CT CR134Feb09 · case CR134Feb09
Case No: 23/CR/Feb09 In the matter between: The Competition Commission Complainant/Applicant And Southern Pipeline Contractors 4° Respondent Conrite Walls (Pty) Ltd 2°" Respondent Panel : Norman Manoim (Presiding Member), Yasmin Carrim (Tribunal Member) Takalani Madima (Tribunal Member) Heard on : 2 -3 August 2010 Order issued on : 29 November 2010 Reasons issued on: 29 November 2010 Reasons for Decision and Order Introduction [1]On 13 February 2009 the Commission filed a complaint referral against Rocla (Pty) Ltd, Southern Pipeline Contractors (Pty) Ltd, Concrete Units (Pty) Ltd Aveng Africa Ltd, Grallio (Pty) Ltd, Cobro (Pty) Ltd, Cape Concrete (Pty) Ltd, Contite Walls (Pty) Ltd, Craig Concrete Products (Pty) Ltd and D&D (Pty) Ltd’ alleging that they had contravened sections 4(1)(b)(i), 4(1)(b)(ii) and 4(14)(b) (iii) "D&D (Pty) Ltd was acquired by Rocla in 2006.
[2]
[3]
[5]of the Competition Act ( “the Act”). The Commission accused them of fixing the selling price and dividing the market for the production of pipes, culverts and manholes. It also alleged that the respondents had engaged in collusive tendering in respect of the supply of precast concrete products to certain suppliers. The Commission asked the Tribunal to order that the Respondents desist from such conduct and for the second to ninth Respondents to pay an administrative penalty equivalent to 10% of each of the Respondents’ annual turnover for the preceding year. This decision relates to only two of the respondents namely Southern Pipeline Contractors (Pty) Ltd and Conrite Walls (Pty) Ltd. Both respondents admitted that they had contravened section 4(1)(b) of the Act but they opposed the Commission’s prayer that we impose a penalty amounting to 10% of annual turnover. The following witnesses were called: Francois Myburgh, General Manager for infrastructure products in the Infraset business e Danie! Michael Greeff, Regional Sales Manager Coastal of Rocla (Pty) Ltd e Stephane Riot, previous General Manager of Southern Pipeline Contractors (Pty) Ltd » Paul Phipps, current General Manager for Southern Pipline Contractors (Pty) Ltd e Arno Louw, Director of EksteenLouw&Badenhorstinc © Donald Robert Speirs , Managing Director of Conrite Walls (Pty) Ltd Background During December 2007 Rocla filed an application for leniency in terms of the Commission’s corporate leniency programme with the Competition Commission regarding its involvement in a cartel in the pre-cast concrete market in South i7] Africa. Rocla, in its application, informed the Commission that it, together with the nine Respondents mentioned above had engaged in conduct whereby it fixed the selling price and divided the market for concrete pipes, culverts and manholes. It also admitted to collusive tendering in respect of the supply of precast concrete products to certain suppliers. The Commission initiated an investigation into the alleged cartel activity on 19 March 2008 and found that the cartel had operated from 1973 until 2007 in Gauteng, KwaZulu-Natal and the Western Cape. !t accordingly filed a complaint referral with the Tribunal on 13 February 2009. Shortly after the Commission filed the complaint with the Tribunal, four of the Respondents entered into consent order negotiations with the Commission. The first to settle was Aveng. It admitted that its subsidiary, Infraset, contravened the Act and agreed to pay a penalty of R46 277 000, being 8% of the turnover attributable to Infraset in the previous financial year, excluding the turnover attributable to paving products. The Commission concluded the second settlement with Concrete Units (Pty) Ltd. It agreed to pay a penalty of RS 763 743, representing 7% of Concrete Units’ turnover for the financial year ending 2008. The third party to settle with the Commission was Cobro Concrete. it agreed to pay a penalty of R4 022 568 which represented 6.5% of its turnover for the financial year 2008. Cape Concrete was the last to conclude a settlement agreement. During August 2010 it agreed to pay a penalty of R4 371 386, representing 7% of the turnover attributable fo Cape Conrite in the 2008 financial year. Southern Pipeline Contractors (“SPC”), Grallio (Pty) Ltd (‘Grallio”) and Conrite Wails (Pty) Ltd did not arrive at a settlement with the Commission. SPC and Conrite Walls acknowledged that they had contravened section 4(1)(b) of the Act but contested the quantum of the penalty sought by the Commission. Grallio denied that it had been involved in cartel activities and sought to oppose the Commission's referral. In light of this the Tribunal ordered a separation of the cases, with SPC and Conrite being dealt with in one matter and Grallio in the other. On 2 and 3 August 2010 this panel heard evidence and argument on the case involving SPC and Conrite. | Grallio’s case was allocated to a different panel.” "Heard on 5, 6 and 12 August 2010 General features of the Cartel
[8]
[10]
[11]The evidence in this case reveals that this was one of the most enduring, comprehensive and stable cartels prosecuted to date. Not only did this cartel precede and succeed the Competition Act, it involved collusion on almost every aspect of section 4(1)(b). Features of the cartel included price fixing, market share allocations, product and geographic market division, customer allocations, bid rigging and inter party terms of trade. It involved collusion at both national and regional levels and operated in such secrecy that members were referred to by number and not name. As a result competition was adversely affected in many markets. Cartel members enjoyed a quiet and hugely profitable life, as evidenced by the drop in prices by between 25-30% post the disbandment of the cartel, some for more than 40 years, at the cost of consumers and taxpayers alike. The cartel operated in the construction industry in the market for the manufacturing of pre-cast concrete products such as concrete pipes, culverts, pre-cast manholes and concrete sleepers at both national and regional levels.° The national cartel commenced operations in 1973 and was founded by Rocla and Infraset (now Aveng), previously known as Fraser Fyfe. Over a period of time and with the influence of its founding members the cartel grew until it included not only large players but also smaller companies who had a reputation for pricing aggressively in the market. It ceased operations in 2007 during which period Rocla applied for leniency from the Competition Commission. Apart from the national cartel, regional cartels were established in Gauteng, Western Cape and KwaZulu-Natal (KZN). The cast of players differed slightly in the regions depending on a particular manufacturer's geographic footprint. However the two key players, Rocla and Infraset were present in each region. Attached hereto as annexure A is a document belonging to Rocla which shows Concrete pipes are used for storm water drainage, sewers and other construction related products. Culverts are conduits used to convey water from one area to another, usually from one side of the road to the other. Precast manholes are working chambers installed at intervals along a pipeline to provide access to a pipeline for inspection and maintenance. Precast concrete sleepers are used in the rail and rail-related industries.
[42]{13]
[14]the players in each region and the arrangements that prevailed among them in February 2002.4 Cartel members met regularly for purposes of sharing information on contracts and, when necessary, to discuss prices. In Gauteng meetings usually took place after the Concrete Manufacturers’ Association meetings on the second Tuesday of each month. In other provinces meetings were not as regular and were held on an ad hoc basis at different venues. One member acted as a banker. The banker’s role was to compile a list of all the contracts available during a specific period.® The information was used to divide contracts among members while also monitoring the number of contracts allocated in order to ensure that they did not exceed their allocated market share in an area. Prices for products were agreed and reviewed twice a year, usually after increases in the costs of raw materials were introduced. Different price lists were also compiled for members as a cover-up. When tendering, the member who was allocated the contract would offer the contractor large discounts while the other members would not offer any or lower discounts in order to ensure that the chosen member won the contract. Members were limited to producing only certain products in certain regions. SPC, for instance, could only produce storm water and sewer pipes with a 300 mm to 1800 mm diameter in Gauteng. Members would also go to extremes to ensure that a firm stayed out of a product market, for instance, D&D and Cobro made monthly payments to Conrite Walls for the so-called “rental” of manhole moulds. Members of the carte! were also allocated geographic markets within which they were allowed to compete. For instance, in Gauteng and KZN the designated areas where members could “compete” were within a 150 km radius around Johannesburg and Durban. Only Rocla was allowed to tender for contracts outside these boundaries. Adherence was monitored by requiring all cartel members, except Rocla, to declare any deliveries outside their designated areas at the monthly meetings. 4 Trial bundle 371, Infraset is in all but the N Province. SThis information was supplied by other cartel members or was advertised in newspaper ads.
[15]
[16](17) Cartel members could not increase their market shares through competition but were allowed to expand through acquisition of an existing competitor. This reinforced Rocla’s stranglehold on the national market. The nature of this cartel is precisely the kind of behaviour the legislature contemplated when it created a per se offence in section 4(1)(b). The harm caused to competition, the economy and consumers alike by cartels of such long duration is difficult if not impossible to measure. This matter also demonstrates the success of the Commission’s leniency policy. But for that policy, the secretive operations of this cartel ~ and others like it - would not have come to light. It is against this background that we are asked to evaluate the culpability of SPC and Conrite. Approach to section 59 {78}
[19]As stated above, SPC and Conrite admit that they have contravened section 4(1)(b)(i), (ji) and (ii). The only issues that remain to be decided are the extent of their involvement, although this is largely common cause, and then the appropriate sanction to be imposed, although it is common cause that an administrative penalty is appropriate. Before we turn to consider their respective roles in the cartel, it would be useful for us to consider the submissions made to us in relation to the application of section 59. Section 59(1)(a) provides that this Tribunal may impose an administrative penalty for a contravention of section 4(1)(b). Section 59(2) provides that a penalty contemplated in sub-section (1) may not exceed 10% of the firm’s annual turnover in the Republic including its exports from the Republic during the firm's preceding financial year. Section 59(3) provides that when determining an appropriate penalty the Competition Tribunal must consider factors such as the nature, duration, gravity and extent of the contravention, any joss or damage suffered as a result of the contravention, the behaviour of the respondent, the market circumstances in which the contravention took place, the level of profit derived from the contravention, the degree to which the respondent has co-operated with the Commission and the Tribunal and whether the respondent has previously been found to be in contravention of the Act.
[20]{21]
[22]The approach taken by this Tribunal thus far to the imposition of an appropriate penalty was set out in Competition Commission v Pioneer Foods (Pty) Lid. ° in that case the Tribunal stated that the primary objective of imposing an administrative penalty was the promotion of deterrence. It moved away from a formulaic consideration of the weighting of the factors listed in section 59(3) on the basis that meaningful distinctions need to be made between different contraventions. Not all the factors listed in section 59(3) would be present in every case or in all types of contraventions nor would they be of equivalent weight in every case. In all its previous decisions where it has imposed an administrative penalty the Tribunal has established what the affected turnover was first and then used that as the ceiling for the imposition of the penalty. This approach came in for criticism by the parties in this case. The Commission argued that there is no legislative basis for the Tribunal to limit the imposition of an appropriate penalty only to the affected turnover’ of a firm as it had done in Pioneer and other previous cases. It submitted that the language of section 59(2) is clear and provides that it should be the firm’s annual, in other words, total turnover that must be the subject of the sanction. Moreover, the Commission argued, the Tribunal had erred in its approach in the Pioneer’ case when it stated that it would require some evidence that a firm’s monopolisation efforts in one product market provided it with some leverage in another before looking to the firm’s total turnover. The language of “leveraging” is used in abuse of dominance cases and conceptually ought not to creep into an analysis of section 4 contraventions. It argued further that our approach may encourage firms to place their business activities in divisions in order to minimise their exposure to sanctions under the Competition Act. (This last submission is rather over-stated. Firms place their business activities in divisions for their own efficiency and profitability motives. Foolish indeed would be a firm to place its operations in multiple divisions simply to limit its exposure to sanction under the Act. It would be cheaper for such a firm to not to contravene the Act.) 6 Tribunal Case No: 15/CR/Feb07 and 50/CR/May08 7 Affected turnover means that turnover of a firm derived from the market or markets or line of business in which the contravention is found to have occurred, in other words the relevant turnover in a particular case. The notion of “affected turnover” does not refer to the turnover of the firm that arose from specific transactions pursuant to the unlawful conduct, but rather to the firm’s turnover arising from the product line(s) in which the contravention occurred. 5 See footnote 6
[23]
[24]
[25][26} The Commission argued that where anti-competitive conduct has benefited firms for more than a decade the approach taken by the Tribunal in Pioneer appears to be under-deterrent in relation to duration. Section 59(2) contains two limitations. It provides that a penalty imposed in terms of section 59(1) may “not exceed 10% of the firm’s annual turnover in the Republic and its exports from the Republic” (first limitation) “during the firm’s preceding financial year” (second limitation). The latter limitation has been interpreted to mean the turnover in the financial year immediately preceding the year in which the contravention has occurred. This means that the turnover in question is limited only to one year’s turnover. An obvious deficiency in this formulation is that it does not take into account contraventions of long duration.® In closing argument the Commission proposed that the Tribunal could utilise affected turnover as a basis for computing the penalty but need not limit the percentage to 10%. A penalty could be as high as 50% of affected turnover provided the final amount did not exceed 10% of total turnover Ms Le Roux, appearing on behaif of Conrite, argued that the Tribunal had erred in its approach by conflating the provisions of section 59(3) with that of 59(2). The correct approach, she argued, is that the Tribunal must first determine whether a penalty is to be imposed under section 59(1). Then it should determine an “appropriate penalty” taking into account the factors listed in section 59(3). After having done that and arriving at a number it should then look to see if the fine thus determined exceeded 10% of the firm’s annual turnover as described. If it did so exceed the fine should be capped at 10%. If not then that is the appropriate penalty. Ms Le Roux appears to be advocating an EU approach to the calculation of the penalty yet, as we note later, Conrite does not apply this approach to its own circumstances. "° Before we turn to the specific circumstances of this matter and a discussion of our approach to section 59, it would be useful to consider approaches in other jurisdictions. A review of some of these reveals that the issue of calculating ° The Commission and Tribunal have in the recent submissions on the Competition Amendment Act sought an amendment to this section for precisely this reason. ‘© tad Conrite consistently followed the EU approach it would have required it to multiply the affected turnover by the number of years of duration of the cartel, instead as we show later it chooses an ad hoc approach of repaying its gains from a pay off from competitors for staying out of the market and adding a 25% premium because as Ms Le Roux herself conceded it would be otherwise under deterrent.
[27]
[28]{29] penalties is not an easy one and differences in approach may be found between enforcement agencies on the one hand and adjudicative agencies on the other. ICN" The International Competition Network (ICN) has compiled a report on the experiences of ICN members in the fight against cartels.'? In general all members agree that the primary objective of an effective penalty is to deter cartel behaviour and secondly, to take away the financial gains that otherwise accrue to cartel members. But these objectives are not so easy to satisfy because it is in most cases difficult to accurately assess and prove the benefits derived from cartel activities. Moreover, imposing very high fines might for example bankrupt companies thereby harming competition while too low fines would not act as a deterrent for other potential cartels. According to the ICN document the most commonly used method to calculate the fine is the determination of a basic penalty or base fine which is then adjusted taking into account mitigating or aggravating circumstances which might be expressed as a multiplier or as the deduction or addition of certain percentages of the base fine. The base fine is calculated by having regard to the affected or relevant turnover. Duration and gravity are the most important criteria. Cartel infringements are classified as minor, serious and very serious. Hard core cartels, such as section 4(1)(b) contraventions, are usually very serious. European Commission In 2006 the EC adopted new guidelines to calculate fines.* The previous guidelines were referred to as the 1998 guidelines. The 1998 guidelines categorised fines according to minor, serious and very serious conduct with a corresponding wide range of basic fines in each category. The aim of the 1998 guidelines was to create transparency in the calculation of fines by the 11 -The International Competition Network is a network of established and newer competition agencies representing about 104 agencies around the world. ? Report prepared by ICN Working Group on Cartels: Defining Hard Core Cartel Conduct, Effective Institutions, Effective Penalties, ICN 4" Annual Conference Bonn, Germany, 6-8 June 2005 5 ‘The Court of First Instance and the Court of Justice are yet to hear a case in which the Commission has used the new 2006 guidelines to calculate a penalty.
[30]Commission so as to stop a growing trend of reduction of fines by the Court of First Instance. The 2006 guidelines focus on the guilty firm’s sales in the relevant market affected by the infringement. They seek to promote a greater deterrent effect and were designed to act as a specific deterrent to infringing firms and a general deterrent to others who are considering entering illegal practices. In order to achieve this the Commission introduced an additional ‘entry fee’ factor to the fine of between 15 — 25% of a firm’s sales in the relevant market, “4 By way of example, to calculate the basic amount of a fine the European Commission wil: [30.1] Take the value of the firm's sales in the last full business year and before tax of goods/services to which the infringement directly or indirectly relates in the relevant geographic area of the European Economic Area, say for example Euro 100 million [30.2] Reflect the gravity of the infringement by taking a proportion, say generally up to 30% of the value of sales as a starting point depending on an assessment of the relevant circumstances of the case (which would include the nature of the infringement, the combined market share of all the firms involved, the geographic scope of the infringement, whether or not it was implemented ~ a global cartel involving most of the players in the market will result in a percentage at the higher end of the scale) for example 30% x Euro 100 million= Euro 30 million. [30.3] Reflect the duration of the infringement by multiplying the amount representing the proportion of the value of sales arrived at in sub- paragraph 2 above by the number of years of participation in the infringement for example 5 years participation x Euro 30 million = R 150 million. [30.4] Include the additional ‘entry fee’ of between 15 — 25% of the value of sales calculated in (1) above where the firm has engaged in cartel 4 The 2006 guidelines are closer in spirit to the US Sentencing Guidelines, which similarly base fines on a proportion (20%) of a firm’s sales in the affected market for the duration of the infringement, adjusting that figure upwards or downwards to reflect culpability, the nature of the conduct, the firm’s involvement in the infringement and their subsequent cooperation with the authorities. 5 Example taken from an information guide by Olswang: New European Commission Guidelines on Competition Fines 10
[31]
[32]
[33]
[34]activities for example 25% x Euro 100 million = Euro 25 million to be added to the Euro 150 million = Euro 175 million (‘basic amount’). The Commission will then adjust the basic amount of the fine upwards in case of aggravating circumstances and downwards to reflect mitigating circumstances. However, if a company terminates the infringement as soon as the Commission intervenes this will not be regarded as mitigating circumstances. If the firm is a repeat offender the basic amount will increase by 100% for each infringement thus if two previous findings of involvement in cartels the fine will increase with 200%. if it is a very large firm with a particular large turnover beyond the sales of the goods/services to which the infringement relates, the Commission may increase the fine to ensure it has sufficiently deterrent effect. The Commission may also increase the amount of the fine in order to ensure it exceeds the amount of gains improperly made as a result of the infringement. The Commission will reduce the fine if it exceeds the legal maximum of 10% of a firm’s total worldwide turnover in the preceding business year. In exceptional cases the Commission may also take into account a firm’s ability to pay where the imposition of the fine would irretrievably jeopardise the economic viability of the firm concerned. '° The 2006 guidelines are closer in spirit to the US Sentencing Guidelines, which : similarly base fines on a proportion (20%) of a firm’s sales in the affected market for the duration of the infringement, adjusting that figure upwards or downwards to reflect culpability, the nature of the conduct, the firm’s involvement in the infringement and their subsequent cooperation with the authorities. 6 Massimo Moto argues that it is a myth that fines will force firms into bankruptcy because the existence of a 10% of worldwide turnover upper threshold on fines limits the possibility that an efficient firm may be forced to exit because it has to pay a fine. Ifa firm does exit it means that that firm could not operate in a competitive market once it is not sheltered by a collusive arrangement. Massimo Motta “On cartel deterrence andfines in the EU”, 12 October 2007 11
[35]UK guidelines The OFT issued guidelines very similar to those of the EC in February 2000, and updated these in 2004 to reflect the OFT’s powers to penalise the infringement of Articles 81 and 82 as well as that of the UK prohibitions.” The approach followed in the UK involves calculating the basic amount of the fine according to gravity and duration, and then applying various adjustments -. [35.1] [35.2] [35.3] [35.4] The basic amount is a percentage up to a maximum of 10% of the turnover of the undertaking in the relevant market in the last financial year. Within this range, the figure chosen will depend on the nature and gravity of the infringement.”® Price fixing, cartels and abuses of dominance with a particular serious effect on competition will be viewed as serious infringements and will attract penalties “at or near 10% of the relevant turnover’. The basic amount is then increased by up to 100% per year or part year for infringements lasting more than one year, i.e. it is multiplied by the number of years. The figure is then increased or decreased to take account of the OFT’s policy objectives, particularly that of deterrence. This adjustment may result in a substantial adjustment of the financial penalty calculated in the earlier steps."° The OFT may then increase or reduce the penalty to take account of aggravating or mitigating factors. Aggravating factors would include: feading or instigating the infringement, involvement of senior management, retaliatory measures taken against other undertakings, not stopping when the investigation started, etc. Mitigating factors would include: participation under severe duress, genuine uncertainty that agreement was infringement, termination when the investigation started, co-operation. 17 The maximum penalty was previously 10% of UK group turnover, for each year of the infringement, for up to 3 years. However, as part of its response to modernisation the UK government aligned its maximum penalty with the maximum penalty for infringement of Articles 81 and 82 of the EC so that the UK maximum penalty is also 10% of worldwide turnover. '8 In establishing the seriousness of the infringement the OFT will consider the nature of the product, structure of the market, entry conditions, effect on competitors and 3" parties and damage to consumers. 1° For example, in the Napp case the basic penalty was increased with an amount of £2m on a basic total penalty of £3.21m. 12
[36]
[37][35.5] Finally the total calculated must not exceed the 10% cap. This adjustment will be made after any reduction made in application of the leniency policy? The OFT will also take into account any penalty imposed by the European Commission or another National Competition Authority, in order to avoid double jeopardy. Nevertheless, the UK Competition Tribunal concluded in the Napp* ‘case that it was not bound by the OFT’s Guidelines. It said that it would make its own assessment by following a “broad brush” approach on whether the penalty was calculated at the correct level and would,S only after reaching its own conclusions as to what the penalty should be, “cross check” its finding by taking into account the OFT’s guidelines. In that case, which concerned excessive and predatory pricing, the Tribunal started its enquiry by considering the case as a whole. {t identified its starting point for calculation of the penalty as the turnover of the firm relevant to the market in which the contravention was found to have occurred (“the relevant turnover’). It then looked for mitigating and extenuating circumstances and concluded that an appropriate penalty would be a fine discounted from the maximum possible penalty. Finally it compared the figure it had arrived at with that proposed by the OFT. In this regard the Tribunal took note of the maximum permissible penalty under the Act and compared (“cross- checked”) its calculation of an appropriate penalty with that calculated by the Director.” US In the US the Sherman Act provides for fines of up to $ 100 million for a corporation for each count on which it is convicted. Larger fines may be imposed under the Comprehensive Crime Control Act and the Criminal Fine Improvements Act which provides that the fine may be increased to twice the gain from the illegal conduct or twice the loss to the victims. As part of their sentencing procedures courts also use the Sentencing Guidelines promulgated 20 4 reduction in fine is available to an undertaking which does not satisfy the conditions for immunity but which provides evidence representing “significant added value” in the investigation and ends its involvement immediately. 21 NAPP Pharmaceutical Holdings Limited and Subsidiaries and Director General of Fair Trading, The Competition Commission Appeal Tribunal Case No 1001/1/1/01, 15 January 2002 2 Paragraphs 497 and following. 13 {38] by the US Sentencing Commission to compute sentences and to evaluate when to depart from those recommended sentences. Under these guidelines the calculation of the fines for corporations starts by determining a “pase” fine. In antitrust cases the base fine is 20% of the volume of affected commerce, which is equivalent to our notion of affected or relevant turnover. The volume of commerce is measured in terms of the goods or services sold by the corporation that are affected by the violation. Once the fine is set, a minimum or maximum “multiplier” is determined by reference to the corporation’s “culpability score”. All corporate defendants start with a base culpability score of 5. The score can be increased if the organization is particularly large or if a high-leve! employee participated in or condoned the conspiracy. The score can also be increased if the organization is a recidivist or obstructed the investigation. The culpability score can be reduced if the corporation had an effective compliance program. Strict standards are however set for proving the effectiveness of the program. The Sentencing Guidelines specify minimum and maximum multipliers that correspond to different culpability scores and by which the base fine should be multiplied, yielding a “fine range”. The court selects a fine from within the range by considering factors such as the organization’s role in the offence and the collateral consequences of conviction. The courts may also adjust the fine range upward or downward taking into account mitigating or aggravating factors such as assistance given during investigation whether the offence involved consumer markets and bribery of public officials.” As can be seen from the above discussion, agencies across the world have had similar challenges. However certain trends can be identified in recent times. The agencies most closely related to ourjurisdiction namely the EC and the OFT have moved towards higher penalties based on worldwide turnover in order to promote effective deterrence. In order to promote transparency and predictability for firms, they have adopted guidelines. These guidelines are not however mechanistically implemented and flexibility is achieved by taking into account mitigating and extenuating circumstances. Courts or adjudicative bodies on the other hand are reluctant to fetter their discretion by confining themselves to an overly “arithmetic” approach. Yet, as we have seen in the Napp case, an adjudicative body finds an arithmetic based approach a useful cross check in determining an appropriate penalty. Despite these differences, % Antitrust Law Developments 6” Edition, Chapter 9D 14
[39]
[40]the guidelines adopted by enforcement agencies in other jurisdictions provide firms and the public alike with some certainty and a degree of predictability. They also serve to enhance transparency. Of critical significance is that the starting point for calculating penalties for all these agencies discussed above is the relevant or affected turnover of that firm and not the total or global turnover. At the same time, the trend in all the jurisdictions, whether or not there is a statutory limitation to the penalty, is to treat duration, extent and gravity of the contravention as serious aggravating factors by which the base fine is multiplied several fold. South Africa The South African Competition Commission has not yet developed guidelines in relation to the calculation of penalties. A recurring sentiment expressed by respondents who wish to plead guilty is that there is no consistency in the approach of the Commission. There may be no basis for this criticism, but in the absence of guidelines when respondents in the same matter are fined differently without the justification being made transparent the Commission makes itself vulnerable to such accusations. A case in point is its approach to the respondents in this referral. The first consent order concluded by the Commission was with Aveng. Aveng admitted that its subsidiary, Infraset, contravened the Act and agreed to pay a penalty of R46 277 000, being 8% of the turnover attributable to Infraset in the previous financial year, excluding the turnover attributable to paving products. The Commission concluded the second settlement agreement with Concrete Units (Pty) Ltd who agreed to pay a penalty of R5 763 743, representing 7% of Concrete Units’ turnover for the financial year ending 2008. The third party to settle with the Commission was Cobro Concrete. It agreed to pay a penalty of R4 022 568.29 which represented 6.5% of its turnover for the financial year 2008. Cape Concrete was the last to conclude a settlement agreement. During August 2010 it agreed to pay a penaity of R4 371 386, representing 7% of the turnover attributable to Cape Conrite in the 2008 financial year. What is not made clear by the Commission in this schedule is whether the fines, apart from the Infraset one, have been calculated on the basis 4See Annexure A of the Commission’s Heads 15
[41]
[42]
[43]of total or affected turnovers. It appears though that all of Concrete Units, Cobro Concrete and Cape Concrete are involved solely or to a large extent in only concrete products.” We must make it plain here that our discussion of the Commission’s approach in settlement agreements concluded by the Commission with other respondents in this matter, does not amount to a fettering of our discretion in relation to the respondents before us in this matter. We simply refer to them to demonstrate the difficulties for the Commission and respondents alike in the absence of guidelines developed by the Commission to have some measure of consistency, on the one hand, and predictability on the other as to the potential exposure of a firm. Guidelines by the Commission would be enormously helpful if they also sought to identify the possible types of extenuating and mitigating factors. For example the Commission could list the duration of the cartel as an extenuating factor warranting a multiplication by the number of years along the lines of the EC or the OFT guidelines. At the same time the guidelines could reflect the unique circumstances of our jurisdiction and the policy imperatives of the Commission at a particular time in history, such as the relative small size of the South African economy, our unique legislative framework and sectors of the economy which are critical to growth and development. After all cartel activity in basic food and health products or bid rigging in the public sector, whether this be construction or education, would have more adverse consequence for consumers than a cartel in the luxury motor vehicle market and the Commission could treat them accordingly. These would be appropriate factors to take into account in a country such as South Africa, where the gap between the poor and the rich widens on an annual basis. In addition the guidelines could provide for discounts for parties who wish to settle early with the Commission and could serve as a useful barometer for this Tribunal.”° The Tribunal on the other hand cannot fetter its discretion by adopting a formulaic approach. After all there may be circumstances in which we may choose not to impose an administrative penalty. Hence it is not open for us to say for example that in all cartel cases or for all respondents, regardiess of the See websites of the respective firms 26 The EC in Cartel cases rewards companies for settling their cases. Parties settling a case, and thereby achieving savings for the Commission and themselves, will receive a 10% discount on their fine. See the EC Settlement Notice. 16
[44][45}
[46]147] circumstances, we will commence at a figure of 30% of the affected turnover. We cannot simply tar all respondents with the same brush. A small player that has been bullied by the larger player who is also the leader of the cartel cannot be treated in the same way as a small player who may hide behind its relatively small size when faced with sanction but whom with knowledge of unlawfulness and with full comprehension was an active participant in cartel arrangements. The Commission however is entitled to publish guidelines as to what level of penalty it would seek against respondents whom it intends to prosecute in our forum or the basis upon which it settles matters with respondents. The approach taken by the Tribunal thus far to the determination of an appropriate penalty has been to identify or connect the turnover of the firm in the line of business or the market in which the contravention has taken place. This has been described as the affected or relevant turnover. While until now the Tribunal has calculated penaities based on the notion of relevant turnover, the Tribunal has never restricted itself to this methodology nor has it held that it would treat all contraventions of the Act in the same manner in all circumstances. In fact the Tribunal has consistently stated that the approach adopted in each case was to serve only as a guideline and that meaningful distinctions need to be drawn between different contraventions in the Act. In Competition Commission v Federal Mogul the Tribunal made it abundantly clear that its approach was meant to be a guideline and that hard- core cartels may well attract the “maximum penalty”. 27 in Competition Commission v SAA,® the Tribunal expressly stated that its approach “was intended as a guideline for the future. However further experience with the Act may indicate that either the weightings are inappropriate or that we have not exhaustively considered all the factors that may exisf’° in Pioneer the Tribunal stated that there was no numerus clausus of circumstances which would cause it to reach for the total turnover of a firm but that each case must be decided on its own facts. The Competition Appeal Court has stated that “section 59 sets out the bare fundamentals of a framework for determining the amount of the administrative penalty’° Thus framework ought not to be conflated with a formulaic 27 59003] 2 CPLR 464 Para 17 2 Competition Commission v South African Airways (Pty) Ltd [2005] 2 CPLR 303 (CT). ? Dara 343 3° Federal Mogul Aftermarket v Competition Commission [2005] 1 CPLR 50 (CAC) 17 148] 149]
[50]methodology. {n order to make meaningful distinctions between contraventions each case must be determined on its own facts.*’ At the same time, as we have seen in all jurisdictions, agencies and courts alike utilise the notion of relevant turnover because it provides a useful and rational basis for computation purposes. Until now the Tribunal has had limited experience with hard-core cartels such as this one. In SAA the Tribunal was concerned with an abuse of dominance case and in Federal Mogul the contravention concerned section 5(1). Although in Pioneer we dealt with a section 4(1)(b) contravention the facts of this case differ substantially from those in Pioneer. There the cartel was often unstable, of short duration and extended over one product and involved limited market division. In this case we are dealing with a cartel pertaining to a range of products in the wider concrete or pre-cast concrete market, and not just one product in the concrete products market. 32 While all of the players were not involved in all of these activities to the same extent at the same time they were all in contravention of section 4(1)(b) in more than one respect and their arrangements extended over many products over a long period of time. It is axiomatic that a stable cartel of long duration involving a multitude of contraventions cannot be treated in the same manner as an abuse of dominance case or an unstable carte! of short duration and which might involve only one contravention contemplated in section 4(1)(b). Section 59, while containing a statutory limit to the amount that a penalty cannot exceed, requires us to give due regard to the nature, duration and gravity of the contravention. We have seen in our discussion above that courts and agencies across the world have sought to deter cartels of long duration by multiplying a base fine calculated on relevant turnover by the number of years that the cartel has been in existence. The reasons for this are obvious. Long standing cartels have the potential to cause long term damage to competition to such an extent as to leave markets permanently disabled. In our view this case requires us to adopt precisely such an approach. We turn to consider the respondents’ cases. 31 See the discussion in Pioneer at para 139ff 32 In that case Pioneer was prosecuted only on its cartel activities in the bread product market not the broader wheat market. However the Commission has initiated complaints against it in the wheat and maize milling markets. 18
[51]
[52]
[53]
[54]155} SPC’s role in the cartel SPC commenced operations in South Africa in 1965 in the business of pre- stressed concrete pipes for use in pressure pipelines. Its primary clients at that time were Department of Water Affairs and Forestry and Rand Water Board. It serviced these clients in this business for more than 20 years. SPC entered the Gauteng market for storm water pipes in January 1994. At that stage it competed with Rocla, Craig Concrete Products and Infraset. After receiving several invitations from Rocla to join the cartel it eventually accepted in October 1994. The cartel’s activities were governed by a document entitled “Modus Operandi”.** In this document cartel members were referred to by numbers and not names. Craig Concrete was number 1, Infraset number 2, Rocla number 3 and SPC number 4. The agreement set out the product ranges that were the subject matter of the cartel, who was precluded from which product markets, geographic demarcations, market share allocations, how future expansions were to be effected, how and when market data ought to be shared, what mechanisms were in place to rectify imbalances on deliveries and pricing, which included prices of delivery; inter party purchases, terms of credit to be extended by members to each other and what level of representation in each organisation was required at meetings where these issues were discussed. In short none of the cartel members conducted business independently from their competitors in the product markets that were the subject matter of the Modus Operandi, whether or not they earned any revenues in those markets. Upon joining the cartel SPC was allocated a share of 12.5% of the sale of storm water and sewer pipes with a 300 mm to 1800 diameter. The agreement only permitted SPC to compete within a 150km radius of Johannesburg. When Concrete Units joined the cartel in 2001, SPC’s share was decreased to 11.75%. In October 2001 SPC acquired Craig Concrete and its market share was increased to 27%. By the time the cartel disbanded SPC was a relatively large player in the Gauteng market with a turnover in the region of R168m in 2008. In terms of the cartel agreement, SPC and Craig Concrete were forbidden to enter the market for culverts and manholes in the Gauteng area. The members 33 Trial bundle 380 19
[56](57)
[58]
[59]supplied products to their customers on the basis of deciared deliveries, the volume of these to accord with their respective allocated shares. Because actual deliveries determined which new contracts could be secured by individual members, each cartel member declared its respective orders and the extent of the previously delivered tonnage. These disclosures were used to determine which members would bid for which contracts. The banker in this cartel was SPC who compiled monthly delivery schedules summarising the members’ declarations of deliveries and secured orders. Prices for products were agreed and reviewed once or twice annually usually following increases in raw material costs. In addition representatives of the cartel met once a month, after the meetings of the Concrete Manufacturers Pipe Division Association for purposes of sharing information on contracts which had been awarded, to confirm the contracts which individual members would attempt to secure and if necessary to discuss prices. Thus the cartel arrangement involved price fixing, geographic market division and product market division over a range of concrete products. Pricing of products, manufacturing and supply of pipe ranges, entry and expansion, bid - rigging, customer allocation, inter-party trade and credit terms were all regulated by the Modus Operandi. The consequence of this extensive collusion was such that none of the cartel members existed in those markets as an independent competitor. SPC was also a party to an agreement concluded for the purposes of collusive tendering in the Gautrain project. According to the Commission SPC, Rocla, Infraset and Concrete Units had concluded an agreement, effective on 4 June 2005, which stipulated, inter alia, that they must obtain each party’s prior written consent before submitting prices or quotations to the Bombela consortium and to agree percentages for dividing the culverts and pipes contracts among themselves. SPC’s submissions SPC did not deny any of the above facts. However in its defence SPC maintained that it had joined the cartel for fear of being undercut and driven out of business. It had never been involved in the manufacture of culverts and was 20
[60]{61] involved in the manholes market to a very small degree as a “sideline’.* it had manufactured manholes for large customers from to time, one of these being in Botswana. Furthermore it had voluntarily withdrawn from the cartel. Moreover, membership of the cartel did not furnish any meaningful benefit to its members and had not done so for some two fo three years prior to October 2007. As far as the collusive tendering arrangement was concerned, SPC admits that it was party to an agreement regarding the provision of precast concrete pipes and culverts in the Gautrain project but argues that, despite the signed memorandum of understanding, the joint venture contemplated in the agreement through which this was to be done was never formed. What eventually transpired is that Rocla and Aveng (Infraset) established a joint venture which purchased pipes from SPC. It is these mitigating factors that it wishes the Tribunal to consider when determining an appropriate penalty. A fair amount of the evidence led by SPC was dedicated to a consideration of SPC’s relevant turnover. Nevertheless Mr Riot testifying on behalf of SPC conceded that the cartel arrangements were correctly described by the Commission,” but persisted with the submission that SPC did not enter the culverts market for its own business reasons. However apart from the evidence of Mr Riot who did not have first-hand information about SPC’s operations in the mid-late 90s, no other evidence in the form of strategic documents supporting this decision was placed before us. As far as the manhole market was concerned Mr Riot conceded that SPC had the capacity to produce these and had done so from time to time but only for large customers such as the Botswana project.*° The extent of SPC’s involvement in collusive activities is relevant to an assessment of which proportion of its turnover constitutes the relevant turnover. The Commission and SPC do not agree on this issue. The Commission submitted that the relevant market was the market for concrete products because the cartel extended over a number of concrete products. Even though SPC had no turnover in respect of a specific product such as culverts, this was because it had agreed to exclude itself from that market. SPC 34 See evidence of Mr Riot Transcript of 2 August 2010 p85 35 Heads paras 3-8 36 He described it as a side line 21
[64]
[65]
[66]did not put up a positive alternative to this but Mr Brassey argued that the relevant market was the market for concrete pipes. It is perhaps easiest to start with the approach advanced by SPC. Although it is common cause that the relevant financial year was 2006 and that it’s annual financial statements at the time reflect its turnover as being R168 825 969. SPC has argued that this turnover figure has subsequently been revised downwards and the revised figure is the correct figure to constitute the cap on the penalty for the purpose of section 59(2). The reason for the reduction was a contention that the turnover as originally stated reflected an amount for the purchase of pipes bought for a customer and then on sold to it at cost pursuant to a contract. SPC refer to this as an “in ~out “amount. Secondly, SPC contended that the affected turnover figure is still lower than this revised total turnover figure and argued for the deduction of two further amounts to arrive at this figure. The first figure to be deducted is the turnover attributable to what it terms its civil engineering activities. It is common cause between SPC and the Commission that this activity should not constitute part of the affected turnover as this entails engineering work on pipes that does not involve either manufacture or concrete products. SPC wants a further deduction of the turnover in respect of the concrete products it supplied for the Gautrain project. SPC had supplied tunnel linings and noise barriers to the Bombela Consortium for the Gautrain Project. While these did not involve pipes, they involved concrete products.*’ It argued that this turnover is not causally linked to the conceded, but in its view abortive, attempt to rig the bidding for this work. The Commission disputes this and argued that there is a causal link between this turnover and the earlier bid rigging agreement and hence the turnover should be considered part of the affected turnover. We consider this factual dispute more fully below. For now it is indicated for the purpose of understanding the different approaches to arriving at the total and affected turnovers respectively. Thus summarising in figures what was set out above: The SPC affected turnover is: 37 See page 4 of Bundle C 22 Total turnover in the 2006 financial statements: R168 825 969 Deduct in-out from total turnover for 2006: R32 371 630 Net total turnover: R135 545 339 Deduct from this civil engineering turnover: R44 935 988 Deduct from this Gautrain contract: R42 834 295 Net affected turnover: R43 684 056 The Commission’s affected turnover is; 168] Total turnover in the 2006 financial statements; R168 825 969 Deduction from this for civil engineering turnover: R44 935 988 Net affected turnover: R123 889 981 It is necessary for us to decide the dispute in respect of total turnover as the imposition of a fine requires us to know what the maximum permissible fine is. We will then consider the dispute as to whether the Gauteng project turnover should be deducted to arrive at the affected turnover. Since there is-no dispute over the deduction of the civil engineering turnover we will assume for the purposes of this case that it is not part of the affected turnover. SPCs Total turnover The basis for excluding the amount for the “in-out” transactions was justified by Mr Louw. Mr Louw from the auditors firm of Eksteen Louw & Badenhorst Incorporated, testified that in his view SPC had erroneously represented this amount as revenue. He submitted that in fact this money was not revenue but was really a cost of sales because SPC had procured steel pipes from a supplier and had supplied these directly to DWAF, its client. Risk and ownership had passed to DWAF immediately and technically this amount ought not to have 23
[69]been reflected in the books of SPC. This was more so that case in light of the fact that SPC had not made a profit on the provision of these pipes. In Mr Louw’s recent professional view, as reflected in his letter addressed to SPC and copied to their legal representatives, the annual financial statements of SPC no longer complied with international best practice because they had included an item as revenue when they ought not to have. See letter attached as annexure 3. In support of this proposition we were referred to a number of invoices purporting to relate to this particular transaction. What Mr Louw could not answer was why he, as an auditor, had previously signed off on these annual financial statements and the reasons for his sudden revision. Nor could he explain the consequences of his newly revised views. Recall that accounting is a double entry system. If an item has been included as revenue, the corresponding cost of purchasing it must be reflected in the cost of sales. Presumably this had already been included in SPC’s cost of sales figure of R117 773 508. That this was so was borne out by the fact that in practice and in the ordinary course of business SPC bought and paid for the steel pipes as evidenced by the invoices, and sold these onto DWAF as part of its contract with DWAF. SPC’s attempts to exclude this item from the notion of “revenue” on the eve of a prosecution in this forum appears rather contrived when previously it had happily reported on this as part of its total turnover for all other regulatory purposes and for purposes of VAT and income tax. Indeed Mr Louw himself testified that his interpretation of the notion of revenue as expressed in his letter dated 16 November 2009 and in his testimony had only recently been sought by the management of spc. Mr Riot under cross examination conceded that they had agreed with DWAF to this particular mechanism in question — namely delivering the pipes and passing risk & ownership directly onto DWAF — in order to obtain a direct benefit in its cash flow. But for this mechanism, DWAF wouid have retained a 10% retention fee. In consequence of this arrangement DWAF would waive the 10% retention which was to SPC’s benefit. In our view this item ought to be included in the total turnover of SPC because this is how it was always treated in the ordinary course of the company’s business for all purposes. 381184. See also his letter dated 16 November 2009 addressed to The General Manager Southern Pipeline Contractors Joint Venture and cc Attorneys Carvalho Hill, Witness Bundle p 109 24
[70]
[71]
[72]
[73]In our view therefore the Commission has correctly considered that the total turnover is the turnover as originally reflected in SPC’s accounts namely R168 825 969. SPC’s Affected turnover It is common cause that SPC, Rocia, Infraset and Concrete Units had concluded an agreement to rig bids for the Gautrain tender. SPC contend that the agreement was never executed and that instead Rocla and Infraset established a joint venture which purchased pipes from SPC. The evidence of Mr Greeff of Rocla was that the original arrangement did not materialise because Rocla and Infraset wanted a larger share of the tender rather than an intention on the part of these firms not to collude. This is supported by a document which reflects a discussion between Rocla, Infraset, SPC and Concrete Units on 2 February 2005. In that document several proposals had been recorded subject to “discussion at industry level’.” The proposals relate to a split of the tender among Infraset, Rocla and SPC and Concrete Units for the supply of pipes and culverts for the Guatrain tender. That document tends to support Mr Greeff’s testimony that the original arrangement did not come to fruition because of disagreement among the members. and that Rocla and Infraset wanted the largest slice of the tender. As it turns out, whatever the nature of the revised arrangement between Rocla and Infraset, what we don’t see in the new arrangements is SPC putting in a tender on its own and competing with Rocla and Infraset. Instead it elected to remain a party to this arrangement albeit as a junior, behind the scenes partner. This subsequent supply arrangement to its erstwhile co-conspirators retains the important element of being executed with in co-operation with and not in competition with competitors thus depriving the customer of the benefit of competing tenders. In our view this amount is legitimately to be considered part of the affected turnover. Accordingly SPC’s relevant turnover figure would be R123 889 981 (total turnover of R168 825 969 less R44 935 988 for civils). * Transcript p 51 Tyjal bundle 372. 25
[74]
[75]
[76]77 Determination of the penalty We now turn to consider the calculation of the penalty. As our earlier comparative discussion has indicated there are a number of approaches to arriving at an appropriate penalty. We have not followed a single approach here but we first consider an arithmetic based approach and then we follow what in Napp was referred to as the broad brush approach i.e. a discretionary approach to the relevant factors, but not one dependent on the application of a prior formula for calculation. We use the arithmetical approach to serve as a cross check of our broad brush approach. As we shall see both lead to the same conclusion for SPC and largely the same for Conrite . The arithmetic approach As discussed previously other jurisdictions have treated duration and gravity as factors contributing heavily in the calculation of penalties and have adopted guidelines for the calculation of penalties. A comparative exercise using the EU and UK methods shows that SPC’s conduct, given the extent and duration of the cartel, would attract the highest penaity. We have found, as shown above, that the affected turnover is R123 889 981. We must then multiply this number by the appropriate base percentage and then by the number of the years of the cartel. Recall that while the SPC’s participation in the cartel endured for 13 years we have elected to limit this period to 10 years, being the period of duration of the Act. By way of example if we adopted the EU methodology in this case we would start with a basic fine of 30% of the affected turnover of SPC. (76.1] 30% (R123 889 981) = R 37 166 994 = base fine. [76.2] Base fine x 10 years (duration) = R371 669 943. Utilising the method used by the OFT method, a cartel of such long duration would produce a fine of 10% of affected turnover multiplied by the duration of the cartel. In SPC’s case the fines would be — [77.1] 10% (R123 889 981) x 10years = R123 889 981. 26
[78]
[79][80}
[81]
[82]Note that both these amounts exceed the 10% total turnover cap of approximately R 16.9 million permitted in our Act. But even if, for argument’s sake, we adopt the lower affected turnover figure contended for by SPC of R43 684 056 we arrive at an amount for the fine that would exceed the total turnover cap: [78.1] EU approach [78.1.1] 30% (R43 684 056) x 10=R 131 052 160 [78.2] OFT approach [78.2.1] 10% (R43 684 056) x 10 = R43 684 056. Although this exercise has not factored in any mitigating discounts it has not included aggravating factors either. What it does show is that the arithmetic approach even on the most favourable scenario for SPC — using its lower base affected turnover and a 10% not 30% initial multiplier results in an amount that far exceeds the permissible cap. These approaches demonstrate that penalties for cartels of long duration will attract a very high base fine. In the context of this case, a base fine of R371m figure represents more than double SPC’s annual turnover of R168m. The arithmetical approaches examined above are reliant on the application of particular formulas. However we could also arrive at similar magnitudes of penalties in our jurisdiction by adopting the broad brush discretionary approach through considering the factors set out in section 59(3) of the Act. Discretionary approach Section 59(3) states- [82.4] When determining an appropriate penalty, the Competition Tribunal must consider the following factors: (a) The nature, duration, gravity and extent of the contravention; 27 (b) Any loss or damage suffered as a result of the contravention, (c) The behaviour of the respondent; (d) The market circumstances in which the contravention took place; (e) The level of profit derived from the contravention; (f) The degree to which the respondent has co-operated with the Competition Commission and the Competition Tribunal; and (g) Whether the respondent has previously been found in contravention of this Act. {83] When we consider the nature and duration of this offence there is very little mitigation to be found in Mr Riot’s submission. Recall that this cartel was in existence for more than 40 years and SPC’s involvement spanned more than a decade and that the modus operandi covered an entire range of products and that a quid pro quo principle applied. SPC was permitted to compete in the designated pipe ranges and benefited from limited price and product competition in that market, provided it did not enter the manhole and culvert market. This agreement resulted in a lessening of competition not only in the pipes market but also in the manhole and culvert markets. Even if we are to accept, for argument’s sake, that SPC may not have wanted to enter the culverts market at first, its on-going commitment to the cartel arrangements ensured that it never considered entering that market despite the fact that it could quite easily have done so. Likewise in the manhole market.*’ Mr Riot’s testimony shows that from a demand side, manholes and pipe networks were closely related in that a large customer would often want both manholes and pipes to be installed.” Indeed SPC’s occasional supply of manholes fo customers outside of the cartel designated geographic areas tends to support the conclusion that it was able to and did indeed supply manholes but in fact chose not to do so because it had adhered to its commitments to the cartel arrangement. Here was a textbook example of a successful firm who could easily enter into related concrete markets but elected not to do so because of its collusive arrangements with competitors. This modus operandi clearly had the impact of raising prices in the 4! See Statement of Stephane Riot Bundle B, par 15 p 67 where he refers to a document called a “Modus Operandi” 7115. Manholes were necessary to access pipe networks. 28
[84]185]
[86]concrete products market. Mr Myburgh from Aveng, testifying on behalf of the Commission, indicated that in his estimation prices of concrete pipes fell between 25-30% after the cartel had disbanded.** However despite this pricing effect, there was no evidence that post the disbandment of the cartel SPC, or any other member of the cartel had re-entered any of the markets they had previously excluded themselves from. This shows that the cartels of such long duration are likely to result in structural changes to affected markets with long term consequences. Firms, but for the cartel, which would have competed in the market, were now unlikely to enter even after the cartel had ceased to exist. The harm caused to competition in such markets is immeasurable The evidence on market circumstances was scanty but we do know from SPC itself is that its major clients included DWAF, Rand Board and various municipalities, all of the projects pertaining to essential services such as water and sewage. These are critical sectors for the growth and development of our country. However the market division and customer allocation arrangements would also have impacted on prices in government infrastructure and construction projects such as stadiums for the World Cup and the Gautrain project. While it is difficult to estimate with precision the extent of damage caused by a cartel of such long duration, such elevated prices in a market where the demand for concrete products was reaching record heights due to thé demand created government projects can only have caused considerable harm to the public purse and ultimately the South African taxpayer. SPC’s stated reason for entering the cartel is that it “feared being undercut and driven out of business”. This fear may keep business persons or executive awake at night but it is nothing more than the fear of the cut and thrust of competition. A fear of competition can never be a justification for engaging in contraventions of the Act. Indeed such conduct should be considered as an aggravating factor. We were referred to an incident by Mr Riot where a customer Cerimele Construction was offered an incentive of R500 000 by Fraser Fyfe (Infraset) to terminate its business relationship with SPC and to accept concrete pipes from cartel members at the same price as had been offered by spc.“ However “This was confirmed also by Mr Speir’s who gave as one of the reasons for his dismal performance in the manhole market was the fact that he was not in the pipe market. Transcript of 2 August 2010, p 42 “ The timing of this is vague but seems to have occurred in the early 1990s. 29 187) (88]
[89]Cerimele Construction refused to take up the proposal from Fraser and maintained its agreement with SPC. Mr Riot in paragraph 10 relies upon this incident as a basis for its knowledge of a cartel. He does not expressly rely upon this as a threat® in our view the incident simply supports the conclusion that SPC was able to withstand its position in the market and was not being threatened directly with a price war. If at that time it was so concerned about predatory tactics by its competitors it could have easily sought the protection of the Commission — even if this was some years after its entry into the cartel — by ‘fessing up. Instead it elected to stay in the comforting arms of its competitors for a period of 13 years amounting to almost all of its active life in the concrete pipes market. Moreover SPC was not a passive cajoled member but instead played an active role as banker for the cartel. As far as the collusive tendering agreement is concerned, bear in mind that the deal was concluded at the highest level of management of the three firms. Mr Riot himself was a signatory to an agreement that he knew was wrong.” Although the original bid rigging agreement was not executed, SPC entered into and benefited from a subsequent supply contract with its erstwhile conspirators. The fact that SPC was a regional player must be borne in mind. However it is not an insignificant player and by its own account has a reputable pipe manufacturing and installation business, boasting special skills in the rehabilitation of steel pipes used in pressure (water) systems. Its customers include entities such as the Department of Water Affairs, the Rand Water Board and other infrastructure entities involved in the roll-out of essential services. The “voluntary withdrawal” of SPC from the cartel is not supported by anything more than a bald assertion. Indeed Mr Riot conceded that the cartel was “not working so well” because Rocla had stopped providing information because its holding company Murray and Roberts was being investigated by the Competition Commission. The message that Rocia was giving to its cartel members was that they should stop this cartel because of the Commission’s investigation into Murray & Roberts.” This was not a voluntary withdrawal motivated by an appreciation of unlawfulness or the desire to compete but rather motivated by the fear of exposure and sanction. 4 Although tries to suggest this in his testimony “6Transcript of2 August 2010 p 80 “7112-114 30
[91]
[92]
[93]We have previously stated that engaging in negotiations with the Commission as to settlement of the matter does not amount to co-operation. Such engagement is only for a respondent's own benefit and not to benefit of the Commission’s investigation. SPC’s evidence that it had gained limited profit from its membership is also not supported by Mr Riot. A graph produced at the hearing purported to show that SPC’s costs were increasing at a higher rate than its profits. However the figures only went as far back as 2002 and none of the underlying material was discovered or provided to the Commission and the Tribunal prior to the hearing. Nor were comparative figures for other cartel members furnished. Mr Riot in any event conceded that all that the graph showed was that costs were increasing at a higher rate than profits. This could not lead to a conclusion that no benefit was derived from its cartel membership. SPC has not been previously convicted of an offence under the Act. In this particular case however, where the contravention occurred for longer than the duration of the Act itself, this factor can hardly be given weighty consideration. SPC had been saved from previous prosecution not because of its innocence but because of its participation in a highly secretive, very stable and enduring cartel in which no member was incentivised to blow the whistle. It was only when Rocla became a reluctant participant, and here we must emphasise not because of some Damescene experience brought on by a desire to do the right thing but because of the Commission’s investigation into its parent company that the cartel became unprofitable and unsafe for the remaining members. Arguably a mitigating factor in favour of SPC is that it was not a leader of the cartel, albeit a committed member. However this must be weighed against its 13 year participation in the cartel, its active role as banker, it’s belated withdrawal, the fact that it colluded on a range of products, that the collusion involved not only price fixing but also market division, customer allocation and bid rigging, the fact that it unlawful conduct had a direct impact on the cost of essential services and public projects, and that it is not a gmail unsophisticated entity but has international shareholders and considerable experience in the markets that it conducts business. Conclusion on SPC 31 {94] {96} Having regard to ail of the above we find that this is a case of great aggravation and no mitigation. This cartel preceded and succeeded the commencement of the Competition Act. Collusion among the members of the cartel extended over a range of product and territorial markets. It even extended to the minutiae of delivery charges and agreement on inter-competitor trade. Virtually no activity was left to competition. As we noted the market divisions that formed part of the cartel arrangements have survived the uncovering of the cartel. There was no evidence that SPC had entered any product or geographical market from which it had excluded itself pursuant to the cartel. Every form of hard core cartel practice was present from market and product division to price fixing and bid rigging. While it was not the leader of the cartel SPC was an active and enthusiastic member for 13 years. Senior management was involved in these activities. As we noted earlier in our comparative discussion duration, gravity and extent ~ all expressly mentioned in 59(3)(a) - are the most weighty factors considered in applying fines to cartels in other jurisdictions, In respect of all three this is the most serious of infringements. Such a case must demand the highest sanction. Hence we conclude that an appropriate penalty in this case is a fine of 10 % of the total turnover of R 168 825 969. The appropriateness of the magnitude (as opposed to the value) of this fine is cross checked by the prior arithmetic approach which brings us to the same conclusion. A penalty in the amount of R16 882 596.9 rounded off to R16 882 597 (sixteen million eight hundred and eighty two thousand five hundred and ninety seven rands) is hereby imposed on spc.* In comparison to the fines imposed in other jurisdictions even a fine of this magnitude seems an under-deterrent for the gravity of SPC’s conduct. Conrite Walls’ role in the carte! (97) Conrite Walls commenced business in 4967 in Durban. It manufactures toilets, steel palisade, concrete palisade, gates and wire fencing. It has never produced or sold pipes, culverts or pre-cast concrete sleepers. It entered the pre-cast concrete manhole market in 1999 and competed with Rocla, Infraset, Grallio, Cobro and D&D in the KZN region. ® 40% of R 168 825 969 32
[98]
[99]
[100]
[104]
[102]According to Conrite Walls the cartel was established towards the end of 2000 or early 2001 when it, D&D and Cobro agreed to divide the Durban market for manhole rings on the basis of each firms’ respective production capacity. Conrite Walls was allocated 15% of the market, Cobro 65% and D&D 20%. The members agreed on pricing guidelines but did not agree on final prices to be paid by customers. Discounts could be given at each member’s own discretion. The members met every two to three months. At these meetings Cobro allocated the available work amongst the members based on their agreed market allocation. Conrite Walls never participated in the national cartel. During 2005 the cartel members agreed that Conrite would exit the manhole market. Following its exit from the manhole market Cobro and D&D divided Contite Wall’s market allocation between them and offered to pay Conrite R30 000 and R10 000 respectively which were so-called “monthly rental’ amounts for usage of its manhole moulds. Fictitious invoices for these amounts were issued by Conrite but in reality these amounted to payments to Conrite to stay out of the manhole market.” These payments were made from August 2005 to February 2008. Conrite does not dispute these facts but avers that it exited the manhole market in July 2005 because it could not tender for much of the work given that it did not have an SABS approval mark. It was also known as a fencing specialist and could therefore not attract a lot of manhole work and its market share never increased beyond 1% in the Durban market. It decided to focus on its core business of manufacturing toilets, walls and fencing. It also argues that because it had not attended any cartel meetings after 2005, its involvement was of limited duration. The Commission however argues that Conrite Wails only ceased its participation in the cartel in February 2008 when it received the last payments in terms of the agreement governing its participation in the cartel agreement re the manhole market. Contrite’s total turnover for the 2008 financial year was R77 405 710. There is no dispute about the correctness of this amount. Conrites’ affected turnover is in dispute. Certain turnover is unrelated to concrete products and that this amount should be deducted from total turnover is not These amounts were equal to Conrite’s nett profit. See transcript page 151. 5°D&D stopped payments in 2007 when it merged with Rocla and Cobro in February 2008. See transcript page 165. 33
[103]
[104]
[105]controversial. More controversial is whether all concrete products should constitute the affected turnover or only that which formed the express subject matter of the cartel agreement. This results in wide gyrations in the figure. If the affected turnover is limited to what is reflected as manhole covers — a fictitious figure as it does not represent any sales at all — the affected turnover is R440 000. Using this as a basis for an arithmetic fine over the six years of the cartel would be wholly under deterrent. Either an amount of R 264 000 on an OFT approach or R 792 000 on an EU approach. Either figure seems wholly under deterrent. If the correct approach is to take all concrete products into account or just one that seems correlated to the exit from manholes — the production of toilet seats at a turnover of R 50 million based on the lower OFT 10% base - the result would be R 30 million or on the EU approach the figure would be rounded of R 90 million®' — a figure vastly exceeding the permissible cap figure of R7,7million This suggests given the widely diverging outcomes that the arithmetic approach is not appropriate on these facts and it wouid be better to approach the matter by considering the factors on a discretionary approach. Indeed, counsel for Conrite acknowledged quite correctly that it would be under deterrent to follow the arithmetic approach and instead opted for a more ad hoc solution. For this reason we have determined the issue following the discretionary basis solely, because the arithmetical result does not yield a useful cross check as it did with SPC Determination of the penalty Mr Speirs, the managing director of Conrite testified that initially his business was in the walling and fencing business. It manufactured pre-cast concrete products as well as some wire and steel fencing products. In relation to the concrete products the business was involved in palisade fencing, manholes and toilet seats but not pipes, culverts or concrete sleepers. In his evidence Mr Speirs attempted to suggest that Conrite was a small operation in KZN and he had somehow “drifted” into the cartel in 2001 for intelligence gathering purposes. He defended his decision to withdraw from the 51 19%(50 million) x 6 = 30 million. 30% (SOmillion) x6 = 90 million. 34
[106]
[107]
[108]manhole market by arguing that Conrite’s share of the manhole market was very small and he was in any event contemplating exiting this market. Mr Speirs claimed that by 2005, Conrite who had only been granted 15% of the market had demonstrated scant progress in the manhole market. His product was still not SABS approved and he seemed to lack the production capacity to take on large orders. in a discussion about Conrite’s poor performance with his erstwhile cartel members, he opportunistically pounced upon the proposal that he withdraw from the manhole market in exchange for payment from his competitors. Mr Speirs and his co-conspirators then embarked on a fraud to hide the true nature of their agreement. Fictitious invoices were made out for payment to Conrite and the manhole mould was physically delivered, but never used, to his competitors’ premises “in case the auditors asked about these payments.” However Mr Speirs could not explain the obvious contradiction in his version namely that if his performance in the manhole market was as dismal as he claimed, why his erstwhile competitors would be willing to pay him to stay out of the market when they could have simply left him to limp along dismally. At first he attempted to credit his competitors with coming up with this idea but after being pressed confesses- [107.1] «It was a collective decision ...of the three people who were involved in this little grouping that this is how it would be done and this is the amount...and we should invoice it on a monthly basis." Moreover he demonstrates a clear understanding of the rationale for these payments. The payments of R30 000 and R10 000 per month were calculated on the basis the market allocations between D&D and Cobro. Mr Speirs, when asked why his competitors would pay him such large amounts given his paltry performance, explains that - [108.1] “ | think possibly the reason that they would have paid me there is because if | had come back into the market at some stage, my pricing structure might not have been as favourable as they would 7161 35
[109]
[410]have liked...and the mere fact that | came in with cheaper prices 53 would have brought all the prices down... He conceded that the money, notwithstanding his earlier remarks that he intended to withdraw from the market in any event, “prevented us from participating in that market’** And finally Mr Speirs, who tells us that he was aware of the “big cartel” between Rocla and Cobro, and was also aware of the Commission’s investigation into Rocla (through D&D after it was bought by Rocla) and Cobro, does nothing when he receives a letter from Cobro’s lawyers stating that they refused to pay these invoices because “they were under investigation by the Competition Commission’. His somewhat bashful explanation for this is that he thought the Commission “would not be interested in small cartels” such as theirs.°® He did not approach the Commission himseif and instead waited until the last minute to make a clean breast of things. After he had agreed to withdraw from the manhole business he was able to focus his productive and management capacity on the toilet seat market which has now become Conrite’s main source of turnover. The agreed exclusion from the manhole market permitted Conrite to focus its productive capacity in the toilet seat and concrete fencing market. The same inputs, skills and mixers were used in relation to the production of concrete fencing and toilet seats.’ indeed if one has regard to its turnover in the toilet seat market we see a phenomenal growth of approximately 635% over a period of from 2006 to 2008 years. Significantly Conrite commences producing toilet seats at approximately the same time it agrees to exit the manhole market. Self-exclusion meaning exclusion by agreement with one’s competitors from a particular product market poses challenges as to the calculation of a penalty — after all there is no turnover earned by the firm from which it had excluded itself. In this particular case, the payments made to Mr Speirs are clearly in lieu of the self-exclusion in the manhole market. However these amounts are static and do not represent actual losses to competition or to the economy as a whole as a result of the seif- exclusion. Indeed Conrite’s counsel conceded that basing a penalty on the 7169-170 #7159 T167 56 7155 and T169 57 Hifferent moulds were required and conceivably there would be a few production peculiarities pertaining to the different products. °8 See Bundle C p 385 36 (1714]
[112]turnover from the manhole market (R400 000) would be under-deterrent and therefore offered to pay all of the payments received plus 25% premium on the amount, bringing it to an amount of R1.35m. In this case we do not see Conrite’s participation as misguided as it purported to be. The nature of the contravention has already been described as the most egregious. Conrite’s participation in this unlawful conduct, at the highest level, persisted from 2001 to 2007. Mr Speirs demonstrated an understanding of the wrongfulness of the cartel members’ activities and clearly benefited from it. Mr Speirs made much of the fact that he wasn't an effective competitor in the manhole market. He wasn’t so precisely because he was party to a comfortable cartel arrangement which was working for him. Why would a cartel member, assured of prices, market shares and customers bother to incur the cost of innovation and hassle of competition? After his self-exclusion from the manhole market, for which he was paid, he was left alone to pursue his interests in the toilet seat market. On Mr Speirs’ own version the payments do not make any sense. Why would Cobro and D&D pay a small and insignificant player to stay out a market when the player presented no threat - on Mr Speirs version - whatsoever? While there was no witness testifying to this effect, it appears from Mr Speirs’ testimony that he seldom encountered his erstwhile cartel members in the geographic markets where he conducted his toilet seat business.* This is remarkable given the massive expansion in turnover that Conrite’s figures show it enjoyed over the period. Ordinarily one would expect rivals capable of producing the same product to enter. The absence of these players and the payments made by them to Conrite suggests that there was a quid pro quo agreement in place between them to the effect that Conrite would stay out of manholes and they would not compete with him head on in the toilet seat market. All the productive and management capacity that had previously been applied to the manhole market could was now applied to the other concrete products produced by Conrite. Not much evidence was led as to the market circumstances in which the cartel operated. However there was no evidence that any of the members of the cartel, post its disbandment had entered any of the markets from which they had previously excluded themselves. However we know from Conrite that Conrite and its erstwhile cartel members were all supplying products to sectors of the So T174 37
[113]
[114]
[115]
[116]economy that were essential to the growth and development of the KZN region and the country in general. Most of these involved government initiatives such as low cost housing and water services. The degree of harm due to the price fixing and market division directly affected the public purse and eventually the taxpayer. Arguably a mitigating factor in favour of Conrite is that it is a relatively small regional player in the market. However while it was a relatively small player in the KZN region, the business had commenced operations way back in 1967, and had presumably gained a great degree of experience, knowledge of competitors and trends in the market. In light of this, Mr Speir’s submission that he went along to meetings with his competitors only for intelligence gathering purposes simply does not ring true. Mr Speir’s non- attendance at meetings after 2005 however is not however a mitigating factor. He was paid his monthly amounts, bar a few short payments and early termination from D&D, by Cobro right until February 2008. Mr Speirs did not withdraw voluntarily, continued with his illegal activity even after hearing that Rocla was under investigation and did nothing when he was advised by Cobro’s lawyer’s that the conduct they were engaged in was illegal. He only approached the Commission after it became clear that Conrite had been cited as a respondent in the matter. Conrite’s co-operation with the Commission and the Tribunal was limited to seeking leniency. The fact that Conrite has not been previously convicted of a contravention under this Act is not a weighty mitigating factor in this case. This cartel was highly secretive, stable and its true nature was concealed from auditors and public alike through fraudulent invoices. Conclusion on Conrite A differentiating factor in Conrite’s case is that the collusion related only to the toilet seat/manhole markets in KZN and was not as extensive as the cartel in Gauteng or the national cartel. Apart from the market division arrangement all other aspects of its cartel arrangements ceased in 2005. However these two factors must be weighed against the nature, gravity and duration of the contravention. The cartel persisted for at least 6 years and extended to price fixing, market share agreements and customer allocation. That the perpetrators 38
[117]
[118]
[419]
[120]
[121]acted in full knowledge of unlawfulness was evidenced by their attempts to cover up their criminal conduct through fictitious invoices. In the context of our legislation having due regard to both duration and extent, we find that an appropriate penalty in this case should be 8% of Conrite’s total turnover of R77405 710 which amounts to R6 192 457 (six million one hundred and ninety two thousand four hundred and fifty seven rands). Order SPC and Conrite Walls are found to have contravened section 4(1)()(i), (ii) and (iii), An administrative penalty of R16 882 597 (sixteen million eight hundred and eighty two thousand five hundred and ninety seven rands) is hereby imposed on SPC being 10% of R168 825 699. An administrative penalty of R6 192 457 (six million three hundred and sixty six thousand five hundred and fourteen rands) is hereby imposed on Conrite Walls being 8% of R77 405 715. The penalties must be paid to the Competition Commission within 20 business days of this order. \| VU 29 November 2010 Yasmin Carrim DATE Norman Manoim & Takalani Madima Concurring Tribunal Researcher: Rietsie Badenhorst For the Commission: Adv O Mooki instructed by State Attorney For the 1* Respondent: M Brassey SC and T Mundell instructed by Carvalho Inc For the 2TM Respondent: MM Le Roux instructed by Prior and Bezuidenhout 39 ANNEXURE A 3 2 ( Zimbabwe 4. Rocla position a) Would stay out # infraset lake care of Hurne to protect our work in Northem Botswana and Mozambique. This not happening. ; ; b) Rocla have held a discussion with Hume ai Murray & Roberts request - 00 action to date. National Gauteng N. Province Mpumalanga “| i 450km radius . infraset Rocia Salberg (manholes) Roca Rocla Infraset Rocla infraset Salberg (Manholes) Southern Pipes Salberg (manholes) Concrete Units (Southern Pipes, \ & ge ¢ Salberg Concrete Units to ra | Middleburg — 150km dit radius) @ é éxf (Also 8. E. to : ns Standerton) LL. o To Kwazulu Natal Eastern Cape ] Western Cape Free State/Central ~| Rocla Rocla Rocla Rocla Infraset Cobro Cobro Infraset (infraset, Gralio Infraset Cape Concrete Southern Pipes and Salberg (manholes) Concrete Units - Concrete Units) to D & D (manholes) (culverts) Poichefsroom Salberg (manholes) Astro (manholes). | [Zambia ; Namibia “TNigera 7. Rocla have a strategy for 4. Rocla Territory 7” What is Infraset’s position? @ entry excluding sleepers but a) infraset stays out will discuss b) Northern Railway — ; Infraset to supply sleepers Will cover Rocla on the rest. ; 4 Rocla’s Attitude ON > oir aa an Souther must be contained domestically. Rocla will react in R.S.A. if Southerns attack Botswana continuously. if no deal in Botswana or Mozambique. Swaziland must be open market and Zimbabwe. Do not wish to rock sensitive boat in R.S.A. with the shortage of volume and the players but [Hill We need firm and irrevocable agreement on offshore issues with Infraset. Recall my basic suggestion to explore joint operations for Africa.
Read from the court's own list. Judgments are official texts; verify against the court's file before you rely on this page. Firms on AttorneyOS get judgments like this one against the matters they touch, in the Monday edition.