Competition Commission of South Africa v Federal Mogul Aftermarket Southern Africa (Pty) Ltd and Others (08/CR/Mar01) [2003] ZACT 43 (21 August 2003)

COMPETITION TRIBUNAL

REPUBLIC OF SOUTH AFRICA

Case Number: 08/CR/Mar01

In the matter between:

The Competition Commission of Applicant

South Africa

and

Federal Mogul Aftermarket Southern Africa (Pty) Ltd

1st Respondent

Federal Mogul Friction Products (Pty) Ltd

2nd Respondent

T & N Holdings Ltd

3rd Respondent

T & N Friction products (Pty) Ltd

4th Respondent

Administrative Penalty and Reasons for the Competition Tribunal’s Decision

(non-confidential version)

Introduction

1. In our earlier decision on the merits of this matter we found that the first respondent had contravened section 5(2) of the Act.[1]

2. We postponed the issue of remedies for further evidence and argument. The only further evidence that we received was an affidavit filed by the first respondent from its managing director Mr. Frederick Nel.

3. The Commission has asked us to impose an administrative penalty upon the first respondent of eight million five hundred thousand Rand (R 8 500 000.00) The Commission had also sought a permanent interdict against the first respondent, but it later abandoned this prayer, so we need only consider the appropriateness of the administrative penalty remedy.

4. The Tribunal’s power to impose an administrative penalty arises from section 59(1) and (2) of the Act which states as follows:

(1) The Competition Tribunal may impose an administrative penalty only –

(a) for a prohibited practice in terms of section 4(1)(b), 5(2) or 8(a), (b) or (d);

(b) for a prohibited practice in terms of section 4(1) (a), 5(1), 8(c) or 9(1), if the conduct is substantially a repeat by the same firm of conduct previously found by the Competition Tribunal to be a prohibited practice;

(c) for contravention of, or failure to comply with, an interim or final order of the Competition Tribunal or Competition Appeal Court; or

(d) if the parties to a merger have –

(i) failed to give notice of the merger as required by Chapter 3;

(ii) proceeded to implement the merger in contravention of a decision by the Competition Commission or Competition Tribunal to prohibit that merger;

(iii) proceeded to implement the merger in a manner contrary to a condition for the approval of that merger imposed by the Competition Commission in terms of section 13 or 14, or the Competition Tribunal in terms of section 16;or

(iv) proceeded to implement the merger without the approval of the Competition Commission or Competition Tribunal, as required by this Act.

(2) An administrative penalty imposed in terms of subsection (1) may not exceed 10% of the firm’s annual turnover in the Republic and its exports from the Republic during the firm’s preceding financial year.

5. At the outset the first respondent has taken a series of constitutional points concerning the competence of this type of remedy and it is these issues, which we first consider.

SECTION A - Constitutional point

6. The first respondent argues that section 59 of the Competition Act is unconstitutional in three respects.

7. Firstly, section 59 of the Competition Act authorises the Tribunal, in the circumstances set out in the section, to impose administrative penalties on firms found to have contravened the Act. These penalties, argues the first respondent, are civil in ‘name’ only, but are criminal in ‘kind’ as they are punitive in nature. For that reason a respondent, in prohibited practice proceedings where a penalty remedy is sought, should have the same rights as an accused person has in criminal proceedings, rights that are guaranteed by section 35(3) of the Constitution. The first respondent argues that on a proper examination these protections are absent from the Act in several material respects. We deal with these in more detail later.

8. Secondly, and we understand this as an alternative to the first point, that even if section 35(3) does not apply to a respondent in prohibited practice proceedings, then section 34 of the Constitution, which provides for the rights of persons to have access to courts, does. Yet again there is, in the view of the first respondent, a want of constitutional compliance, as the Tribunal is not an “independent and impartial tribunal or forum” as required by that section.

9. The third submission is that section 59 is substantially irrational and accordingly unconstitutional to the extent that it provides that a firm’s turnover should be used as a basis for determining a maximum penalty in terms of that section.

10. The Commission and the amici [2] submitted that all three constitutional points were without merit and that we should reject them. The Commission, however, argued that we have no jurisdiction to determine the constitutionality of our own statute, as we are not a court with the status of a High Court.[3] The first respondent argued that all it sought was for us not to exercise our powers in terms of section 59. The tribunal is permitted, argues the first respondent, to decide whether section 59 is unconstitutional. If we decide it is not, we must not implement it.

The amici in this respect took the side of the first respondent and, whilst conceding that there was no direct authority on this point, argued that some other decided cases might be persuasive, including a decision of the Canadian Supreme Court in Douglas /Kwantlen Faculty Association v Douglas College where the court held that:

“A Tribunal must respect the Constitution so that if it finds invalid a law it is called upon to apply, it is bound to treat it as having no force and effect.”[4]

11. As we have decided that the section is not unconstitutional we do not need to decide this jurisdictional point, and we have proceeded on the assumption that the first respondent and the amici are correct in this respect. In any event, we believe that should this matter be taken to another forum, that forum may find it’s useful to have the perspective of the body responsible for adjudicating competition matters.

Background

12. Prior to the Competition Act coming into force in 1999, its predecessor, the Maintenance and Promotion of Competition Act (Act No. 96 of 1979) had a bifurcated approach to what the present Act defines as prohibitive practices. Under the old Act, some restrictive practices were susceptible to civil remedies, which did not include penalties, whilst others, including resale price maintenance, were made offences and could only be enforced by criminal law.

13. There were few if any criminal prosecutions under the repealed 1979 Act. It is not hard to understand why. Competition cases are difficult to conduct not only because they are fact intensive, but also because they involve the application of both law and economics. Neither the Department of Justice nor the SA Police Service have people with any special skills in this area - nor would it have been worth their while securing them, since under the old Act the number of cases requiring prosecutions was too insignificant to warrant the investment. This led to the demise of enforcement and, not surprisingly, when the new Act was proposed, Parliament, in the explanatory memorandum, stated that decriminalizing restrictive practices was a deliberate policy choice to improve enforcement. The administrative penalty became a feature of the new Act. What the Act sought to achieve was to improve enforcement by making a specialist agency and adjudicative tribunal solely responsible.

14. At the same time the legislature intended to decriminalize anticompetitive actions, believing that in modern law administrative

penalties sufficed. [5]

“Infringement of competition legislation will not be subject to criminal sanction, the only exceptions being in respect of breaches of confidence, hindering the administration of the Act and failures to attend when summoned and to answer truthfully to the Commission.”[6]

15. This is the background to the present section 59, which provides an administrative penalty as a possible remedy to the Tribunal in the event of a finding that a respondent firm has contravened certain provisions of the Act. [7]

The Respondent’s Argument in Terms of Section 35 of the Constitution

16. Section 35 (3) of the Constitution states:

Every accused person has the right to a fair trial, which includes the right-

(a) to be informed of the charge with sufficient detail to answer it;

(b) to have adequate time and facilities to prepare a defence;

(c) to a public trial before an ordinary court;

(d) to have their trial begin and conclude without unreasonable delay;

(e) to be present when being tried;

(f) to choose and be represented by, a legal practitioner, and to be informed of this right promptly;

(g) to have a legal practitioner assigned to the accused person by the state and at the state expense, if substantial injustice would otherwise result, and to be informed of this right promptly;

(h) to be presumed innocent, to remain silent, and not to testify during the proceedings;

(i) to adduce and challenge evidence;

(j) not to be compelled to give self-incriminating evidence;

(k) to be tried in a language that the accused person understands or, if that is not practicable, to have the proceedings interpreted in that language;

(l) not to be convicted for an act or omission that was not an offence under either national or international law at the time it was committed or omitted;

(m) not to be tried for an offence in respect of an act or omission for which that person has previously been either acquitted or convicted;

(n) to the benefit of the least severe of the prescribed punishments if the prescribed punishment for the offence has been changed between the time that the offence was committed and the time of sentencing ; and

(o) of appeal to, or review by, a higher court.

17. Measured against section 35(3) of the Constitution, the first respondent submits that the Competition Act fails to comply with that

constitutional standard in three respects:

1) The Tribunal is not a ‘court’;

2) A respondent is not afforded the benefit of the more exacting standard of proof in criminal proceedings, viz. proof beyond a reasonable

doubt. Although section 35(3) is silent on the burden of proof, the first respondent argues that this right has been read into the section by the Constitutional Court interpreting sections 35(3)(h), (i) and (j) [8]; and

3) A respondent is deprived of its right to silence.

18. Since all parties before us are agreed that we are not a court, at least in the conventional sense implied by section 35, and that proof in our proceedings is based on a balance of probabilities[9], the first respondent is correct that, at least in these two respects, the Tribunals procedures fall short of section 35’s strictures. It is less clear whether the Act can be read as to deprive a respondent of a right to remain silent although the first respondent concedes that this was not an issue in the instant case. [10]

19. Nothing turns on determining this latter aspect, because if the first respondent is correct about the first two aspects, its argument must succeed.

20. In order to determine this we have to decide –

1) whether section 35 of the Constitution is intended to apply to Tribunal hearings for which an administrative penalty may be imposed ?

2) if not, why not; and then further;

3) if not, whether our proceedings meet the standard set out in section 34 of the Constitution.

21. The success of the first respondent’s section 35(3) argument is dependent on the notion that the effect of section 59 is punitive in the criminal law sense – hence a contravention of the Act that leads to the order of such a remedy is the analogue of a criminal court imposing its stricture on a miscreant, at least in substance if not form, and that accordingly, if the demands of the Constitution are not to be frustrated the legislation must provide the contravening party no less rights than would their criminal law counterpart.

22. There is no authority presently in our law to support such a view and hence the first respondent’s case is so heavily reliant on European case law for its conclusions.

23. On the other hand, as the amici point out, what South African case law there is, albeit not directly in point, tends to indicate our courts’ reluctance to expand the notion of who is an accused person beyond its conventional notions in criminal law.

24. The leading case is Nel v Le Roux NO and others[11], where the Constitutional Court had to decide whether the provision in the Criminal Procedure Act, which allows a judicial officer to imprison a recalcitrant witness, was unconstitutional because it did not meet the requirements afforded to an accused person in terms of the Interim Constitution. Amongst the grounds on which that section was attacked was the fact that the provision was incompatible with an accused person’s right to be presumed innocent and remain silent (section 25(3)(c)), and the privilege against self-incrimination (section 25(3)(d)).

25. Whilst the Court held that such a person was entitled to procedural fairness, the examinee was not entitled to section 25(3) rights,

“….for the simple reason that such examinee is not an accused facing criminal prosecution. The section 189(1) proceedings are not regarded as criminal proceedings, do not result in the examinee being convicted of any offence and the imprisonment of an examinee is not regarded as a criminal sentence or treated as such.” [12]

26. This is where the first respondent ingeniously endeavours to use a case ostensibly against it to bolster its own argument. The first respondent argues that the lesson from Nel’s case is not that the Court is not prepared to extend the rights of an accused person to a non-criminal trial context. Rather the test resides in understanding the objective of the sanction. Reliance for this is placed on a remark by the Court later in the same passage where it states:

“If, after being imprisoned, an examinee becomes willing to testify this would entitle the examinee to immediate release; in American parlance such examinees ‘carry the keys of their own imprisonment in their own pockets’. The imprisonment provisions in section 189 constitute nothing more than process in aid of the essential objective of compelling witnesses who have a legal duty to testify to do so; it does not constitute a criminal trial nor make an accused of the examinee.”

27. Thus, the first respondent argues, Nel turns on the fact that the object of imprisonment of recalcitrant witnesses is coercive, not punitive. This case it argues, is

therefore authority for the proposition that it is the ‘object’ of the sanction that is determinative and if the object is to punish as opposed to coerce, section 35(3) rights are still of application.

28. This is an interesting gloss on the Nel case but the first respondent cannot get round the clear language of the Court that the rights were not triggered for:

“the simple reason that an examinee is not an accused facing criminal prosecution.” [13]

29. The first respondent also seeks to rely on the Baloyi[14] case to reinforce this gloss. The question there was whether the alleged violator of an interdict in terms of the Domestic Violence

Act (no 116 of 1998), who faced conviction and imprisonment (and a fine), was an accused person as contemplated in terms of section

35(h) of the Constitution. The Court held that such person was. The Court distinguished Nel’s case on the basis that the objective was:

“..not to coerce the will to desist from on-going defiance, but to punish the body for completed violation; and the convicted person carries no keys in his pocket – indeed there is nothing in the Act to suggest that he can be released early if either the complainant wishes or the judicial officer so decides.”[15]

30. However it is quite clear that Baloyi was not concerned with an administrative penalty whose objective is punitive. It was rather concerned with a statutory offence for which conviction and imprisonment were consequences for the transgressor. This is illustrated by the following passage from the decision:

“The language of the Act is clear. Section 6 is headed Offences and Penalties and says that a person who contravenes an interdict “shall be guilty of an offence and liable on conviction” to a fine or imprisonment for a period not exceeding twelve months. Section 3(4) states that the judicial officer shall enquire into the alleged breach and “(b) convict the respondent of the offence contemplated in section 6.”[16]

The European Convention cases

31. Devoid of any local authority that supports its proposition, the first respondent has turned to European case law for assistance. In terms of Article 6 of the European Convention a person charged with a criminal offence is afforded certain minimal rights. [17]

32. It appears that in an evolving jurisprudence over some years, that the majority of the decisions handed down by the European Court of Human Rights, have been at pains to ensure that Article 6 is not too easily evaded, leaving member states with no fairness standard to adhere to in proceedings that may be seriously invasive of citizens’ rights. For this reason they have been at pains to find proceedings, despite their outward administrative law trappings, as criminal in nature and hence subject to the purview of Article 6.

33. As stated for instance in the Engel case, which dealt with disciplinary proceedings against military conscripts in the Netherlands armed forces, member states are free to classify anything as ‘criminal’ subject to the rights that the Convention protects. The converse, that is, classifying something as disciplinary instead of criminal is not subject to the same latitude.

“The converse choice, for its part, is subject to stricter rules. If the Contracting states were able at their discretion to classify an offence as disciplinary instead of criminal, or to prosecute the author of a mixed offence on disciplinary rather than on the criminal plan, the operation of the fundamental clauses of Article 6 and 7 would be subordinated to their sovereign will. A latitude extending thus far might lead to results incompatible with the purpose and object of the Convention. “[18]

34. The Engel case illustrates this distinction. The violation of Article 6 in this case was limited to the fact that the military discipline proceedings had taken place ‘in camera’ and hence did not constitute the public hearing that the Article requires.[19]

35. Our Constitution on the other hand, by virtue of section 34, ensures that fairness is not jettisoned from dispute resolution simply because a procedure is not characterized as criminal. In our law ‘non-criminal’ disputes must still comply with section 34. The right to a public trial, the issue in the Engel case, is, in terms of our Constitution, guaranteed not only in criminal proceedings (section 35(3)(c)) but also by section 34.[20]

36. The court in Engel developed three criteria for determining whether a person facing a proceeding was the subject of a ‘criminal charge’.

(a) Whether the provision defining the offence belongs to the criminal system of the respondent State.

(b) The nature of the offence.

(c) The degree of severity of the penalty that the person concerned risked incurring.[21]

37. Noteworthy, in relation to the latter criteria, the court said the following:

“In a society subscribing to the rule of law, there belongs to the ‘criminal’ sphere deprivations of liberty liable to be imposed as a punishment, except those which by their nature, duration or manner of execution cannot be appreciably detrimental. The seriousness of what is at stake, the traditions of the Contracting states and the importance attached by the Convention to respect for the physical liberty of the person all require that this should be so.”[22]

38. It was argued for the government in a later United Kingdom case that this aspect of Engel, namely, the absence of any threat of imprisonment, is a powerful indicator that proceedings in which only a penalty may be imposed do not give rise to a ‘criminal charge’.[23]

39. However the Strasbourg Court has gone considerably further down the road on these matters and subsequent cases indicate that the court would find a ‘criminal charge’ in cases where penalties are sufficiently burdensome even though they may not entail the loss of liberty.

40. This expanded notion of “criminal” has not been uncontroversial, involving, in certain of the cases, spirited dissents. In dissenting the judges examined the fact that in certain countries a policy of decriminalisation had taken place for sound reasons. Among the reasons proffered are the removal of a moral judgment about the affected person’s behaviour and enhancing the efficiency of the legal system by removing minor offences from the criminal justice system and thereby reducing backlogs. Thus, decriminalisation, far from being a retrogressive policy that requires the redress of the Convention to correct, is a trend in modern societies faced with complex regulatory regimes.

41. A good example of this approach is found in the dissenting opinion of Judge Bernhardt in Ozturk v Germany[24]:

“ Thus the real problem in my opinion is whether the decriminalisation here under consideration is a legitimate exercise of national determination and whether it is in conformity with the object and purpose of Article 6. My answer is in the affirmative. The reasons for removing some minor offences from the field of criminal law, and for providing special sanctions and procedures for them, can hardly be considered unfounded or disguised.”

42. Another of the dissenting judges, Judge Matscher had this to say in the same case:

“I shall merely point out that decriminalisation is something very different from a mere switch of labels. Social changes and new attitudes, as well as technical and economic circumstances,

are leading states to reassess the elements which go to make up criminal offences; thus certain comparatively minor offences, which

nowadays are very common, have been removed from the criminal sphere and classified as regulatory offences. This has important

consequences, which obliges us, in my view, to conclude that the nature of the offence itself has changed. The moral verdict is no longer the same, in other words, a regulatory offence no longer carries the blame which attaches to a crime; the court’s

decision is not entered into the criminal record; nor do regulatory offences carry a more severe penalty in the event of recidivism, this being another feature of criminal law; investigatory measures are also limited - there may, for example, be none of those restrictions on the person’s liberty which apply in criminal proceedings (neither police custody, nor detention on remand, nor the interception of communication may be ordered).The sanctions, too, are fundamentally different. There is no imprisonment.” [25]

43. Notwithstanding this dissenting opinion, the Strasbourg jurisprudence was further developed. In the case of AP, MP and TP v Switzerland [26] the court developed its notion of the nature of the offence in an important way. The first respondent has latched on to this decision and it is fundamental to its argument.

44. The majority of the court said the following:

“ As regards the nature of the offence, it is noted that tax legislation lays down certain requirements to which it attaches penalties in the event of non-compliance. The penalties, which in the present case take the form of fines, are not intended as pecuniary compensation for damages but are essentially punitive and deterrent in nature.[27]

45. Note that the argument of the first respondent in this case has been that the punitive and deterrent character of the administrative penalty under the Competition Act is what imbues it with its ‘criminal ‘ character. Yet in the same case, again a dissenting opinion by Judge Baka, who is joined by Judge Bernhardt, adopts an entirely different stance on this point:

“ I consider that the fine imposed by the authorities on the heirs in the instant case was fiscal in nature and not criminal. Such types of fine are designed to prevent tax evasion. In so doing their main purpose is to protect the financial interests of the state and in a broader sense those of the community. Their undeniably severe punitive character is not just to punish for the tax, which was withheld, but also to deter the offender, through the imposition of a financial penalty, from committing further offences and to deter other taxpayers from possible tax evasion in the future… It is more justified, however, to point to the fact that while the incorrectly declared income was significant and the imposed fine ‘not inconsiderable’ no entry was made in the criminal record of P’s heirs, thus excluding the assumption that the fine was criminal in character.”[28]

46. Thus, unlike the majority who bundle the deterrent and punitive aspects of the penalty, Judge Baka separates them and finds only the latter indicative of the existence of criminal character. He goes further to discern the dominant purpose for the fine and, finding it not to be punitive even though it has some aspects of punishment, concludes that it is not criminal.

47. The critique of the majority approach in Strasbourg has not been confined to some eloquent dissents. In the UK in the Han case the majority of the Court, albeit following Strasbourg, did so with the greatest reluctance as appears from their decision. [29]The court found that although their system of administrative penalties for certain tax contraventions was not consistent with the Convention in terms of the European case law, it was not an unfair system. Unlike the UK court we are not obliged to follow the dominant European approach if we feel that the dissenting arguments are more forceful and in accordance with our own system.

48. The impact of the Strasbourg jurisprudence and the Han case impacted directly on competition jurisprudence in the NAPP [30] case recently heard by the United Kingdom’s Competition Commission Appeal Tribunal (CCAT).

49. The issue arose as to whether the burden of proof should be criminal or the civil standard in a case for which a fine was being sought. The CCAT held that, although it was, in following Han, bound to find that its proceedings were subject to Article 6, and were, in this sense, ‘criminal’, it did not follow that:

“…these proceedings must be subject to the rules that apply to the investigation and trial of offences that are classified as criminal

law offences for the purposes of domestic law.”[31]

50. The decision goes on to state:

“In our view it follows that neither Article 6 nor, the Human Rights Act 1998, in themselves oblige us to apply the criminal standard of proof as established in domestic law in cases where the Director seeks to impose a financial penalty in respect of alleged infringements of the Chapter I or Chapter II prohibitions under the Act.”[32]

51. From the cases that we have considered we would make the following observations:

1) The Strasbourg Court, faced with the Hobson’s choice of, on the one hand, applying Article 6 of the European Convention according to a very broad sweep set of criteria with the unintended consequences potentially flowing from this, or, on the other hand, having no procedural protections at all for many administrative proceedings, has opted in favour of an extended application.

2) These concerns, as the amici point out, are less pressing in our constitutional dispensation because the procedural protections of sections 33 and 34 remain applicable to non-criminal proceedings. Accordingly, the concern in Europe to develop an extensive notion of what is criminal in order to ensure that citizens have some measure of procedural protection is more compelling in Europe than it is in our circumstances.

3) Yet, even within the European jurisprudence there is a history of strong dissents that suggest that the expansive notion of what is criminal is problematic for nation states.

4) Note moreover that Article 6 of the Convention is less demanding than our section 35(3) - section 59 would certainly withstand scrutiny under Article 6. Noteworthy is the fact that in the Napp case, despite the CCAT finding that the Article 6 applied, it did not consider itself bound to adopt the criminal law burden of proof. As that Tribunal observed:

“Neither the ECHR itself nor the European Court of Human Rights has laid down a particular standard of proof that must be applied in proceedings to which Articles 6(2) or(3) apply, and still less that the standard should be that of proof “beyond a reasonable doubt” which is not a concept to be found in the domestic systems of many of the signatory states.”[33]

5) The minority decisions, in particular the Azturk and AP, MP and TP cases, distinguish between the deterrent and retributive purpose of penalties, holding that only the latter gives the penalty a criminal character. This distinction is relevant as it emerges again in other decisions and literature that we consider below.

The double jeopardy cases

52. Both the first respondent and the amici relied on several decisions in North American jurisprudence involving pleas of double jeopardy. In most jurisdictions this plea is typically raised when a person alleges that he or she is being tried twice for the same crime. The court in dealing with this plea has to develop an approach to classification, which helps it decide whether both proceedings are criminal in nature, in which case the plea succeeds or whether there is some distinction that renders the one proceeding non-criminal, and hence the plea fails. The first respondent and the amici commended us to consider the cases on this theme as they believe that the jurisprudence that has developed around this classification, will aid us in our effort in deciding whether section 59 imbues our proceedings with a criminal character.

53. In the case of Canada, the Supreme Court has developed a test to decide when a proceeding may be barred in terms of Section 11(h) of the Canadian Charter, their double jeopardy provision. The test laid out in the case of Wigglesworth[34] states that an offence falls under section 11(h) if:

i. The proceedings are by their very nature criminal proceedings

ii. The punishment invoked involves the imposition of true penal consequences.

54. According to Wilson J writing for the majority:

“.. a true penal consequence which would attract the application of section 11 is imprisonment or a fine which by its magnitude would

appear to be imposed for purpose of redressing the wrong done to society at large rather to the maintenance of internal discipline

within a limited sphere of activity.”[35]

55. Applying the test the court held that a disciplinary action brought against a policeman for assault in terms of the police code, a so-called ‘service offence’, did not bar subsequent criminal proceedings for the same assault because the fine imposed was designed to achieve a particular private purpose namely, discipline in the police force, and not to redress harm done to society as a whole.[36]

56. The Wigglesworth test was applied in a subsequent case, Shubley,[37] where a prisoner had been disciplined for an assault and was placed in solitary confinement and thereafter charged in criminal

proceedings. Again the question was whether subsequent criminal proceedings should be stayed on the grounds of double jeopardy. The majority of the court applying the test held that the prison disciplinary proceeding did not stay the criminal case. The minority,

including Judge Wilson who had written the Wigglesworth decision, also applied the test but came to the opposite conclusion – the punishment they held was a true penal consequence and thus the accused was being punished twice. The question arises, what is a true penal consequence? The majority said that while a person such as a doctor or a policeman may be punished in both a disciplinary and retributive way, they might not be punished twice in a retributive way for the same offence.

57. What emerges from both these decisions is the recognition that punishment can have a retributive element and that it is this characteristic

that makes a proceeding susceptible to being classified as criminal in nature. Whether the test otherwise recommends itself is open to question given its unpredictable outcome in Shubley. But the test seems to be designed to address specifically the problem of double jeopardy in disciplinary hearings that are followed by a criminal prosecution. It does not answer directly the question of whether the sanction itself was unconstitutional if not imposed by a criminal court. At best the cases suggest the government must choose its procedure.

58. In the United States, the Supreme Court has had a more robust approach to the problem. In United States v Ward [38] the court held that:

“Only the clearest proof will suffice to override legislative intent and transform what has been denominated a civil remedy into a criminal penalty “ [39]

59. Ward was cited with approval in Hudson, the most recent Supreme Court decision on double jeopardy to which we were referred, where the court held that even in a case where a civil penalty is indicated, it would have to enquire further as to whether it was so punitive either in purpose or effect.[40] In making this determination the court would apply the factors set out in an earlier case, Kennedy v Mendoza-Martinez,[41] which it said provide useful guidance as to whether a sanction was civil or criminal in nature. These, the Court held, are:

1) whether the sanction involves an affirmative disability or restraint(that is, loss of liberty )

2) whether it has historically been regarded as a punishment

3) whether it comes into play only on a finding of scienter

4) whether its operation will promote the traditional aims of punishment

5) whether the behaviour to which it applies is already a crime

6) whether an alternative purpose to which it may be rationally connected is assignable to it

7) whether it appears excessive in relation to the alternative purpose assigned.

60. In Hudson the court observed that no one factor is controlling as they may point in different directions. [42]

61. The court in Hudson differed from an earlier decision in Halper[43] where the emphasis was on the constituent elements of punishment, namely retribution and deterrence. Commenting that Halper’s deviation from longstanding double jeopardy principles was ill-considered, the Hudson court stated:

“As subsequent cases have demonstrated, Halper’s test for determining whether a particular sanction is ‘punitive’,

and thus subject to the strictures of the Double Jeopardy Clause, has proved unworkable. We have since recognized that all civil penalties have some deterrent effect. (The court then cites various cases)..If a sanction must be totally remedial i.e. entirely non-deterrent to avoid implicating the Double Jeopardy Clause, then no civil penalties are beyond the scope of the clause.” [44]

62. Later the court developed this theme when it stated:

“But the mere presence of this purpose is insufficient to render a sanction criminal, as deterrence “may serve a civil as well as criminal goals…. For example the sanctions at issue here, while intended to deter future wrongdoing, also serve to promote the stability of the banking industry. To hold that the mere presence of a deterrent purpose renders such sanctions ‘criminal’ for double jeopardy purposes would severely undermine the Government’s ability to engage in effective regulation of institutions such as banks.”[45]

63. Although Hudson deals with the Double Jeopardy Clause in relation to civil penalties, an earlier decision of the Supreme Court in Atlas Roofing [46] held that an administrative penalty imposed for deterrent purposes withstood the due process requirements of the seventh amendment.

Atlas Roofing was concerned with the issue of whether an administrative hearing in which a civil fine could be imposed violated the right to a jury trial in terms of the seventh amendment. The court held that:

“When congress creates new statutory rights it may assign their adjudication to an administrative agency with which a jury trial

would be incompatible, without violating the Seventh Amendment’s injunction that jury trial is to be preserved in ‘suits at common law’. Congress is not required by the Seventh Amendment to choke the already crowded federal courts with new types of litigation or prevented from committing some new types of litigation to administrative agencies with competence in the relevant

field.”[47]

64. The first respondent relies on Halper as authority for the proposition that the existence of a punitive element renders a proceeding criminal in nature. Apart from the fact that Halper is no longer authoritative on this point, the problem with this reliance on the double jeopardy cases is that they deal with a

different problem of characterization to the one we have here, as we pointed out with the Canadian cases. They say, in effect, that when the government has available to it a criminal remedy and also a remedy that, although civil in appearance, has penal consequences, it cannot utilize both less it risk the successful invocation of double jeopardy. The cases do not decide that the impugned civil type remedy is unconstitutional because it fails to provide the transgressor with the procedural protection to which a criminal accused is entitled - they only require the government to make a choice of which procedure it wishes to utilize and Halper is clear on this.

65. A court’s answer to the question of when a proceeding exposes a person to double jeopardy may not be the same as its answer to the question of whether the proceeding affords the person adequate procedural protection given the nature of the proceeding. It is entirely probable that a court, anxious not to expose a citizen to overreach by the government’s many tentacles, may approach the first question more expansively than it would the second.

66. Thus, the case law, whilst useful in introducing certain tests and principles, cannot be decisive in respect of the issue we are called on to decide.

67. For this reason we see value in reframing the question that we have to decide:

Reframing the Question

68. Properly reframed the question is whether the remedy that section 59 affords the Tribunal is subject to adequate procedural protection for a respondent, given the nature of the sanction.

69. Despite the fact that in the case law that we have considered the courts have applied different classification tests or, even when applying the same tests, have come to opposing conclusions, one theme remains constant: the extent of procedural protection afforded is a function of the extent to which a proceeding or penalty encroaches upon a person’s rights. A continuum exists between these two values and, classic criminal law proceedings aside, when one moves along the continuum no bright line clearly demarcates when additional procedural protections must be afforded. The answer lies in each case in an examination of the procedure, its remedies and their purpose in each case. [48]

70. The case law shows that courts no longer have a bipolar view of the law – that something is either criminal or civil. With the increasing trend of modern governments to extend administrative law and, at the same time, utilize administrative bodies to enforce it, that division has blurred. The courts recognize that administrative bodies will be enforcing the law through hybrids that are neither wholly criminal nor civil. Administrative bodies not only have the power to enforce the law within their domain but in some areas, competition law being one, are given their own tribunals to enforce it. In our constitution this trend is explicitly recognized in sections 33 and 34 of the Constitution.

71. Allied to this growth in the use of administrative institutions has been an increase in administrative, as opposed to criminal, penalties. As Professor Mann has put it at page 1847:

“The expansion of civil sanctions is part of a more general phenomenon that has characterized law in recent history. There has been a constant search for new ways to achieve social ends through the legal system.”[49]

72. Since the advent of the post-1994 order in South Africa we too have seen the legislature making increased use of administrative penalties as part of the enforcement of legal norms. The Competition Act is unexceptional in this respect. [50]

73. The reasons for this are not hard to discern. When governments reserve new areas for public regulation they opt for enforcement by specialist, expert bodies capable of resolving issues speedily and efficiently. It is clear that, in so doing, they must have regard for a person’s constitutional rights. Precisely how government must guarantee those rights, so that they are not sacrificed on the altar of administrative efficiency, is less clear. The answer, in our view, lies in situating the transgression and its remedy in their proper place on the continuum.

74. The procedural ‘deficits’ in this case, as identified by the first respondent, are the lower burden of proof, the alleged lack of a right to silence and the fact that the tribunal is neither a court nor independent. We deal with the independence issue in the next section. We will deal with the issues of burden of proof and the status of the Tribunal relative to those of a court as part of an examination of where Competition Act transgressions, and the section 59 remedy lie on the continuum and then, in that context, examine whether these deficits are constitutionally fatal.

75. When deciding the classification of a transgression and, thus, the degree and content of procedural protection to which it is entitled, three broad areas of agreement are discernible from the case law referred to above.

76. These three broad areas seem to be:

1) What is the nature of the transgression? In other words is the transgression, by reference to its history and what it seeks to prohibit, typically of a criminal nature?

2) What is the nature of the penalty?

3) Is there a rational connection between the conduct the legislature seeks to prohibit and the sanction it imposes? The greater the disjuncture, the more likely is the sanction to be punitive in nature and, hence, the greater the degree of procedural protection that must be afforded.

77. In particular we ask whether section 59 is a criminal remedy and thus subject to procedural protection provided for in Section 35(3) of the Constitution. Our analysis will show that we have concluded that section 59 does not require the Competition Act to provide the heightened criminal law protection afforded by section 35(3) of the Constitution, and that although the procedure must comply with section 34 of the Constitution, it complies with that standard.

1) Nature of the transgression

78. As we indicated earlier, the legislature enacted the present Competition Act with the express intention of decriminalizing competition

transgressions. It did so because of the enforcement failure associated with its predecessor where an array of contraventions were dependent on criminal law enforcement remedies.

79. Apart from the brief, and inglorious period of Government Notice 801 of 2 May 1986, which converted certain competition transgressions to crimes, there is no history to suggest that these contraventions have long been treated as criminal in our law.

80. In foreign jurisprudence the situation varies, although the modern trend is again away from the use of criminal law as a tool of enforcement against restrictive practices. The exception in some jurisdictions has been enforcement against cartels, where in the United States and Canada, and recently the United Kingdom, criminal law is used. [51] The reason for this is the concern that even huge fines were inadequate remedies against cartels and that only the prospect of

imprisonment of senior executives would be an adequate remedy against corporate miscreants. This remedy thus places this species of transgression at a different point on the continuum and for this reason adherence to criminal due process standards is appropriate in those jurisdictions seeking imprisonment as a remedy.

81. Most competition law transgressions, however, fit uncomfortably into the cloth of the criminal law. Firstly the state of mind of the transgressor, whether we term this element ‘intention’ or ‘scienter’, is not typically an element of a prohibited practice.[52] What the Act refers to as prohibited practices are typically acts by firms with market power who act in a manner that utilizes that power to exploit consumers or exclude competitors. What the Act is concerned about is the ‘effect’ of a firm’s conduct in the market place, not what it intended by that conduct. Hence the leitmotif of the word “effect” to be found in each of the core prohibitions.[53]

82. The Competition Appeal Court has recently observed in relation to dominance that:

“The concept of abuse is an objective one…”[54]

83. In doing so they followed the approach of the European Court of Justice in Hoffman-La Roche and Co AG v Commission of the European Communities[55] where that court had noted:

“The concept of abuse is an objective concept relating to the behaviour of an undertaking in a dominant position which is such as to influence the structure of the market where, as a result of the very presence of the undertaking in question, the degree of competition is weakened and which, through methods different from those which condition normal competition in products or services on the basis of the transactions of commercial operators, has the effect of hindering the maintenance of the degree of competition still existing in the market or the growth of that competition.” [56]

84. The first respondent has tried to argue that because the penalty in section 59 may only be imposed in limited cases, those which are so called ‘per-se’ violations and those where there has been a repeat of the same conduct, an element of knowledge of wrongdoing, of intent, is imputed as a threshold condition for the imposition of the penalty. On this basis, it is argued that key elements that are part of a criminal finding are present and, so, the protections available to a criminal defendant are required if penalties for these contraventions are to be imposed.

85. However, this argument confuses two notions, on the one hand, transgressions for which proof of a mental element is a requirement and on the other hand, transgressions, which are so notorious as being anticompetitive, that firms ought to know that they are wrong. In the case of the former, failure to prove mens rea must result in an acquittal. However, in the latter, proof of intent is not essential - it is rather the nature of the transgression itself and not the subjectivity of the respondent’s mind that makes the penalty competent.

86. The Act makes only a limited number of transgressions susceptible to a administrative penalty because it limits penalties to acts that are known to be anticompetitive precisely in order to reduce the incentive on the part of potential transgressors to engage in them. If a penalty was applied to all forms of anticompetitive behaviour, including those at the cusp of legality, lawful or even pro-competitive behaviour may be disincentivised. For this reason the penalty regime is limited to those deemed notoriously

anticompetitive, the so-called per se transgressions, and to the repeat by a firm of conduct previously found anticompetitive. This approach is therefore premised on an incentive-based model not a criminal one.

87. Absence then of proof of a mental element as a component of a transgression is an indication of the fact that it is not criminal in nature. [57]

88. As Areeda points out:

“The most fundamental reason for limiting intent inquiries is that the relevant intent – intent to harm one’s rival – is impossible to distinguish from the intent to behave competitively in a broad range of situations ….In such a setting an ‘intent’ not to harm a rival is tantamount to an intent not to compete.”[58]

89. In addition competition transgressions contain no moral or normatively condemnatory aspect. That is because the conduct itself is typically only objectionable when performed by firms with market power. It is hard to apply some normative standard to conduct whose legality is dependent on the protagonist’s share of the relevant market. Thus when it comes to an abuse case, the same act perpetrated by company A, which has market power, may be a prohibited practice, but is not in respect of firm B if it lacks market power. In addition prohibited practices are usually a matter of degree. Apart from the per se contraventions, most conduct that transgresses is a matter not of the inherent nature but of degree. Thus ‘recommending’ a minimum resale price is lawful, ‘maintaining’ it is not.

90. When we deal with the nature of the penalty we will take this theme of morally neutral contraventions further.

91. With criminal cartel prosecutions as the one exception, most jurisdictions enforce competition law violations with a lower burden of proof than they would in criminal matters. This is not surprising given the nature of the contraventions, which, as we have already stated, typically depend on an element of market power. A finding of market power requires an analysis of the relevant market, which in turn requires an analysis of whether certain products are substitutes for one another and therefore part of the relevant market. To require that, in a dominance case, the Commission must, for example, prove beyond a reasonable doubt, that a cola drink did not compete with other carbonated drinks, would effectively render the section unenforceable. That kind of certainty cannot be attained in such an evaluation, which must then, by its very nature be established on the balance of probabilities rather than on the elimination of doubt.

92. In the United Kingdom, as we noted earlier in the Napp case, the CCAT, held that the burden of proof in Competition cases involving penalties is the civil standard of proof although the CCAT cautioned that it needed to be strong and compelling evidence. This conclusion was reached notwithstanding the fact that Article 6 of the European Convention applied to the proceedings and that they were for that purpose to be regarded as ‘criminal ‘.

93. In Canada dominance cases are assessed by the civil standard, and whilst it is conceded that they do not impose penalties for these, they do impose far ranging remedies including divestiture - remedies firms may find far less palatable than penalties.[59] In the United States dominance is an infringement of section 2 of the Sherman Act, a criminal statute, but the Department of Justice can proceed civilly for a breach and it frequently does so in monopolization cases, even though its remedies here are equitable

rather than civil. Nevertheless in the United States private litigation plays a major role in ensuring compliance with the antitrust laws and for this reason plaintiffs are rewarded with treble damages. As Areeda states:

“Such a remedy not only compensates private persons for their injuries, but gives them a powerful financial incentive to enforce the antitrust laws.”[60]

94. Later he goes on to observe that although the private plaintiff need not act in pursuit of the public interest,

“Nonetheless the most obvious feature of the treble damage action is punishment “[61]

95. Treble damages are awarded on the basis of the civil burden of proof.

96. In South Africa the legislature’s choice of a civil as opposed to a criminal remedy arose out of a concern that competition violations were insufficiently policed under the old Act. This is specifically mentioned in the preamble of the Act:

“That apartheid and other discriminatory laws and practices of the past resulted in …… inadequate restraints against

anti-competitive trade practices… .”

97. The literature on economic regulation suggests that lawmakers typically prefer to make transgressions of the law subject to a less demanding civil standard, as they fear that there may be under-enforcement if the activity was criminalized. The history of the failure of our predecessor the Competition Board in this respect is a matter of public record and tends to validate the theory in the literature. There is little doubt that our legislature would not contemplate reverting to the criminal law to act against anti-competitive conduct given our past experience.

98. Not only may the use of a criminal standard in this respect lead to an under-deterrence problem from the perspective of the enforcer, it may also from the perspective of the target lead to an over-enforcement problem. Our legislature, whilst regarding competition law infringements as very serious, does not believe that they should be visited with the wrath and consequences of the criminal law. Furthermore they believe that these transgressions are most efficiently policed and enforced by specialist agencies not by our already crowded courts.

99. We conclude that neither the nature of the prohibitions, their history, the legislative intent or policy informing their treatment, suggests that prohibited practices are, in substance, akin to crimes.

2) Nature of the penalty

100. It could of, course, be argued that, even if competition law transgressions are best left as civil law violations, it does not follow that they should then be visited by ‘punitive’ remedies such as an administrative penalty, when other remedies are available. It is true that the Act does conceive of a range of remedies, other than the administrative penalty, for a violation, but not every case commends itself to such a remedy, nor may they prove to be a sufficient deterrent. It would significantly undermine the effort to curb anti-competitive practices if all that a miscreant firm was faced with after being found in violation was a cease and desist order.

101. Hence for a species of violations the Act makes the imposition of a penalty competent. Since, as discussed earlier, many antitrust contraventions may not be clear to firms, the Act limits the imposition of a penalty to those which are either sufficiently well known to be anti-competitive or where a firm is a repeat offender, that is, to circumstances where the transgressor ought to have known that the particular conduct was unlawful.

102. The case before us is an example of why a penalty is the only proper remedy. If the remedy were limited to an interdict, it would pay firms to transgress the section until detected and then to agree to abide by a cease and desist order. This is clearly an unsatisfactory

situation.

103. Furthermore, penalties may loom larger in the choice of remedies sought by the Commission than may be the case in respect of their counterparts in the better-resourced developed countries. Complicated behavioral remedies, which would entail ongoing monitoring and enforcement over a period of time, may be beyond the resources of the Commission. A penalty as a deterrent raises no such problems.

104. It is as well to examine now an issue, which we avoided in the previous discussion on the case law, namely the question of whether punishment has a single or multifaceted objective. We are of the opinion that much of the case law on which the respondent sought to rely failed to make this distinction, which is absolutely crucial in ascertaining the appropriate procedural safeguards a remedy attracts.

105. Writers on the subject have repeatedly distinguished between various goals for the imposition of punishment.

106. Yeung argues that traditional criminal law offers five theories for punishment, namely retribution, deterrence, incapacitation,

rehabilitation, restitution, and compensation.[62] She states that of these, only retribution and deterrence are relevant to regulatory penalties and compete as a possible theoretical basis for them. While Yeung argues that a penalty regime should include both elements and suggests that ultimately the two are impossible to untangle, other writers such as Rubin[63] suggest, after considering the same theories of punishment that:

“Of these purposes, deterrence would appear to be the dominant one in the case of administrative remedies. Compensation and inducement to take legal action are essentially irrelevant, while punishment, a puissant but ambiguous idea, turns out to be little more than deterrence when it is factored into its part.”[64]

107. He develops this ambiguity theme by looking at the way the notion has been used in the literature and he then comes to the conclusion that:

“Punishment as a reason for imposing administrative fines, is a complex concept that must be parsed to be assessed with any clarity. ..For present purposes, it is useful to distinguish among different functions of punishment, the classic ones being deterrence and retribution.

108. He goes on to argue that it is difficult to conceive that legislatures intend that administrative agencies’ civil penalties are intended to be retributive because society does not construe these institutions sufficiently highly to bestow them with the regard to serve this purpose. As he puts it in a memorable phrase:

“It seems unlikely that most people believe that so biblical a purpose as retribution should be assigned to these technocratic, unpoetic

institutions.”[65]

109. He concludes that:

“ We are left with one dominant purpose for administrative penalties namely deterrence. This term is itself derived from criminal law however; a more administrative one is law enforcement, or compliance. The obvious purpose of imposing penalty on a private actor is to induce that actor to obey a legal rule. The message is not the moral judgment ‘you are a bad person’ but rather the pragmatic instruction ‘cut it out’. This distinction can be derived from the nature of the sanctions, the nature of the underlying violation, and from the conception of administrative law.”[66]

110. As Feinberg has noted, penal sanctions for regulatory offences do not constitute punishment, even though the imposition of such sanctions visits ‘hard treatment’ on the transgressor, because they do not convey “reprobative symbolism” which he regards as the essence of punishment.[67]

111. Mann notes the growth in our legal systems of what he terms ‘deterrence ideology’, which, he argues, has led to a decrease in the difference between the purpose of criminal and civil law:

“Deterrence ideology with its philosophical background of law and economics, became a significant causal factor in the growth of punitive civil sanctions.”[68]

112. Case law also distinguishes between deterrent and retributive effects of punishment. In one jurisdiction, Australia, which is the system closest to ours (indeed our section 59(3) appears to have been modeled on their section 76 of their Trade Practices Act), the courts have repeatedly emphasized this point.[69]

113. In NW Frozen Foods v ACCC [70] the Federal Court of Appeals held that the purpose of the administrative fine is not punishment but deterrence. The court quotes with seeming approval from French J in TPC v CSR Ltd:[71]

“ The principal and probably the only object of the penalties imposed by section 76 is to attempt to put a price on contravention that is sufficiently high to deter repetition by the contravenor and by others who might be tempted to contravene the Act.

114. They then quote Lee J (293) who said that:

“ the object to be served by s 76, namely, to promote competitive conduct in trade or commerce by use of penalties sufficient to deter acts that would tend to be destructive of such competition”[72]

115. The mantra of the duality of deterrence is repeated in ACCC v Rural Press and Others[73] where the court stated:

“ I am also mindful in assessing the level of the pecuniary penalty, that it is directed to deterrence of acts that would tend to be destructive of competition. That deterrence has two facets. They are the deterrence of the individual contravener from engaging in conduct in contravention of the Act in the future, and the deterrence of other traders from engaging in conduct in contravention of the Act.”[74]

116. Whilst some Australian judges have expressed reservations about whether deterrence is the sole criteria, this does appear to be the dominant view.[75]

117. We now examine the section itself. In the first place we note that section 59(2) places a ceiling or cap on the maximum penalty that can be imposed. This is set at 10% of the offending firm’s turnover during the firm’s preceding financial year. The fact that it is set out in this way and not, as in section 74, which provides a set maximum penalty for criminal offences, suggests that the legislature intended the penalty to be relative to the economic impact of the respondent in the market, that is, the greater the turnover of the firm, the greater the potential for it to be penalized.[76] This relative, as opposed to absolute cap, is more consistent with a deterrent than a punitive model, because given that a deterrent model is incentive-based it requires greater administrative penalties to be imposed on larger players.

118. None of the factors set out in section 59(3) is inconsistent with a deterrence model, although, we acknowledge that it is not pure in this respect. All the factors listed either deal with the information the Tribunal would need to address in setting a administrative

penalty or in order to achieve the two aims of deterrence namely, deterrence of the respondent and deterrence of others contemplating

the same conduct.

119. Even a factor such as the degree of co-operation with the Commission and Tribunal, although superficially more associated with the mitigation notions of criminal law, has an incentive logic to it, and thus is consistent with the goal of deterrence as it incentivises firms to settle with the authorities.

120. Notably absent from the criteria are any notions of moral judgment or censure characteristic of criminal law. As Coffee[77] points out one of the features of the criminal law is its deliberate intent to inflict punishment in a manner that maximizes stigma and censure.

121. Nor are there any consequences of the penalty apart from the possibility that it may be taken into account if the firm was again subject to a penalty. But this again is perfectly consistent with a deterrence model, that is, if someone has been awarded an administrative

penalty, and again contravenes the Act, the penalty may have to be increased the next time around as compliance had been insufficiently

incentivised.

122. Nor do penalties impose upon a firm any disability as a criminal conviction sometimes does. [78] Further the legislature’s choice of the word ‘penalty’ as opposed to ‘fine’, the latter a term usually

associated with the language of criminal law, is again indicative of its civil aspect.[79]

123. We do not suggest that section 59 exhibits a ‘pure’ deterrent model devoid of any retributive element. We do however

suggest that it is by far the dominant element, and that retribution is insufficiently a determinant in the final mix to alter the categorization of the penalty into a punitive one.

124. For this reason the sanction imposed by section 59(3) is insufficiently retributive in character to render it punitive in nature in a manner that requires the heightened protections afforded by section 35(3).

3) Rational connection

125. We discuss this aspect in relation to the broader rationality challenge, which we deal with below, so it is unnecessary to repeat these issues here.

Section 34 of the Constitution (the fairness standard)

126. It is noteworthy that the first respondent’s attack in terms of section 34 of the Constitution has been confined to the issue of independence. It has not suggested that, were section 34 to be the governing section, the Act fails to provide the necessary degree of procedural fairness required by the Constitution. We submit that the Act does so and, apart from the deficits explicitly mentioned earlier, would even comply with section 35(3). One of the most notable features of the new Act which has not been previously

mentioned is the deliberate separation of what was once a single competition agency with combined investigative and decision making

functions into two separate bodies. In prohibited practice cases such as the one in casu the Commission acts as the prosecutor but cannot adjudicate, and the Tribunal adjudicates but cannot prosecute. Both bodies are institutionally separate as we elaborate on below. Secondly, the Act has an elaborate set of procedures, from the requirements for a complaint referral, to the nature of the way hearings are conducted, designed to ensure compliance with the Constitution’s mandate of fairness.

127. A prohibited practice hearing, even where a section 59 penalty is competent, finds itself located on the continuum at a place where the level of procedural protection afforded is consonant with the severity of the proceeding.

Section 34 of the Constitution (the independence issue)

128. The first respondent’s second attack is based on section 34 of the Constitution. This section states that:

Everyone has the right to have any dispute that can be resolved by the application of law decided in a fair public hearing before a court or, where appropriate, another independent and impartial tribunal or forum.

129. As already noted, the first respondent does not appear to contend that it did not receive a fair hearing in terms of our proceedings.[80] The Act and the Rules clearly make ample provision for a fair hearing and were the Tribunal not to follow such a proceeding it would clearly expose itself to review.

130. What the first respondent relies on is that the Tribunal is, in its opinion, not an independent and impartial tribunal or forum. As we understand it the nub of the issue is not impartiality but independence. The first respondent asserts that the Tribunal is not independent as it forms part of the executive and that Tribunal members should be viewed in the same way as public servants.

131. A reading of the Competition Act shows that this contention is not well-founded. The legislature has taken several steps to ensure the independence of the competition institutions not only from the executive but also from one another.

132. In terms of section 20(1) of the Act the Competition Commission:

“(a) is independent and subject only to the Constitution and the law;

(b) must be impartial and must perform its functions without fear, favour or prejudice.

133. Section 20(3) goes on to state that:

“Each organ of state must assist the Commission to maintain its independence and impartiality, and to effectively carry out its powers and duties.”

134. These provisions apply equally to the Tribunal in terms of section 26(4).

135. Apart from this unambiguous statement in the legislation, which mirrors the language of section 34 of the Constitution, by its specific reference to the institution’s independence and impartiality, the Tribunal is functionally independent of the executive in the sense that none of its decisions require consultation with or the ratification of the executive.[81]

136. In the case of the Tribunal additional measures are taken to ensure that members have relative security of tenure. Tribunal members are appointed by the President on the recommendation of the Minister of Trade and Industry for a period of five years.

137. During the period of their appointment they may only be removed from office if they become disqualified or in the limited circumstances set out in terms of section 29(5)(b) namely –

“(i) serious misconduct;

(ii) permanent incapacity; or

(iii) engaging in any activity that may undermine the integrity of the Tribunal.

138. Only the President on the recommendation of the Minister can remove a Tribunal member from office.

139. In the Van Rooyen[82] case the Court observed that:

“There is a difference between being nominated by the executive to perform a duty which calls for an independent decision and being chosen by the executive to perform a duty in accordance with its wishes. If the power to recall is subject to objective criteria consistent with the Constitution, that power is not constitutionally objectionable “[83]

140. In our view, the appointment and removal of the Tribunal members, in terms of the Act, conforms with this observation by the Court. Thus although members are subject to appointment by the executive it is clear from the Act that members are appointed to perform an independent function. Secondly, as far as removal is concerned the power to recall is subject to objective criteria.

141. The fact that members are appointed by the executive has been held not to contravene the separation of powers in the Van Rooyen case:

“The mere fact, however, that the executive and the legislature make or participate in the appointment of judges is not inconsistent with the separation of powers or the judicial independence that the Constitution requires.”[84]

142. In the Freedom of Expression case Hlophe ADJP (as he was then) observed:

“It is well established that there are degrees of independence. Indeed it is not every tribunal that can be as completely independent as a court of law is expected to be. The independence of courts of law and administrative bodies cannot be measured by the same standard.”[85]

143. As far as financial independence is concerned, section 34(2) states that:

“The Minister may not during the term of office of a member of the Competition Tribunal reduce the members salary, allowances or

benefits.

144. The financial provisions that relate to the Tribunal and Commission that are set out in section 40 and 41 of the Act indicate that the institutions are:

(a) financed by Parliamentary appropriation:

(b) have their own accounting officer

(c) Report to Parliament annually on their activities and finances.

145. These provisions indicate that the institutions report to Parliament and not to the executive albeit that the Minister is their conduit to Parliament.

146. In our view these features of the Act, coupled with the manner in which the Tribunal functions, indicate a relationship of independence from the executive that meets the requirements of section 34.

147. The constitutional attack in respect of this aspect must accordingly also fail.

Rationality argument

148. The final attack on section 59 is that it is irrational because it is turnover based. The argument is that it is well known that whilst a firm may have a large turnover its level of derived profit from these activities may be low. Hence a firm with a large turnover but low profits is susceptible to a higher fine than one with larger profits but a low turnover. The irrationality, the first respondent alleges, is that the firm, which has profited less from its ill-gotten gains faces a potentially more formidable fine. There is thus an irrational disjuncture between the firm’s sin and the sanction.

149. Whilst the economic observation that there is no consistent correlation between levels of profit and turnover is correct, this does not the make the use of turnover irrational.

150. The Act, as we stated earlier, is centrally concerned with regulating the abuse of market power. One of the consequences of market power is the ability it gives a firm to price above the competitive price. That so-called monopoly price is reflected in an increase in turnover for the firm - a turnover greater than would have been in a competitive market. From the perspective of the consumer the harm of the anti-competitive pricing has been the increased price paid to the firm and hence its increased turnover. The fact that the firm has not internalized that market power efficiently into profit in each case is a secondary issue. Indeed this is why it is located as a factor in section 59(3), not 59(2) i.e. we assess the loss to consumers in setting the cap in section 59(2), we assess the gain to the firm under section 59(3). The approach is, on the contrary, perfectly rational.[86]

151. Although the first respondent had raised the irrationality argument in its initial heads of argument, following the amici‘s heads, which dealt comprehensively with this issue, the argument was neither pursued in its second set of heads nor in the oral submissions, although we were told that it had not been abandoned.

152. In our view the case law the amici have referred us to strongly supports their major contention that all rationality review requires is that legislation has a purpose that is not unconstitutional and that it is rational to believe that the legislation will advance this purpose. The existence of a less invasive means to achieve a government purpose is at constitutional law a question of proportionality and not a question of rationality.[87]

153. We find nevertheless that the 10 % turnover threshold is a rational one in the Act’s schema but that even if we are wrong in this respect that given the high threshold for rationality review set by our courts, the attack on section 59 by way of a rationality review is unpersuasive.

Conclusion

154. We find that a section 59 penalty does not constitute the type of punishment which necessitates the constitutional protection of

section 35(3) given the restricted scope our courts have given to the application of that section. We find further that prohibited practice cases in the Tribunal are subject to sections 33 and 34 of the Constitution. Measured against the procedural requirements of those sections, the Act contains adequate and appropriate procedural protections to respondents to ensure fairness, given the nature of prohibited practices and the penalties that may be imposed. We also find that the Tribunal, although not ‘an ordinary court’ for the purpose of section 35(3), is an independent and impartial tribunal for the purpose of section 34.

155. We further find that the penalty contemplated in section 59 is not irrational.

156. Accordingly all three of the first respondent’s constitutional points fail.

Per: N. Manoim

Concurring: D. Lewis and M.T.K. Moerane

SECTION B – THE ADMINISTRATIVE PENALTY

157. The Commission has asked that we exercise the powers granted us under Section 59(1)(a) of the Act to impose an administrative penalty on the first respondents who have been found to practice resale price maintenance, a contravention of Section 5(2). The Commission

had initially also asked for interdictory relief but this was withdrawn.

158. Contraventions of Section 5(2) constitute a species of price fixing, the vertical variant of that set of contraventions generally construed as the most anti-competitive of business practices. It is for this reason that resale price maintenance is prohibited per se, why, in other words, it permits of no pro-competitive defence. And it is for that reason that the Act provides for the imposition of administrative penalties on first-time Section 5(2) offenders, a level of deterrence reserved for a mere handful of other transgressions.

159. We interpret the legislature’s intention in designating a particularly powerful sanction for this class of contraventions as, firstly, a manifestation of the serious nature of the contravention, of its unequivocally deleterious impact on competition.

Secondly, we understand the particular treatment of these contraventions to manifest the degree of ‘nakedness’, as

it were, of the restraint in question, of the widespread appreciation and understanding in the business and general community that Section 5(2) contraventions offend the fundamental and most widely understood tenets of competition law and economics. As such, they are to be distinguished from a range of other potential contraventions which must be subjected to a ‘rule of reason’ analysis, that is, to a complex legal and economic analysis, that may or may not establish a contravention.

160. We have no doubt that it is appropriate that an administrative penalty be imposed in this instance. As noted, it is not for nothing that the Act specifically empowers us to award penalties to first-time contraventions of Section 5(2) – it is a serious

contravention and one proscribed by most antitrust statutes.

161. That having been said, we are, in determining the size of the administrative penalty, in relatively uncharted territory. There is,

ineluctably, a significant element of discretion at work in arriving at a decision regarding the size of the penalty. [88] However discretion is never unfettered – it must be exercised justly and reasonably, and the basis for the decision must be clearly set out. There are, moreover, as we indicate earlier, two provisions in the Act that actually lay down the bare fundamentals of a framework for determining the size of the administrative penalty. These are Sections 59(2) and 59(3). For ease of reference we reproduce them again:

162. Section 59(2) provides:

An administrative penalty imposed in terms of sub-section (1) may not exceed 10% of the firm’s annual turnover in the Republic and its exports from the Republic during the firm’s preceding year.

163. Section 59(3) provides:

When determining an appropriate penalty, the Competition Tribunal must consider the following factors:

(a) the nature, duration, gravity and extent of the contravention;

(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.

164. It has been argued by the first respondent with, it appears, some support from the Commission, that the turnover figure referred to in Section 59(2) should not influence the size of the actual penalty.[89] In other words, it is argued that the turnover threshold is an upper limit, pure and simple, and should only come into play if the amount arrived at after taking into consideration the factors listed in Section 59(3) exceeds the specified threshold. Mr. Brassey, for the first respondent, has cautioned us against adopting a formulaic approach to the determination of penalties when the Act itself does not provide such a formula. The size of the penalty, on this contention, should be determined after consideration of the Section 59(3) factors alone.

165. However, in our view, this is an unrealistic proposition. It would be extremely difficult to specify the appropriate level of an administrative penalty without recourse to any reference point. It would leave the determination of the level of the penalty as an exercise of ‘pure’ discretion and would, as such, provide no certainty. Indeed this appears to have been explicitly recognized by the Commission which favours following a central aspect of the US approach by designating the ‘volume of affected commerce’ as the appropriate base level. There would appear to be no meaningful difference in principle in using the reference point provided for in the Act – the 10% of turnover – rather than utilizing the US reference point. Under the circumstances then it is surely preferable to draw as much guidance from our own Act as is possible.[90]

166. Given that we have identified deterrence as the primary purpose of the imposition of administrative penalties, and that the deterrence

element must have some relationship to the harm inflicted by the prohibited practice, it is obvious that we must have regard to the maximum level of the penalty as our point of departure. If this were not the case it would make it worth it for many firms to run the risk of incurring a penalty and to treat it, as some courts have referred to it, as a license to do business.[91]

167. Our view is that the upper threshold of 10% of turnover should be reserved for the most egregious contraventions and an absence of mitigating factors. By way of example, a hard-core cartel in a significant area of commerce and in the investigation of which the parties have refused to co-operate with the authorities may well attract the maximum penalty. Using this approach we will examine each of the considerations that the Act requires us to factor into our decision, although not all of the considerations listed in the Act will bear with equal weight upon every decision regarding the scale of an administrative penalty. We then proceed to examine those considerations that do bear on our determination of the level of the administrative penalty to be imposed in this matter.

168. Before considering each of the factors listed in Section 59(3) we must determine the year referred to in Section 59(2). The first respondent correctly observes that the Act does not define the date by reference to which the ‘preceding financial year’ is to be calculated.

169. The first respondent contends for the year preceding the contravention. While there are alternatives – the year preceding the handing down of this decision is one possible alternative – we need not decide this point, as we will accept the first respondent’s assurance which was not disputed by the Commission that we can accept these figures as correct whichever year is contemplated.[92] Had this not been common cause we would have had to decide the matter, but we need not. [93]

170. The first respondent also insists that the turnover figure should refer only to sales of the product in the relevant market – that is, Ferodo brake pads. The statute clearly specifies ‘the firm’s annual turnover in the Republic and its exports from the Republic’ as the relevant reference point. In Europe the European Court of Justice has confirmed that the basis of their penalty in Article 15(2) is the entire turnover of the group on all products worldwide, but observed that the Commission habitually reduced the penalty where the infringement concerned only one line of business among others. [94]

171. In our view that is the proper approach. There is no warrant to interpret the legislation more restrictively than its clear language

suggests, however we should exercise a discretion to set the maximum threshold at a lower level based on the turnover of the line of business where the infringement occurred in the appropriate circumstances. That discretion would be exercised by examining the

relationship of the affected turnover relative to the total turnover, and then determining where the penalty cap should be set, given the object of deterrence.

172. For the purpose of this case we will exercise that discretion in favour of the first respondent, and base the maximum threshold on the turnover in the infringing line of business only. This amount is [confidential] which is 10 % of the turnover for the Ferodo

products.[95] (The alternative would have been to take the first respondent’s entire turnover in the Republic, a figure of [confidential] – yielding a maximum permissible penalty of [confidential].

The nature, duration, gravity and extent of the contravention

173. We have already indicated our views on the nature and gravity of the contravention. We have underlined that resale price maintenance is a species of price fixing, the most egregious category

of anti-competitive practice. As elaborated above, the Act explicitly indicates the serious view that the legislature takes of the practice of resale price maintenance.[96] However, that being said, resale price maintenance is generally not considered in the same light as the notorious horizontal varieties, hard-core price fixing or market sharing cartels or bid rigging. Indeed, anti-trust enforcement has, in recent years, tended to adopt a relatively benign attitude towards vertical agreements in acknowledgement of economic theory and research that has demonstrated the pro-competitive core that often lies at the heart of what, at first blush, may appear to be anti-competitive arrangements.

174. However, minimum resale price maintenance – the fixing by manufacturers or wholesalers of a floor price in respect of the onward sale of their products – is still widely construed to be an unequivocally anti-competitive variety of vertical arrangement and this, as already noted, is reflected in the particular status that this contravention is accorded in the Act. Indeed the first respondent has persistently raised a defence that, ironically, serves to illustrate precisely why resale price maintenance retains its pariah status even in the midst of a general reconsideration of the impact of vertical arrangements. We refer to the first respondent’s insistence that its objective in setting minimum resale prices was to prevent the ‘commoditisation’ or, as it sometimes terms it, the ‘prostitution’, of its product.[97]

175. The first respondent insists – and we have no reason to dispute this claim – that considerable resources have been devoted to establishing the Ferodo brand and that the maintenance of the brand’s distinctive reputation would be undermined were its distributors to compete for their market share by engaging in price competition. In other words, a certain cachet, a signal of quality, attaches to the brand and consumers purchase it for this reason. The superior reputation of the brand may not only support sales volumes of Ferodo products but it may allow a premium on the price of the product and hence the producer’s margins relative to that of other braking products. A drop in price would distort the signal of superior quality that the manufacturers wish to transmit to the market. The value of the brand, were it to enter price competition, would be reduced to the level of products that do not enjoy its reputation and it would henceforth be condemned to compete on price alone thus impacting negatively on the margins to be earned from the product in question.

176. We understand the first respondent’s concern. It is perfectly legitimate to utilize brand reputation as the core of a strategy

aimed at maintaining the price of the product and the margins earned. Hence many producers strive to maintain the reputation

of their brands, the better to maintain a price level and/or sales volume higher than that of functionally competitive products. Indeed vertical agreements, which may on the face of it appear to be restrictive, are often tolerated precisely because they add value to a particular brand and thus strengthen inter-brand competition. These may include exclusive distribution arrangements or the specification of certain performance standards and marketing support in the distribution and sale of the brand.

177. In other words the brand’s reputation must be supported by pro-competitive means, through, for example, product innovation or by a marketing strategy that adds value to the brand in question. It would, generally, be perfectly acceptable to insist that wholesalers adhere to certain standards in the handling of the product and in after-sales service and support. Strategies of this type, strategies that seek to distinguish the product or brand in question from competing products or brands, legitimately permit the maintenance of the price of the product and the margins that attach to it. They, in other words, legitimately introduce an element of segmentation into the market for, in this instance, braking products, with the prospect that the brand enjoys something akin to a temporary monopoly in the segment that its marketing strategy or product development has created for itself.

178. However, the objective of maintaining the reputation of a brand cannot be legitimately pursued through a system of artificial price supports effected through the agency of an agreement between, as in this instance, the producer and the wholesaler. Indeed a large reason for anti-trust’s aversion to price supports of this nature is precisely that it constitutes an alternative to other pro-competitive strategies aimed at distinguishing the product or brand from its competition. However, resale price maintenance is the path chosen by the first respondents and it is for this reason that the practice has been found to be in contravention of Section 5(2) of the Act. This, as we have repeated ad nauseam, amounts to nothing more or less than price fixing and, as such, is a contravention of considerably gravity. That, as the first respondent insists, it is intended to maintain the value of brand, does not excuse the practice – on the contrary it clearly reveals its anti-competitive core.

179. So much for the nature and gravity of the contravention – it is a specie of price fixing and, as clearly indicated by its treatment in the Act, a particularly grave contravention. What of its duration?

180. The first respondent – and, it appears, the Commission – have taken the view that the contravention has endured for a short time only.[98] The first respondent points out that the period for which the complainant’s discount rate was cut lasted for a matter of weeks. The first respondent points out further that we have acknowledged in the reasons for our decision that there was no intention that the reduction in the complainant’s discount should be of a lengthy duration, that it was instituted merely to bring the complainant back into line and that once this objective had been achieved by the complainant raising its price – or, what is the same thing, reducing the discount - to its customers that it, the first respondent, would be willing to restore the discount to the complainant. However, this argument misconstrues the nature of the contravention in question.

181. The first respondent’s contravention does not reside in the reduction of the discount available to the complainant. The cutting

of the discount was merely the sanction utilized to compel the complainant to restore the fixed price. The contravention resided

rather in the fixing of the resale price.

182. Unfortunately evidence has not been led that explicitly establishes the length of time for which the price fix had been in existence, nor, indeed, whether in the wake of the imposition of the sanction, the fixed price has simply been restored and continues to prevail. Indeed had persuasive evidence to this effect been led, the Commission may well have been able to sustain its initial request for interdictory relief.

183. We do know for certain that from the time that the complainant entered the market it had expected that, given the volumes that it expected to achieve, and that it indeed very quickly succeeded in achieving, it would receive a 47,5% discount off the first respondent’s

list price and that it would on-sell the product at a discount of 35% off the manufacturers list price. So strong was this

convention, that the complainant, and, it has been established, a number of the other large traders in Ferodo products, immediately

reacted to Midas’ decision to embark on a price war, to violate, in other words, the established pricing convention.

184. Indeed, the strength of the convention, or, what amounts to the same thing, the certainty of retribution, for a time successfully constrained Midas’ competitors from reducing their own prices in response to Midas’ initiation of the price war. In other words, they allowed Midas to take custom away from them because they firmly, and correctly, believed that Midas was breaking the well-established rules of the game and so their initial response was in the form of a plea to Federal Mogul reinstate the rules. While this strongly suggests that the price fix had been in operation for some considerable time, long enough to have become an established practice in the market in question, it probably does not constitute evidence that enables us to establish conclusively that the duration of the contravention extends beyond the period from which the complainant had entered the market.

185. Indeed, the first respondent insists that, not only do we not know the duration of the price fix, but, that the only evidence of a contravention on the part of the first respondent is the evidence relating to the application of a sanction – the cut in the discount. There is, argue the first respondents, no evidence that they fixed the price but merely that they applied a sanction to a party that failed to apply the agreed price. In Nel’s words, ‘no case was made out or evidence led at the hearing to the effect that FM was responsible in any way for setting this price convention.’

186. However this argument is thoroughly unpersuasive. Federal Mogul applied a sanction in order to compel those trading in Ferodo products to adhere to a pricing convention that applied in respect of those products. Who else, would they have us believe, established

this convention if not the brand owners themselves? It may well be that the convention was established and policed in collaboration

with other brake producers, that, in other words, the convention was a form of horizontal price fix. This we do not know. However, we do know that Federal Mogul applied the sanction and it makes no sense whatsoever that they applied the sanction in support of a convention imposed upon them from elsewhere.

187. We know, too, with certainty that the contravention – the resale price fix – applied to all Ferodo braking products. The Commission would have been well advised to investigate the relationship between the prices of Ferodo braking products and those

of other braking products. It may then have established that Ferodo products - a powerful presence in the market – played the role of a price leader. Had this been established the extent of the contravention would have been very far-reaching indeed. On the present evidence, however, we are only entitled to conclude that the contravention fixed the price of Ferodo products, a leading braking products brand with a significant market share. We do not know the extent to which this price fix influenced the price of braking products generally. We return to this point later.

Any loss or damage suffered as a result of the contravention

188. Both the Commission and the first respondent have similarly misconstrued this issue. The Commission, in its assessment of the damage, has sought to establish that the first respondent’s contravention resulted in the complainant’s demise. The first respondent, for its part, insists that it was the complainant’s tardy payment record and, in particular, the level of debt that it assumed in consequence of its final purchase order from the first respondent, that caused it to exceed its credit limit. It was this that gave rise to the first respondent’s decision to put the complainant on stop supply and that, ultimately, forced the latter to exit the market. Accordingly, the first respondent insists that the discount cut, which, as we have pointed out, was of short duration, caused little loss to the first respondent. Damage to the first respondent was induced by the cessation of supplies and that this in turn was caused by the complainant’s payment record.

189. It is reasonably clear that, in administering the discount cut, the intention of the first respondent was to discipline the complainant

and to secure his adherence to the pricing convention not necessarily to drive him from the market. The first respondent’s

decision to cease supplying its products to the complainant was clearly precipitated by the latter’s final purchase order and the key role played by that purchase in raising the complainant’s level of indebtedness to the first respondent. On the other hand, it is equally clear that the reduction in the complainant’s discount rate was a key factor in a chain of events that led to its demise, even if we accept that the factor that ultimately precipitated the complainant’s demise was caused by his level of debt rather than by the discount cut per se.

190. However, it is, in our view, not necessary to determine the role played by either of these factors, namely the discount cut and the complainant’s payment record, in the demise of the complainant. These factors may well assume central significance should the complainant elect to pursue a claim for private damages. However, from our perspective, the damage to the complainant is only significant to the extent that the weakening or the exit of a robust competitor does undoubtedly diminish the level of competition in the market for braking products.

191. Our task is rather to assess the damage to competition, the ‘public’ damage, as it were, wrought by the practice of resale price maintenance. There is no simple way of measuring this. Firmly established economic theory tells us that a range of investment decisions are distorted – and a misallocation of economic resources occurs – in consequence of producers fixing a supra-competitive price for their products. In this instance the supra-competitive price fix will secure a supra-competitive return to investment in braking products and, as a result, the direction of scarce investment resources to that activity when the competitive margins do not warrant that level of investment. This excessive investment in braking products will occur at the expense of investment in more deserving projects, ‘more deserving’ because on a level playing field, a playing field that is not skewed by an illegal price fix, they would secure a relatively higher rate of return than would investment in braking products.

192. On the other hand, in this particular instance, the price fix may well direct capital away from investment in innovation in braking

products. We have noted the first respondent’s proclaimed desire to maintain the reputation of its brand. It could do this through product innovation or through investing in superior service to its customers. However, it has chosen to protect its brand’s reputation by placing a floor beneath its price thus potentially resulting in a sub-optimal level of investment in improving the quality of its product.

193. Expressed in the clinical language of micro-economic theory this may be construed as a peculiarly ‘victimless crime’, a ‘mere’ statutory transgression that, while giving offence to economic theorists, visits no harm on flesh-and-blood citizens. However, this is not correct. It is those who have paid more for Ferodo products than they would otherwise have paid who are immediately damaged. And, equally, it is those who are effectively forced to accept an inferior product, a product that has not had the benefit of continuing investment in quality, that are damaged by the contravention.

194. Bear in mind, also, that by the first respondent’s own admission – a fact unfortunately not verified by the Commission – Ferodo has a market share of slightly below 30%. We note again that should the first respondent’s price fix have acted to support prices of competing braking products, then the extent of damage increases significantly. The first respondent will undoubtedly protest that there is no evidence of price leadership on its part. However, it requires no great insight to predict that a product, which the first respondent itself proclaims to be a leading brand and which has at least a 28% market share, will influence the price of competing products. Certainly, it is perfectly reasonable to assume that, had the distributors of Ferodo products been permitted to engage in a ‘price war’, the inevitable response of producers and vendors of competing braking product would have been to decrease their prices as well.

195. In our view, the probability is that the likely damage wrought by the first respondent’s practice of resale price maintenance is considerable and that this damage significantly compounds the harm caused by the exit of a robust competitor from the market for the sale of braking products.

The behaviour of the first respondent

196. The aspect of the first respondent’s behaviour that is most pertinent to the calculation of the administrative penalty undoubtedly centers upon the level of co-operation with the competition authorities. This is fully elaborated below.

197. The first respondent insists that the company should not be prejudiced by the behaviour of two employees – Moll and Taylor – who were most directly responsible for the offensive conduct. These employees did not represent the ‘controlling mind’ of the company. Nel, on the other hand, did control the company and, he avers, has never done anything to associate himself with the conduct of his two senior employees.[99]

198. However, the Commission points out that Nel has not, to this day, repudiated Moll and Taylor. To the contrary he explicitly endorsed the incriminating letter drafted by Moll. Also recall that Taylor was dispatched by Nel to communicate the news of the discount cut to Erasmus in the wake of a meeting between Erasmus, Taylor and representatives of Midas. It seems most probable that it is at this meeting where the disciplinary strategy was devised.

199. Note also that Moll and Taylor are not merely overenthusiastic sales representatives, anxious to meet their targets and, in the process, transgressing the boundaries of company policy and the law. They are senior executives of the first respondent reporting directly to Nel, the Managing Director. We are satisfied that the first respondent’s ‘controlling mind’ was thoroughly engaged with this contravention.

The level of profit derived from the contravention

200. Similar considerations apply to the calculus of profit derived from contravention, as to the quantum of harm or damage generated.

The first respondent has again attempted to link the profit derived from the contravention to the sanction, to the discount cut.

They then insist that they derived no profit from their action, that if any profit accrued from this then it did so in Midas’ favour rather than Federal Mogul’s. Again this misconstrues the contravention.

201. Federal Mogul has profited – for all we know, still profits – from a distribution scheme which maintained prices at a supra-competitive level and which reduced the requirement to invest in building the quality and reputation of the Ferodo brand as part of a strategy to maintain margins. We are obviously not able to put a precise figure on this. However, we are able to observe that its 28% market share coupled with the importance that the first respondent itself ascribes to brand development suggests that the profits derived from the anti-competitive conduct – measured as the sum of the price premium and the reduced investment in brand and product development - are significant.

202. The Commission avers – and we agree – that ‘the benefits may not have been in rands and cents but rather in the relationship with Midas in setting the tone for the relationships

with the independents’.[100] That is to say, the first respondent’s objective in practicing resale price maintenance and in punishing transgressors was not directed at the prospect of short-term pecuniary gain, but rather at keeping an entire anti-competitive pricing scheme in place.

The market circumstances in which the contravention took place

203. We have already ventilated our concerns at the market circumstances in which the contravention took place. We repeat: on the first respondent’s own admission, Ferodo is a major brand with a significant market share. It stands to reason that Ferodo is in a powerful position to influence the pricing and other competitive behaviour of its competitors. If it introduced a scheme that ensured that its prices were maintained and that reduced the necessity for capital investment, we can reasonably hypothesize that this decision fed into the decisions of competitors. By contrast a luxury good producer with a relatively small market share who wished to maintain the cachet surrounding his product, may fix resale prices and, as such, would contravene the same provision of the Act, but the market circumstances would serve to mitigate the consequences of the contravention. In these market circumstances, the converse is true – the impact of the contravention is exacerbated.

204. The first respondent has painted a picture of a highly volatile market, one in which “irrespective of dealers’ complaints, FM had a personal interest in preventing price wars of this type”.[101] As already noted it had a general interest in ensuring that its product was not ‘commoditised’ or ‘prostituted’.

Competitors were taking advantage of the first respondents’ preoccupation with its attempts to restructure its corporate and distribution structure and it did not want additionally to be burdened with the trauma of its distributors engaging in a price

war. It is, however, not at all clear why these conditions or ‘market circumstances’ should mitigate the anti-competitive conduct of the first respondent and thus the scale of the administrative penalty. These activities are part of the process of doing business, which does not afford the luxury of a ‘time-out’ while one or other market participant prepares itself for robust competition in the future. Indeed this vulnerable moment is precisely the moment when an able competitor

would attack. It is certainly not the time to allow a competitor the shield of an anti-competitive restraint.

The degree to which the first respondent has co-operated with the Competition Commission and the Competition Tribunal

205. The first respondent’s conduct in this matter leaves much to be desired.

206. While the first respondent is correct in pointing out that the Commission’s initial query received a prompt response, this appears increasingly to either reflect ignorance or arrogance, rather than a willingness to co-operate. In the event Moll, a senior executive of the first respondent, wrote a letter so incriminating that one must either believe that he was entirely ignorant of widely accepted rules governing business conduct or that he arrogantly believed his company to be beyond the reach of the fledgling competition authority.

207. However, even if we were to accept that Moll’s prompt reaction reflected an initially co-operative response from the parties, any semblance of co-operation ended when it became clear that the Commission was not going to be fobbed off by Moll’s response. Indeed it was immediately replaced by the most brazen obstructionism. This is particularly manifest in the problems encountered by the Commission in its efforts to identify the correct first respondent in this matter. Note the following chronology:

· Pee Dee Wholesalers lodged its complaint against Federal Mogul Aftermarket Southern Africa (Pty) Ltd (“Federal Mogul Aftermarket SA”) early in November 1999.

· On 19 November 1999 Federal Mogul Aftermarket SA replied to a letter of the Competition Commission and informed the Commission that the complainant “has been dealing with Federal Mogul Friction Products (Pty) Ltd. Federal-Mogul Aftermarket Southern Africa (Pty) Ltd is the company

within the Federal-Mogul group that is in the process of taking over the marketing and sales of the friction products.”

· The Commission proceeded with its investigation and on 7 February 2001 filed a complaint referral against Federal Mogul Aftermarket SA with the Tribunal.

· However, on 25 April 2001 Federal Mogul Aftermarket SA’s attorneys informed the Commission that its client was not the entity with which the complainant had been dealing. According to them the complainant had dealt with T&N Friction Products (Pty) Ltd trading as Ferodo, a distributor of Ferodo’s products and advised the Commission to withdraw the complaint against its client. Based on this the Commission instructed its attorneys to investigate the information furnished with the Registrar of Companies. It was found that Federal-Mogul Friction Products had bought T&N Friction Products in 1998.

· Although evidence submitted to the Commission indicated that Federal Mogul Aftermarket SA was the correct party cited as the respondent the Commission decided to apply to the Tribunal for the joinder of the 2nd – 4th Respondents.[102] Subsequent to hearing this application, as well as a counter challenge brought by the 2nd to 4th respondents that the

Tribunal is not competent to order the joinder of respondents in such proceedings, the Tribunal ordered that they be joined. Nel

acknowledged at the Hearing held on 27 August 2002, that he was aware at the time when his attorneys wrote the 25 April letter that, although T&N Friction Products held Pee Dee’s account at the time, Federal Mogul Aftermarket SA, in fact he himself, was since June 1999 responsible for any decisions concerning the account. The hearing of the complaint referral proceeded on the basis that only the first respondent is the appropriate respondent in the proceedings.

· Subsequent to the joinder hearing the Commission filed an amended complaint referral. The Respondents were late in filing their amended answering affidavit stating that the Commission’s amended complaint referral was misfiled. As a result the Respondents’ attorneys were not aware of its existence and had answered the wrong complaint referral when they eventually got around to doing it late in December 2001, instead of the end of November 2001. The amended answering affidavits were only filed by the end of January 2002.

208. While we reluctantly concede that this type of point-taking obstructionism is standard fare in the practice of litigation, it clearly cannot be mistaken for co-operation. The Commission was obliged to spend considerable time and resources to identify the proper ‘respondent’ a fact that later was conceded when it became clear to the first respondent that the Commission was determined to pursue the complaint and was not going to be fobbed off by an evasive strategy.

Conclusion and Finding

209. We have considered the factors set out in Section 59(3) of the Act. We are, in our view, dealing with a grave contravention of the Act. On the present evidence we know that the contravention had endured from the time that Erasmus entered the market and, although it may not be unreasonable to infer a longer duration, we give the first respondent the benefit of any residue of doubt and so conclude that the duration of the contravention was short. The first respondent’s share of the market and the leading character of its brand lead us to conclude that the contravention was of a far-reaching extent. It is not possible to calculate precisely, the loss of damage suffered as a result of the contravention, nor the level of profit, which accrued to the first respondent. However, the nature of the product and the first respondent’s position in the market enables us to conclude with confidence that the damage wrought to the competitive fabric of the market was significant. While the first respondent has not previously been found in contravention of the Act, it has not co-operated with the Commission in its investigation – indeed it has resorted to the expedient of legal technicality and plain deceit to throw the investigators off course.

210. Given that we have found that the threshold for the maximum penalty is the turnover in Ferodo products and that is set at [confidential], we find that three million Rand (R 3 000 000.00) is an appropriate penalty having regard to all the factors set in section 59(3).

211. We make the following order:

1) The first respondent is ordered to pay an administrative penalty of three million Rand (R 3 000 000.00).

2) The penalty must be paid to the Commission within 21 business days of the date of this order.

Date 21 August 2003

D. Lewis

Concurring: M.T.K. Moerane and N. Manoim

[1] Tribunal Case No: 08/CR/Mar01 of 28 January 2003

[2] When the first respondent raised the constitutional points in its initial heads of argument the Tribunal decided that the points of law were of such importance that it would be appropriate to brief amici curiae to make submissions on these aspects. Advocates Gilbert Marcus SC and Mathew Chaskalson were briefed.

[3] Section 170 of the Constitution states that courts of a status lower to that of a High Court may not enquire into or rule on the

constitutionality of any legislation.

[4] 77 DLR (4th) 94 (SCC)

[5] Note the motivation for a decriminalized administrative fine system in relation to tax violations in the United Kingdom by the Keith report. “We have noted … the high resource cost of the investigation of fraud to the criminal standard, and the understandable constraints this imposes on the investigation of the smaller frauds……. In those cases where “civil” investigation techniques suffice to secure evidence of the true extent of the fraud, the process is an economical one at least by comparison with the cost of a comparable criminal investigation. The investigation of acts of dishonesty in relation to tax matters in a “civil” style, reinforced by inducements, rather than as criminal offences under the Judge’s Rule, are as such generally welcome to our witnesses and we heard no consistent body of criticism of the lower civil burden of proof in such cases as being unfair to the taxpayer.” This extract from the Keith Report which investigated the VAT civil penalty code is referred to in the case of Han & Yau Martins and Martins Morris v Commissioners of Customs and Excise [2001] EWCA Civ 1040 (3rd July, 2001), which we discuss later in our reasons.

[6] See Explanatory memorandum to the Competition Bill, [B98-98], page 68.

[7] Note that an administrative penalty is only appropriate for contraventions that are set out in section 59(1).

[8] See S v Zuma 1995(2) SA 642 (CC) at para 33

[9] See section 68 which states: “In any proceedings in terms of this Act, other than proceedings in terms of section 49C or criminal proceedings, the standard of proof is on a balance of probabilities.”

[10] Section 56(2) of the Act can certainly be read to protect this right.

[11] 1996(3) SA 562 (CC)

[12] ibid at 571 D-E

[13] ibid at 571 E-F, par [11]

[14] S v Baloyi 2000(2) SA 425 (CC).

[15] See Baloyi supra, para 22.

[16] See Baloyi supra, para 22

[17] The full name is the European Convention for the Protection of Human Rights and Fundamental Freedoms. Article 6 states in its opening sentence: “ In the determination of his civil rights and obligations or of any criminal charge against him, everyone is entitled to a fair and public hearing within a reasonable time by an independent and impartial tribunal established by law.”

[18] Engel and Others v The Netherlands (No 1) [1976] ECHR 3; 1 EHRR 647, 8 June 1976, para. 81, page 678

[19] The article does provide for exceptions to a hearing in public but none of these had been pleaded by the government. See Engel page 681.

[20] Section 34 of the Constitution refers to a “fair public hearing”. The right is also protected in the Competition Act. (See section 52(2)(a)).

[21] See Engel para. 82

[22] See Engel para. 82

[23] See argument for counsel referred to in Han, paragraph 73.

[24] See Ozturk v Germany [1984] ECHR 1; 6 EHRR 409, page 437, para 2

[25] See Ozturk page 434 para3

[26] 26 EHRR page 541.

[27] See AP, MP and TP v Switzerland page 558-9 paragraph 41.

[28] See AP, MP and TP v Switzerland, the dissenting opinion of Judges Baka and Bernhardt, page 562.

[29] The court at paragraph 74 indicated that it was reluctantly persuaded because in its view it was persuaded by the Keith Report that the VAT Civil penalties scheme was a legitimate balance between the interests of Customs and Excise and the taxpayer in avoiding the travails of a criminal prosecution and the stigma of conviction for a crime involving dishonesty.

[30] Napp Pharmaceutical Holdings Ltd v Director General of Fair Trading, Case No 1001/1/1/1/01, 15 January 2002 unreported.

[31] See Napp para 101

[32] See Napp para 103

[33] See Napp para 102

[34] Wigglesworth v The Queen 32 CRR 219 (SCC)

[35] See Wigglesworth page 237

[36] See Wigglesworth page 237

[37]R v Shubley 46 CRR 104 (SCC)

[38] 448 US 242, 1980

[39] Ward at 249

[40] Hudson v US 522 US 93, 99 (1997)

[41] (1963) 372 US 144, 9 L Ed 2d 644, 83 S Ct 554

[42] ibid Hudson, page 460

[43] US v Halper 490 US 435

[44] Hudson pages 460-461.

[45] Ibid at 463

[46] 430 US 442 (1977)

[47] See Atlas Roofing at 455.

[48] The idea of the concepts existing in a continuum arises from an article by Professor Mann in the Yale Law Journal, June 1992 entitled

“Punitive Civil Sanctions: The Middleground Between Criminal and Civil Law”.

[49] See fn. 47 supra.

[50] See for example the Mine, Health and Safety Act 29 of 1996 (amended by Act 72 of 1997), The Basic Conditions of Employment Act, 75 of 1997, The Employment Equity Act, 55 of 1998, section 45(1) of the Pharmacy Act 53 of 1974 as amended and section 102 of the Telecommunications Act, 103 of 1996 as amended.

[51] Nevertheless Areeda acknowledges that a basic policy dispute about antitrust law concerns the appropriateness of criminal sanctions. See Areeda Hovenkamp, Antitrust Law, ¶ 300-397, ¶303b3, page 31. Areeda also points out that because the courts have been usually dealing with Sherman Act cases in civil proceedings they have not felt constrained as they might have been in developing the common law of antitrust, (¶303b4). Note that in the United States although the Sherman Act, its principal antitrust statute, is criminal the Department of Justice may use it to proceed, both civilly or criminally. The well-known Microsoft case is an example of the DOJ proceeding with a civil case for monopolization under the Sherman Act.

[52] See for instance Yeung who states: “While criminal offences typically depend on some proof of mens rea, be it intent, knowledge or recklessness, in order to establish the commission of an offence, regulatory violations often discard mens rea so that a regulatory offence may be committed, without the need to establish any mental element on behalf of the offender. In other words liability is visited on the offender for breach of regulatory law without proof of the offenders culpability or subjective wickedness.” See Karen Yeung, “Quantifying Regulatory Penalties: Australian Competition Law Penalties in perspective” Melbourne University Law Review, August 1999, page at 459.

[53] See these references in section 4(1)(a), 5(1), 8(c),8(d), 9(1)(a).

[54] Patensie Sitrus Beherend Beperk v The Competition Commission and others, CAC Case No: 16/CAC/Apr02, page 29

[55]Case no 85/76[1979] ECR 461, para 89.

[56] Hoffmann-La Roche supra para 91

[57] Recall the test in Kennedy v Mendoza –Martinez supra.

[58] See Areeda op cit ¶ 311(d)

[59] See section 79(2) of the Canadian Act. According to Paul Collins and D. Jeffrey Brown, until 1986, monopolization was a criminal offence under the Act. The ineffectiveness of the former Canadian provision led to its replacement with a new civilly reviewable provision for abuse of a dominant position..” See Paul Collins and Jeffrey Brown , National antitrust law in a continental economy: A comparison of Canadian and American antitrust laws, 65 Antitrust Law Journal 495 at 530.

[60] See Areeda ¶ 303d.

[61] See Areeda ¶ 303d

[62] See Yeung op cit 445. Yeung refers to retribution as desert.

[63] See Punitive Damages: Reconceptualizing the Runcible Remedies of Common Law, Edward L. Rubin, Wisconsin Law Review 1998

[64] Rubin op cit 136.

[65] Rubin op cit 140.

[66] Rubin op cit 142.

[67] We rely on Yeung op cit footnote 73 for her summary of his views in this way. She sources the view to Joel Feinberg, “The Expressive Function of Punishment” (1965) 49 The Monist 397.

[68] See Mann op cit 1846-7.

[69] Although the Courts impose these penalties they do so on the basis of the civil burden of proof. The other difference is that the cap for the penalty is set by an amount stipulated in the statute and not by reference to turnover.

[70] 1996 71 FCR 285

[71] See NW decision at 292, 1991 ATPR 41 076.

[72] There are various other quotes by Lee J in other decisions which are referred to, all of which emphasise the deterrent effect

[73] [2001] ATPR 41-833

[74]ibid para 12

[75] In a dissenting judgment in NW Frozen Foods case Carr J says that on his view of the cases punishment may still be one of the purposes of section 76.

[76] As Korah has explained the European model from which the 10 % turnover penalty is derived, the test is intended to reflect the economic power of the firms responsible (See EC Competition Law and Practice, 7th Edition 2000, by Valentine Korah, page 179 quoting the Court of First Instance.)

[77] See John C Coffee. Jr: Paradigms Lost: The Blurring of the Criminal and Civil Law Models and what can be done about it, Yale Law Journal, June 1992

[78] Section 28(3)(d) of the Competition Act contains a typical example of this kind of disability. The section provides that a person is ineligible to serve as a Tribunal member if the person has been convicted of an offence committed after the Constitution of the Republic of South Africa, 1993 (Act No. 200 of 1993), took effect, and sentenced to imprisonment without the option of a fine.

[79] See for instance Robert F. Blomquist who writes on this distinction: “The purpose of a penalty is also to inflict a pecuniary punishment for the violation of a law; however, the technical definition of a penalty is ‘that which is demanded for a violation of a statute, which violation may or may not be a crime.’ This technical distinction between a penalty and a fine survives in statutes that allow a penalty to be recovered in a civil action but require that a crime be punished with a fine or imprisonment.” See Blomquist , Rethinking the citizen as prosecutor model of environmental enforcement under the Clean Water Act, 22 Georgia Law Review 337,1998 at 360.

[80] See transcript of 23 April 2003, page 31, line 20

[81] Contrast this with the Tribunal’s predecessor the Competition Board whose decisions were recommendations only and required Ministerial assent to have the force of law. See Maintenance and Promotion of Competition Act, section: 6, 12, 13 and 14.

[82] Van Rooyen and Others v The State and Others (General Council of the Bar of South Africa Intervening) 2002 (5) SA 246 (CC)

[83] ibid Para 93.

[84] Van Rooyen supra para 106. The Court refers to the fact that in its first certification judgment it held that the executive could have retained the power to appoint judges (and magistrates) itself, without infringing the institutional independence required by the

Constitutional principles. (Para 109)

[85] Freedom of Expression Institute and Others v President of Ordinary Court Martial , and Others. 1999(2) SA 471 (C) at 483

[86] Note that even the choice of 10% as the measure has some pedigree in the field of competition economics. When assessing whether a firm has monopoly power economists apply the test of the hypothetical monopolist – could the firm raise its price by 5-10% without it not being feasible for it to do so. (See Competition Law, fourth edition by Richard Whish, page 28)

[87] The amici relied on the following cases for this proposition: Prinsloo v van der Linde 1997 (3) SA 1012 (CC) at para 36, East Zulu Motors (Pty) Ltd v Empangeni/Ngwelezane Transitional Local Council and Others 1998 (2) SA 61 (CC) at paras 24 and 30, S v Lawrence; S v Negal; S v Solberg 1997 (4) SA 1176 (CC) at paras 41-46.

[88] In Australia, whose section 76 as we mentioned earlier is the one that influenced our section 59(3), the Federal Court has stated that: “the fixing of the quantum of penalty is not an exact science. It is not done by the application of a formula, and, within a certain range, courts have always recognized that one precise figure cannot be incontestably said to be preferable to another.” See TPC v TNT [1995] FCA 1166; (1995) ATPR 41- 375 ,40,165(Burchett J)

[89] For Commission’s view see Transcript, 7 April 2003, p6

[90] Indeed the ‘volume of affected commerce’ is, on the Commission’s own version, itself measured by reference to turnover. See Transcript p7

[91] See R. v Browning Arms Company of Canada Ltd., (1974) 18 C.C.C (2d) 298 (Ont.C.A.), a judgment often referred to in sentencing jurisprudence in Canada, according to Davies, Ward & Beck in Competition Law of Canada, where the Court stated at 303 that: “The section’s role in the protection of free competition has commercial importance and breach of it has important economic implications and

consequences. The aspect of deterrence to others who might be tempted to commit similar offenses is clearly an important factor to be taken into account. The penalty imposed must not be so small as to be regarded by the accused or other corporations as a license fee for carrying on business in a manner contrary to law.” Note that, although resale price maintenance is regarded as a criminal offence in Canadian Competition law, the principle remains relevant for our purposes.

[92] See transcript page 39-40.

[93] It appears that the European Union, which also caps fines at 10% of the firm’s total sales in its previous financial year interprets this as previous to the imposition of the fine, not to the cessation of the infringement.

[94] See the European Court of Justice as articulated in Musique Diffusion Francaise (100-103/80 ECR 1825), at para 121: ‘It follows that, on the other hand, it is permissible, for the purpose of fixing the fine, to have regard both to the total turnover of the undertaking, which gives an indication, albeit approximate and imperfect, of the size of an undertaking and of its economic power, and to the proportion of that turnover accounted for by the goods in respect of which the infringement was committed, which gives an indication of the scale of the infringement. On the other hand, it follows that it is important not to confer on one or other of those figures an importance disproportionate to the other factors and, consequently, that the fixing of an appropriate fine cannot be the result of a simple calculation based on total turnover.’

[95] The Ferodo turnover according to the first respondent’s figures which the Commission does not dispute is R 60,318,000.00

[96] In adopting this approach we are at one with the European approach where, according to one authoritative text placed on record by the respondents: ‘…a multitude of factors are weighed by the Commission. These factors can broadly be summarised under three headings: the type of infringement, the importance and structure of the market and the defendant’s position in it, and the defendant’s conduct.’ On the question of the type of infringement, the writers note further that ‘clear-cut hard-core violations are viewed seriously and attract a heavier fine’. (Ritter, Braun and Rawlinson EEC Competition Law: A Practitioner’s Guide (Kluwer) at page 683. Pertinent to the case before us, another European authority notes that the EU authorities view ‘retaliatory measures against other undertakings with a view to enforcing practices which constitute an infringement’ as an ‘aggravating circumstance,’ which will increase the basic amount of the fine. (Valentine Korah – Cases and Materials on EC Competition Law at p234)

[97] See First Respondent’s Affidavit Regarding Administrative Penalty (affidavit of Frederik Johannes Nel) para 24.3

[98] Indeed the Commission formally concedes that the duration of the contravention is ‘short, being under a year’. (Transcript p15)

[99] See First Respondent’s Affidavit regarding administrative penalty (affidavit of Frederik Johannes Nel), para 25.

[100] Transcript p13

[101] See First Respondent’s Affidavit Regarding Administrative Penalty (affidavit of Frederik Johannes Nel), para 24

[102] Federal Mogul Friction Products (Pty) Ltd, T & N Holdings Ltd