Naspers Ltd and Electronic Media Network Ltd & Another (23/LM/Feb07) [2008] ZACT 10 (24 January 2008)

[1] On 8 October 2007 the Tribunal unconditionally approved the transaction whereby Naspers Limited (“Naspers”) acquired sole control of Electronic Media Network Limited (“M-Net”) and SuperSport International Holdings Limited (“SuperSport”). The reasons for this decision follow.

[2] The primary acquiring firm is Naspers Limited (“Naspers”), a public company listed on the JSE Securities Exchange Limited. The parties submitted that Naspers Beleggings Limited and Keeromstraat 30 Beleggings Limited are the firms that directly or indirectly control Naspers.

[3] The primary target firms are Media Network Limited (“M-Net”) and SuperSport International Holdings Limited (“SuperSport”). M-Net was launched in 1986 as the first subscription television service in South Africa. M-Net currently has an array of general entertainment and niche channels and broadcasts to more than 1.3 million viewers in more than 40 countries across the African continent and its adjacent islands. SuperSport’s operations include the acquisition, packaging and scheduling of sports related franchising and merchandising, sports competitions, sports package tours and ownership of sport related assets such as professional sports teams. M-Net is jointly controlled by Naspers and Johnnic Communications Limited (“Johncom”). M-Net controls Oracle Airtime Sales (Pty) Ltd (“Oracle“). SuperSport is also jointly controlled by Naspers and Johncom.

[4] In terms of the proposed transaction Naspers will acquire sole control of M-Net and SuperSport. The majority direct shareholder of M-Net/SuperSport is a holding company called MNH Holdings (1998) (Pty) Ltd (“MNH”). This holding company’s shareholding in SuperSport is currently 52.6%. MNH is currently owned 50% by Naspers, 47.5% by Johncom and 2.5% by Natal Witness Investments (Pty) Ltd (“the Natal Witness”). The current direct shareholding in M-Net/SuperSport is as follows: MNH 52.66%; Naspers 23.33%; Johncom 13.56% and MultiChoice Africa (Pty) Ltd (“MultiChoice”) 10.45%. The direct and indirect interests in M-Net/SuperSport are accordingly as follows: Naspers 60.12%; Johncom 38.56 and Natal Witness 1.32%. Naspers, Johncom and the Natal Witness are parties to a shareholders’ agreement in respect of MNH (“the MNH Shareholders’ agreement”), which also regulates their relationship in respect of M-Net/SuperSport. Naspers exercises managerial control over M-Net/SuperSport, however both Naspers and Johncom, through their respective interests and shareholding arrangements, jointly control M-Net/SuperSport. The merger results in a change from joint to sole control. Therefore the proposed transaction entails the acquisition by Naspers of all Johncom’s direct and indirect interests in M-Net/SuperSport.

[6] The parties have submitted that the rationale for the transaction is the need to improve the empowerment shareholding in M-Net/SuperSport which decreased as a result of the unbundling of Johncom from Johnnic Holdings Limited, and a desire to simplify the corporate structure.

[7] During its investigations of the proposed transaction the Competition Commission (“the Commission”) received objections from print media companies alleging that Naspers would, post- merger, behave in a manner which is likely to substantially prevent or lessen competition in the market for advertising in print media. The print media companies were Caxton and CTP Publishers and Printers Ltd (“Caxton”) and Independent News and Media (South Africa) (Pty) Ltd (“Independent Group”).

[8] Caxton’s submissions to the Commission are contained in a letter received from its attorneys dated 30 March 2007. In this letter Caxton submitted that the merger would enable Naspers to leverage its dominant position in pay TV in support of its print media interests by engaging in the practices of cross subsidisation, foreclosure and bundling. This would ultimately retard the ability of Naspers’ competitors in the print media market to act as an effective competitive constraint and, so, portend a substantial lessening of competition in the print media market.

[11] On the 2nd July 2007, Caxton brought an application before this tribunal for leave to intervene and participate in the merger proceedings in terms of section 53(1) (c) (v) of the Competition Act 89 of 1998 (‘the Act’) read with rule 46 of the Competition Tribunal Rules. We were notified by the Independent Group that it did not intend to participate further in the merger proceedings. While the application to intervene was not opposed by the merging parties, they did contest the scope of intervention requested by Caxton. The intervention application was accordingly set down for hearing.

[12] Caxton is a publisher and printer of books, magazines, newspapers and commercial print in South Africa. It controls one regional daily newspaper and many regional and community newspapers. Caxton also publishes a number of magazines which are distributed nationally and it also has its own distribution division.

[13] The merging parties did not oppose the application brought by Caxton. However they argued that the applicant’s scope of intervention should be limited by permitting it to raise only issues relating to vertical foreclosure and mixed bundling. The merging parties argued that the applicant should not be permitted to raise cross-subsidisation as an additional competition issue.

[14] After hearing the intervention application the Tribunal granted Caxton intervention rights insofar as its allegations of bundling and foreclosure were concerned, but declined to permit it to include the allegation relating to cross subsidisation in the scope of its intervention.

[15] On 7 August 2007 Caxton launched an urgent review application before the Competition Appeal Court (“CAC”). Caxton argued that the Tribunal had committed a reviewable irregularity by placing limits on the scope and ambit of intervention. The review application was heard on 10 September 2007 and on 4 October 2007 the CAC dismissed Caxton’s application on the basis that Caxton failed to show that the Tribunal, in making the aforesaid order, acted ultra vires, exceeded its jurisdiction or that it improperly exercised its discretion.

[16] This transaction involves the acquisition by Naspers of sole control of M-Net and SuperSport. As elaborated above, while, pre-merger, the target firms, M-Net and SuperSport, were managed by Naspers, control of the companies was shared between Naspers and Johncom. The transaction effects a change from joint control by Naspers and Johncom to sole control by Naspers. At the risk of stating the obvious, in examining the impact of this merger on competition our task is to establish whether any likely change in competitive conditions, in particular a likely substantial lessening of competition, is attributable to the removal of Johncom from the joint control structure. That is, the predicted change in the competitive landscape must be merger specific; it must be shown to have emanated from the transaction.

[17] From the perspective of pay TV, or, indeed, TV generally, there are no horizontal competition issues at stake in this transaction. The acquiring firm, Naspers, is, as already stated, simply increasing its stake in the two target companies, M-Net and SuperSport, and, in the process, moving from a position of shared control to sole control. There are no additional pay TV assets or any other media assets that are being acquired. Indeed to the extent that there are any horizontal issues implicated in this merger at all, these would ordinarily be grounds for celebration in the ranks of those charged with promoting competition. This is because both Naspers and Johncom do separately own a powerful stable of competing assets in the magazines and newspaper segments of the broad media market. Moreover, Johncom holds a significant stake in Caxton, another powerful participant in printing and in the publication of newspapers and magazines. In other words this transaction effects the structural separation of Naspers from two of its most significant competitors in print media, namely Johncom and Caxton. Nor does the transaction give rise to any vertical integration. However, far from celebrate, we are asked to prohibit the transaction.

[18] The reason why this apparently counter-intuitive response is urged upon us arises precisely from the potential combination of, on the one hand, Naspers’ print media assets and, on the other, the unrestrained control of the pay TV assets, M-Net and SuperSport, that it will acquire in consequence of the transaction. It is argued that whereas in the pre-merger joint control structure Johncom’s presence will have restrained Naspers from deploying the TV assets to support its print media assets, post-merger Naspers will be at large to deploy its powerful and newly unrestrained position in pay TV to bolster its already powerful competitive position in the print media market and, particularly, in the magazine market.

[19] Why would Johncom have restrained Naspers in this way? The only inference that can be drawn from this apprehension, is that an unrestrained Naspers will be prepared, temporarily at least, to compromise profit maximizing behaviour in its management of its pay TV assets (in which Johncom has a substantial economic interest) in order to bolster its print division (in which Johncom has no interest whatsoever except qua competitor). An unrestrained Naspers would, the argument must logically assume, do this, would be prepared to sacrifice profit maximization in pay TV, in the interests of the greater Naspers group. On the other hand, Johncom, as a rational profit maximiser and with no interest in the fortunes of the Naspers group beyond their shared interest in M-Net and SuperSport, would not support the deployment of its TV assets in this way and so, it is argued, its removal from the control structure in consequence of the transaction enhances the strategic possibilities available to an unfettered Naspers. These new strategic possibilities, it is further contended, are likely to be deployed to undermine the competitive structure of the magazine market and would ultimately give rise to anti-competitive outcomes in that market.

[20] The Commission’s report reveals that it has been alive to these concerns. However, as noted above, it does not envisage that they give rise to the likelihood of a substantial lessening of competition. It concludes that post-merger Naspers has neither the incentive nor the ability to utilize its pay TV assets in a manner that will compromise the competitive structure of, or competitive outcomes in, the magazine market. Caxton, a significant player in the magazine market, thinks differently. It apprehends that the merger will sorely affect its magazine interests, to the extent of possibly driving it from the market, and so it has intervened in the determination of this matter.

[21] There are, as we have said, no horizontal issues at stake in this transaction. A rather lame attempt has been made to construct some semblance of a vertical issue but, although we deal with this briefly, we think that it has so little merit that even Caxton, who is responsible for placing this issue before us, does not seem to have persisted with it. We are then left with the potential ‘portfolio effects’, with the prospect that a firm – Naspers – involved in multiple media markets, will deploy its assets in one of those markets, the pay TV market, to advance its position in a second market, that for the publication of magazines. It is common cause that these – pay TV and print media – constitute separate markets although both sell advertising frequently to the same advertisers.

[22] As already noted, Caxton was permitted to intervene in respect of its allegation that Naspers will utilise its post-merger position in pay TV to engage in the practices of foreclosure and bundling in order to bolster its position in the print media market. We will confine ourselves to an examination of these theories of harm. With respect to foreclosure, it is alleged that M-Net/SuperSport will refuse to make available advertising slots for magazines to Caxton and, presumably, to other print media groups that are in competition with Naspers. With respect to bundling and, particularly, mixed bundling, Caxton alleges that M-Net/SuperSport and Media24, the Naspers print media division, will make available bundles comprising TV advertising slots and advertising space in their magazines to prospective advertisers. We will focus on these allegations and particularly that of mixed bundling which was the only anti-competitive prospect seriously pursued by the intervener and in other submissions made to the Commission.

[23] Media markets are appropriately characterized as two-sided markets. That is to say there are, on the one hand, markets for the sale of the media products, for sales of magazines and newspapers and for the sale of viewership subscriptions for pay TV. For those whose television viewing is confined to the free-to-air (FTA) channels there is the payment of a license fee. Those consumers who choose to purchase a pay TV subscription are effectively obliged to pay a public TV license and, hence, to ‘purchase’ access to FTA TV. That is to say, while all of those who have access to FTA do not necessarily have access to pay TV because of the additional subscription and dedicated equipment that is required, those who have access to pay TV necessarily have access to FTA TV.

[24] Then there are the markets for the sale of advertising space on these varying media. There is an interface between the two sides of the market to the extent that the volume, price and content of advertising are powerfully influenced by the number and the demographic character of viewers and readers of the various media.

[25] There are two broad media markets implicated in this transaction. These are the print media markets which comprise magazines and newspapers. And there is the television market which is populated by FTA channels and pay channels. The acquiring firm – Naspers – is active in both broad markets. It owns a significant stable of magazine and newspaper titles. And it shares control, with Johncom, of M-Net and SuperSport and also of Oracle, the company which sells advertising slots on M-Net and SuperSport as well as on certain of the other channels that are part of the Dstv pay satellite bouquet provided by Multichoice. The transaction will, as already noted, result in Naspers acquiring sole control over M-Net and SuperSport.

[26] The intervener, Caxton, is active in the print media markets. It owns significant printing capacity as well as a wide array of magazine and newspaper titles. It contends that Naspers will leverage its newly unrestrained control of the pay TV businesses of M-Net and SuperSport in support of its already powerful position in the print media markets, and, particularly, in the magazine segments of the print media market.

[28] Pay TV refers to subscription based television services. Currently there are two pay TV platforms in South Africa (the M-Net terrestrial platform and the Multichoice digital satellite platform). Pay TV in South Africa started in 1986 with the launch of a single terrestrial pay TV channel (M-Net). In 1995 a multi-channel satellite subscription service, Dstv, was launched consisting of fourteen channels, with M-Net and SuperSport as its primary content providers.

[29] MultiChoice offers premium digital bouquets consisting of some 74 video channels, 65 audio channels and 8 data channels, as well as premium analogue terrestrial services consisting of sport and movies. M-Net produces 15 entertainment channels for broadcast across the African continent. M-Net has output deals with all the major film and television studios, enabling it to screen quality movies, series and miniseries. produces nine sports channels for the platform.

[30] There are currently two FTA operators in South Africa. These are the state controlled South African Broadcasting Corporation (“SABC”) which broadcasts on three domestic channels namely SABC1, SABC2 and SABC3. The SABC also broadcasts SABCAfrica, a news and entertainment channel, to the rest of the continent. The second FTA operator is e-TV which was launched in 1998. The station carries a mix of news, sports and entertainment.

[31] Caxton contends for separate markets for the sale of advertising on pay TV, on the one hand, and FTA TV, on the other. While it is not always easy to discern the basis for Caxton’s arguments, it is fair to say that an extremely limited factual basis is established in support of Caxton’s contention that there are two markets for the sale of advertising on TV. While we will not make a definitive finding on the relevant market for advertising on TV, there are strong prima facie grounds for suspecting that the market definition is significantly more complex than that contended for by Caxton.

[32] Hence we note the contention – one that appears to be common cause – that advertising flighted on pay TV, is directed at high income groups, who constitute the large bulk of consumers of pay TV. It is a contention that appears to be supported by the evidence presented to us. However in deciding the relevant market – and, in particular in deciding whether the sale of advertising on pay TV and FTA belong in separate markets or in the same market – we would have to consider whether it is appropriate to treat FTA TV as homogenous from, inter alia, the perspective of the income levels of those who view the respective FTA channels which, as we have noted, is considered to be the defining characteristic of pay TV viewers. But this is clearly not the case. At the most general level, language, for example, distinguishes the three SABC FTA channels as well as the e-TV channel from each other and language coincides to a significant extent with income levels.

[33] Thus while a more rigorous analysis of the relevant markets may well conclude that SABC1 and pay TV belong in separate relevant markets, it may equally conclude that SABC3 and pay TV occupy the same market. By the same token, the market will also clearly be impacted upon by the programming content on the respective channels. Hence the evidence shows that the Afrikaans news programme on SABC2 is a particularly highly sought after slot for advertising directed at high income viewers. A rigorous examination of the markets may then conclude that while the sale of advertising on selective programmes flighted on SABC2 belong in the same relevant market as that occupied by pay TV and SABC3, this does not apply to all advertising slots on SABC2. Again we may find that while the same advertisers contend for slots during significant sports events on the pay TV channels and on whatever FTA channel (the three SABC channels and e-TV) these significant sporting events are flighted, this may not reflect advertising demand for, say, an evening movie channel on pay TV and a late afternoon soap opera programme slot on SABC1. Note the following extract from the evidence of Mr. Peter McKenzie, the CEO of Oracle:

A lot of our advertising, whilst the number, the cost of the spot, the efficiency of the spot, the cost per point that we talked about yesterday are very important, a lot of it is event driven and that is particularly so in the case of sports. So a high profile sport event where you have got very limited inventory, I mean you can’t stop Sunday night’s Rugby semi-final to drop a commercial in when you want. Yeah, that is going to mean that the price of that spot is going to be greater.

[34] These issues pose major questions regarding the boundaries of the relevant market. While the merging parties have at least addressed these complexities in the testimony cited above, they still seem to contend for a single market for TV advertising. Caxton, for its part, has made no attempt to confront these complexities and so, all evidence to the contrary notwithstanding, it still insists on wholly separate FTA and pay TV advertising markets. While we cannot, on the evidence submitted to us, definitively identify the boundaries of the relevant market, it is clear to us that both the merging parties’ contention for a single market for advertising on pay TV and FTA as well as Caxton’s insistence that the sale of advertising on pay TV and the sale of advertising on FTA TV belong in separate relevant markets, are likely to be significant over-generalisations that are not based on a rigorous analysis of the facts.

[36] This view is then supplemented by the apparently unsupported, and contradictory, assertion that advertising on pay TV is a ‘must have’, that is, that a product directed at high income consumers and that relies on TV and print advertising, ‘must’ purchase slots on pay TV if it is to successfully market its product. We cannot identify the evidential basis for this assertion, which appears to be undermined by Marshall’s insistence that advertisers will purchase ‘both or neither’. The evidence does reveal that many of the advertisers that utilise slots on (selected) FTA channels and programmes do indeed also purchase advertising time on pay TV (just, indeed, as do these advertisers also appear to purchase advertising space on the full array of competing magazine offerings).

[37] However, on the basis of this evidence it is not clear why pay TV is more appropriately bestowed with a ‘must have’ status than are selected FTA channels and programmes, or, for that matter, than are the various magazines in which they all appear to advertise. All that is clear is that if demand for the two products is complementary – we heard much about gin and tonic in the hearings – then neither can acquire a ‘must have’ status that does not belong equally to its complement. This is, by and large, what the evidence suggests although Mr. McKenzie testified that there are many more instances of advertisers utilising only FTA TV (to the exclusion of pay TV) than the converse.

[38] We are, in any event, not at all certain that this observation suggests that the products are complementary or that they necessarily belong in separate markets. It may suggest no more than that a significant portion of the considerable cost of running an advertising campaign on television is in the production of the advertising material itself and it makes basic commercial sense to then ‘sweat’ the costly product to the fullest extent possible, to achieve, in other words, maximum coverage of the material produced. Nor are we convinced that, despite the amount of hearing time devoted to this issue, that anything of significance turns on it.

[39] Certainly it does not suggest – as does demand complementarity – that the one product (advertising on pay TV) cannot be consumed, and so will not be purchased, without the purchase and consumption of its claimed complement (advertising on FTA TV). Indeed we have been presented with no reason or evidence for believing that were the cost of advertising on pay TV to be increased by a significant amount that advertisers would not move more of their TV advertising spend to those segments of FTA TV whose viewership matched the demographic segment that watch pay TV. Given that there is no evidence suggesting collusion between pay TV and FTA TV, it is reasonable to interpret the proximity of the pay TV and SABC3 advertising rates as evidence that they are substitutes for one another, that is, that they belong in the same market.

[40] The identification of the relevant market for the sale of advertising on TV does have bearing on the adjudication of this merger. If there is a separate market for the sale of advertising on pay TV – that is, if an increase in the relative price of advertising on pay TV did not cause customers to substitute this for advertising on FTA TV – then we may justifiably conclude that M-Net/SuperSport is clearly dominant in its market and, this, as we shall outline below, may have implications for the attractiveness of a pay TV/magazine bundle, the more so if the contention that advertising on pay TV is a necessity, a ‘must have’, for advertising directed at high income consumers were to be accepted.

[41] However, the evidence seems to suggest that both pay TV and FTA TV enjoy pricing power in relation to advertisers at certain times and on certain programmes. This is likely to derive, to some extent, from the relative scarcity of these slots, a scarcity that is heavily influenced by the regulatory regime that lays down the maximum time that may be devoted to advertising. We note that pay TV appears to be allotted fewer advertising minutes than is FTA TV. But, as mentioned above, the tendency to advertise on both pay and FTA TV also probably derives from the necessity to amortise the significant costs sunk in the production of TV material over as many eyes as possible.

[42] We should emphasise that evidence of pricing behaviour indicates a close correlation between the prices charged to advertisers by pay TV and SABC3. This appears to bear out our suggestion that the FTA market is segmented and should not be viewed as a homogenous entity. It also appears to indicate keen competition for advertisers between those television platforms (such as pay TV and SABC3) with a predominantly high income viewership.

[43] In short, on the evidence presented we are unable to arrive at a firm decision on the relevant market. Although we are inclined to believe that there is a degree of substitutability between advertising on pay TV and advertising on the high income segments of FTA, the evidence presented does not permit us to make a definitive finding on the relevant market but does rather suggest that, at least as long as their prices are broadly comparable, advertisers will utilise both platforms. However, as outlined below, when we proceed to examine the likely competition implications of this transaction we will assume that Caxton’s contention that there is a separate relevant market for pay TV in which the merging parties are dominant is correct and we will show that, even under this highly favourable assumption, a persuasive case for apprehending the likelihood of a significant lessening of competition post-merger is not made out.

[45] Naspers’ print media interests are housed in Media24 which is a large publisher of magazines and newspapers as well as one of the largest printers and distributors of magazines and related products in Africa. Media24’s main competitors in South Africa include Caxton Printing and Publishing, the Independent Group and Johncom.

[46] It was submitted by Ms Patricia Scholtemeyer, CEO of Media24 Magazines that, as to frequency of publication, magazines should be placed in two broad categories, namely weekly and monthly magazines. She indeed appeared to suggest that these occupied distinct relevant markets.

[47] Ms. Scholtemeyer further contended that the other major category refers to content and the consumer at whom this content is directed. Here the main categories of magazines are women’s interest, men’s interest, general interest, home, parenting, youth, sports, motoring, travel, finance and various niche categories such as computing and agriculture. She argued that a niche brand will generally utilise a niche magazine in order for it to reach its target market. According to Ms Scholtemeyer the main drivers of competition are content, visibility and availability.

[48] Again the relevant market has not been rigorously identified by the various participants involved in the assessment of this merger. Hence both appear to analyse the magazine market as though it were a single homogenous market, despite the evidence cited above regarding distinct categories of magazines. Hence with reference to the various consumer niches referred to it is clearly unlikely that an increase in the price of gardening magazines would cause the readership to substitute in favour of car magazines. The segments may of course not be quite as strongly maintained – or, at least may be significantly broader – where the sale of advertising is concerned, although, again, common sense suggests that there is, from the advertisers’ perspective, a strong element of segmentation between these various magazine categories.

[49] Again we decline to make a definitive finding on the relevant market or markets for the sale of advertising space in magazines. We do note however that if we view the magazine market as a single market – and none of the submissions have taken another view even though the evidence does appear to cast doubt upon it - then Naspers is clearly dominant. In the weekly magazine segment, Ms. Scholtemeyer puts Naspers’ share at approximately 80%. The question that immediately arises is why, if Naspers/Media24 was intent on mounting a predatory strategy in order to strengthen its position in the magazine market which it already dominates, it would have had to turn to its pay TV business for assistance. Naspers/Media24 seems, in other words, well placed to predate in the magazine market without the costly and time consuming expedient of taking Johncom out of the pay TV business and without constructing the elaborate and, as we shall see, complex, bundles between advertising on pay TV and magazine advertising that are at the heart of the case against the merger.

[50] As outlined above, there is no horizontal overlap in this transaction and, thus, no likelihood that harm to the competitive structure or competitive outcomes will emanate from that direction. There are other possible sources of harm, several of which were urged upon the Tribunal by Caxton and these we have examined. These are an allegation of likely vertical harm arising from foreclosure by M-Net and Supersport. Secondly there are alleged portfolio effects arising from the likelihood of mixed bundling.

[51] Caxton presented evidence and argument in respect of its claimed fear of possible post-merger foreclosure. It argued that advertising on pay TV played an important role in promoting magazine sales and, particularly, in the promotion of new titles. Caxton averred that Naspers would deny its competitor’s access to its pay TV platform for the purpose of advertising their publications. Although it attempted to make something of these allegations in its initial submissions to the Tribunal, Caxton does not, in the end, appear to have persisted with these inasmuch as they merited no mention in its final argument. This is not surprising because none of the evidence presented bears out the contention that Caxton has made much use of TV advertising in general, much less pay TV in particular, to promote the sale of its magazines. Even if we assume that Caxton’s contention that there is a separate relevant market for advertising on pay TV is correct, there is no evidence that advertising on pay TV constitutes an important input in the market for magazine publication.

South Africa: Competition Tribunal

Naspers Ltd and Electronic Media Network Ltd & Another (23/LM/Feb07) [2008] ZACT 10 (24 January 2008)