Critics of amendments to equality and unfair discrimination act have a poor understanding of their meaning

23 June 2021 by Pierre de Vos

The Promotion of Equality and Prevention of Unfair Discrimination Act (Pepuda), passed in 2000, imposes two distinct obligations on the state and on private parties. 


First, it prohibits the state and private parties from unfairly discriminating against anyone and from subjecting individuals to hate speech. Second, it imposes obligations on the state and on private parties to promote the achievement of equality.

While the first part of the act came into effect in 2003, the second part, dealing with the promotion of equality, was never implemented.

The Amendment Bill proposes changes to both the first and second parts of Pepuda, raising hopes that the government may finally be ready to implement the second part of the act. 

As this part of the act imposes duties on private parties to dismantle the structural or systemic barriers faced by those who find themselves on the wrong side of any power hierarchy, it is not surprising that groups like the Institute of Race Relations (IRR), the National Employers Association of South Africa (Neasa), Freedom of Religion South Africa (Forsa), and political parties such as the Democratic Alliance (DA) and the African Christian Democratic Party (ACDP) have launched scathing attacks on the proposed amendments.

As far as the first part of Pepuda is concerned, the Amendment Bill expands the definition of “discrimination” to bring it in line with the equality jurisprudence of the Constitutional Court. It does so by making clear that an intention to discriminate is not a requirement to show that discrimination had occurred; and that discriminatory harm includes the causing of prejudice and the undermining of the dignity of an individual, if it is based on one or more of the prohibited grounds such as race, gender and sexual orientation.

It also, superfluously, inserts a provision that discrimination on a particular ground need not be “the sole or dominant reason for the discriminatory act or omission”. (The latter is superfluous as the reason for a discriminatory act or omission has never been relevant to determine whether discrimination, in fact, occurred.)

Critics of this amendment suggest it somehow broadens the scope of discrimination by eradicating the “intention requirement”, and by accepting that if a “person who alleges discrimination says they feel prejudiced, or that their feelings have been hurt, then that is enough to establish discrimination”. (The latter claim is made by For SA, probably part of its hustle for donations from that part of the population that believes critical race theory will destroy the world.) 

But these claims are obviously false and it is difficult to see how anyone with a glancing knowledge of the law could have honestly made them.

In our law, proof of intention has never been required to establish discrimination. 

For example, in Pretoria City Council v Walker, the Constitutional Court — interpreting the non-discrimination provision in the Constitution — stated that “proof of such intention [to discriminate] is not required in order to establish that the conduct complained of constitutes discrimination”. The court also confirmed in that case that discrimination “must be determined objectively in the light of the facts of each particular case”. 

Interpreting and applying the definition of discrimination in Pepuda, the Constitutional Court further held in MEC for Education: Kwazulu-Natal and Others v Pillay, that it would be guided by its interpretation of the non-discrimination clause in the Constitution, and confirmed that discrimination is established objectively, looking at all the available evidence to determine whether the act or omission impacted on the complainant in an impermissible manner.

But even if one ignores the jurisprudence, the claims made are ludicrous as the wording of the proposed amendment (“causes prejudice”) clearly requires an objective assessment, based on the available facts, to determine whether the complained-of act or omission caused prejudice related to a prohibited ground such as race, gender and sexual orientation. 

Non-lawyers sometimes make this mistake, not realising that different treatment cannot be discrimination unless it is linked to the person’s race, sex, gender, sexual orientation or other prohibited grounds.

Anthea Jeffery, head of policy research at the IRR, also suggests that the proposed amendments would expand the grounds on which discrimination can occur, expressing concern that the act empowers equality courts to decide, in any future case, that socioeconomic status should be recognised as a prohibited ground.

This is rather odd, seeing that both section 9(3) of the Constitution and Pepuda already allow the courts to identify additional (or analogous) grounds of discrimination. (See, for example, the Constitutional Court judgments of Hoffmann v South African Airways — identifying HIV status as such a ground — and Larbi-Odam and Others v Member of the Executive Council for Education (North-West Province) and Another — identifying foreign citizenship as such a ground.)

Jeffery also seems to be unaware that the equality court has already identified socioeconomic status as an additional ground on which discrimination can occur. 

Thus, in Social Justice Coalition and Others v Minister of Police and Others (Case number: EC03/2016, delivered on 14 December 2018), the Western Cape High Court, sitting as an equality court, held that the unequal distribution of police human resources in the Western Cape (with more resources allocated to affluent areas) discriminated against residents of Khayelitsha, among others, on the overlapping grounds of race and poverty.

In doing so, the court specifically referenced the provision in Pepuda identifying “socioeconomic status” as a possible additional ground of discrimination.

It is also very important to note that establishing that discrimination had occurred is only the first step in determining whether the complainants will win their case. The second step is to determine whether the discrimination is fair (and thus lawful) or unfair (and thus unlawful). 

The court will consider a number of factors listed in section 14 of Pepuda to make this determination, but no amendment to section 14 is proposed. All this means is that proposed amendments to the first part of Pepuda should be considered as tweaks to update the legislation and bring it in line with Constitutional Court and other jurisprudence.

The Amendment Bill proposes more significant changes to the second part of Pepuda, which is aimed at promoting equality. These proposed amendments are unfortunately not well drafted. The proposed text is also rather vague about the exact nature of the duties to be imposed on private parties. (Here I will only focus on the provisions relating to private parties as these are the provisions that the IRR and Neasa object to most strenuously.)

The proposed amended section 24 of the act imposes a duty (in general terms) on all private parties not only to “promote equality”, but also (a proposed insertion) to “eliminate discrimination”. 

As the IRR points out, it is odd that the proposed text imposes a duty to eliminate “discrimination” instead of “unfair discrimination”, and I assume this is a drafting mistake. 

Without the “unfairness” requirement, the section could be read as requiring the elimination of all kinds of distinctions based on prohibited grounds such as race, gender and sexuality — even when those distinctions are objectively justifiable or necessary for redress purposes. (Of course, Jeffery does not mention, or is unaware of, the latter point, perhaps because the IRR would probably welcome the abolition of affirmative action.)

The duty to “promote equality” must be read with the proposed expanded definition of equality, which would include: “(a) the full and equal enjoyment of rights and freedoms as contemplated in the Constitution; (b) equal right and access to resources, opportunities, benefits and advantages; (c) de jure and de facto equality; (d) equality in terms of impact and outcomes; and (e) substantive equality”.

What is left unstated in this definition is who must be compared to whom to assess equality. In the context of Pepuda, the obvious answer is that different groups (based on race, sex, gender, sexual orientation and other irrelevant characteristics) are being compared to one another.

For some reason, this proposed definition has caused consternation among some critics of the Amendment Bill — many of them proponents of “colour blindness” and/or proponents of an “equal opportunity society” — partly because it envisages the creation of a society in which individuals will enjoy an “equal right and access to resources, opportunities, benefits and advantages”, regardless of the individual’s race, sex, gender, sexual orientation or other irrelevant characteristics.

The proposed expanded definition of “equality” therefore provides guidance on what the promotion of equality should aim to achieve. 

Put differently, what might a South African society look like if systemic inequality were eliminated and the social and economic success, as well as the status, of each individual were no longer partly determined by the accident of their race, sex, gender, sexual orientation or other irrelevant characteristics. 

As this suggests what a society might look like in which race, sex, gender and sexual orientation truly no longer mattered, it is not clear why conservative critics are horrified by the proposed definition.

However, on a generous reading, this criticism might not be based on the vision of equality reflected in the definition, but rather on opposition to the idea that anyone — including private persons and companies — could ever be expected to do anything to promote its achievement. 

This view would make sense if one believed in “trickle-down” equality, or that the current racialised and gendered inequality in South Africa is entirely unrelated to the social system that privileges men over women and white people over black people.

Which is exactly why this view does not make sense — is there really any credible argument to be made that the gender pay gap in South Africa is unrelated to patriarchy, or that the obligations imposed on employers by the Employment Equity Act helped to reduce this gap by 24% over 30 years? This is a point that I hope to come back to in a future post.

Having said all this, there is much room to improve the proposed bill, especially those provisions dealing with the obligations imposed on private parties to promote equality. These provisions are far too vague and leave too much discretion to the minister to flesh out as the government sees fit.

Given the government’s refusal over the past 20 years to implement even the very weak provisions of the existing act related to the promotion of equality, I fear that this may signal a pre-emptive retreat by our government, based on a reluctance to upset big business and other private institutions.

But I would be happy to be proven wrong.

Source: Daily Maverick at