---

Latest updated version 11 September 2024

 

IN THE COMPETITION APPEAL COURT OF SOUTH AFRICA

 

Case No.246/CAC/Jun23

REPORTABLE: YES

OF INTEREST TO OTHER JUDGES: YES

REVISED.

26 April 2024

 

In the matter between:

 

THE COMPETITION COMMISSION                                    Appellant

 

and

 

WACO AFRICA (PTY) LTD                                                 First Respondent

TEDOC SGB CAPE JV                                                       Second Respondent

SUPERFECTA SGB CAPE JV                                            Third Respondent

MTSWENI SGB CAPE JV                                                   Fourth Respondent

TEDOC INDUSTRIES (PTY) LTD                                        Fifth Respondent

SUPERFECTA TRADING 159 CC                                       Sixth Respondent

MTSWENI CORROSION CONTROL (PTY) LTD                Seventh Respondent

 

Coram: Manoim JP, Siwendu AJA, Murphy AJA

Judgments: Murphy AJA (minority) [1- 116] ( Manoim JP, separate concurring, (majority)[ 117-145])

Heard on: 1 February 2024

Decided on: 26 April 2024

Summary: Competition Act 89 of 1998 – section 4(1)(b) – whether bidding consortiums responsive to tender contravene- horizontal or vertical agreement considered – partnership law interface with competition law – hub and spoke agreements and intra-enterprise conspiracy considered

 

On appeal from the Competition Tribunal:

1.                 The appeal is dismissed.

2.                 There is no order as to costs.

 

JUDGMENT

 

Murphy AJA

Introduction

 

1.                 In February 2018, the appellant, the Competition Commission, (“the Commission”), following an investigation referred a complaint to the Competition Tribunal (“the Tribunal”) against Waco Africa (Pty) Limited (“Waco”) and three joint ventures (“JV’s”) in which Waco was a partner through its local operating division SGB- Cape (referred to collectively as “the respondents”). The three JV’s are Tedoc SGB- Cape Joint Venture, (“the Tedoc JV”), Superfecta SGB-Cape Joint Venture (“the Superfecta JV) and Mtsweni SGB-Cape Joint Venture (“the Mtsweni JV”). The complaint also cited Waco’s joint venture partners (“JVP’s”) individually, namely Tedoc Industries (Pty) Ltd (“Tedoc”), Superfecta Trading 159 CC (“Superfecta”) and Mtsweni Corrosion Control (Pty) Ltd (“Mtsweni”). Waco has been referred to throughout the proceedings by the name of its South African operating division SGB-Cape, and we will continue to do so.

 

2.                 The Commission contends that the respondents, acting in concert in relation to a tender issued by Eskom in 2015 (“Corp3130”), contravened section 4(1) of the Competition Act 89 of 1998 (“the Act”), which in relevant part reads:

 

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

 

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

 

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

 

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

 

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

 

(iii)  collusive tendering.”

 

3.                 Section 4(1) of the Act only prohibits restrictive practices: i) between competitors in a horizontal relationship; and ii) which have unjustifiable anti- competitive effects (in terms of section 4(1)(a) of the Act) or are per se prohibited (under section 4(1)(b) of the Act).

 

4.                 The Commission maintains that the respondents engaged in price-fixing and collusive tendering in contravention of section 4(1)(b)(i) and section 4(1)(b) (iii) of the Act and thus requested the Tribunal to impose an administrative penalty against them in accordance with section 58(1)(a)(iii) read with section 59 of the Act. On 30 May 2023, the Tribunal handed down its decision dismissing the complaint, holding that there had been no contravention of section 4(1)(b)(i) or (iii) of the Act primarily because the Commission had “failed to prove the first leg of the requirement of section 4(1)(b)” that SGB-Cape and the three JVs were competitor firms and thus in horizontal relationship with each other. The Commission has appealed against the decision of the Tribunal on numerous grounds, not all of which require determination for the purposes of the appeal. We dismiss the appeal. However, as appears from the concurring judgment of my colleagues, Manoim JP and Siwendu AJA, we differ in our reasons for doing so.

 

The factual background

 

5.                 Waco was founded in 1952 and has since grown into an international conglomerate operating in Africa, Australasia and the United Kingdom. SGB-Cape is a business division of Waco focused on the provision of scaffolding and insulation services. It has rendered industrial scaffolding and heat insulation services to Eskom since 1970 and has sophisticated expertise in working on high-pressure, high- temperature vessels such as boilers. Scaffolding is required at Eskom’s power stations to provide access to generating units, boilers and turbines, for the purposes of maintenance and repair. The work is technically complex and involves high levels of skill, competence and experience. Thermal insulation is equally complex and requires similar experience, skill and careful design of the insulation material and its application.

 

6.                 Prior to 2007, SGB-Cape provided services to Eskom at 9 of its then 11 coal- fired power stations. Thereafter, pursuant to Eskom’s strategy of reducing its dependency on suppliers, the number of power stations serviced by SGB-Cape was reduced to four (Matla, Kriel, Lethabo and Grootvlei) in terms of a tender (ENK275) awarded to it in 2010.

 

7.                 At around this time Eskom was under pressure to demonstrate increased commitment to broad-based black economic empowerment (“BBBEE”) in its procurement policies. It accordingly informed SGB-Cape that it needed to enhance its empowerment profile to maintain ongoing commercial opportunities. SGB-Cape then formed separate partnerships with Tedoc and Superfecta, two black women owned (“BWO”) companies, for the purpose of jointly supplying goods and services to Eskom.

 

8.                 Tedoc is a BWO personnel service provider specializing in outsourcing and placement of labour and payroll administration. It has had a relationship with SGB- Cape since 2008. SGB-Cape formed a JV with Tedoc (“TSJV”) in 2011 in order to tender for the scaffolding services at Eskom’s Kusile power station. In terms of the TSJV, Tedoc supplied labor services (recruitment of workers and payroll administration) while SGB-Cape contributed the scaffolding and project management. In 2012, after Eskom informed SGB-Cape that a power station serviced by it would be reallocated to another supplier if it did not improve its BBBEE level to above 50%, SGB-Cape approached Superfecta (a BWO which had worked with SGB-Cape at Koeberg power station in insulation) to provide labour services at the Kusile power station. Thereafter, in 2014, SGB-Cape formed an incorporated JV, Octorex (Pty) Ltd (“Octorex”), with Tedoc and Superfecta as JVPs (each holding a 25% shareholding) to service the insulation contract at Kusile. The JV was on a similar basis to the TSJV formed with Tedoc in 2011 for the scaffolding services. By 2015 SGB-Cape had improved its empowerment status to a level 2 BBBEE rating, with its black ownership status having increased from 16.49% in 2013 to 43.06% in 2015. Its BWO status increased from 4.53% in 2013 to 14.83% in 2015.

 

9.                 In March 2015 Eskom issued the invitation to tender in respect of Corp3130 for the supply, transportation, delivery, insulation and dismantling of scaffolding and thermal insulation at all of its 15 power stations. The invitation to tender permitted bidders to bid for all 15 power stations or only a number. In terms of Clause 4 of the invitation to tender Eskom reserved the right to appoint one or more bidders upon evaluation of the bids, was not bound to accept the lowest of any tender and reserved the right to enter into negotiations with any one or more of the bidders. The invitation to tender was ambiguous in relation to the BBBEE requirements. At page 5 of the invitation it stated that “preference shall be given to Black Owned companies with at least 51% Black Ownership”. However, confusingly, at page 11 the invitation stated that 51% was an objective criterion in terms of PPPFA[1] for the award of the tender.

 

10.              On 17 March 2015, Mr. Naicker of SGB-Cape emailed Ms. Mzileni of Eskom attaching a completed acknowledgement of tender form in relation to Corp3130, reflecting that it was from SGB-Cape; Tedoc SGB-Cape JV; Superfecta SGB-Cape JV; and Hygitech SGB-Cape JV (later substituted by Mtsweni SGB-Cape JV). Mr. Naicker explicitly recorded that SGB-Cape would be submitting four tenders “with joint venture partners as listed above”.

 

11.              A tender clarification meeting was held at Eskom on 26 March 2015 and was attended by SGB-Cape executives as well as Mr. Mtsweni of Mtsweni Corrosion Control, the JVP in Mtsweni SGB-Cape JV. Questions raised by attendees led to the clarification that 51% black ownership was a requirement for the tender and thus cleared up the ambiguity in the invitation to tender. Later on 26th March 2015, after the bid clarification meeting, Mr. Naicker wrote to Ms. Mzileni indicating the intention of SGB-Cape to use JVs and requesting additional time inter alia to arrange and prepare joint venture agreements.

 

12.              SGB-Cape then engaged in internal discussions and the evaluation of strategic partners for the purpose of considering different JV options. Internal correspondence reflects that SGB-Cape considered eight different options. The first option was for SGB-Cape to tender on its own, but it was recognised that its bid would not meet the 51% black ownership criterion. The other options were bids by JVs including: a single bid by the Octorex JV (involving Tedoc and Superfecta, each as 25% shareholders); or other JVs with different JVPs, including the second, third and fourth respondent JVs. The prices of the bids of the various possible JVs would be higher than the stand- alone SGB-Cape bid.

 

13.              SGB-Cape ultimately decided to submit four bids. The first by SGB-Cape alone; the second by the Tedoc SGB-Cape JV; the third by the Superfecta SGB-Cape JV; and the fourth by Mtsweni SGB-Cape JV – which was black youth owned (“BYO”). In his testimony before the Tribunal, Mr. Visagie explained that the rationale for SGB- Cape proceeding in this way was twofold: i) to enable Eskom to “mix and match” power stations with SGB-Cape or a JV which enjoyed enhanced empowerment credentials in three different respects (black-owned, BWO or BYO); and ii) to increase the total number of power stations it would be involved in servicing. The SGB-Cape stand-alone bid did not meet the 51% black ownership requirement and was non-compliant on that basis alone.[2]

 

14.              None of the JVPs had previously been awarded any scaffolding or insulation projects directly at Eskom or any other industrial client. However, as discussed, Tedoc and Superfecta had partnered with SGB-Cape in Octorex and Tedoc alone at Kusile.

 

15.              There were different reasons for selecting each of the three JVPs to partner SGB-Cape in the three distinct JVs. Tedoc was selected because it was BWO with proven capacity to provide the contract administration and labour sourcing services needed to fulfill the tender. It had a database of candidates with experience in the provision of scaffolding services. It could not have bid alone as it lacked the technical know-how and expertise in relation to the project management of complex scaffolding and insulation services. The Tedoc SGB-Cape JV was constituted on 24 April 2015 with a 49% shareholding by Tedoc and 51% by Waco. This resulted in the JV enjoying 70.9% black ownership and 51.7% BWO.

 

16.              Superfecta was selected because it too was BWO and also had worked with SGB-Cape on other projects as a supplier of labour and other human resources services. It had a database of clients with experience in the provision of heat insulation services. The Superfecta SGB-Cape JV was constituted on 24 April 2015 with a 45% shareholding by Superfecta and 55% by Waco. This resulted in the JV enjoying 68.7% black ownership and 53.2% BWO.

 

17.              Mtsweni was selected because it was BYO. It also had no experience with the work required under the tender but was primarily a provider of employment and transport services. The owner of Mtsweni, Mr. Sibusiso Mtsweni, had worked for SGB- Cape as a quality controller and had performed roof repairs as an independent contractor at Kendal where SGB-Cape erected scaffolding. It was thus in a position to provide contract administration and labour sourcing services needed. The Mtsweni SGB-Cape JV was constituted on 24 April 2015 with 40% shareholding by Mtsweni and 60% by Waco. This resulted in the JV enjoying 65.8% black ownership and 40% BYO.

 

18.              Mr. Visagie approached each of the JVPs separately with distinct JV proposals. The invitation from SGB-Cape did not disclose the identity of the other JVPs and each JVP was unaware of the nature of the other JVs and their bids. Tedoc and Superfecta agreed to participate in the bids and to receive a labour administration fee of […]%. Mtsweni negotiated a […]% labour administration fee. The labour administration fee is a percentage calculated taking into account: i) the industry standard for this type of project; ii) the type of skills needed to be sourced and their scarcity; iii) the number of employees needed for the project; iv) BEE credentials; and v) the quality of the services provided. The labour administration fee is intended to cover the JVP's overheads, including costs such as vehicles needed, telephones, staff complement required for the project, the number of site offices etc.

 

19.              Internal correspondence between 17 April 2015 and 20 April 2015 confirmed the intention not to pursue a JV similar to Octorex which was used in the Kusile tender – being a single bid by a single JV constituted by SGB-Cape and two JVPs (Tedoc and Superfecta). The preference was clearly for three distinct bids by different JVs each involving a partnership between SGB-Cape and a single JVP. Waco believed that the three bids gave Eskom the opportunity to develop three “black industrialists” as opposed to one. It also enhanced the possibility of SGB-Cape winning work at more power stations.

 

20.              On 21 April 2015 the board of directors of Waco resolved to submit a tender for the contract. The resolution authorised Mr. John Falconer (the Commercial Director at SGB-Cape) to complete the tender documents. On the same day, the boards of Tedoc, Superfecta and Mtsweni passed resolutions deciding to: i) enter into a joint venture agreement with SGB-Cape; ii) submit a tender in the name of the JV for Corp3130; and iii) authorise Mr. Falconer to sign the tender documents. It appears from emails dated 23 April 2015 addressed by Mr. Naicker of SGB-Cape to Mr. Jiyane at Tedoc and Ms. Peters at Superfecta that the resolutions were drafted by SGB-Cape and sent to the JVPs for resolution and signature.

 

21.              There had also been some discussion about the possibility of Mr. Mtsweni submitting a separate bid. On 21 April 2015, Mr. Visagie sent an e-mail to Mr. Falconer and others at SGB-Cape, which he copied Mr. Mtsweni, in which he stated:

 

“Sibusiso (Mr. Mtsweni) confirmed telephonically that he understands our requirement that if we enter into a joint venture for the Eskom maintenance tender that Mtsweni Corrosion Control will not submit a tender on their own. This to avoid possible conflict of interest. Sibusiso agreed to this.”

 

22.              Mr. Mtsweni did not testify before the Tribunal. However, he confirmed in his witness statement that he had attended Eskom’s clarification meeting on 26 March 2015 and was actively looking to partner with other suppliers and to participate in the tender. None of his efforts progressed to any formal proposals to bid for the tender jointly with others.

 

23.              On 22 April 2015, Mr. Falconer addressed an important e-mail to his colleagues at Waco and SGB-Cape dealing with a proposed discount structure. It provides insight into the strategy of SGB-Cape in relation to the tender and its hope that it would receive contracts at seven of Eskom’s power stations either on its own or as a participant in the three JVs. It proposed discounts of between 1.5% and 7.5% depending on the number (between 4 and 7) of power stations awarded and would apply to all of the bids involving SGB-Cape “whether on our own, or in joint venture with a BEE partner.”

 

24.              SGB-Cape concluded JV agreements with each JVP on 24 April 2015 for the purposes of supplying “goods” and “related services” to Eskom in terms of Corp3130. The JV agreement explicitly recognised that: i) the JVP possessed the requisite BEE credentials; ii) the JVP had proven expertise and experience in the provision of labour and associated employment practices; iii) SGB-Cape conducts business as a provider of scaffolding, thermal insulation and corrosion protection services and asbestos removal, including all related technical and operation services; and iv) the JVP and SGB-Cape had agreed, through the medium of the JV, to execute and complete the “project” for their joint benefit. The joint venture interest of the parties was defined as the respective participation ratio which each party had in the JV – being their shareholding. Clause 15 of the JV agreement provided that any profits or losses of the JV would be shared by the parties, pro rata to their participation ratios in the JV.

 

25.              In terms of the JV agreement, the JV subcontracted to SGB-Cape the entire supply of goods and related services (project management services) to be supplied by the JV to Eskom. SGB-Cape would supply all technical and operational services, including management of labour (whether sourced themselves or provided by the JVPs), the necessary scaffolding, insulation materials and all financial, project, and commercial management. The JV then subcontracted to the JVP the provision of skilled, semi-skilled and unskilled labour (related services) to be supplied by the JV to Eskom. The consideration payable to the JVP by the JV in respect of the supply of related services was the amount as agreed between SGB-Cape and the JVP of the consideration payable by Eskom to the JV. It was agreed to pay Tedoc and Superfecta a labour administration fee of […]% and Mtsweni […]%.

 

26.              The invitation to tender for Corp 3130 had various requirements. Each bidder was required to complete and submit proposals in relation to SHE (safety, health and environment), commercial, financial, technical and quality. In general terms, this required a determination of the total cost of meeting the specifications of the tender and to ensure recovery of the business expenses and profit.

 

27.              SGB-Cape unilaterally prepared and completed the bid documents for itself and on behalf of the three JVs for submission to Eskom. The estimating team at SGB-Cape worked on the bid documentation for all four bids and determined the tender rates, including discounts to be offered to Eskom, and finalised the mandatory commercial, financial, technical, SHE and quality proposals. Mr. Visagie completed the technical aspects of the JV bids, while Mr. Naicker and Mr. Falconer compiled the other aspects. Thus, SGB-Cape provided the SHE plan, baseline SHE risk assessment, asbestos contractor approval certificate, relevant information in relation to the quality management system, a quality control plan, draft inspection and test plans, detailed estimates of the quantities of goods and the nature of services required to fulfill the tender, and similar related information.

 

28.              The bids thus contained virtually identical commercial, technical, financial, SHE and quality documents. The transport factors and price escalation formulas, in respect of both labour and non-labour components, were identical in all four bids.

 

29.              Apart from providing the corporate documentation required by SGB-Cape to complete the respective bids, the JVPs did not participate in any aspect of the bid preparation process including the pricing of the labour services. The JVPs were not aware of the rates, costs and profits of SGB-Cape; nor were they aware of the JV’s combined rates. The JVPs did not have: i) an SHE policy, plan and assessment; ii) an approval certificate as an asbestos contractor; iii) the financial profile or means to capitalise the project; and iv) relevant quality accreditations. The JVPs had no ability to do design of specialised scaffolding structure drawings and the engineering certification of such or to erect scaffolding to a minimum height of 80m. Likewise, they had no experience in construction or maintenance at power stations or the ability to do specialized lagging, cladding, asbestos removal and storage. Only SGB-Cape could meet these requirements.

30.              The labour rates reflected in the bids were calculated using as a base rate the rates paid by Eskom at the four power stations at which SGB-Cape was providing services. Certain of these rates were regulated by the Steel and Engineering Federation of South Africa (“SEIFSA”) and the Metal and Engineering Industries Bargaining Council (“MEIBC”). Amounts were added to the base rate for skills and unemployment levies, leave allowances, provident fund contributions and other benefits, fees and levies. The recovery of the costs of medicals, communication, transport of personnel, and personal protective equipment were also added to the labour cost and calculated on an average hourly cost per worker recovery.

 

31.              The goods and materials rates in the bids were based on information from SGB- Cape’s procurement department which sourced material prices from various suppliers of the specialised and technical material required by the tender specifications. The suppliers have existing relationships with SGB-Cape that enabled competitive pricing. Different mark-ups, depending on the likely volume of a particular material to be used, were added to the prices. The JVPs had no input on these matters either as they had no knowledge of or relationship with the suppliers.

 

32.              SGB-Cape, having thus by agreement compiled the bids and determined the bid price for each of the four bids, submitted the bids to Eskom on 28 April 2015. Each of the bids was on its own an offer in respect of all 15 Eskom power stations. Mr. Falconer signed the bids on behalf of SGB-Cape and the three JVs. The JVPs were not aware that SGB-Cape had submitted a stand-alone bid outside of a JV and only became aware that SGB-Cape had submitted other JV bids later during the bid evaluation process. Hence, SGB-Cape was in possession of the full pricing information of each of the four bids. SGB-Cape did not pass any pricing information from any of the JVPs to the other JVPs. Nonetheless, Mr. Falconer of SGB-Cape had knowledge of the pricing information of each bid.

 

33.              As mentioned earlier, the stand-alone bid by SGB-Cape was not compliant with the requirements of the bid as SGB-Cape was not 51% black owned. Across the line items of the bid, the SGB-Cape sole bid was consistently the cheapest of the four bids. Tedoc SGB-Cape JV had a level 1 BEE status with 70.9% black ownership and 51.7% BWO. Its prices were […]% higher than SGB-Cape’s bid in respect of every line item. Superfecta SGB-Cape JV had a level 1 BEE status with 68.7% black ownership and 53.2% BWO. Each of the line item prices in the bid was identical to the prices of the Tedoc SGB-Cape JV bid and thus […]% more than SGB-Cape’s stand-alone price. Mtsweni SGB-Cape JV had a level 1 BEE status with 65.8% black ownership and 40% BYO. Each of the line item prices in this bid was […]% more than the other two JVs and […]% more then SGB-Cape’s stand-alone prices.

 

34.              The factual witnesses who appeared on behalf of SGB-Cape before the Tribunal explained that the mark-ups above the SGB-Cape stand-alone bid were necessary to recover the additional costs of the JVPs (in addition to the labour administration fee of […]% and […]% paid to the JVs) and in order to mitigate the risks arising from administering and managing the JVs. SGB-Cape justified the price differential between the bids on the basis of three key facts: i) the JVs would require greater oversight and management than SGB-Cape’s stand-alone bid to ensure that each JV performed to Eskom’s requirements; ii) Mtsweni was an untested JVP, which created a risk for delivery to the specifications of the tender when compared to the proven track record of Superfecta and Tedoc; iii) and each JV exceeded Eskom’s 51% black ownership goal which justified a premium and enabled Eskom to allocate power stations in a manner that satisfied its price and transformation preferences.

 

35.              An additional […]% mark-up was added to all costs across the four bids. Mr. Visagie testified that […]% of the […]% (described as a “rough calculation”) was attributable to SGB-Cape's overheads, which then left […]% “fat in the margin”. However, the JVP's made no input into this cost either.

 

36.              Section 17 of the Commercial Proposal of each of the four bids dealt with “Price Basis”. This clause cross-referenced the other bids as follows:

 

“We wish to secure at least six Power Stations i.e. existing four Power Stations plus an additional two…SGB-Cape is currently carrying out the maintenance work on the Grootvlei, Matla, Kriel, and Lethabo power stations, and has an operations and administration establishment near Kendal Power station to run and administer all of the work done on these four stations. It would be relatively simple for SGB-Cape to upgrade this establishment, and this would enable us to handle the maintenance on more than four power stations close to the area concerned, and we are therefore in a position to offer discounts based on us being awarded four or more of the power stations close to our current Kendal establishment. We are therefore prepared to offer you a discount structure as follows:

 

·                    On the basis that we are awarded the four power stations we are currently on, we offer 1,5% discount to al tendered rates.

 

·                    If we are awarded a fifth station, with this being either Hendrina, Kendal or Duvha, we will offer a 3% discount to al tendered rates.

 

·                    If we are awarded fifth and sixth station, with these again being two of the above three stations, we will offer a 5% discount to all tendered rates.

 

·                    If we are awarded all seven stations, being the current four and the above three, we will offer a 7,5% discount of all tendered rates.

 

All of the above discounts are applicable to any of the offers which have been made to Eskom which involve SGB-Cape, whether this is on their own, or in joint venture with a partner – e.g. if SGB-Cape are awarded Grootvlei and Matla, Tedoc SGB-Cape JV is awarded Kriel and Lethabo, and Superfecta SGB-Cape JV is awarded Duvha and Kendal, then 5% discount would be applicable to the tendered rates from all three offers.”

 

37.              In May 2015, Ms. Asanda Mzileni, Senior Advisor, Group Commercial in Eskom’s procurement department, conducted an initial evaluation of the 31 bids received for Corp3130. She noticed the significant similarities between the four bids and that they had all been signed by Mr. Falconer, who she knew from her previous dealings with SGB-Cape. She saw the bids as different bids by four competitors and was concerned that the JVs and SGB-Cape had shared pricing information. She believed that the bids were anti-competitive (price-fixing) as SGB-Cape could crowd out other competitors. It was conceivable for the bids after evaluation to rank first to fourth - meaning that SGB-Cape could benefit SGB-Cape disproportionately. However, under cross examination during her testimony to the Tribunal, Ms. Mzileni readily conceded, more than once, that nothing in the bid documentation was misleading. The bid documents all reflected the identity of the bidders and there was no apparent attempt “to pull the wool over the eyes” of Eskom or to create a false impression. Ms. Mzileni had known Mr. Falconer for a number of years, had worked with him in the past and immediately saw and recognised that he had signed each bid on behalf of SGB-Cape and the three JVs.

 

38.              Ms. Mzileni raised her concerns internally at Eskom leading to an investigation by Eskom’s Forensic and Anti-Corruption Department, which on 8 September 2015 concluded that the price coordination amounted to collusive tendering and that “the difference on rates between SGB and the JVs were consistent which shows that the difference was well calculated and SGB was made the lowest”, which conduct, in its view, amounted to price-fixing. On this basis, the investigators recommended that Eskom management consider disciplinary action against SGB-Cape and the JVs for collusive tendering and that the matter be reported to the Commission in terms of the Act.

 

39.              On 25 November 2015, management at Eskom made a submission to the Exco Procurement Subcommittee and the Board Tender Committee seeking a resolution to negotiate and conclude contracts for Corp3130 with five successful bidders: Southey Contracting (Pty) Ltd, Electro Heat Energy (Pty) Ltd, Oram Industrials, FBC Johannesburg Scaffolding JV and Kaefer Energy Products (Pty) Ltd. Management also requested a mandate to engage in post-tender negotiations with the identified bidders. Eskom then requested SGB-Cape and the three JVs to extend the validity of their bids for Corp3130 and to agree to an extension of the validity date to 30 June 2016 – presumably to allow for the post-tender negotiations. This is the first occasion on which the JVPs learnt of the other bids involving SGB-Cape.

 

40.              Subsequent litigation delayed the award of the contract to the identified bidders for some years.

 

The proceedings before the Tribunal

 

41.              On 18 March 2016, Eskom filed a complaint against the respondents for collusive tendering and price-fixing in relation to Corp3130. On 13 March 2017, in response to an apparent request from the Commission for an extension to investigate the matter until 30 September 2017, Eskom’s attorneys recorded that the complaint had been referred a year earlier and that the length of time taken by the Commission to investigate the matter had prejudiced Eskom and in the light of that they were instructed on behalf of Eskom to withdraw the complaint referral.

 

42.              Subsequent to Eskom’s withdrawal of the complaint, the Commission nonetheless decided to continue with its investigation and also initiated a complaint to cover the individual JVPs in addition to SGB-Cape and the three JVs. The Form CC1 dated 14 March 2017 particularised the complaint as follows:

 

“The Respondents are alleged to have entered into an agreement and/or arrangement to collude, by discussing and coordinating the preparation and presentations of their respective bids to Eskom in respect of an Eskom tender, with tender number Corp3130 – a tender which called for the supply, transportation, delivery, installation and dismantling of scaffolding and thermal insulation for all the 15 Eskom coal-fired power stations, in possible contravention of section 4(1)(b)(i) and (iii) of the Competition Act 89 of 1998, as amended.”

 

43.              The Commission later sought an order from the Tribunal declaring that the respondents had contravened sections 4(1)(b)(i) and (iii) of the Act and an order declaring Waco and the three JVPs liable for the payment of an administrative penalty equal to 10% of their respective annual turnover in terms of section 58 read with section 59 of the Act.

 

44.              On 23 August 2018, pursuant to an earlier order of the Tribunal, the Commission filed a supplementary affidavit clarifying that its allegation of collusive tendering was not founded on the agreement to form the JVs but on the mandate given to Mr. Falconer to prepare and finalise the tender documents for all four bids.[3]

 

45.              The Tribunal heard evidence and argument over ten days during June 2022 and August 2022. Four factual witnesses testified, namely: Ms. Mzileni, on behalf of the Commission; and Mr. Visagie, Mr. Falconer and Ms. Peters (the owner of Superfecta) on behalf of the respondents. There are no consequential disputes relating to the relevant facts. The Tribunal also heard evidence from two expert witnesses: Mr. Jason Aproskie - a Principal Economist at the Commission; and Mr. Stephen Malherbe – a competition economist employed by Genesis Analytics who testified on behalf of the respondents.

 

46.              The Commission argued before the Tribunal that SGB-Cape was a competitor to each of the three JVs and the JVs were competitors to each other. Each JV could alone perform in terms of the tender and were formed for that purpose. The Commission identified the joint decision taken on 21 April 2015 by SGB-Cape and the JVs to mandate Mr. Falconer to prepare the bids and fix the prices of the bids (rather than the formation of the JVs) as the impugned conduct (“the first impugned agreement”) constituting price-fixing and collusive tendering. The mandate to Mr. Falconer to decide on the pricing of all four bids, the Commission submitted, amounted to “perfect coordination” of bid prices between four competitors. Each of the four bidders, through Mr. Falconer and SGB-Cape, was aware of the prices of the other bidders. Accepting that the JVPs could not price key aspects of the bids, there remained room to independently price at least three elements of the bids: firstly, the […]% or […]% labour administration fee paid to the JVP; the […]% or […]% mark-up added to the JV’s labour prices; and a […]% mark-up on total costs added by SGB-Cape.

 

47.              Mr. Aproskie testified that the respondents’ counterfactual of a single joint bid contradicted SGB-Cape’s stated objective of winning more of the Eskom work and would have presented a greater risk of non-success. The risk of non-success would naturally have created an incentive to price more competitively and the coordination evident in the characterised conduct resulted in the loss of this competitive advantage. Coordination on pricing was not required for SGB-Cape to provide its set of transformation options to Eskom. Independent pricing would have incentivised the JVPs to cut margins or negotiate down prices for technical services in order to win the work, or at least more of the work. Moreover, SGB-Cape and the JVs did not fully disclose to Eskom the true nature of the arrangements and the price coordination. The four bids were presented as independently priced bids when in fact they were not. This amounted to collusive tendering.

 

48.              In addition to its contention that there had been price-fixing and collusion between SGB-Cape and the JVs, the Commission also maintained that the agreement of 21 April 2015 between SGB-Cape and Mtsweni in terms of which a competitive  restraint was imposed on Mtsweni not to compete for the tender independently (“the second impugned agreement”) also amounted to collusive tendering.

 

49.              The respondents argued before the Tribunal that the JVPs were in effect suppliers to the JV, and the relationship between SGB-Cape and those firms was a vertical one. Absent the JV arrangements no competitor was removed from the process and the mandate to Mr. Falconer was an integral part of the agreement to establish the JVs. Moreover, the circumstances did not reveal a collusive purpose. Eskom had signaled to the market and to SGB-Cape specifically that in addition to technical requirements, it intended to bring the non-price transformation imperatives into account. Each JV offered distinct transformation objectives. The multi-bid strategy would not have resulted in higher prices, poorer quality of service or reduced technical excellence. Nor did it remove competitors from the process. There was no adverse outcome for Eskom or competition. None of the common features of bid-rigging (collusive tendering) was present in this case. There was no conspiracy to create anti- competitive outcomes, such as creating a supra-competitive price or reduction in quality or capacity. The core logic of price-fixing or bid-rigging was absent in this case.

 

50.              The respondents argued further that the counterfactual contended for by the Commission, namely, multiple bids but with independent pricing of certain of the labour rate components by each JVP in its respective JV bid was not plausible from a practical perspective. None of the three JVPs would have been able to complete its own bid, let alone in a competitive fashion. Only SGB-Cape was able to determine the pricing of the materials and services. In so far as the JVPs had some knowledge of the labour components of the bid, as the applicable labour rates (“rate to man”) had been set through collective bargaining, legislation and prevailing labour practices, there was limited scope for independent pricing. Likewise, while basic labour costs were in some respects discretionary, elements such as accommodation, medical, PPE etc. on a per hour basis were not discretionary but “passed through”. The same consideration applied to the non-work wage package (NWWP) which included contributions for skills levy, unemployment insurance fund, provident fund, leave and bonus provisions and like contributions.

 

51.              The Tribunal held that the JVPs, Superfecta and Tedoc, were not in horizontal relationship and concluded similarly in relation to Mtsweni. It rejected the hypothesis that Mtsweni was capable of bidding on its own or potentially through subcontracting with other competitors. In relation to the JVs, the Tribunal held that it was insufficient to find that the four bidders (the JVs and SGB-Cape) were competitors solely on the basis that they had submitted bids for the same tender. It rejected the Commission's contention that the restrictive practice was the provision of the mandate to Mr. Falconer after the JVs were set up. In its view, the relevant agreements were the JV agreements concluded for the sole purpose of bidding for the tender.

 

52.              The Tribunal also rejected the Commission's argument that the three JVs were separate firms that could have done the work independently and were thus competitors on that basis and the counterfactual of four independent bids. Even if it was technically feasible to duplicate estimating and accounting resources, the Tribunal considered that it “would not be commercially rational for a firm to establish four entities in which it is the controlling mind or majority shareholder, and then allow each one of them to bid independently in competition with each other, without having some influence in the pricing decisions of those firms, for the same piece of work or tender. If it failed to exercise some influence in the pricing decisions of these entities - either by providing guidelines or determining the margins - it would run the risk of cannibalizing its own margins.” Accepting that SGB-Cape was the single controlling mind, it followed that SGB-Cape could not collude with itself. The correct counterfactual in its view was a single bid from SGB-Cape likely in a JV with one or more JVPs.

 

53.              The Tribunal thus concluded that the JVs and SGB-Cape were not horizontal competitors. It also held that the conduct impugned by the Commission, the granting of the mandate to Mr. Falconer, could not be characterised as collusion (in terms of section 4(1)(b)(iii) of the Act). The basis of this latter finding is not clear, but it seems that the Tribunal was influenced by the fact that there was some measure of transparency, SGB-Cape had been encouraged by Eskom to come up with transformation solutions, and the procurement legal framework did not expressly prohibit the submission of multiple bids by the same bidder. However, the Tribunal neglected to specifically consider whether the alleged conduct constituted price-fixing as contemplated in section 4(1)(b)(i) of the Act. In the final analysis, the Tribunal held that absent a horizontal relationship there had been no contravention of section 4(1)(b) of the Act.

 

The appeal

 

54.              The Commission noted its appeal on 21 June 2023 and filed a notice of appeal setting out extensive grounds of appeal. It contends essentially that the Tribunal erred by: i) finding that the JVs were not competitors; ii) rejecting the counterfactual of four separate bids; iii) failing to accept that it was possible to price the four bids more independently by giving the JVPs a greater say in pricing the labour component of the bid; iv) finding that the formation of the JVs was the relevant agreement rather than the agreement of the JVs with SGB-Cape for it solely to complete the bid documents and determine pricing; v) holding that Mtsweni was not a potential competitor; and ultimately vi) finding therefore that the four bidders were not in horizontal relationship meaning there had been no contravention of section 4(1)(b) of the Act.

 

55.              The respondents align with the reasoning of the Tribunal. They contend that the three JVs were constructed by SGB-Cape for the specific purpose of Corp3130 and absent the decision to create JVs for this bid, they would not have existed. The JVs and SGB-Cape were accordingly not in horizontal relationship and the correct counterfactual therefore was a single bid as opposed to four independent bids. Additionally, for the reasons stated by the Tribunal, Mtsweni was not a viable competitor and the restraint upon him did not amount to collusive tendering. In reality the JVPs were suppliers in the downstream market for labour recruitment, administration and supply and thus the relationship was vertical in substance.

 

56.              The respondents further argued, that given their transparency and the disclosure of the essential elements of the scheme in their bid to Eskom, there was no collusive tendering. The core logic of price-fixing and collusive tendering was absent here as there was no conspiracy to create anti-competitive outcomes such as bidding at a supra-competitive price or restricting quality. With regard to the price coordination undertaken by Mr. Falconer, the respondents essentially submitted that the coordination had redeeming efficiency gains and transformative benefits of an order that weighed against prohibiting the conduct per se. The notion advanced by the Commission that the JVPs could have acted more independently in negotiations regarding the mark-ups and the labour administration fee was implausible given the relationships and the overall purpose of the scheme.

 

A horizontal relationship

 

57.              The purpose of section 4(1)(b) of the Act is to capture anti-competitive conduct which is so egregious that no efficiency defence is permitted. To that end it is necessary to characterise the impugned conduct. The purpose of characterisation is to determine the potential for the impugned conduct to result in harm to competition. Therefore, the inquiry under section 4(1)(b) examines firstly whether the parties are “firms ..in a horizontal relationship” – that is, competitors, and, if so, secondly, whether the case involves price-fixing, the division of markets or collusive tendering within the contemplated scope of section 4(1)(b).[4]

 

58.              The determination of horizontality is a mixed legal-factual question akin to an in limine determination of a condition precedent. It is in essence a jurisdictional point. The Tribunal and this court only have jurisdiction to determine if a harmful restrictive practice of collusion or price fixing has occurred if it is established that the agreement or practice was concluded or engaged in between parties who can rightly (legally) be described as “competitors”. A horizontal relationship is defined by section 1 of the Act simply as “a relationship between competitors” as opposed to a vertical relationship that is defined as “a relationship between a firm and its suppliers, its customers or both”.

 

59.              Methodologically, in accordance with our prevailing jurisprudence, the question of whether firms are competitors also involves an examination of the counterfactual position (where there is no impugned agreement in place) to the existing factual position (where the agreement is in place) to determine whether the agreement itself resulted in harm to competition or not and, therefore, whether the conduct should be per se prohibited. The question posed in this counterfactual analysis is thus simply whether the parties were potential competitors in the absence of the impugned agreement. If the answer to the question is in the affirmative, then competition may have been harmed. If the answer is in the negative - i.e. the two firms would not have been competitors absent the agreement - then the agreement could not have harmed competition and would not fall within the scope of the per se prohibition.[5]

 

60.              The Tribunal held that the relationship between the JVs and SGB-Cape was not horizontal but vertical because: i) the JVs were constructed for the specific purpose of the tender; ii) SGB-Cape was the controlling mind of the four bids and thus could not collude with itself; iii) the counterfactual was therefore a single joint bid; and iv) but for the decision to create the JVs the bids would not have existed.

 

61.              The finding of the Tribunal on the question of horizontality, in my view (which is not shared by my colleagues), is erroneous for a few reasons.

 

62.              The first question is whether SGB-Cape and each of the JVs are firms in a horizontal relationship as contemplated in section 4(1)(b) of the Act. Section 1 of the Act defines a “firm” to include a person, partnership or a trust. Waco Africa (Pty) Ltd, as an incorporated company, constitutes a person and thus a firm in terms of the Act. The question then is whether the JVs were also “firms”. The term “joint venture” usually describes an enterprise jointly embarked upon with the object of making a profit. As such, the term has been regarded as synonymous with a partnership.[6] A partnership is a legal relationship arising from an agreement between two or more persons to contribute to an enterprise with the object of making profits and to divide such profits.

 

63.              The JV agreements between SGB-Cape and the three JVPs all indicate in their preamble that the parties agreed to execute and complete “the project/enterprise” (Corp3130) for their joint benefit. In clause 6 of the JV agreements the parties agreed that the funds required for the purposes of the joint venture would be raised from bank facilities and loans by the parties pro rata to their participation ratios; and clause 15 provided that any profits and losses of the joint venture would be shared by the parties pro rata to the participation ratio. The agreed contribution of each partner to the JV was the supply of different goods and services to the JV.[7] Each of the minority JVPs held between 40 and 49% interest in each of the respective JVs. Each JVP provided labour services to the JV which represented at least 50% of the tender value. Substantial participation of each JVP was thus a requirement of the partnership. Hence, SGB-Cape and each of the JVPs to the JVs brought something to the JV partnerships; the business of each JV was to be carried on for the benefit of both parties; and the object of the JV was to make and share profits. Accordingly, the provisions of the JV agreements leave no doubt that the relationships constituted by the JV agreements were indeed partnerships. For that reason, each of the JVs constituted a firm as defined by section 1 of the Act.

 

64.              The next question is whether the JVs were competitors with each other and with SGB-Cape (firms in a horizontal relationship); or can we regard the JVs as mere suppliers or customers to SGB-Cape in vertical relationship. The term “competitor” is not defined in the Act. In ordinary parlance, a competitor is a rival that competes in a competitive situation with others striving to achieve the same goal or a specific economic benefit. The terms “supplier” and “customer” in the definition of a vertical relationship are likewise not defined in the Act. A supplier is an individual or entity that provides goods or services to another party under the terms of a contract. A customer is an individual entity that purchases goods or services from a supplier in exchange for consideration. The JVPs were not suppliers to SGB-Cape. They were suppliers to their respective JVs. The question then is: were the JVs competitors with each other?

 

65.              If we were to accept the finding of the Tribunal that the JVs were essentially SGB-Cape in another guise, we would in effect be holding that the JVs were not firms as contemplated in section 1 of the Act but were shams. None of the witness accepted this was the case.

 

66.              Mr. Malherbe, the respondents’ expert, also rejected the notion that the JVs were shams or expedient mechanisms at the service of the “single controlling mind” of SGB-Cape. When specifically asked by counsel on behalf of the Commission whether the JVs were “real economic entities” or in effect shams, he replied that they were “real entities” which “were undoubtedly going to be production mechanisms with different components operating in the real world with each of those different parties…having a real role.” When asked in a follow up question whether each of the JVs would have been capable of executing the tender had it received an award, Mr. Malherbe replied in the affirmative.

 

67.              Mr. Malherbe’s answers are consistent with the principles of the law of partnerships. Whilst a partnership is not a legal persona, but merely an association of persons, it is nonetheless a firm as defined in section 1 of the Act and possessed of qualities akin to legal personality. In Potchefstroom Dairies v Standard Milk Supply Co.[8] the Court held:

 

“It makes little difference whether we regard a partnership as a persona or whether we regard it as a contractual compound of several personae. In the one case the firm name if properly signed is the signature of the persona, in the other case it is the signature of the contractual compound. And the distinction between the two seems more academic than substantial… a partnership through a corporate individual is so far analogous to a persona that it may be called a quasi-persona… For many purposes it has been or is treated as having a persona of its own; and particularly in relation to commercial transactions is this a convenient and succinct way of regarding its position.”

 

68.              It is moreover an established principle that partners are able to bind themselves to a partnership creditor (in this case Eskom) in such a way that each partner is liable in solidum to the creditor for payment of the whole of the partnership debts during and after the subsistence of the partnership.[9]

 

69.              It is common cause that each JV (as opposed to the individual JVPs), by reason of the respective contributions of SGB-Cape and the particular JVP, was capable of providing the entire suite of services and the required goods that Corp3130 required. Each JV made an offer to Eskom to provide the relevant goods and services and certified that it was capable of delivery. It follows that Eskom could have awarded contractual work at one or more of the 15 power stations under the tender to any one, some or all of the JVs and SGB-Cape. The four bidders were in competition with each other for the same work.

 

70.              In the light of that, it seems to me erroneous to conclude that the relationship between the JVs and SGB Cape was “vertical” as defined in section 1 of the Act. A vertical relationship, as set out above, has a specific statutory, legal meaning. It is a relationship between a firm and its suppliers and customers. Most certainly, as discussed but worthy of emphasis, the JVPs were in vertical relationship with the JVs. SGB-Cape provided both services and goods in relation to scaffolding and insulation to each JV in which it was involved vertically as a JVP. Likewise Tedoc, Superfecta and Mtsweni provided labour services to each JV in which they were involved as JVPs. The JVPs, including SGB-Cape, were suppliers to the JVs. However, the three other JVPs provided no goods or services to SGB-Cape and were thus not suppliers to it. They supplied their services only to the JVs in which they were involved. And, as already said, the JVs were competing against each other and SGB-Cape (in its stand- alone bid) for the work on offer under Corp3130.

 

71.              The Tribunal's finding that the correct counterfactual was one combined bid, and not four separate bids is, therefore, in my opinion, not sustainable in law or on the facts as it ignores the express provisions of the Act, established legal principles and the explicit terms of the bids which in law could have obliged each JV to perform the requirements of the tender. Had the intention been to submit a single bid, then SGB- Cape, as the “single controlling mind”, could have done that in a vehicle similar to Octorex which created vertical relationships between the JVPs and the single incorporated Octorex JV vehicle formed in that instance. Such an option was indeed considered by SGB-Cape for the purposes of Corp3130 but was rejected for reasons that remain opaque.

 

72.              As Mr. Ngcukaitobi SC correctly submitted on behalf of the Commission, a firm should not be permitted to submit a bid on the strength of the combined economic, technical and black ownership components of both constituent firms viewed together, but then disavow the separate legal and economic entity on the basis of a form over substance argument that the JVs were not firms as defined in order to avoid legal consequences. Once the bidders created an appearance of legitimate competition, it is impermissible for them to attempt to escape the legal implications of their conduct. To uphold the Tribunal’s finding on horizontality would not only do violence to the provisions of the Act and the principles of the law of partnership, as said, by in effect holding that the JVs were shams, but also to the terms of the bids submitted by each JV in which they individually undertook to deliver the goods and services of the tender.

 

73.              The counterfactual analysis also requires determination of whether the parties were potential competitors in the absence of the impugned agreement. The agreements (or concerted practices) identified by the Commission as the restrictive horizontal practices in issue in this case were the mandates given by each of the JVs to Mr. Falconer of SGB-Cape to complete and submit their bids and the price coordination that this conduct involved. The respondents and their witnesses proceeded on a different assumption - that the relevant agreement or practice was the agreement between SGB-Cape and each JVP to form the JVs. The Tribunal essentially accepted the respondents’ position and concluded that absent the formation of the JVs (of which the mandate given to Mr. Falconer was an integral part) there were no competitive firms in horizontal relationship and in the final analysis dismissed the complaint on that basis.

 

74.              In terms of section 49B of the Act, the Commission is entitled to initiate a complaint against an alleged prohibited practice. The complaint initiated by the Commissioner in this case is that set out in the Form CC1 issued and signed by him on 14 March 2017. The Form CC1 focused on the mandate given to Mr. Falconer and alleged that the respondents had entered into an agreement and/or arrangement to collude, “by discussing and coordinating the preparation and presentations of their respective bids to Eskom”. There can thus be no doubt that the restrictive practice targeted by the Commission was the coordination of the bids (the first impugned agreement) rather than the formation of the JVs. In my view, the Tribunal erred in holding otherwise. Hence, (applying the counterfactual analysis) absent the coordination arising from the mandate given to Mr. Falconer (the first impugned agreement), SGB-Cape and each JV remained in a rival position and capable of bidding for the work and receiving it from Eskom. SGB-Cape (in its stand-alone bid) and each JV stood in horizontal relationship with one another and the conclusion that the relationship was in substance vertical is incorrect.

 

75.              In the result, in my view, the Tribunal erred in its ultimate finding that the complaint fell to be dismissed on the basis that there was no restrictive practice between firms in a horizontal relationship. My colleagues in the majority disagree with my conclusion for the reasons set out in their concurring judgment below. They prefer the view that the relevant relationship was indeed vertical. The appeal must therefore be disposed of on the in limine basis that as there was no agreement or practice of firms in a horizontal relationship, the prohibition against price fixing and collusive

tendering has no application in this case. I, however, on my approach, am obliged to address these issues and turn now to do so.

 

The scope of the prohibitions in section 4(1)(b) of the Act

 

76.              The question then is whether the impugned agreement between the firms in horizontal relationship, the four bidders, amounted to a restrictive horizontal practice involving collusive tendering or price-fixing as contemplated in section 4(1)(b) of the Act.

 

77.              Categorising an agreement or practice as having the character of prohibited “fixing a purchase or selling price”, “dividing markets” or “collusive tendering”, as contemplated in section 4(1)(b) of the Act, requires the decision-maker to confer an analogous character on the agreement or conduct in question. Therefore, it is necessary to envision the qualitative character of the kind of agreements and practices intended to be prohibited. Characterisation in this sense involves run-of-the-mill statutory interpretation to give content to the scope of the prohibition and will naturally draw on the experience of other legal systems. Moreover, “whether conduct is of such a character that no defence should be entertained is informed both by common sense and competition economics”.[10]

 

78.              Nonetheless, determining whether impugned conduct falls within the scope of the per se prohibitions in section 4(1)(b) of the Act is not a straightforward matter. The problem arises from the difficulty in maintaining a clear distinction between the per se prohibitions in section 4(1)(b) of the Act and the kinds of conduct prohibited in section 4(1)(a) of the Act which permits a rule of reason analysis of any justification for the apparently anti-competitive effects when weighed against technological, efficiency or other pro-competitive gains (“the efficiency defence”).[11] There is often “no bright light separating per se from rule of reason analysis”. Per se prohibitions also require research into market conditions before reaching a presumption of wholly unjustifiable anti-competitive conduct.[12]

 

79.              The cardinal difference between a per se and a rule of reason standard depends on how much evidence is required to decide whether an impugned agreement is prohibited. A rule of reason standard requires detailed analysis of market power and its effects, as well as the consideration of possible efficiency gains, and thus usually involves an expensive and lengthy trial. Competition law thus draws a distinction between conduct which is almost always harmful (per se prohibited) and conduct which is only potentially harmful (inviting rule of reason justification). Price- fixing, division of markets and collusive tendering (bid-rigging) are almost always harmful. Rule of reason justifications of such conduct, requiring usually a lengthy trial, invariably will shift the balance of convenience in favour of the defendants and may undermine legislative policy to constrain cartel-like behaviour.

 

80.      In American Natural Soda Ash Corporation and Another v Competition Commission of South Africa and Others[13], (“ANSAC”) the Supreme Court of Appeal (“SCA”) set out the approach to be followed for determining whether conduct falls within the scope of the prohibitions enacted by section 4(1)(b) of the Act. The starting point is to recognise that the juxtaposition of section 4(1)(a) to section 4(1)(b) means that section 4(1)(b) is aimed at imposing a per se prohibition in which the “efficiency defence” expressly contemplated by section 4(1)(a) cannot be raised. The essential enquiry is to establish whether the character of the conduct complained of coincides with the character of the prohibited conduct: and this process necessarily embodies two elements. One is the scope of the prohibition: a matter of statutory construction. The other is the nature of the conduct complained of - this is a factual enquiry.

 

81.              With regard to the first element of the process – the matter of statutory construction, the wording of section 4(1)(b) of the Act, though concise and clear, requires fuller exegesis with reference to the context, the purpose of the prohibitions and the intention of the legislature. For procedural and jurisdictional reasons specific to the case, the SCA in ANSAC did not provide precise guidance on what in the second element of the process, the mandatory factual enquiry, should entail. The reasoning of the SCA in ANSAC indicates that it accepted that evidence of the nature, purpose and effects of the agreement (beyond the terms of the agreement) may be relevant and admissible in determining whether the impugned conduct falls within the scope of a per se prohibition. There is accordingly no basis for restricting the relevant evidence in the factual inquiry solely to the terms of the impugned agreement or practice. Evidence of the impact of the impugned conduct on competition (obviously relevant to the evaluation of the efficiency defence under section 4(1)(a) of the Act) may also be relevant and admissible in characterising conduct for the purposes of section 4(1)(b) of the Act.

 

82.              Moodaliyar and Weeks in their discussion of the method of characterisation,[14] and in support of the contention that evidence of impact is relevant in characterising per se prohibitions, refer to a comparable line of reasoning followed in Massachusetts v Board of Optometry[15] where the court asked whether the restraint is inherently suspect or of the kind that appears likely, absent an efficiency justification, to restrict competition and decrease output. For example, horizontal price-fixing and market division are inherently suspect because they are likely to raise prices by reducing output. If the restraint is not inherently suspect, then rule of reason must be employed. But if it is inherently suspect, then the per se analysis still asks if there is a plausible efficiency justification for the practice. Such an efficiency defense is plausible if it cannot be rejected without extensive factual inquiry. If it is not plausible, then the restraint can be quickly condemned. But if the efficiency justification is plausible, further inquiry is needed to determine whether the justification is really valid. If it is, it must be assessed under the full balancing test of the rule of reason without further inquiry- there are no likely benefits to offset the threat to competition.[16]

 

83.              While the nature and purpose of the alleged conduct (considered relevant by the SCA in ANSAC) may be more or less self-evident on the basis of straightforward information, the same cannot easily be said about the effects of the agreement. The danger in allowing evidence on effects, as just intimated, is that it will likely blur the line between the per se and rule of reason standard. However, it is possible to perform initial characterisation on the limited facts descriptive of the impugned conduct, as well as the court’s “experience and knowledge regarding a broader class of cases”. Conscious of the danger of blurring per se and rule of reason analysis, courts or tribunals conducting an initial characterisation, and in order to give effect to the advantage of the per se process (the avoiding of lengthy and complex trials of fact), must rely on a common sense understanding about price-fixing, dividing markets and collusive tendering. And, therefore, any evidence on effects relied on for the purpose of characterisation should be limited to only “low cost information that clearly contradicts the general experience and knowledge about the broader class of conduct” or judicial notice of notorious legal, social and economic aspects.[17]

 

84.              On this basis, a court may then elect a summary disposition of the case based on initial characterisation that the complained of conduct can be presumed to be of such a nature that it should be prohibited per se. Only where there is uncertainty regarding evidence of the purpose and impact of the conduct, should consideration be given to fuller evidence supplementing and refining the initial characterisation, especially in more complicated cases where an initial characterisation may not be straightforward. At this point, the court must avoid moving too fully into the territory of rule of reason analysis by ensuring focus on the effects of the impugned conduct merely to characterise it, rather than a full blown justification analysis, blurring the line between per se and rule of reason, with implications, for the “front loading” of costly information by defendants, thereby undermining the effectiveness of the per se standard.

 

Collusive tendering

 

85.              With this analytical framework in mind, I turn to consideration of whether the impugned agreement and practice in this case constituted collusive tendering.

 

86.              Collusive tendering (bid-rigging) is a particular form of price-fixing within the context of tenders involving potential bidders agreeing amongst themselves to collude and coordinate their bids in order to determine the winner at a particular price. Usually what happens is that the producers of the goods or the service providers in the market collude and decide amongst themselves which of them will submit the lowest and, therefore, the likely successful bid. The purpose of such a scheme often is to allocate tenders between tenderers over a period of time so that each obtains a share of the market. This results in artificially high prices. As such, collusive tendering is a form of unethical allocation of markets and ultimately price-fixing in the context of tenders.[18]

 

87.              Collusive tendering takes various forms,[19] for example: i) cover bidding - when competitors agree to submit bids that are intended to be unsuccessful (artificially inflated price or non-compliance with mandatory requirements), so that the chosen one of them can win the contract;[20] ii) bid suppression - when one or more competitors who otherwise would be expected to bid, or who have previously bid, agree to refrain from bidding or withdraw a previously submitted bid so that the designated winning competitor’s bid will be accepted; iii) bid rotation - potential competitors agree amongst themselves to submit bids but for one of them to submit the lowest bid so that each bidder can take turns at winning a contract on a rotational basis;[21] iv) bid shopping - the practice of divulging a tenderer’s bid to other bidders before the award of a contract in order to secure a lower bid; v) collusive subcontracting – where a low bidder will agree to withdraw its bid in favour of the next low bidder in exchange for a lucrative subcontract that divides the illegally obtained higher price between them; and vi) repetitive bidding aimed at helping members of a bid-rigging agreement to allocate contracts among themselves. These forms of collusion are not exhaustive; collusive tendering can take complex forms and is not static in that new forms evolve over time. However, what all these forms of collusion have in common is that they are surreptitiously aimed at increasing market share and maximising profit by lessening competition.[22]

 

88.              The use of the adjective “collusive” connotes an element of secrecy, deceit or surreptitious conduct. The Concise Oxford Dictionary defines “collusion” as “a fraudulent secret understanding; especially between ostensible opponents as in a lawsuit”. Applying the principle of statutory construction that words in a statute should be given their ordinary meaning, on the assumption that the legislature specifically used the words in their popular sense, unless the context or the subject matter clearly shows that they were intended to be used in a different sense,[23] it follows that there should be an element of deceit, secrecy or surreptitious dealing before impugned conduct can constitute collusive tendering. The public call for tenders implies a requirement that they be competitive and individualised and is aimed at discovering objectively which producer or service provider can offer the best goods or services at the best price. Any form of surreptitious cooperation between tenderers subverts that goal and breaches the terms of the invitation to treat.[24]

 

89.              All forms of collusive tendering, therefore, typically have two things in common: i) the aim to mislead purchasers by creating the appearance of competition to conceal secretly inflated prices; and ii) an agreement which predetermines the winning bidder thereby restricting competition among the conspiring bidders.

 

90.              The Tribunal’s finding in this case in relation to the issue of collusion, and thus collusive tendering, is cryptic and devoid of coherent analysis – presumably because it ultimately disposed of the case on grounds of a lack of horizontality. Having concluded that the JVs and SGB-Cape could not be classified as horizontal competition, it nonetheless went on to opine that since the JVs were constructed for the specific purpose of the tender, with SGB-Cape as the single controlling mind, SGB- Cape could not collude with itself and that the granting of the mandate to Mr. Falconer could not be characterised as collusive tendering in terms of section 4(1)(b)(iii) of the Act.

 

91.              The Commission argued that once each JV granted Mr. Falconer (SGB-Cape) the right to compose its bid, they committed an act of collusion. The object of the impugned agreement was to distort competition in that it created a facade of four separate firms in competition with each other. Yet those firms had coordinated their prices through various acts and did not disclose the price coordination to Eskom, which the Commission said was deceptive by creating the illusion of disclosure where none existed. Taking into account the purpose of the overall scheme which was to ensure SGB-Cape, facing a threat to its business, retained work at a number of power stations, the collusion was aimed at retaining its market share. These actions, taken together, it submitted, were distortive of competition and established a contravention of section 4(1)(b)(iii) of the Act.

 

92.              The Commission's arguments are unpersuasive. As the respondents correctly argued, having regard to the nature and purpose of the impugned agreement, as well as its terms and the manner of the bids’ presentation, it is more than evident that there was no attempt at collusion on the part of the respondents. The purpose and rationale of the JV agreements, or, more pertinently, the mandates given to Mr. Falconer, were not to harmfully limit output or raise prices, but rather to provide an expanded offering addressing Corp313O's key non-price factors, being Eskom’s empowerment criteria. The purpose of the impugned agreement aligned in an important respect with some of the key purposes stipulated in section 2 of the Act, namely, to promote and maintain competition in South Africa in order inter alia to ensure that small and medium sized enterprises have an equitable opportunity to participate in the economy and to promote a greater spread of ownership, in particular, to increase the ownership stakes of historically disadvantaged persons.

 

93.              The nature and narrower purpose of the impugned mandate to Mr. Falconer, the respondents maintained, were predicated upon the assessment that SGB-Cape was the only entity with the full information required to complete the bid in a manner addressing the conditions of the tender. Whether or not that narrower purpose led the respondents into price-fixing is a matter to which I return later; suffice it now to say that the blatant, transparent and easily detectable manner in which the mandate was executed by Mr. Falconer militates against any intention to deceive Eskom – and thus the possibility of collusive tendering.

 

94.              Of particular importance in relation to the question of collusive conduct, the express joint discount offered in each bid made it clear that the four bids were related and submitted jointly on shared information. Each bid made explicit reference to the other bids twice. Section 17 of the Commercial Proposal of each bid cross-referenced the other bids and openly stated that the discounts were applicable to any of the offers made to Eskom which involved SGB-Cape, on its own, or in a JV with a JVP. This points to an absence of any intent to mislead Eskom about the nature of the proposals – a proposition with which Eskom, through Ms. Mzileni, agreed. Moreover, the discount offers were made in a context in which SGB-Cape (as the incumbent service provider to Eskom in relation to the advertised goods and services for more than a decade) had been in ongoing discussions about how best to meet Eskom’s BEE requirements. It had sought to do so to the satisfaction of Eskom through a variety of JV vehicles such as Octorex. The discount clause alone, understood in the context of the ongoing discussions with Eskom, permits a finding on the probabilities that there was no attempt at deceit or surreptitious dealing.

 

95.              Furthermore, SGB-Cape’s acknowledgement form of 17 March 2015 indicated that it intended to submit a stand-alone bid and three distinct JV bids with identified JVPs. SGB-Cape’s name was clearly depicted on all the bids, in the names of the JVs, their logos and letterheads and all relevant documents. Furthermore, a basic reading of the bids disclosed that there was a deliberate and express uniformity of the bids signed by one person, who was known to the relevant persons dealing with bids at Eskom, through previous ongoing dealings. There were identical SHE, quality, financial, commercial and technical documents; and the structures across the JV agreements were identical.

 

96.              Any reasonable person reading the bids, therefore, given their nature, form and content, would have seen the obvious coordination that had taken place. Ms. Mzileni on receipt of the bids immediately noticed the significant similarities between them and readily conceded during cross examination that nothing in the bid documentation was misleading. The bid documents all reflected the identity of the bidders and there was no apparent attempt to create a false impression. Ms. Mzileni had known Mr. Falconer for a number of years, had worked with him (and SGB-Cape as the incumbent service provider for over a decade) in the past and immediately saw and recognised that he had signed each bid on behalf of SGB-Cape and the three JVs. Yet she sought no clarification from Mr. Falconer, opting rather to suspect the worse, claiming at the same time that she had no right or obligation to seek clarification. That is not correct. Clause 20 of the Standard Terms and Conditions explicitly permits Eskom to obtain from a tenderer clarification of any matter in the tender which may not be clear. Had Ms. Mzileni done that, the object of the four bids may have been clarified and the complaint of collusion avoided.

 

97.              The Commission’s point that SGB-Cape ought to have made fuller disclosure in the four bids of the price coordination undertaken by Mr. Falconer is a good one. None of the bids expressly disclosed that the pricing of each bid was not independently calculated, or that their pricing was relative to SGB-Cape’s bid price as the base price. But it does not follow from the inadequate disclosure that the respondents engaged in surreptitious deception. The application of basic arithmetic would have revealed and did in fact show the difference in price and the reasons for it. There is accordingly an insufficient evidentiary basis to conclude that the perhaps feckless and incomplete disclosure was intended to conceal any price inflation or favour a predetermined winner among the competitors. There was no deception or harm.

 

98.              Moreover, as Mr. Malherbe explained in his evidence, the core logic of collusive tendering was absent here. There was no attempt by SGB-Cape to create the illusion of competition. There was no conspiracy to create anti-competitive effects or impact - such as bidding at a supra-competitive price or restricting quality or capacity. None of the mechanisms typically used to achieve the effects of bid-rigging was present. There was no intended clear winning bid. There was no agreement between the JVs and SGB-Cape to share profits of any successful bids with unsuccessful bidders. It was not possible to determine that any particular JV bid would be favoured as a result of the combination of price and non-price factors. The SGB-Cape bid was the best on price but known to SGB-Cape to be doomed to failure by virtue of it not meeting the black ownership requirements – suggesting the possibility that it was imprudently submitted as a reflection of the baseline from which the other pricing was done. The bids of the JVs on the other hand had a combination of factors that could not predictably guarantee success in securing the tender in any order of preference.

 

99.              As between the JVPs, each was unaware of the JVPs participating in other JVs. It is true that SGB-Cape was in possession of the full pricing information of each of the four bids. But it did not pass any pricing information from any of the JVPs to the other JVPs. And, as Manoim JP succinctly explains in the concurring judgment, there is no evidence of any harm to competition occasioned by Mr. Falconer being in possession of that knowledge.

 

100.          Therefore, the impugned agreement or conduct in this instance lacked the features normally associated with collusive tendering and did not cause any apparent harm. There is insufficient evidence to conclude that the respondents aimed to mislead Eskom by creating the appearance of competition to conceal secretly inflated prices or that any agreement existed which predetermined the winning bidder thereby restricting competition. The impugned agreement accordingly did not fall within the scope of the per se prohibition. In the premises, despite its erroneous reasoning, the Tribunal did not err in concluding that there was no collusive tendering in contravention of section 4(1)(b)(iii) of the Act.

 

Price-fixing

 

101.          Can the first impugned agreement be characterised as per se prohibited price- fixing under section 4(1)(b)(i) of the Act? The Tribunal made no definitive pronouncement on the matter. However, it reasoned that even if one were to assume that it was technically feasible to duplicate estimating and accounting resources that SGB-Cape utilised to develop the bids, it would not have been “commercially rational for a firm to establish four entities in which it is the controlling mind or majority shareholder, and then allow each one of them to bid independently in competition with each other, without having some influence in the pricing decisions of those firms, for the same piece of work or tender”. This reasoning discloses that the Tribunal did not consider the impugned agreement, the price coordination, to be price-fixing.

 

102.          Price-fixing is an agreement among competitors to raise, fix, or otherwise maintain the price at which their goods or services are sold. The per se prohibition of price-fixing targets schemes which force consumers to pay more for goods through the structuring of a market favouring providers or producers despite their not adding value. Price-fixing necessarily involves some form of coordination between competitors for the supply into the market of their respective goods or services with the design of eliminating competition regarding price. That is achieved by the competitors (often collusively) fixing their respective prices in some form, by setting uniform prices, or by establishing formulae or ratios for the calculation of prices, or by other means designed to avoid the effect of market competition on their prices.[25] Price-fixing can take many forms, including: i) establishing or adhering to price discounts; ii) holding prices firm; iii) eliminating or reducing discounts; iv) adopting a standard formula for computing prices; v) maintaining certain price differentials between different types, sizes, or quantities of products; vi) adhering to a minimum fee or a price schedule; and vii) fixing credit terms.[26] In many cases, participants in a price- fixing conspiracy will establish some type of mechanism to make sure that all potential competitors adhere to the agreement.

 

103.          The price-fixing proscribed by section 4(1)(b)(i) of the Act is “directly or indirectly fixing a purchase or selling price or any other trading condition”. Unlike collusive tendering there is no distinct requirement of an element of surreptitious dealing for an agreement or practice to constitute price fixing. Direct price-fixing can often be transparent, while indirect price-fixing will usually be circuitous or surreptitious.[27] While price-fixing inevitably involves price coordination by competitors, it does not follow that price-fixing has necessarily occurred whenever there is an arrangement between competitors that results in their goods reaching the market at a uniform price. So called “naked price-fixing” is always manifestly anti-competitive. Yet many pro-competitive horizontal arrangements may have an ancillary effect of fixing prices – such as the “hub and spoke” arrangements discussed by Manoim JP in the concurring judgment. The SCA acknowledged in ANSAC that price-fixing may “be limited to collusive conduct by competitors that is designed to avoid competition, as opposed to conduct that merely has that incidental effect”. Literalness when interpreting the phrase “price- fixing” can be “overly simplistic and often overbroad”.[28] An overly broad per se prohibition against price-fixing arrangements may end up unduly targeting pro- competitive arrangements and prudence may require distinguishing covert arrangements from overt arrangements and naked price-fixing from price-fixing which is ancillary to some pro-competitive or efficiency objective.[29]

 

104.          In ANSAC the SCA clarified the approach to the scope of the legislative prohibition against price-fixing as follows:

 

“There can be little doubt that an agreement by competitors that has as its specific design the elimination of price competition (the essential characteristic of a cartel) constitutes direct price-fixing as contemplated by the statute. Where competitors have reached an agreement to set uniform prices, without more, all that might be required in order to establish a transgression of s 4(1)(b) is to produce their agreement, because its very terms may admit of no conclusion but that it was designed to eliminate price- competition…But indirect price-fixing presents greater complexity. It is not difficult to envisage conduct by competitors that is designed to eliminate price-competition indirectly, by shifting the supply of competitors’ goods to a separate entity that is under their control, and which purports to set the price for the goods…But not every arrangement between competitors entailing the ultimate supply of goods necessarily falls into that category. It is, for instance, not difficult to envisage a bona fide joint venture that is embarked upon by competitors for a legitimate purpose, through the vehicle of a separate entity, which must necessarily set a price for goods that it supplies (emanating from the competitors) merely as an incident to the pursuit of the joint venture…There is in our view no a priori reason to assume that such an arrangement constitutes prohibited price-fixing as contemplated by s 4(1)(b) of the statute… [T]hen it might well be necessary to enquire beyond the mere terms of the competitors’ agreement in order to establish whether it is or is not merely a sham ”[30]

 

105.          The scope of the prohibition of indirect price-fixing, as with collusive tendering, therefore, requires consideration of evidence regarding the nature, purpose and effects or impact of the impugned agreement or practice. Thus, in characterizing the first impugned agreement the principal question is whether it had “as its specific design the elimination of price competition (the essential characteristic of a cartel)”, or if it was embarked upon by competitors for a legitimate purpose to set a price for the goods or services merely as an incident to the pursuit of a joint venture or similar arrangement. Once again, for the reasons already discussed, caution must be observed to avoid straying too deeply into rule of reason analysis.

 

106.          The Commission’s concern about price-fixing in the present case related to the pricing method and coordination in pricing each of the bids. Each JV separately authorized SGB-Cape and Mr. Falconer to set the bid prices. Mr. Falconer attended to the bid preparation and set each JV's bid price and mark-ups. SGB-Cape created its own base price with substantial mark-ups, and then added an additional mark-up for each JV. Each firm, through Mr. Falconer, participated in the setting of the prices of the other firms. This, in the opinion of the Commission, created an appearance of competition, while in effect, there was none.

 

107.          There appears to be no reason to object to the minority JVP in a JV assigning to the majority JVP a vertical authority to set prices for the services and goods it will input into the JV. The problem here though goes beyond that and is different. Here we have the majority JVP of one JV setting rates across other competitor JVs. On the face of it, this would seem to be price-fixing. But the question is whether it is price-fixing that deserves to be characterised as prohibited per se or is the conduct of such an order that it should have been subject to rule of reason justification under section 4(1)(a) of the Act?

 

108.          While the Commission conceded that only SGB-Cape could have made any input into the pricing of the key scaffolding and insulation goods and services provided, it contended that the JVPs were in a position to independently price elements of the labour costs. Each JVP had expertise and could provide inputs on the cost of significant labour components of the bid but was prevented from doing so by SGB- Cape’s determination of those costs. It essentially argued that had the JVPs been left to price independently, they could have priced more competitively and come up with a lesser amount in relation to certain of the elements, primarily the labour administration fee, the […]% mark-up on top of the total costs and the mark-ups of […] to […]% on the labour prices depending on the JV.

 

109.          The respondents convincingly answered these contentions in some detail. The purpose of the JV agreements and the mandate to Mr. Falconer was not to limit output or raise prices but rather to provide an expanded offering addressing the non-price factors, namely the BBE criteria. The unilateral formulation of the bid strategy and compilation of the bid documents, including the setting of the prices, by SGB-Cape for all the JVs was justified by the fact that it was the only entity capable of efficiently pricing the bids. The setting of the overall price of the bid was complex and entailed determining the total cost of the undertaking required for the tender and ensuring that a sufficient margin was built into each bid to recover costs and account for associated risks. The labour rates were regulated by industry bodies, such as the MEIBC, the relevant bargaining council with jurisdiction, or in terms of legislation and existing practice. Many of the other associated labour costs, the non-work wage package, the add on burdens of accommodation, medical, PPE etc. were simply passed through and in respect of which there was no discretion. In any event, the incentive of all parties would have been to keep all the passed through prices as low as possible, and SGB- Cape and the JVPs would have shared that intention. Thus, the only relevant costs that were not fixed or passed through were the mark-ups and the labour administration fee. The mark-ups were to recover overheads and the additional costs of the JVPs above the labour administration fee and to mitigate risks arising for administering and managing the JVs; and the labour administration fees were in fact independently negotiated with the relevant JVPs.

 

110.          As the Tribunal noted it would have been burdensome to set up independent processes in relation to the negotiation of these elements of the price and in some respects would have involved SGB-Cape undercutting its own bids in the other JVs. In the circumstances, the independent pricing process envisaged by the Commission would have achieved little from an efficiency perspective. To the extent that there may have been some scope for negotiation of the mark-ups, it is doubtful that the failure to have those negotiations outweighed the efficiency gains.

 

111.          Although this might have been better tested in a rule of reason analysis under section 4(1)(a) of the Act, from the perspective of characterisation it cannot be convincingly stated that the agreement was of a character devoid of all redeeming features so as to be prohibited per se. Any price-fixing this approach entailed was not covert “naked price-fixing” but was ancillary to the overt efficiency gains. The bids had positive effects aligned with the objects in section 2 of the Act regarding the promotion of opportunities for small and medium sized enterprises and a greater spread of ownership to historically disadvantaged persons.

 

112.          It bears repeating that SGB-Cape disclosed its involvement in the three JVs, albeit somewhat inadequately. Eskom was nonetheless aware of SGB-Cape’s involvement and any confusion regarding its strategy ought to have been clarified during the tender evaluation process. SGB-Cape was primarily intent on meeting the BEE criteria in the legitimate hope of obtaining more work. There was nothing wrong with SGB-Cape seeking to obtain as much work as possible by employing the strategy that it did. No independent alternative bid by other bidders in the market was prevented or impeded by the unilateral formulation of the four bids. There was also no harm to competition as no competitor or competitive constraint was removed because of the formation of the JVs or the pricing coordination.

 

113.          Furthermore, the conditions of Eskom’s invitation to treat provided that upon evaluation of the tenders received, contracts could be concluded with one or more tenderers; and Eskom was not bound to accept the lowest of any bid and reserved the right to enter into post tender negotiations with any one or more of the bidders. Therefore, the offers contained in the bids in response to Eskom’s invitation to treat did not fix any price immutably as there were further processes that would ultimately have determined the final price. The price-fixing or coordination involved here was thus qualitatively different to the ordinary kind of price-fixing where competitors covertly agree to set the prices of goods directly sold into the market. In the premises, the first impugned agreement did not constitute price-fixing and was not in contravention of section 4(1)(b)(i) of the Act.

 

The Mtsweni restraint

 

114.          The Tribunal did not err in holding that Mtsweni was not a potential bidder for Corp3130, and that the restriction placed on Mtsweni did not have the effect of removing a competitor from bidding on the tender. Mtsweni had no previous experience or track record in the type of scaffolding and insulation work requested in Corp 3130. Nor did it have the necessary certifications. There were no realistic grounds to assume that Mtsweni had sufficient resources to enter the market and was thus a potential competitor on its own.[31] There was accordingly no horizontal relationship. Mtsweni would not have been eligible to bid as the lead partner in a JV as it could not deliver all the services and there is no evidence of any kind that Mtsweni had an alternative JV opportunity in which it could have been the minority JVP. Moreover, the second impugned agreement was a legitimate restraint in the circumstances in that it aimed rationally at ensuring Mtsweni’s efforts were directed appropriately to the JV of which it was a JVP and not in competition with it.

 

Costs

 

115.          The respondents maintain that the Commission did not serve the public interest when it engaged in hopeless litigation that has had significant ramifications for the respondents. It argues that the entire case was essentially a waste of resources and that a costs order would bring some measure of discipline and restraint where it is needed.

 

116.          While there may be some basis to the contention that the Commission was somewhat zealous in pursuing this matter, the arrangement which the respondents concluded was indeed unusual and on the face of it had an element of price coordination and price-fixing that justifiably required investigation and adjudication. In the circumstances, fairness does not require a costs order against the Commission.

 

Murphy AJA

 

Manoim JP (Siwendu AJA concurring) (separate concurring decision)

 

Introduction

 

117.          I have read with great interest the very thorough judgment of Murphy AJA. He has captured the essence of the issues the case raises, and I can do no better than focus on one issue where I see a simpler path to deciding the matter, although I come to the same outcome that he did, that the appeal should not succeed.

 

118.          I have referred to the respondents in the same way as Murphy AJA has; SGB- Cape (is the first respondent), the JVs to refer to the second, third and fourth respondents) and where I refer to the minorities in these JV’s, the fifth, (Tedoc), sixth (Superfecta) and seventh respondents (Mtsweni) I refer to them as the JV partners or JVPs.

 

119.          I pick up from one of the last remarks Murphy AJA makes in the costs section. He notes that the arrangement that the respondents had in place was unusual. I agree with this observation and that seems to go to the heart of the matter in analysing whether the arrangements infringed section 4(1)(b) of the Competition Act 89 of 1998, (the Act).

 

120.          The Tribunal decided the matter on the basis that there was no horizontal arrangement because SGB-Cape as controller of all four bids, could not compete with itself. That, although not stated, is the application of the Copperweld doctrine where it was held that a firm could not compete with itself.[32]

 

121.          I do not consider the Copperweld doctrine would apply on the facts of this case. The JV partners were not analogous to wholly or controlled subsidiaries of SGB-Cape. Prior to the tender they were separate autonomous entities, and the tender did not change this. They continue to be separate autonomous entities. Nor were the JV’s mere shams and in this respect, I would agree with Murphy AJA.

 

122.          SGB-Cape has argued that the solution is to rely on characterisation, a doctrine that this court has accepted since the ANSAC decision[33]. However, that pre-supposes the presence of a horizontal arrangement or what facially appears to be one, and I do not believe this case needs to rely on characterisation if the feature of a horizontal agreement is absent, as I go on to find is the case here.

 

123.          Murphy AJA finds that there is a horizontal agreement but that it lacks the features normally associated with collusive tendering (the section 4(1)(b)(iii) count, and that in respect of the section 4(1)(b)(i) count, the price fixing was “qualitatively different to the ordinary kind of price-fixing where competitors covertly agree to set the prices of goods directly sold into the market.”

 

124.          He gets to the existence of a horizontal agreement by applying the common law approach to partnerships. Here the component parts of the JV are subordinated to the overall partnership constituting the firm in question. He makes a most elegant argument for this position, but in my view the case can be decided more simply on the basis that the Commission has not established the prerequisite for any case under section 4(1)(b) to succeed namely the existence of a horizontal agreement. That is because I approach the case by considering the component parts of the JVs – the joint venture partners as the primary vehicles for the substance of the analysis, whilst the form it took viz. joint venture partnerships, should be kept subordinate to this in analysing the competition implications of the arrangements.

 

Commission’s case

 

125.          The reason this case has proved so difficult for the Commission to prosecute is that it lacks any coherent theory of harm normally associated with collusive tendering or price fixing as Murphy AJA ‘s decision makes clear. What is more unusual is to have a collusive tendering case, also have as an alternative charge, price fixing. Experience shows that where one is dealing with a case concerning the alleged suppression of competition amongst bidders, if the authority cannot establish that the firms were engaged in collusive tendering it is unlikely to succeed in establishing that they were, in the alternative, engaged in price fixing. If the were, this would be collusive tendering.

 

126.            I deal with the pricing harm first. Here the Commission relies for its case on harm; to compare what happened with the tender, in contradistinction to what might have happened in two counterfactual situations had the bids been constructed differently. But the two candidate counterfactual theories of harm make no sense on the facts.

 

127.          The first was that the bids could have been prepared in a non-collusive manner, with each bid being sealed from the content of the others. Here the Commission posited the possibility of three separate bids, albeit completed by SGB-Cape personnel, but with a Chinese Wall erected between them so each bid was the product of an independent SGB-Cape team. The Tribunal correctly rejected this as impractical and lacking commercial reality. The second theory was a single bid in which all three joint venture partners had a stake. This doubtless meant that SGB would have to dilute its stake but there is no guarantee that it would have been willing to do so.

 

128.          But regardless of the commercial practicality of these alternatives there is no evidence that had the bids been restructured in any one of these ways, this would have led to lower prices for the four bids.

 

129.          The second theory of harm advanced by the Commission related to the collusive tendering charge under 4(1)(b)(iii). Here the argument was that by supplying multiple bids (four in number), the JV’s and SGB had attempted to ‘crowd-out’ from consideration, other bids which may have proved more competitive. Whilst ‘crowding- out’ may well be an acceptable theory of harm in a collusive tendering case, there needs to be a factual basis to it. For instance, that the tenderer had decided to limit the number of bidders it would consider, for the tender and that failing the achievement of its minimum number, it might decide not to proceed with the tender or re-open it. Thus, a bidder concerned that there were insufficient bids which might lead to the customer opening a new tender, might submit additional ‘sham’ bids so that the customer had the illusion that it had sufficient number of genuine bids and would not cancel the tender.

 

130.          This was the position in the US case of Reicher on which the Commission relied.[34] But in Reicher the court found that the customer who had put out the tender required two firms to bid, otherwise it would withdraw the tender. Aware of this Reicher arranged another company, which potentially could have bid, but had at that stage chosen not to, to give it (Reicher) a signed blank bid which the latter then sent in. This kept the tender alive. On appeal the court held that:

 

“….in a bid rigging conspiracy, the determination of a per se antitrust violation depends on whether there was an agreement to subvert the competition, not on whether each party to the scam could perform. The Lab was obliged to consider all the bids it received, and entitled to assume that bids actually submitted were bona fide. Had it not received Giolas' bid it necessarily would have changed the OCA specifications in such a way as to obtain competition for a rebidding process, or it would have negotiated a contract with Reicher's company. In a negotiated contract it would have scrutinized the costs in a manner presumed to be unnecessary when there are competitive bids. Thus, notwithstanding Giolas' undisputed incapacity, through their agreement Reicher and Giolas were able to manipulate the bidding and lull the Lab into the belief it had the benefits of true competition. Having bid on the job and having created the appearance of legitimate competition in an open bidding process, they cannot now escape the inevitable conclusion of dirty dealing by denying that they were competitors.” [35]

 

131.          However, there is no such evidence in the present case. Eskom did not place a ceiling on the number of bids it would accept and in fact many firms bid. There is no evidence that Eskom would have withdrawn the tender if insufficient bids were submitted or that this was the purpose of the SGB-Cape’s strategy of submitting four bids.

 

132.            It is because it has no plausible theory of harm that the Commission has struggled to articulate the terms of the alleged horizontal agreement in this case.

 

133.          In its final heads of argument presented to this court, the Commission argued that the horizontal agreement was constituted both expressly and tacitly. It was expressly created through the mandates given by each JV partner to SGB to complete their bid documents which included prices. The tacit agreement amounts to the same thing. On this argument each JV by allowing SGB to prepare bid documents that included prices entered into the impugned horizontal agreements.

 

134.          But it is not clear on what basis the respective vertical agreements each JV partner had with SGB, had a horizonal component. The Commission in these heads makes clear that it does not consider that the creation of the JV’s breached section 4(1)(b), nor the fact that SGB held a majority interest in all four bids. Rather, its focus is on the fact that post the creation of the JVs, SGB was given the price mandate. But this separation of before and after is artificial. Neither of the passages from the record that the Commission cites in furtherance of the proposition supports this bifurcation of the mandate into pre- and post-tender.

 

135.          However, I accept that there is a line of authority in both US and European law, that recognises that parties can enter into a horizontal agreement, indirectly, through a mandate given to a third party. In this case SGB-Cape would be the third party. These agreements are now termed ‘hub and spoke’ agreements, although as pointed out in the literature, they were recognised as horizontal long before the label was attached.[36]

 

136.          A clear explanation of the issue is set out in a submission made by the European Union to an Organisation for Economic Co-operation and Development (OECD) forum discussing hub and spoke arrangements:

 

“Hub-and-spoke arrangements presuppose at least two information exchanges: one between spoke A and hub B, and another one between hub B and spoke C. Such vertical information exchanges, even when relating to strategic information, may often be legitimate and thus legal business practices. The difficulty, therefore, is in identifying the situations in which such legitimate practices amount in reality to an illegal horizontal agreement or concerted action. It is therefore of particular importance to assess the context behind the information exchanges between spoke A and hub B and between hub B and spoke C. This assessment is case-specific and will notably have to take into account the level of awareness of the spokes regarding the information exchanges between other spokes and the hub and whether they have distanced themselves from the behaviour of the hub. This raises the question as to the requisite standard of proof for horizontal collusion in such cases and notably whether it is necessary to show that the spokes know that the information exchanged with the hub is being passed on further to other spokes.”[37]

 

137.            In a hub and spoke arrangement the spokes (in this case the JV partners) do not need to be shown to have agreed directly with one another. Rather, it suffices that a mandate is given to the hub (in this case SGB-Cape) which facilitates an agreement by exchanging competitively sensitive information with another competitor. In this case this competitively sensitive information was pricing.

 

138.          As Whish and Bailey note in their treatise there is one case where such a finding has been made:

 

“In SIA Remonts v Konkurrences padome[38] the [European] Court of Justice established that an undertaking, X, could be liable for collusive tendering where a third party service provider Y, brought about this infringement by exchanging commercially sensitive information with competitors of X if X could reasonably have foreseen that Y would share that information and X was prepared to accept the risk that this entailed.”[39]

 

139.          But the only party in possession of the sensitive pricing information was SGB- Cape. It did not glean any sensitive pricing information from any of the JV partners that it passed on to the others. To the extent that different management fees were charged there is no evidence that SGB-Cape exchanged this pricing amongst the JV partners.

 

140.          The evidence is that the bids went in without the JV partners having any knowledge of the other bids. Thus, on the pricing theory of harm, even applying a hub and scope analysis, there was no exchange of confidential information, express or tacit, and hence no evidence of any horizontal agreement between the JVs. Or as was succinctly put in one of the US cases, it was a hub without a rim.[40]

 

Analysis

 

141.          The probabilities in this case favour the version that SGB-Cape embarked on a strategy not to collusively rig the tender but to present itself in the best position to win back power stations that it had historically lost. As Murphy AJA outlines in his history section, SGB-Cape had in the past been the largest incumbent service provider to Eskom, servicing 9 out of 11 power stations. Gradually over time, due to the social imperatives facing Eskom on procurement this position declined. At the time of the tender, SGB-Cape had been reduced to four stations.

 

142.          When Eskom decided in 2015 to issue tender for its 15 power stations, SGB- Cape was faced with a dilemma – it might lose its current power stations due to its unfavourable BBEE shareholding, whilst at the same time it sought to restore some its past glory by winning more power stations from its present four, out of the fifteen on offer in the tender. What SGB-Cape believed was that if Eskom was driven to make its award premised on fulfilling or being seen to fulfil its social imperatives, then it should design its bid accordingly. Recognising that three categories had been signalled, its calculation was that if it could present a bid that met each category. At the same time, it offered a sweetener – if Eskom accepted more than one bid it would discount the prices of its bids.

 

143.            Although the JVs are constructed as partnerships, that area of law is not the way to analyse the issue which is where I must respectfully part ways with Murphy AJA careful analysis of its implications.

 

144.          While it is correct that the definition of a firm in the Act ‘… includes… a partnership” that does not negate the need to evaluate the competition consequences of the constituent firms who may constitute a partnership for bidding purposes. If that was not the case competitors could bid jointly and hide behind the cloak of partnership to disguise a collusive design. This is not to say that the JV’s are a sham. I agree with Murphy AJA that they are not. Only that it is necessary to analyse what role the constituent members of the respective JV’s may have played in relation to the tender.

 

145.          This is not to engage in piercing the corporate veil as it would be in a company law case. Rather it is part and parcel of competition analysis to examine incentives. For this reason, de-constructing a commercial arrangement to see what the incentives are, is a necessary part of this analysis. This is why merger control emphasises control and engages in looking at a control structure until the ultimate controller is identified.[41] But his approach is not confined to merger control. It is also why section 4(1) refers to an ‘association of firms’. It is because this recognises that the constituent parts of an association which might present as a single legal person comprises competitors as its members. In Copperweld the court in rejecting what it termed the intra-enterprise conspiracy theory (that a company could be found to conspire with its subsidiary remarked: “The intra-enterprise conspiracy theory looks to the form of an enterprise’s structure and ignores the reality” ... and then stated:

 

“If antitrust liability turned on the garb in which a sub-unit was clothed, parent corporations would be encouraged to convert subsidiaries into unincorporated divisions”.

 

146.          Whilst in Copperweld the court looked at substance over form to avoid over enforcement of the antitrust laws, so one must avoid the same error in under enforcement. For this reason, it is necessary to the analysis to look behind the partnership of the JV’s to examine their constituents.

 

147.          Analysing the facts from this perspective the following are relevant. The JV partners who were constituent parties in each JV, other than SGB-Cape, are not competitors in the same line of industry as the latter. Individually none of them could be bidders for the tender. They are labour brokers not equipped with the necessary experience or skills to compete in that market. Thus, including any of them in a bid does not give rise to bid suppression of a potential competing bid.

 

148.          Second the JV partners did not have access to any of the confidential pricing information. All that information was SGB-Cape’s and so its use in the multiple bids had no spoke and hub consequence in the sense envisaged in VM Remonts.

 

149.          Thirdly, the evidence is that the JV partners had no knowledge of any bid other than their own. At most they were aware from the letter sent by SGB-Cape to each that it was planning to also bid with other partners. Who they were and what that entailed was not revealed to them. The only JV partner who testified, said she was not aware of any of the other bids until the Competition Commission commenced its probe.

 

150.          From this it is clear that the mandate given to SGB-Cape was only vertical. There is no evidence that the mandate had any horizontal implication. Nor on the facts can it be inferred that there was an implied horizontal agreement.

 

151.          Without a horizontal agreement there can be no finding against the respondents in terms of either section 4(1)(b)(i) or4(1)(b)(iii).

 

152.            As far as the Mtsweni restraint and the costs are concerned, I am in full agreement with the conclusions of Murphy AJA on these issues and have nothing further to add.

 

153.          I concur with Murphy AJA that the appeal must accordingly be dismissed with no order as to costs.

 

The following order is made:

 

154.          The appeal is dismissed.

 

155.          There is no order as to costs.

 

Manoim JP

 

Siwendu AJA

 

DATE OF HEARING:          1 February 2024

DATE OF JUDGMENT:       26 April 2024

 

FOR APPELLANT:             Adv T Ngcukaitobi SC Adv H Drake

 

INSTRUCTED BY:              Ndzabandzaba Attorneys Inc.

 

FOR RESPONDENTS:        Adv M Le Roux SC

                                         Adv K Turner

 

INSTRUCTED BY:              Werksmans Attorneys

---