Solidarity obo Van Zyl v KPMG Services (Pty) Ltd and Others (JR 960/12) [2013] ZALCJHB 251; (2014) 35 ILJ 1656 (LC) (10 October 2013)

REPUBLIC OF SOUTH AFRICA

THE LABOUR COURT OF SOUTH AFRICA, JOHANNESBURG

JUDGMENT

Reportable

Case no: JR 960/12

In the matter between:

SOLIDARITY OBO VAN ZYL ................................................................................APPLICANT

and

KPMG SERVICES (PTY) LTD ...............................................................FIRST RESPONDENT

COMMISSIONER NKGOENG N.O ...................................................SECOND RESPONDENT

COMMISSION FOR CONCILIATION,

MEDIATION AND ARBITRATION ........................................................THIRD RESPONDENT

HEARD: 10 JULY 2013

DELIVERED: 10 October 2013

SUMMARY: Test for review of credibility findings in arbitration awards

___________________________________________________________________

FOURIE, AJ

Introduction

To what extent will shortcomings or errors by an arbitrator in the assessment of credibility of witnesses, render the award reviewable? That is the key issue for determination in this review application.

The facts

The applicant (Van Zyl), a Trainee Accountant at KPMG, was alleged to have used a racial slur (relating to a controversial political figure) during a conversation with colleagues while driving to lunch. One of the passengers in the car, Ms Wendy Marais, complained to Van Zyl’s manager about the use of unacceptable language by KPMG personnel. This prompted an investigation, and in due course Van Zyl was charged with misconduct. Van Zyl protested his innocence, but was found guilty and was dismissed. Only Marais and Van Zyl testified to the incident at the disciplinary inquiry.

At arbitration, in addition to the evidence of Marais (who testified to the incident) and Van Zyl (who denied it), two other occupants of the vehicle at the time that the alleged racial slur was uttered, testified. Both testified that they did not hear Van Zyl use the racial slur or the context in which he was alleged to have used it, and would have heard it, had it happened.

The arbitrator was thus faced with mutually irreconcilable and contradictory versions. Given the nature of the alleged misconduct, and the obvious lack of evidence aside from the viva voce evidence of those persons in the car at the time, the finding as to whether the words were said would be decisive as to the outcome of the arbitration proceedings.

The key finding in the arbitration award reads as follows:

‘38. However, it is my findings that the evidence of the respondent’s witness, Wendy Marais was more probable than that of the applicant and his witnesses. The evidence of the applicant’s witnesses was that they did not hear the use of the ‘K’ word. The applicant on the other hand could not advance any reason why Wendy would fabricate a story against him. The applicant conceded that there was no bad blood between him and Wendy.

39. Based on the above, it is my conclusion that the respondent has successfully discharged the onus of proof that the applicant was guilty of the said charge and thus makes the dismissal of the applicant by the respondent substantively fair.’

The law on reviews of credibility findings

For several years the Labour Courts have held that the duty of an arbitrator when confronted by two irreconcilable versions, is

essentially the same as that of a Judge in a trial court, as expressed by the Supreme Court of Appeal in Stellenbosch Farmers' Winery Group Ltd and Another v Martell Et Cie and Others1 :

'To come to a conclusion on the disputed issues a court makes findings on (a) the credibility of the various factual witnesses; (b) their reliability; and (c) the probabilities. As to (a), the court's finding on the credibility of a particular witness will depend on its impression about the veracity of the witness. That in turn will depend on a variety of subsidiary factors, not necessarily in order of importance, such as (i) the witness' candour and demeanour in the witness-box, (ii) his bias, latent and blatant, (iii) internal contradictions in his evidence, (iv) external

contradictions with what was pleaded or put on his behalf, or with established fact or with his own extracurial statements or actions,

(v) the probability or improbability of particular aspects of his version, (vi) the calibre and cogency of his performance compared to that of other witnesses testifying about the same incident or events. As to (b), a witness' reliability will depend, apart from the factors mentioned under (a) (ii), (iv) and (v) above, on (i) the opportunities he had to experience or observe the event in question and (ii) the quality, integrity and independence of his recall thereof. As to (c), this necessitates an analysis and evaluation of the probabilities and improbabilities of each party's version on each of the disputed issues. In the light of its assessment of (a), (b) and (c) the court will then, as a final step, determine whether the party burdened with the onus of proof has succeeded in discharging it. The hard case, which will doubtless be the rare one, occurs when the court's credibility findings compel it in one direction and evaluation of the general probabilities in another. The more convincing the former, the less convincing will be latter. But when all factors are equipoised probabilities prevail.'

This position seems to have been consistently adopted since at least 2007 (see Vodacom Service Provider Co (Pty) Ltd v Phala NO and Others (2007) 28 ILJ 1335 (LC), Lukhanji Municipality v Nonxuba NO and Others (2007) 28 ILJ 886 (LC)) and has been applied fairly consistently since then (see for example, Sasol Mining (Pty) Ltd v Nggeleni NO and Others (2011) 32 ILJ 723 (LC), SATAWU obo Semate v CCMA & Others, Unreported Labour Court decision, Case JR341/2010, 18 December 2012).

In Sasol Mining, Van Niekerk J, after quoting the abovementioned passage from Stellenbosch Farmers' Winery, held as follows:

‘[9] One of the commissioner’s prime functions was to ascertain the truth as to the conflicting versions before him. As I have noted, this much the commissioner appears to have appreciated. What he manifestly lacked was any sense of how to accomplish this task, or which tools were at his disposal to do so. The commissioner was obliged at least to make some attempt to assess the credibility of each of the witnesses and to make some observation on their demeanour. He ought also to have considered the prospects of any

partiality, prejudice or self-interest on their part, and determined the credit to be given to the testimony of each witness by reason of its inherent probability or improbability. He ought then to have considered the probability or improbability of each party’s version.’

While the general approach adopted above may be sound, in my view it needs to be exercised with some caution, given the limited grounds of review available in our law, and the need to recognize and maintain the fundamental (and important) distinction between appeal and review proceedings. To simply apply the onerous test applicable to appeals against findings of fact by trial courts, to reviews of arbitration awards, without careful appreciation of the special nature of review proceedings, could confuse the distinction between appeal and review proceedings.

It should always be borne in mind that the review powers of a Court are limited to those contained in section 145(2) of the LRA, and the ground of reasonableness as set out in the judgment of the majority of the Constitutional Court in Sidumo2. As the Supreme Court of Appeal stated in Samancor3:

‘It is trite that an appeal does not lie against the award of an arbitrator. Even if the reviewing court believes the award to be

wrong, there are limited grounds upon which it is entitled to interfere. Section 145 of the Labour Relations Act 66 of 1995 permits the Labour Court to set aside an award for one or other defect stated in s 145(2) – none of which are now applicable. But it was recognised in Sidumo v Rustenburg Platinum Mines Ltd, adopting what was held in Carephone (Pty) Ltd v Marcus NO, that an award may also be set aside if it is one that ‘a reasonable decision-maker could not reach’, and it was on that basis that Samancor sought to have the award set aside.’.

The LAC in Bestel v Astral Operations Ltd and Others4 summarised the test for review as follows:

‘It is important to emphasise…that the ultimate principle upon which a review is based is justification for the decision as opposed to it being considered to be correct by the reviewing court; that is whatever this Court might consider to be a better decision is irrelevant to review proceedings as opposed to an appeal. Thus, great care must be taken to ensure that this distinction, however difficult it is to always maintain, is respected.’ (Own emphasis).

Recently, in Herholdt v Nedbank Limited5 the Supreme Court of Appeal re-emphasised the primacy of the Sidumo-approach to review proceedings, and rejected the argument (and line of authorities in the Labour Court) that, based on a dictum in a minority judgment in Sidumo, that the failure by an arbitrator to take into account relevant facts could result in a reviewable irregularity in the process which in itself justified setting aside the award, regardless of the reasonableness of the outcome. The SCA stated the correct review test as follows:

‘ In summary, the position regarding the review of CCMA awards is this: A review of a CCMA award is permissible if the defect in the

proceedings falls within one of the grounds in s 145(2)(a) of the LRA. For a defect in the conduct of the proceedings to amount to a gross irregularity as contemplated by s 145(2)(a)(ii), the arbitrator must have misconceived the nature of the inquiry or arrived at an unreasonable result. A result will only be unreasonable if it is one that a reasonable arbitrator could not reach on all the material that was before the arbitrator. Material errors of fact, as well as the weight and relevance to be attached to particular facts, are not in and of themselves sufficient for an award to be set aside, but are only of any consequence if their effect is to render the outcome

unreasonable.’ (Own emphasis)

In the wake of the SCA’s decision in Herholdt, it is necessary to consider the approach to be adopted by a reviewing court when assessing the reasonableness of an arbitration

award in which credibility findings play a key role in the outcome, as in this case.

In my view, the correct approach to assessing credibility findings at the stage of review applications, is as set out by Redding AJ in the unreported decision of Transnet Ltd v Gouws and Others6:

11. The proper approach of a court (or arbitrator) which is called upon to determine which of two mutually destructive versions should be accepted was related in the judgment of Stellenbosch Farmers Winery Group Limited and other v Martell et Cle and others 2003 (1) SA 11 (SCA) at 14 J – 15 E, where Nienaber JA is reported as follows:

… [see quote above]

12. That judgment emphasises the interrelationship of credibility of the witnesses, their reliability and the probabilities. However, it is be borne in mind that the ultimate decision which a court, or an arbitrator (as the case may be) must determine is whether on the issue in question the party which bears the onus has discharged it. In cases concerning the fairness of dismissals under the LRA, the party which bears the onus of justifying the dismissal is the employer. The question which the arbitrator must ask in discharging its duties as such is where the probabilities lie. If the probabilities favour the employer, it may well discharge the onus of proving the dismissal was fair. If they do not, the employer may fail.

13. The importance of credibility has sometimes been over-estimated. An assessment of evidence on the basis of credibility only, without

regard for the underlying probabilities, is inappropriate – indeed, it constitutes a misdirection. (See Medscheme Holdings (Pty) Limited v Bhamjee 2005 (5) SA 339 (SCA) at 345 A)

14. The proper approach was expressed by Eksteen J in National Employers General Insurance Co. Limited v Jagers 1984 (4) SA 437 (E) at 440 D-H, which is to the following effect:

‘It seems to me, with respect, that in any civil case, as in any criminal case, the onus ordinarily only be discharged with adducing credible evidence to support the case of the party on whom the onus rests. In a civil case the onus is obviously not as heavy as it is in a criminal case, but nevertheless where the onus rests on the plaintiff as in the present case and where there are two mutually destructive stories, he can only succeed if he satisfies the Court on a preponderance of probabilities that his version is true and accurate and therefore acceptable, and that the other version advanced by the defendant is therefore false or mistaken and fails to be rejected. In deciding whether the evidence is true or not, the Court will weigh up and test the plaintiff’s allegations against the general probabilities. The estimate of the credibility of a witness will therefore be inextricably bound up with a consideration of the probabilities of the case and, if the balance of probabilities favours the plaintiff, then the court will accept his version as being probably true. If, however, the probabilities are evenly balanced in the sense that they do not favour the plaintiff’s case any more than they do the defendant’s the plaintiff can only succeed if the Court nevertheless believes him and is satisfied that his evidence is true and that the defendant’s version is false.’

…

18. The second respondent did not pertinently address the question of the credibility of the three witnesses he mentions – Mr Van Rensburg, the first respondent, and the first respondent’s wife, Mrs Gouws. It is this failure which draws the criticism of the applicant in these proceedings. It has been submitted that there ought to have been a discrete enquiry by the second respondent into the question of each witness’s credibility.

19. As I have indicated above, the important question which had to be tackled by the arbitrator was whether the employer, on a preponderance of probability, had established that the first respondent had received cash bribes from Nu-Liner. The key question for him was which version was more probable. He was able to reach a decision on the probabilities without having to have regard to the credibility of each witness. It is quite possible for evidence to be assessed purely on its probability, assuming for the purposes of that assessment that the witnesses who testified were credible. It is not necessary for a judicial officer or arbitrator to find a witness not to be credible in order to find that his evidence is not probable. In this regard, there is an informative and authoritative article by the former judge of the Appellate Division, H.C. Nicholas “’Credibility of Witnesses’” (1985) 102 SALJ 32”.

20. In my view, the failure by the arbitrator to make a pertinent finding on credibility does not demonstrate that he failed to understand the proper approach to the assessment of conflicting evidence. The arbitrator appears clearly to me to have understood that his primary task was to resolve the conflicting versions by having regard to the balance of probability. He applied the correct judicial technique in this regard. Accordingly, his failure to address the credibility of each witness and comment thereon is not a fatal flaw which would entitle the applicant to review of his award.” (Own emphasis).

It should also be borne in mind that errors of fact by an arbitrator, particularly in respect of facts that he or she is empowered to determine (such as findings of fact on the probabilities), will not usually give rise to a valid ground of review. As pointed out by the Supreme Court of Appeal in Dumani v Nair and Another7:

[29] … 'Recognition of material mistake of fact as a potential ground of review obviously has its dangers. It should not be permitted to be misused in such a way as to blur, far less eliminate, the fundamental distinction in our law between two distinct forms of relief: appeal and review. For example, where both the power to determine what facts are relevant to the making of a decision, and the power to determine whether or not they exist, has been entrusted to a particular functionary (be it a person or a body of persons), it would not be possible to review and set aside its decision merely because the reviewing Court considers that the functionary was mistaken either in its assessment of what facts were relevant or in concluding that the facts exist. If it were, there would be no point in preserving the time-honoured and socially necessary separate and distinct forms of relief which the remedies of appeal and review provide.'

….

[33] For these reasons, even if there were a misdirection by the presiding officer in regard to the evidence of Claassen, the convictions would not be reviewable on the ground of material error of fact, nor under the guise of the provisions of s 6(2)(e)(iii) of PAJA, viz 'because irrelevant considerations were taken into account or relevant considerations were not considered'. That leaves the following grounds of review relied upon by the appellant, namely that the presiding officer acted arbitrarily (based on s 6(2)(e)(vi) of PAJA) and that the presiding officer's decision was so unreasonable that no reasonable person could have reached it (based on ss 6(2)(f)(ii)(cc) and (h) of PAJA). (The alleged misdirection to which I have referred would be relevant, if established, to the latter ground in considering whether, on the facts before the presiding officer as disclosed in the record, no reasonable person could have found the appellant guilty.) These grounds are dealt with in the judgment of my colleague Theron JA in whose judgment I concur.’ (Own emphasis)

In summary, complaints about errors made in the treatment and assessment of evidence by an arbitrator will normally be bound up in an attack on the overall reasonableness of the result, and will not in themselves constitute separate grounds of review to be determined independently from the result (as they did under the ‘latent irregularity’ or ‘dialectical unreasonableness’ approach).

While arbitrators should always aspire to meet the exacting standard set by the Supreme Court of Appeal in Stellenbosch Farmers Winery for the proper assessment of conflicting versions by a finder of fact, an arbitration award that does not live up to this standard will not automatically be subject to review. Arbitrators are empowered to deal with the dispute with a minimum of legal formalities, their decisions are immune from appeal, and the legislature has set a high bar for reviewing arbitration awards. Errors committed by an arbitrator in the assessment thereof will not necessarily vitiate an award.

Application to the facts

The arbitrator’s reasoning leaves much to be desired. There is no indication that he undertook an assessment of the credibility of other witnesses, including the applicant. Aside from explaining why he considers it unlikely that Marais’ evidence was fabricated or tainted by improper motive, the arbitrator provides no further reasons for his conclusion on the probabilities.

When measured against the requirements set in Stellenbosch Farmers Winery, as adopted by the Labour Court in cases such as those mentioned above, it is clear that the award falls short of the mark. This does not, however, mean that a Court sitting on review can intervene without considering whether the arbitrator’s conclusions were reasonable, in light of the issues and the evidence, and not only on the reasons provided by him (See Herholdt at para 12). It must also be borne in mind that is not strictly necessary for an arbitrator to find that a witness was not credible, in order to find that his version was not probable (See Transnet above).

The arbitrator was alive to the key dispute, and applied his mind to it. The arbitrator’s observations on the lack of any indication as to why the complainant would have fabricated a complaint against the applicant, are material to the overall probabilities. The record indicates that the Marais initially did not mention Van Zyl in her email, but merely complained about the use of inappropriate language by KPMG personnel, with the intention that clerks be reminded to think carefully before speaking, as unprofessional conduct and thoughtless remarks could have serious consequences for KPMG in its client relations (such as with the Auditor General’s office, where these individuals were working at the time). It was only after KMPG prompted her for further information as to the incident that she mentioned Van Zyl. KPMG then proceeded to take disciplinary steps against Van Zyl out of its own accord, and not at Marais’ insistence. Marais testified that she never envisaged or intended that her complaint would result in Van Zyl’s dismissal. Marais appears to have been reluctant to become involved in the disciplinary and subsequent arbitration proceedings, but did defend her initial complaint, and provided what appears to be cogent and credible evidence of the incident at both the disciplinary hearing and the arbitration. Attempts during cross-examination to challenge her credibility or to impugn her motives were not successful.

The conclusion reached by the arbitrator, that Marais’ version was, on the probabilities, more likely, is in my view a finding that a reasonable arbitrator could have reached on the totality of the evidence before him.

A further ground of review related to the alleged inconsistent application of discipline by KPMG. It was alleged that KPMG had failed to take disciplinary action in respect of similar incidents in the past. This review ground can be disposed of speedily.

22.1. Firstly, there is no clear evidence of KPMG ignoring complaints of similar misconduct. On the contrary, Mr Jeremy van Niekerk, managing partner of KPMG’s Nelspruit office, testified that the company views such conduct in an extremely serious light, and that if a complaint is made, it will be investigated and disciplinary action will follow.

22.2. Secondly, the nature of the misconduct is so serious as to warrant summary dismissal, regardless of past oversights by an employer. Racist hate speech can simply not be tolerated in an open and democratic society. This principle has been confirmed and applied by our Courts on numerous occasions. See for example the recent judgment of Gaibie AJ in Modikwa Mining Personnel Services v Commission for Conciliation, Mediation and Arbitration and Others (2013) 34 ILJ 373 (LC) at para 24-26, and the authorities cited therein.

Due to the measure of uncertainty in the law on review of arbitration awards based on credibility findings, I am not inclined to order either party to pay costs.

The review application is dismissed, with no order as to costs.

__________________

Fourie, AJ

Acting Judge of the Labour Court of South Africa

Appearances:

For the applicant: Ms S Greef, Solidarity

For the third respondent: Attorney B Bleazard

1 2003 (1) SA 11 (SCA) at para 5:

2Sidumo and Another v Rustenburg Platinum Mines Ltd and Others (2007) 28 ILJ 2405 (CC)

3National Union of Mineworkers v Samancor Ltd (625/10) [2011] ZASCA 74 (25 May 2011) para 5.

4 [2011] 2 BLLR 129 (LAC) at para 18.

5701/2012 [2013] ZASCA 97 (5 September 2013) at para.25.

6Case JR206/09, Labour Court, 25 April 2012

7Dumani v Nair and Another 2013 (2) SA 274 (SCA) at para 29-33