American Natural Soda Ash Corporation and Another v Competition Commission of South Africa (554/2003) [2005] ZASCA 42; [2005] 1 CPLR 1 (SCA); [2005] 3 All SA 1 (SCA); 2005 (6) SA 158 (SCA); 2005 (9) BCLR 862 (SCA) (13 May 2005)

Last Updated: 8 June 2005

THE SUPREME COURT OF APPEALOF SOUTH AFRICA

REPORTABLE

Case number: 554/03

In the matter between:AMERICAN NATURAL SODA ASH CORPORATION First applicant CHC GLOBAL (Pty) Ltd Second applicant andCOMPETITION COMMISSION OF SA First respondentBOTSWANA ASH (Pty) Ltd Second respondentCHEMSERVE TECHNICAL PRODUCTS (Pty) Ltd Third respondentMINISTER OF TRADE AND INDUSTRY Fourth respondentBEFORE: MPATI DP, CAMERON, NUGENT,

CONRADIE JJA and COMRIE AJA

HEARD: 24 FEBRUARY 2005DELIVERED: 13 MAY 2005Constitution – appellate court structure – section 168(3) – appeal lies from Competition Appeal Court to Supreme Court of Appeal – section 173 – court’s inherent power to protect and regulate own process – requirement that special leave to appeal be obtained imposed – Competition Act 89 of 1998 – Section 3(1) – Act applying to all economic activity having effect within Republic –adverse effect not a precondition for jurisdiction of Competition Commission – Standing before Competition Commission – Act’s provisions to be interpreted widely – no need for complainant to show particular damage – Section 4(1)(b) – admissibility of evidence before Competition Commission – Commission first to construe provisions of s 4(1)(b) to establish width of statutory prohibition – ruling on evidence premature – matter remitted to Commission to determine meaning of s 4(1)(b) and to determine what evidence, if any, is admissible to establish whether agreement at issue falls within prohibition in s 4(1)(b).

JUDGMENT

CAMERON AND NUGENT JJA:

[1] This is an application for leave to appeal against an order of the Competition Appeal Court (the CAC) in October 2002, [1] dismissing an appeal from orders of the Competition Tribunal (the Tribunal) made on 27 March 2001 and on 30 November 2001. The parties’ dispute concerns the importation from the United States of soda ash (an ingredient essential to the manufacture inter alia of glass). The applicant (Ansac) is a non-stock, non-profit Delaware corporation formed by five United States soda ash producers in the early 1980s to export their product abroad. (The second applicant is Ansac’s local distributor: we refer to it with Ansac.) Within the United States, the creation of Ansac and its operations would have been illegal under the 1890 Sherman Antitrust Act, [2] but in 1918 Congress granted export-directed cartels exemption from the antitrust legislation. [2] The question the application raises is to what extent Ansac’s activities run afoul of the South African Competition Act 89 of 1998 (the Act). That question was raised formally in October 1999, when the second respondent, a Botswana producer of soda ash (Botash), and its South African distributor, the third respondent (Chemserve), launched an application for interim relief against Ansac before the Tribunal. (We refer to those respondents together as Botash.) Botash charged that Ansac was contravening the Act’s prohibition on restrictive horizontal practices. These are found in s 4:

(1) An agreement between, or a concerted practice by, firms, or a decision by an association of firms, is prohibited if it is between parties in a horizontal relationship and if [3] –

(a) it has the effect of substantially preventing, or lessening, competition in a market, unless a party to the agreement, concerted practice, or decision can prove that any technological, efficiency or other pro-competitive gain resulting from it outweighs that effect; or

(b) it involves any of the following restrictive horizontal practices:

(i) directly or indirectly fixing a purchase or selling price or any other trading condition;

(ii) dividing markets by allocating customers, suppliers, territories, or specific types of goods or services; or

(iii) collusive tendering.’

The Act defines ‘horizontal relationship’ as ‘a relationship between competitors’ (s 1).

[3] The parties soon found themselves caught in a procedural bog. The details have already been reported [4] and we mention only the essential features. Two months after Botash’s opening salvo, Ansac launched an application against Botash, charging predatory pricing in violation of s 8 of the Act. [5] The parties withdrew their contesting challenges when the Competition Commission (the Commission) (which chapter 4 of the Act gives extensive power to initiate anti-competitive measures and investigate and evaluate alleged contraventions), [6] itself concluded that Ansac was engaging in prohibited conduct and filed a complaint with the Tribunal, only to withdraw it and file a fresh referral two months later. It was these proceedings that Botash joined when it secured the Tribunal’s leave to serve an intervening complaint on Ansac. [4] But the bog only deepened, because a year after the first application was launched, the parties were unable to agree on a statement of facts for the Tribunal, and in January 2001 Ansac applied for the complaint to be dismissed on various grounds that are not now relevant. In the reasons the Tribunal gave for its ruling made on 27 March 2001, it recorded that at a ‘pre-hearing’ it convened in relation to those issues, it requested the parties to prepare argument on the question, ‘does s 4(1)(b) allow for an efficiency defence’ because ‘the conclusion would determine whether this evidence could be led at the hearing’. (It seems that the evidence that the Tribunal had in mind was evidence that Ansac wished to lead to establish that it was a ‘legitimate cost-saving efficiency-producing joint venture’, whose savings enabled it to market North American soda ash in Southern Africa more cheaply than local competitors.) On 27 March 2001 the Tribunal rejected Ansac’s objections to the complaint, and also ruled that ‘evidence concerning any technological, efficiency, or other pro-competitive gain that might be admissible in terms of section 4(1)(a) is inadmissible in terms of section 4(1)(b).’ We deal more fully below with the meaning and effect of that ruling. [5] Eight months later, in a second ruling delivered on 30 November 2001, the Tribunal dismissed two ‘exceptions’ that Ansac had taken to the complaint. The two points concerned the scope of the Act’s territorial application; and the question whether Botash had legal standing to become a complainant when its complaint made no allegation that it had suffered particular harm from Ansac’s activities. The Tribunal rejected all of Ansac’s contentions. [6] These three rulings – on the inadmissibility of certain evidence regarding an alleged s 4(1)(b) contravention; on the scope of the Act’s application; and on Botash’s standing – the CAC upheld in dismissing Ansac’s appeal. [7] An attempt by Ansac to appeal directly to this court without obtaining the CAC’s leave foundered when this court held that such leave was required. [8] (We refer to this court’s judgment in the leave to appeal application as ‘ Ansac (1) ’). Leave was then sought from, and refused by, the CAC, resulting in the present petition for leave to appeal, which the judges who considered it referred for oral argument with the direction that the parties should be prepared, if called upon to do so, also to address the merits of the proposed appeal. [9] [7] Before we deal with the substance of the application, it is necessary to consider this court’s jurisdiction to hear the appeal.

This court’s jurisdiction to hear the appeal

[8] Section 62 of the Act deals with appeals from the CAC. It specifies first matters in respect of which the Tribunal and CAC ‘share exclusive jurisdiction’ (s 62(1)). These include (subject to limited exceptions) the interpretation and application of restrictive horizontal practices (s 62(1)(a)). Section 62(2) then confers additional (non-exclusive) jurisdiction on the CAC over the question whether action or proposed action by the Commission or Tribunal is within their respective jurisdictions (s 62(2)(a)); any constitutional matter arising in terms of the Act (s 62(2)(b)); and the question whether a matter falls within the Tribunal’s or the CAC’s exclusive jurisdiction (s 62(2)(c)). [9] Section 62(3) is the critical provision. It provides that the jurisdiction of the CAC –

‘(a) is final over a matter within its exclusive jurisdiction in terms of subsection (1); and

(b) is neither exclusive nor final in respect of a matter within its jurisdiction in terms of subsection (2).’

Section 62(4) provides expressly that, subject to leave to appeal being obtained (s 63), an appeal from the CAC lies to this court or the Constitutional Court (the CC) ‘in respect of a matter within its jurisdiction’ in terms of s 62(2) – in other words, in respect of matters over which the CAC has non-exclusive jurisdiction, including constitutional questions.

[10] As in the case of the relevant sections of the Labour Relations Act 66 of 1995 , [10] these provisions undoubtedly constitute a statutory endeavour to vest partial final appellate jurisdiction in the CAC. The effect of s 62(3) regarding appeals to the CC is uncontroversial, since it allows appeals on ‘any constitutional matter’, and under the Constitution the CC’s sole jurisdiction is in such matters. [11] No impairment of constitutionally derived appellate power is thus apparent. More difficult is the Act’s seeming attempt to limit appeals to this court. [11] In National Union of Metalworkers v Fry’s Metals , [12] which was argued before the same panel in the same week as the present application, we held that –

11.1 Any legislative endeavour to vest final appellate jurisdiction in an appeal court other than this court has to be judged in the light of the appellate structures created by the Constitution; 11.2 The Constitution provides not only that this court ‘may decide appeals in any matter’, but that it ‘is the highest court of appeal except in constitutional matters’ (s 168(3)): this provision superseded both the statutory and common law sources of this court’s jurisdiction, and there can be no reason to give it less than its full meaning in relation to both constitutional and non-constitutional matters; 11.3 The Constitution’s typology of final appellate courts is exhaustive: it does not envisage other final appeal courts with authority equivalent to that of this court and of the CC; 11.4 This court’s appellate powers do not derive from any particular statute, but from the Constitution itself; 11.5 The Constitution does not envisage that legislation can assign the jurisdiction of this court piecemeal or wholesale to other specialist tribunals with final appellate jurisdiction; 11.6 The legislature may create rights that are not appealable; but once appellate jurisdiction falls to be exercised, this court is empowered to exercise it finally (apart from the CC), since final appellate tribunals with authority similar to this court are not envisaged in the Constitution.

[12] These conclusions govern the present matter. They lead to a similar outcome. The issue in NUMSA v Fry’s Metals was the appellate structures created by the Labour Relations Act 66 of 1995 (the LRA). The relevant provisions of the LRA are replete with admonitions that they are ‘subject to the Constitution’. This particular phraseology is not manifest in the Act.

But its absence is of no significance. This is for two reasons. First, s 1(2)(a) of the Act provides expressly that the Act ‘must be interpreted –

‘in a manner that is consistent with the Constitution’.

That governs the entire Act and each of its provisions. Second, it is a principle of statutory interpretation – by now above debate or citation of authority – not only that all legislation must be interpreted in the light of the Constitution, but that ‘legitimate interpretive aids’ must, where possible, be employed to avoid a finding of unconstitutionality. Only if this is not possible should a statutory provision be found unconstitutional.[13]

[13] In accordance with the Act’s own injunction, it must be interpreted consistently with the Constitution. In accordance with sound constitutional hermeneutics, its provisions should if possible be interpreted so as not to render them unconstitutional. This result is attained if the appellate structures the Act creates are read in conjunction with and in conformity with those the Constitution establishes. Those structures must, it follows, be read as adjunct to, and not exclusionary of, the Constitution’s appellate structures. No express provision in the Act prevents this, and constitutional principle requires it. [14] The apparent attempt to vest exclusive jurisdiction in the CAC in respect of the interpretation and application of chapters 2, 3 and 5 of the Act can and must thus be read so as to be consistent with the Constitution, and the finality conferred on the CAC by s 62(3)(c) is thus subordinate to the appellate powers the Constitution confers on this court. It follows that this court has jurisdiction to consider the substance of the application for leave to appeal. [15] This conclusion does not involve a finding of unconstitutionality, but derives from an application of the Constitution’s provisions to the appellate structure created by the Act, and from following the Act’s own injunction as to its interpretation.

The test for leave to appeal

[16] The Act’s provision dealing with leave to appeal, s 63, covers only the right to appeal from the CAC in terms of s 62(4). In other words, the statute’s leave to appeal mechanism deals only with appeals from matters where the CAC exercises its non-exclusive powers. Section 63 does not deal with the right to appeal to this court conferred by the Constitution, in conjunction with which, as we have held, the Act’s own appellate structures must be interpreted. [17] In Ansac (1) , the applicants claimed that the Act’s provisions conferring exclusive final jurisdiction on the CAC were unconstitutional, and thus that they were entitled to note an appeal directly to this court without seeking leave from the CAC. This court refused the order. The basis for doing so was that ‘even if the applicants’ attack on the constitutionality of the attempted jurisdictional ouster succeeds, the need for leave to appeal will remain and will extend, on the excision of the wording complained of, to all appeals from the Competition Appeal Court’. [14] [18] It was not argued that, nor did the court consider whether, the Act’s appellate provisions should be interpreted consistently with the Constitution in a manner that avoided the need for any excision on the ground of unconstitutionality. Our present finding that the provisions can be so read thus raises a question that was not before the court in Ansac (1) , namely what procedure should govern an appeal to this court on a matter that the Act’s express leave to appeal provision does not cover. [19] In NUMSA v Fry’s Metals [15] we held that –

19.1 this court’s inherent constitutional power to protect and regulate its own process [16] empowers it to require applicants for leave to appeal from a specialist appellate tribunal to demonstrate, in addition to a reasonable prospect of success, that there are ‘special circumstances’ indicating that a further appeal should lie; 19.2 the benefit of institutional expertise, and the imperative of expedition, strongly indicate that the path to this court from a specialist tribunal should not be untrammelled; 19.3 leave to appeal is necessary to protect the process of this court against abuse by appeals that have no merit, and it is in the interests of justice that the requirement of special leave be imposed, for if appeals were allowed without trammel, the expeditious resolution of disputes would be unconscionably delayed, and the justified objects of the statute impeded.

[20] For the reasons set out in NUMSA , we come to the same conclusions here. Leave to appeal from this court is required before an appeal may be prosecuted from the CAC on the matters set out in s 62(1), and special circumstances must exist before this court will grant leave. [21] As we observed in NUMSA (para 43), the procedures for applying for leave to appeal, and the factors relevant to obtaining special leave, are well-established. [17] The criterion for the grant of special leave to appeal is not merely that there is a reasonable prospect that the decision of the CAC will be reversed – but that the applicants can establish ‘some additional factor or criterion’. One is where the matter, though depending mainly on factual issues, is of very great importance to the parties or of great public importance. In applying this criterion, this court must be satisfied, notwithstanding that there has already been an appeal to a specialist tribunal, and that the public interest demands that disputes about competition issues be resolved speedily, that the matter is objectively of such importance to the parties or the public that special leave should be granted. [22] We emphasise once more that the fact that applicants have already had a full appeal before the CAC will normally weigh heavily against the grant of leave. And the demands of expedition add further weight to that. [23] We now deal with the merits of the application.

The territorial application of the Act

[24] Section 3(1) of the Act provides that it applies (subject only to collective bargaining-related exceptions) ‘to all economic activity within, or having an effect within, the Republic’. Ansac’s argument was, and is, that the ‘effect’ the statute contemplates must be an adverse effect, whose nature must be established before it can be said that the Act applies. Because neither the Commission nor Botash in its intervention alleged that the Ansac agreement has negative or deleterious effects within the Republic, Ansac urged that the complaint be dismissed. The Tribunal rejected this contention. The CAC gave it comprehensive consideration, but also rejected it. [18] [25] Before us, Ansac did not dispute that the statutory phrase ‘an effect’ was wide and unqualified. But it persisted in the contention that s 3(1), when placed in its proper context and purposively interpreted, had to be read as bringing only anti-competitive activity within its purview. If, therefore, Ansac urged, the Tribunal examined the Ansac membership agreement in the context of evidence relating to its competitive gains and found that its effect was pro-competitive, it should decline entirely to assume statutory jurisdiction over Ansac’s activities since this would best promote the Act’s objectives. [26] The argument flies in the face of the plain meaning of the statute’s wording. ‘Effect’ is not only neutral, but extremely wide. Standing without qualification, it necessarily embraces both the benign and the malign. It is hard to avoid the conclusion that this is deliberate. The Act’s language elsewhere is pregnant with words and concepts that convey the negative effects of what it sets out to prohibit. The absence of any such terminology in the application provision must be accorded its proper significance. [27] Ansac’s argument moreover requires that words be added to s 3(1) when there is no discernible justification for doing so. Well-known canons of statutory interpretation inhibit the court’s power to do this. Words cannot by implication be read into a statute unless the implication is necessary in the sense that without it effect cannot be given to the enactment as it stands. [19] Not only is there no evident justification for the super-addition Ansac’s argument requires, but the statute can be given fully coherent effect without it. [28] Ansac’s contention moreover has the anomalous consequence that, if it were adopted, the Tribunal and the Commission would have no jurisdiction to conduct any of their activities until they had established that the economic activity at issue had a negative or deleterious effect within the Republic. A long contestation about the statute’s applicability would ensue before the Act’s institutions could assume jurisdiction. That, manifestly, is to approach the structure and operation of the Act, and the functioning of its institutions, from the wrong end. [29] The correct approach – which the wide and unqualified wording of s 3(1) requires – is that all effects are captured, but that the statute enjoins only those that are adverse. We agree with the CAC, for the reasons fully set out in its judgment, that the ‘effect’ the Act contemplates must be such that it falls within the regulatory framework created by the statute, whether anti-competitive or not. This inquiry, as Malan AJA pointed out –

‘does not involve a consideration of the positive or negative effects on competition in the regulating country, but merely whether there are sufficient jurisdictional links between the conduct and the consequences. ... The question is ... one relating to the ambit of the legislation: the Act in the matter under consideration, its regulatory “net”, concerns not only anti-competitive conduct but also conduct the import of which still has to be determined.’[20]

Ansac’s contention must for these reasons be rejected.

The intervenor’s legal standing

[30] The CAC found that even though Botash did not allege that it had suffered individual harm because of the performance of the Ansac agreement, it had the standing necessary to seek an order against Botash interdicting the continued performance of that agreement.

The CAC’s conclusion was based on a careful analysis of the Act’s provisions. [21] The CAC gave particular weight to the Act’s complaint procedure, [22] which provides that, in addition to the Competition Commissioner, ‘any person’ may submit a complaint against an alleged prohibited practice to the Commission (s 49B(2)(b)), whereupon the Commissioner ‘must direct an inspector to investigate the complaint as quickly as practicable’ (s 49B(3)). The CAC also gave weight to the fact that a complainant may apply to the Tribunal for interim relief (s 49C). [31] The CAC took further into account in deciding that Botash had standing that the statute casts the right to participate in hearings of the Tribunal widely. The complainant has this right not only if it referred ‘the complaint’ to the Tribunal (s 53(a)(ii)(aa)), but also if in the opinion of the presiding member of the Tribunal the complainant’s interest is not adequately represented by another participant (s 53(a)(ii)(bb)). [32] It is true, as Ansac pointed out on appeal, that s 53 specifically sets out the participation rights of the complainant, and that one of the factors the statute requires the Tribunal to consider in granting interim relief is ‘the need to prevent serious or irreparable harm’. [33] It does not follow from these facets of the statute, however, that an intervening party must (as Ansac urged us to find) show the ordinary common law prerequisites for obtaining relief. As the CAC rightly pointed out, the orders the Tribunal can make in response to the referral of a complainant are ‘of a limited kind to be made in the public interest’. From this the CAC inferred that a complainant need not show that it has suffered particular damage. [23] [34] We agree with the CAC’s conclusion. Ansac’s argument seeks to conclude from the limited express rights the Act confers on a participant in a hearing that the Act requires an intervenor to comply with the strict common law requisites for interdictory relief; but this is to overlook the significance of the fact that a broad ambit of participatory rights is created in the first place. Ansac likewise underscores that an applicant, to obtain interdictory relief under the Act, must place on the scale the risk to it of ‘serious or irreparable damage’; but ignores the fact that obtaining such relief may not be an intervenor’s sole interest in the proceedings. [35] We see no reason to circumscribe narrowly the right to intervene in proceedings under the Act. We therefore conclude that the absence of a claim of particular damage on the part of Botash is no bar to its title to claim relief as an intervenor.

Admissibility of evidence about the nature and effect of the agreement

[36] The greatest part of the parties’ dispute, and of the argument before this and the other forums, related to the ruling concerning the admissibility of evidence. What complicates the matter is the lack of consensus about the effect of the Tribunal’s ruling. [37] It is clear from its juxtaposition with s 4(1)(a) that s 4(1)(b) is aimed at imposing a ‘per se’ prohibition: one, in other words, in which the efficiency defence expressly contemplated by sub-para (a) cannot be raised. The reason for the blunt terms of sub-para (b) is plain. Price-fixing is inimical to economic competition, and has no place in a sound economy. Adopting the language of United States anti-trust law, price-fixing is anti-competitive per se . All countries with laws protecting economic competition prohibit the practice without more. The fact that price-fixing has occurred is by itself sufficient to brand it incapable of redemption. The Tribunal has found that once the conduct complained of is found to fall within the scope of the prohibition , that is the end of the enquiry. There is no potential for a further enquiry as to whether the conduct is justified (an enquiry of the kind that is envisaged by s 4(1)(a)), and evidence to that end is not relevant and thus inadmissible. It is this finding that the Competition Appeal Court upheld; and it is clearly correct. Indeed, none of the parties to the appeal suggests otherwise. [38] Yet there is no consensus between the parties as to whether the Tribunal’s ruling was limited to that. While the respondents contend that the Tribunal’s ruling upheld by the CAC only excludes evidence that is tendered to establish an ‘efficiency defence’ of the nature contemplated by s 4(1)(a) (an outcome not controversial before us), Ansac submits that the ruling goes much further. It contends that the Tribunal has also precluded evidence that is relevant to ‘characterising’ its conduct and thus to determining whether or not it falls within the scope of the legislative prohibition in sub-para (b) at all. [39] That lack of consensus is not altogether surprising because Ansac’s argument before the Tribunal, as recorded in its heads of argument (and repeated before us) was not directed to the question whether conduct prohibited by s 4(1)(b) could be justified by evidence. It was directed rather to the question whether evidence was admissible to determine whether Ansac’s conduct is prohibited at all: in other words, whether the Ansac agreement constitutes price-fixing as prohibited by the Act. Only in the alternative did Ansac submit that, if its conduct did not fall foul of s 4(1)(b), but was sought to be brought within the separate prohibition in s 4(1)(a) (which the Commission has not yet tried to do), evidence would be admissible to justify its conduct as envisaged by that sub-paragraph. [40] The Tribunal appears to us to have elided these two separate submissions and thereby misdirected its enquiry. It seems to have been of the view that Ansac sought to advance the evidence in order to establish that its conduct, though falling within s 4(1)(b), is nevertheless justifiable by criteria of the kind contemplated by s 4(1)(a) when that was not Ansac’s contention. That the Tribunal’s ruling posits that no evidence except the terms of the agreement in question is relevant (and thus admissible) to the question whether s 4(1)(b) has been contravened is evident from the following passage from its reasons:

‘[T]hose who set themselves the task of impugning agreements thus described in Section 4(1)(b) do not have to establish any deleterious impact on competition. All that has to be established is the existence of an agreement embodying the features detailed in Section 4(1)(b)(i)-(iii).’ (Emphasis added.)

Moreover, the terms of its ruling are sufficiently expansive to exclude all evidence relating to the purpose and effect of the agreement. Its ruling (which we repeat for convenience) was that:

‘On the argument we requested on section 4(1)(b) we find that evidence concerning any technological, efficiency, or other pro-competitive gain that might be admissible in terms of section 4(1)(a) is inadmissible in terms of section 4(1)(b).

In the reasons it gave for its later ruling of 30 November 2001, the Tribunal explained its earlier ruling as follows:

‘The panel held that Section 4(1)(b) required no showing of anti-competitive effect and that it permitted of no efficiency defence [the defence allowed for by s 4(1)(a) – the mere fact of the agreement was sufficient to condemn it.’

This lends support to Ansac’s contention that the ruling was intended to exclude all evidence except the terms of the agreement.

[41] The Tribunal’s ruling, particularly in the context of the reasons it gave, is open to the construction that (perhaps inadvertently) it has precluded evidence even if the object of advancing it is to demonstrate that Ansac’s conduct does not fall within the prohibition in s 4(1)(b) at all. To that extent its ruling was in our view premature and therefore incorrect. This ruling the CAC endorsed. In this in our view it fell into the same error. [42] But even if the ruling is no more than ambiguous, and was not intended to have that effect, it is clearly desirable that there should be clarity on the issue, bearing in mind the uncertainty that clearly exists, and the enormous expense this uncertainty has already entailed. [43] We pointed out earlier that an agreement that involves, amongst other things, price-fixing, is prohibited by s 4(1)(b), and nothing can be advanced to justify it. But when has prohibited price-fixing occurred? This is not always simple to determine. In the United States the condemnation of price-fixing arises from judicial interpretation of s 1 of the Sherman Act. [24] In the European Union, in Australia, and in this country it is decreed by legislation. [44] In the United States the enquiry is approached by ‘characterising’ the conduct complained of to determine whether it constitutes that form of conduct that the courts have through case precedents labelled ‘price-fixing’ but have not comprehensively defined. In this country, where the prohibition is decreed by legislation rather than by judicial intervention, the prohibited form of conduct must be established by construing s 4(1)(b). [45] Once the ambit of sub-para (b)’s prohibition has been established the enquiry can move to whether or not the conduct in issue falls within the terms of the prohibition. That is a factual question that must be answered by recourse to relevant evidence. [46] There is in principle no reason why the enquiry should not be conducted in reverse. The enquirer might choose first to identify the true character of the conduct that is the subject of the complaint, and only then turn to whether the conduct (so characterised) constitutes price-fixing as contemplated by s 4(1)(b). (This is how the enquiry is conducted in the United States, though there the two elements tend to be elided, because the scope of the prohibition is itself a matter of judicial rather than legislative determination.) [47] Whichever approach is adopted, the essential enquiry remains the same. It is to establish whether the character of the conduct complained of coincides with the character of the prohibited conduct: and this process necessarily embodies two elements. One is the scope of the prohibition: a matter of statutory construction. The other is the nature of the conduct complained of: this is a factual enquiry. In ordinary language this can be termed ‘characterising’ the conduct – the term used in the United States, which Ansac has adopted. [48] Price-fixing necessarily contemplates collusion in some form between competitors for the supply into the market of their respective goods with the design of eliminating competition in regard to price. That is achieved by the competitors collusively ‘fixing’ their respective prices in some form. (By setting uniform prices, or by establishing formulae or ratios for the calculation of prices, or by other means designed to avoid the effect of market competition on their prices.) [49] But while price fixing inevitably involves collusive or consensual price determination by competitors, it does not follow that price fixing has necessarily occurred whenever there is an arrangement between competitors that results in their goods reaching the market at a uniform price. The concept of ‘price fixing’, both in lay language and in the language that the Act uses, may, for example, be limited to collusive conduct by competitors that is designed to avoid competition, as opposed to conduct that merely has that incidental effect. [50] As the majority of the United States Supreme Court pointed out in Broadcast Music, Inc v Columbia Broadcasting System, Inc 441 US 1 (1978) at 9:

‘Literalness [when interpreting the phrase ‘price-fixing’] is overly simplistic and often overbroad. When two partners set the price of their goods or services they are literally “price fixing,” but they are not per se in violation of the Sherman Act...Thus, it is necessary to characterise the challenged conduct as falling within or without that category of behaviour to which we apply the label “per se price fixing.” That will often, but not always, be a simple matter.’

[51] What is important for the present proceedings is that the nature of the prohibiting source in this country – a legislative injunction against a certain form of conduct – makes it impossible to conclude the enquiry into whether particular conduct is prohibited without at some stage determining the scope of the legislative prohibition. And unless that determination is made, it is not possible to predict what evidence will be relevant or irrelevant to the factual part of the enquiry. [52] There can be little doubt that an agreement by competitors that has as its specific design the elimination of price competition (the essential characteristic of a cartel) [25] constitutes direct price-fixing as contemplated by the statute. Where competitors have reached an agreement to set uniform prices, without more, all that might be required in order to establish a transgression of s 4(1)(b) is to produce their agreement, because its very terms may admit of no conclusion but that it was designed to eliminate price-competition. [53] But indirect price-fixing presents greater complexity. It is not difficult to envisage conduct by competitors that is designed to eliminate price-competition indirectly, by shifting the supply of competitors’ goods to a separate entity that is under their control, and which purports to set the price for the goods. If that separate entity is no more than the alter ego of the individual competitors in association, who are in truth consensually fixing their prices through the medium of that alter ego, then no doubt the façade behind which they are acting can be stripped away to reveal the reality of the arrangement (collusion by two or more competitors designed to ensure that their respective goods reach the market at non-competing prices). [26] [54] But not every arrangement between competitors entailing the ultimate supply of goods necessarily falls into that category. It is, for instance, not difficult to envisage a bona fide joint venture that is embarked upon by competitors for a legitimate purpose, through the vehicle of a separate entity, which must necessarily set a price for goods that it supplies (emanating from the competitors) merely as an incident to the pursuit of the joint venture. [55] There is in our view no a priori reason to assume that such an arrangement constitutes prohibited price-fixing as contemplated by s 4(1)(b) of the statute. (We emphasise that we make no finding as to whether or not it is.) If, on a proper construction of s 4(1)(b), such an arrangement does not constitute prohibited price-fixing, then it might well be necessary to enquire beyond the mere terms of the competitors’ agreement in order to establish whether it is or is not merely a sham: to establish, in other words, whether the vehicle for the joint venture is in truth a single entity supplying its own goods to the market (albeit that the source of the goods is the competitors) for which a price must necessarily be set by the joint venture vehicle; or whether the vehicle for the joint venture is merely a cloak for what is in truth collusive action designed to ensure that the goods of competitors are supplied to the market at non-competitive prices. [56] What is critical to the present application is that the determination of what evidence is admissible depends on the scope of the legislative prohibition. Until there is clarity on what the legislation prohibits (and on what is not prohibited) it is premature to rule on what evidence might or might not be relevant and admissible to determine whether the prohibited conduct has occurred. [57] The parties are agreed that the Tribunal has yet to determine that issue. The Tribunal has not yet in express terms construed s 4(1)(b) and established its scope (nor what falls outside its scope). Nor is the scope of the prohibition in our view self-evident. The Competition Commission, in its submissions before us, has recognised some of the absurdities that would follow from a construction of s 4(1)(b) that prohibits all consensual conduct by competitors that ultimately produces a uniform price for goods emanating from them. The Commission has for this reason been constrained to read words into the statute to avoid the absurdities. [58] If the statute prohibits all consensual conduct amongst competitors that has the effect of creating uniform prices for their goods in the market, then the only evidence relevant to the enquiry is no doubt evidence that establishes the existence of a consensus having that effect. But if the prohibition is more restricted, then plainly the terms of the agreement alone might not be decisive. [59] We are not called upon in this application to give meaning to the prohibition and indeed it is not permissible for us to do so. The jurisdiction of this court, as we have pointed out, is confined to considering appeals, which contemplates the existence of an order or ruling by another court on the issue under appeal. [60] We do not suggest that the evidence Ansac seeks to lead is necessarily admissible. We hold only that it is premature at this stage to make a finding as to what evidence is or is not admissible, so long as the characteristics of the prohibited conduct have not been established by construing the statute. It is for the Tribunal to consider, in the manner and in accordance with such procedure as it may decide, to what extent evidence may be admissible to establish whether the Ansac agreement falls within the prohibition contained in s 4(1)(b). [61] To the extent that the Tribunal’s ruling is confined to precluding evidence purporting to justify conduct prohibited by s 4(1)(b) its finding is correct, and the CAC correctly dismissed the appeal against it. But to the extent that its ruling precludes evidence to ‘characterise’ the conduct in issue in order to determine whether or not the s 4(1)(b) prohibition covers that conduct at all, its ruling was premature and thus incorrect and is liable to be set aside. [62] Our findings that the CAC’s conclusions relating to jurisdiction and standing are unassailable clearly constitute a sufficient basis on which to refuse leave to appeal on those issues. Our finding relating to the remaining issue (the ruling on the admissibility of evidence), however, is of sufficient importance, both for the proper determination of the present dispute and for the future application of s 4(1)(b), to justify our intervention to correct the Tribunal’s findings insofar as it is necessary to do so. [63] As for the costs of the application, Ansac has failed in its contentions regarding the territorial application of the Act, and Botash’s legal standing. But it has had some success regarding what may be seen as the parties’ principal dispute – the admissibility of evidence to characterise the Ansac agreement and Ansac’s projected activities within South Africa. The parties are therefore invited to submit written argument as to what costs order would be most appropriate in this court and in the forums below. [64] There is a remaining observation. The present proceedings underline the need for care to be taken when isolating issues and dealing with them separately from the remaining issues. We repeat what was said by this court in Denel (Pty) Ltd v Vorster [27] in a related context:

‘[I]t is appropriate to make a few remarks about separating issues. Rule 33(4) of the Uniform Rules – which entitles a court to try issues separately in appropriate circumstances – is aimed at facilitating the convenient and expeditious disposal of litigation. It should not be assumed that that result is always achieved by separating the issues. In many cases, once properly considered, the issues will be found to be inextricably linked even though at first sight they might appear to be discrete. And even where the issues are discrete the expeditious disposal of the litigation is often best served by ventilating all the issues at one hearing, particularly where there is more than one issue that might be readily dispositive of the matter. It is only after careful thought has been given to the anticipated course of the litigation as a whole that it will be possible properly to determine whether it is convenient to try an issue separately. But where the trial court is satisfied that it is proper to make such an order – and in all cases it must be so satisfied before it does so – it is the duty of that court to ensure that the issues to be tried are clearly circumscribed in its order so as to avoid confusion... [A]nd when issuing its orders a trial court should ensure that the issues are circumscribed with clarity and precision.’

[65] We make the following order: 1. The application for special leave to appeal against the order of the Competition Appeal Court insofar as it dismissed the appeal against the findings of the Tribunal relating to jurisdiction and standing is refused. 2. The application for special leave to appeal against the order of the Competition Appeal Court insofar as it dismissed the appeal against the ruling of the Tribunal relating to the admissibility of evidence is granted. 3. The appeal succeeds to that extent and the order of the Competition Appeal Court is set aside. In its place there is substituted:

‘(a) The Tribunal’s ruling that evidence is not admissible to justify conduct falling within the prohibition contained in s 4(1)(b) stands. (b) The Tribunal’s ruling, to the extent that it excludes all evidence relating to the nature, purpose, and effect of the Ansac agreement, is set aside.’

4. The parties are invited to submit written argument as to the appropriate costs order, in this court and in the courts below.

E CAMERON & RW NUGENT

JUDGES OF APPEAL

CONCUR:

MPATI DP

CONRADIE JA

COMRIE AJA

[1] American Natural Soda Ash Corporation v Competition Commission 2003 (5) SA 633 (CAC) (Malan AJA, Davis JP and Jali JA concurring).[2] Sherman Act, 15 USC § 1: ‘Every contract, combination in the form of trust or otherwise, or conspiracy, in restraint of trade or commerce among the several States, or with foreign nations, is declared to be illegal. Every person who shall make any contract or engage in any combination or conspiracy hereby declared illegal shall be deemed guilty of a felony, and, on conviction thereof, shall be punished by fine not exceeding $10 000 000 if a corporation, or, if any other person, $350 000, or by imprisonment not exceeding three years, or by both said punishments, in the discretion of the court.’ § 2 makes monopolizing trade a felony.[3] Section 3 of Act 39 of 2000 amended the Act by moving the italicised words, which had been in sub-para (a), to the end of the opening portion of the provision. When the proceedings commenced the Act was in its unamended form, but nothing turns on this and we give the post-2000 wording.[4] American Natural Soda Ash Corporation v Competition Commission 2003 (5) SA 633 (CAC); American Natural Soda Ash Corporation v Competition Commission 2003 (5) SA 655 (SCA).[5] Section 8 of the Act prohibits ‘abuse of dominance’, making it unlawful for a dominant firm (defined in s 7) to charge an excessive price to the detriment of customers, and related exclusionary acts. [6] In terms of s 21(1)(g) of the Act the Competition Commission may refer matters to the Competition Tribunal and appear before the Tribunal.[7] American Natural Soda Ash Corporation v Competition Commission 2003 (5) SA 633 (CAC).[8] American Natural Soda Ash Corporation v Competition Commission 2003 (5) SA 655 (SCA).[9] Supreme Court Act 59 of 1959 s 21(3)(c)(ii). [10] See the judgment of this court in National Union of Metalworkers of South Africa v Fry’s Metals (Pty) Ltd (case no 026/03, delivered on Tuesday 12 April 2005) paras 9-33.[11] Constitution s 167(3)(b): CC has jurisdiction ‘only’ in ‘constitutional matters, and issues connected with decisions on constitutional matters’.[12] Judgment of this court dated Tuesday 12 April 2005, paras 5-33.[13] See National Coalition for Gay and Lesbian Equality v Minister of Home Affairs 2000 (2) SA 1 (CC) paras 23-24 and particularly National Director of Public Prosecutions and another v Mohamed NO [2002] ZACC 9; 2002 (4) SA 843 (CC) para 33.[14] 2003 (5) SA 655 (SCA) para 16.[15] Judgment of 12 April 2005, paras 34-44.[16] Constitution s 173: ‘The Constitutional Court, the Supreme Court of Appeal and the high courts have the inherent power to protect and regulate their own process, and to develop the common law, taking into account the interests of justice’.[17] They are set out in the Supreme Court Act 59 of 1959 and in the decisions of this court, including Westinghouse Brake & Equipment (Pty) Ltd v Bilger Engineering (Pty) Ltd 1986 (2) SA 555 (A) 564H-565E.[18] 2003 (5) SA 633 (CAC) paras 7-21.[19] Rennie NO v Gordon NO 1988 (1) SA 1 (A) 22F, per Corbett CJ, adopted in Bernstein v Bester [1996] ZACC 2; 1996 (2) SA 751 (CC) para 105, per Ackermann J.[20] 2003 (5) SA 633 (CAC) para 18.[21] 2003 (5) SA 633 (CAC) paras 2-5.[22] We refer, since it makes no material difference, to the provisions of the Act as amended by Act 39 of 2000.[23] 2003 (5) SA 633 (CAC) para 4.[24] Quoted in note 2 above.[25] Concise Oxford English Dictionary: ‘an association of manufacturers or suppliers formed to maintain high prices and restrict competition’.[26] Compare 91/301/EEC: Commission Decision of 19 December 1990, relating to a proceeding under Art 85(1) of the EEC Treaty (IV/33.016 – Ansac), Official Journal L 152, 15/06/1001 p 0054 – 0060.[27] 2004 (4) SA 481 (SCA) para 3, per Nugent JA.