The Constitutional Court and "representivity" in Barnard

In the light of the central and recurring importance of the concept of (demographic) representivity in the practical experience of application of the constitutional principle of affirmative measures within the equality discourse, it was disappointing to read the judgments in Barnard (CC), and to note what I will describe as the very haphazard and superficial treatment of it by the various judges. This is surprising. As I will note later, it is clear that nearly all of the judges in this matter ultimately rejected Ms Barnard's complaints on the basis of the apparent justification for the SAPS affirmative action measure – the pursuit of representivity. For now, though, I will just include a brief montage of the few instances where the judges referred to representivity, and I will then briefly examine whether we can find any real guidance on the meaning and role of this concept within the continuing affirmative action debate.

One possible reason for the rather disappointing coverage of the concept of representivity in Barnard (CC) may be due to the complainant herself (or her legal team). Moseneke ACJ tells us that, in oral argument, Ms Barnard "jettisoned her detailed attack against the Employment Equity Plan and the Instruction as unjustifiable infringements of her equality protection because they amounted to racial quotas or racial norming or racial profiling".⁶⁴ However, in the light of the importance of the concept (and how the Employment Equity Act fits within the broader ideology of the demographic transformation of the public service and our other workplaces), I would suggest that the court was morally bound to consider this much more meaningfully and in more detail.

Be that as it may, the first real encounter we have with the role of representivity in Barnard (CC) is found in paragraph 66 of Moseneke ACJ's majority opinion, although the learned judge's views on its role are less than clear. He tells us that the employment equity plan in this case obliged the National Commissioner to take steps to achieve the targets set, and that the Commissioner was "within his right and indeed duty to take steps that would achieve the set targets".⁶⁵ While the judge tells us that "the implementation of a valid plan may amount to job reservation if applied too rigidly",⁶⁶ he was of the opinion that this was not the case here, for the following reason:

[O]ver-representation of white women at salary level 9 was indeed pronounced. That plainly meant that the Police Service had not pursued racial targets at the expense of other relevant considerations. It had appointed white female employees despite equity targets. Had the Police Service not done so, white female employees would not have been predominant in any of the levels including salary level 9 nor would they have been able to retain their posts.⁶⁷

This is interesting, in two ways. In the first place, Moseneke ACJ implicitly endorses demographic representivity in the context of target – setting (a specific group can only be "over-represented", and such "over-representation" can be "pronounced" only if measured against some standard – and demographic representation in terms of the population demographics was the express standard employed by the SAPS). In the second place, the judge's remarks appear to provide some distance between the "equity targets" (an affirmative action measure) and past unfair discrimination as a cause for any such "over-representation". He tells us that SAPS has appointed white female applicants despite such targets (ie he is not concerned that their over-representation is due to their past undue preference in employment, but rather he observes that even after the implementation of an affirmative action measure they were still appointed). On the role or importance of representivity, however, we find nothing more in the majority opinion in Barnard (CC).

Van der Westhuizen J came teasingly close to providing us with a clearer picture of the meaning and the proper role of representivity, but this also (I would submit) eventually came to nought:

Before focusing specifically on the facts of this case, it must be pointed out that equality can certainly mean more than representivity. Affirmative measures seek to address the fact that some candidates were not afforded the same opportunities as their peers, because of past unfair discrimination on various grounds. By focusing on representivity only, a measure's implementation may thwart other equality concerns. For example, if a population group makes up 2 or 3 percent of the national demographic, then, in an environment with few employees, the numerical target for the group would be very small or even non-existent. If a candidate from this group is not appointed because the small target has already been met, this may unjustly ignore the hardships and disadvantage faced by the candidate or category of persons, not to mention the candidate's possible qualifications, experience and ability.

And:

"Although equality can manifest in various forms, in the context of this case it takes the form of representivity. By appointing Ms Barnard, her designated group would have been significantly over-represented and her appointment would have aggravated racial inequality."⁶⁸

Also, the judge's example highlights the clear disconnect between representivity based on population demographics and the recognition of past disadvantage as well as the pursuit of (substantive) equality. There is, as per his example, not necessarily any correlation between the two, and I will suggest that in a scenario such as that referred to in the example the equating of representivity and demographics to (assumptions of) disadvantage would not only serve to further disadvantage the relevant member of a designated group (the point the judge seems to be making), but it would also allow an employer to ignore true (even proven) disadvantage and to refrain from pursuing affirmative action as a constitutionally-mandated form of remedial or restitutionary action in favour of the relevant group/member of a group. Surely this kind of outcome – compare what happened in Naidoo - must be absurd (perverse) when considered against the backdrop of the provisions of the equality guarantee, and the necessity for the pursuit of substantive as opposed to formal equality prized so highly by our Constitutional Court. In paragraph 183 of the judgment, van der Westhuizen J again confuses the reader with the statement that "[t[]he goal of equality is being promoted in this case through representivity". In the context of his (disappointing, as observed earlier) consideration of the potential impact of the measure on service delivery, the judge then also refers to "[w]hen a balance does have to be struck between efficient service delivery and equality in the form of representivity …"⁷⁰ – again, is he saying that representivity is a form of equality? To just confuse things even more, it would seem, he then continues to note that in Ms Barnard's case "any possible negative impact on service delivery was overshadowed by the fact that her appointment would have significantly aggravated unequal representation at salary level 9".⁷¹ I will be generous and attribute this use of the word "unequal" to a typo; if not, what would this terminology imply? "Equal" representation would require all race groups (to just use race as an example) to be, well, equally represented in the workplace.⁷² This again would be something very different from what the EEA requires.

We also do not find much guidance in Barnard (CC) about the meaning, role and importance of representivity (or its interaction with equity and equality) amongst the other judgments of the court. Cameron J, Froneman J and Majiedt AJ, in their separate concurring judgment with that of Moseneke ACJ, had the following to say:

If a decision-maker does not justify how he or she balances the important considerations of representivity and service delivery, remedial measures will suffer an invidious gloss. A decision-maker could prize representivity over service delivery without sufficient regard to the specific facts of a case. This would suggest that representivity is always more important than the quality of service provided by a public body. But this is a false choice. There is no evidence that we must sacrifice the quality of our public bodies to achieve the important goals of representivity and to redress past disadvantage. [My emphasis]⁷³

Apart from telling us that representivity is both an important "consideration" and "goal", what we find here is an indication that it is something different from redress. The learned judges are at pains to refer to the "important goals of representivity and to redress past disadvantage". So, upon my reading, these are two separate and different concepts. Of course, common sense and a dictionary also tell us that these are not one and the same thing (and, as mentioned, we see them also textually separated or distinguished in section 2 of the EEA). Finally, Cameron J et al also seem to imply that there is some link between representivity and the redress of past disadvantage, but just like so many others (including the drafters of plans and policies on public service transformation) they seem to base this on some unspoken assumption. I would suggest that this is evident in their following observation:

[The National Commissioner] was entitled to prefer racial representivity over gender representivity, provided he had a justification for that decision. In other words, it is not necessarily an injury to dignity to view a person only through the lens of one ground listed in section 9(3), provided the reason for doing so is to redress historical inequality.⁷⁴

The underlying assumption (that preferring one form or manifestation of representivity over another requires justification, and that such justification would/must be the redress of past disadvantage) is clear; whether there is any real basis for it is less clear. And that is not really the point. The point is that these judges owe us (and Ms Barnard) more of an explanation, and less in the way of having to draw inferences. So, from all of this the most I can say is that Barnard (CC) tells us the following:

1. 
   equality does not equal representivity, and these two concepts are simply not the same thing;
   
2. 
   even though that may be the case, representivity may be a form of equality; and
   
3. 
   while representivity is an important goal and consideration, it is something distinct from the redress of past disadvantage.⁷⁵
   

By now it must be abundantly clear that I have a rather large bee in my bonnet over representivity. Others may argue, however, that this lack of clarity (could it be dissembling) by judges is all fine and well; that the Constitutional Court was not called upon to provide us with any definitive guidance on the meaning, etc of (demographic) representation in the context of our equality discourse. They may argue that this is much the same as the issue of the meaning and role of quotas (which Moseneke ACJ told us, expressly, did not require examination in Barnard – of course, I do not agree). But they would miss the point. What was said about representivity by the various judges can hardly be classified as being obiter, if we consider how the separate concurring judgments all seemed to, ultimately, base their approval of SAPS's treatment of the complainant in this matter on its pursuit of (demographic) representivity in its workforce.

A case in point, and one issue that stands out for me, is that even van der Westhuizen J, who arguably (and as noted above) was the single member of this court most willing to consider the impact of the measure on those disadvantaged by it, relied on the promotion of representivity to justify the potential impact of the implementation of the measure on Ms Barnard's dignity. The judge formulated two questions (or factors) to assess the impact on the complainant within this context, and then tells us that neither of the two negative impacts hinted at by these questions had actually materialised – importantly, note that the reasoning here relies almost exclusively on the promotion of representivity and/or the role that the concept played in the design and implementation of the SAPS measure at issue in this case:

Was the impact on Ms Barnard's dignity reasonable and justifiable in light of the goal of substantive equality? I consider two factors. First, she treated [sic] as a mere means to achieve an end? Did the decision reduce her to a member of an underclass to the extent that her place in society and in the Constitution is denigrated? Even the perception of this may threaten the pursuit of our constitutional goal of a society in which everyone, regardless of their differences, is equally valued and at home. Second, does the measure's implementation amount to an absolute barrier to her advancement? If a measure is used to obliterate a person's chances at progressing in her chosen career, it would not pass constitutional muster. It would constitute an impermissible barrier to an individual's ability to "develop [her] humanity [and] 'humanness' to the full extent of its potential". The Act indicates a cognisance of the dangers of establishing "an absolute barrier to the prospective or continued employment or advancement of people".

Neither is present in this case. Ms Barnard failed to secure appointment because there was over-representation of people from her designated group. Had this over-representation not been present, the policy would not be a bar – let alone an absolute one – to her (or any other similarly qualified white woman's) appointment. Ms Barnard's career advancement within SAPS was not destroyed. The Employment Equity Plan has specific targets for different occupational levels and is flexibly used to cater for over- and under-representation. This flexibility ensures that she can be promoted to a higher occupational level should representation targets allow. By the time the case reached this Court, she had been promoted, albeit to a different department. The goal of equality is being promoted in this case through representivity.' [My emphasis]⁷⁹

Is it just me, or is there something rather incongruous in this reasoning? Does this amount to saying the following: "There is a measure here that is being challenged on the basis that it does not allow for the appointment of someone who is white. Ms X was not appointed because she was white. This is not a problem, however, because if she had been black she would have been appointed."? I am probably wrong; van der Westhuizen J is an eminent jurist, and one of the judges of our highest court. But surely we deserve more clarity than this.⁸⁰

That being said, however, and without wanting to sound overly simplistic in the analysis of this judgment, van der Westhuizen J spent fifteen paragraphs discussing the role of dignity in this case, and then found that it was trumped by representivity. And yet, as said earlier, the learned judge does not spend even one paragraph explaining to us the meaning and relevance of representivity in the application of an affirmative action measure in order to promote substantive equality. And this, to me, is especially strange when one considers the central importance that he attached to the balancing exercise that is inherent in constitutional adjudication (and which, in fact, forms the basis for his suggested standard of the review of affirmative action measures which should include the consideration of proportionality):

No [constitutional] provisions may be interpreted in isolation and no right protected and enforced without regard to other rights. Especially the exercise of one constitutional right may often have to be balanced against another. Courts are regularly called upon to do so thoughtfully and candidly. To a considerable extent, this is what constitutional adjudication is about.⁸¹

Yet we do not find an explanation of the meaning, content and proper (constitutional) role of the central concept or interest which the judge feels trumped Ms Barnard's right to dignity in this case. As a reader of the judgment I do not find it more helpful than the extremely one-sided judgment of Mlambo J in Barnard (LAC), which, as already noted, Malan has characterised as "nothing but a dreadful miscarriage of justice".⁸² The big difference, of course, is that even though we are no closer to discerning the meaning of representivity within the equality jurisprudence, the concept has apparently been given constitutional approval by the judges of our highest court. No future judge, outside of this court, who might feel more inclined to interrogate these issues, can now overturn this implicit approval of representivity as an apparent (although highly suspect) new constitutional goal. The drafters of the EEA (and the SAPS and doubtless a large number of other human resource managers in the public service) must be delighted. I am less so.

The vagueness of the language and reasoning of these judges is sorely problematic, not only because of the lack of clarity that Barnard (CC) has provided us on all the above-mentioned issues and questions, but also because such uncertainty also leaves just too much room for dishonesty. We have seen that the Constitutional Court found that the SAPS measure did not constitute an absolute bar to the appointment of Ms Barnard or any other white female. It is interesting, considering this, that the reality seems to have been quite different. In a recent speech the CEO of Solidarity,⁸³ Dirk Hermann, recounted the following exchange that occurred in testimony before the Labour Court in Barnard (during the leading of evidence of a SAPS human resources professional, Superintendent Ramathoka):

According to Ramathoka, an interviewing panel was not allowed to recommend people from the non-designated group if that group happened to be overrepresented. People from that group could apply for a position, but they would not be recommended.

"What is the point in allowing them or inviting them to apply?" asked Advocate Grogan.

And then followed Ramathoka's astounding reply that, in terms of the Employment Equity Act, an absolute ban on applications was not allowed. Whites could therefore submit an application; however, strict enforcement of the SAPS's plan meant they would not be recommended.

"That's cynical," Grogan said.

The judge was also confused: "Sorry, I just need to clarify something. Superintendent, I am confused. Are you testifying that the policy and process of the SAPS, once a plan has been issued which shows an overrepresentation in a particular position, is an absolute rule? That the SAPS will not, and may not, appoint a person from a non-designated group to that position, as an absolute rule?"

"It is an absolute rule, my lord," Ramathoka replied.⁸⁴

It appears – and this is worrying - that the judges of our Constitutional Court may be easily fooled. It might be apt to quote the words of American economist Thomas Sowell – referring to affirmative action for African-Americans as "the grand fraud". This was expressed, of course, in the American context, but see if you can spot its relevance closer to home:

No issue has been more saturated with dishonesty than the issue of racial quotas and preferences … Many defenders of affirmative action are not even honest enough to admit that they are talking about quotas and preferences, even though everyone knows that that is what affirmative action amounts to in practice … When any policy can only be defended by lies and duplicity, there is something fundamentally wrong with that policy. Virtually every argument in favor of affirmative action is demonstrably false. It is the grand fraud of our time … Affirmative action is great for black millionaires but it has done little or nothing for most people in the ghetto. Most minority business owners who get preferences in government contracts have net worth's of more than one million dollars. One of the big barriers to any rational discussion of affirmative action is that many of those who are for or against it are for or against the theory or the rationales behind group preferences and quotas. As for facts, the defenders simply lie.⁸⁵