Watch here.Listen here.In this interview with Chris Steyn, Hermann Pretorius of the Institute for Race Relations (IRR) unpacks his polling results on South Africa’s women voters from 2021 to 2026, including the key finding that women voters are keeping the African National Congress (ANC) ahead of the Democratic Alliance (DA). “..all the years from 2021 till now, the ANC has an advantage among women voters.. if only men voted in an election in South Africa in 2026, the ANC and the DA would be essentially tied at 26-27%. If only women voted, the DA would be on 26% and the ANC would be on 37%.” Asked for his predictions of the outcome of the November Local Government Elections, Pretorius estimates that the ANC could get as much as 50% in Mangaung, while DA in Johannesburg could get between 42 to 44% and the ANC between 20 to 23%; the DA in Tshwane between 44 to 47% and the ANC between 24 to 27%; the DA holding Cape Town with between 58 and 61% of the vote; the DA in Ekurhuleni garnering as much 43% of the vote with the ANC down in the mid-low 20s and the Freedom Front Plus at between 9 and 11%; and in EThekwini the DA and MKP possibly tying at 30%, with the IFP at about 14%, and the ANC at around 10%. “But I think the number to watch is MK versus DA. Those I think in an urban-suburban setting are the two parties that are likely to be at each other's throats for that first spot.”.Sign up for your early morning brew of the BizNews Insider to keep you up to speed with the content that matters. The newsletter will land in your inbox every morning on weekdays. Register here.Support South Africa's bastion of independent journalism, offering balanced insights on investments, business, and the political economy, by joining BizNews Premium. Register here.If you prefer WhatsApp for updates, sign up to the BizNews channel here..Edited transcript of the interview.Chris Steyn (00:01.046)What is the political state of play with regards to women voters? Let's speak to Hermann Pretorius of the Institute for Race Relations. Welcome, Hermann.Hermann Pretorius (00:11.714)Morning, Chris.Chris Steyn (00:13.964)You've done polling from 2021 to 2026, and that shows that women voters are keeping the African National Congress ahead of the Democratic Alliance. Please share those details with us.Hermann Pretorius (00:29.418)So every year the IRR Commission's opinion polling and we look at three things: political voting intentions, race relations, attitudes, experiences, and policy priorities. And it struck me as I was looking through our historical data that we've got the perfect set of data that covers this period of 2021 to 2016, this municipal local governance election cycle.Why don't we try and see patterns in these? Because usually I'm so focused on the immediate findings that you forget that you actually have a storyline that expands itself. So this time I decided no, let me look at that storyline. And to make sure that we could make credible conclusions, you needed to make sure that within the data there was a subset that was number one consistent and number two large enough so that you don't sort of look at a group that is thirty respondents the one year, twenty the next, forty the next, because that really doesn't guide you in the direction of insight. And so naturally I thought, well, it's women's month in August. I'm looking at the polling. Why don't I look at women voters? And I didn't expect to make any real significant findings, but then the one finding that really grabbed my attention that you shared now is that it is female voting: They are keeping the ANC ahead of the DA. And this isn't just from our one sort of collection of data from 2026, it's all the years from 2021 till now, the ANC has an advantage among women voters. Not a fixed advantage of exactly the same size, but it's always present and it is remarkably consistent.Chris Steyn (02:26.071)exactly what is the figure that you estimate.Hermann Pretorius (02:29.932)11 points, roughly 10 or 11 points. And if you could look at it this way, if only men voted in an election in South Africa in 2026, the ANC and the DA would be essentially tied at 26-27%. If only women voted, the DA would be on 26% and the ANC would be on 37%. So not only does the ANC have a 10-11 point advantage amongst women over the DA, it also has it in turn over ANC women to ANC men. And the bulk of the report is actually dedicated to figuring out why this might be the case. Now I go through a few explanations that fail, that don't really account for what we see in the data. And then we land in a space where I think it is fascinating and an underappreciated area of research that women voters and male voters experience government policy differently.Corruption is an issue that more men care about than women. Illegal immigration, more men care about it than women do. Unemployment, both genders care very much about it, but men slightly more. However, if you then look at gender-based violence, housing, service delivery in the form of water and electricity, there you start finding issues that still matter to both genders, but women care more about them. That's where the politics become interesting, not as identity politics, but as issue-based politics that two groups of different people experience differently.Chris Steyn (04:10.029)Now the African National Congress has also lost a lot of support. How did the genders fare in that loss?Hermann Pretorius (04:18.862)What is remarkable, and if I glance up here, do excuse me, I just have my presentation and my data here to make sure that I am not sort of thumbsucking the figures here. What's interesting is that men and women since 2021 have left the ANC at roughly the same rate.But the ANC started this six year period, this local government cycle, on 55% of the female vote and ends it on 37% of the female vote. Where men began it on 44% in terms of the ANC vote, and the ANC ends it on 27%. So roughly an equal dip, but it's that built-in advantage that the ANC came into the cycle with that is still present and that is that dividing factor currently still between you know the first place and second place within our politics.Chris Steyn (05:19.703)What about the Democratic Alliance? How do the genders show up in their support?Hermann Pretorius (05:24.674)So you can group the parties into three categories male dominated parties, gender parity parties, and female dominated parties. The DA is one of three gender parity parties, I would say. The other is the EFF and then the Patriotic Alliance. They roughly have as many male voters as they have female voters. Then you've got your male dominated parties, MK, IFP, Freedom Front Plus, and that sort of sociologically it does quite make sense. These are more traditionalist parties, perhaps with sociological elements of patriarchal culture. Not that I'm making a judgment on that. I think it's just fair to say that one shouldn't be surprised that these traditionalist parties have this male dominance in their support level. And then in the female-led parties, that category only has one party in it, and that's the ANC. So it is quite a remarkable thing that there must be a reason for it. It is consistent, it is present. And if we add the three layers, or the two layers plus a third, number one, we've got the fact of this consistent advantage, number two, we've got the data on differing issue priorities. And then the third layer that I try to add in my analysis is to see if there is something in the ANC's political or governmental record that aligns with female issues that allows them to have this advantage, and there the household and issues related to the household, to children, to the ability to go from month to month, as many many households in South Africa are women-led. I think that was the thing that appeared in the data to show: look, this isn't an unthinking female advantage to the ANC, there's some real service delivery importance, impressive delivery points that ten, fifteen years ago probably built in that ANC female advantage among specific voters that has been quite stubborn in terms of the competition for other parties.Chris Steyn (07:41.272)Did your polling take into account race? The race of the women? And what did that show?Hermann Pretorius (07:45.445)Yes. Yes. The danger here is that methodologically I really wanted to go there. But you're in a dangerous situation if you start halving your sample. And you can still sort of if you go from 600 or 1,000 respondents to 500 or 300; statistically it still means something. But if you then go into the four racial categories and you take four hundred, three hundred people and you divide them into four, then you're starting to work with very small subsets that could just totally mislead you. So for that purpose, I am working on a next report specifically looking at township voters and rural voters. Who make up the two largest residential categories in our samples so that we can look at the racial component. But here, what stood out for me is exactly the absence of race. That if only men, all men, black men, white men, coloured, Indian, all men voted, you would get this result of a rough tie between the ANC and the DA. So it's almost worth excluding race in this specific report to draw attention to the fact that, you know, to point a phrase, there's a gender…it's not a pay gap, it's a voting gap. And if you leave a race out of it, if you leave class out of it, we still have this reality that we need to find an answer to: what is it that women and men respond to differently in our politics?Chris Steyn (09:20.824)You don't think that women are more forgiving towards the failures of the African National Congress and are hanging on to hope for longer than the men.Hermann Pretorius (09:31.077)That's possible, but I don't think it's visible in the data. And the first point I would say to counter that argument is the rate of decline of men and women in the ANC is roughly equal. That trajectory is almost parallel. So women and men have departed the ANC at the same rate. So you don't get a slower trickle off for women. And then secondly, a fascinating component is we also asked…Hermann Pretorius (10:04.558)…in our polling, are you more satisfied with your life than you were five years ago? And then we also ask it prospectively in five years' time for the next five years, are you more or less hopeful? And men and women don't have a consistent trajectory on that, that might explain this. If there was this idea of hope that the ANC might recover, might find itself, and that was a female trait in our electorate, then in terms of satisfaction, we might see that appear either retrospectively or prospectively, and we don't. In fact, one of the explanations for this difference between men and women voters that I exclude because it doesn't explain what we see in the data, is this satisfaction metric. For example, in 2021, men and women had a very, very different satisfaction score in comparison to 2026. And yet the size of the advantage that the ANC has between the genders remained consistent. So the satisfaction picture changed, it almost inverted itself. But the ANC still maintained that support. So sentimentally, you know, characteristically there might be an argument to be made for patience or hope, you know, typically the more mothering characteristics one might think. But I don't think the data shows it.What's more interesting is to look at, or not more interesting, but more explanatory, is if you look at the years 2002 to 2017, and you take male and female-led households as two groups, and you see at what rate water and piped water and mains electricity access improved for those two households, you get double digit…Hermann Pretorius (12:02.806)…improvement for female-led households in both power and water, but low single-digit improvement among men. So quite clearly female households had a better experience in terms of getting service delivery their household needs met in a crucial 15-year period. And I think that might perhaps be the explanation for that latent ANC bulkhead.Chris Steyn (12:33.262)And what would be the explanation for that? Better service delivery to female led households.Hermann Pretorius (12:41.306)Probably two twofold. Number one is the age old Apartheid era hangover of male migratory work. We still have this pattern where men from rural South Africa urbanise at a faster rate than women do because women stay behind with the kids and just sort of take care of the homestead or the household. And the rates of service delivery improvements differed because in the cities you already had higher levels of service delivery. So the improvement was more negligible and more invisible. But in rural areas, the improvement was drastic and visible. You could go from, you know, walking to a river to get water to a borehole to tapped water, and the change would be incredibly visible. So I think the first factor is just the residential divide, that where these improvements happened drastically, women were perhaps more the dominant lead in households. And then the second factor is just the biology that women are stuck with more household responsibilities than men. Men can sort of be in a relationship, a child is born, and men can go off, where the household pretty much accrues, if we see the household as sort of the family unit, accrues naturally by you know human nature to the female focus. And we see that in South Africa, we've got male-led households, female-led households, dual-parent-led households, grandparent-led households, and child-led households. And of five categories, female-led households in South Africa are 42% of households. So that is a very large chunk of these five categories. And if you have those households and you add the layer of residential service delivery trajectories, you end up in a situation…Hermann Pretorius (14:54.652)…where women likely saw the improvement of piped water and electricity mains to a more noticeable, more memorable extent during this period of frankly surprising and overlooked ANC delivery.Chris Steyn (15:12.878)The local government elections are in November. Are you ready to make predictions?Hermann Pretorius (15:21.108)So predictions are… I will try, I will try, but I must say I wouldn't want necessarily to be held to every digit that I'm about to share. So taking into account that question of rural suburban, this division that we've spoken about, there is possibly an indication to believe that more developed industrialised metros and municipalities will see better DA performance for a whole range of issues, but including the fact that those places are where the ANC's female-led advantage will likely be weakest.It's one of the factors that's contributing to the ANC being a rural party, because the rural household is still a rural household. And I think if you want to sort of look at the metros, look at which ones have the largest sort of rural character. I think Mangaung comes to mind, Buffalo City comes to mind. Those I think the patterns show us the ANC are likely to get 50% plus support, of course, factoring in the Mangaung administrative disaster of not submitting necessarily the full candidate list. So that might throw my Mangaung understanding a bit, you know, up into the air. So that's the first one. Are these metros more rural-based voters likely to be the ANC stronghold? Hermann Pretorius (17:39.757)But you ask for predictions; I would say this is where we might end up. DA Johannesburg, 42 to 44%. ANC Johannesburg, 20 to 23%. DA Tshwane 44 to 47%, ANC Tshwane 24 to 27%.And then the big one that you should keep an eye on: Cape Town I think the DA holds comfortably with, you know, let's say between 58 and 61% of the vote. Nelson Mandela Bay, I don't even pretend I'm going to understand because it's such a complex electorate. But the Metro I think people should really keep an eye on is Ekurhuleni. Ekurhuleni, I think is going to be the big surprise story from this election where the DA has a very good candidate, running a very good campaign and I think we might be looking at a DA on 43% in Ekurhuleni, and ANC down in the mid-low 20s and perhaps even a Freedom Front Plus 9, 10, 11 points. So we all think about Tshwane as the place where the DA-Freedom Front Plus Coalition is, you know, the natural thing. Don't rule out Ekurhuleni to have even a clearer DA Freedom Front Plus coalition picture than Tshwane might have.Chris Steyn (19:24.856)What about KZN, where former president Jacob Zuma reigns…in his own mind at least.Hermann Pretorius (19:34.32)Yeah, it is a question of so many variables in in KwaZulu Natal because you've got two parties who are Zulu nationalist parties who are competing for the Zulu nationalist voteThen you've got the ANC that has almost stabilised at its lowest point of support. I mean, they fell so catastrophically bad in 2024 that it's difficult to see them losing further vote share. And the real question is, of course, the suburban advantage that the DA might have. Now, I think we're looking at an underperformance by MK in EThekwini because of its suburban rather than rural character. So I don't think people should expect the same strength of MK as they had in 2024.And I think there might be essentially a tie, and this might be a very bad call, it might age horribly, but I think the D A and MK in EThekwini might end up essentially tied at 30%. Then you've got the IFP possibly 14-ish percent, ANC 10-ish percent, so the GPU constellation can emerge.Hermann Pretorius (21:06.756)But I think the number of course to watch is MK versus DA. Those I think in an urban-suburban setting are the two parties that are likely to be at each other's throats for that first spot.Chris Steyn (21:22.282)Thank you, what fascinating polling and fascinating predictions. Thank you. That was Hermann Pretorius of IRR speaking to BizNews and I am Chris Steyn Thank you, HermanHermann Pretorius (21:37.219)A great pleasure.