---

REPUBLIC OF SOUTH AFRICA

IN THE COMPETITION APPEAL COURT OF SOUTH AFRICA

CAC Case No.: 144/CAC/Aug16CT

CT Case No.: 019950

REPORTABLE: Yes

OF INTEREST TO OTHER JUDGES: Yes

DATE: 14/9/2017

In the matter between:

lsipani Construction (Pty) Ltd                                                                             Appellant

and

The Competition Commission                                                                        Respondent

JUDGMENT

Vally AJA

Introduction

[1] lsipani Construction (Pty) Ltd (lsipani) appeals against a decision of the Competition Tribunal (Tribunal) imposing a single penalty of R21 783 153.40 on it for contravening s 4(1)(b) of the Competition Act 89 of 1998 (the Act). The contravention  took the form of the appellant providing  another firm, one of its competitors, with a cover price. This constitutes unlawful collusion. It is often described as " collusive tendering" or " bid-rigging". A cover price is a price that a firm desiring to win a tender provides to another firm that does not. The firm that is provided with a cover price would bid for the tender at a price higher than the cover price, thus ensuring that it would fail in its bid and, hopefully, the bid of the firm that provided the cover price would succeed. Firms that provide and receive cover prices succeed in deceiving the entity issuing the tender into believing that the tender process was free, fair and honestly competitive.

[2] lsipani and another firm, Neil Muller Construction (Pty) Ltd (NMC) both operate in the construction industry. They compete with each other in the open market. At least that is what they are supposed to do. But they found that colluding with each other could at times be more useful, less problematic, more certain and a more profitable way of doing business. After all, who needs the untidy, cumbersome, painful and anxiety-filled uncertainties that are part of doing business in a free market where all are supposed to compete fairly and equally? The promotion and development of such a market may be aimed at preventing corruption but unfortunately not all businesses, as the evidence in the case illustrated, have inoculated themselves from engaging in collusive practices. These practices, in my view, are a form of corruption. They, like those engaged in by members of cartels, or firms engaged in predatory pricing, or those electing to charge excessive prices, ultimately distort markets to the detriment  of the public interest.  The harm  these  practices  cause is often  not insignificant. The law takes a dim view of such practices precisely because of the harm it causes and because it erodes the moral fabric of our society.

[3] In this case, lsipani was found to have engaged in unlawful practices on two occasions. In terms of the Act it should be accountable for both. The Tribunal recognised this and took it into account when considering the punitive action that should be taken. However, the punishment it imposed was a consolidated one. This provoked a cross-appeal by the respondent, the Competition Commission (Commission). It asked that this Court set aside the decision of the Tribunal and refer the matter back to the Tribunal for a fresh consideration of the punishment - one that does not consolidate  the  two punitive measures into a single one.

Facts of the case

[4] The Commission accused lsipani of unlawfully providing two cover prices to NMC in relation to construction projects in Stellenbosch. The first concerned the construction of a private building, and the second involved alterations and additions to a multi-storey building at the University of  Stellenbosch's (University) Faculty of Engineering. The cover prices were provided in August 2010 and November 2010. The Commission invited lsipani to settle the matter with itself. lsipani admitted that it acted wrongfully by contravening s 4(1)(b) of the Act on two occasions. It agreed to engage the Commission on the possibility of concluding a settlement agreement wherein it would accept responsibility for its unlawful conduct and pay a fine for each of the  contraventions. Unfortunately, it was unable to agree with the Commission that it should pay an administrative penalty for each of the contraventions, or to accept the Commission's proposal on the amount it should pay in administrative penalties, which it found to be excessive. This prompted the Commission to refer the matter to the Tribunal where it could be dealt with through the adversarial process catered for by the Tribunal's rules.

[5] At the Tribunal, lsipani, through its managing director, Mr Jandre Arangies (Arangies), admitted that during the period August to November 2010 it provided two cover prices to NMC. Despite its admission lsipani initially attempted to escape liability for its conduct by pleading that the complaint initiated by the Competition Commissioner (Commissioner) did not mention it by name and therefore was not a complaint against it. But this plea was eventually abandoned and it elected to challenge only the quantum of the combined administrative penalty that the Commission sought to impose upon it. To understand its challenge, it is necessary to have regard to the details of the conduct to which it admitted.

[6] The details are uncomplicated. During August 2010 a Mr Paul Symington (Symington) of NMC asked a Mr Weyers Willemse (Willemse) of lsipani to provide him with a cover price for a private building project in Stellenbosch. Willemse obliged. However, lsipani, which was desirous of winning the bid, did not succeed in doing so as there were other competitors who bid independently of lsipani and NMC and one of them succeeded. During November 2010 a Mr Barend Badenhorst (Badenhorst) of NMC contacted Willemse to provide him with a cover price for the project at the Engineering Faculty of the University. Again, Willemse obliged. This time, too, lsipani failed to secure the contract as another firm won the bid.

[7] At the Tribunal lsipani elected to present viva voce testimony focussing on the quantum of the administrative penalty for which it should be liable. Arangies presented the testimony. He stated that the practice of providing a cover price was so pervasive in the industry that lsipani did not see itself as acting wrongfully by engaging in it. However, since lsipani has learnt that it  acted unlawfully, and given that it did no more than furnish NMC with "a simple cover price" , ultimately gaining nothing from so doing save that NMC was retained on the client's tender list, it asked that the administrative penalty be one that was not unduly harsh. By remaining on the tender list NMC was able to secure a right to participate in any tenders the two clients may in future issue.

[8] Arangies drew the attention of the Tribunal to the following data which he extrapolated fr

2009 2010 2011 2012 2013 Average Turnover 345834474 270406394 255714620 232725998 355805942 292097486 Cost of sales -295017700 -23667 402 7 -2309828 44 -207713338 -318809399 - 257839462 Gross Profit 59816774 33732367 24731776 25012660 36996543 34258024 Gross Profit % 14,69% 12,47% 9,67% 10,75% 10,40% 11,73% Profit for year 18923 45 0 9037216 3711479 5921832 13195648 10157925 Nett Profit% 5,47% 3,34% 1,45 % 2,54 % 3,71% 3,48% Dividends declared   1794908 901083 694100 531341   Current Assets   37605583 51907629 40256493 96939495

om the accounting records of lsipani:

Current liabilities	47368624	5117 3631	34455521	79 2341 95
Current Ratio	0.79	1.0 1	1.17	1.22

[9] It is to be noted that while the 2014 turnover of lsipani was not included in this extrapolation it was furnished in the record[1], as it is certainly relevant for the determination of an appropriate penalty.   The above table does not   reflect

that lsipani declared a dividend in the 2009 financial year. However, Arangies admitted that lsipani had declared a dividend of R9m in that financial year. He said this was done in a quest to meet its BBBEE (Broad Based Black Economic Empowerment) targets. Some part of this dividend went into a Special Purpose Vehicle so that it could improve lsipani's BBBEE status. The monies from the dividend were intended to benefit its employees as well as some community­ based groups. In this regard it has allowed for its Employees' Trust and Foundation to acquire a 25% ownership stake in the company. He did, however, acknowledge that one of the objectives of upgrading its BBBEE standing was to improve its chances of acquiring work from, amongst others, government-based projects.

[10] Arangies further claimed that lsipani's average profit for the five years 2009 - 2013 was only R10 157 925.00. This represented a profit margin of 3.48% of turnover. This is to be compared with the average profit margins in various related industries during this period. These were:

[10.1] mining and quarrying                                    16%

[10.2] manufacturing                                               5%

[10.3] electricity, gas and water supply                   25%

[10.4] trade                                                              5%

[10.5] transport, storage and communication          7%

[10.6] real estate and other business services        12%

[10.7] community social and personal services       11.5%

[11] lsipani also drew attention to the dividends it declared over the five-year period as well as to its current ratio during the same period. A current ratio is the ratio of current assets to current liabilities, arrived at by dividing the current assets by the current liabilities. A current ratio below one indicates that the firm will not be immediately able to pay all its current liabilities from the current assets. The current ratio, lsipani contended, is important for it provides a  guiding measure of a firm's ability to secure and perform future work. It reveals the liquidity of the firm and provides an insight into the ability of the firm to carry the expenses necessary to provide the work while awaiting payment- payment always following completion of all or part of the work. Rarely, if ever, does payment in construction projects precede performance by a contractor.

[12] In these circumstances, contended lsipani, it would be inappropriate for the Tribunal to impose an administrative penalty equivalent to 10% of its average turnover for the period 2009 - 2013. Such a penalty would be burdensome to the point of being seriously debilitating on its future business. It would result in the current ratio reverting "to a figure probably around the 0.79 mark of 2010" , which would make it difficult for it to carry out a large project. It would not have sufficient running capital to maintain the project bearing in mind that current liabilities have to be settled on a regular basis.

[13] lsipani went on to point out that the practice of cover pricing, pernicious as it may have been, was restricted to contracts where the construction companies were invited to tender for the contract (tender-based contracts). The practice was not to be found in negotiated contracts. The latter contracts simply provided no scope for parties to defeat the system of free competition and undermine the value and benefits that accrue therefrom. For this reason it would be only correct to utilise the turnover it acquired from tender-based contracts in order to determine what a fair and appropriate administrative penalty should be. The breakdown of its turnover during the three years of 2009-2011 was:

Turnover	2009	2010	2011
Tender-based contracts	163,925,155.15	180,162,052.50	205,695,883.55
Negotiated contracts	181,909,318.67	90,244,341.50	50,018,736.14
TOTAL	345,834,473.82	270,406,394.00	255,714,619.69
Percentage
Tender-based	47%	67%	80%
Negotiated	53%	33%	20%

[14] It is not clear why it supplied the breakdown of its turnover figures for these three years only. Nevertheless, the average of its total turnover for the three years was R290 651 829.17, whereas the average of its turnover acquired from tender-based projects for the same period was R183 261 013.74. If the penalty is to be based on the average turnover of this three-year period, then according to the logic propounded by lsipani it is the latter and not the former average that should be the benchmark for its determination.

[15] It argued  that  the  two  contraventions  for which it is being  held accountable did not subvert or undermine the public interest as lsipani did not acquire any benefit from the two cover prices having lost both bids. This, lsipani argued, must militate towards a lower penalty than that sought by the Commission. Finally, it argued that a penalty of 10% of turnover for the 2014 financial year as specified in s 59 of the Act as the upper limit is one that should only be imposed in cases where the unlawful conduct was egregious. This is  not so in this case, it contended. In essence, lsipani submitted that there were a number of mitigatory factors that bore significant weight to propel the administrative penalty away from the upper limit of 10% of total turnover  for 2014 as sought by the Commission.

[16] In contrast, the Commission took the view that the profitability as well as the motives of lsipani were irrelevant for purposes of determining the administrative penalty. It took the view that there is sufficient protection in the upper limit to prevent it from asking for, or the Tribunal from imposing, a penalty that is overbearing. Accordingly, the focus for determining an appropriate penalty should be the turnover in the year 2014. As far as the mitigating factors are concerned, these are not to be found in the financial data released and relied upon by lsipani but in the considerations highlighted in s 59(3) of the Act.

The approach of the Tribunal

[17] The first issue the Tribunal focussed on was whether it should impose two penalties rather than one as lsipani has admitted to two contraventions. It recognised that it could impose two penalties but, as mentioned above, it decided to consolidate the two into a single one. It reasoned thus:

"[19] We agree with the Commission that each instance of a cover price constitutes a separate self-standing infringement of the Act. This is consistent with the approach followed by the Office of Fair Trading and the Competition Appeal Tribunal in Kier Group PLC and others v Office of Fair Trading where each of the Appellants were fined separately for each cover pricing infringement.

[20] However, we have the discretion based on the  facts  of each  case in the interest of fairness and the doctrine of proportionality to decide how to levy an appropriate administrative penalty pursuant to s 59(3) of the Act. In this case we have decided to levy a single administrative penalty in respect of two separate incidences of cover pricing. In our decision we were cognizant of. the fact that there was a second, separate contravention of the Act in our final calculation of   the single penalty amount."

[18] In addressing the issue of an appropriate penalty the Tribunal took guidance from its approach in an earlier case, which approach was sanctioned by this Court. The earlier case was The Competition Commission v Aveng (Africa) Limited t/a Steeledale and others.[2] In that case the Tribunal came to the conclusion that the approaches adopted in previous cases by itself and by this Court, especially in Southern  Pipeline  Contractors  v  Competition  Commission[3]had evolved into a single approach consisting of six steps. The six steps to  be followed in terms of this approach are:

1.      Determine the affected turnover in the relevant year of assessment.

2.      Calculate the base amount, which is the proportion of the turnover referred to in 1.

3.      If the contravention exceeded  one year, the  amount obtained  in  2 should be multiplied by the duration of the contravention.

4.      If the amount obtained in 3 exceeds the upper limit prescribed in section 59(2) of the Act, then it should be rounded-off.

5.      Take aggravating and mitigating factors to either  increase or decrease the amount acquired at the end of 4.

6.      Round the amount off so that it does not exceed the upper limit prescribed in s 59(2) of the Act.

[19] According to Aveng the affected turnover " is based on sales of the products or services that can be said to have been affected by the contravention."[4] The year would be the last financial year of the period during which the contravention occurred, or to put it differently it would be "the most recent financial year in which there is evidence that the firm participated in the contravention.''[5] On the face of it, it would appear that performing the task required in step 1 would be fairly simple and straightforward. However, in this case there was a controversy between lsipani and the Commission as to which year should be regarded as the base year. lsipani contended that it should be 2011 as the contraventions took place August and November of 2010 and its Tribunal to hold that it be 2012, i.e. the financial year, which according to lsispani's own version, is when the contravening practice ceased.

[20] Upon carefully scrutinising the evidence, the Tribunal found that the Commission's submission was correct and held that it would have to take 2012 to be the base year.

[21] The Tribunal proceeded to consider which turnover of 2012 should form the foundation of the ultimate penalty imposed. lsipani claimed that it was only the turnover resulting from tenders that should be taken into account, as opposed to turnover arising from negotiated contracts. The Commission maintained that it should be total turnover as that is what the legislature intended. The Tribunal agreed with the Commission. It drew from this Court's approach in Southern Pipeline Contractors. It was also the approach  adopted by the Tribunal in its own judgment  in Southern  Pipeline Contractors.[6]  On this logic it concluded that the foundation should be the total turnover for 2012, which was R232 725 998.00. This, according to it, was the affected turnover.

[22] Having established the affected turnover (or the foundation amount), it turned its attention to the base amount (Step 2). The task required of it here is to calculate a proportion of the affected turnover, between 0% and 30%. On this issue, the Commission called for a base amount of 17% for each contravention, while lsipani  suggested  that it  be 10%  for both contraventions.  In terms of Aveng[7] the Tribunal was required to consider three principal factors: (i) the nature, gravity and extent of the contravention; (ii) the loss or damage that resulted from the contraventions; and, (iii) the market circumstances in which  the contraventions occurred. These factors are taken verbatim from s   59(3)(a), (b) and (c) of the Act.

[23] On this aspect of its task, the Tribunal commenced by noting that cover pricing is a form of bid rigging. Bid rigging, because of its uncomplicated nature, is particularly harmful. Offenders can engage in it without much effort and with  a reasonable amount of confidence that it would not be detected or exposed. It will only be eradicated, if ever, if the deterrence factor is sufficiently weighty. For this to be so, the possibility of facing a heavy penalty if caught is necessary to match the benefit of potential gain from the unlawful practice. This should make it very risky to engage in the practice. It must have a deterrent effect. Only then can the practice be eliminated.

[24] The Tribunal reasoned that as the practice was pervasive in the industry and caused a great deal of damage to the interests of society as a whole, the base amount should have been 30%. However, because lsipani did not benefit from the conduct (as it lost the bid), it would be appropriate to apply a base amount of 12%. Doing the arithmetic (12% of R232 725 998.00) the Tribunal came to a figure of R27 927 119.76 as the base amount.

[25] Executing the task in Step 3, the Tribunal came to the conclusion that a multiplier of 1.3 should be applied. This finding was based on the acceptance of the evidence of Arangies that the actual contraventions occurred during August 2010 and November 2010, but lsipani only ceased engaging in the practice of giving or receiving cover prices in November 2011. If this evidence is accepted  it means that factually the contravention exceeded one year, i.e. August 2010 - November 2011. The contravention lasted longer than a year - one year and three months to be exact. Therefore, according to the Tribunal, the multiplier to be applied should be 1.3. Multiplying 1.3 to R27 927 119.76, an amount of R36 305 255.68.

[26] Moving on the Step 4, the Tribunal found that the turnover for the year of 2014 should be taken into account to determine the 10% limit prescribed in section 59(2) of the Act. This is in stark contrast to the submission of lsipani, which was that the turnover of 2011 should be taken into account. lsipani was particularly concerned that since the transgressions of August and November 2010 it had increased its turnover substantially. Nevertheless, the Tribunal chose the 2014 turnover because it was the financial year that preceded the finalisation of the matter before it. There is logic to this choice. It derives from the application of the stare decisis principle. This Court pronounced that the 10%  limit  on the turnover  of  a  firm's  local  and  international  sales  of " the

preceding year' prescribed in s 59(2) of the Act refers to the year "preceding that in which the penalties are imposed. "[8]

[27] As for the mitigating and aggravating factors that need to be examined in Step 5, the Tribunal looked at the issues identified in s 59(3)(e), (f) and (g) of  the Act - i.e. the level of profit derived from the contravention; the degree to which the lsipani had co-operated with the Commission and with itself; and whether lsipani had previously been found in contravention of this Act. It found as mitigating factors the fact that lsipani was a first offender; that it did not directly profit from the two contraventions; that lsipani had admitted to acting unlawfully; that an undiscounted penalty would undermine lsipani's attempt to improve its BBBEE status and that such a penalty would diminish lsipani 's chances of future work. It weighed these against the aggravating factors, which were that a high level employee (a procurements director) was involved in the transgressions and that there were two separate transgressions. After " weighing up" all these factors it decided to grant a discount of 40% to lsipani. The 40% reduction resulted in the penalty decreasing from R36 305   255.68 to R21 783

153.40. As this is less than 10% of lsipani's 2014 turnover there was no need for a further rounding-off.

[28] The cross appeal

[29] The decision to consolidate the two contraventions into one for purposes of calculating an appropriate administrative penalty provoked an appeal from the Commission. It is referred to as a cross-appeal as it was delivered (filed and served)  after  lsipani  had  already  delivered  its  notice  of  appeal.  It  is   the Commission's case that the Tribunal should have imposed two instead of a single administrative penalty.

[30] The power of the Tribunal to impose a penalty on an errant party is one that lies within its discretion. It is a discretion that is wide and cannot be fettered even by its own Guidelines or policies.[9]

[31] The Commission had asked the Tribunal to declare that lsipani had contravened s 4(1)(b) of the Act by engaging in collusive tendering "during the period August to November 2010"[10] , and that the Tribunal declare further that lsipani " is  liable for payment  of an administrative penalty  equal  to 10% of its turnover in terms of s 58(1)(a)(iii) read with section 59(2) of the Act."[11]

[32] We know that lsipani confessed to two contraventions, one in August 2010 and one in November 2010. However, the call for a penalty of 10% of lsipani's turnover does not specify that it should be 10% of the turnover for each of the contraventions. Mr Ngalwana for the Commission pressed us to read this call for "a penalty equivalent to 10%" to mean a call "for a penalty equivalent to 10% for each contravention". He pointedly informed us that even if we were unwilling to read paragraph 2 of the motion as he encouraged, we nevertheless should bear in mind that in law the Tribunal was obliged to impose a separate penalty for each infraction by lsipani. Its failure to do so was a misdirection so severe that it was actually an error of law. The only way to remedy this error was to refer the matter back to the Tribunal for it to properly apply its mind to issue of an appropriate penalty for each of the infractions. Should the Commission succeed on this that would be end of the matter. It would  be neither necessary nor appropriate for this Court to highlight its view on the appropriateness of the quantum of any penalty that should be imposed lest it unduly influenced the decision of the Tribunal. The Tribunal should be allowed to consider the matter free of any constraints, direct or indirect.

[33] We know from [20] of the Tribunal's decision[12] that it understood its unfettered discretion to include the power to consolidate more than one penalty, in this case two, into a single one. From the Tribunal's perspective, as long as the ultimate penalty fell within the scope of s 59 of the Act, it would have acted within its statutorily conferred powers. It is empowered, in terms of s 59(1)(a)[13] to impose a penalty for a contravention of s 4(1)(b); the penalty must take   into account the factors and circumstances listed in s 59(3)[14]  and must comply with the restriction set in s 59(2)[15] of the Act. As long as the performance of its task was consistent with the provisions of these sub-sections, its manoeuvrability was elastic enough for it to combine two or more penalties into one. Viewed from a different angle, as long as the single penalty took account of all the contraventions and as long as the penalty took full account of everything said in s 59 of the Act it acted intra vires.

[34] The Commission, however, contended that embedded in the logic of the Tribunal lays an error of law. The error, according to the Commission, is this: the text of s 59(1) clearly refers to the power to impose "an administrative penalty only for a prohibited practice''. The reference to the prohibited practice is in the singular and not in the plural - it does  not  say  that  the  Tribunal  may impose a penalty for "prohibited practices". Thus, on a literal interpretation of the section, there can only be one penalty for a single prohibited practice. This is a legal requirement. By extension, if there are two prohibited practices there has to be one penalty for each, as each prohibited practice stands alone and must be accorded its own weight and value. By failing to recognise this legal requirement the Tribunal erred in law.

Did the Commission follow the correct procedure by claiming relief on  the grounds of an error of law through a cross appeal?

[35] At the hearing Mr Fagan for lsipani contended that there was no error of law committed by the Tribunal, but even if there was the Commission was not entitled to bring its complaint in the form of a cross-appeal.  It should have brought this complaint by way of review proceedings, citing the Tribunal as a party so that the Tribunal could be given an opportunity to explain its reasoning. This contention was not raised or foreshadowed in lsipani's heads of argument. As a result the Commission was given no opportunity to deal with it prior to the hearing. The Court too was taken  by  surprise.  lsipani's written submissions were submitted on 3 March 2017, while those of the Commission were  submitted on 20 March 2017. The hearing was on 29 March 2017.  lsipani did  not file supplementary  written submissions. Yet it raised the contention during   its oral submissions and expected this Court to accept and deal with it. I am  firmly of the view that it is not acceptable for parties to ambush each other and the Court in this way. It makes for poor administration  of  justice.  Mr  Fagan having raised the issue did not refer us to any authority  that  lays  down  a principle that a challenge to a decision of an administrative body, bearing the same power and status as the Tribunal, on the basis that it erred in law cannot  be brought by way of an appeal, or why a challenge of this nature can only be brought by way of review proceedings. Consequently, it is left to this Court to undertake the necessary research in order to justly determine whether the Commission should be non-suited for electing to follow an inappropriate procedure.

[36] There are many errors of law that have been brought to appellate courts by way of appeal. This is natural. After all, an incorrect inference (adverse or otherwise) drawn from established facts is an error of law. Similarly, an incorrect legal conclusion drawn from established facts, from an admission of inadmissible evidence or a failure to admit relevant admissible evidence are all errors of law. This list is not exhaustive. The law reports are replete with examples of such errors and most of them enter the portal of a higher (appellate) court by way of an appeal. In this case, the contention of error of law flows directly from the decision of the Tribunal and from the papers. No extraneous evidence is relied upon or required. Hence, I am of the view that the submission of Mr Fagan that the Commission approached this Court through an incorrect and inappropriate procedural route is without merit.

[37] Furthermore, having been taken by surprise, Mr Ngalwana for the Commission decided to provide supplementary written submissions after the hearing. Unfortunately, Mr Fagan did not do the same. Mr Ngalwana submitted that the Commission is entitled to approach this Court by way of cross-appeal as its complaint against the Tribunal is a legal issue not requiring any further evidence. However, he submitted that if it was wrong on this then this Court should take the approach adopted in a majority judgment of the Constitutional Court. The approach adopted was one that moved away from deciding a dispute through the narrow lens of "formalism". The judgment he drew attention was that of Cameron J, with the concurrence of Moseneke DCJ, Froneman   J, Khampepe J, Yacoob J, (the majority judgment) in KwaZulu-Natal Joint Liaison Committee.[16]

[38] In KwaZulu-Natal Joint Liaison Committee, the applicant sought relief to compel the MEC for Education in KwaZulu-Natal (the MEC) to pay subsidies to each of its members, who were all independent schools, on the basis that there was a bilateral agreement between it and the MEC obliging him to pay these subsidies. Cameron J found that no bilateral agreement was proved. However the obligation to pay the subsidies arose from "broader public law and regulatory grounds rather than bilateral agreement."[17]  The learned judge came to this conclusion despite the fact that the pleaded case was that  a    bilateral agreement had come into effect. The "public Jaw and regulatory grounds" that provided the ratio decidendi for the decision were not pleaded at all. The Court had this to say about that:

"[67] The applicant concedes that it did not specifically argue for relief on the basis of the obligation created by the regulation. However, its founding affidavit expressly invoked item 195 of the Norms, and reliance on statutory obligations was foreshadowed in its papers."

"[68] In support, the amicus submits that a claim arising from the Norms read with the KZN regulations is adequately pleaded. And, it says, the evidence on record lays a sufficient basis to find the applicant schools had a right to be paid the first-quarter tranche due on 1 April 2009. Further, this court has previously adopted remedies for a situation where a claim is apparent from the papers and the evidence, even if it was not the cause of action expressly advanced or argued. With this I agree. As in the cases the amicus mentions, there is no prejudice to the respondents here." [18]

[39] Having found that the applicant had succeeded on " broader public law and regulatory grounds" only, it was granted relief that was consistent with this limited success. Had it succeeded on the grounds of breach of a bilateral agreement the relief it would have secured would have been broader.

[40] Froneman J, in a judgment that concurred in part and dissented in part with the majority, elaborated on the theme eschewing formalism when deciding issues in a case:

"[79] How important is the legal label one attaches to a set of facts upon which a party relies for a remedy under the law? Not decisively so, I would suggest, in a matter where the facts are not essentially disputed and no material prejudice to any party flows from whatever label is assigned to them by the formality of the law. This is that kind of case, but the opposing parties urged us to attach different labels to the facts upon which relief was sought, and determine the outcome according to the label. The invitation should be resisted -   substance should count, not form.

[86] The applicant brought its application by way of notice of motion. Even if it chose review of administrative action as the formal label it was not obliged to use rule 53 - the procedural mechanism for a review - for that purpose. The rule exists principally in the interests of an applicant, and an applicant can choose to waive a procedural right. In this case, where a litigant brings proceedings against the state, 'the latter can always, in answer to an ordinary application, supply the record of the proceedings and the reasons for its decision.' There was thus nothing in the form of the proceedings in the high court that prevented the first and second respondents from producing the record of the budget allocation and decision-making in regard thereto, or anything else they considered relevant. They could have done it whether the claim was based in contract or in administrative law. The blame for their failure to do so cannot be laid at the applicant's door."[19]

[41] However, for the sake of completeness it must be said that Froneman J came to the conclusion that the applicant (KwaZulu-Natal Liaison Committee) had placed sufficient undisputed facts before the Court to show that it was entitled to relief sought in the law of contract. The learned judge found that  the MEC's conduct  was  such  that  it  resulted  in  him  assuming  a    contractual obligation.[20] On this basis, he dissented with the order, for according to him the order did not dispense complete justice.

[42] It bears mentioning that there were three minority judgments in Kwa-Zulu Natal Liaison Committee - penned respectively by Nkabinde J, Zondo J (with Mogoeng CJ and Jafta J concurring) and one jointly by Mogoeng CJ and Jafta J. Nkabinde J took the view that a party is bound by its pleadings and if the relief sought is not covered by its pleadings, it must fail.

[147] The purpose of pleadings, as dealt with by the Supreme Court of Appeal in Minister of Safety and Security v Slabbert, ([2010] 2 All SA 474 (SCA)) albeit in action proceedings, is apposite. The court remarked:

'The purpose of the pleadings is to define the issues for the other party and the court. A party has a duty to allege in the pleadings the material facts upon which it relies. It is impermissible for a plaintiff to plead a particular case and seek to establish a different case at the trial. It is equally not permissible for the trial court to have recourse to issues falling outside the pleadings when deciding a case.

There are, however, circumstances in which a party may be allowed to rely on an issue which was not covered by the pleadings. This occurs where the issue in question has been canvassed fully by both sides at the trial.'

Notably, in casu, the existence of a contract or quasi-contract was neither pleaded nor agreed upon by the parties nor was the broader public law ground pleaded." [21]

[43] It bears reminding that there is a wealth of well-established authority supporting this dictum of Nkabinde J.[22] that a Court is only entitled to decide "the case that the applicant brought the respondents to court to answer'.[23]

[44] We are bound by the majority judgment in KwaZulu-Natal Liaison Committee and have to agree with the Commission that its choice of cross­ appeal as the procedure for its complaint of error of law should be entertained if we are to adopt a " non-formalistic" approach. It has placed all the material relevant to this issue before this Court and dealing with it by way of cross­ appeal causes no prejudice to lsipani. In any event, I had already found that the procedure adopted by the Commission was not inappropriate.

[45] Further, I must emphasise that it was inappropriate for a party to raise its objection in the manner that lsipani did. The fact that I dealt with the objection does not mean that in future such conduct will be allowed. On the contrary, my firm view is that any party which fails to raise an objection in its written submissions (be they the initial or the supplementary ones delivered before the hearing) should not ordinarily be allowed to raise the said objection or take the said point in its oral submissions at the hearing. It must show exceptional circumstances to have the objection or point considered.

Is the Commission correct that the Tribunal erred in law?

[46] Based on the facts and arguments presented in this case, I am not  entirely convinced that the Commission is correct that the Tribunal erred in law. The one peg on which it hangs the argument is s 59(1)(a) of the Act. It is true this section is cast in the singular - "a practice" - but this is not decisive of the issue. Section 6(b) of the Interpretation Act 33 of 1957 provides that "words in the singular number include the plural, and words in the plural number include the singular'.[24] There is no suggestion that s 6(b) of the Interpretation Act is inapplicable in this case. Accordingly, the usage of phrase "a practice" may well be there for no reason other than linguistic consistency.

[47] In my judgment this is an issue that does not need to be definitively decided here as the facts of this case show that the approach adopted in evaluating the gravity of each of the two offences and the attitude of lsipani in both instances is identical. The result that follows, whether one treats it as two smaller individually decided penalties or a single large penalty, which in effect is the sum of the two penalties, is the same. I elaborate on this later when dealing with the appeal of lsipani. For this reason it is not necessary to definitively decide whether the Tribunal is not entitled at all to combine the two offences  into one for purposes of determining an appropriate penalty or penalties. Perhaps the facts that reveal themselves in other cases will warrant the adoption of the interpretation favoured by the Commission, and the establishment of a principle that it is always necessary to deal with each offence on its own merits but this is not the case.

[48] The Commission went on to say that in this case the Tribunal imposed a single penalty for the August 2010 contravention and used the November 2010 penalty as an aggravating factor. Hence, to the extent that it has discretion to impose a single penalty, it exercised the discretion injudiciously. In contrast, it should have imposed two penalties and then adjusted the sum of the two penalties in order to meet the requirements of proportionality and fairness. But it is here that the logic propounded by the Commission becomes problematic. If the end result of the combined penalty has to be one that is proportionate and fair then there should be no problem with imposing a single penalty, as long as there is no absolution inherent in the penalty for any contravention of which the firm was found guilty. There could be a legitimate argument to this effect if the single penalty imposed was too low considering that the firm is being punished for more than one contravention. In other words, the ultimate penalty must be fair and proportionate to the number of offences. In this case, the Tribunal asserted that it was. It said:

"In our decision we were cognizant of, the fact that there was a second, separate contravention of the Act in our final calculation of the single penalty  amount."[25]

[49] The penalty in other words was a cumulative one. The penalty was R21 783 153.40. Given that the gravity of each offence was of equal weight and the time they were committed was in the same financial year, the Tribunal could have decided to impose half of this penalty for each of the infringements and still have achieved what it believed was a fair and proportionate penalty overall. This is demonstrated further below when I look at the question of what I believe is a fair and appropriate penalty for both offences. For the moment it bears noting that the Commission does not say that the penalty of R21 783 153.40 was inappropriate for both penalties.  It cannot ask, and is not asking,  for A penalty of R21 783 153.40 for each offence. It is not even asking for a penalty higher than R21 783 153.40 for both offences. It merely asks for the matter to  be referred back to the Tribunal to be re-considered afresh on the basis that it should impose a separate penalty for each contravention. In other words, the error of law complained of is one on the merits and not one where the Tribunal failed to understand the nature of its discretion. But, we would only refer the matter back to the Tribunal if we were to find that the error it is supposed to have committed could be remedied with an instruction to issue two penalties and if the combined penalty would be different from the present one. Then too, we would also only send it back if the result would be indeterminate. I deal with this later.

[50] In other words, even if there was an error of law as the Commission contends, it must show that the error was material and resulted in an order (outcome) that is wrong. An appeal is directed at the order, not the   reasoning underlying the order. Hence, not every error of law is appealable or reviewable.[26]

The appeal of lsipani

[51] As the appeal of lsipani focussed on each of the steps followed by the Tribunal in the process of calculating the appropriate penalty it is necessary  to examine the Tribunal's approach in each step and this is what I now propose to do.

Which turnover should be accepted as "affected turnover''? (step 1)

[52] The Commission took account of the averment in the answering affidavit of Arangies as well as of his viva voce evidence before concluding that the base year should be 2012. In his answering affidavit, Arangies avers:

"lsipani ceased any involvement in cover pricing in response to the Commission 's letter of 23 November 2011.

[53] Viva voce he indicated that it was possible for lsipani to have reciprocated the favour granted to it by NMC by engaging in illegal conduct to the benefit of NMC. However, he did not go so far as to say that this actually happened. Even though there was no evidence of actual contraventions by lsipani post November 2010, it has to be accepted that according to its own version lsipani saw nothing wrong with providing or receiving cover prices until 23 November 2011 and it was only as of that date that it ceased engaging in this practice.

[54] In essence, the evidence is two-fold: there were actual contraventions by lsipani in August and November 201O; and, there were possible contraventions by lsipani between November 2010 and 23 November 2011. As regards the latter it is important to bear in mind that while there is no evidence of any other actual   contravention   post   this   period   there   is   an admission   that the contraventions only ceased as from 23 November 2011. Faced with this evidence the Tribunal came to the following conclusion:

"Thus given the evidence that the conduct took place in the year 2010, but  was  only  ceased  in November  2011,  the  turnover  for  the 2012 financial year should be used for calculating the penalty."[27]

[55] The Tribunal can hardly be faulted for taking Arangies at his word when he said that lsipani only ceased engaging in providing or receiving cover prices on 23 November 2011. His viva voce testimony did not contradict this. At best for lsipani the evidence was equivocal. He admitted that between November 201O and November 2011 lsipani may have furnished NMC with a cover price or received one from NMC. Arangies did not specifically and directly aver that the contravention ceased in November 2010, but he did aver directly and specifically that it did cease as of 23 November 2011. In my view, the Tribunal's assessment of the evidence is faultless. In any event if the turnover of 2011 was

taken into account as per the submission of lsipani it would ultimately result in a higher penalty as the 2011 turnover was higher than the 2012 turnover.[28]

[56] The question of whether only the turnover earned from tender-based contracts should be taken into account does not arise as the figure for the 2012

turnover is not split into tender based and non-tender based turnover.[29]

The calculation of the base amount (Step 2)

[57] In step 2 the Tribunal determined the base amount to be 12% of the affected turnover. It came to this amount by holding that cover pricing is a form of bid rigging and bid rigging is particularly harmful to the public interest, which calls for an appropriate deterrent-directed penalty. Whilst the Tribunal was not wrong to see cover pricing as a form of bid rigging, it bears reminding that not all cases of cover pricing are actually directed at, or achieve, the rigging of a bid to the extent that it affects the price that is eventually paid by the tenderer. Some of it, especially "simple"  cover pricing  rarely does.[30]

[58] Thus, while not all cover pricing is less serious than bid rigging " simple" cover pricing often is. In this case we are dealing with that only. The cover price provided to NMC by lsipani did not result in the rigging of the bid. It could have, but as lsipani lost the bid it did not. When I say it could have, I have in mind the situation where lsipani's bid was inflated because it believed that its only real competitor was NMC, and as NMC was provided with a cover price then in effect it no longer faced any real competition, and it could, without fear of losing the bid, inflate its price. Had it done so and had it won the bid, the tenderer would be no wiser and would merely have suffered the prejudice. As that did not occur in this case a base amount of 12% of the affected turnover, in my view, is a bit on the steep side. I think 10% would be more appropriate. This is what lsipani argued for at the Tribunal.[31] Thus, I would reduce the base amount from R27 927 119.76 (12% of R232 725 998.00) to R23 272 599.80.

The multiplier if the contravention exceeded one year (Step 3)

[59] The Tribunal came to the conclusion that a multiplier of 1.3 should be applied. This conclusion arose from its acceptance that the contraventions took place during August 2010 and November 2010, but lsipani only ceased engaging in the practice of giving or receiving cover prices in November 2011 resulting in the contravention exceeding one year, i.e. August 2010 - November 2011. However, even on the Tribunal's finding it erred in its calculations. The multiplier of one year and three months is 1.25 and not 1.3 as arrived at by the Tribunal. The Commission agrees that the Tribunal erred in this step. Hence, applying the factor of 1.25 to the base amount of R23 272 599.80, the amount should be R29 090 749. 75.

Rounding-off (Step 4)

[60] There is no need to round this amount-off (Step 4) as it is less than 10% of the 2014 turnover. The 2014 turnover is R378 807 544.00.

Mitigating and aggravating factors (Step 5)

[61] The Tribunal looked favourably at lsipani because: it was a first offender; it did not directly profit from the two contraventions; that it admitted to acting unlawfully; and, that an undiscounted penalty would undermine lsipani's attempt to  improve  it BBBEE status. It was however disturbed by the fact that a high level employee (a procurements director) was involved in the transgressions and that there were two separate transgressions. Taking all of this into account, it reduced the penalty by 40%.

[62] Arangies' evidence under cross-examination by Mr Quilliam for the Commission revealed:

"Quilliam: Now, when you say you were aware of the simple cover pricing arrangement in the industry and it was so pervasive that one of your colleagues actually studied it before the introduction of the Competition Act, I believe you mentioned a textbook in 1985, how else did you become aware of this practice? When you entered the industry, was this practice explained to you by a colleague or did you just observe it in the industry. How did you become aware of how this practice worked?

Arangies:        Yes, I just experienced it.

Quilliam:            just experienced it from observing other firms.

Arangies: No, but I didn't work for other firms. I've worked for lsipani since I was in matric. So I just experienced it in our firm.

Quilliam:          Within lsipani itself?

Arangies:       Yes, then called Van der Sluys

Quilliam:   Then Van der Sluys, I understand. By observing it  within  your firm, did you have a clear understanding of who the other firms were that specifically were involved in this practice and who have decided not to be engaged in this practice?

Arangies :        No, I don't know.

Quilliam:  No, so there was no meeting in your memory that all the   firms came together and decided the rules of the game for cover pricing, simple cover pricing in the construction industry?

Arangies:        No, never.

Quilliam:    To your mind was there any overarching agreements by   any other firms, which may or may not have included lsipani, on how simple cover pricing would work and how tenders would be allocated ... I beg your pardon, how tenders would be dealt with in terms of simple cover pricing arrangement?

Mr Arangies: It might have been. I've read in the newspapers, but certainly us not.

Quilliam: So, you weren't aware of a single or a  number  of  agreements between any number of firms that set the rules for simple cover pricing and how certain tenders would be dealt with under the simple cover pricing arrangement.

Arangies:         No, not at all. ..."

[63] In his answering affidavit Arangies says this:

"There are two reasons, I believe, why lsipani did not realise that it had contravened the Act. The first is because the practice of providing cover pricing was so pervasive in the construction industry that it did not strike lsipani and (certainly not me) as being possibly unlawful." (Emphasis added)

[64] An analysis  of  the  evidence  in  toto  demonstrates  that  his viva voce testimony was far from satisfactory. Viva voce he says this: the practice of providing a cover price, though illegal, was pervasive in the construction industry; he, himself had experienced it but he does not know of any firms that were engaged in the practice. This at the very least is bewildering. He cannot at once say that the practice is "so pervasive in the construction industry" and he has experienced it but he does not know of any firms that engaged in it. For him to be able to confidently say that the practice is "so pervasive in the construction industry'' he must know of other firms engaging, or having engaged, in it. Moreover, he goes further and says he had experienced it. In which case he cannot avoid having to identify where, when and how he had experienced it so that he can safely and fearlessly conclude that "so pervasive in the construction industry'' . Hence, when he asked to identify any other firms engaged in it he cannot say he does not know of any and expect that answer to be reconcilable with his earlier answers as well as his averment in his answering affidavit. Put differently, if he wanted to hold on to the answer that he does not know of any firm  other  than  lsipani  engaging  in  the  practice  then  he  had  no  basis for claiming that it is " pervasive in the construction industry'' and more importantly, he could not say he has experienced it.

[65] It may be that he intended to convey the fact that it was only pervasive within lsipani (previously operating under the name and style of Van der Sluys). But if that is the case, then again he had no basis for saying that the practice was " pervasive in the construction industry'' as lsipani was only one firm in the industry. Furthermore, if the practice was a pervasive part of lsipani's modus operandi then he should know of firms other than NMC who colluded with lsipani in providing cover prices. Yet he claimed that lsipani only engaged in cover pricing on the two occasions that the Commission was able to find out about - i.e. the August and November 2010 tenders in Stellenbosch. The two occasions do not make for a " pervasive practice" . In essence, there is clearly a rational disconnect in his evidence.

[66] When pressed on this issue further with a direct and simple question he claimed to be ignorant. The question and answer were:

"Quilliam:  What I'm trying to  distil here is whether  there was any   other instances where lsipani requested a cover price from someone such as NMC or provided a cover price to anyone else"

Arangies:        I'm not sure"

[67] Unfortunately, this answer, too, is problematic. At the commencement of his evidence and in his answering affidavit Arangies adamantly averred that the only two cases of illegal conduct that lsipani had engaged in were those identified by the Commission - the August 2010 and the November 2010 cases involving NMC. If this was the case, then it is not clear why Arangies equivocated in his answer.

[68] On  balance  then,  either  Arangies   was  very  confused   or    he  was economical with the truth. It is difficult to see how a person in his position, a managing director of a very large firm making a turnover of a quarter to a third   of a billion rands a year, could be confused about a simple matter of whether his firm was engaged in the illegal acts of providing, or receiving a cover price  in  any case other than the two identified by the Commission. Moreover, he could  not be confused about whether the practice was pervasive in the industry or not. He claimed to have been in the industry his entire working life, which exceeds a period of twenty-five years, and that he was more than just familiar with the operations and practices that endure in the industry. In fact, he placed his reputation and experience in the industry before the Tribunal and asked that it take him at his word because these were extensive. In these circumstances , it cannot safely be concluded or inferred that he was confused about whether the practice of cover pricing was pervasive in the industry or not. A factual issue of this nature is too simple for him to have been confused about.

[69] In my judgment, the only conclusion to be drawn from this contradictory and equivocal evidence is that Arangies was not candid with the Tribunal. As a result, lsipani (as he was its sole witness, and it based its entire case on his testimony) had relinquished any claim it may have had for leniency from the Tribunal.   Despite  this, the Tribunal decided to show considerable leniency   by discounting the penalty by 40%. The Tribunal was not parsimonious with the discount - 40% is no miserly discount. I am of the view that it falls within the margins of a fair and reasonable discount given lsipani's lack of willingness to expand on its contention that the practice was pervasive for fear of revealing names of other transgressors. By refusing to do this, lsipani essentially informed other firms in the industry that it could be trusted never to reveal names of any of them if it knew that they have engaged in, or knows that they are engaging in, illegal conduct. Its reticence to reveal names translated into an unwillingness to cooperate with the Commission.

[70] Mr Fagan placed substantial emphasis on the fact that no one suffered any loss as a result of the unlawful conduct. This, he contended, should result in lsipani receiving a hefty discount when determining the final penalty in step 6. This issue however, has been addressed in Step 2 where the base amount has been reduced from 12% to 10%. In fact. lsipani argued that the base amount should be reduced to 10% of the affected turnover precisely because the bid was not successful and therefore no harm was suffered by the University. It was successful in  this argument.[32]   It should not be asking to benefit twice  from a single fact.

[71J Mr Fagan SC emphasised the fact that lsipani derived  no  financial benefit from the illegal conduct. This, of course, redounds to the benefit of lsipani -  it is a factor of some mitigatory value. But this has been catered for  in the decision of the Tribunal when it awarded a discount of 40% to lsipani. It is no mean discount and I hold that it is appropriate in this case. The discount of 40% in my view is substantial and fair in the circumstances of this case. While I am not insensitive to the fact that lsipani did not win the bid, I do not overlook the fact that its intention was to win the bid and it was willing to act unlawfully to do so. Its failure to win the bid does not diminish its culpability. Hence, I see no misdirection on the part of the Tribunal in awarding a discount of 40% to lsipani.

[72] Applying a 40% discount to the amount of R29 090 749. 75 arrived at in my calculation at the conclusion of step 3, the penalty becomes R17 454 449.76.

[73] To return  to  the  cross  appeal:  If  the  penalty  of  R17  454 449.76  is imposed  for  the  first  offence  - August  2010  -  only  then  the  process  of determining an appropriate penalty for the November 2010 offence has to be resumed. This would be done by repeating the 6-step approach outlined above. In many respects the factors referred to in [52] – [72] would be restated. The outcome would, at best, be marginally different. The two offences are closely related, if not identical, in gravity. The period in which they were committed is a single financial year (so the affected turnover (step 1) would be the same; the factors to be taken into account in the rest of the steps would be almost identical (there may a minor tweaking in the analysis of the second contravention), and finally the upper limit (step 6) would be the same. Assuming for analysis purposes only that the penalty  for the second offence (November 2010) comes to the same amount as arrived at for the first offence, i.e. R17 454 449.76. If each is imposed separately then the cumulative amount would be a penalty of R34 908 899.52. Before the order is made it would, in terms of s 59(2), be necessary to re-visit Step 6 to ensure that it does not exceed 10% of lsipani's annual turnover for 2014.

[74] The Commission agreed that if the cumulative penalty were to be 10% of the 2014 turnover it would be too high for these two offences as they simply do not constitute "egregious conduct."  The 10% upper limit “is reserved  for the most egregious conduct” where there is an "absence of any  mitigating factors."[33] The cumulative  penalty  of  R34  908 899.52  translates  into  9.2%  of the  2014  turnover.  This, too,   would  be  inordinately   high.  It would  not   " be proportional in severity to the degree of blameworthiness of the offending party, the nature of the offence and its effect on the South African economy in general and consumers in particular'[34] ,  but it must be high enough to have a deterrent effect. In this case, a penalty of 5%, or thereabouts, of the 2014 turnover for both of the contraventions would be a fair and reasonable penalty. It would take note of the gravity of the offences; it would satisfy the need to punish the offender sufficiently to ensure that the requirement of deterrence is not lost, and it would be fair to the offender. In short, it would serve the interests of justice.

[75] The amount I have arrived at is R17 454 449.76. This constitutes 4.6% of the 2014 turnover. Whether one divides this equally into two to take account of each contravention and then order lsipani to pay a penalty of R8 727 224.88 for each contravention or to a pay a total of R17 454 449.76 for both contraventions is really of no moment.

[76] Finally, the penalty of R17 454 449.76 that I have arrived at is  significantly  lower  than  the one imposed  by  the  Tribunal,  which is R21 783 153.40. It is startlingly different and allows for this Court to exercise its very limited powers on this issue and interfere with the Tribunal's discretion. While the only difference between the approach I adopt and the one adopted by the Tribunal relates to the base amount (Step 2), however, the end result is a difference of R4 328 703.64. This is a substantial amount of money. In this circumstance referring the matter back to the Tribunal would be costly, time consuming and of no benefit to the parties or the Tribunal.

Conclusive remarks

[77] At the hearing we received a stridently expressed submission from Mr Fagan criticising the Tribunal for, according to him, "mechanistically applying the six-step approach" devised by the Tribunal in Aveng. This " mechanistic application" he contended resulted in a disproportionate penalty. However, as the discourse between him and the Court progressed the criticism really melted away. When pressed to identify what exactly the Tribunal did wrong he said its choice of 12% of affected turnover as a base amount (in Step 2) was too high, and its discount of 40% (in Step 5) was too mean. Thus, the complaint was no longer that the six-step approach was "mechanistically applied' in this case, or that it ought not to have been followed in this case, it was that its application could have produced a more favourable result for lsipani had the Tribunal been slightly more lenient towards it in taking into account the required factors in Steps 2 and 5.

[78] It needs to be stressed that the exercise of determining an appropriate administrative penalty is, like sentencing in a criminal matter, case-specific. It is not, and can never be, scientific. When different penalties in different cases are compared, there will at times be surprise  and even disbelief. Some of this will  be justified. In order to avoid arbitrariness and capriciousness to creep into the decision-making it is important that the Tribunal and even this Court apply an objective approach to the matter. The six-step approach devised by the Tribunal in Aveng goes a long way towards achieving this. It has built-in flexibility; its application requires rigorous and comprehensive thought; it requires the full­ reasoning underlying the determination to be articulated; it provides substantial material to the affected parties to challenge the decision on objective grounds should any of them be aggrieved; it allows non-parties to understand the approach and logic of the Tribunal (and of this Court); it respects the legislature's intention by heeding the factors outlined in s 59 of the Act, and finally, but most importantly, it advances that which is  central  to  our constitutional order - the rule of law. Applied carefully and rigorously the resulting penalty is bound to be proportionate. Proportionality, in my view, is not a self-standing consideration. It is a function of the combined and duly weighted factors referred to in s 59 of the Act. It is for these reasons, too, that the Commission should take advantage of the six-step approach when considering what would be a fair and reasonable settlement with parties who wish not to engage in a bruising adversarial tussle.

[79] No doubt in time the approach will be refined or even modified, but for  now it provides a good working basis. There may at some future point be cases where it cannot or ought not to be followed as its application, or the outcome of its application, would not serve the interests of justice. In that case it should not be followed, but full reasons for adopting this course must be provided and the approach that is followed must strive to achieve all the benefits of the approach referred to in the previous paragraph.

Costs

[80] lsipani has been successful in the appeal. The Commission has not been successful in the cross-appeal. However, the Commission's opposition to the appeal was not without merit, nor was its cross-appeal. As with lsipani, the Commission brought helpful written and oral submissions that were of great assistance in the formulation of the ideas expressed in this judgment. Both legal teams are thanked for this. It is important not to mulct a statutory body which opposes an appeal and stands by or defends " honest and reasonable decisions made in the public interests."[35] In the circumstances, I believe that it would be just and fair for each party to pay its own costs of both the appeal and the cross-appeal.

[81] I had completed this judgment on 10 April 2017. It is unfortunate that it  took so long to be handed down. The judgment and the order I propose dissents from that of the majority. This is regrettable. It is my view that the penalty imposed by the majority is too low given the conduct of lsipani. I also hold that it is too low to have any dissuasive force. In my view the penalty I propose takes account of all the considerations that need to be taken into account and that it balances all the interests and the relevant factors, such as the interests of lsipani. the conduct of lsipani, the gravity of lsipani's offence, the interests of the Commission and the public interest in general, especially its interest in a corrupt-free market.

Order

[82] This is a minority judgment. However, I would have made the following order:

1.    The order of the Tribunal is set aside.

2.    The appellant is found to be guilty of two counts of contravening section 4(1)(b)(iii) of the Competition Act 98 of 1998.

3.    The appellant is ordered to pay an administrative penalty in the sum of Seventeen million four hundred and  fifty-four thousand four hundred and forty-nine rands and seventy-six cents (R17 454 449.76).

4.    The cross-appeal is dismissed.

5.    Each party is to pay its own costs.

Victor AJA (with Davis JP concurring)

[83] I have had the pleasure of reading the judgment of my colleague Vally AJA. Regrettably I cannot agree with our brother.

[84] The issues for determination in this appeal is whether the penalty was too high and in the cross appeal whether the infringements should be considered as two separate infringements thus affecting the penalty imposed.

Cover pricing in the context of the Construction Industry at the time of the infringement

[85] The cover price infringements occurred in 2010. Mr Arrangies conceded in cross examination that it only became clear to him after receipt of the Commission's letter of 23 November 2011 that cover pricing was a problem. According to his testimony had he been approached to provide a cover price up to that point he would have found nothing legally untoward about cover pricing. This acknowledgement must be considered in context before adverse inferences can be drawn. At that time cover pricing was endemic, regarded as legitimate  in  the  industry  and  was  pervasive  throughout  the  industry. This perception was pervasive in the United Kingdom as well as confirmed in Kier Group pie & another v Office of Fair Trading [2011] CAT.

[86] At the time the cover price was a reciprocal ad hoe favour extended to a competitor. In this case there was no central control or orchestration by companies in respect of tenders as to amount price fixing and other cartel activities. There was no form of compensation for providing the cover price. Despite extensive cross examination at no stage was the Commission able to establish that the infringement was intentional and made in order to distort competition. At the time companies were fearful of being taken off the tender  list if they did not submit a tender. This was the same position for lsipani. This led to a situation where companies would supply a tender price even if they would not have been able to take on the work.

[87] The evidence of Mr Arrainges can be characterised as amounting to a general ignorance that cover pricing was an infringement. This was consistent with so many others in the construction industry not only in South Africa but also in the UK. As is described in Keir, supra, as recently as 2006 cover pricing was included in textbooks as anything other than a normal practice. Whilst cover pricing is clearly an infringement its level of egregiousness in the hierarchy of infringements  has to be taken into account: that is made clear in Keir:  'simple cover pricing the respondent has already made its own unilateral decision not to compete in the work before the request for a cover price is made'.

[88] In considering an appropriate penalty it is necessary to see whether the end result is proportional. The analysis in Kier is helpful in order to assess the proportionality of the ultimate outcome. Cover pricing may deceive the customer about the extent of the competition and could have anticompetitive effects. It  is clear from the decision in Keir, that it ranks low in the pantheon of anti­ competitive offences.[36]

[89] The facts do not support a conclusion that lsipani was aware of the illegality and intentionally engaged in the infringement resulting in a restriction or distortion of competition. There are a number of other reasons why the penalty imposed is startlingly inappropriate. In this case the high point of the evidence established that the conduct of the lsipani was committed negligently and not intentionally. It is negligent because lsipani ought to have known that the conduct would result in a restriction or distortion of competition.

[90] This court can interfere with an administrative penalty on appeal if the discretion was exercised on the wrong principle See MacNeil Agencies v Competition Commission 121.CACJul12. That having been said, it is obviously necessary to impose a fine sufficient to deter lsipani and other companies to refrain from the cover price infringement.

[91] Regard ought to be had to comparative penalties in cases of cover pricing.  The  test is  whether  the  result  is  fair  and proportional. In Stanley's Removals [37] , for eight infringements of cover pricing the Tribunal imposed a penalty of 3.5% of its turnover. In cases where the infringer consented to a penalty in terms of the fast track process, the following fines were imposed: Competition Commission v Civcon Construction 1% of its turnover payable in six instalments; Competition Commission vs Giuricich Coastal Projects Pty Ltd 1.25% of annual turnover payable in six monthly instalments. The Tribunal imposed a penalty 9.4% of lsipani's turnover thereby characterising it as close to the most egregious infringement and out of line with the other penalties for infringements which were more serious.

[92] Once the indication is that the penalty is not proportional, it is necessary to consider where the Tribunal erred. The appropriate starting point would be to consider each of the indicated steps.

Step One

[93] lsipani raises two  issues  in relation to Step One. The first is whether  2011 or 2012 should be the appropriate year of assessment. The second issue raised in step one is the failure by the Tribunal to take into account the difference between negotiated and non-negotiated turnover. In relation to the appropriate year issue, Mr Arrangies conceded that he would have assisted NMC with a cover price until the Notice of Infringement. Clearly his intention to continue to participate in cover pricing continued until the notice of infringement of 23 November 2011. It is undisputed that there was no further infringement.

[94] The second issue raised in Step One was the failure by the Tribunal to differentiate between negotiated and non-negotiated tenders. In Omnico (Pty) Ltd & another v Competition Commission and others [38] this court made it clear that affected turnover was to be taken into account when imposing a penalty. In this case the affected turnover must be the non-negotiated tenders. The schedule shows that the non-negotiated tenders amounted to R205, 605,883.55 for the year 2011. The interpretation of affected turnover is also applied in UK construction companies which switch between non negotiated and negotiated tenders. In this case there was no evidence that lsipani could switch from a non-negotiated to negotiated tender. In G F Tomlinson Group Ltd & another v Office  of  Fair  Trading  and  similar  cases[39]   dealing  with  negotiated  and  non­ negotiated tenders, affected turnover was not taken into account where an entity could switch from a non-negotiated tender to negotiated tender. I agree with lsipani's submission that in this case there can be no switching within the same project and therefore affected turnover must be taken into account. The affected turnover principle has also been established in Omnico.   [40]

Step Two

[95] The Tribunal used 12% as it base amount. It is unclear how it arrived at this percentage. Clearly cover pricing is at the lowest end of egregious conduct. The Tribunal applied a percentage of 12% of the 30% it could have applied. lsipani's conduct falls to be assessed at most at one third of the 30%. 12%  is too high a starting point when considering the nature of this infringement. If the starting point is too high in step two then the end result will result in a distortion  of the end result.

Step 3

[96] The Tribunal erred in using a multiplier of one year and four months (1.3). The actual duration was from August 2010 to 23 November 2011. This is one year three months which should be a multiplier of 1.25 of the affected turnover.

Step Four.

[97] lsipani  submitted   that  the  appropriate   year  to  decide  the  'preceding financial year 'for purposes of Step 4' is 2013. The fine was imposed on 18 July 2016. There is no evidence to suggest that the lsipani engaged in cover pricing  for a long duration. It is also not lsipani's fault that the matter was finally adjudicated 5 years after the infringement. There is no  suggestion  that  the lsipani delayed matters save that it failed to agree to  a consent  order.  In this case the failure to agree a consent order should not result in unduly punishing lsipani. Its right to test whether cover pricing was an infringement was not opportunistic in the light of the wide spread practice and it being taught at University level. In considering whether  the statutory  cap  had been exceeded the   Tribunal   chose   2014   which   is   a   year   with   a   higher   turnover     of R378 807 544. Notwithstanding the factors submitted  by  lsipani justifying  the relevant year to be 2013 ultimately there is no reason to disregard the choice of 2014 by the Tribunal.

Step Five

[98] Within the context of cover pricing being an infringement it was necessary for the Tribunal to take all factors into account and attribute fair weight to all the factors. The Tribunal found the aggravating factors included the role of the procurement director and that Mr Arrangies was aware of the fast track settlement process. As stated earlier this contravention was not done intentionally. Although senior personnel were involved there is no evidence to suggest that they did so with the intention to distort competition. The weight attached to the second infringement as an aggravating feature must be considered in the light of the continuous belief (although incorrect) that cover pricing was not illegal. The tribunal reduced the penalty by 40%. There are a wide range of mitigating factors some of which go to the core of what ought to be taken into account for mitigation. In considering that only 40% was allowed for mitigation , it means that the aggravating factors played a far more significant role for the Tribunal. Less weight was attached to the mitigating factors. In my view this is inconsistent with arriving at a proportional result.

[99] There  are many  mitigating  factors  of a  substantial  nature.  Significantly this would include the successful continuation of the business and the preservation  of jobs.  lsipani cooperated.  It has a very  low level of profit  margin to turnover. The profit margin was lower than the industry norm for that period. In 2011 it was 1.45 %. In Mac Neil Agencies (Pty) Ltd [2013] 2CPLR 416(CAC) this court found that a low profit margin to turnover was relevant. This approach was also adopted in the cases of Kier Group, Barrett and Tomlinson, supra.

[100] The BBBEE initiative would be set back as no dividends could be declared to repay the loan of the BBBEE participants. A high penalty would impact negatively on the future of contracting opportunities especially with Government projects. lsipani must have sufficient capital to retain future work. It wanted to achieve a certain current asset level so as to be registered as a CIDB grade 9. It had to have at least R40 000 000 capital available. An overly harsh penalty would set back the goal of moving to a CIBD 9 level. It would set back the BBBEE initiative. In my view this is an important consideration. There is no evidence to support the conclusion that lsipani could receive sponsorship to reach CIBD level 9 classification.

[101] In summary, this court has previously cautioned against the trap of applying a mechanical interpretation to the six steps. The six step is a framework. Cognisance must be taken of the context in which the contravention took place, the nature of the infringement and the nature of the product and its effect on consumers. The scheme was not part of a scheme centrally controlled over a period of time. There was no evidence of lsipani's conduct inflating the level of the winning bids. There were no compensation payments. lsipani was not a repeat offender in the true sense. In all probability the cover pricing would result in a losing bid.

[102] From my analysis it follows that cover pricing does not stand to be regarded as equivalent to hard core bid rigging. Cover pricing was widespread and endemic throughout industry. Ultimately the penalty must have a deterrent effect and yet be proportional.  A deduction of 60% is fair in the circumstances.

Summary on the penalty

[103] The affected turnover is R205, 605,883.55. In applying a base amount of 10% this equals R20 560 588. The correct multiplier should be 1.25% which produces a  figure  of  R25 700 735.  Rounding  off  the  2014  turnover  of R378 807 544 by 10% amounts to R37 880 754. There is no need to round off in this step. A discount of 60% applied to R25 700 735 means that R15 420 440 stands to be deducted from R25 700 735.  The  appropriate  penalty  is  thus R10 280 295 which in my view is sufficiently severe without being disproportional.

The Cross Appeal

[104] The Commission sought a single penalty in its Notice of Motion. The Tribunal does have a discretion to impose a single penalty for more than one infringement. The imposition of a single penalty for multiple infringements has been applied in other cases. See Stanleys supra. In my view there is no basis to overturn the Tribunal's finding that the two infringements must be considered separately for the purpose of imposing the penalty.

In the result the following order is made

1.       The appeal is upheld

2.       The fine is changed to R10 280 295.

3.       The Competition Commission is ordered to pay the costs.

4.       The cross appeal is dismissed with costs

_________________

DAVIS JP

_________________

VICTOR AJA

_________________

VALLY AJA

Date of hearing:                                    30 March 2017

Date of judgment:                                 14 September 2017

For the Appellant:                                 Adv E Fagan SC

Instructed by:                                        Werksmans Attorneys

For the Commission:                            Adv V Ngalwana SC with Adv A Bodliani and Adv L Quilliam

Instructed by:                                        Ndzabandzaba Attorneys

---