---

REPUBLIC OF SOUTH AFRICA

THE LABOUR COURT OF SOUTH AFRICA, DURBAN

JUDGMENT

Reportable

Case no: D253-03

In the matter between:

CAPTAIN M MUNSAMY ...........................................................................................Applicant

and

THE MINISTER OF SAFETY AND SECURITY ...........................................First Respondent

THE SOUTH AFRICAN POLICE SERVICES .......................................Second Respondent

Heard: 25 to 28 June 2012

Receipt of written closing submissions: January 2013

Delivered: 3April 2013

Summary: In the absence of a rational coherent employment equity plan permitting such, an employer may not prefer one group of designated employees over another who are supposedly over-represented. Discrimination based on race unfair.Requirements for rational employment equity plan.

JUDGMENT

WHITCHER, AJ

Introduction

1. In early 2000, 195 vacant promotional posts were made available in KwaZulu-Natal at the level of Superintendent to bring the total number to 479. The applicantapplied for posts459, 463 and 493.

2. The applicant was recommended for appointment to post 459 but was not appointed on the basis that Indian males were over-represented and Africans were under-represented at the level of Superintendent. The second respondent [“the respondent”] pleaded that the applicant lacked relevant experiencefor post 463 and that the successful candidate, Ms Drotsky, a white female, was appointed on the basis of ‘representivity’ and that the candidate with the highest points, B Gumbi, was appointed to post 493.

3. The issuesare whether the respondent’s decision which was based on affirmative action was in line with a defensible employment equity plan, and, as such, constituted fair discrimination against the applicant and whether the respondent had a fair reason for not appointing the applicant to post 493.

General background facts

4. On 27 June 1997, in a collective agreement on employment equity, the respondent and organised labour, SAPU, POPCRU, PSA and NPSU, agreed that the respondent shall strive to attain a minimum of50% Black people and 30% women at management levels by the year 2000. In order to meet theobjectives of the policy, the respondent was to use the population distribution based on the 1996 National Population Census by the year 2005. The parties agreed that an employment equity plan would be drafted in accordance with the LRA¹and that ‘a mandated committee shall be constituted by the National Negotiation Forum’ for this purpose.

5. In early 2000, the promotion process in question began. On 13 March 2000, the national office directed that the criteria for selection must include ‘representivity’ and that numerical targets of 70% black, 30% white and 50% women must be achieved as far as possible, taking into consideration the composition of the provinces.

6. A provincial selection panel in KwaZulu-Natal considered the applications and made recommendations to a national selection panel. Each candidate, in addition to being scored in relation to his or her own apparent expertise and experience, was awarded a score for representivity. Indian males and white males were given 5.5, whilst African males were given a score of 8.

7. The applicant was shortlisted for posts 459 and 493, but not for post 463. The provincial panel recommended that he be appointed to post 459.The province’s recommendationsand a summary of theirnumerical achievements were transmitted to the national office on 27 June 2000.

8. Their recommendations achieved numerical targets of 73,85%black candidates and 26,15% white candidates in respect of the 195 posts. This likewise achieved the 50%/50% ratio set out in the 1997 agreement and the 70%/30% ratio set out in the March 2000 directive.

9. Therecommendations werenot approved by the national panel. The panel, which included two representatives each from SAPU andPOPCRU,decided on 13 July 2000 that since the respondent had submitted an Interim Employment Equity plan to the Department of Labour on 1 June 2000, the provinces must review their recommendationsin accordance with the planand strive to achieve targets of 70% Blacks (African, Coloured and Indians), 30% Whites and 30% females and to improve representivity amongst black males and females and white females. The demographics provided by the 1996 Census were to be used as guidelines in this regard.

10. The directive to KZN had an attachmentheaded ‘Numerical Goals – KwaZulu-Natal Demographics’.The document was authored by the respondent’s Employment Equity Division. The goals were extrapolated utilising the 1996 National Statistics and applying such statistics to the respondent’s apparently existing workforce.

11. The manner in which such demographics were purportedly applied was as follows:

1. The number of promotional posts (195) was initially divided, in terms of race, by the 70% designated / 30% non-designated ratio.

2. Thereafter, the national statistics purportedly derived from the 1996 census were applied to the 70% of the posts which were benchmarked for black people, and within that each of the racial groups were allotted the number of posts which accorded with the national demographics thereby achieving a total number of posts made available per racial group.

12. The numerical goals reflected that Africans were to be awarded 192 posts and the Indian male incumbency was to be reduced to a total of 17 and the female incumbency, 7.

13. In consequence of the directive from the national panel, the KZN provincial panel set about realigning its initial recommendations to the figures contained in the attached ‘Numerical Goals – KwaZulu-Natal Demographics’ document.

14. Its decision to recommend the applicant to post 459 was substituted with a recommendation that one Captain Zakwe be promoted in his stead. Ms Drotsky, a white female, who, like the applicant, had not been shortlisted for post 463, was recommended for the post in the review. Zakwe and Drotsky were appointed to the posts.

15. The applicant testified that he had at the time over 25 years of service and had performed some of the most important functions attached to post 459, including the inspections of 25 stations, which Zakwe had not performed. In response to a respondent witness’s suggestion, he said he should not be forced to relocate anywhere else in the country in order to obtain a promotion. He was born and bred in KwaZulu – Natal and a move outside the province would necessitate him leaving his home, his extended family and having to incur the expense of accommodation in another province.

16. According to the submitted Interim Employment Equity Plan and Report–

1. The respondent had consulted all organised labour, employees who are not represented by unions and employees from designated and non-designated groups butthe extent of the consultations had been very limited.²

2. Each province had their own plans because it was impractical to develop a single plan that will make provision for the exclusive circumstances in the provinces.

3. The executing authority and management of the respondent had set the numeric targets set out in the plan.

4. Overall targets for the duration of the plan in contrast to targets for recruitment for the duration of the plan were set.

5. For ‘overall targets’ at management level, targets were set at 50% designated/50% non-designated for the year 2000; 60%/40% for the year 2001. For other levels, the targets were 50%/50% for the year 2000; 70%/30% for the year 2001.

6. For ‘recruitment at all levels for the duration of the plan’ the targets were 70% designated / 30% non-designated and 70% male / 30% female.

7. In a separate section, the plan set out the overall number of whites [male and female], Indians [male and female], Coloureds [male and female] and Africans [male and female] it intended to have in the whole of the respondent by the end of December 2004.

17. It is noted here that the plan was submitted on 1 June 2000 to the Department of Labour, but on 29 May 2000 the Provincial Commissioner of SAPS: KwaZulu-Natal sent a document headed ‘Employment Equity Plan: KwaZulu-Natal’ to the head office of the respondent. According to the document, as at 29 May 2000, the respondent in KZN had not conducted a workplace audit, prepared draft numerical targets, set up a consultative forum or consulted with any role players on these issues. They hoped to complete these processes by the end of 2000.

The law

18. Employers are obliged to make the workplace equitably representative and may use discriminatory affirmative action measures to do so.³An employer may, however, not prefer one group of designated employees over another group of designated employees who are supposedly over-represented in the absence of proper proof of such representativeness and a valid employment equity plan which permits the action of the employer. The LAC recently affirmed this rule in SAPS v Solidarity on behalf of Barnard ⁴when it held that the failure by the SAPS to appoint a recommended white female candidate did not constitute unfair discrimination where white females were over-represented in the level of the advertised post andthe failure to appoint was in line with a rational, coherent employment equity plan intended to redress inequitable representation in the workplace.⁵

19. The issue is what constitutes a rational coherent employment equity plan and action in line with such a plan.

20. The answer is to be found in the provisions of the Employment Equity Act, 1998 and the Code gazetted to regulate the preparation and implementation of employment equity plans, as set out below. The Act and Code were in operation at the time of the promotion process in question.

21. In terms of section 15 (3) of the Act, legitimate affirmative action measures include preferential treatment and numerical goals, but exclude quotas. The concept of preferential treatment and numerical goals to be achieved within a certain period of time differs quite significantly from the concept of quotas. The imposition of a strict quota is a rigid measure requiring a certain fixed proportion or percentage to be included whereas preferential treatment and goals is more flexible allowing the achievement of objectives over a period of time.

22. Subject to section 42, nothing in section 15 of the Act requires a designated employer to establish an absolute barrier to the prospective or continued employment or advancement of people who are not from the designated groups or who are from a designated group that happens to be over-represented in the relevant occupational level.

23. In terms of section 16 of the Act, the employer is obliged to take reasonable steps to consult with all designated and non-designated groups on the preparation and implementation of an employment equity plan. The parties invited to consult, and thus the consultative forum/s, should reflect the interests of all employees. The Guideline speaks of a forum being broadly representative of allinterests.⁶ For instance, if the employer chooses to consult with or invites only certain unions and the membership of these unions, collectively, is mainly made up of one particular racial group, it logically follows that there has not been proper consultation. Similarly if the employer fails to meaningfully attempt to consult members of non-designated groups on the preparation and implementation of an employment equity plan, therehas been no proper compliance with the Act and Guidelines.

24. The employer is only obliged to provide a proper opportunity for consultation. It cannot be held against the employer if any particular interest group failed to take up the opportunity of consultation when invited to do so. Guidelines of the Code provide that when a representative union or body refuses or fails to take part in the consultation process, the employer must record the circumstances in writing.

25. The consultations need not result in agreement on the issue for consultation and it remains the employer’s responsibility in terms of the Act to proceed with implementing employment equity.

26. According to guideline 8.4.2, in developing the numerical goals, the following factors, inter alia, must be taken into account-

1. the degree of under-representation of employees from the designated groups in the different occupational levels;

2. present and planned vacancies;

3. the demographic profile of the national and regional economically active population;

4. the pool of suitably qualified people from designated groups from which the employer may reasonably be expected to promote or appoint employees; and

5. the number of present and planned vacancies that exist in the various categories and levels, and the employer’s labour turn-over.

27. These factors are also taken into account when the department of labour is determining whether the employer is properly implementing equity.⁷

28. According to section 17, and, implicit in sections 17, 20 and 42 of the Act and guidelines 8.4.2 of the Code, the issues for consultation must include the workforce profile analysis, including the relevant information and demographics and the extent of under-representation of employees from the designated groups in the different occupational categories and levels. It must also address what sort of affirmative action measures are to be applied, including demographic guidelines and goals and whether, in these respects, national and/or regional demographics will be used, time periods and what constitutes reasonable progress over the duration of the plan.

29. Based on the above provisions, where an employer used affirmative action measures to prefer one designated group over another who were supposedly over-represented, the employer must prove the following toestablish that its conduct was in line with a defensible employment equity plan:(i) that there was an over-representation of the discriminated against group and an under-representation of the preferred group in the level of the post in question: this requires the conduct of a proper workplace profile audit; (ii) that the measure is sufficiently coherent and not open to arbitrary application or abuse; (iii) that the measure is permitted by the Act; (iv) an equity plan that permits the disputed measure, either expressly or by clear implication; (v) that the measure is intended to correct inequitable representation in the workplace; and, (vi) that the measure arose out of proper consultations, i.e. therehad been proper consultation on the particular measure.

30. Whether a court may be called upon to adjudicate the fairness, in the wide sense of the word, of ‘measures’, especially where they havebeen the subject of proper consultation and agreement between management and labour in the consultative forum and where the aggrieved parties have the right to advance their interests through political and industrial action, is debatable but as will be seen is not an issue that needs to be determined in this case. It is not necessary for the resolution of this dispute.

Analysis of the case

Measures not in line with an employment equity plan

31. In the first phase of the process, the assessment of the applicant included the attribution of a score for ‘representativeness’. Indian males, as members of the designated group, were attributed the same scores as white males, the non-designated group. The provincial panel was not furnished with any demographics by the national office until 19 July 2000, some months after the initial allocation of points.

32. When canvassed with them, not one of the respondent’s witnesses were able to point to any equity plan which permitted such a point system or provide any explanation as to how such a point system was established, or what informed the actual allocation of specific points to the individual race groups.This system was thus arbitrary and lacked any rational basis.

33. Notwithstanding the arbitrary and irrational allocation of points, the applicant was recommended for promotion to post 459. He was, therefore, not materially prejudiced by the system. The absence of material prejudicedoes not detract from a finding of unfair discriminationbecause, while prejudice is highly relevant, the main mischief guarded against in the right not to be unfairly discriminated and solatiumawarded in unfair discrimination cases is the infringement of the right to dignity.The central role of dignity as a substantive test of process in unfair discrimination is relevant. The constitutional court has emphasised that a person’s constitutional right to dignity is infringed where he or she is unfairly discriminated against on the basis of race or gender.⁸In the circumstances, the use of such an arbitrarymeasure against the applicant constituted unfair discrimination.

34. The main measure that led to the de-selection of the applicant was the document headed ‘Numerical Goals – KwaZulu-Natal Demographics’ authored by the respondent’s Employment Equity Division.

35. In the document, the national populationdemographics, based on the 1996 census, were applied to 70% of the posts which were benchmarked for black people, and within that, each of the designated groups were allotted the number of posts which accorded with the national demographics thereby achieving a total number of posts available per racial group.Fixed quotas for the promotion process were essentially established.The numerical goals reflected that Africans were to be awarded 192 posts. As there were only 195 posts available for promotional purposes, this target reflects that almost 100% of the proposed posts were to be given to African applicants.

36. Prior to the promotion process, 95 Indian males occupied positions at salary level 9 – 10, whilst 11 Indian females did likewise. In terms of the respondent’s proposal embodied in the Numerical Goals document, the Indian male incumbency was theoretically to be reduced to a total of 17 and the female incumbency, 7.

37. To achieve this, if the respondent had followed the Numerical Goals document strictly, it would have been required that no single Indian person was to be promoted.

38. The approach also, potentially, has the result that Indian members of SAPS in KwaZulu-Natal and Coloured members in the Northern and Western Cape would thereby have to undertake provincial migration in order to either become employed or be promoted; an issue that was raised by the applicant.

39. In terms of the 70%/30% measure, 30% of the posts remained reserved for whites, while the partial application of national demographics, more particularly national population demographics, only to the designated group, led to 3% being reserved for Indians, a designated group.

40. According to the respondent’s witnesses, the creation and application of the numerical goals and measures described above and used to de-select the applicant were purportedly based on the Interim Equity Plan, the 1997 collective agreement and agreement between management and labour at the national panel meeting in July 2000. Moreover, the plan was derived from proper consultation with its employees.

41. The respondent’s witnesses were unable to direct this Court to any evidence, other than their say so and the statements made in the plan, that the Interim Plan had been a product of proper consultations. It is unthinkable that no corroborating documentary evidence would be available to sustain such a claim.

42. Their claim is directly contradicted by the letter of 29 May 2000 from the KZN province to the national office. This letter makes it clear that, as at the 29 May 2000, one day before the Interim Plan was submitted to the Department of Labour, no reliable workforce or demographic profile of the province had been prepared or consulted on and that there had been no consultation on affirmative action measures and numeric goals relevant to the province. It was only by the end of the year 2000 that the province anticipated that these issues would have been compiled and consulted on.

43. Furthermore, according to the Interim Employment Equity Plan Report, each province was supposed to prepare its own plan, ‘because it was impractical to develop a single plan that will make provision for the exclusive circumstances in the provinces.’

44. In the report, the extent of consultations is described in the form of various percentages, mostly all below 20%, but no evidence was led as to what informed these figures and what they actually meant.

45. The 1997 collective agreement clearly did not constitute consultation on the Interim Equity Plan. According to the agreement, the parties still intended to consult on and prepare an employment equity plan in accordance with the Employment Equity Act.

46. The agreement between management and two representatives each from POPCRU and SAPU during the meeting of the national panel in July 2000 to assess the appointment recommendations from the provinces does not constitute consultation contemplated by the Act for the purposes of preparing an employment equity plan.

47. Even if it is accepted that the Interim Equity Plan had been the product of a proper consultation process, the respondent was unable to direct this Court to any part of the plan which mentioned and entitled the respondent to use and apply the affirmative measures described above in the promotion process, namely to establish quotas, to use numerical targets based on national population demographics in a province, to apply demographics or national demographics in the awarding of posts to the designated group and to apply demographics or national demographics in the awarding of posts to the designated group in circumstances in which 30% of the posts remained reserved for whites while only approximately 3% remained reserved for Indians who are designated as beneficiaries in the Employment Equity Act.

48. The respondent submitted that these imperatives are contained in the 1997 collective agreement because it states that:

‘In order to meet the objectives of this document, the Service shall strive to attain representativeness which reflects the population distribution based on the 1996 National Population Census, by the year 2005.’

49. The assertion is unsustainable. This general statement in the agreement cannot be interpreted to contain the specific measures in question. Fairness and the principle of legality require that such drastic measures should be expressly and clearly stated in an equity plan.

50. The agreement also stated that the target to be achieved by 2000 was 50% Black / 50% White people and that ‘representivity means that all racial groups of the South African community shall be broadly mirrored in the various activities of the Service while taking into consideration gender and disability.’

51. The assertion contradicts the national panel’s directive to the provincial panel in March 2000 to take into accountthe ‘composition of the provinces’ in the promotion process.

52. The agreement was also not drawn up in terms of the Employment Equity Act. The parties stated in the agreement that a proper equity plan was still be prepared in accordance with the relevant law.

53. In any event, the respondent put up the June 2000 plan as their Employment Equity Plan. The national panel in their meeting of July 2000 stated this plan must be used in promotions.Any independent agreement at this meeting on numerical goals and affirmative action measures is irrelevant. The purpose of the meeting was to assess promotion recommendations and an agreement in this context between two representatives each from two of the many unions in operation at the respondent does not amount to a consulted plan contemplated by the Employment Equity Act.

54. In the circumstances, the respondent has failed to prove that the discrimination against the applicant was in line with an employment equity plan that had been the subject of proper consultation and that the measures relied upon to de-select the applicant were permitted by theemployment equity plan relied upon.

55. Moreover, the application of national demographics, only to the designated group, in circumstances in which 30% of the posts remained reserved for white people while less than 3% were reserved for Indians, a designated group, logically lacked fundamental rationality.

56. It would be absurd to expect an employer to produce perfectly accurate workplace profile audits and resultant numerical targets. However, the extent of the discrepancies as shown by the applicant and set out below, brings into question the coherency and rationality of the numerical targets relied upon by the respondent to de-select the applicant.

57. The Employment Equity report submitted by the KwaZulu-Natal province on 29 May 2000 makes it clear that when the respondent submitted its equity plan to the department of labour on 1 June 2000, the province had not at that time completed its workforce profile and the persal system used to calculate numeric goals was unreliable. The province stated that it was therefore unable to set objective and rational numeric targets. It was only by the end of the year 2000 that KwaZulu-Natal anticipated that it would have compiled reliable statistics and a demographic profile of its workforce. Also to be completed only by the end of that year was the determination of numeric goals and strategies to address under-representivity as identified in the analysis which was still to be completed.

58. The respondent stated that the numerical targets were to be based on the National Demographics of the Economically Active Population of South Africa but the demographics of the general population was used. The respondent’s testimony that the statistics for the economically active population was the same as the general population is clearly incorrect, logically, and as was demonstrated by documentary evidence of the 1996 Census handed in by the applicant.

59. The document produced by the respondent shows that in 1996, the economically active Indian population comprised 9.2% of the population of KwaZulu-Natal, while the document produced by the applicant established that it was in fact 15.29%.

60. There were 195 promotion posts to be filled. 30% were to be reserved for the White population group. 30% of 195 posts is 58.5 posts. In the numerical goals document, a total of 63 posts were to be allocated to White people. This represents 32% of the available posts, being 2% in excess of the guideline.

61. As per the national demographics which were allegedly relied upon, the percentage which ought to have been applied to each of the designated groups was as follows:

• African – 76.7%

• Coloured – 8.9%

• Indian – 2.6%

62. These percentages, applied to the promotional posts which remained available to the designated groups, result in the following:

• African – 101.7 posts

• Coloured – 11.8 posts

• Indian – 3.5 posts

63. As it was, the numerical goals reflected that Africans were to be awarded 192 posts (being 131 posts to be allocated to males (proposed 145 less the existing 14) added to 61 posts to be allocated to females (proposed 62 less the existing 1). This represents 90 posts more than ought to have been allowed for in terms of the National Demographics which were allegedly applied.

64. As to the posts which ought to have been allocated to the Coloured population, which ought to have been 12 on the above calculations, only 4 were allocated in terms of the numerical goals.

65. It is not possible to say that, but for these errors or discrepancies, the applicant would not have been de-selected. The respondent contended that since it was not disputed that Indian males were in any event over-represented at the relevant level in KZN and furthermore that his non-appointment to post 459 was in order to redress under-representation of African males at the relevant level, the respondent did not unfairly discriminate against the applicant. In this regard, the respondent referred to section 195 of the Constitution which provides inter alia that public administration must be broadly representative of the South African people.

66. Clearly, proof of over-representation and an intention to correct inequitable representation is not the primary determinative in cases of this kind as established by the SCA in Gordon v Department of Health: KwaZulu-Natal⁹and the test set out in paragraph [29] above. The case referred to by the respondent, SAPS v Solidarity on behalf of Barnard (supra) re-affirmed this as well. According to the respondent’s own reading of the case, the LAC held that the failure by the SAPS to appoint a recommended white female candidate did not constitute unfair discrimination where white females were over-represented in the level of the advertised post andthe failure to appoint was in line with a rational, coherent employment equity plan intended to redress inequitable representation in the workplace.

67. Aside from the fundamental inapplicability of the affirmative action measures applied against the applicant, the applicant submitted that the appointment of an African male in post 493 was irrational because it contradicted the respondent’s reason for failing to appoint the applicant, namely to facilitate a broadly representative workforce at all levels of the respondent. Save for one white male, the station was occupied entirely by African males and females. The applicant contended that in light of the objective of the respondent, there was no reason why a suitably qualified candidate, such as the applicant who had already been recommended for appointment, could not have been placed in the post.

68. This submission goes to the rationality of the use and application of national demographics per se. Whether the courts may adjudicate such a complaint is debatable, but it is not necessary to decide this matter since this case has been resolved on other more fundamental grounds.

Conclusion

69. The application of affirmative action measures which resulted in the applicant being denied promotion to post 459 was not in line with a defensible employment equity plan and as such the conduct of the respondent against the applicant was unfairly discriminatory.

Post 493

70. I accept the respondent’s submission that the applicant failed to show that his non-appointment to this post was unfair.

71. The applicant was ranked fourth out of four candidates that were shortlisted. The successful candidate, Gumbi, scored 64.31 and the applicant scored 59.28. The respondent established that, even if the discriminatory points given for representivityis excluded from the scoring,Gumbi’s score remained higher than the applicants’, even if marginally so. There can therefore be no suggestion that Gumbi’s appointment could be regarded as one based on race. Moreover, during cross-examination, the applicant disavowed any allegation that Gumbi was not a suitable candidate.

Appropriate relief

72. The applicant achieved the rank of Major on 1 May 2011. He is seeking to be remunerated the difference between the salary he actually earned and the salary he would have earned had he been promoted on 1 July 2000, until 1 May 2011, which includes the difference in bonuses payable, as set out in the schedule he prepared. The amount in question is R333 421.00.In additionthe applicant seeks compensation in the maximum amount permissible in terms of section 194(4) of the LRA on the basis that, having been denied a promotion in 2000,he was prejudiced in that he was unable to apply to be promoted to any further ranks because he was unable to jump levels in the ranking structure when applying for promotion.

73. Regarding the first part of the relief sought, although the applicant was recommended for appointment by the provincial panel, the final decision fell with head office. It is, therefore, relevant whether he would most probably have been appointed if this recommendation had not been cancelled on the basis of unfair discrimination. This Court can safely find that the applicant would probably have been appointed because head office would have had to have a good reason to reject the recommendation. Since head office did not prove that it potentially had a good reason, outside of unfair discrimination, to reject the recommendation, the applicant is entitled to the relief sought.

74. I find no reason to grant the applicant compensation over and above of what he actually lost in terms of remuneration. I have taken into consideration that when the unfair conduct occurred, namely in the year 2000, the development of affirmative action policies in terms of the Act was relatively new and the respondent did not have the advantage of the guiding principles which have since been established by case law.

75. Order

1. The application of affirmative action measures by the second respondent which resulted in the applicant being denied promotion to post 459 was not in line with a defensible employment equity plan and as such the conduct of the respondent constituted unfair discrimination.

2. The second respondent is directed to pay the amount of R333 421.00 to the applicant, Mr Munsamy, within two months of the delivery of this judgment.

3. The second respondent is directed to pay the applicant’s costs in this suit.

__________________

Whitcher AJ

Acting Judge of the Labour Court

APPEARANCES:

For the Applicant: Advocate K Allen

Instructed by: Norton Rose, Durban

For theSecond Respondent: Advocate NH Maenetje SC

Instructed by: State Attorneys, Durban

---