South Africa: Competition Tribunal

You are here:
SAFLII >>
Databases >>
South Africa: Competition Tribunal >>
2011 >>
[2011] ZACT 73
| Noteup
| LawCite
South African Breweries and Others v Competition Commision (134/CR/DEC07) [2011] ZACT 73; [2011] 2 CPLR 403 (CT) (16 September 2011)
Download original files |
BEFORE THE COMPETITION TRIBUNAL OF SOUTH AFRICA
CT CASE NO: 134/CR/DEC07
SOUTH AFRICAN BREWERIES LIMITED …......................................................First Applicant
APPOINTED DISTRIBUTORS …..................................................................Second Applicants
(2nd -14th Respondents in the main matter)
and
COMPETITION COMMISSION ….............................................................................Respondent
In the matter between:
COMPETITION COMMISSION …...................................................................................Applicant
and
SOUTH AFRICAN BREWERIES LIMITED ….......................................................1st Respondent
AFRICA’S BEER WHOLESALERS (PTY) LTD …..............................................2nd Respondent
BOLAND BEER DISTRIBUTORS (PTY) LTD …..................................................3rd Respondent
ERMELO BEER WHOLESALES (PTY) LTD …....................................................4th Respondent
GREYTOWN BEER DISTRIBUTORS (PTY) LTD …............................................5th Respondent
MAKHADO BEER WHOLESALERS (PTY) LTD …..............................................6th Respondent
MIDLANDS BEER DISTRIBUTORS (PTY) LTD …..............................................7th Respondent
MKUZE BEER WHOLESALERS (PTY) LTD …...................................................8th Respondent
SOUTHERN CAPE BEER DISTRIBUTORS (PTY) LTD …..................................9th Respondent
STEFQUO (PTY) LTD ….....................................................................................10th Respondent
VRYHEID BEER DISTRIBUTORS (PTY) LTD …...............................................11th Respondent
MADADENI BEER WHOLESALERS (PTY) LTD …...........................................12th Respondent
WESTONARIA BEER DISTRIBUTORS (PTY) LTD ….......................................13th Respondent
THOHOYANDOU BEER DISTRIBUTORS (PTY) LTD …...................................14th Respondent
Panel :Norman Manoim (Presiding Member), Yasmin Carrim (Tribunal
Member), and Merle Holden (Tribunal Member)
Heard on : 4-6 April 2011
Decided on : 7 April 2011
Reasons issued on : 16 September 2011
Reasons: Dismissal Application
______________________ _________________________________________________
Introduction
In this application the respondents seek dismissal of the complaint referral on grounds of lack of jurisdiction, because the complaint referral is not founded on the complaint it purports to be based on.
The first respondent, South African Breweries Limited (‘SAB’) and, collectively, the second to fourteenth respondents, the latter whom, for convenience, we will refer to as the ‘appointed distributors’, have each brought a separate application for dismissal. As they traverse the same points of law it is convenient to deal with them in the same set of reasons.
We have found for the respondents and gave an order to this effect on 7 April 2011. Although we gave a short statement at the time of granting the order, we indicated that the reasons would follow later. In this decision in Part A, we set out firstly our reasons for coming to that conclusion, based on existing jurisprudence from higher courts that is binding on us. In Part B, we indicate our concerns as to why this jurisprudence may need re-consideration to avoid injustice to complainants and the public interest as represented by the Competition Commission (“the Commission”).
PART A
Background
On 25 November 2004 the complainants in this case filed a complaint with the Competition Commission.
The complainants are a group of companies, comprising both retail and wholesale liquor operations, located primarily in the Eastern Cape. The complainants are led by a group of companies that comprises wholesale and liquor outlets known as the Big Daddy’s Group.1 The head of the group, Nicholas Peter Pitsiladi (“Pitsiladi”) is the signatory to the document that sets out the complaint. Pitsiladi states that he has been authorised to represent all the complainant firms.2
The complaint relates to the complainants’ wholesale businesses. In the liquor industry the role of wholesalers has been to act as the middleman between manufacturers and outlets that sell to the public, be they retailers, hotels, restaurants and taverns.
Wholesalers typically receive liquor consignments in bulk from manufacturers and then deliver liquor in smaller quantities to their customer base. Whilst manufacturers have a national footprint, wholesalers usually deliver products to a more confined geographical area.
Historically, wholesalers received a discount on the retail price from the manufacturer and they would use this discount to fund the costs of their distribution business, whilst keeping some of the margin for themselves as profit.
What triggered the complaint were events in 2004 following the introduction of the new three tier licence system under the new Liquor Act 59 of 2003. Under the new system, according to the complainants, the position of traditional wholesalers changed, as manufacturers could now get distribution as well as manufacturing licences. Although not entirely clear from the complaint, it seems that SAB stopped offering a discount to wholesalers and sold its products to them at the same price as it sold to the retail segment. Since retailers were the wholesalers’ customer base, they could hardly sell to them at a higher price than they would get if they bought directly from SAB. This, the complainants felt, would spell the death knell for wholesalers as SAB was a dominant supplier of liquor.
The appointed distributors, the remaining respondents in this case, also serve a distribution function. Unlike other firms performing this function as wholesalers, the appointed distributors operate in terms of a different business model. They receive a fee for distribution services, unlike the wholesalers, but are also restricted to certain territories in which they are obliged to operate and to serve any customer who placed an order with them, are restricted to only distributing SAB products and are subject to certain performance standards. They all operate in terms of a contract with SAB, which whilst not uniform throughout, contains the same standard features outlined above.
A central part of the dispute in the main case was whether the appointed distributors and the wholesalers performed an equivalent function. We do not need to decide that dispute for the purpose of this decision, but this background is necessary for an understanding of the issues in the dismissal application.
Prior to lodging the complaint with the Commission, Pitsiladi had an exchange of correspondence with SAB in which he expressed his concern about the elimination of the discount and threatened to take up the matter with the Competition Commission. SAB was adamant it was acting within the law and refused to accede to the demands.3
Pitsiladi, dissatisfied with this response, duly lodged the complaint on 25 November 2004. The complaint consists of a covering form CC1 – the standard form complainants are required to complete in terms of the Commission’s rules - and an annexure ‘B’ to it, which sets out the description of the complaint.4 In paragraph 11 of this annexure, reference is made to correspondence alleged to have accompanied the complaint:
“In this regard, annexed hereto are examples of a letter dated 21 September 2004, addressed to SAB and a letter dated October 2004, sent by SAB to Mr Pitsiladi.”
It is common cause that the referral is based on this complaint and that there was no initiation by the Commission or further initiation by the complainants that preceded the referral.
Regrettably the ambit of the content of the complaint is not common cause, as there is a dispute about whether the letters the complainants purport to attach to the description of the complaint, referred to in the paragraph quoted above, were indeed attached, with the version filed with the Commission. To fuel the confusion further it appears that Pitsiladi has confused the dates of the letters he refers to in paragraph 11.5
The controversy about what constitutes the complaint became apparent only in the dismissal application proceedings. When the applicants brought their respective applications they attached the complaint i.e. the form CC1 and Annexure B, but not the correspondence supposedly annexed to it. The Commission in its answering affidavit queried why the correspondence had not been annexed. SAB alleged in its reply that if letters had been attached to the complaint, the Commission had not supplied them to it (SAB) when it made discovery. It had attached the discovery version of the complaint- the one sans annexures - to its founding affidavit. SAB then requested the Commission to furnish it with the letters.
The facts get messy hereafter. It is common cause that the Commission no longer has a version of the complaint with the letters annexed to it. It is not clear if it ever had it in this form.6 This is because due to the passage of time, none of those responsible for the Commission’s investigation were still in its employ at the time this application was heard nor could the letters be found on its electronic filing system and thus neither a human or electronic recollection exists to explain whether the correspondence was ever received with the complaint. When challenged by SAB to prove it possessed them at the relevant time, it was unable to do so.
The Commission relied instead on an affidavit from Pitsiladi, who stated that to the best of his knowledge the letters had been supplied at the time. But he too could not supply records from his side that could establish this fact conclusively.
Why is the issue of whether the letters were attached to the complaint such an important issue in the contestation, given that we have copies of the correspondence, since SAB received them at the relevant time and discovered them? Largely because in one letter Pitsiladi sets out his firm’s case in respect of the price discrimination, as between itself and the appointed distributors, in the form eventually taken up by the Commission in its referral. But this formulation is not followed in the body of the complaint i.e. Annexure B, for reasons that are not clear. The case law, which we discuss later, requires the complaint to be a unitary document. Thus, whether the letters were attached to it at the relevant time, matters.
While it is common cause that the Commission does not have the contested correspondence presently in its possession and probably did not have it when the time came to produce documents for discovery, it does not follow that they were never filed with the complaint originally in 2004. It is clear from a reading of the complaint that the complainants intended to annex the letters. It is also clear from the covering letter that the complainants’ attorneys wrote to the Commission that the faxed copy would be followed up with an original which was to be delivered. Thus to conclude that the letters were never attached, we have to assume that there was negligence on their attorneys’ behalf in both despatching a faxed and subsequently a hard copy version and secondly, that the Commission, if it had received both versions of the complaint without the letters, despite the express reference to their existence in Annexure B, was negligent in not requesting them.
This requires too many assumptions to be made and so we find that it is probable that the letters did accompany at least one version of the original complaint. As long as one version dispatched, contained the contested letters, that suffices to make them part of the complaint.
Nevertheless even if the contested correspondence is assumed on a balance of probabilities to have been annexed to the complaint, the case would still fall to be dismissed. We explain why this is so later, based on the existing case law in respect of complaints, but first it is necessary to consider this jurisprudence to understand how it has led us to this conclusion.
Legal Framework and Case law
Before a prohibited practice reaches the Tribunal for a hearing it goes through three stages. First, it has to be the subject of a complaint brought at the instance of a complainant or the Commission, secondly it has to be investigated by the Commission and thirdly, it has to be referred to the Tribunal at the instance of either the Commission or the complainant.
This process involves two documents which have juristic consequences; the complaint initiation, which to avoid confusion, because of the similarity of their names, we shall refer to from now on as the ‘initiation document’, and the complaint referral, which we shall refer to from now on simply as the ‘referral’.7
Their juristic distinction is important as we discuss more fully below. For now it suffices to state that the initiation document commences the investigation process of the Commission into a complaint against a prohibited practice, the referral, which comes later, commences the litigation process before the Tribunal.
The crucial issue in this case is the relationship between the initiation document filed by the complainants with the Commission and the referral made by the Commission commencing the current proceedings in the Tribunal. The case law of the superior courts requires the case made out in the referral to be founded on a prior initiating document or series of initiating documents.8 (Recall that in this case there has been no separate referral by the Commission. The only initiating document before us is that of the complainants.) To the extent that a referral contains allegations not founded in the initiation document preceding it, it is jurisdictionally incompetent. The courts have approached this issue very strictly as we shall see.
The first case in which these issues came up before the Competition Appeal Court (‘CAC’) was Glaxo.9 In that case the complainant had referred its own complaint after the Commission had elected not to refer it. The issue was whether all the allegations contained in its referral had been made in the initiation document filed with the Commission. The respondents had argued that certain prohibited practices alleged to have been perpetrated by them in the referral, were not contained in the initiation document and for that reason should be struck out.
The case is significant for a number of reasons. In the first place the court held that the initiation document is “... the jurisdictional fact or precondition which must be satisfied before the Tribunal can exercise powers over a respondent.”10 The court held that in the absence of particulars relating to such conduct to the Commission, “... any exercise of power by the Tribunal in relation to such prohibited practice will be invalid.”11
The crucial passage for purposes of this decision then follows:
“The proper approach is to determine first what conduct is alleged in the complaint and what prohibited practices such conduct may be said to invoke or be rationally connected to. Then consideration is given to the referral to see whether the conduct alleged is substantially the same.”12
How does one determine whether conduct alleged is ‘substantially the same’? For most cases this will probably not present a difficulty, but at the margin it may present great difficulties, particularly when, as in the present matter, the complainant complains, but the Commission refers. Is a complaint that a dominant manufacturer has denied me a previously enjoyed wholesaling discount, substantially the same as the case where on referral, the issue is the denial of the discount for some classes of distributors, but not others?
In Glaxo the court grappled with this problem by attempting to harmonise two conflicting notions. On the one hand, sensitive to the need to be fair to complainants, the court held complainants are not required to:
“... pigeonhole the conduct complained of with reference to particular sections of the Act. What is required is a statement or a description of prohibited conduct.... The complainant need only identify the conduct of which it complained.”13
It noted that section 49B is not prescriptive about how a complaint may be initiated and that this then ran throughout the complaint procedures, “... the object being to enable complaints to be lodged without the need for procedures that are too technical and/or formalistic.”14
But having made these concessions to informality the court nevertheless held that in stating the conduct (as opposed to pigeonholing it to sections of the Act):
“There must be a rational or recognisable link between the conduct referred to in the complaint and the prohibitions in the Act, otherwise it will not be possible to say what the complaint is about and what should be investigated.”15
So what then is the link in this chain between the non-pigeonholed conduct, but conduct rationally or recognisably linked to the prohibited practice contained in the referral? The court answered this by resorting to well known concepts of civil law and said it was the “...facta probanda necessary to establish a prohibited practice.”16
Applied to the facts of this case what the court seems to be saying is that to found the Commission’s present section 9(1) claim, it was not necessary for the complainants to allege that SAB had contravened section 9(1) of the Act. It was however necessary for the initiating document to contain the facta probanda of price discrimination. Thus it would have been necessary, following Glaxo, for them to say, at a minimum, that; SAB was a dominant supplier of alcohol; the complainants were supplied by it at a price which was discriminatory, as some purchasers, the appointed distributors got preferential discounts denied to them; the supply transactions to them and the appointed distributors were equivalent; and that there was harm to competition as a result of the discrimination.
Quite how formally this approach has been applied emerges from the facts in Glaxo. Both the court and the Tribunal agreed that the claims made out in the referral for excessive and predatory pricing were not mentioned in the initiating document and therefore could not be referred. They came to a different conclusion on a third claim.
The complainants had referred a claim that the respondents had refused them access to an essential facility, (section 8(b)). In their initiating document this was confined to a refusal to supply (section 8(d)(iii)). The respondents had argued that it was not competent to plead the access to essential facility claim in the referral, when the initiating document was confined to an allegation of a refusal to deal.
The Tribunal held that this reference to a refusal to deal in the initiating document sufficed jurisdictionally for a claim of denial of access to an essential facility to be made in the referral. The Tribunal held that the two forms of conduct were “...often two sides of the same coin”.17 The court disagreed with this conclusion. Applying its facta probanda approach it held that the two concepts were not “conceptually similar”. What this illustrates is that despite the gesture to informality in dispensing with pigeonholing, the court still applied a very strict standard to the concept of facta probanda.
The case law in this area did not develop further until it emerged, belatedly, as an issue in the Woodlands case which was considered by the Tribunal, CAC and the SCA.18 The case concerned the validity of two summonses issued by the Commission pursuant to an investigation of firms in the milk industry. Two firms involved in milk processing had received summonses from the Commission in terms of section 49A of the Act. Both summonses required executives of the firms to appear for interrogation, whilst the one firm was required to produce documents as well. Neither firm had been mentioned in the Commission’s initiation document that had preceded the summonses. The two firms later became respondents in a complaint referral that also implicated a number of other processors in the industry inter alia in cartel activities. The two respondents advanced a number of challenges both to the validity of the summons and the referral. The Tribunal had ruled that both summonses were invalid and for that reason did not pronounce on the validity of the referral; the CAC on appeal held that the one summons was valid, but the other not.
The CAC did not address the validity of the referral in its decision. However in the application for leave to appeal it did.
The CAC upheld its validity, addressing in the main, the contention that the referral was not jurisdictionally competent, as although a cartel in the milk industry had been alleged the identities of all the members had not been disclosed in the initiating document. The CAC made the point that in cartel cases the identity of all the members may be hidden at the time of the complaint initiation and hence a complaint into an industry as a whole was competent.
The matter went on appeal to the SCA. The SCA upheld the appeal, finding that the initiating document was invalid, and in doing so made several findings that became crucial to informing the approach of the CAC in subsequent matters where the validity of referrals were challenged. They are also central to the respondents’ arguments in favour of dismissal in this matter.
The SCA decision approached the validity of the initiation document in two respects; one is more general and the other more specific. In asserting the more general principle the SCA held that as the actions of the Commission in respect of prohibited practice cases may lead to punitive measures including administrative penalties, which the court characterised as bearing a close resemblance to criminal penalties, its powers must be interpreted in a way that least impinges on a firm’s constitutional rights to privacy, a fair trial and just administrative action.19
More specifically the court held that there was no reason why an initiating document requires any less specificity than a summons. Here the court is clearly referring to a summons in terms of section 49A, i.e. one to compel documents from a person or to require them to be interrogated.20
It then held that if this was the case: “It must survive the test of legality and intelligibility.” The court explained that the reason for this is that the scope of a summons may not be wider than the initiation:
“The Act presupposes that the complaint [initiating document] (subject to possible amendment and fleshing out) as initiated will be referred to the tribunal. It could hardly be argued that the commission could have referred an investigation into anticompetitive behaviour in the milk industry at all levels to the tribunal.”
The court went on to say that this did not mean that the Commission could not, subsequent to a validly initiated investigation, obtain information about other transgressions, but that if it did, it was fully entitled to use the information for the purpose of amending the complaint or the initiation of another complaint and fuller investigation.
The first case in which the CAC had to apply the approach of the SCA in Woodlands was the Netstar case.21 Although the validity of the scope of the initiating document does not seem to have been in issue in Netstar, the court nevertheless opined on it.
Here the court adopted the strictest position thus far on the relationship between the initiating document and the referral, holding that if the original ground for complaint had related to a ‘prohibited agreement’ in terms of section 4(1), the Tribunal could not determine it on the basis that there was a ‘concerted practice’ or vice versa. The language of section 4(1) refers to a restricted horizontal practice as being: “An agreement between, or concerted practice by...”
The court referred to Woodlands and its likening of an initiating document to a summons. However in Netstar the court seemed to understand the reference to be that of a civil summons commencing proceedings, because it went on to say:
“This is not to say what is required of a complaint is the level of precision that is demanded in pleadings. Where the complainant is a lay person that would be to demand more than can reasonably be expected. What is required is that the conduct said to contravene the Act must be expressed with sufficient clarity for the party against whom that allegation is made to know what the charge is and be able to meet and rebut them.” (Our emphasis)22
The court went on to say that even though the Tribunal adjudicates over issues which are more general and less clear cut than in a civil case this did not mean that:
“...broad and unspecified generalities should take the place of a properly articulated complaint before the Tribunal to which the target of the complaint can respond. ...There is no excuse for resorting to vague generalities instead of formulating complaints accurately and in detail in accordance with the provisions of the Act.” 23
The difficulty with this paragraph is that although commencing with discussion of the requirements of an initiating document it seems to elide that with the requirements for a referral by its reference to a ‘properly articulated complaint before the Tribunal to which the target of the complaint can respond’.
This elision is furthered in the Yara case as we go on to consider. In Yara the procedural issues were similar to those in the present matter. A private complainant had initiated a complaint alleging that Sasol, a dominant firm, had abused its dominant position as a supplier of certain chemical products, alternatively contravened section 9(1). The Commission then referred the complaint on this basis. Sometime after the complaint had been referred, the Commission sought to amend its referral, by including two additional respondents and alleging that the two respondents and Sasol had contravened section 4(1)(b) of the Act.
The respondents objected, alleging that the amendment sought to introduce material not covered by the original complaint. The original complaint they contended was limited to Sasol allegedly contravening sections 8 and 9. But the matter was more complicated than this. The complainant had in an affidavit annexed to its initiating document alleged that the three firms were part of a cartel. The court held that the complainant had intended to submit information relating to the three firms in its affidavit but did not agree that these allegations were meant to constitute a:
“... distinct complaint in the sense of a separate cause of action within the complaint; as opposed to further information concerning the initial complaint.24
Later the court said:
“The allegations of cartel activity are bald and not supported by any detail. I can only conclude therefore that the allegations fall under information submitted under section 49(B)(2)(a) of the Act.”25
The CAC held that the Commission could not amend an initiating document as there was no procedure in the Act providing for this.26 Thus contrary to what the SCA seemed to hold in Woodlands, an initiating document, in terms of this decision, cannot be amended.
This finding is directly relevant to our present case because one of the arguments relied on by the Commission is that if the contested correspondence does not form part of the initial complaint, then it could be construed as amending it. From Yara it appears this argument cannot prevail.27 If the initiating document is not capable of amendment then no subsequent correspondence from the complainant could be of any relevance in widening the complaint.
The next implication of this decision pertinent to the present case is the court’s emphasis on the intention of the complainant. The Act distinguishes between situations when members of the public make submissions to the Commission as complainants (section 49B(2)(b)), and when they make submissions of ‘information’ merely as informants (section 49B(2)(a)). The distinction matters because the complainant has certain rights and its submission has a certain juristic status – it is an initiating document, whereas the informant has no rights nor does its submission have any juristic status.28 How then does one tell if a submission comes from a complainant or an informant?
This problem first arose in Clover, a case considered by the CAC prior to Yara.29 In Clover the question was whether a letter written to the Commission by a member of the public constituted a complaint or the submission of information. The distinction mattered as if it was the complaint, the person submitting it was the complainant and not the Commission, and accordingly, the time period for referring the complaint had prescribed. The court held, upholding the earlier Tribunal decision, that the person in question was not a complainant as she had no intention of being a complainant. The conclusion that she lacked the intention to be a complainant was based on a detailed analysis of her letter as well as the circumstances under which it was received by the Commission.
In Yara - although making no reference to Clover - the intention test was applied again, when the court had to decide whether the submissions regarding alleged collusion in the initiation document constituted merely ‘information’ submitted or were part of the complaint.
The court, as we explained earlier, decided they were not part of the complaint based on their interpretation of the complainant’s intention. But in this respect, in applying the intention test, the court was going further than it had in Clover. Whilst Clover used the intention test to decide identity; i.e. whether the person making the submission was a complainant or an informant, Yara applied the intention test to the content of what was common cause to the complainant’s submission, to determine in what respects parts of the document could be classified as constituting a complaint and what parts could be regarded as mere submissions of information. The former constituted part of an initiating document, the latter, since only information, did not.
As the court expressed it:
“It is not the absence of an express mention of section 4(1) that is the relevant point but a clear absence of any intention on Nutri-flo’s part to be a complainant in respect of a price fixing complaint.”30
The court then went on to examine what the purpose of the CC 1 form that is used for lodging initiation documents, is:
“In competition cases, the parties look to the CC1 Form for details of the complaint(s) against them. Therefore, if it appears in the CC1 Form together with accompanying statements, where relevant, that no complaint lies against a particular party, such a party may assume that it is not a true party to the proceedings. It is therefore improper to bring such a party within the ambit of the complaint by way of either a referral or an amendment thereto.”31
The Yara decision thus represents the apogee of the strict approach to the relationship between the initiation document and the referral. It makes two important additions to the post-Woodlands case law. The Commission may only refer that part of the submission from the complainant that it intended to complain about and not those that constitute mere information, and secondly, it offers a further rationale for the strict approach; because the initiating document is what respondents look to for details of the complaint against them, it is improper to bring a party before the Tribunal by way of a referral if the details of the complaint in the referral are not found in the initiating document.
We now apply this case law to the facts of this case.
Is the complaint contemplated by the complainants, the complaint that has been referred by the Commission?
Before we can decide whether the complaint contemplated by the complainants in the initiating document is the one referred by the Commission, we have first to decide what the parameters of their initiating documents are. This is a matter of both factual and legal dispute. Only once the dispute around the parameters has been determined, can we go onto the next step which is to perform the comparative analysis.
Parameters of the initiating document
SAB argues that to identify the complaint we must look no further than the description of the complaint that appears as annexure B to the initiation document. The Commission argued that the complaint is wider than Annexure B and that in this regard we can have reference to; (i) correspondence that was annexed to the complaint and (ii) correspondence subsequent to the filing of the complaint.
The reason for this debate is that if Annexure B is read in isolation from this correspondence, it does not clearly contain support for the practices that the Commission has referred. If we have regard to the correspondence – that subsequent to the complaint and that annexed to it in the disputed correspondence, arguably there is a foundation for it.
The Yara case is the authority for deciding both these issues. On the authority of Yara the subsequent correspondence, i.e. the correspondence which was sent after submission of the initiating document, was considered not part of the complaint. The complaint is not a process – it is a singular event – it is the CC1 and the documents annexed to it containing the statement of facts. If the complainant wishes to expand on a complaint already submitted, as we understand Yara, it must file another complaint.
If a complainant submits further documentation to the Commission, after filing the initiating document, this would be considered as information submitted, not a complaint. Yara, as we noted earlier, holds that the Act provides for no procedure for the amendment of an initiating document.32
In this case to the extent that subsequent correspondence may be read to amend Annexure B, we find that such amendments are not competent and cannot amend the initiating document. We are thus confined to considering Annexure B and the annexed correspondence, in order to consider the parameters of the complaint.
Since the annexed correspondence is only relevant to found the section 9 price discrimination count, we can conveniently consider the other counts first, as the Commission does not seek to rely on the annexed correspondence for their foundation.
The claims made out under sections 4(1) and 5(1) of the Act are based on the same facts and the latter is pleaded as an alternative to the former. The Commission alleges that the agreements between SAB and the appointed distributors, which carve out exclusive territories for the latter, amount to either a horizontal agreement between competitors not to compete, because SAB is seen as a competitor of the appointed distributors in the market for the distribution of liquor, or in the alternative, a series of vertical agreements between SAB ‘qua manufacturer’ and the appointed distributors ‘qua distributors’ in which exclusive territories are carved out, thus lessening intra-brand competition in SAB products.33
SAB argues, and we agree correctly, that there is no mention of either of these counts in annexure B. Granted in a concluding paragraph the complainant says if the main complaint does not constitute a contravention of section 8, then the same conduct may contravene sections 4 and/or 5. But this does not help the Commission. Mere recitation of the sections does not make out a case which is not there, just as the absence of their mention would not negate a case that was made out.
The initiating document neither expressly nor implicitly refers to the contracts SAB has with appointed distributors, nor to their exclusive nature, which is the nub of the section 4(1) and 5(1) cases in the referral. Put another way, although the initiating document could be read to describe a vertical foreclosure case, it is not the one the Commission advances. Pitsiladi is concerned that the elimination of the wholesale discount will destroy the viability of the wholesale business and arguably, although he does not say this expressly, lead to a reduction of intra-brand competition in the sale of SAB products. The Commission does not focus on the elimination of the discount for its vertical case; it is concerned with the nature of the vertical effects of territorial exclusivity, although it does regard the elimination of the discount as a factor aggravating the weakness of intra-brand competition, but this concern is incidental and not fundamental to its case.34
Nor are the appointed distributors mentioned on the CC1 as parties against whom the complaint has been laid, which as we have seen from the case discussion above is a requirement.35 Although the theory of harm that the Commission wishes to advance in respect of these two claims emerges tangentially from the factual milieu provided in the complaint, it is not founded in the complaint if we apply a facta probanda approach nor if we apply an intentional approach.
There is therefore no basis in Annexure B for these two counts. The same can be said for the retail price maintenance claim. There is no mention of this in annexure B.
This leaves the price discrimination count in terms of section 9(1). This has been the main focus of the Commission’s evidence during the course of our aborted hearing.
The case for dismissing the price discrimination count is not as clear cut as with the other three. This is so for two reasons. The language in Annexure B is opaque and some formulations may permit for different constructions. Further, at least one letter in the annexed correspondence does contemplate the factual issues that later gave rise to the price discrimination case – hence SAB’s vigorous challenge as to whether it was indeed annexed.
The SAB gloss on Annexure B is that SAB is trying to eradicate wholesaling in its products and to take this market over for itself. Whilst this is a self-serving reading, it is not inaccurate. In prefatory paragraphs Pitsiladi has explained how the new licensing regime enables SAB the manufacturer to expand to become SAB the wholesaler. In the key paragraph the allegation is that SAB has decided to do away with all other wholesalers in the country and to: “eventually control themselves the entire wholesale division of selling beer within South Africa.”
On this scenario, the case is not about the discrimination in favour of the appointed distributors, but the elimination of wholesaling in its products. In paragraph 10 of Annexure B, Pitsiladi goes on to refer to the sale of beer by SAB the manufacturer to “SAB the wholesaler”. The phrase is open to contending interpretations. Does he mean to restrict this to SAB in its wholesaling function or incorporate the appointed distributors into this notion? Unfortunately he is not clear on this point and ordinary interpretation would suggest what he has in mind is SAB performing a distribution function in respect of its products. It is common cause that SAB distributes most of its own products through its depots, so the ordinary interpretation does not lead to absurdity.
But there is also textual support from his own usage of terminology that Pitsiladi does not contemplate the appointed distributors when he refers to ‘SAB the wholesaler’ because in other places he uses different terminology to describe the former. In one of the letters annexed to the referral he refers pertinently to the 9th respondent, which is one of the appointed distributors and he describes the firm as one of the ‘existing independent distributors”.36 Given that this letter precedes the complaint, it is likely, that had he meant to refer to the appointed distributors, he would have used the term that he used in his prior correspondence i.e. ‘independent distributors’.
It would be fair to assume from this that he is aware of the distinction between the appointed distributors and the SAB distribution depots and that he does not apply the phrase, ‘SAB the wholesaler’ to both indiscriminately. We can further assume that he uses this terminology consistently. If this is so, then the reference in Annexure B to ‘SAB the wholesaler’ must be taken not to be a reference to the appointed distributors and is thus not a source or foundational fact for the price discrimination case in the text of Annexure B.
That takes us to the next question. If the text of Annexure B cannot found the price discrimination claim, can the correspondence annexed to Annexure B? The fact that the allegations may be contained in a letter attached to the complaint and not in the complaint itself does not matter. In Yara the court accepted that an annexed affidavit could form part of the complaint.37 Furthermore, as we noted earlier, the Form CC1 requests the complainant to attach relevant correspondence to it.38 Therefore the fact that a complaint may be contained in a letter attached to the summary of the complaint is not decisive in concluding that it does not form part of the complaint. We therefore need to examine the content of the letter to see if it is intended to be part of the complaint.
We examine again the undated letter from Pitsiladi to SAB that we referred to earlier. Note that although undated, it is likely from dates on the other correspondence to assume that this letter preceded the filing of the initiating document by about a month. We set out below the relevant paragraphs from it:
“Since our discussion I have sought legal advice and it is our company’s contention that SAB is not entitled to dominate the market to the extent of supplying its product to its own distributor divisions/wholesalers with its concomitant special benefits and then on-selling to the other distributors/wholesalers at the same price that the SAB distributor wholesaler is selling to retailers. We also contend that SAB is not entitled to favour certain distributors/ wholesalers with special prices to enable them to supply at the SAB retail price list but refuse those same prices to other independent distributors/wholesalers.”39 ( Our emphasis)
A textual analysis of this paragraph reveals the following:
Pitsiladi first complains about the denial of the wholesaler’s discount ( ‘the first complaint’);
Pitsiladi also complains about price discrimination in favour of the appointed distributors (‘the second complaint’) (Underlined extract in the quote);
he regards them as discrete issues. We say this as he prefaces the second complaint with the phrase “We also contend.....40; and the first complaint is repeated in Annexure B, but not the second.
The question then is why has Pitsiladi carried over only the first complaint and not second complaint into annexure B, written only a month later? Was this intentional or negligence or simply poor drafting? SAB suggests that because it had written a letter in answer to this letter, which preceded the initiating document, and in which the allegations of price discrimination were refuted, it was clear that this explanation satisfied Pitsiladi, at least at the relevant time, and hence he did not intend to complain about price discrimination in Annexure B.41
We are not in a position to speculate on whether this is so or not. But we do not need to know why it was omitted to interpret the initiating document; the fact is it was. Given that the first complaint (denial of the discount) was carried forward to the initiating document, but not the second ( price discrimination), we can infer, applying the intention approach that for whatever reason, the complainants as represented by Pitsiladi, did not intend to complain about price discrimination in the text of the complaint as set out in Annexure B.
This does not resolve the issue entirely and another issue remains. Why then was the letter attached at all; if that is left of the complaint was repeated in Annexure B, surely this would render its attachment superfluous? SAB answers this by asserting that the letters are annexed as proof of only one contention, namely that set out in paragraph 11 of Annexure B, where Pitsiladi asserts that despite his prior correspondence with SAB, it had declined to desist in what he termed “its anticompetitive actions”. SAB argues that in the context of what is stated in paragraph 11, the reason the letter is attached is to show that Pitsiladi had attempted to resolve the complaint, but despite his request, SAB persisted in its conduct. It is not intended as an elaboration of the complaint set out in Annexure B. It serves only this limited purpose.
Applying the intention test to the language, this argument is a reasonable one. We therefore find that Pitsiladi, and thus the complainants, had set out their full complaint in annexure B, and that the attached correspondence was not meant to expand its ambit, rather to illustrate that they had tried to resolve the complaint but unsuccessfully. On this basis there is no clear intention to persist with the price discrimination complaint to the Commission.
We therefore find that Annexure B contains the ambit of the complaint and that none of the counts contained in the present referral can be found in it. The complaint made about SAB in the initiating document has not been the one referred by the Commission. Further, that the initiating document does not contemplate any other firm other than SAB, in the limited extent mentioned, as a respondent and therefore, on that ground alone, there is no case made out against any of the appointed distributors.
Section 50(3) argument
We must now consider an alternative argument advanced by the Commission. The Commission argues that even if we cannot identify the foundation of the present referral in the initiating document, it is nevertheless entitled to add particulars to the complaint without having to initiate its own complaint.
In this regard the Commission relies on section 50(3)(iii) which states:
“50(3) When the Competition Commission refers a complaint to the Competition Tribunal in terms of subsection (2) (a), it-
may-
refer all the particulars of the complaint as submitted by the complainant;
refer only some of the particulars of the complaint as submitted by the complainant; or
add particulars to the complaint as submitted by the complainant; and
must issue a notice of non-referral as contemplated in subsection (2)(b) in respect of any particulars of the complaint not referred to in the Competition Tribunal”.
The Commission argues that it is not bound to the four corners of the complaint as submitted, but can expand the complaint in the referral without the need for further initiation in terms of this section. The respondents give a restrictive interpretation of this sub-section. They argue that it permits the Commission to add detail, but not new causes of action.
The Commission in our view correctly, points out that the term ‘particulars’ is used uniformly throughout section 50(3). Section 50(3) contemplates four different scenarios. The term must therefore be given a consistent interpretation for all the scenarios. As the SCA has pointed out in Woodlands, albeit in a different context, the same term must be given a consistent interpretation in a statute.42
If particulars meant only detail this would not make any sense when used in relation to the rest of section 50(3). To take one example the Commission may issue a non-referral in respect of particulars (section 50(3)(b)). If this meant only detail, it makes no sense, since we know that with a non-referral of particulars a complainant can bring its own referral. (Section 51(1)). It cannot refer detail – it can only refer a cause of action or claim. Particulars must mean facts akin to a claim. Similarly, the section says the Commission may only refer ‘some’ of the particulars. Again if particulars are something susceptible to referral or non-referral they must refer to something susceptible to being a claim, not mere detail.
The Commission’s interpretation is also fortified by the logic of how the Act works. The Commission receives a complaint and then investigates it. It is highly likely that in the course of the investigation the Commission may identify further information not covered by the complaint or may re-characterise what has been identified by the complainant in a different manner. What economists might refer to as altering the theory of harm as initially articulated by a complainant. Provided the Commission’s addition does not travel too far from the original complaint this seems to be a legitimate course of action.
Previously we have taken this approach as we did in our decision in Yara.43 However the higher courts have not followed this. Although no discussion on the ambit of section 50(3)(iii) is evident in the decisions, we assume that the CAC disapproves of this approach as it was expressly mentioned in our reasons in Yara, but not mentioned in the appeal decision.
Although we would otherwise have dismissed the respondents’ application based on this argument, we find we are bound by CAC precedent and for this reason it cannot avail the Commission. For the reasons set out above we find that we do not have jurisdiction to consider the present referral. Therefore the application is granted.
Costs
Although SAB sought costs in this matter we find no merit in this application. In the first place the CAC has already determined this point in Omnia Fertilizer Ltd vs the Competition Commission and upheld a prior Tribunal decision that costs cannot be sought against the Commission or by it in prohibited practice cases. 44 Secondly, even if that approach is to change, SAB has not adequately explained why this application has been brought so late in the matter. What is clear is that the respondents were influenced by the release of the Yara judgment in bringing these applications as they closely followed on its release.45 However as they also sought to rely on earlier decisions, Woodlands (delivered on 26 August 2009) and Glaxo (delivered on 6 December 2006) before this case commenced), we do not feel they are entitled to costs, assuming such a costs order against the Commission is competent, as the applications could have been brought much earlier.
PART B
Reappraisal of the approach to the jurisprudence in respect of initiating documents
Although the Tribunal is bound by precedent created by higher courts we are also the only full-time adjudicative institution created by the Act. As such it is sometimes incumbent upon us to comment on issues that go beyond the resolution of a particular case when we have concerns about the effectiveness of the workings of our system.
We believe that entirely unwittingly, decisions that impact on the legal requirements for a valid referral based on the prior complaint, have threatened to undermine the rights of complainants and the public as represented by the Commission to get access to justice.
We have seen an increasing number of cases in which applications for dismissal or striking out have been made in prohibited practice cases that will never be decided on their merits. As we observed earlier in a short statement when we gave our order dismissing the present case, this is a matter for concern:
“We regret that this case has not been brought to finality so it could be decided on the evidence and not a point of jurisdiction. The complainant will not know whether its complaint was well founded and the respondents will not have an opportunity to clear their names from accusations of anticompetitive conduct.”46
For this reason we have taken the unusual, but we hope not perceived disrespectful, step of alerting higher courts to the problems created by the jurisprudence in the hope that they find this consideration helpful, should these issues come before them again.
When the complaint procedure was designed in terms of the new Act, private and public enforcement was contemplated. At the same time the policy consideration was to have the jurisprudence created in specialist bodies, comprising the three competition authorities the Commission, the Tribunal and the CAC.
If competition litigation was to be confined to specialist bodies it had implications for the system of public and private enforcement of competition law. The legislature opted for a system with several features; first, the reservation of competition jurisprudence on the merits to the exclusive preserve of the competition authorities; second, the prioritising of public enforcement over private enforcement, by giving the public prosecutor in the form of the Commission a preferential right to prosecute matters under the Act; third, recognition that private enforcement had a role to play in enforcing competition and for this reason private claimants should be able to enforce their rights to relief and damages; fourth that civil courts were the best institutions to determine quantum of damages .
For this reason private litigation in the High Court was limited to an assessment of the amount and the awarding of damages arising out of a prohibited practice, but did not involve decisions on the merits which were reserved for the competition authorities.47 Because a determination of the merits was a prerequisite for a damages claim, a method had to be devised for determining this aspect of private claims within the competition system, whilst at the same time not compromising the public enforcement preference of the Commission.
A compromise system was developed to mediate these interests. This led to the design of the complaint procedure, allowing complaints to be driven by private complainants, subject to a waiting period for the Commission to investigate and determine whether it would prosecute a complaint itself and the provision for alternatives, if it did not, which would still allow private complaints to be determined on their merits by the Tribunal and the CAC.
The first alternative is where the Commission decides not to refer the complaint. The Act provides for what it terms a ‘non-referral’ which can be either actual or deemed from the effluxion of time. The consequence of a non-referral is that the complainant is restored its private right and given the title to prosecute the claim itself. The second alternative is that the Commission can elect to prosecute parts, but not the entire complaint, leaving the residue to follow the non-referral route.
Even if the Commission has wholly referred a claim, the complainant can under circumstances still apply to intervene in a case (section 53(1) (c)).
The system also recognises that complaints are not always driven by private parties and that sometimes private parties may want to be informants not complainants. For this reason it recognised the private parties’ rights to be informants (section 49B(2)(a)) and the Commission’s right to self-initiate a claim (section 49B(1)).
A complainant was given the right to pursue interim relief (section 49C) and protected in relation to the terms of a consent order (section 49D(4)). The reason such a complex edifice was created was to regulate the inter-relationship between the complainants’ rights to pursue relief vis a vis the Commission’s duties as prosecutor of first instance in matters involving competition law infringements.
Whilst the existence of a complaint was a necessary prior jurisdictional fact to referral, its content was intended to regulate this inter-relationship and not meant to confer rights upon a respondent, which the respondent could then rely upon to have a complaint dismissed or struck out. Yet this is precisely what has happened. Whilst we have no decided case in which we have had to adjudicate on a dispute between a complainant and the Commission over ownership of the referral, there have been several challenges to date at the instance of a respondent and many more are pending at the time of these reasons.48
The initiating document has become transformed into something it was never intended to be. The complaint, in of itself, does not require a respondent to answer. There is no obligation for it to do so under the Act, and no procedure for this under the rules. It may ignore a complaint without any legal prejudice to its rights. Secondly, the initiating document may only partially find its way into the eventual referral. Whatever was disposed of, unless referred by the complainant in the event of a private referral, does not see the light of day before the Tribunal. The initiating document is therefore not a pleading or an initiation document in legal proceedings. There is not even a requirement that a respondent know of its existence before the case is referred.
Yet the implications of the present case law – although the cases never say this - are that unless the initiating document is as close in precision to an eventual referral, respondents will pick holes in it and the referral will be susceptible to dismissal or striking out. 49
Does this imply that an initiating document has no legal requirements at all other than that it exists? No, it does not. At a minimum, the complaint must describe conduct that is susceptible to prosecution as a prohibited practice. Next, it must be sufficiently clear to meet three implied requirements of the Act.
First, the Act requires that a complaint is sufficiently clear so that one can appreciate if it is the initiative of the Commission or the complainant in the event of a dispute between the two. Second, that for the purpose of section 67(2), which deals with a limitation of actions, one must know whether the initiating document was the complaint now being pursued against a respondent for the sake of interpreting this section sensibly if there were a later challenge of double jeopardy. Third, as the CAC recognised in Glaxo, a complainant should not be allowed to keep part of the complaint in its pocket, get a non-referral and add to the complaint that which was never told to the Commission. If the Commission was deprived of its entitlement to be the prosecutor of first choice the complainant would not be entitled to proceed with the action until subsequently non-referred.
This can of course still lead to enormous complexity. Glaxo’s solution was to require that the facta probanda of the referral be found in the initiating document. But in Glaxo, the court, mindful of the fact that complainants were entitled to lodge a complaint without requiring expert legal advice to do so, recognised that a plaintiff does not have to pigeonhole its complaint to sections of the Act.
The problem is reconciling these two approaches. If alleging the facta probanda is a prerequisite for validity, this presupposes that a complainant has an appreciation of the subtleties of the Act to perform this exercise - all that is not required of it, is the insertion of the section number. But in practice appreciation of these subtleties is far from simple. The discussion on the difference between the facta probanda of access to an essential facility and a refusal to deal by the court in Glaxo, seems to illustrate this concern. That seems an exercise in pigeonholing.
But experience has shown that this is not the only problem. The next aspect which goes beyond correct pigeonholing is the problem complainants frequently have of seeing the tip of the iceberg, but not what lies beneath the sea. As in Yara, a complainant observes high prices, considers from its knowledge of the industry that the source of this high pricing is market failure manifested by some anticompetitive practice of uncertain provenance. It may have reason to think that the high price is the product of unlawful unilateral conduct by a single firm, when in fact, and unbeknown to it, beneath the ocean lurks a cartel hidden from view. The complainant naturally cannot identify the cause of the high price, only its manifestation– even if it identified a possible cartel, it may not know its exact nature and its constituent members.
The case law suggests that if the Commission identifies the true cause as a cartel in the course of its investigation, it must institute a separate complaint de novo. If the existence of the cartel is only revealed after the Commission has already referred the case on some other premise – say as an abuse of dominance – the Commission would not be able to amend its referral, but would have to initiate a new complaint. If, prior to referral, the Commission can add to the particulars of the complaint by relying on section 50(3)(a)(iii), as we discuss more fully below, this problem can be obviated without the need for a fresh initiation which creates the problems of whose complaint it is. It would still be the complainant’s complaint that is ‘referred’ subject to the time period requirements set out in the Act.
We would concede that if the existence of a cartel only became apparent later, post referral, that it would be difficult to find a solution, short of a fresh initiation, if the complaint had been based not on a cartel but an abuse of dominance. That is because post referral, the section 50(3)(a)(iii) opportunity may have gone for adding particulars because it seems that this power must be exercised at the time of the referral. Nor could one rely on the initiating document for a solution. If the initiating document was confined to an allegation of an abuse of dominance against a single firm it would be difficult to characterise the later cartel case as constituting ‘substantially the same conduct’ as the abuse complaint, even if they have the same outcome, namely, a supra-competitive price. That might be an unavoidable consequence of the present complaint based system, but at least these outcomes would be rare.
The other problem with the facta probanda approach is that it fails to recognise that characterising a competition complaint is a complex question that even experts let alone complainants may not agree upon. This is because prohibited practice cases are for the most part effects based. This means that characterising the harmful conduct has to be based on eradicating the effect and this may not be so easy to determine at the stage of initiation. Market power, which is the root cause of anti-competitive outcomes, can be exercised by a single dominant firm, what is termed unilateral conduct or by firms co-ordinating their behaviour. It is not always self-evident that the exercise of market power is caused by one or the other without further investigation.
In the case of cartels even an investigation may not lead to a proper identification of the conduct and hence the utility of the Commission’s leniency policy which induces cartelists to break ranks to avoid the imposition of a penalty. It is not as if a complainant has not identified conduct that may qualify as a prohibited practice. It is simply that they have incorrectly identified the root cause of an anti-competitive outcome.
But even if one correctly identifies the problem as one of single firm conduct the facta probanda approach requires the complainant to describe precisely what competition harm the single firm conduct involves. This is not a straight forward task. Take the situation where a dominant firm in an upstream market is also the supplier to its rival in the downstream market of a key input. Its downstream rival complains that the input is being priced at a supra-competitive price making it unable to compete. This set of facts, could lead to several candidates for a cause of complaint. It could be an excessive price, a margin squeeze, a constructive refusal to deal, denial of access to an essential facility, price discrimination or all of the above. Without further analysis it would be difficult for either the complainant or the Commission at the time of the submission of the complaint to be able to come to a more discerning conclusion of its classification. For this reason only the process of analysis which is premised on further investigation would shape the conclusion in preference of one theory of harm as opposed to the other. Yet on the facta probanda approach all these claims have distinct facta probanda.
The complainant’s initiating document might be read to incorporate all of these, none of these or some in preference to others. The Commission is presented with a dilemma. If it refers one or more of these possible claims without self-initiating, it risks a jurisdictional challenge that the referral is not properly founded in the initiating document.
Given the case law the Commission would probably, as a matter of caution, err on the side of self-initiation, rather than take forward the complainant’s initiation, in cases where it feels uncertain if the referral will be founded in the initiating document. Proponents of the strict approach might see no problem with this. However there is; to the extent that the Commission over corrects and usurps the private party’s complaint as its own, it detracts from rights the Act gives to the private party.
Whilst the Commission is investigating its complaint and has not self-initiated, a complainant enjoys a right to control the timing of the referral.50 If the Commission has not referred its complaint within one year after submission, the complainant can elect to give the Commission an extension or get its non-referral and prosecute the claim itself.
If the Commission self-refers, the complainant loses this right. Or it could lead to disputes between the complainant and the Commission as to who owns the referral. A further problem if the Commission takes over the complaint is that the complainant may be deprived of its right to bring interim relief as it may no longer be regarded as the complainant since its complaint has been superseded by the Commission’s own initiation.
On the other hand if the Commission is allowed to add particulars to the complaint we have only one initiation document and it becomes less important as to who owns it. Whether the Commission added to what was not there or referred what was there, is no longer a relevant or interesting question. This does not prejudice the respondent who must meet the referral at whosoever’s instance.
Thus the Commission’s argument advanced for interpreting section 50(3)(a)(iii); allowing it at referral stage to add particulars to a claim without having to self-initiate, is both a coherent and fair way of interpreting the legislation and one that respects a complainant’s and the public’s rights to access to justice . The use of section 50(3) also helps resolve interminable debates about whether as part of a complaint properly made, some facts submitted constitute information rather than the complaint. If the Commission can add particulars, this unfruitful exercise in classification can be avoided.
Where the Commission has initiated its own complaint it can of course cure any defects in the lack of congruence between the initiating document and the referral by a last minute supplementary initiating document which closes the gaps. Such a step it seems can validly be performed, even at the 11th hour. 51
Post Glaxo other rationales have been given in the case law for adopting the strict approach. Whilst these cases all appear to approve Glaxo, they offer their own rationale and not that given in Glaxo for the strict approach. Recall that in Glaxo the rationale was to prevent a complainant from keeping some part of its complaint in its pocket, thus depriving the Commission of an opportunity to investigate it.
In Woodlands the argument is that as the initiating document precedes an investigation, and that a summons involves the Commission exercising rights of discovery and interrogation, the initiating document must be intelligible in the sense that this term has been used in the case law in relation to search warrants.52 However, this leaves out a crucial step. The Commissioner, who is required to authorise a summons in terms of section 49A of the Act, is required to apply a fresh consideration to the issue of the summons and this creates a distinct and new jurisdictional fact.
The Commissioner has to decide what documents to require, from whom and for what purpose. The recipient of the summons need not even be potential respondent – simply someone believed to have the information required. The summons involves the exercise by the Commission of a policing power, albeit more limited than a search warrant, and therefore its terms need to be more strictly construed than those of an initiation document which does not clothe the Commission with any policing power. But it does not follow that because the issue of a summons has as a prerequisite the initiation of a complaint that the requirements of a valid initiation document need to be the same as those for a valid summons. Each must be judged according to its own requirements and the fact that they must occur in a particular sequence, does not alter their distinct and separate legal requirements.
The next rationale for the strict approach and which was advanced during this hearing is that the initiating document must have a proper basis as it involves state action against a firm and the exercise of public powers against it in respect of an alleged unlawful act. For this reason, so the argument goes, the initiating document must meet the requirements of the strict approach, as firms are entitled to know that the state is investigating them and if so, about what.
This approach gets some recognition in the recent case of Loungefoam where the court held:
“The consequences of a public charge that a firm is guilty of anti-competitive conduct are potentially far-reaching. Considerable reputational damage may flow from being charged with anti-competitive conduct. The ability of the public, via the lens of the media, to distinguish between an allegation of anti-competitive conduct and proof that anti-competitive conduct has occurred is by no means clear.53 If nothing else the firm so charged must devote resources that would otherwise be directed elsewhere to defending itself including, in many instances, fighting a public relations battle in trying to clear itself of these charges. As the Commission’s own statistics quoted in paragraph [47] show many charges of anti-competitive conduct prove on investigation to be unfounded. In addition the investigative phase enables the target firm and the Commission to arrive at a suitable consent order in terms of s 49D that largely obviates the need for a protracted hearing before the Tribunal.54
But later in the same decision the court goes on to say:
“This Court has also stressed that the focus of the complaint should be the conduct that is said to be anti-competitive.55 We have also recently emphasised that all that is required is that the conduct said to contravene the Act be expressed with sufficient clarity for the party against whom that allegation is made to know what the charge is and be able to prepare to meet and rebut it,56 bearing in mind that the competition issues upon which the Tribunal is called to adjudicate may be broader, more general and less clear-cut than those that arise in a conventional civil case in the High Court.57 This gives a broad scope to the Commissioner in formulating the terms of a complaint initiation. In a cartel case, where new participants may be discovered as an investigation progresses, the Commissioner may be justified in couching a complaint initiation in fairly broad terms covering a number of market participants, on the basis of circumstantial evidence to be construed in the light of the pattern that cartel activity takes, even if the Commissioner lacks information directly implicating a particular firm, but that is not what has happened here. ”58
There are two problems with this approach as we can see by the way the court in Loungefoam, sensitive to the Commission’s concerns, tries to reconcile the manner in which the Commission can rectify its problems in initiating complaints validly, with the rights of privacy and dignity of the respondent firm.
The court recognises that the Commission may not know all the facts at the time it commences an investigation. For this reason it may, the court suggests, cast its net more widely, and include as subjects of the investigation, firms it suspects from circumstantial evidence, may be part of a cartel. Yet if the purpose of strict approach is to minimise the possibility for random initiation against innocent parties, this approach seems to encourage it, rather than frustrate it. The strict approach incentivises the Commission to widen its net of suspected firms thus adding to the possibility of an innocent firm being mentioned in an initiating document.
But apart from creating this paradox, there is a far more fundamental problem with this rationale for the strict approach. The act of initiation of a complaint has been transformed into something it is not. It may precede a summons and a complaint referral, but it should not be confused with either.
The Commission cannot simply with an initiating document search premises or subject persons to interrogation or seize documents. All these require an additional step, be it by way of summons or search warrant. Nor as we remarked earlier is a respondent obliged to respond to an initiation document. Thus to elide the requirements for a valid initiating document, with these other processes, is to mischaracterise it and accordingly its consequences. In this respect it seems that the exigencies of a competition prosecution are more demanding than their counterparts in criminal law despite the fact that these are civil proceedings. In a criminal proceeding an accused cannot rely on some difference between the charge sheet in court and the initial complaint as recorded in the police station register to have a prosecution dismissed.
Further, the arguments about reputational damage to a firm should not be overstated. The courts have recognised that the privacy rights of a corporation do not require the same protection as those of individuals.59 Respondents in prohibited practice cases are firms, typically not natural persons and frequently well resourced corporations with access to the media. To liken their considerations for privacy to that of the individual citizen would be erroneous. The public too can be assumed to be able to distinguish between a complaint that is being investigated against a firm, one that is referred and one in which a firm is found to have contravened. Thus the degree of opprobrium that might attach to the news that a firm is being investigated for a prohibited practice should not be overstated
In the same vein to liken prohibited practice contraventions to crimes as a means of justifying the strict approach undermines an important policy consideration of the legislature in the Act, to decriminalise anticompetitive conduct to enhance enforcement. Granted administrative penalties that can be imposed may be severe, but as has been pointed out in previous decisions, the differences between the treatment of a crime and a contravention of the Act are significant not least because no remedy under the Act leads to imprisonment or the deprivation of liberty of an individual.60
Finally, the strict approach ignores the fact that the initiating document marks the beginning, not the end of the investigation. To suppose that it must resemble the referral in its facta probanda suggests that the Commission possesses full knowledge of a prohibited practice at the start of a process which is meant to inform it. It seems to render the act of investigation akin to trial preparation, rather than investigation of whether there is in fact a cause of action and what it is. The strict approach is based on the questionable assumption that full knowledge is available to the complainant be it the Commission or member of the public at the moment of initiation. Experience shows that complainants frequently mischaracterise the source of a complaint even though they may correctly identify an anticompetitive outcome. The strict approach either penalises them for their error or incentivises as we noted the Commission to take over their complaint.
One of the reasons that complaints are so difficult to analyse is that anti-competitive outcomes are also consistent with normal activity absent a prohibited practice. For instance conduct that constitutes an abuse of dominance when performed by a dominant firm is not unlawful when done by a non-dominant firm. High prices are usually the most common manifestation of a prohibited practice that customers complain about to the Commission. Competitors complain about some change in behaviour by a rival. However neither of these manifestations of anti-competitive conduct is necessarily proof of its existence. High prices may be a response of markets to a shortage in production or spike in demand. A change in a dominant firm’s behaviour may be the sign of pro-competitive rather than anticompetitive behaviour. This means that private complaints in particular may misdiagnose the facts or miss the point entirely. Hence the Commission needs to take what is given to it, even though lacking precision and particularity, and ascertain if a prohibited practice lurks beneath the facts and if so, what it is.
The danger with the strict approach is that it will too frequently require the Commission to refashion the complaint as its own. That for the reasons we gave earlier defeats one of the important objectives in the Act in giving complainants some control over the outcome of their complaint.
Respondents are not prejudiced if the strict approach is jettisoned. They need only respond to a complaint referral not the initiating document. Strict adherence to proper pleading in referrals has always been a feature of litigation in these proceedings. The history of this case has demonstrated that. The real reason many respondents have sought to rely on the strict approach has been to stop proceedings against them without having to put up a defence on the merits, because they anticipate that if freshly initiated, a claim may have prescribed or even if not, that the Commission will be disinclined to recommence the process de novo.61 That is not a proper basis for preferring the strict approach.
The strict approach will have two outcomes, neither of which is desirable and in the interests of justice; the Commission will become over inclusive in its initiation policy – no firm will be ignored if susceptible of possible suspicion and no theory of harm will be excluded from possible consideration; respondents, lured by the enticing prospect that a case can be dismissed without ever going to the merits, will be incentivised to dissect the minutiae of initiation documents to find them wanting.
Access to justice and indeed the rule of law become casualties. Granted the public are not entitled to have complaints which are not related to prohibited conduct under the Act prosecuted. Granted they are not entitled to expect to have every complaint they make remedied. But they are entitled to expect if they make a complaint under the Act to the Commission, even if lacking in legal and economic articulacy, that it will be judged on its merits.
For this reason a doctrine of interpretation needs to be applied in difficult cases that departs from discerning whether precise facta probanda emerge from a complaint to found a referral, to reading it in an inclusive manner that allows complaints to be heard on their merits and not be dismissed for want of some earlier precision. This is a doctrine that approaches the interpretation of initiating documents in favour of upholding validity of the referral which it underpins rather than dismissal in marginal cases; that favours the right of complainants to have access to justice.
Conclusion
Jettisoning the strict approach need not be as drastic a departure from current case law as it might seem. For instance the test in Glaxo that the conduct is substantially similar can remain the filter for the determining whether a referral is based on the initiating document. What needs to be reconsidered is how this test is applied in practice.
Facta probanda are a useful tool in applying this test, but they must not be used as conclusive of whether the test is met. If the facta probanda that appear in the initiating document underpin the referral that should serve to dispense with a challenge to it. However, even if they do not, it should not follow that the complaint referral should be dismissed on that basis alone; rather the enquiry should go further. Greater appreciation needs to be given to the fact that substantial similarity does not require complete resemblance. Because a given set of facts alleged in a complaint, may be capable of different conclusions of its competitive harm, courts should be slow to dismiss complaint referrals on jurisdictional grounds because of an apparent lack of alignment in facta probanda. As long as an inexact document is lined up for comparison with a later, more precise one, it will always be found wanting. Yet this is exactly what the facta probanda and strict approaches require the adjudicator to do, as they compare the initiating document with the referral, with, as we have seen, unfortunate results.
The intention test remains useful for determining where, as in a ‘Mrs Malherbe situation’, one is not certain if a submission is a complaint. It is less useful in interpreting the language of what is otherwise unambiguously a complaint, because that leads to a dissection of the language of a complaint in an unhelpful manner that offends against the notion of an access to justice and incentivises opportunistic behaviour by respondents.
The Commission’s power to add particulars to a complaint in section 50(3)(a)(iii) needs to be appreciated against this backdrop, as a useful means by which it adds coherence to a complaint, which is either improperly formed or because made without investigation, may be incorrect in some of its assumptions or wanting in an economic analysis of a theory of harm. Many of the difficulties identified in the case law would be overcome if this power to supplement was given the reading suggested by the Commission in this matter but not yet recognised by the courts.
In cartel cases, there is no reason why the initiating document should have to identify all members of a cartel prior to the referral, for a valid referral against the firm. Whilst in most cases this can be accomplished in view of the leniency policy, it is not always so. Paradoxically, it is the more egregious cartels which retain their covert nature, which will benefit from this formality. It should be noted that contrary to what is stated in Woodlands, an investigation into a prohibited practice does not necessarily mean an investigation into a particular named firm.62 Indeed the Act used to state in terms of the old section 44 that :
“A complaint against a prohibited practice by a firm may be initiated by the Commission or submitted to the Competition Commission in the prescribed manner”.
When the Act was amended, section 49B(1) became the equivalent section. This time the legislature referred to a “complaint against a prohibited practice” dispensing with the phrase “... by a firm”. This seems to suggest that the focus of a valid complaint is the conduct and not the identification of specific firms. Whilst in dominance cases the conduct and the identity of the firm would go hand in hand with the complaint, the case of the cartel is less evident, given their complexity.
Technical failures in the initiating document, of the sort which might require strict adherence in pleadings, need to be reconsidered as a basis for dismissing claims against particular respondents incorrectly cited in the initiating document, but not in the later referral. Confusion between a holding company and its subsidiary or between a division and the firm to which it belongs frequently occur. This type of error is usually only apparent after the firm has filed its answer. Requiring the firm to recommence an initiation and then to amend its complaint referral, if prescription has not intervened in the interim, is formalism at its most absurd.
Whilst there may still be some hard cases that are difficult to resolve, adherence to some of these approaches will at least eliminate some cases which ought to be tried on their merits falling by the wayside on preliminary objections.
It needs to be borne in mind that disputes in the Tribunal are subject to section 34 of the Constitution of the Republic of South Africa which deals with access to courts.63 This requires that access to justice be granted to all participants – those who seek to bring a complaint and those who seek to defend themselves against a complaint. The right of access to justice by a complainant is no less compelling a consideration than the right of a respondent to a fair administrative process. But at the stage where a case is dismissed on jurisdictional grounds there is no symmetry in the consideration of the balance between the two competing rights. A fair process right is enjoyed once a referral is made. A right to access to justice is wholly negated if a matter cannot even to proceed to trial.
Two of the elements of a functioning civil justice system, as famously laid down by Lord Woolf in his report on the reform of the civil justice system in the United Kingdom, is that a system of justice must be understandable to those who use it and second, that it provide as much certainty as possible as the nature of particular cases allow.64 The strict approach does not meet that standard. Complainants do not understand now what is required of them as it has become too complex and accordingly the outcomes of the system are uncertain.
No doubt the complainants in this case will feel Lord Woolf’s test has not been met and they may well be right.65
______________________
Norman Manoim
Concurring: Yasmin Carrim and Merle Holden
Tribunal Researcher: Londiwe Senona
For the First Respondent : D Unterhalter (SC), P McNally (SC), A Cockrell (SC) and L Sisilana, instructed by Bowman Gilfillan.
For the Second to Fourteenth Respondents : J Wilson instructed by Deneys Reitz
For the Commission : A Gotz and R Mkhabela instructed by
Mkhabela Huntley Adekeye Attorneys.
1They number some 20 firms, and are listed in the complaint. (See record page 14-15). There are 15 further complainants, who are not part of the Big Daddy’s Group, and who are separately listed in the complaint. (See record pages 15-17).
2See paragraph 3, record page 17.
3See undated letter from SAB, record page 76
4The form itself instructs complainants on how to complete it. Attaching to it a “... concise statement of the conduct that is the subject of the complaint” is one of the prescribed instructions. The complainant as well is encouraged to attach to the CC 1 “... any relevant documents ...”
5See footnote below.
6The Commission supplied one letter from SAB in reply to an earlier letter from the complainant. But the copy of the SAB letter in the Commission’s possession appears to have been received by it after the complaint was lodged. This appears from later correspondence between the complainant and the Commission in which he states that he ‘again’ encloses the letter.
7The initiation document is referred to in section 49B of the Act and the referral in section 50 and 51 of the Act.
8The Woodlands case (Woodlands Dairy (Pty) Ltd and Another vs The Competition Commission, Case No: 88/CAC/Mar09) seems to suggest that there could be more than one or the one could be amended. The Yara decision (Yara South Africa (Pty) Ltd vs The Competition Commission and Others, Case No: 93/CAC/Mar10) seems to contemplate that amendment is not authorised by the Act.
9National Association of Pharmaceutical Wholesalers and Others vs Glaxo Wellcome (Pty) Limited and Others, Case No: 45/CR/Jul01; GlaxosmithKline South Africa (Pty) Ltd vs David Lewis N.O and Others, Case No: 62/CAC/Apr06.
10Ibid see paragraph 29 of the CAC decision. In Woodlands on the same subject the SCA held: “ The question arises whether there are any jurisdictional requirements for the initiation of a complaint by the Commissioner. I would have thought, as a matter of principle, that the commissioner must at the very least have been in possession of information concerning an alleged prohibited practice which, objectively speaking could give rise to a reasonable suspicion of the existence of a prohibited practice. Without such information there could not be a rational exercise of the power.” See Woodlands paragraph 13.
11Ibid paragraph 30.
12Supra fn 9, paragraph 33. Note in this passage the Court cites with approval from the decision of the Tribunal which was being appealed. In that passage the Tribunal held the test was held to be whether the conduct in the referral was ‘substantially the same as that alleged in the referral’. See paragraph 88 of the Tribunal decision.
13Ibid paragraph 15.
14Ibid paragraph 14.
15Supra, fn 9, paragraph 16.
16Ibid paragraph 19. Facta probanda are generally defined as “facts which must be proved in order to disclose a cause of action”.
17See Tribunal Glaxo decision supra, paragraph 96.
18See supra fn 8 for CAC decision. See also Supreme Court of Appeal decision, Case No: 105/2010, and Tribunal decision, Case No: 103/CR/Dec06.
19See supra fn 18, Woodlands Supreme Court of Appeal decision, paragraph 10.
20In later decisions this seems to have been misunderstood to apply to a summons commencing action in the High Court.
21Netstar (Pty) Limited and Others vs Competition Commission and Another, (Case No: 99/CAC/May 10; 98/cac/May10; 97/CAC/May10).
22Ibid paragraph 27.
23See footnote above.
24See supra fn8, Yara decision, paragraph 30.
25Ibid, Yara Paragraph 33.
26Ibid, Yara paragraph 39. “There is no provision in the Act for amendment of a complaint. Rule 18 of the rules for the conduct of proceedings in the Tribunal only provides for amendment of a referral. I can only conclude that the legislature intended that complaints be initiated or submitted as provided for in section 49(B)(1) and 49(B)(2) of the Act. Further, the legislature must have intended that the Commission only refer to the Tribunal such a complaint as initiated or submitted to it. Consequently only the particulars of the complaint as submitted by the Nutri-Flo should have been referred to the Tribunal. The information relating to cartel activity and collusion was not intended by Nutri-Flo to be a complaint.”
27We consider this more fully later in the decision.
28See the definition of complainant in the Act which refers to a complainant as “...a person who has submitted a complaint in terms of section 49B(2)(b)”. See also sections 49C, 49D(4), 50, 51 and 53 of the Act, for the rights given to the complainant.
29Clover Industries Limited and Another vs The Competition Commission and Others, Case No: 78/CAC/Jul08.
30Yara paragraph 35.
31Yara paragraph 38.
32This is contrary to what was expressed in Woodlands where the SCA contemplated the possibility of amending an initiating document. See supra fn 18.
33See complaint referral paragraphs 7-8 and in particular 7.7 for the horizontal allegation paragraph 8.5 for the vertical allegation.
34See complaint referral paragraph 8.6, where the Commission concludes that the anticompetitive effects of the agreement are aggravated by various factors, one of which is stated in 8.6.3 to be “SAB’s practice of not offering trade discounts and delivery compensation fees to independent distributors such as the wholesalers.”
35See Yara decision, paragraph 25 where the court noted that firms which the Commission sought to introduce as respondents by way of an amendment to the referral had not been mentioned as respondents on the CC1.
36Undated letter addressed to SAB from Pitsiladi, record page 105.
37See Yara decision, paragraph 28. “ However the form CC1 is not the only document that may contain details of the complaint. The form provides for attachment to it of: ‘any other relevant document...’
38See fn 4 supra.
39Record pages 105-6.
40Record Page 106
41See undated letter from SAB to Pitsiladi record page 46 which refers to a letter from Pitsiladi of 11 October. In the version of this letter at page 114, this is the same letter that appears at 105 of the record, we can see at the top that it was faxed on 11 October 2004. The paragraph in question states: “A small amount of production is distributed through independent distributors, who have been specifically appointed to provide distribution services for SAB. These distributors make no margin in respect of products sold through them – they are paid a warehousing and distribution fee for their services...”
42See Woodlands Dairy (Pty) Ltd and Another vs The Competition Commission, Case No: 105/2010 [2010] ZASCA (13 September 2010), paragraph 18.
43In the case between the Competition Commission vs Yara South Africa (Pty) Ltd and Another, Case No: 31/CR/May05, paragraph 40, the Tribunal said that: “The Commission is not only empowered in section 50(3)(a)(iii) to add further particulars to a complaint submitted by a complainant at any stage of its
investigation but is also enjoined to investigate a complaint “if new facts came to light.”
44Case No: 77/CAC/Jul08, paragraph 20.
45The CAC Yara decision was delivered on 14 March 2011.
46Tribunal statement dated 7 April 2011.
47Section 62(5) read with section 65(6) - 65(10) of the Act.
48Eg. Paramount Mills (Pty) Ltd Complaint, Case No: 15/CR/Mar10; Arcelormittal Complaint, Case No: 61/CR/Sep09; Computicket Complaint, Case No: 20/CR/Apr10; Astral Operation Complaint, Case No: 74/CR/Jun08, Pentel Complaint, Case No: 27/CR/Apr11.
49See Yara – the respondent looks to the initiating document to see the complaint against it.
50Section 50(2) read with 50(4) and (5).
51See Loungefoam and Others vs The Competition Commission and Others, Case No: 102/CAC/Jun10, at paragraph 55, which appears to approve a contention made in these terms by the tribunal in the decision subject to appeal.
52See for instance Powell NO and others v Van der Merwe NO and others 2005 (5) SA 62 (SCA).
53The Supreme Court of Canada has recently drawn attention to the problem that an accused person may face if they are unable to ensure that ‘the public will not be influenced by untested, one-sided and stigmatising information bearing on issues that are often irrelevant to guilt.’ Toronto Star Newspapers Ltd and others v R (in right of Canada) and others; Canadian Broadcasting Corporation and others v R and another [2010] SCC 21; [2010] 1 SCR 721; 2011 (1) CLR 1 (SCC) para [51].
54Loungefoam paragraph 49.
55Glaxo Wellcome (Pty) Ltd and Others v National Association of Pharmaceutical Wholesalers [2002] ZACAC 3, paras [15] to [19].
56The essential question is whether the issue was raised with sufficient clarity not whether it was described by a term understood in the area of competition law. See Senwes Ltd v Competition Commission of South Africa [2009] ZACAC 4, paras [27] to [43].
57Netstar (Pty) Ltd and Others v Competition Commission South Africa and Another [2011] ZACAC 1, para [27].
58Loungefoam , paragraph 53.
59In Zuma case [S vs Zuma 2009 (1) SA 1 (CC), paragraph 77] the Constitutional Court held “… in Thint's case we are dealing with the search of the offices of a company. As a corporate entity, Thint does not bear human dignity and thus its rights of privacy are much attenuated compared to those of human beings...” See footnote 65 of the Tribunal Woodlands decision, Case No: 103/CR/Dec06.
60See Federal Mogul decisions, first by the Tribunal (Case No: 08/CR/Mar01) and subsequently, Federal-Mogul Aftermarket South Africa (Pty) Limited v Competition Commission and Another, Case No: 33/CAC/Sep03.
61Section 67(1) states “A complaint in respect of a prohibited practice may not be initiated more than three years after the practice has ceased.”
62If Woodlands is correct it may mean that if in a cartel case only some firms are identified in the initiating document and at some time later, others are and thus a new initiation is required, but this happens three years after the conduct has ceased, the latter firms cannot be prosecuted because of prescription, but the former firms who were mentioned in the earlier initiation (assuming this to be within three years of the cessation of the conduct) would. This seems anomalous given that they are participants in the same conduct. Noteworthy, is that section 67(1) refers, like section 49(B)(1), to the practice not the firm whilst section 67(2), which is dealing with limitations on bringing a subsequent referral, refers to complaint being referred against “ ...any firm...” This fortifies the suggestion that an initiating document refers to a practice but does not require mention of the firm whilst the referral for obvious reasons does. If the legislature had intended the mention of the identity of the respondent firm to be a prerequisite of a valid complaint, the language of 67(1) and 67(2) would surely be the same.
63Section 34 states that “everyone has the right to have any legal problem or case decided by a court or an independent body.”
64Rt. Hon. the Lord Woolf, Master of the Rolls, Final Report to the Lord Chancellor on the civil justice system in England and Wales (July 1996). This passage has been cited with approval by the former Chief Justice Sandile Ngcobo in a conference address entitled “Enhancing Justice: the search for better justice”, 6-10 July 2011.
65Note that both complainants in this matter and in Yara were legally advised in filing their respective complaints. It has become a matter not simply of access to legal expertise, but specific competition law expertise, invariably economic expertise as well.