---

COMPETITION TRIBUNAL

REPUBLIC OF SOUTH AFRICA

Case Number: 14/IR/NOV99

In the matter between

DW Integrators CC Claimant

and

SAS Institute (Pty) Ltd Respondent

Decision on Application for Interim Relief in terms of Section 59 of the Competition Act, 89 OF 1998

________________________________________________________________

Introduction

1. This case is concerned with the complex interface between anti-trust and intellectual property – we are being asked, in the name of anti-trust, to oblige the respondent, SAS Institute (SAS), a large software firm and an uncontested owner of valuable intellectual property, to issue a licence in its intellectual property to the claimant, DW Integrators (DWI), a firm that provides consulting services to the licensees of SAS software programs. The services provided by DWI and other service providers essentially enable SAS’s clients to adapt the SAS software to their specific needs. The claimant avers that it cannot provide its services effectively without itself possessing a licence in SAS software and that SAS, by refusing to issue a licence to DWI, is preventing the latter from participating in the market.

2. In other words, the claimant avers that it is being excluded from the market by acts perpetrated by a dominant firm and, accordingly, that the respondent is in violation of Section 8(c) of the Competition Act which provides that it is an offence for a dominant firm to engage in an exclusionary act if the anti-competitive effects of that act outweigh any associated efficiency gains. Moreover, the claimant alleges that the respondent’s software to which, it claims, it is denied access, is an essential facility insofar as it is, in the words of the Act, ‘an infrastructure or resource that cannot reasonably be duplicated, and without access to which competitors cannot reasonably provide goods or services to their customers’. Accordingly, the claimant alleges that the respondent has thereby placed itself in violation of Section 8(b) of the Act, which prohibits a dominant firm from denying a competitor access to an essential facility. A violation of Section 8(b) cannot be countervailed by efficiency gains – it is, in other words, per se illegal.

3. The claimant has submitted a complaint along these lines to the Competition Commission. In addition, the claimant has asked the Tribunal to make an order in terms of Section 59 that will, in the interim, provide relief from the transgressions allegedly perpetrated by the respondent. This is the matter with which the Tribunal is presently seized. In order to grant interim relief the Tribunal must be satisfied that a restrictive practice exists; that, in the absence of an order, the claimant will incur irreparable harm or that the purposes of the Act will be frustrated; and that the balance of convenience favours the granting of an order. The Tribunal must be satisfied on all three counts failing which it is not entitled to make an order in terms of Section 59.

Background

4. The respondent, SAS Institute (Pty) Ltd, is a locally registered wholly-owned subsidiary of the SAS Institute Incorporated, a private company incorporated in the USA, and is the licensed South African distributor of its US parent’s software. The software in which SAS specializes is known as information delivery software. This type of software is used to store and manage large sets of data and is typically licensed to large companies who pay initial licence fees that often run into millions of rands.

5. DWI describes information delivery software as a specialized product that should be distinguished from three other categories of software: Personal productivity software, operational application software and transactional database software. Moreover, DWI alleges that SAS software is unique. It relies on SAS’s marketing material for substantiation of this contention. For example, it quotes SAS’s claims that SAS is the “only end-to-end solution for managing, organising and exploiting data throughout your business” and that SAS “provides the only suite of tools that allows administration of data warehouses across the enterprise”.

6. In 1995 DWI’s predecessor began providing consulting services to licensed users of SAS software in South Africa following suggestions to this effect by SAS. These services included advising on the use of SAS software, installing and customising the software, developing turnkey applications to be used in conjunction with the software, providing general support services for the software, and training staff in the use of the software.

7. SAS and DWI formalised their relationship in 1996 by entering into a ‘Quality Partner Agreement’ the purpose of which, as recorded in the preamble to the agreement, was to form the basis of a close working relationship between the parties. On the one hand, the agreement bestowed certain privileges on DWI as a quality partner. For example, SAS undertook in the agreement to recommend the consulting services of quality partners to its customers and agreed to give quality partners access to its customer mailing lists. On the other hand, it contained provisions aimed at ensuring that the quality partner’s staff were properly trained and provided a satisfactory service to SAS licensees. The agreement was valid for 1 year, renewable by agreement. Either party could terminate the agreement on 30 days written notice.

8. The Quality Partner Agreement was conditional on DWI’s holding a valid licence for SAS software in terms of a SAS’s master licence agreement. DWI duly concluded a software licence agreement with SAS in August 1996.

9. In August 1997, SAS advised DWI that it would be terminating the Quality Partner Agreement with effect from 5 September 1997. In its letter of termination, SAS ascribed its decision to terminate the agreement to “…unwarranted, unfounded, incorrect and unprofessional and slanderous statements [by] DW Integrators’ directors and contractors, with regard to SAS’s employees, products and services.” The letter further stated that various statements made by DWI’s staff were having a negative effect on the SAS Institute’s relationship with its customers. DWI in turn attributed the breakdown in its relationship with SAS to three factors: first, that SAS was unhappy with the fact that DWI was advising clients to buy only those modules of SAS software that met the clients’ technology requirements, and not the full SAS software suite; second, that DWI was advising clients to buy the software products of SAS’s competitors in cases where, in DWI’s opinion, the alternative software better suited the needs of the clients; third, that SAS started providing comprehensive support services directly to licensees, and thus began to feel threatened by DWI because it considered DWI to be its competitor in the market for support services.

10. SAS denied these allegations. It submitted that in principle it did not object to DWI’s recommending other software than SAS software, but questioned DWI’s competence to do so given the fact that DWI staff were not conversant with the full range of SAS’s software. In addition, it denied that it intended competing in the market for support services, stating that it considered the provision of consulting or support services not to be its core preferred business. It pointed out that there were a number of other consulting firms providing similar services to DWI whose participation in the market SAS did not object to and in fact encouraged.

11. Following attempts by DWI to convince SAS to restore DWI’s status as a quality partner in terms of the original agreement between them, SAS proposed a new draft Quality Partner Agreement, which DWI refused to sign because it felt that its terms were anticompetitive and threatened DWI’s independence as a software consultant. For example, the new draft required the quality partner to promote SAS software and to submit to SAS for prior review and approval all its advertising and other promotional and display material relating to services in support of SAS software.

12. A few months later, SAS cancelled its software licence agreement with DWI after DWI had failed to pay its licence fees despite several reminders. The parties are at odds as to whether SAS’s cancellation complied with the terms of the licence agreement. At any rate, DWI tried to convince SAS to change its mind and tendered payment of the licence fee. SAS, however, rejected DWI’s late payment stating that it would not accept payment “prior to being convinced that DWI’s business supported the best interests of SAS Institute”. DWI interpreted this statement to support its contention that SAS was denying DWI a licence in order to exclude it from the market unless it agreed to go along with the allegedly anticompetitive terms of SAS’s new draft Quality Partner Agreement.

13. DWI maintains that without a SAS software licence it is unable to provide adequate consulting services to its clients. It alleges that access to its clients’ software is scant consolation, since its clients do not always have spare computers available for DWI consultants to work on. Furthermore, it says that without its own licence it is unable to run training courses for its consultants, which impairs its ability to keep its consultants properly trained. As a result, it runs the risk of losing some of its clients.

Interim Relief

14. As mentioned above, the Tribunal may only grant an interim relief order in terms of Section 59 if it is satisfied that a prohibited practice has occurred, that an interim order is necessary to prevent irreparable harm or to prevent the purposes of the Act from being frustrated, and that the balance of convenience favours the granting of the order. Once the Tribunal has determined that each of these conditions has been met, it may grant an order for interim relief although it must ensure that the terms of the order are indeed interim in nature, that is that they do not inadvertently have final effect.

15. The respondent pointed out that the claimant had not made any allegation in its founding affidavit in respect of the balance of convenience and argued that the application should fail on this ground alone. While it is striking that the claimant failed to deal with this very important interim relief requirement explicitly, we are of the opinion that we have been given enough general information in this case to form an opinion in respect of this requirement. We are therefore not prepared to dismiss the claimant’s application merely because it failed to aver that the balance of convenience was in its favour.

16. Per definition an application for interim relief is decided without the advantage of a full investigation by the Competition Commission. Moreover, although not precluded from hearing oral evidence, in proceedings of this nature the Tribunal is, for the most part, obliged to base its decision on the papers submitted. Adjudication on contested evidence generally requires the benefit of the full investigation and the taking of oral evidence and this limits the Tribunal to a decision based on uncontested evidence contained in the papers.

17. Accordingly, the Tribunal will not grant interim relief lightly. The evidence upon which it must rely is limited, and, although underpinned by a rich legal and economic theory, anti-trust adjudication is enormously influenced by the facts particular to each case. This is particularly so when efficiency arguments have to be evaluated as provided for in Section 8(c). Conversely, when, as in Section 8(b), no defence is provided for, the Tribunal is required to be particularly confident of its facts before granting interim relief. Interim relief is a powerful instrument of the Competition Act, and though of vital importance in the context of anti-trust enforcement, must be approached with care.

18. Caution is particularly well-advised when dealing with the interface between anti-trust and intellectual property. We concur with the much-cited decision in Atari Games Corporation v Nintendo of America Inc (897 F.2d 1572 (Fed. Cir. 1990), which warns that “the danger of disturbing the complementary balance struck by Congress is great when a court is asked to preliminarily enjoin conduct affecting patent and antitrust rights. A preliminary injunction entered into without a sufficient factual basis and findings, though intended to maintain the status quo, can offend the public policies embodied in both the patent and anti-trust laws.” (at 1577).

Arguments in Limine

19. The respondent raised a number of points in limine in the answering affidavit filed in its defence. It has persevered with only two of these points. Firstly, it objects to the claimant’s software importation and distribution activities, which it alleges are activities that fall beyond the scope of the claimant’s founding statement. It contends that the claimant’s present application is aimed at protecting these unauthorised activities and consequently that if we were to allow the application we would be encouraging the claimant to continue to act beyond its chosen scope. This is clearly not the case. The claimant’s importation and distribution activities have no bearing on the relief sought in this application. If we were to find that the respondent’s refusal to licence is an infringement of the Competition Act, our finding would stand irrespective of whether the claimant is engaging in other activities not covered by its founding statement.

20. The respondent’s second point in limine is the argument that because the claimant’s request for the renewal of the licence and for a new licence was made and rejected before 1 September 1999, the date on which the Competition Act came into operation, the Competition Tribunal has no jurisdiction to hear this application. We are not persuaded by this argument either. The alleged anti-trust violation – the respondent’s refusal to grant a licence – is ongoing and therefore falls within the jurisdiction of the competition authorities under the Competition Act. It is irrelevant when the respondent first refused to grant a licence or what form the refusal took – the refusal continues.

The Continued Validity of the Licence Agreement

21. The principal relief that the claimant seeks from the Tribunal is a declaration that the licence agreement which the respondent refused to renew remains in full force and effect and is binding upon the claimant and respondent, provided the claimant pays the relevant licence fee. To grant an order in these terms, the Tribunal would have to find that the respondent’s refusal to renew the licence constituted a breach of contract. An enquiry into whether the respondent breached the contract is not a competition law enquiry. The competition law issue here is rather whether the respondent’s refusal to grant the licence, whether by way of renewal of the existing licence or the issuing of a new one, is an abuse of dominance in terms of the relevant provisions of Section 8 of the Act. If we find that Section 8 has been transgressed, we could order that the respondent be granted a licence, but we could not declare that this should be by way of renewal of the existing licence agreement rather than under a new licence agreement.

Abuse of a Dominant Position – the relevant market

22. We are then left with the allegation that the respondent, by refusing to enter into a new licence agreement, is abusing a dominant position and this in two ways: firstly, by perpetrating an exclusionary act in violation of 8(c); secondly, by denying the claimant access to an essential facility in violation of 8(b).

23. A necessary preliminary in establishing abuse of dominance, is establishing dominance and, in order to do this, the relevant market has to be identified. The evidence and arguments of the parties is not helpful in identifying the relevant market with the requisite degree of confidence. In interim relief proceedings where, without the benefit of the Commission’s investigation, the views of the parties are all that the Tribunal has to rely upon, the effect of the inability of the parties to establish the relevant market is particularly debilitating.

24. The claimant holds that the relevant market is the market for information delivery software and, conceivably, because of the allegedly unique qualities of the respondents product, the market for SAS information delivery software. The claimant claims that the respondent is dominant in the former market – the market for information delivery software -both internationally and domestically; and obviously that it is a monopolist in the latter market, the market for SAS information delivery software. It then identifies a ‘sub-market’ - the market for servicing SAS information delivery software - arguing that SAS is leveraging its monopoly, or, alternatively, its dominant position in the primary market in order to limit competition in the ‘sub-market’.

25. Little concrete evidence is presented in support of these various claims. We reject the contention that SAS is a monopolist. It is common cause that there are other information delivery software products available. While we are prepared to accept that none of these are homogenous ‘commodity-type’ products and that the commercial strategy of participants in this market is to continually distinguish, primarily through innovation, its particular offering from that of its competitors, we have no reason to believe that these products cannot be substituted for each other. The evidence relied upon for asserting the uniqueness (read ‘monopoly position’) of the SAS product are the boasts made in SAS promotional literature, where, per definition, SAS would be most inclined (and feel most free) to proclaim the technological uniqueness of its product.

26. That having been said, SAS clearly occupies an important place – indeed the pre-eminent place – in the global market for information delivery software. The claimant, drawing once more on SAS promotional literature, mentions a figure of 50%. This is called into question by the respondent’s rejoinder to the effect that this figure only incorporates the market shares of those firms dedicated to the production of information delivery software and, accordingly, that it does not include the market shares of some very powerful firms – IBM for example – who produce a wider range of software. While the respondent has been extremely imprecise in its rejoinder, by pointing to the casual basis whereby the claimants have arrived at the figure of 50%, the real possibility that the actual figure falls below 45% must be considered. This is important because immediately the market share falls below 45%, the Act requires an assessment of market power in order to make a finding of dominance. While again less than absolutely conclusive, the respondent’s argument calls into question the contention that SAS exercises market power. We are impressed by the respondent’s uncontested claim that information delivery software is accounting for a rapidly growing share of the total software market and, accordingly, that this will attract the attention of other software firms. When one considers the resources – both financial and human capital – available in this industry and the consequent likelihood of rapid new entry, it seems unlikely that SAS would be capable of exercising market power. In the absence of market power, the claimant would, in order to establish dominance, have to demonstrate conclusively that SAS’s international market share exceeds 45%. We do not believe that the claimant has discharged this onus where the international market is concerned.

27. The claimant appears to be on stronger ground when they argue that SAS accounts for a share of the local market for information delivery software considerably in excess of 50%. While the respondent’s rejoinder to the effect that SAS occupied a small share of the potential information delivery software market is rejected (market share is properly identified in relation to the existing not to the unquantifiable potential market), we are not satisfied that the claimant has discharged its onus to establish conclusively the respondent’s share of the South African market. However, more damaging to the claimant’s argument is its failure to establish the relevance of the South African, as opposed to the international, market for information delivery software. The nature of the product, the scale of the individual licence contracts, and the scale and multinational character of the typical customer for information delivery software suggests that the search, by the customer, for the best and most cost-efficient product is not likely to stop at South Africa’s borders – in other words we have no reason to believe that there is a national market for information delivery software and every reason to believe that it is a global or international market that is relevant for the purposes of this enquiry.

28. The respondent, for its part, holds that the market for servicing information delivery software in South Africa is relevant. It points out that it does not compete with the claimant in the market for information delivery software, but only in the South African market for the servicing of this software. The respondent points out that in this market – even if confined to the servicing of SAS information delivery software alone – the respondent, far from being dominant, has a smaller market share than the claimant.

29. We do not accept the respondent’s argument here. An abuse of dominance is generally perpetrated by a dominant firm in a particular market vis-à-vis customers or suppliers in markets down or upstream of the market in which the alleged perpetrator is dominant. In other words to establish dominance in the market for information delivery software would be a legitimate basis for examining possible abuse downstream in the servicing of this software. The question of course is whether dominance has been established in the relevant global market for information delivery software. As already elaborated, we are not satisfied that this has been established.

30. There is one remaining issue relevant to the question of establishing dominance and that concerns the claimant’s argument that, because it has established itself as a specialist service provider for SAS software, the respondent effectively enjoys the power of a monopolist in relation to the claimant, regardless of whether SAS is actually in a monopoly position in relation to the market for information delivery software (it clearly is not) or, indeed, whether it is dominant in the market for information delivery software (which we conclude has not been conclusively established). We have to tread carefully here. Were this argument to be accepted too easily it would in effect mean that any distributor or supplier or service provider that attached itself to a particular brand would be absolved of the necessity to establish dominance, but would simply have to establish that an abuse took place. A similar argument is raised in the case of franchising where the cost to the franchisee of switching from an established franchise into a new franchise relationship is prohibitive, thus according the existing franchisor effective dominance, despite the putative existence of alternative franchising opportunities. This is referred to as ‘relational dominance’. It is a controversial concept. The US courts have for instance not developed a clear and unambiguous approach to this issue. See for example Siegel v Chicken Delight 664 F.2d 43 (9^th Cir. 1971), in which the court based its analysis on a market narrowly defined as the franchisor’s line and the subsequent US Supreme Court decision in Eastman Kodak Co. v Image Technical Services, Inc. [1992] USSC 73; 504 US 451, 467 – 79 (1992), in which the court recognised that switching costs could limit the relevant market to the seller’s line of products. In contrast, the courts in a number of more recent cases have interpreted the Kodak decision narrowly and have refused to accept a market definition limited to the franchisor’s product line – see for instance Digital Equipment v Uniq Digital Tech [1996] USCA7 45; 73 F. 3d 756, 762 (7^th Cir. 1996) and other cases surveyed in McDavid and Steuer’s article in the Antitrust Law Journal, vol. 67 of 1999). Although the concept of ‘relational dominance’ might possibly be applicable in this case, we have not been provided with either a sufficient factual or conceptual basis to identify dominance purely on the basis of the relationship between the parties.

31. Our conclusion then is that dominance has not been established in the market for information delivery software, and nor has dominance been established in the relationship between the parties. There is accordingly no further basis for examining the alleged restrictive practice because it is framed as an abuse of a dominant position. Accordingly, the application for interim relief is dismissed because it has failed to establish the existence of a restrictive practice.

Irreparable harm/frustrating the purposes of the Act

32. The claimant’s failure to establish dominance, much less an abuse of the dominance alleged, means that the Tribunal is not required to examine the additional conditions that must be established if the Tribunal is to grant interim relief. We note, however, that the claimant has equally failed to establish that, in the absence of interim relief, irreparable harm will result or that the purpose of the Act will be frustrated.

33. There is clear evidence to the effect that successful providers of services to SAS clients operate without the benefit of a licence. These providers use the software licences of their clients to provide the required service. There is, to be sure, evidence that the lack of a licence will inconvenience DWI. Its ability to provide training to its consultants may suffer somewhat as will its ability to provide emergency services. However, should this matter come to full trial and should the claimants prevail at that stage, the inconvenience will be temporary. What is clear is that the claimant is, in the interim, capable of carrying out its core service functions and that it is, in fact, presently providing these services to its clients. Establishing irreparable harm requires stronger evidence than this. The claimant itself makes the highly qualified claim that, absent an order for interim relief, ‘there is a very real risk that that the claimant may breach a contract with a client, which could result in such client electing to cancel the contract’. We concur with this assessment of the potential harm and it falls well short of what the Tribunal requires to make a finding of irreparable harm.

34. Nor has the claimant established that the purposes of the Act will be frustrated. This, too, is a sterner test than that apparently assumed by the claimant. It requires more than a simple restatement of the purposes of the Act and an accompanying assertion that the respondent’s actions are at odds with these purposes. It requires tangible, demonstrative evidence that the administration and reputation of the Act will be compromised by a failure to obtain interim relief. The claimant has not discharged this onus.

Finding and order

35. The claimant’s request for interim relief is dismissed.

36. The claimant is ordered to pay the respondent’s costs in the application on the scale as between party and party, including the costs of one counsel and one attorney.

___________________________ ________________

D.H. Lewis Date

Presiding Member

Concurring: M.G. Holden and U. Bhoola

---