20 %.

(12) Between 35 and 40 %.

(13) Between 10 and 15 %.

(14) Between 15 and 20 %.

(15) Between 15 and 20 %.

(16) Between 10 and 15 %.

(17) Between 30 and 35 %.

(18) Between 30 and 35 %.

(19) Between 20 and 25 %.

(20) Between 35 and 40 %.

(21) Between 10 and 15 %.

(22) Between 5 and 10 %.

(23) Between 15 and 20 %.

(24) Between 45 and 50 %.

(25) Between 1 and 5 %.

(26) OJ No L 356, 5. 12. 1992, p. 1.

(27) Between 10 and 15 %.

(28) Between 20 and 25 %.

(29) Between 10 and 15 %.

(30) Between 15 and 20 %.

(31) Between 1 and 2 %.

(32) Between 2 and 3 %.

(33) Between 2 and 3 %.

(34) Between 0 and 1 %.

(35) Towels other than Always.

(36) Between 10 and 20.

(37) Between 10 and 20.

(38) Prices for towels vary significantly relative to prices for tampons in the different Member States and do so independently of their relative penetration levels.

(39) Between 25 and 30 %.

(40) Between 60 and 65 %.

(41) Between 15 and 20 %.

(42) Between 30 and 35 %.

(43) Between 10 and 15 %.

(44) Between 30 and 35 %.

(45) Between 1 and 5 %.

(46) Between 5 and 10 %.

(47) Between 10 and 15 %.

(48) Between 5 and 10 %.

(49) Between 5 and 10 %.

(50) Brands with a market share in excess of 3%.

(51) Between 30 and 35 %.

(52) Between 15 and 20 %.

(53) The ratio of VPS's branded product prices to its distributor brand prices is between 2:1 and 2.5:1.

(54) Between 80 and 85 %.

(55) Between 60 and 65 %.

(56) Between 80 and 85 %.

(57) Between 65 and 70 %.

(58) Between 10 and 15 %.

(59) Between 35 and 40 %.

(60) Between 35 and 40 %.

(61) Between 30 and 35 %.

(62) Between 35 and 40 %.

(63) Between 5 and 10 %.

(64) Between 15 and 20 %.

(65) OJ No L 116, 12. 5. 1993, p. 21.

(66) Nielsen does not cover certain important retail outlets (it estimates its coverage factor in Germany to be 76 %). P& G have therefore assumed that relative market shares of suppliers of Nielsen stores are identical to those of the market as a whole, except in respect of Aldi, which sells only own-label goods. Furthermore, to the extent that competitors of P& G produce own-label products, these are included in 'others' in Nielsen data. J& J and P& G do not produce own-label fempro products.

(67) Exact market shares deleted as a business secret.

(68) Between 60 and 65 %.

(69) Between 40 and 45 %.

(70) Between 65 and 70 %.

(71) Between 50 and 55 %.

(72) Between 10 and 15 %.

(73) Between 5 and 10 %.

(74) Between 5 and 10 %.

(75) Less than 1 %.

(76) Between 15 and 20 %.

(77) Between 25 and 30 %.

(78) Between 35 and 40 %.

(79) Exact market shares deleted as business secret.

(80) Between 75 and 80 %.

(81) Between 65 and 70 %.

(82) Between 1 and 5 %.

(83) Between 1 and 5 %.

(84) Less than 1 %.

(85) Between 70 and 75 %.

(86) Between 65 and 70 %.

(87) Less than 1 %.

(88) Between 10 and 15 %.

(89) Between 20 and 25 %.

(90) Between 35 and 40 %.

(91) Between 30 and 35 %.

(92) Between 30 and 35 %.

(93) Between 20 and 25 %.

(94) Between 15 and 20 %.

(95) Between 25 and 30 %.

(96) Period.

(97) Multiple.

(98) Multiple.

(99) Between 1 and 5 %.

(100) Between 10 and 15 %.

(101) Between 5 and 10 %.

(102) Between 5 and 10 %.

(103) Between 5 and 10 %.

(104) Between 15 and 20 %.

(105) Between 70 and 75 %.

(106) Herfindahl-Hirschman-Index of concentration in Germany by value as calculated by the Commission: - pre-merger: 2 670, - post-merger: 4 110.

(107) Between 1 and 5 %.

(108) Between 35 and 40 %.

(109) Between 30 and 35 %.

(110) Between 15 and 20 %.

(111) Between 70 and 75 %.

(112) Less than 1 %.

(113) Less than 1 %.

(114) Between 15 and 35 %.

(115) Less than 50 %.

(116) Less than 50 %.

(117) Between 10 and 15 %.

(118) Between 25 and 30 %.

(119) Between 1 and 5 %.

(120) Between 10 and 15 %.

(121) Between 20 and 25 %.

(122) Between 1 and 5 %.

(123) Between 75 and 80 %.

(124) Less than 1 %.

(125) Less than 1 %.

(126) Between 5 and 10 %.

(127) Between 10 and 15 %.

(128) Between 45 and 50 %.

(129) Between 30 and 35 %.

(130) Between 60 and 65 %.

(131) Between 40 and 45 %.

(132) Between 10 and 15 %.

(133) Between 5 and 10 %.

(134) Between 1 and 5 %.

(135) Exact market shares deleted as business secret.