CC and Computicket

The judgment

The court's own file

COMPETITION TRIBUNAL OF SOUTH AFRICA Case No: CROO8Apr10 In the matter between: THE COMPETITION COMMISSION Applicant And COMPUTICKET (PTY) LTD Respondent Panel : Norman Manoim (Presiding Member) Yasmin Carrim (Tribunal Member) Andreas Wessels (Tribunal Member) Heard on : 4-6, 9-11, 13, 16, 17, 30, and 31 October 2017; 22 and 23 February 2018 Order issued on +21 January 2019 Reasons issued on +21 January 2019 Decision and order Introduction 14. In this case the Competition Tribunal (“Tribunal”) was asked to determine whether a dominant firm in the market for the provision of outsourced ticket distribution services to inventory providers for entertainment events (inter alia), abused its dominance by securing exclusive agreements with its clients. 2. The case against the respondent, Computicket (Pty) Ltd ("Computicket"), an outsourced ticket distribution service provider, was referred to the Tribunal in April 2010, by the Competition Commission (‘Commission’), following a series of complaints that were lodged by Computicket's competitors. For purposes of these reasons we shall refer to outsourced ticket distribution 1 service providers such as Computicket or its competitors as “outsourced ticket distributors”. Relief sought 3. It is common cause that the exclusive contracts that the Commission seeks to impugn were still in existence at the time the complaint was referred on 30 April 2010. It was for this reason no doubt that the Commission, as part of its relief, sought orders from the Tribunal that were consistent with this approach. Thus the relief sought in relation to the exclusive provisions included the following: 3.1.A declaratory order that the provisions contravene the Act for the period from 1999 to date of the order; 3.2. An order that the exclusivity agreements are void and of no force or effect; 3.3. An order interdicting Computicket from entering into any further exclusive contracts with inventory providers in the relevant market. However, the Commission at the commencement of this hearing abandoned the prayers for an interdict and the voiding of the exclusivity terms. It has also confined its prayer for a declaratory orderto the period September 1999 to December 2012. Thus, the relief now sought by the Commission is confined to a declaratory order and the imposition ofan administrative penalty. Scope of the hearing 6.During the evidence of the first factual witness, Computicket questioned the case it was meant to meet after an attempt by the Commission's counsel to widen its case by adducing testimony that Computicket’s conduct was exclusionary, even in circumstances where the contracts did not contain exclusivity clauses. Computicket's legal team argued that the Commission's case was confined to that made out in the pleadings viz. only the contracts containing exclusivity provisions. The Commission contended that its case was not confined to the contracts but also how the behaviour of Computicket in the market place reinforced the exclusionary nature of the contracts. The Tribunal was asked by Computicket to rule on the matter. We ruled in Computicket's favour and the case was confined to the issue of whether the exclusivity clauses in the contracts had an exclusionary effect.* Our reason for doing so was simple. The contracts had been the basis of the Commission's case from the beginning until this moment during the hearing, and it would have been unfair at that late stage for Computicket to have to meet an additional allegation of exclusion for which it had not come prepared. Procedural background 9. 10. 11. The origins of this case date back to February 2008, when a rival of Computicket known as Strictly Tickets CC (“Strictly Tickets’) laid a complaint with the Commission. This complaint was followed by complaints from four other firms. The complainants were Soundalite CC, KZN Entertainment New and Reviews CC, L Square Technologies, and Ezimidlalo Technologies cc? The Commission decided to consolidate these complaints as they raised overlapping issues and this led to the present complaint referral which was filed with the Tribunal on 30 April 2010. The hearing in this matter commenced more than seven years later on 4 October 2017. The long delay can be attributed to a lengthy and litigious history between the parties over discovery of documents followed by an unsuccessful administrative law challenge to the Commissioner's decision 1 See transcript page 435-436 for order. 2 These complaints were lodged at various times between 3 March 2008 and 7 September 2009. See Complaint referral paragraphs 6 to 14, record pages 9-10. 3 to refer the complaint in terms of section 50(2) of the Act.5 The history has been more fully set out in our prior decision on the review and it is not necessary to repeat it now. This decision explains the gap in time between the lodging of the complaint, now more than 10 years ago, and the commencement of the hearing. The hearing 12. 13. 14. 3 Our decisi The hearing lasted 13 days and apart from the oral and written testimony of witnesses, the record comprised the investigation record of the Commission, which included questionnaires sent out to several industry participants and discovered documents emanating from Computicket. The Commission led two factual witnesses; the first being Mr Gary Charne, the joint owner of Strictly Tickets, the first complainant in the matter and a competitor of Computicket. The second witness was Mr Beard Jay, at one time the Chief Executive Officer of the Johannesburg Civic Theatre (“Joburg Theatre"), a customer of Computicket. The Commission also provided a witness statement for Daryl Keith Baruffol, Ticketing Manager of Cricket South Africa (Pty) Ltd ("CSA") but did not call him to testify. However as both sides have relied on his witness statement to support different propositions, we have referred to it when relevant. Computicket led only one factual witness Mr Kurt Drennan, the General Manager of Computicket. Both sides also called economic experts. Dr. Liberty Mncube ("Mncube”), the Commission's Chief Economist was the economic expert witness for the Commission and Professor Nicola Theron (“Theron”) of Econex, a private sector economic consultancy, was the economic expert witness for Computicket*. Computicket challenged Mncube's independence in this matter and argued that his evidence should not be admitted. We deal with this issue below. this matter is reported in the dismissal application Computicket vs Competition ‘Commission, CROOBApr10/DSM022May11. “ Both experts had generated their own data from discovered documents. 4 Admissibility of expert evidence 15. 16. 17. 18. Computicket has argued for the exclusion of the expert evidence of Mncube on the grounds that he is not independent.5 Initially the challenge to his evidence was on the basis that he was an employee of the Commission and part of the investigation team that investigated the case against Computicket. During oral argument Computicket walked back from this proposition and confined its criticism only to his role in the investigation and the fact that according to it, he did not concede points that fairly, as an expert, he should have. Let us consider the facts first and then the legal test to be applied. Mncube was involved in the investigation of the complaint during the phase prior to the referral, although he says he was not one of the principal investigators. He was involved in the formulation of a questionnaire that was sent to the industry participants which has been referred to several times in this decision. He was also one of the authors of a report to the Commissioner which recommended, at the end of the investigation, that the complaint against Computicket be referred. The leading case on the subject is National Justice Compania Naviera SA v Prudential Assurance Co Ltd also known as “The /kerian Reefer’.® In that case which has been cited with approval by the Competition Appeal Court (‘CAC’) in the Saso/’ case, the court set out the duties of an independent expert. Relevant to this case is the following remark: “Expert evidence presented to the court should be, and should be seen to be, the independent product of the expert uninfluenced as to form or content by the exigencies of litigation”. ® See paragraph 69.7 of Computicket's heads of argument. © National Justice Compania Naviera SA v Prudential Assurance Co Ltd (‘The Ikerian Reefer’) [1993] 2 Lloyd's. 7 Sasol Chemical Industries Ltd v Competition Commission, Case No 131/CAC/Jun14, judgment of 17 dune 2015. 19. 20. 21. 22. 23. 24, Phipson cites a passage from a more recent United Kingdom case, Armchair Passenger Transport Ltd, where Nelson J elaborated on these issues further, in particular stating: “The questions to be determined are whether: (a) the person has relevant expertise; and (b) he is aware of his primary duty to the Court if they give expert evidence, and are willing and able despite the i tion_wit ic th thé duty.” (Our emphasis). In this regard it is clear that there has been no challenge to Mncube’s expertise and second that he is aware of his obligations to the Tribunal; he stated this unequivocally during his testimony.* He went as far as stating that if there was any conflict between his duty to the Tribunal and his duty to the Commission he knew where his duty lay. As he put it: “My employer has put me up to advise the Tribunal. They have not stated what type of advice they expect.’? Nor does it matter that he is an employee of the Commission. As Woolf M.R. held in Field v Leeds Council, as summarised again by Phipson; “ ...the simple fact of employment did not disqualify the employee from acting as an expert for his employer. An employee was capable of being independent.” Moreover, there is no suggestion in this case that Mncube was to receive any incentive dependent on the outcome of this case being decided in the Commission's favour.'? This then leaves the issue of his employment confined to the question of whether his involvement in the investigation has compromised his independence. Two facts were relied on; his involvement in the preparation of the questionnaire and the report that was served before the Commissioner to decide whether to refer the matter. No authority was advanced to suggest that an expert in this type of case should not be involved in the obtaining of the relevant evidence. The ® He stated this several times in his testimony. See transcript pages, 933, 934, 938 and 939. Transcript 937. ‘© Transcript 937. questionnaire was not sought to obtain a particular outcome but to get industry views. The fact of the matter is that Computicket placed great reliance on some of these answers itself in the course of the hearing and relied on them in its final heads of argument, which suggests that the questionnaire was relevant and unbiased. It is frequently the task of expert economists in these matters to help advise on what data need to be gathered and once gathered to analyse it. Theron herself engaged in investigating facts for the purpose of her presentation as she testified. For instance, information was obtained from third parties regarding the size of the market (reports by PWC on the size of the market)" and viewing of websites of other potential competitors,2 25. The mere fact that Mncube was engaged in preparing this questionnaire does not on its own compromise his independence. 26. The report given to the Commissioner was produced during the course of the litigation relating to the review of this referral so the Tribunal had a chance to consider it during the course of that litigation. This was unusual as such internal reports are not normally discoverable in these proceedings. For that reason if there was to be any evidence of a lack of independence, then a report that was not expected to be discovered, but was, might constitute the strongest evidence of that fact. Yet that report in our view was carefully considered and fairly stated the issues. In our decision on the review we remarked that": “Even though the report was the only document that served before the decision makers we are satisfied that it set facts and conclusions that constituted a proper basis for reaching a determination that a prohibited practice had been established” 27. Our decision in the review matter was not appealed by Computicket and it stands. * Exhibit 13 slide 21 % Ibid slides 29 and 30. *3§e@ Computicket vs Competition Commission, page 25 paragraph 96 of the decision, Case No. CROOBApr10/DSM022May11 28. As with the questionnaire there is nothing in that report that compromises Mncube's independence." 29. The test in /kerian, and the cases subsequent is not simply that the product of the expert was done in the course of litigation, but that it was influenced by the exigencies of the litigation. This means that the party seeking to have the evidence ruled inadmissible must show evidence of this influence that compromised the expert's independence. Counsel for Computicket was doubtless aware of this and for this reason in a lengthy cross examination he sought to elicit answers from Mncube that might have revealed evidence of these exigencies. As the transcript reveals he did not extract them. Mncube did not promote the case for referral. He was clear that his primary duty was to the Tribunal. His employer understood this as well. He was under no pressure to recommend that the case be referred nor once it had been referred, to ensure that the Commission was successful. His evidence on these issues was not discredited. 30. Finally, the challenge was also based on Mncube's seeming refusal to make concessions on certain factual issues. The main complaint was that he had not conceded that two rivals had effectively entered during the period. This objection is too without foundation. The fact that an expert does not concede certain points might at best open him up for criticism, it does not serve to make the expert's testimony inadmissible. Second, we have found that Mncube's view in this regard accorded with the evidence in the record during the period we have found to have constituted the complaint period."S 31. Insofar as he was then not willing to comment on Computicket's evidence concerning entry after the complaint period, which is where the nub of the criticism lay, his reasons for doing so were sound. He explained that this period had not been analysed as it lay outside of the complaint period - and +4 In any event Mncube testified that he was not the author of the report, the investigating team was; he was responsible for the economic analysis in it. Transcript page 950. As he put it, he did not advocate the case for referral. %8 Recall that Ticketpros only entered the market in 2004 and Webtickets we have found was not an effective competitor prior to 2010. 32. 33. he could not therefore comment on whether this constituted a proper counterfactual. We find that there is no proper basis for challenging Mncube’s independence as an expert witness in this matter. The request to rule his evidence as inadmissible is refused. Nor is there any basis for Computicket's alternative request that if it is admissible, Theron’s evidence be preferred to that of Mncube where they are in disagreement.'® As these reasons indicate, there are some instances where we have preferred the evidence of the one expert over the other. We have not adopted a blanket approach to prefer the one's views over the other. To do so would be a misdirection. The objection to the acceptance of Mncube's evidence, based on an alleged lack of independence, thus fails. Factual Background 34. 35. 36. Since much of the factual record is common cause we set out these facts first. Computicket, the respondent firm in this case, was acquired by its present owner, the well-known retailer the Shoprite Group, in 2005. The business has had various owners since its establishment in 1971. Computicket describes itself as a “...centralised real-time distributor of ticketing services’."” The witnesses in this case have generally referred to firms which offer this service as an “outsourced ticket distributor" or ‘OTD' and this is the convention we will follow in these reasons. What these firms do is to sell tickets on behalf of providers of entertainment to members of the public. The providers range from theatres, concert promoters to sports stadia. In this decision we will refer to them by the generic term inventory provider or “IPs”. "8 Paragraph 69.8 of Computickel's heads of argument. *7 See answering affidavit, paragraph 23.3 record page 68. 9 37. In the period we are concerned with the OTDs all operated from some form of computer platform which has become relatively more sophisticated since the pioneering days of Computicket in the 1970's. 38. Ironically, Computicket, once the innovator that introduced the concept of outsourced ticket distribution to the South African market, stands accused in this case of using exclusivity to prevent more innovative competitors from gaining a foothold in the market because it had not kept up with technological developments that threatened its business model."® 39. The remuneration model of the OTDs differs, but essentially their source of revenue are fees charged either to the IP or the member of the public who purchases the ticket or is split between both. For this reason, the OTD service is classified as operating in a two-sided market. This is because it has two customers: the IP which wishes to sell tickets to its event and uses the services of the OTD to do so, and the end customer who wishes to purchase tickets to attend the event. As we discuss later in these reasons, Computicket argues that the two-sided nature of this market is relevant to assessing whether there has been an abuse of dominance. The Commission argues that although this is a two-sided market this feature does not alter the analysis. Merger with TicketWeb 40. Although exclusive agreements are the central focus of this case, for most of its existence, Computicket did not operate through exclusive agreements with IPs. According to Computicket these were first introduced in 1999. The Commission alleges that this was done in response to the entry of a competitor known as TicketWeb. TicketWeb, then owned by a music promotion company, had entered the market in 1998. It first partnered with music retailer Musica and later partnered with Edgars, which had a larger retail footprint."9 1® This was the thrust of the evidence of Gary Chane. 19 See witness statement of Bernard Jay paragraphs 10 to 14. 10 41. 42. 43. 45. According to Bernard Jay (“Jay"), who at the time had worked for the music promoter that owned TicketWeb, Computicket lost market share to it. Computicket's response to TicketWeb's entry was thus to enter into exclusive agreements with some providers. These exclusive agreements were to differ from the later ones during the complaint period in two respects. First, they were for a short duration, four months or less.” Second, there were fewer contracts then in place. The Commission claims that Computicket discovered 77 contracts relating to the period 1999-2001. Jay's evidence that Computicket lost market share to TicketWeb may well be explained by the difference in pricing. The Commission compared TicketWeb's prices with those of Computicket and concluded that the latter's standard commission during the period was 4% compared with that of Computicket whose commission on average was 5%.24 This did not mean that TicketWeb was always cheaper than Computicket during this period. The TicketWeb fee for its outlets and call centre was higher than for its online sales and it also increased its fee in November 2000. At some ticket prices Computicket was more competitive? The Commission's point here is that during the period that TicketWeb operated, it imposed a competitive constraint on Computicket. Jay stated that despite this, TicketWeb's market share declined. According to him Computicket offered to undercut TicketWeb if inventory providers agreed to sign exclusivity agreements. In 2002, Computicket acquired TicketWeb. According to Jay, Computicket continued to offer exclusivity contracts after the acquisition and extended them to IPs which were previous clients of TicketWeb. He explained its rationale for doing so: 2 See Expert statement of Commission paragraph 27. 2" See Expert statement of Commission paragraph 115. 22 See Expert statement of Commission Figure 4. nn 47. 48. 49. 50. 51. “| believe Computicket did this to maintain its dominant position in the market and to deter future potential competition such as TicketWeb”.23 Jay's opinion on this is fortified by Computicket's own internal documents. Prior to the merger with TicketWeb an internal recommendation was prepared in Computicket. In the document entitled Project Symphony the following observations are made under the heading Benefits of the merger: “CTK removes its main competitor whose presence has resulted in a reduction of commissions and service charges over the past two years”. And later on in the same document: “CTK can return to the position as the one stop shop for all the consumers ticketing needs, removing confusion in the minds of consumers regarding where to acquire pre-event tickets.’ Price Increases Post-merger with TicketWeb, Computicket implemented two price increases. The first was in April 2002 and the second around mid-2003. Cumulatively these price increases over the period led to a price increase of between 33% to 100% depending on the ticket price.25 Theron does not seriously dispute these figures. The dispute is over what conclusions to draw from this. The Commission first relied on these figures in anticipation of a dispute over market definition. That point is now moot as Computicket accepts the Commission's market definition and that it holds a dominant position in that market.26 2 Jay witness statement paragraph 14. 2 Record page 2893. 25 See Expert statement of Commission paragraph 119. 28 Part of this dispute was over whether firms that self-supplied for example, movie houses that do their ‘own on-line bookings, formed part of the relevant market. The Commission contended that they did not while Computicket contended for their inclusion. Computicket has since accepted the Commission's approach. 12 52. 53. 54. 55. 56. 57. There is also a difference of opinion between the experts as to whether the period of competition between the two firms prior to the merger, represented normal competition in this sector and would thus represent price levels in a proxy for a competitive market - as Mncube testified - or was a price war— as Theron suggested. What is more fruitful to glean from this history is that it explains why Computicket changed the nature of its exclusive contracts at a later stage. The short-term contracts had not led to the elimination of a competitor, hence requiring the merger with the competitor to increase profits, this could explain why the contracts became longer in duration and more aggressive when the next rival entered the market. Merger with Shoprite The next rival to enter the market was a firm called Ticket Shop. Interestingly it was owned by Shoprite, now the current owner of Computicket. Ticket Shop had a brief existence. It entered the market in late 2004 and exited in 2006.” Nevertheless, according to the Commission, Ticket Shop “acted as a competitive constraint on Computicket” during this brief period.?8 However, the competition between the two firms did not last long. Again, a merger changed the market dynamic. This time however Computicket was the target not the purchaser. In mid-2005 Shoprite purchased Computicket from its then owner MWEB, a subsidiary of the Naspers group. The merger was notified to the Commission and was approved. A condition that Shoprite imposed on Computicket as a condition of the sale was crucial to the facts of this case and what later transpired. Shoprite required a profit guarantee from the sellers. In order to ensure they could meet this 2 Competition Commission expert report paragraph 127. 28 The Commission relies for this on internal documents from Computicket. See Commission's export report paragraph 128 and Table 4 which quotes from an internal report by Computicket in December2005 which attributed a net margin decrease from 33% to 28% as “a result of increased competition from the Shoprile ticketing system.” 13 59. 60. 61. guarantee, the sellers, i.e. MWEB, decided to extend the ambit of Computicket's exclusive contracts. The new exclusive contracts were longer in duration and had various other features not in the earlier contracts as we discuss more fully below. The merger with Shoprite was implemented in mid-2005. The Ticket Shop brand was removed by 2006. Since 2005, and up until now, Shoprite has continued to be the owner of the Computicket business. After 2005, some new OTDs entered the market. This fact is common cause. What is not common cause is how they fared during the complaint period. For the Commission they were unsuccessful entrants, unable to challenge Computicket's dominant position, because of the exclusionary effects of the exclusive contracts. Computicket argues that certain of these firms have been successful entrants, whilst those that failed have done so for reasons that cannot be attributed to the effects of the exclusive contracts. Since these contracts are at the heart of this case we now turn to consider both their terms and manner of their implementation. The contracts 62. 63. 64, The exclusive contracts have evolved both in form and duration since Computicket first introduced them in the late 1990's. Mr. Drennan from Computicket who had joined the company in 1996 becoming its financial manager in 2000, testified that Computicket introduced the exclusive contracts in response to the entry of TicketWeb.?9 These initial contracts according to the Commission were typically for periods of four months or less. The exclusivity clause in those contracts stated as follows: % Transcript page 614 to 615. 14 “Client agrees that Computicket's appointment to sell tickets on its behalf for the Event is exclusive and Computicket alone shall sell tickets to the Event or performance to the exclusion of any other person other than Client (and in that regard only to the extent agreed to in writing by Computicket).” 65. We will refer to these as ‘first generation’ agreements to distinguish them from the agreements that were introduced later in 2005 with enhanced exclusivity features. 66. In 2005 Computicket changed the terms of its exclusive agreements. As noted earlier, these changes came about because MWEB, which was selling Computicket to Shoprite, wanted to protect itself, because it had given revenue warrantees to Shoprite to underpin the purchase price. According to Drennan’s testimony the instruction to them from Naspers, MWEB's parent, was to “...secure our stock fora longer period oftime." 67. However, the agreements remained in this form beyond the guarantee period. In other words, under Shoprite’s ownership and up until at least the end of the complaint period, these agreements remained in force. The key clause is contained in clause 2.3 of the agreements. This states: “For the duration of this Agreement, Client appoints Company [i.e. Computicket], which accepts the appointment, to be Client's exclusive ticketing agent for all Events, and Client agrees, for the duration of this Agreement not to instruct or allow any other party to accept bookings or sell or distribute tickets to any Event without the written consent of Company.” (emphasis added) 68. Although the 2005 agreements and the first-generation agreements both had event exclusivity at their core, there are also some notable differences. 69. First the duration of the agreements. The 2005 agreements were for a minimum of three years (as opposed to four months) and contained a default annual renewal clause. The effect of this clause is that if neither party ® Transcript page 520. 15 70. 71. 72. cancelled three months prior to the expiry of the existing agreement, it would be renewed for another year, by default. Moreover, even during the renewal period the default position was that unless the contract was cancelled by the customer the contract was extended for a further year.3" Second, the first-generation agreements referred to a single event — in the renewed agreements the exclusivity pertained to all events by that client. But even the term ‘event’ acquired an expanded meaning. It went beyond the events hosted by the client and included that of any third party, ina venue owned or leased by the client. This was defined as follows: “Every event or performance organised, staged or managed by Client itself, or an association with a third party; or by a third party in a venue owned or leased by Client (or otherwise subject to the control of Client), and referred to in the applicable Event information Sheet which Client is, in terms of this Agreement, obliged to submit in terms of the Agreement.” A feature of both the first generation and 2005 agreements was an ‘all or nothing’ policy adopted by Computicket. In terms of this policy unless the client agreed to the exclusivity there would be no agreement available to them.22 It seems that conflicts with IPs over enforcement of this provision did not arise during the period the first-generation contracts were in place. Note this is the period prior to the Shoprite takeover. After the takeover, as we later consider, these conflicts emerged. It is unclear from the record whether this can be attributed to the change in management or the length of the contract term. When Mr Drennan was asked this pertinently by Mr Wilson for the Commission, he was unsure: 3'See Bundle H1 page 465 Annexure 3 of the contract which states: “This Agreement shall commence on the date of Client's signature hereof and shall continue for an initial period of three years, and unless terminated at the end of the initial period by either party giving the other three mont termination, the Agreement shall continue for successive periods of one year each, subject to the right of either party to terminate the Agreement at the end of each successive year by giving three months’ written notice oftermination prior thereto.” 82 Transcript page 699. written notice of 16 73. 74. Mr Wilson: Did the old Computicket have a more lenient approach on exclusivity? Mr Drennan: “As | said luckily, we didn’t really have to test that so it's very difficult to say that there wasa specific different mandate and that was really just in terms of the duration of contracts were very short. Yes, it sounds like a lot, but a show typically takes four to six months to take place and roll out, so | can't say with definitive (sic) that it was, would have been worse or better before that.’93 He did however testify to a far more hands-on involvement by Shoprite as ‘opposed to the prior attitude of Naspers. He said under the old management they used to report to the board once a year while Shoprite took a more active role in management. These facts are important to one aspect of the Commission's case. The Commission argued that even the short-term contracts had an exclusionary effect. Computicket, whilst denying that its contracts had an exclusionary effect, argued in the alternative that even if post 2005 the contracts were exclusionary, there is no evidence that the first-generation contracts were. Thus, both the duration and ambit of the exclusivity, and whether and how it was enforced, are relevant to making this determination, as we discuss later when we examine the test for exclusionary effects. Exclusionary effects 75.The Commission's main case is that Computicket contravened section 8(d)(i) of the Act. Although it also relies on sections 8(c) and 5(1) of the Act, the provisions of 8(d)(i) contain a reverse onus, which means it makes it easier for the Commission to prove its case. Thus, if it fails under this sub- section, it would not succeed under the other two. We will therefore only consider whether the Commission has established a contravention of section 8(d){i). This section provides as follows: °° Transcript page 702-3. % Transcript page 703. 7 76. 77. Itis prohibited for a dominant firm to- (d) engage in any of the following exclusionary acts, unless the firm concerned can show technological, efficiency or other pro-competitive gains which outweigh the anti-competitive effect of its act; (i) requiring or inducing a supplier or customer to not deal with a competitor; Mr Kushke for Computicket had argued that this section was not applicable in cases involving contracts. His argument was that since section 5(1) refers expressly to the word agreements, which section 8 does not, then any alleged prohibited practice implicating an agreement, should be prosecuted under section 5(1) and not section 8(d)(i). °° There is no basis for this argument either in the text of the Act or in its logic. The difference between section 5 and section 8 is the requirement for dominance to be established in the latter unlike in the former. While it is correct that under section 5(1) the impugned practice must involve an agreement, this does not mean we read out an agreement from the application of section 8 simply because it does not expressly use this term. All this means is that the terms of section 8, whilst narrower in respect of the class of respondent (only applying to dominant firms) is wider in concept of the practice than section 5. There was no reason for the legislature to indicate expressly that the concept of abuse under section 8 contemplated both agreements and practices not founded on an agreement. This would have been superfluous. What the legislature chose to do was to narrow the ambit of section 5. It does not follow that by doing so it also narrowed the ambit of section 8 which addresses the behaviour of dominant firms whose actions have more serious consequences for competition than non-dominant firms. °5 Section 5(1) provides that “An agreement between parties in a vertical relationship is prohibited if it has the effect of substantially preventing or lessening competition in a market, unless a party to the agreement can prove that any technological, efficiency or other pro-competitive, gain resulting from that agreement outweighs that effect.” 18 78. 79. 80. 81. 82. 83. But the argument also makes no economic sense. Indeed, it is hard to see how section 8 would be of much effect if it excluded any practice implicating an agreement. A review of the sub-paragraphs of sub-section 8(d), suggests that an agreement is central to most, if not all of the practices prescribed. By way of illustration the phrase “selling goods or services” introduces both sub- paragraphs (iil) and (iv) whilst buying-up introduces sub-paragraph (v). None of these provisions could be sensibly interpreted and applied if they excluded the notion of an agreement. The same can be said of section 8(d)(i). It too has the concept of an agreement as its central focus. ‘Requiring or inducing a supplier or customer’ is redolent of the suggestion that this pressure is exerted, whilst not always, at least often through the provisions of a contract. Nor is there any sensible reason why this should not be the case unless section 8 is to be stripped of its very essence. Indeed, many of the decided cases on abuse have had at their core, the exercise by a dominant firm of its power through the medium of a contract; e.g. the SAA cases.°° Computicket was not able to refer us to any authority from another jurisdiction, in support of this proposition. It can safely be rejected as without substance. Both the Commission and Computicket are agreed on the legal principles in approaching the interpretation of section 8(4)(i).°7 That argument can be summed up as follows: The Commission bears the onus to establish that the contracts constitute an exclusionary act. That is established if the Commission establishes that the contracts in question “... require or induce a supplier or customer not to deal with a competitor.” We have also held in SAA that: 38 Competition Commission vs South African Airways (Ply) Ltd Case No. 18/CR/Mar01 (‘SAA(1)). See also Nationwide Airlines (Pty) Lid and another vs South African Airways (Pty) Ltd (B0/CR/Sept06) (‘SAA(2)’). 57 Commission's heads paragraph 14, Computicket heads paragraph 44. 19 “If the conduct meets the requirements of the definition, we then enquire whether the exclusionary act has an anticompetitive effect. This question will be answered in the affirmative if there is (i) evidence of actual harm to consumer welfare or (ii) if the exclusionary act is substantial or significant in terms of its effect in foreclosing the market to rivals.’8 84. We will now apply this approach to examine whether the Commission has established all the elements of the contravention. Dominance 85. The first requirement for liability in terms of section 8(d)(i) is that the firm concerned is dominant. This fact is now common cause although it was not until the commencement of these proceedings. In the complaint referral the Commission alleged that Computicket had a market share of over 95%. Computicket denied this allegation although it put up no market share of its own.” 86. However, at the commencement of proceedings counsel for Computicket confirmed that it would accept that it was dominant throughout the complaint period. This was also the approach adopted by Theron‘? During the hearing Mncube presented updated figures based on documents discovered by Computicket and he showed that during the period 2005 to 2009 Computicket's market share ranged from 95% to 99.1%.4> 87. It is also common cause that Computicket’s annual turnover throughout the complaint period exceeded the threshold set out in terms of section 6 of the Act. 88. We find that Computicket was a dominant firm for the purpose of section 7 of the Act throughout the complaint period in the OTD market. 38 See Competition Commission v South African Airways (Pty) Limited (18/CR/Mar01) at para 132. 2 See complaint referral paragraph 34. See answering affidavit paragraph 35.1. ** Transcript page 11. Transcript page 1332. “3 Mncube economic presentation, Exhibit 3, slide 23. 20 89. 90. 91. 92. Exclusionary Act The agreements in question, it is common cause, are at least facially exclusive. They prohibit inventory providers who are Computicket's customers from utilising the services of a competitor for the duration of the contract without the written consent of Computicket. This meets the legal requirement of the definition set out in section 8(d)(i). Applying the test in SAA The main dispute in this case relates to three issues of difference. First, has the Commission discharged its evidential onus of establishing an anticompetitive effect? Second, even if it has for some part of the complaint period, has it done so for the full complaint period, which is on the Commission's version a period from 1999 to 2012. Third, assuming the Commission has discharged this onus for some, or all of the complaint period, this does not end the matter. We have to consider whether the conduct complained of nevertheless results in any “... technological, efficiency or other pro-competitive gains that outweigh the anti-competitive effect.” Here the Act makes it clear that the evidential burden rests on the dominant firm. As a short hand we will refer to this from now on as the efficiency defence. Anticompetitive effects The primary anticompetitive effect which the Commission contends for in this case is that the agreements had a substantial exclusionary effect on rivals by foreclosing the market to them throughout the complaint period because they were not able to compete for sufficient inventory (i.e. tickets) to reach the scale needed to compete effectively in the market. But the Commission also alleges other harms. Inventory providers who were the customers on one side of the market were prevented from contracting with other suppliers and this reduced their ability to sell all their inventory. Second the Commission contends that in real terms, booking fees have 21 93. 94, 95. 96. 97. 98. increased.“* The Commission also alleges that as an indirect effect of the foreclosure innovation was stifled in this industry as new technologies which would have threatened the business model of Computicket were marginalised to the detriment of both inventory providers and consumers. Foreclosure of rivals There is no dispute about which firms were in the relevant market during the complaint period. There were very few entrants and due to the mergers mentioned earlier, at least two exits. For the Commission this sparsity is indicative of the exclusionary nature of the contracts. Computicket does not dispute that competitive entry has been limited. What it does dispute is whether this is attributable to the exclusive contracts. Rather, Computicket explains this on the basis of the superiority of its business model, its strong brand and credibility with suppliers and its investment in its technology. Ata theoretical level, both economists were agreed that if there was a robust counterfactual this would answer the question of the contracts causative effect on foreclosure.4® What is meant by the counterfactual analysis is to ask the question — what would have happened in the market if the impugned practice — in this case the exclusive contracts — had not been in place. Theron and Mncube agree that this is the right question to ask; they disagree about how to apply this model to the facts of this case. Mncube argues that the correct counterfactual is what would have happened between 1999 and 2010 if the exclusive agreements had not been in place.4® Theron argues that only the period after 2012 can be considered a true counterfactual, as this was the period in which we see new entry and lower #* Mncube presentation, ibid, slide 90. 4° Exhibit 13, Theron’s presentation, slide 6. “Experts agree: Counterfactualis the correct approach but difficult to construct.” “® Mncube presentation, op cit slide 85. 22 99. 100. 101. 102. 103. prices, despite the fact that the exclusive agreements were still in place. In other words, what Theron is arguing is that if the contracts were exclusionary we should still be seeing anticompetitive effects in this period. Instead she argues we are seeing the benefits of competitive entry and lower prices. Mncube’s counterfactual differs from that of Theron both in its time period and its construction. He first makes the practical point that there is no data for the period prior to the first introduction of the exclusive contracts that followed the entry of TicketWeb in 1998. Thus, the pre-1999 period cannot be used as a counterfactual as Computicket had neither exclusive agreements or any competitors at the time. What Mncube relies on as a proxy instead is a period where he says there was vigorous competition. This is the period 1999-2001 which is the period when TicketWeb entered the market. This is a period of aggressive pricing by both firms. He compares this to the period after that and compares levels of profitability and prices in those periods to the later post 2005 period. Thus, he suggests that this period when exclusive contracts were in place albeit less extensive, and of shorter duration, can be used as a proxy for a counterfactual. Theron counters this by querying this logic. The Commission, she notes relies on this period (1999-2001) as part of the complaint period. If it is part of the complaint period, she argues, because some exclusive contracts were in place then it does not serve as a useful counterfactual. Theron's candidate counterfactual is the period after 2012. Here she inverts the counterfactual by choosing a period, not where there are no exclusive contracts — but rather where they exist, but the foreclosure effects according to her do not seem apparent. In other words, Theron is saying if the practice continues, but not the outcomes, then the practice is probably not exclusionary because if it were we should expect to see its outcomes throughout i.e. beyond 2012. Mncube says this attempt at creating a post 2012 counterfactual is flawed. His argument is the fact that outcomes may have improved post 2012, tells 23 us nothing. We don't know if outcomes would have been even better if the practice had ceased. Since we don’t know the answer to that question we don't have a benchmark against which to assess this outcome — thus this analysis is not fruitful. Furthermore, to the extent that this period reflects improved entry and potentially lower prices, there may be other explanations for this — he posits for instance that the development of the internet in the later period may have had an effect. 104. We agree with Mncube's critique here. There are two objections to this selection of the post 2012 as a useful counterfactual. 105. The first is the evidence for why it is a good counterfactual i.e. controlling for ‘all things being equal’, is not robust. This is because we have very little evidence about what was happening in the market during the post 2012 period compared to what we had during the complaint period. Computicket did not call as witnesses any new entrant or absent such a possi ly, any expert who had studied this market during the period. 106. The second objection is that this period never formed part of Computicket’s original defence in the matter. Recall had this case gone ahead at the close of pleadings, when it was due to start it would have been completed, in all likelihood prior to 2012 — indeed it is even post the close of pleadings which closed in 2010 and thus is not a period where the Commission has been able to construct a rebuttal case if necessary. 407. The Commission has thus not had an opportunity to investigate this period. Mncube in his evidence made it clear that the Commission had ended its investigation in November 2009.7 The Commission no longer relies for its relief on an interdict; if it was, the situation may have been different because the continued existence of the practice would have been relevant to the relief sought. If parties were able to prolong the period for the introduction of evidence to create a suitable counterfactual then cases would be interminable. At some time the door must close. “7 Cross examination of Mncube transcript 977-8. 24 108. We thus find that the introduction by Theron of this post 2012 period to constitute a relevant counterfactual is procedurally irregular and in any event, insufficiently robust. 109. We return to Mncube's candidate for the counterfactual. This first period 1999-2001 showed entry and competitive outcomes notwithstanding the presence of some exclusive agreements entered into by Computicket. This suggests two possible alternatives. The first is that the theory that the exclusive agreements were exclusionary is flawed. Recall Theron suggests that the period cannot be both a counterfactual and a period of exclusion at the same time. It has to be one or the other. We disagree with her that the 1999-2001 period does not have evidentiary value as the best available proxy for a counterfactual. This is because although there were some exclusive agreements there was a significant difference between the nature of the exclusive contracts in the two periods which explains the differences in outcomes in the 1999-2001 and post 2005 periods. 110. First, the contracts in 1999-2001 were in place for limited time periods; mostly less than four months*®. Second, there were far less contracts in place. As shown in the table below, the overwhelming majority of contracts discovered by Computicket for the period 1999 to 2012 were exclusive. In the period 1999-2001 there were 72 contracts discovered which had information on exclusivity of which 71 were exclusive. The number of discovered contracts however increase drastically in 2006 following the take- over by Shoprite, from 58 exclusive contracts in 2005 to 303 in 2006; 323 in 2007 and 431 in 2008. We further note below that the duration of the contracts changes significantly from 2005. * See Commission's expert report figure 7 expert witness file page 72. 25 Table 1: Exclusivity clauses in Computicket’s discovered contracts‘? Period Discovered with Exclusive Non- % | Information Exclusive | Exclusive Exclusivity 1999-2001 | 7 72 71 LL 4 99% 2002 141 139 138 1 99% 2003 129 129 128 1 90% 2004 106 81 80 1 99% 2005 109 58 58 O 100% _2006__| 324 _305___|__303 2 99% 2007 325 324 323, 1 100% 2008 494 433 431 100%_ 2009 103 64 64 0 100% 2010 36 34 34 0 100% Total 1639 1630 | 9 99% 1999-2010 | aan eee I 111. In terms of the duration of the contracts, of the 77 contracts discovered for the period 1999-2001, 72 had information on exclusivity and only 7% had a duration of more than four months. Compare this to the post 2005 period in which between 97%-100% of the contracts were multi-period, i.e. had a duration of more than four months.5 112. Third, Shoprite Checkers was a more hands on manager of the Computicket business than MWEB, as Drennan testified. As a result enforcement of the terms and conditions was a key part of Shoprite Checkers’ strategy when a new entrant in the market emerged as is evidenced most clearly by the experience of Strictly Tickets. Moreover, the customers who were targeted were, according to the evidence of Chame of Strictly Tickets, the type who * See Mncube slide 44, Table 8. 59 See Mncube slide 45, Table 9. 26 113. 114, 115. 116. could give regular and repeat business to a new entrant. In particular theatres were, the record shows, the subject of aggressive and threatening correspondence from Computicket reminding them of their exclusive agreement and threatening consequences. As a result according to Charne the inventory providers were reluctant to deal with his firm despite being satisfied with his service. On his evidence the sole reason that his firm was unable to succeed in the market was the existence and enforcement of the exclusive agreements®". Computicket contends that Strictly Tickets remained in the market until 2016. If the enforcement of the agreement was exclusionary it questions how it was able to remain in the market long after the period in which the evidence of enforcement took place. It is true that the enforcement period ran from 1999 to 2012 largely and thus some years prior to Strictly Tickets’ exit. However Strictly Tickets was for most of the period a small player which never regained its market in the theatre segment that it was expanding into in 2006. Whilst Strictly Tickets was a modest competitor compared to the resources at Computicket's command it did introduce a number of innovations that made it attractive to inventory providers and final consumers.5? First end customers could make use of its ticketless technology in 2004. The ticket would be sent by SMS to the customer's mobile phone after booking. This was an important advance on the Computicket booking system which required the customer to go into a Shoprite store to collect tickets even though they might have booked them online. Paperless technology was only introduced by Computicket in 2011, seven years after Strictly Tickets. Second, it was user friendly. Charne testified that if the customer was struggling to make a booking then the system could detect this, and they would contact the customer and enquire if it needed help. 5" Transcript page 225. 2 Charne operated from a garage in his home and it seemed had only one other staff member other than himself. 27 117. The evidence that Strictly Tickets was a competitive threat is not solely dependent on Charne's testimony. Documents in the record shows that theatres valued Strictly Tickets and wanted to use its services. These theatres included Victory Theatre, Liberty Theatre on the square, Heritage Theatre and the Dockyard Theatre to name a few®, 118. The response of these theatres is that they saw Strictly Tickets as an outlet for ticket sales and did not want to be limited to one supplier. What these inventory suppliers valued was expanding their providers so more tickets could be sold. 119. Computicket's response to this was twofold. First, that having two inventory providers was confusing and could lead to duplication. The second response was that inventory suppliers preferred using the services of Computicket. This last response is easily disposed of. If inventory providers preferred to deal with Computicket — and we accept that there were many that did — then Computicket did not need exclusive agreements to retain their business. 120. The second argument requires more consideration of the evidence. The use of more than one OTD can either relate to a single event or to the performances of a provider across events. 121. The Commission contends based on Charne’s evidence that it is possible for more than one provider to sell tickets for the same event. After all the Commission points out, the exclusive agreements still allow the inventory provider to sell its own tickets. This means the logistics of having more than one provider are not insurmountable. 122. Computicket, relying on Drennan's evidence, painted a world of chaos in which patrons landed up in seats not next to their friends and where gaps appeared in seating in some areas unfilled, whilst in others seating was cramped. He referred to this colourfully as the Swiss cheese effect preparing a drawing (see Exhibit |) where this chaos was graphically reflected. But Drennan's horror case was a deliberate caricature which could easily be *° Transcript page 58 and page 62-63. 28 avoided. Nor would such chaos occur in venues where there was no specific seating allocation or for outdoor concerts. Indeed, Charne hit back and claimed that even where Computicket had exclusivity to an event there had been double booking®. According to Charne this was due to a flaw on the Computicket system as double booking did not take place on the Strictly Tickets system. He claimed that this was because this is a volume-based business i.e. that once stock has been allocated to the OTD, tickets sold would immediately become blocked off on the system to avoid double booking. In response to the issue of fraud, Charne submitted that this was ‘one of the reasons why Strictly Tickets’ technology was superior to that of Computicket at the time as it prevented any fraudulent activity such as the duplication of tickets from taking place as each ticket had a unique 2D barcode.5> 123. Darryl Baruffol, the ticketing manager of Cricket South Africa, and a former employee of Computicket, said in his witness statement that there was no problem in appointing more than one ticket provider. All that has to be done is to block off the tickets which each provider has.®° 124, Although Drennan may have exaggerated this risk we will accept that Computicket may have a reputational interest in insisting on exclusivity fora particular event. The argument here is that if the other OTD did not deliver, some reputational damage might extend to Computicket in the minds of consumers, if it was also selling tickets for the same event. However, that does not justify requiring exclusivity for all a provider's events. Indeed, nothing would incentivise an OTD more than to know if it did not deliver, it might not win the right to sell for the customer’s next event. Also, not every event would carry this risk. The greatest reputational risk for an event is carried by the inventory provider. Presumably if having more than one OTD led to a risk of chaos they would be the best judge of this and decide on Transcript page 56-57. 5S Transcript pages 54-59. 58 Baruffol witness statement record page 29 paragraph 8. We discuss this more fully later. 29 125. 126. 127. 128. 129. exclusivity voluntarily without having this choice imposed on them by prior contract as is the case with the exclusive agreements. Entry of rivals Oneof the principal areas of disagreement between the parties was whether effective entry had been foreclosed during the relevant period (1999 -2012) due to rivals being unable to achieve sufficient scale in the market. According to the Commission one reason why the market was fragmented was because Computicket's contracts were staggered, which they indeed were. This meant that Computicket's rivals would not be aware of when individual contracts would expire. Whether or not Computicket deliberately adopted this strategy to exclude rivals or not, is not at issue. As Mncube stated the fact that Computicket sequenced its contracts in this way meant that it would have been difficult for an entrant to attract scale and become an effective competitor to Computicket, because not all contracts were expiring at the same time, reinforcing Computicket's incumbency advantage*”. According to Theron approximately 20-25% of the market became contestable each year which in her view was large®*. However, she stated that one has to move beyond simply looking at the percentage of the contracts which become available each year to the evidence of entry in order to consider the anticompetitive effects of the contracts®, There are two problems with Theron's approach to staggering. The first is that since contracts provided for default renewal of one year on a continuous basis, the periods of exclusion may have extended beyond three years. Secondly, her figure of 25% is simply based on the numbers she assumed must terminate each year not those that in fact did in practice. 57 Transcript page 861. *8 Transcript page 1272-1273. © Transcript page 1274. 130. 131. 132. 133. 134, 135. Her main counter to the staggering argument was to look at entry during the period. We discuss below that contrary to Theron’s argument, entry during the period was constrained thus validating the staggering argument made by Mncube. Mention has already been made of three of the firms that existed but exited during this period. TicketWeb and Ticket Shop were by virtue of mergers absorbed into Computicket. They therefore do not represent examples of effective entry during the complaint period. Strictly Tickets modest entry and demise have been described above. For reasons we explain later we regard the relevant period as ending by 2010. Three other firms are mentioned as entrants during this period i.e. prior to 2010. They are Webtickets which entered in 2007, Ticket Connection which entered the market in 2008 and Ticket Space which entered in 2009. © The common feature of each of these firms is that they were associated or were owned by a bricks and mortar retailer. Since the lack of such an outlet was one of the reasons Computicket attributed to Strictly Tickets’ failure, it is important to evaluate, if despite having the benefit of national retailers as an outlet, these firms could succeed in the market. Ticket Connection was the least successful of all three entrants. On paper its prospects looked good. It could distribute tickets through the internet, phone bookings and retail outlets belonging to Mr Price. Yet it ended in the same year it entered. The question is why? According to the information it submitted to the Commission during its investigation, it organised one concert for singer Josh Groban, which ended in total failure when the concert was cancelled, and it had to reimburse the patrons who had bought tickets. It was unable to do so and went into liquidation. © Theron slides page 26. 31 136. Computicket latched on to this fact to make two points. First that entry was possible to organise even a large scale event; and second that the firms’ exit was not due to the exclusivity clause, but the misfortune of the cancellation of the concert. However, Lisa Kuhle of Ticket Connection, who provided these facts to the Commission, also went on to say that the other reason for its exit was that it could not expand and get new inventory providers because most of the potential inventory providers were tied to the exclusive contracts. 6 137. In cross examination Mncube was criticised for not having obtained more information about Ticket Connection. This observation is correct, but no other facts about Ticket Connection were put to Mncube other than those the Commission had obtained. Presumably if this firm had, up until its demise, been gaining customers or market share at Computicket's expense, it would have been well aware of this. We know from other documents in the record discovered by Computicket that its employees were vigilant in monitoring the progress of rivals. 138. Whilst bad luck may have partially contributed to the firm’s downfall, we cannot ignore its evidence that the exclusionary contracts had played a role in the difficulty of its survival. Put differently, the correct way to look at this is to ask whether, despite some bad luck, the firm may have survived in the market, in the absence of the exclusionary contracts. On Ms Kuhle's version the firm could not. This version, despite first appearing in the Commission's investigative report, which Computicket was given early access to long before the hearing, was not contradicted by it. & 139. Very little was said about Ticket Space. Theron mentions in her slides that it entered the market in 2009 but that is all. Mncube had not heard of it and this again was a subject of criticism by Computicket. This seems unfair as ©! See J2 page 676, and transcript pages 1055- 1058. © As a result of lengthy litigation in which Computicket sought to review the Commission's decision to prosecute it the CAC ordered production of the Commission's investigative record. This report containing the exact same references to the information submitted by Ms Kuhle appears in the investigative report and internal document produced by the Commission's investigative team prior to the decision made to refer the present complaint. 32 140. 141. 142. 143. beside from Theron’s slide, Mncube did not know much more - it was not mentioned by Mr Drennan. Nor was its existence put to Mr Charne or Mr Jay the Commissions’ industry witnesses. Its apparent anonymity suggests that it too had entered but did not disturb or disrupt the market in any serious way. Of the three, Webtickets was the firm most likely to succeed in the OTD market and it is still in existence today. However, at the time, according to the Commission, Webtickets was still a minor player. Theron disputes this. Her one slide has a heading, “Webtickets — not just a small internet player’ She then cites five examples of its customers. However, of the five, only two may have been customers during the complaint period. The first is the Table Mountain Cable way. According to the slide “21 May 2009 — 800 000 tickets per annum." When challenged by the Commission as to her source for the 800 000 tickets sold per annum, Theron could only rely on a document in the record that made a reference to this figure. However, as the Commission pointed out, this figure was part of a proposal document which Table Mountain Cable Way had made to Webtickets i.e. this did not reflect actual sales. Theron conceded that she had not done anything else to verify this figure, which if correct would have made this one of the most significant IP customers to have.65 Theron conceded that at the time Webtickets had entered the market in an admissions market as opposed to a reserved seating market. She also conceded that they hadn't started as a large player but that their association with Pick n Pay helped them to grow and, in her words, “they managed to gain scale.’® © Exhibit 13 slide 30. ® Ibid, slide 29. ®5 Transcript 1466-7. ® Ibid 1465. 33 144. 145. 146. 4147. 148. Theron is not able to provide any evidence regarding Webtickets’ fortunes from 2007 when they entered into the market until 2009. She provided a snap shot of its website dated September 2009 which contained the logos of what appeared to be customers at the time. They were she conceded all open admissions events or customers, although she contended that this did not mean it could not gain scale and exercise a constraint on Computicket.6” Because Theron's evidence here is reliant on odd references in the record, she was not able to explain if Webtickets was an effective entrant during the complaint period. The factual witnesses who were present, Charne and Jay, considered it a marginal competitor at the time. Charne described it then as “... small and floundering around looking for inventory, getting small stuff here and there.’®®, More tellingly when asked if the name Webtickets rang a bell as a competitor he stated: “perhaps now but not at the time’®. Nor was the entry of Webtickets familiar to Jay at the time. He testified that the first time he had heard about it was “..af some point in 2013 or 2014 as a potential ticketing engine for Pick n Pay.’ He went on to say that he did not think the venture with Pick n Pay ever took place.”! During the course of its investigation the Commission had asked industry players about WebTickets. The survey was sent out in March 2009 which meant that impressions were contemporaneous with the complaint period. This is particularly important in the case of TicketWeb now much more of a presence than it was then. The Commission reports that 20 IPs responded to this questionnaire. The one question asked was are there any viable competitors to Computicket. Eleven respondents answered that they were not aware of any. Four identified Strictly Tickets (Charne's business) although only one said they used them. Only two mentioned Webtickets; one © Ibid page 1468 © Transcript page 132. © Ibid page 132. Transcript page 366. TM Ibid, 34 indicating that they were not a viable competitor and the other that they were only viable for small venues.”2 149. Drennan who was Computicket's only factual witness was asked by his counsel to describe Webtickets as he knew it in 2010. Drennan testified that Webtickets did not have the ability to manage reserved ticket venues and that it had focused on what he termed the admissions market by which he meant a venue for which reserved seating was not a requirement.’3 150. If Webtickets was considered a serious threat at the time, one would have expected Drennan to have stated this and Computicket to have discovered contemporary internal documents discussing this; neither emerged, suggesting as Charne and Jay testified, that it was a marginal player at the time and did not act as a constraint on Computicket. That this firm later after the complaint period emerged as a more serious competitor, is an interesting fact, but not one which either expert or any factual witness has analysed for us. 151. Thus, at least until 2010, Webtickets, is like the other candidate competitors in the OTD market at the time, a small player which had entered the market but had not expanded sufficiently to constitute a significant competitive threat. Evidence of implementation 152. Shoprite’s stewardship of the Computicket business saw the aggressive implementation of the exclusivity clauses once Strictly Tickets had emerged as a potentially serious competitor. It was not this way from the beginning. According to Charne, when Strictly Tickets entered the market in 2004 it was. able to share inventory with Computicket for some providers.” 153. However matters soon changed. Strictly Tickets had targeted theatres as the best way to enter the market. Partly this seems to have been because 7 See Commission expert report Record pages 50-52, paragraphs 154-5, ® Transcript page 470. 74 Charne witness statement record page 40 paragraph 13. 35 Charne had close personal connections with theatre owners but also because theatres represented continuous repeat business and hence income for an OTD. 154. Not surprisingly this is where Computicket's most aggressive enforcement efforts were directed. The first incident we have evidence of was when the Old Mutual Theatre was informed that its events were no longer to be listed on the Computicket website and that it would not be reinstated until it stopped dealing with Strictly Tickets.’5 155. There followed a litany of such threats to theatres. The Commission in its expert report identifies eight IPs, mostly theatres, who were subject to these threats. Threats ranged from delisting on the Computicket website, to removal of equipment on site, damages claims and refusal to renew agreements unless there was future compliance. 7° The font of this policy appears from the evidence given to the Commission, to be from a Mr Hayes, the then general manager of Computicket— a Shoprite man who seems less concerned with his relationships with IPs than had Alfie Reid, the National Sales manager and Drennan who presumably because of their pre-Shoprite involvement in the business had stronger personal customer relationships. 156. The Commission's case is not confined to the personal testimony of Charne ‘or what reports he had received from his potential customer. The discovered documents detail these threats made.”” Thus from at least December 2006, up until at least September 2009, there is evidence of threats made by Computicket personnel to enforce its exclusivity. 7° 157. Nor was the enforcement aimed just at theatres. It included music promoters (Showtime Management)’ and event organisers such as Whiskey Live. 75 Charne witness statement record page 41 paragraph 18.1. 78 Commission's expert report, witness file section 8.7 pages 81-82. 7 The Commission sets out a number in a table. See expert report pages 84-86. %8 Ibid page 86 where there is reference to a communication from Computicket to the organisers of the Whiskey Live Festival threatening a damages claim when the IP said it wanted to partner another supplier “who is wiling to customize software to address our ticketing needs.” 7° Transcript page 313. 36 158. Bernard Jay who was the then manager of the Johannesburg Theatre had signed a non-exclusive agreement with Computicket in 2006. But in 2009 when the theatre gave three months’ notice to terminate the contract it sought another one-year contract with Computicket on a non-exclusive basis. Computicket refused and Jay had to sign a one year exclusive agreement.® 159. Nor were customers always aware of the obligation to cancel expressly to avoid default renewal. One hapless IP according to the letters in the record was surprised to discover his contract had been rolled over when he objected to threats of enforcement of exclusivity*’. Status quo bias or inertia has long been recognized in behavioral economics. For many reasons, and to the detriment of consumers, a lack of attention to detail such as automatic renewal clauses can often be exploited by companies. As stated by Thaler and Sunstein “.when renewal is automatic, and when people have to make a phone call to cancel, the likelihood of renewal is much higher than it is when people have to indicate that they actually want to continue...’®, Automatic renewal clauses or defaults therefore act as powerful tools to attracting and maintaining a firm's market share. 160. Despite the fact that the Commission was alleging that there was a deliberate strategy to target IPs willing to use the services of rivals, Computicket did not call any witness who from the correspondence would have been relevant to rebutting this issue; i.e. Alfie Reid the author of most of this correspondence and who had supplied a joint written witness ‘statement with Drennan and Hayes who is mentioned as the decision maker in Reid's correspondence and in internal emails. 161. In the joint witness statement the response to this issue is bland. “In each of the instances, Computicket gave its clients a choice to sell through © Jay witness statement paragraph 18 page 50. © Bundle K2, page 582-583. © R Thaler & C Sunstein, Nudge: Improving Decisions about Health, Wealth and Happiness (2008). 37 Computicket or to switch to Strictly Tickets. The choice was the clients’ to make.."83 162. They then asserted that the fact clients chose to use Computicket and not Strictly Tickets was because the client believed that the latter would give them a better service than the former. Conclusion 163. The evidence of entry during the period is limited and ineffectual and consistent with the Commissions’ theory of harm. 164, Our conclusion is that based on the evidence we have, there was limited market entry during the period 2005 and 2010, a period which at the beginning and thereafter coincided with the period of the introduction of the longer-term exclusivity contracts and Computicket's aggressive enforcement of its rights under these contracts. 165. No other theory for why entry was so limited and ineffectual has been offered to rebut this conclusion. The best offered, was of entry post 2010, but this does not negate what was happening in the market for at least five to six years prior to this. Further, and importantly, this period coincided with a post 2010 rise in demand and hence supply of IPs and a rise in ticketing solutions that did not require the customer to pick up tickets from a physical outlet. What Charne described as the tedium of going to Shoprite to collect tickets. Indeed Computicket's business model was particularly reliant on customers collecting tickets from Shoprite stores. It was put to Charne in cross examination by counsel for Computicket that 80% of its revenues were generated through its physical distribution network. 166. What seems to have changed in the market in the period after 2010 was the rise in internet usage and developments in technology by rivals. This view ®3 Combined Witness Statement on behalf of Computicket, para 9.1. ® Drennan, Reid, Joint witness statement, record page 20 paragraph 9. ® Transcript page 134. 38 was best expressed by Darryl Baruffol, the ticketing manager of Cricket South Africa who stated: “The new system that has been developed for CSA would not have been feasible in the past, and it is only the increased availability of open-source software and increased internet usage in South Africa in the recent past that have made it viable now.’®6 167. Significantly he goes on to state: "Previously, CSA was dependent on Computicket as a result of its ability to reach a wide range of customers via its retail network and call centre.’®” 168. Baruffol's views on this matter can be taken seriously as he knows the industry — he was for four years an employee of Computicket.** 169. There is also evidence of new entry at this time. Five firms entered in 2010, four in 2011, one in 2012, two in 2013 and one in 2016 which is where the record ends.®? 170. We find on the basis of the evidence we have considered above that exclusionary effects are evident with sufficient robustness for the period 2005 to 2010. In the next section when we deal with anticompetitive effects we will consider whether the evidence for these effects exists during this period. The economic analysis of exclusion 171. Theron criticised the Commission for adopting a form-based rather than an effects-based approach. As we understand it, what she was arguing was that the fact that a dominant firm has exclusive agreements with customers, is insufficient on its own to constitute evidence of exclusion — one must have evidence of effects. The Commission, she stated, had not presented evidence of effects for the complaint period prior to 2005. She fairly & Baruffol witness statement record page 35 paragraph 34. * Ibid, & Ibid record page 35 paragraph 3. *® Theron slides page 26. 39 conceded that the Commission had presented evidence of exclusion after 2005. 172. Inthe period after 2005 she and Mncube differed in their interpretation of the evidence and what they chose respectively to rely on. 173. Since we have not found sufficient evidence of exclusion during the period prior to 2005, we do not need to discuss the theoretical debate between form and effects any further. 174, In the post-2005 period, we have found that entry during this period was limited and ineffectual as we discussed earlier. Why then on the facts is there a difference between the pre and post 2005 periods if Computicket had exclusive agreements in both periods? The answer lies both in the differences in the contract terms and their enforcement. Post 2005, the contracts were for three years and could be extended by default beyond that for a year at a time. The contractual terms were also more extensive in their scope. Moreover, although we do not know how much of the market was foreclosed by these contracts, we know from absolute numbers that far more were in existence than in the earlier period, thus limiting rivals’ opportunities to the more lucrative customer base, particularly the repeat business afforded by some providers such as theatres. 175. Finally, and this is the most significant difference, exclusivity was aggressively enforced from 2005 after Shoprite took over, particularly when new entrants emerged. The take it or leave it threat was particularly effective. By contrast, we have no evidence of enforcement in the prior period. 176, The earlier period however does stil have evidential value. This is the period Computicket had to compete with TicketWeb and its profitability suffered. Its pricing power was only restored after it had merged with its rival TicketWeb. This is evidenced by the two price increases that followed in 2002 and 2003. What is likely is that Computicket learnt during this period that the first- % See Exhibit 13 slide 4 where she slates: “The main point ofdifference on the form based/likely analysis is the extentof foreclosure especially during the 1995-2005 period.” 40 generation exclusive agreements did not protect it from rival entry, and that competition had proved costly. Mergers would not always be a solution to restoring pricing power, given the requirement to get them approved by the ‘Commission. An improved exclusive contract was thus a rational response. This is not mere speculation. When Computicket's erstwhile owners wanted to secure their asking price for the sale of the business to Shoprite they had to give a profit warrantee. They realised that improving the duration and nature of the exclusive agreements would be the best way to protect the profitability of the business going forward and hence protect them from liability for the guarantees. Thus the seller's appreciation of the history of the market and market dynamics were inherited by the purchaser going forward. 177. The second theoretical point ofdifference between Theron and Mncube was over the nature of the OTD market as a two-sided market. 178. According to one scholar, Giacomo Luchetta, the earlier literature has defined two aspects to a two-sided market. % First a platform exists into which two different types of users enter into a single transaction which takes place through a platform. In this case the two types of users would be the IPs who want to sell tickets to their event and the other is the final consumer wishing to purchase the tickets to the event. Both make use of the OTD, here Computicket, as the platform for their interaction. The second aspect of the two-sided market is that the numerosity of each group creates an externality or in plain English, a benefit for the other. 179. Theorists of the two-sided market suggest that it is necessary to analyse this externality effect before condemning behaviour as abusive. 180. Theron invokes this theoretical concept as she says the Commission has failed to appreciate how a two-sided market could constrain a firm’s market power. Thus in some of the classic examples of two sided markets, a credit 5 Giacomo Luchetta, “Is the Google platform a two-sided market? Journal of Competition Law and Economics, September 2013 , 10(1) 185 al 188. 8 Luchetta, op cit, page 192. 41 181. 182. 183. 184. 185. card company needs both a critical mass of users who use the card and merchants who will accept it. How was this concept apt to the exclusive agreements with IPs in this case? Apart from its invocation this was not clear from Theron's analysis which seemed to be more directed as criticism of the Commission for not having dealt with it. Yet it was Theron who introduced this concept into the analysis to show its relevance. We cannot see how she has done this. Mr Wilson for the Commission asked her pertinently in cross-examination whether Computicket required the exclusive agreements to prevent its platform from falling apart as otherwise it might not have minimum efficient scale. Theron's answer was to deny that she was suggesting that Computicket was close to a point of “implosion” but she stated that if you are losing customers ‘on the one side you will lose them on the other.®3 ‘As we understand her answer she is not justifying the exclusivity on the basis of ensuring Computicket's continued participation in the market. What then might be the benefits? In an analysis of two-sided markets the OECD has recognised that a two sided platform may have two benefits.“ The first is that it may link interrelated products together thus providing a benefit to diverse customers.® It is difficult to see how the final consumer benefits from this here. Are consumers of plays benefited by having all plays on the Computicket website or do they go to other media to find out what they wish to see and then to the OTD just to get tickets? There is nothing in the evidence which suggests that Computicket provides this benefit to consumers. ® Transcript page 1367. **Two-Sided Markets, OECD Policy Roundtables (2009). % Two-Sided Markets, OECD Policy Roundtables (2009). 42 186. 187. 188. 189. 190. Does the IP benefit from this? There is no evidence that the IP benefits in any way from their capture or as the OECD using more technical language states “...they enable the platform to internalize externalities.” The other positive justification for exclusive contracts in two sided markets identified by the OECD is that: “..exclusive contracts may be pro- competitive if they allow entrants to attain critical mass at the expense of the incumbents.” But the reverse is the case here. Computicket is the incumbent and it enjoyed a near monopoly position at the time it introduced the three-year version of the exclusive contracts in 2005. We can exclude from consideration that exclusivity is justified for Computicket to attain critical mass to enter. As said above, Theron posited a theoretical argument that a two-sided market could limit the market power of the incumbent firm. She said that you must consider customers of both sides, i.e. the inventory provider on the one hand but also whether the price is becoming so expensive that the ticket buyer (i.e. end customer) is no longer going to buy the ticket. This she argued is a constraint on the market power and the exercise of that of the incumbent on the platform.°° She said, "/ can't just increase the booking fee to the point where the tickets will become prohibitive and people will leave the platform."*7 However, Theron's argument remained theoretical. She did not credibly explain her theory in the context of the characteristics of the outsourced ticket distribution services market. End customers in this market buy a very specific, bespoke product, for example tickets for Snow White playing at a specific theatre or an event like a Lady Gaga performance. Given Computicket's exclusive contracts with inventory providers during the relevant period, combined with Computicket's effective enforcement of its “all or nothing” policy, the end customer is a price taker in this market since Transcript, page 1266, lines 7 - 19. 57 Transcript, page 1484, lines 8 - 10. 43 it has no alternative platform/ provider to switch to if itregards the ticket price as too high. Furthermore, Computicket's fees are not transparent to the end customer and therefore this does not influence the purchase decision, i.e. the end customer only sees the final price of the ticket. Importantly, an end consumer that is dissatisfied with the final ticket price will have to forgo of the product since the tickets for that specific show or event would not be available on any alternative platform. However, end customers could, for ‘example, in certain cases if they regard the price of a certain class of ticket as too high, “buy down” rather than leave the platform altogether, i.e. purchase a cheaper class of ticket for the same show or event. Theron presented no evidence of likely end-customer behavior in the relevant market or sales or customer evidence to show which and when (i.e. at what price or price increase) end customers would forgo of the product entirely and leave the Computicket platform. 191. More relevant is whether two sided aspects of this type of market raises barriers to entry for rivals by preventing them from achieving sufficient scale, which is what Charne's evidence amounts to. Here again the OECD paper is instructive: “/t is possible for a two-sided platform to use exclusive contracts fo exclude competitors. However, the welfare consequences of these contracts are not clearly harmful. Exclusive contracts may foreclose the market in a socially harmful way if one firm has exclusivity over most or all of the market and the exclusivity is persistent.’®(Our emphasis) 192. Although it is important when faced with a two-sided market to consider if one's conclusions should change — and here Theron is correct- in this matter the two sided nature of the market has if anything raised barriers to entry for rivals by depriving them of access to the scale and quality of inventory that they needed to successfully enter the market to take on a well-entrenched, well resourced, dominant incumbent. Anticompetitive effects 5° OECD paper op cit. 193. 194. Following the approach in SAA we now enquire whether the exclusionary act had an anticompetitive effect. First, we look at whether there is evidence of actual harm to consumer welfare. In a useful article in which they argue for the utility of continuing to use consumer welfare as an antitrust standard, Melamed and Petit argue for what they term “recalibrating” the standard to avoid the problem of false negatives. False negatives come about when one assumes mistakenly that there is no antitrust harm when in fact there is. Interestingly the authors offer examples of where false negatives may occur include exclusionary agreements. They explain this as follows: “There are two elements to an antitrust violation: bad conduct and more market power than there would be absent that conduct. Bad conduct is, to oversimplify, conduct that does not reduce costs or price or increase output or product quality (including innovation). Such conduct can create or increase market power... by weakening competitors and thus decreasing market rivalry, such as by tying arrangements or exclusive dealing — only, in other words by undermining the competitive process. If we believe that consumer or economic welfare is more likely to be harmed by false negatives than false positives in general or in certain kinds of cases, the balance embedded in antitrust doctrine can be recalibrated. The recalibration could take the form of increased recourse to presumptions or incipiency tests in merger or unilateral conduct assessment (e.g., the Philadelphia National Bank presumption),28 new threshold rules for specific restraints (¢.9., exclusive dealing agreements longer than X years shift the burden of proofto the defendant); relaxing evidentiary requirements, such as belief that Brooke Group requires meticulous proof of both below- cost pricing and recoupment in the same (monopolized) market and/or in the short term; and changing the conduct requirements themselves (e.g., refusals to deal and patent manipulation).” 28 (own emphasis) This extract usefully summarises the approach of how to analyse the Commission's theory of harm in this case. Computicket has used the exclusive contracts to weaken rivalry by raising the barriers to entry of competitors and thus increase its market power. 45 195. 196. 197. 198. 199. 200. 201. Because of this market power, certain, to quote the authors, ‘bad conduct’ manifested itself during the complaint period. These according to the Commission are (i) a lack of entry and hence, derivatively, a weakening of rivalry; (ii) higher prices; (ii) reduced supply to IPs; (iv) a degradation in quality and (v) a lack of innovation. The evidence of the lack of entry has been considered in the previous section and need not be repeated. Prices The experts could not agree on the pricing evidence. Each presented data for different periods based on different assumptions. Mncube's approach was to compare a period where there was competition, with a period where there was not, and to conclude that the difference in prices, and profitability indicated an anticompetitive effect. He chose the period 1999-2002, the period when TicketWeb was competing in the market, as the period of normal competition. When he compared prices in this period to those after the merger with Computicket in 2002, he found prices had escalated substantially — there were two steep increases in 2002 and 2003, followed by, post 2005 a less steep, but still steady increase. His conclusion was that the exclusionary agreements had kept out rivals and Computicket could exercise pricing power to bring prices to a supra-competitive level and to retain that pricing. This approach was criticised by Theron, who suggested that the Commission had relied on nominal pricing and that increases would have looked less drastic if the Commission had controlled for inflation in its pricing data i.e. real prices. Responding to this Mncube recalculated his figures, and argued that even in real terms, the prices reflected returns that were ‘supra-competitive. Theron had presented tables (slides 36 and 55 of Exhibit 13) where she showed what percentage of the ticket price the fees represented during the period 1999 to 2010. These figures showed that in percentage terms the 46 booking fee had not substantially increased as a component of the overall ticket price. This analysis was also challenged by the Commission. The Commission points out that the fact that the booking fee might not increase substantially as a component of the overall ticket price is not the correct analysis. The example was given of a car valued at R 1 million. If its insurance went up from R 1000 to 2000 that would not represent a substantial increase in the overall price of the car, but it is a 100% increase in the cost of insurance.°? 202. Next Theron argued that the TicketWeb period was not a representative benchmark, as this was a period characterized by a price war, not normal market activity. Thus, she argued Mncube’s benchmark was not normal competition but a period of competitive attrition. In cross-examination the Commission suggested to Theron that it was erroneous to assume that because in this period Computicket was making an accounting loss this meant that it was selling its entertainment tickets at a loss — a fact she conceded.1" 203. Thus far we have presented Theron’s critique of Mncube. But she went further and presented her own pricing data. Included in her data were prices for the post 2012 period, up until 2016. Mncube countered by arguing that this data fell outside of the complaint period (recall it ends on the Commission's version in 2012) and thus the economics of this period had not been properly analysed. Recall this was much the same argument made out about the barriers to entry discussed earlier. 204. Nevertheless Theron’s table included pricing effects during the complaint period as well. Her most important effort was to mine the Computicket data base for trends in the booking fees and commission per ticket in real terms for the period 2001 to 2016. This data showed she stated in her report “...that Transcript page 221 cross examination of Theron. 109 For instance in the Project Symphony strategic document recommending the Ticket Web merger the author reflects that both firms “are making losses with only moderate prospects of making a profit of any significance. The ticketing market is too small and under too much pressure {o sustain two competitors in the short to medium term. See Project Symphony deal sheet, record page 2891. 10 Transcript pages 1480-1482. a7 205. 206. 207. 208. 209. 210. there has been no consistent increase in average prices (booking fees and commissions) paid by inventory providers, and indeed from 2005 these prices have consistently declined.1°2 Indeed, her graph shows a decline in average booking and commission fees for the very period we are concerned with here — 2005 to 2010. This would appear to negate the Commission's case for anticompetitive pricing effects. Mncube however said this data suffered a crucial methodological error. Theron's pricing table had failed to take into account what he termed the change in the composition of demand. What he meant by this was that Theron's figures did not demonstrate what she wanted to show — that prices did not increase significantly during the period and for a period decreased. Theron's method was to divide the “... total service charge per transaction by the total number of tickets sold per transaction.”This was to get to the average booking fee per transaction. But Mncube said if for some reason demand for tickets of low prices increased then the average ticket price decreases. He showed in his slides how even if the firm had increased its prices if there was a change in the composition of demand her method would show that prices had decreased — of course a fallacious outcome."°? Theron conceded this problem existed with her analysis." But she stated that this was still the best possible analysis.1°> Mncube then recalculated the fee increases using as he stated a method that avoided the composition of demand problems. Here he stated that the average annual fee increase per ticket type ranged from 11% for a ticket priced R60 in 2002 to 52% fora ticket priced R300 in 2002.16 ‘2 Econex report witness statement file page 155, paragraph 13 and figure 1. 13 See Exhibit 3, slides 86-7. 104 Exhibit 13 slide 46. 1% Ibid, 3 46. ide 91. See also exhibit 10 A where these figures were re-calculated for the R 60 ticket. 48 211. He also critiqued Theron's decision to exclude the top 5 IP customers from her figures. The reason she did so was to deal with rebate arrangements with certain clients, such as concert promoter Big Concerts where the ticket price was rebated to the IP provider. Her reasoning was that these ostensibly large fees needed to be excluded to avoid distortions in the data." Her figure 29 for instance shows an increase in booking fees from -27% to 137%.108 242, However, Mncube argued that excluding the top five clients — which Theron did for the whole period of 2001 to 2016, the distortion appears to be only in 2015 - biased the estimates of booking fees downwards. He redid these figures to include the top 5 clients to show that the booking fees with them included were in fact higher.1° 213. Although Mncube had to perform some recalculations to provide for inflation, and his choice of the TicketWeb period as his counterfactual, may be subject to some criticism, nevertheless the probabilities favour his version of the pricing effects. 214. The evidence that post-merger with TicketWeb, Computicket was able to increase prices substantially was not contradicted. The fact that this pricing even if at a level more modest than the Commission suggests, was maintained throughout the period till 2010 suggests that Computicket's post- merger pricing power was sustained and not contested down from those levels and may even have increased further. Even Theron's slides on Computicket's income and profitability show that the firm's figures improved steadily during this period. For instance, she shows the steep increase in entertainment fee income from financial year end March 2002 (around R18 million) to financial year end 2011 (around R80 million)'®. The same trend is visible for entertainment profits on fees which go from a loss-making figure 10” Econex report, witness statement file, pages 234-5, paragraphs 247-8, 10° [bid, figure 29. 109 Exhibit 3, slide 88. 180 Exhibit 13, slide 51. 49 of R7 million in the 2002 financial year to a profit of R 27.6 million in financial year 2011.11 215. The pricing evidence is thus on a balance of probabilities consistent with the Commission's theory of harm. 216. But pricing evidence is also relevant to the issue of entry. In a contestable market one would expect if the incumbent firm is enjoying high prices and superior profits that this would be conducive to attracting effective entry or expansion from rivals. We see no evidence of either during this period suggesting that there were barriers preventing or limiting effective entry and expansion - most probably due to the exclusive agreements. 217. Our conclusion on the pricing evidence is that it is consistent with the theory of an anticompetitive effect. Certainly, there is no evidence during the period from 2005 to 2010, which shows that any entry had threatened Computicket’s pricing power, particularly when compared to the period of competition with TicketWeb. Supply decrease 218. The evidence of a reduction in supply was largely anecdotal. Important here was the evidence of Charne who contended that inventory suppliers were not able to sell as much inventory as they would have liked because they were tied to a single OTD. Charne made convincing argument that for many events the increase in the number of sellers must logically increase the probability that more tickets would be sold. Certainly, the views of some of the theatres canvassed by the Commission support this.""2 Lack of innovation and quality 219, The OTD market has the potential for technological innovation to improve the customer experience. When Computicket first introduced an outsourced computerised ticketing service in August 1971 it distinguished itself as an TM' Ibid, slide 53. Note that after the 2011 financial year the profit on entertainment declines steadily to a negative figure in 2015 and 2016 although overall the business remained profitable, presumably from ticket sales on travel, busses etc. +2 Transcript page 297-299. 50 220. 221. 222. 223. 224. innovator. According to Drennan the firm achieved “... a world first in providing successful electronic reservation services”. \ts concept was chosen as one of the Top Ten of Great South African inventions.*"3 Since that time advances in technology have changed this market from the provision of physical tickets to be purchased from bricks and mortar outlets to booking online with the functionality for tickets to be printed at home or downloaded onto a cellphone. Mr Charne testified that Computicket had not kept up with innovations that rivals had introduced. Even in 2004 when he entered, he had the technology to allow customers to purchase tickets online which could be downloaded on to a cell phone obviating the necessity to obtain printed tickets from an outlet. Other rivals have introduced this as well. Mr Drennan was defensive on this aspect. He suggested that physical outlets were still necessary as many customers did not have smart phones or access to home computers. He indicated that the soccer market customer was an example of this. He also attempted to counter Charne’s evidence by ‘suggesting that if a cellphone’s screen was damaged scanning a bar code at the venue was not feasible. He claimed that Computicket had made investments in new technology and was able to provide home ticket printing functionality. He did not indicate when this innovation had come into being nor how extensive it was. Certainly, and he conceded this, some major events still required customers to collect their tickets from a Shoprite outlet. However, in the case of one event, he suggested that this arrangement was the IP providers’ preference.'"4 While we are sceptical about the claims made by Drennan there is certainly evidence that rivals were offering greater innovation than Computicket was 113 See Drennan witness stalement, record page 2 paragraphs 3.2 and 3.3, 14 See transcript page 815. 51 225. 226. 227. 228. 229. at the time. This claim does not rest solely on the testimony of Charne but also others such as Baruffol. There is an explanation for why Computicket lacked the incentive to innovate. This is because its business model was developed by Shoprite whose strategy was to use the OTD business to drive feet into its stores via its money market counters"5. The evidence of this strategy together with the anecdotal evidence suggests that it is probable that the exclusive contracts had as an effect a slow up take in the market of innovative technology that was available but not fully implanted during the complaint period. Even though Theron did not concede that the exclusive agreements had led to a suppression of innovation she did not disagree with Mncube that innovation is a relevant consideration for the purpose of analysing anticompetitive effects. Quality of service to customers While much of the evidence of customer satisfaction coincided with effects of a lack of innovation and the inconvenience of having to go to Shoprite stores others complained about aspects of service level. Jay for instance identifies the culture as “arrogant and unfriendly” at Computicket which he testified coincided with its acquisition by Shoprite. Computicket relied on comments from some IP customers to the Commission who seemed satisfied with Computicket's service. But this is not the issue. There is no evidence that IP customers considered that service levels were better as a result of the exclusivity. Put differently those positive about Computicket’s services did not link this causally to the presence of exclusivity. Conclusion 5 Transcript page 452-453. 52 230.Whilst some evidence of anticompetitive effects was inconclusive (service quality), there is sufficient other evidence to suggest that the exclusive agreements had resulted in anticompetitive effects. The strongest evidence was that of foreclosure of the market to effective competition during the complaint period. Evidence concerning supra competitive pricing effects, a decrease in supply by inventory providers, a reluctance by Computicket to timeously make use of available advances in technology and innovation and a lack of choices for end customers, was consistent with the Commission's theory of harm. The cumulative effect of all these factors suggest that the Commission has established a case of anticompetitive effect on a balance of probabilities. Efficiency defence 231. 232. 233. The onus to prove an efficiency defence as we indicated earlier rests, in an 8(d)(i) case, with the respondent. Computicket largely relied on the evidence of Theron to establish this aspect of its case. Theron identified from the literature four reasons that could justify the existence of exclusive agreements on efficiency grounds. They are: 233.1. Client specific investment (e.g. hardware, software, seat plans, marketing contribution, etc); 233.2. Free rider risk. By this is meant that other OTD will benefit from Computicket's investment for the client; 233.3. Reduction in costs associated with splitting of the inventory; and 233.4. _ Lower transaction costs for consumers."7° “18 See Exhibit 13 slide 59. 53 234. 235. 236. The Commission accepts that in certain instances an exclusive agreement maybe justified on efficiency grounds. It does not accept they are justified in this case. According to Mncube the literature on the subject suggests exclusive agreements are justified when they protect the seller's (in this case Computicket) incentive to invest. If the seller fears that its investment will be appropriated by a rival — what is often termed in the literature as a “free ride” then it will have less of an incentive to invest in something that may be of benefit to the customer. However, not all investments by a seller justify concerns about a free ride. Mncube derives from his reading of the literature three circumstances when exclusivity is justified to achieve these efficiencies. They are: 236.1. The investment is non-contractable. In less technical language this means that if the seller can anticipate what its investment in the customer's service will be, it can provide for this expense by specifying it in the contract. It does not need to provide for exclusivity. Only if for some reason this expense can't be calculated ex ante, would an exclusivity clause be justified. 236.2. Investment is specific to that customer and can't be used for another. If the investment is not customer specific the supplier is free to use it for its other customers and hence can realise its investment. An example of a contract that was customer specific was the contract that Computicket entered into with the Natal Sharks Rugby Union. Here there was a requirement from the customer for specific services for which Computicket designed a bespoke contract to realise this investment. 236.3. Investment has external effects on competitors of the seller increasing the value of trade between the seller's rivals and its customers i.e. the free rider problem. 54 237. Theron accepts this reading of the theory in the literature; but the two economists differed over whether the three criteria are to be considered cumulative. For Mncube, all three criteria must be met for the efficiency to be recognised, whilst Theron argued that even if only one ground is met this suffices. 238. It is not necessary for us to resolve this dispute in this case. Computicket meets none of these tests. This is because it has applied the exclusive provisions in each contract it has with its IPs regardless of what type it is or what its needs are. The standard terms were applied for the standard length of time to all providers. Nor was Computicket able to provide any documentary evidence to support that these contracts were motivated by any of the efficiency concerns set out above. 239. Where Computicket did vary its exclusive agreements, it did so by way of contract. An example of this is an agreement it had with the Sharks Rugby Union who had specific needs. Here they were accommodated by an agreement proving and costing these bespoke obligations by Computicket. This is a perfect example of how an investment in the services offered to a particular IP, could be made contractable. 240. The only potential case for an efficiency that was made out was Drennan’s testimony that exclusivity was necessary to avoid the splitting of inventory when two OTDs sold tickets for the same event. Some witnesses denied splitting was a problem. For instance, Baruffol of CSA explained that there was no difficulty splitting inventory between different providers: He explained: “... all that has to done is to block off certain tickets for sale by each provider which cannot then be sold by other providers."117 241. Although the Commission contended that exclusivity was not justifiable even fora single event, we have not gone that farto reject this explanation. This however does not help Computicket discharge its evidential burden given the nature of the contracts in this case in the post 2005 period. Even if 177 Baruffol witness statement record page 29 paragraph 8. 55 242. 243. exclusivity could be justified for a single event, that does not justify exclusivity applying to all of an IPs events, and for three years of exclusivity renewable for another year by default. Finally, we turn to Computicket's evidence of how the exclusivity agreements were extended in 2005. This was not premised on any efficiency justification. Rather, as we discussed earlier, this came about because MWEB needed to secure the sale value of Computicket to Shoprite because it had given profit warrantees. This is no efficiency justification; its real purpose was to protect Computicket's pricing power, by foreclosing vital inventory from rivals and thus erecting barriers to entry. This in tum secured for MWEBa higher sale price for the business. The onus was on Computicket to justify an efficiency defence. It has not been able to discharge this onus. Conclusion 244. Remedy 245, 246. We find that Computicket's exclusionary terms in it contracts with providers constituted a contravention of section 8(d)(i) for the period mid-2005 to 2010178, There was some reluctance on the part of Computicket to have to deal with the issue of remedies until we had made a finding on the merits. However, we find no substance in this argument. Most of the debate around the appropriate remedy turned on which section of the Act we made our finding on (there is no administrative penalty applicable for a first contravention if the finding had been made in terms of section 8(c)) and the duration. Neither of these issues required that we make our finding first. Computicket was able to prepare its argument on penalties and did so. On duration we have largely found in its favour. We are of the view that we are in a position TM® Data presented has been on an annual basis i.e. for 2005. Nevertheless, for the purposes of pricing we date the period from mid-2005 which coincides with the Shoprite takeover. 56 247. 248. 249. to determine the remedy in this decision and to delay further an already lengthy process would not be in anyone's interests. Declaratory order There is no dispute that a declaratory order is competent if we find a contravention of section 8(d)(i). The only dispute in this regard was about the duration of the contravention. We have found this period for reasons explained above to be the period from 2005 to 2010. Accordingly, this declaration is made in paragraph 1 of our order. Penalty Both parties applied the six- step methodology set out in the Aveng"’® case. The Commission recommended a penalty of R 21 099 100 (twenty-one million ninety nine thousand and one hundred rand). Computicket, if we found a contravention, recommended a penalty of R 10 454 200 (ten million four hundred and fifty four thousand and two hundred rand). Interestingly although the parties used different base years for the affected turnover (the Commission used 2016; Computicket, 2010, resulting in a lower figure) and differed in the duration (Commission thirteen years, Computicket four), their final calculations, before considering aggravating and mitigating factors were fairly close. The Commission imposed no premium for aggravation but allowed for no discount for mitigation and hence its R 21 million figure; Computicket arrived at a figure of R 20.9 million but argued that it should enjoy a 50% discount for mitigation, and hence its figure is approximately half of that of the Commission's. Step one: affected turnover The affected turnover is the turnover of the affected commerce in a relevant year of assessment. Given that we have found that we only have evidence of the contravention persisting until 2010, we agree with Computicket's approach that this is the relevant year for this assessment.'”° In 2010 it is "8 Competition Commission vs Aveng (Africa) Limited and Others Case No. 84/CR/Dec09. 120 The Commission had argued for the continuation of the contracts into 2012 and applied this as the base year. 37 common cause that Computicket’s turnover in respect of entertainment ticket sales was R 52 271 000.12" ‘Step two: base percentage 250. The base amount is calculated in terms of the Aveng methodology of a percentage of the affected turnover from 0 to 30%. The base turnover is determined on the basis of the following factors; nature, gravity and extent. Computicket argued for a base amount of 10%. The essence of this argument was that IP customers would have wanted to use Computicket regardless of the exclusivity provisions and that nevertheless, post 2011, other providers have gained market share. 251. The Commission argued for a base turnover of 30%. It contended that the transgression was egregious and was enforced contrary to the wishes of Computicket's IP customers. It described the contracts as “.. baldly exclusionary in nature and without any legitimate basis.”"22 252. Certainly, the exclusivity was enforced with great aggression, particularly to thwart the entry of Strictly Tickets in and around 2004. We however have insufficient evidence to come to any conclusions for the period post 2010. Although, it appears that there may be a change in market conditions post 2010, we do not know the reasons for that. We consider a base amount of 20% to be appropriate. 253. On this basis the base amount would be R10 454 200.00 Step Three: Duration 254. Once one has calculated the base amount it is multiplied by the years of the contravention. Here the view of the period of the contravention varied considerably. The Commission argued it had lasted for 13 years (1999- 2012) while Computicket argued that at best it had lasted between 3 and five years and had as a compromise suggested a multiplier of 4. Computicket argued that the period should only run from its adoption of the three year 121 Computicket heads paragraph 186.3 and record Bundle G page 356. "22 Commission heads paragraph 331. 58 255. 256. 257. 258. 259. contracts (circa mid 2005) but noted that the period of enforcement of them had been limited to between 2007 and 2008. Since we have found the contravention to have lasted from mid-2005 to 2010, we propose to multiply the base by 4.5. This leaves a figure of R 47 043 900. Four: Appli if The penalty may not exceed 10% of the respondent firm's prior annual turnover. Here the approach taken by the Commission and Computicket has a surprising twist. The Commission takes as the relevant year the one ending in June 2016. Here the annual turnover was R 210 991 000. Computicket applied the turnover of its 2017 financial year. This amount was R 224 782 000 and thus higher than the figure the Commission relies on. We will accept the Commission's base figure since this benefits Computicket. On this approach the cap on the penalty may not exceed R21 099 100. Of course, this arithmetical consequence, may appear to render much of the earlier debate over the extent of the duration and the base percentage as academic. This is partially true, but there remains a debate over mitigation and it is important for this purpose to see where the penalty may have been, but for the consequences of the cap. Step five aggravation and mitigation Computicket argued that its penalty should be halved. The first part of the argument appears to be addressed to denying the existence of any aggravating factors. Here Computicket defended its aggressive pre-hearing litigation stance, by arguing that it was ultimately successful in obtaining the documents it had sought. That of course is true, but its ultimate prize was to successfully review the Commission's decision to refer and in this it was unsuccessful. Nevertheless, the Commission has not chosen to regard this litigious aspect of its behaviour as an aggravating factor and so we will not take it into account against Computicket. 59 260. 261. 262. 263. ORDER Computicket then sought to motivate the 50% discount on the amount thus far calculated (recall this was on their 2017 turnover of R 20 908 400.00) by alleging that consumers and IP providers benefited from its services. This is not the point. The point is whether they would have benefited by competition for these services and obtained lower prices, better quality service and more innovation without the exclusive contracts. The question is not what the market would be without Computicket but what would it be like without the exclusionary conduct. Here the answer is clear - both consumers and IP providers would have been better off, and potential rivals would not have been excluded or prevented from effectively competing and expanding in the OTD market. There is no basis to recognise any of these factors as mitigating its conduct in this case. The fact that there appears to be more entry in the period after 2010 was not due to any mitigating conduct on Computicket's behalf. We also bear in mind that the base has already been halved due to the ceiling placed in section 59(2). However, we do take into account the fact that Computicket has not previously been found in contravention of the Act and we have applied a minor discount off the cap of R 21 099 100.00 and rounded off the penalty to R 20 million.’?3 This is the amount for which Computicket is liable. Computicket has contravened section 8(d)(i) of the Act, for the period mid 2005 -2010. Computicket must pay an administrative penalty of R20 000 000 (Twenty Million Rand). Computicket must make payment of the administrative penalty within 60 business days of this order. 3 This is in effect a discount of approximately 5.2% 60 4. There is no order as to costs. 21 January 2019 Date 21 January 2019 Mr Andfges Wessels Date Ms Yasmin Carrim concurring. Tribunal Researcher: Caroline Sserufusa Tribunal In-House Economist: Karissa Moothoo Padayachie For the Commission: J. Wilson SC and P. Ngcongo, instructed by State Attorney. For the Respondent: L.S. Kuschke SC and M.J. Engelbrecht, instructed by Werksmans Attorneys 61