Affirmative Action and UCT – the debate

23 April 2007

Professor David Benatar unleashed a storm of debate at UCT with his inaugural lecture on Justice, Diversity and Affirmative Action, delivered on 11 April. The discussion fanned out beyond UCT the following day when the Cape Times ran an article in which Benatar outlined part of his arguments on why he believes affirmative action does not succeed. And on 16 April, Benatar and deputy vice-chancellor Professor Martin Hall headlined a public and sometimes heated debate on the issue. That debate forms part of UCT's Respect campaign, launched recently. Today, Monday Paper runs Benatar's piece as it appeared in the Cape Times, alongside a self-penned summary of Hall's response. In keeping with the traditions for debate, Monday Paper has agreed to allow Benatar a rebuttal to Hall's response.

"Affirmative action" not the way to tackle injustice

By Professor David Benatar

A partheid's legacy looms large over contemporary discussions of (racial) affirmative action in South Africa, rousing passions on both sides of the debate.

Proponents of affirmative action view opponents as, at best, insensitive to the enduring effects of generations of racism or, at worst, racists seeking to preserve their own privilege.

Opponents of affirmative action charge that affirmative action perpetuates two fundamental errors of apartheid - an insistence on classifying people by "race", and distributing some benefits (at least partially) on the basis of "race".

I agree that much opposition to affirmative action stems from sinister sources. Then again, much enthusiasm for affirmative action results from vested interest. However, there are also principled friends and principled foes of affirmative action.

The crucial question, therefore, is not whether there are knaves who share one's view about affirmative action, but rather whether one's view is supported by the best arguments. An ethical assessment of affirmative action is not possible without understanding what the policy is. Yet the term "affirmative action" is notoriously vague.

Not all the policies that it describes are controversial. In its mildest form, affirmative action involves taking positive steps to avoid discrimination, to ensure that opportunities are open and available to all and that fair standards of selection are used.

This form of affirmative action, which we might call "equal opportunity affirmative action", is a badly-needed corrective to unfairness, much of which is hidden rather than obvious. Equal opportunity affirmative action, unlike all other forms of affirmative action, does not require us to enter the fetid terrain of racial classification and racial preference. It seeks to undo unfairness whoever its victims might be.

Nobody committed to fairness could take exception to affirmative action in this form.

The controversy arises once affirmative action requires a system of racial classification and a mechanism for assigning individuals to racial categories, favouring some people merely on the basis of such assignments.

There are at least three forms of affirmative action that presuppose racial classification and racial preference. The mildest of these is what we can call "tie-breaker affirmative action".

According to such a policy, if two candidates are equally qualified, preference should be given to the one from a designated (racial) category. Here "race" is given some, but only a little, weight.

Greater weight is given to "race" by what I call "strong-preference affirmative action", a policy that allows preference to be given to a less qualified candidate from a designated racial group. This form of affirmative action comes in various strengths.

The mildest forms allow preference to be given only to marginally less qualified candidates, while the more severe forms allow preference to be given to significantly less qualified (or even unqualified) candidates.

Even more extreme than strong-preference affirmative action are "set-asides", where only people from designated categories are considered.

The more weight an affirmative action policy attaches to "race", the weaker a candidate of the desired "race" can be in other ways while still obtaining the place or position for which he or she is competing.

In saying this, I am not making a value judgement, but stating a scientific truth. The same would be true if one accorded special weight to eye or hair colour, height, age, national origin or any other such attribute. That is simply how weights and scales work.

Although it is a simple fact, it has implications for an ethical assessment of affirmative action.

The more weight an affirmative action policy attaches to "race" the less easily it can be justified. In other words, it is less implausible to attach minimal weight to "race" than it is to accord it significant weight.

In South Africa a range of weights are attached to "race". However, since at least some positions are set-aside for "blacks", affirmative action in this country sometimes attaches immense weight to "race".

Arguments for the controversial forms of affirmative action fall into one of two broad categories - arguments that affirmative action is necessary to rectify injustice, and arguments that affirmative action is necessary in order to produce some good consequence, usually by increasing diversity.

I shall focus here on arguments of the first sort, because they are the arguments most commonly invoked in South African defences of affirmative action.

The sentiment underlying rectification defences of affirmative action is entirely understandable. The massive injustices of the past, the effects of which are still manifesting, call out for rectification. The question, however, is whether affirmative action is either the effective or appropriate means of rectification that its defenders assume it to be. The answer, I'm afraid, is negative.

The central problem is that affirmative action (in its controversial forms) favours people of a designated "race" whereas rectification requires bestowing a compensatory benefit on those who have been treated unjustly.

Defenders of affirmative action assume that these categories are either co-extensive or at least sufficiently overlapping to treat "race" as a proxy for disadvantage. There are numerous problems with this assumption.

For example, the rectification imperative is strongest for those who are most disadvantaged by injustice. Yet these people are least likely to benefit from affirmative action.

They are so disadvantaged that they often cannot even begin to compete for the positions that beneficiaries of racial affirmative action obtain. Those most likely to benefit are those who were either minimally disadvantaged or not disadvantaged at all (in the relevant ways) or those who, although once disadvantaged, have already been compensated.

Consider, for example, affirm-ative action in academic staff appointments. Those deprived of adequate primary or secondary education certainly do not benefit from affirmative action in this sector.

Those "blacks" who do benefit are those who already have higher, especially doctoral, degrees. However, their having such degrees is an indication either that they were not (that) disadvantaged - for example, they may be the children of exiles and were educated at excellent universities abroad - or the opportunity to obtain a higher degree was itself a rectificatory intervention to compensate for what would otherwise have been the unfair deprivation of that opportunity.

Moreover, although moderate disadvantage is also deserving of rectification, there are some "whites" who have suffered such disadvantage. Given South Africa's past, we should expect that there are many more "black" victims of social injustice. However, if one's motivation is the rectification of injustice, then one should be concerned with rectifying all (equivalent) injustices irrespective of the "race" of the victim.

Why use "race" as a proxy for disadvantage when one can focus directly on disadvantage?

Using "race" as a proxy is a kind of racial profiling, which defenders of affirmative action typically take to be unacceptable in other contexts, such as blood donation.

The architects of affirmative action in South Africa have done their best to insulate it from legal challenge. Instead of allowing the courts to consider whether affirmative action laws and policies are compatible with the right to equality, judicial questioning of the matter has been forestalled by a constitutional provision permitting affirmative action.

Despite this, however, affirmative action, in its controversial racial preference forms, has an inherent weakness that is routinely overlooked.

It simply cannot be implemented unless there is a way to assign people to racial categories.

However, there is no clear criterion for making these classifications and the reason is not hard to fathom. Official state classification or a standard objective classificatory norm would be far too reminiscent of apartheid.

The alternative possibility is self-classification. Yet the self-classification criterion is not officially espoused for the obvious reason that deciding whether one will be the beneficiary of affirmative action cannot be left in the hands of those who will either receive or not receive the benefit depending on how they classify themselves.

Thus the whole affirmative action enterprise seems predicated on the citizenry's acquiescence to a self-classification that is in keeping with how they were or would have been classified by the apartheid state.

If citizens refused to self-classify in accordance with this principle and decided to classify themselves arbitrarily or subversively, or not at all, the government would either have to abandon racial preference or embrace the racial pseudo-science that underpinned apartheid.

The government would have to start judging, for example, whether the offspring of an "African" and a "white" is an "African", a "white" or a "coloured", irrespective of how that person identifies himself and irrespective of whether that person identifies himself racially at all.

This is a distasteful, messy business. Nor is it clear that this has anything to do with rectifying injustice. If the offspring of a "white" South African is presumed not to be disadvantaged, why, if he is also the offspring of a "black", is he considered (educationally) disadvantaged, and thus a fitting beneficiary of affirmative action?

Proponents and opponents of racial preference can agree that injustice must be rectified.

Once that is recognised, opponents of affirmative action cannot be dismissed as the moral cretins that defenders of affirmative action often allege them to be.

There are excellent reasons to be deeply suspicious of affirmative action, only some of which I have outlined here.

Rejecting this policy does not in the least diminish the duty to address injustice, either past or present. Indeed it might help focus the mind on more appropriate methods of rectification.

back to top

Creative Commons License This work is licensed under a Creative Commons Attribution-NoDerivatives 4.0 International License.

Please view the republishing articles page for more information.