Six days before the vote, four serious pollsters could not agree on who was winning Johannesburg. Gareth van Onselen, whose Victory Research polls have drawn fire from the ANC and ActionSA alike, told Alec Hogg why the metro numbers diverged, why turnout could swing a party by up to five points, and why the real result would only emerge from two weeks of coalition bargaining after 4 November. He set out the test he believed eThekwini would pose for every other metro, the squeeze facing the DA in Cape Town, an idea for running Gauteng's three metros as one, and the ANC strongholds that, if they cracked, would mark a change not seen since 1994. He also had a blunt verdict on Herman Mashaba's claim to Soweto, and a word on the eight weeks his own Joburg tap had run dry..Listen here.Edited transcript of the interview.Alec Hogg (00:01) Welcome to the BizNews Edge. Well, six days out, four serious pollsters, but in Johannesburg they can't agree who's winning. One has the DA cruising, another had the ANC in front. So who's right? And what should you be looking for when the results start rolling in on the 5th of November? My guest has built the most detailed by-election tracker in the country. His firm runs the polls the ANC loves to hate. And not just the ANC: ActionSA as well is not too mad about what his polls say. Gareth van Onselen is with us. He's the top pollster, for my money. Wonderful to have you with us, Gareth. He's also, by the way, the author of Inside Politics. It's a must-read blog. And he's the CEO of Victory Research. Okay, so we did that bit.You said to me off screen, when I said to you this is the most important interview this week, that I'm just flattering you. But my goodness, Gareth, I'm not the only one waiting to hear how you're reading this, because for many of us it seems a little bit confusing. Overall, I suppose we could start there. The polls seem to be kind of as one on the ANC, between 31 and 32% nationally, and the DA at 25 to 26% nationally. They seem to get it right there. But when it comes down to the individual races, there are lots of changes. Why would that be, for someone who doesn't really know this stuff?Gareth van Onselen (01:36) Well, look, the sample when you do a national survey is just much bigger, and much more reliable and much more stable. And because you're dealing with a much bigger population, there's much less fluidity, and so the bands in which parties operate tend to be more stable. And that's probably why the national big picture appears more coherent. Metros are, A, a smaller sample, but B, they're far more volatile, and there is more movement in them.They are also the primary target of this election and of election campaigns. So it's where political campaigning is focused. It's not focused at a national level, and if anything is going to have an impact, it's at metro level.Alec Hogg (02:21) Now, Herman Mashaba did take issue with you last time. You were right, he was wrong. Are you reading him the same way? Well, many seem to say that ActionSA is not a spent force, but certainly not as strong as in 2021.Gareth van Onselen (02:37) Well, no. Actually, the last SRF poll which we did, which was their most substantive poll, had a national sample of around 3,000 and 800 for three of the big metros. It had ActionSA on 3.4%, I think, nationally, at a sort of 40% turnout. So that would be growth of almost a percent or so nationally for ActionSA. Certainly a fundamental improvement from 2024.It does, according to that poll, have a problem in Joburg, though, which puts it sort of eight points down from where it was in 2021. And that's a fundamental problem for ActionSA, because that's a core pillar of their brand: Joburg, and Mashaba and his mayoralty there. But it's up in Tshwane. It seems to be doing well there, on about 12% or so, which it can take some encouragement out of. But on the whole, the poll does show some growth for ActionSA.Alec Hogg (03:41) Interesting that you see it as encouraging. Those numbers, they would find pretty disastrous within the organisation, because they are aiming much, much higher. I guess you have that often, because politics, as someone was saying to me yesterday, is all about emotion, not really about logic, and you've got to play in logic.Gareth van Onselen (04:03) Well, one must also differentiate between what political parties' private ambitions are and what their public ambitions are. This election, like every election, has been rife with insane political predictions from politicians, most of which downplay their opposition and play up their own support, which is part of the game. Although they do really stretch the bounds of credulity with some of these things.So it's very difficult for us to gauge internally what a success or a failure is for a party, because we don't know what the actual internal measurements are. If you hold them against their external communication, well, every single party in the country is going to fail profoundly, because they all seem to be expecting 100% of the vote. I think one of the real pragmatic tests is power, which, you know, we're going to have to rely on the coalition negotiations to determine. Who wins that battle has got nothing to do with polls. That's got to do with your ability to negotiate a position, and we need to see how that plays out.But sorry, just a final point on that. I think one of the things that tends to happen in local government elections is that the polls are quickly forgotten in the wake of coalition negotiations. In a national election, success or failure is pretty much determined by your percentage and whether you're up or down. In local government elections, you can do extremely poorly but negotiate some position of power. Coalition negotiations overtake the media for two weeks and become the be-all and end-all of the election. And once that is done, what people remember is whether you're in government or not. So any party can make up for a poor performance by being able to deliver a disproportionate amount of power in a coalition agreement.Alec Hogg (05:56) Help us through the big ones where there is a lot of confusion, and maybe we can start in Johannesburg. We had an impassioned plea yesterday from probably the top fund manager in South Africa on BizNews TV, Cy Jacobs, saying he's got behind the Helen Zille campaign. Please, it's the last chance for Joburg; if we don't get out and vote, and if Helen doesn't run the place, he sees nothing but disaster. That's usually a very rational mind. Has Cy got reason to be concerned?Gareth van Onselen (06:31) Well, I think everyone's come to the conclusion, and certainly all the polls, not just the SRF polls, suggest that there is no realistic prospect of a majority government for the DA. But there is a very powerful, probably the pre-eminent, outcome that the DA will end up with the largest share of the vote in Johannesburg. And so much of the effort now, among people who support the DA, seems to be towards maximising its advantage as the biggest party rather than a majority party, the logic being that it will enter coalitions with the most informal power.And that is something to take very seriously. A lot is said about the degree to which the DA is alienated from other parties and doesn't get on with them. But the truth of the matter is that the kind of informal power that an actual election result delivers counts for a lot in a negotiation. Politicians talk a good game publicly, but privately they understand that you can write off 40% of the electorate, or 35%, or whatever the DA gets in Johannesburg, but it comes at a great electoral cost. And your ability to justify that is extremely limited. Those things count for much more behind closed doors than they do publicly. And that's why there is this massive emphasis.So I think the closer the DA gets to 40%, the more informal power it will have in those negotiations, and the more difficult it is for other parties to make a rational and acceptable case to the public as to why it should be excluded. I mean, the DA is not MK.Alec Hogg (08:18) Gareth, what about the turnout? How important is that? And why is it that national elections are so much more appealing to voters, in terms of actually going to the polls, than local elections, when local elections are actually where you should be paying more attention?Gareth van Onselen (08:37) So that's a very important question. Turnout is the most critical thing in a local government election. It is the decisive factor. And it can shift any particular party's support by up to five points, if it's able to get its voters out at a disproportionate rate. So if your people turn out on election day, if you run an effective campaign that motivates them at a disproportionate rate, there is a huge practical benefit to it.The fundamental trend, however, is that turnout is in decline, and therefore it is getting harder and harder to do that, because the pool of active voters is shrinking all the time, and your ability to generate enthusiasm is more and more limited. I think Johannesburg is exceptional amongst all metros in this regard, because of the impact of Zille's campaign. It is disproportionate. You can see a fundamental change in direction in the DA's track. Long term, it's struggled for many elections in Johannesburg and has in fact been in decline. And the fact that it has grown means that there is a disproportionate amount of enthusiasm in that metro, and its ability to deliver a disproportionate turnout is much higher, and the DA will be banking on that. So it's absolutely critical.Why people vote more in local government elections than national elections: it's a complex answer to that question. If you had to put a gun to my head, I would just say it's laziness. People just don't take it as seriously as they do national government politics. I think they don't understand the potential impact of local government on their lives. They see things in very grand terms. And local government is the thin end of the wedge. Policy can have massive implications, obviously, at national level, but a lot of it is implemented at provincial and local government level. And I think a lot of people miss that.Alec Hogg (10:39) Not sure whether you've been into the granularity of something as far down as uMngeni, but the fact that Chris Pappas has done such an amazing job at uMngeni by any standards, and pretty much everyone agrees, and yet he might still lose it to MK, who have targeted that as their big opportunity, would seem to suggest that emotion really does rule, and maybe tribalism as well plays a big part and will play a big part in these elections.Gareth van Onselen (11:10) Well, I'm not a big believer in this idea that MK lives or dies by Jacob Zuma. I think MK is actually about Zulu nationalism, which has been a force in KwaZulu-Natal politics for 30 years. You forget that the IFP had 50% in the early nineties in KZN. That was slowly eroded away with the rise of Jacob Zuma. But the ANC actually just accommodated that Zulu nationalism. It never actually diluted the impulse. And all that's happened with Zuma's breakaway is that Zulu nationalism now has a very powerful exemplar in Jacob Zuma.And I think if he were to die or retire or give up, or be thrown out, whatever happens in MK, all you need is someone else who is able to capture and represent that particular cultural-political force in KZN, and they will have a huge amount of support. And that's very difficult for any party that is not an ethnic nationalist party, or demagogic in any way, to counter, because that is where the emotion that you point to comes in. It is by its nature an appeal to identity, as opposed to a set of values and principles.And that is a very hard argument to win. In order to win it, you need to identify a series of values and principles that people hold as strongly as, or more strongly than, that kind of innate sense of belonging. That is incredibly difficult to do in politics, and those kinds of issues do not grow on trees. And that's the fundamental challenge that I think any party that doesn't have Zulu nationalism underpinning it in KZN struggles with.Alec Hogg (13:05) Mm. Even Chris Pappas, although he's done so well.Gareth van Onselen (13:09) Yes. Look, they won uMngeni on a knife edge, so it's not like they're losing some massive chunk of support. It was always going to be a heartfelt battle. And they won it while MK didn't exist and the ANC was in decline. And one of the benefits for the ANC of having accommodated Zulu nationalism was that it nullified the effect of parties like the NFP and the IFP. Once it was let loose and could be identified in and of itself as a force, it became far more attractive and powerful, and it is far more powerful outside the ANC than inside the ANC. And that's as big a problem for the DA as it is for the ANC, which has been absolutely decimated in that province.But look, it's not all doom and gloom from the DA's perspective. It's doing very well in the tracking in eThekwini. Its long-term trajectory in KZN is slow but upwards, and it seems like a big growth area. But uMngeni is an outlier, and its ability to secure a majority there is going to be very difficult.Alec Hogg (14:20) Hm. Moving to eThekwini, there are some who say it's home and hosed for the MK party.Gareth van Onselen (14:30) Well, I think most of the polls have MK... The first thing you have to understand about polling in KwaZulu-Natal, and this was demonstrated in 2024, is that it is a very volatile province and metro, because it is undergoing a sort of fundamental change and restructuring of the landscape, which is playing out slowly over time, and the chips haven't settled yet. So what you see is a huge amount of movement between political parties.And obviously the limitation of any poll is that it is a snapshot in time. That's fantastic if the market is stable; it gives you a relatively good read on where the lines are headed. But if the market is volatile, like it is in KZN, you really need a tracking poll, which we don't have this election, to give you a sense of where things are headed. That said, MK looks particularly strong in eThekwini, as we sort of knew it would be after the 2024 elections. It doesn't look like it's quite set for 50%. It seems to be mid-40s.But the thing that's interesting about eThekwini from a coalition perspective is that it's likely to replicate what happened in KwaZulu-Natal. And a principle has already been established there in terms of coalitions, which is that all parties seem willing to do everything in their power to keep MK out. It's actually extraordinary. There's a 1%, or one-seat, threshold in KwaZulu-Natal, and that has forced all these parties to work together. If that principle has already been established, it's difficult to understand why it wouldn't be established again in eThekwini. It would certainly be a hard sell to voters to say we've got one provincial set of values and one metro set of values.That's not to say South African politics, which is defined by hypocrisy and double standards, is not capable of that, but it's a hard sell. And so for me, eThekwini is kind of the lighthouse on the hill, in so far as if that principle holds into the local government elections and it defines coalition arrangements in eThekwini, then why should those arrangements not apply elsewhere, to other metros? So eThekwini is the test of the belief of political parties as to whether or not MK should be kept out of government. And so I think it's a real litmus test for the other metros.Alec Hogg (16:54) It's quite a critical question as well, though: how well have the DA and the ANC worked in eThekwini? Because if they're getting along there, the template's been established. Why not Joburg? Why not Nelson Mandela Bay? Why not many other parts of the country?Gareth van Onselen (17:13) Yeah. You see, one of the problems with South African politics is that the analysis of it is highly tactical. So what tends to happen is people see what the lay of the land is like after an election, and then they say, well, this is the obvious path, or this is what you can do, or this is how you can negotiate for as much or as little power as possible.But you forget that it's fundamentally important for a political party not to be entirely pragmatic. They have to be strategic. And to do that, you actually have to have a clear idea, before the election results, of what it is you want to achieve. Then you need to look at the lay of the land, and then you need to make a decision as to whether or not your outcome is compatible with the lay of the land. I understand compromise is part of the game. Obviously these are not absolutes. But more or less, you need to be able to align whatever coalition agreement you want with your long-term strategy.That is a particularly difficult question for the DA and the ANC. On the one hand, they have, as in KwaZulu-Natal, established a principle and value already at national level, and so it would be odd if it wasn't replicated in the local government election. You'd need a different explanation for that.The one potential explanation, although it does introduce tensions into the GNU, is that there would be a fundamental differentiator in the power dynamic. At national level, the ANC is the majority party, which suits its disposition. It believes it is of the people and for the people and represents the people, and acts in quite a divine manner, to use a religious philosophy, and I think is quite arrogant in that regard.At local government level, the DA would be the majority party. And that's never happened in South Africa before, where the ANC has been a minority party in a coalition, or maybe once, in KZN in the early nineties, but it is largely untested. And how the ANC would respond to something like that, where it needs to take an informal lead from another party that is bigger than it, would be a fundamental test of that relationship, and how stable it is and how mature it is. And I don't know how that would play out, but it would be a very serious concern for me if that is the approach those two parties want to take.Alec Hogg (19:35) What about their followers in the ANC and the DA? How open are they, or how supportive are they, of those two parties working together?Gareth van Onselen (19:45) Well, the polls suggest that there's a lot of support for an ANC-DA agreement. There are upsides and downsides for both parties and both sets of voters. But first of all, it doesn't actually matter what the arrangement is: South African voters historically and traditionally really love the idea of cooperation. One of the reasons why the DA did so exceptionally well in 2000, when it formed, was because it united the opposition, quote unquote, and got a huge amount of support in that regard. People really like that kind of thing.But those perceptions are informed by reality, and the more fractious a real-world relationship is, like the Government of National Unity, the more it begins to affect people's perception of working together. People's relationship to coalitions in the abstract is not great. People don't trust them and are far more hesitant about them than they are about the idea of political cooperation. They haven't had a good experience of actual coalition government. So while voters on both parties' sides are favourable towards them working together, there is this juxtaposition where coalitions are frowned upon.And being able to bring those two things closer together depends on a good experience, which I don't think the GNU has delivered yet to a great degree, and certainly coalitions haven't. And that's the fundamental problem: the gap between ideals and reality.Alec Hogg (21:25) Moving to the Cape, starting at Nelson Mandela Bay. It seems, from most of the polls, to be a dead heat there between the ANC and the DA in that metro. How would you see, with all your political insights, that working out? More of the same, or perhaps a fresh start?Gareth van Onselen (21:48) Yeah, I've done quite a lot of in-depth analysis of the VD-level data in Nelson Mandela Bay. And the truth about Nelson Mandela Bay is that there are a series of strongholds there, ANC and DA strongholds, which seem impenetrable for the other party. And there is a very small number of fluid VDs. And so what tends to happen, and you've seen this in the last three or four elections in Nelson Mandela Bay, is that it is a competition of turnout rather than an ability to win over new voters.So given the DA's track record in local government elections, it is likely to perform better on the turnout front. But I don't think what you're going to see is any fundamental shift in the core support base in Nelson Mandela Bay, and the differential will be really small, most likely in the DA's favour, if previous trends are anything to go by. And that won't be enough for either party to get to 50%. It might deliver the DA a marginal majority, but you'll face the same kind of situation you do in other metros.Alec Hogg (22:59) And moving to the Western Cape. The PA, we haven't mentioned them yet.Gareth van Onselen (23:04) Well, the PA is not really a factor in the City of Cape Town. It is a factor, but only in so far as every small party is a factor in the city. And the DA's problem there is a kind of death by a thousand cuts. There are a lot of small parties, each taking a tiny chunk out of the DA. It's not losing a substantive amount of support to anyone in particular. That story is different in the rural Western Cape, outside the metro, where the PA is much stronger and a more consolidated threat.But in the metro, the fundamental problem is that the DA has to fight a battle on about ten different fronts, against ten very small parties, each of whom just want a percent, and that has been enough to hurt its overall average. And then the DA's latest tracking poll, which needs to be read with some caution because it's got a 21-day moving average, which News24 did report but no one's picked up on. That means there are 21 days of data in that average. So it's not an immediate read of the environment; it's a read of the average of the last four weeks. So you just need to understand what you're looking at. Anyway, it puts the DA on 51%, and this is the DA's own internal polling data.That is obviously a knife-edge scenario, and the DA, you can be sure, will be pouring gazillions into its squeeze message, which is exactly to counter what I've said: this kind of "stop the chaos, reduce the impact of a multitude of small parties with no real influence, and back us". That message has been very powerful in the past. The DA needs to squeeze 2 or 3% out of those smaller parties to get into a kind of safe zone of 55, 56%. We'll only know on election day how effective it's been, but it's under big pressure in that metro.Alec Hogg (24:55) Mm-hmm. You have been sharing your by-election tracker with us, and there's a lot of work that goes into that. If you take that into account and then have a look at the rest of the Western Cape, are there some surprises that are going to come out at us after election day that no one's really factoring in yet?Gareth van Onselen (25:18) I don't think there'll be surprises if you use the 2024 election result as your guide. There will be surprises if you use the 2021 election as your guide. The PA, for example, really exploded in 2024. And all things being relative, it will probably replicate that performance in 2021 in the rest of the Western Cape, but I don't think it will exceed it dramatically. So it will appear to have grown substantially compared to 2021, but the trend was established in 2024.And that will be a big concern to the DA and the ANC, both of which are hurt by the PA. The ANC is a bit of a lost cause in the Western Cape, but the DA in particular, in certain core municipalities... Yeah, that is a fundamental problem that it has failed to solve thus far.Alec Hogg (26:17) Gayton McKenzie told us that he believes they're going to win in George.Gareth van Onselen (26:23) Geez, I haven't looked at George. I couldn't say on that. They've done very well in by-elections there, but it's a very select number of by-elections in which they compete, which tend to be overwhelmingly coloured wards. Its performance outside of that... Obviously, you need to win everywhere to win a majority. But without a doubt, I would think it would be a substantive player in whatever government needs to be negotiated after the election in George. It definitely has a very strong base there.Alec Hogg (26:57) And is it likely that within the Western Cape you'll see a lot of governments that formerly were majorities becoming coalitions?Gareth van Onselen (27:08) Well, the projections have something like a hundred hung councils expected. I would imagine a vast majority of those will be in KZN, where, again, if the 2024 result is replicated, the ANC is going to be decimated.Interestingly, in eThekwini, if you look at the polls as a whole, they do show the ANC recovering a bit. The earliest polls had it at around 8% in eThekwini, and it is up to 14% now, which obviously is a relative improvement. It's still a massive slaughtering of the ANC's vote in that metro. But its fundamental goal will, ironically, be the same as MK's goal everywhere other than KZN, which is to get enough to be able to force some power in a negotiation.But I think the ANC is reconciled to the long-term trend: the best it can hope for is to be a majority party in coalition negotiations. Not a majority party, the biggest party in coalition negotiations. And its days of being able to dictate terms in municipalities are getting more and more limited.Alec Hogg (28:26) Hm. I kept the Freedom Front Plus for last, because Corné Mulder was in our studio a little while ago, and he said everything is showing him they're going to have a very, very good election. His words. And I guess where they might score best is in Tshwane. So perhaps you can give us your thoughts on how the Freedom Front Plus has been polling, and what it looks like in Tshwane.Gareth van Onselen (28:52) So the last SRF poll, which was, as I said, its most substantive out of the three it did on this election, had, I think, the FF Plus on around 3% or so, which would be about a percent's growth from 2021, down from 2019, when it got about 4% or so. There is a possibility it could get up to that. There's no one particular place where the polls suggest the FF Plus is shooting the lights out in terms of urban votes. It might be doing very well in rural areas. It's pretty stable in Tshwane, at around 8% or so. I think it got 7.5% in the last election. So stable and solid. It's got a small showing in Ekurhuleni.The problem with the Freedom Front Plus for me is that it depends on your frame of reference. That quote from Corné Mulder is interesting, because the party's been around for 30 years and it has never really breached 2% in any sustainable way. Now, you come from the business world. If there was a business that never had more than 2% profit for 30 years... It's not even profit, but it's just a political disaster in big-picture terms. It's one of the biggest political failures in South African history.If your goal is to win national power and win majorities and grow your vote, it's a disaster zone. So what is your measure of the Freedom Front Plus? Is a very good election growing by a percent, from two to three? In which case, well done, man. You really shot the lights out. If your goal is to actually be a powerful political force in South Africa, it's just a joke.Alec Hogg (30:37) Tshwane, how do you see that ending?Gareth van Onselen (30:41) Well, it's exactly the same situation, in terms of the lay of the land, as Johannesburg and eThekwini. The polls have the DA in first place, the ANC in second place, and MK and the EFF kind of third and fourth. And what's interesting about Tshwane and Johannesburg and Ekurhuleni is that, for the first time post-ANC majorities, the lay of the land is going to be more or less the same.There is a real environment for a consolidated coalition agreement across all three metros. No political party that I've seen has actually put this idea forward, but it seems pretty obvious to me. What you want is a situation where all three executive mayors, with similar administrations behind them, can sit down and have a proper conversation about economic reform, integration, infrastructure planning and a long-term strategy to turn all three metros into an economic hub.And the benefit of that is, A, it reduces political competition, so you don't have different metros pursuing different goals at odds with one another. But B, your ability to improve a single metro is dramatically improved if its surrounding metros are on the same page. So I don't really regard the three Gauteng metros as fundamentally different in any way when it comes to whatever the post-election discussions are going to be. I think they should be approached as one, and I think the lay of the land is going to be the same in all three of them.Alec Hogg (32:21) Gareth, when you look at the direction of travel of South Africa after this election, will it be perceived as positive? Or what would make it perceived as positive? And I say this because it is important. We have economic growth which is once again below population growth, so GDP per capita is contracting. We have a rand that's stable but not really that strong, and we have unemployment which is untenable. There are many eyes looking at this reform story, which is starting to look a little bit more positive. But many eyes are watching this election and will be hoping to get some picture out of it which shows the direction of travel is at least going the right way.Gareth van Onselen (33:05) Yeah. Look, you're the expert on this, not me, but I would imagine most investors are long on South Africa, not short. And so their view, particularly with regard to the local government elections, is whether or not it fits into a long-term trajectory towards reform and economic stability and growth. Local government elections are very unsettling, because they are so fluid and dynamic and complex, and there are so many different hubs, to leave any analysis or forecast of what's to come with a coherent picture of what it means for the long-term view.And so I suspect, from an economic point of view, people will obviously want more stability, which is the key thing, and a sign of coherence in terms of policy and governments. They don't want a wide range of a hundred different governments doing a hundred different things, particularly at metro level.But the fundamental outcome of it will be the implications for 2029. Almost immediately after this election is complete, that will become the discussion and the determining factor for investors. And it is no easier a question. The DA will highly likely, if these polls hold, finish within 10 points, and there's a scenario where it finishes within five points of the ANC. And for the first time ever, the DA will be able to say in a meaningful way that the ANC can be beaten. That's a message that has always been projected 10 years into the future, but now can be brought home in a very real way. And that introduces even more uncertainty.So I think it's just a shuffling of the chairs on the deck, unfortunately, for those people with an economic view on South Africa. And whatever they take out of these local government elections is going to be made redundant on November the 5th, and they will have something new to worry about.Alec Hogg (35:07) And you, what are you going to be looking at on November the 5th?Gareth van Onselen (35:11) I'm going to be checking whether I have water in my tap on November the 5th. If I do, that will be a good thing, seeing as I haven't had it for eight weeks. No, from a personal perspective, as a resident of Joburg, one wants to see a fundamental change in service delivery. That is the profound problem. You drive on the roads and they're broken. Water doesn't come out of the tap. Electricity is constantly cutting out. These are very simple symptoms for a resident, but underneath them lies a fundamental crisis with infrastructure.And what you want to see is a long-term, medium-term and short-term strategy to address that. You want to see the ability to raise capital, because it's going to need a lot of it to fix these problems. And you want to see a government that is, A, good at managing expectations, because these problems are not going to be fixed overnight, but B, excellent at communication. This balance, between "this is going to take a long time to fix" and constantly reassuring and proving to the public that something is being done, is a very difficult balance to hold. You need to do both those things relentlessly, all the time. And yeah, that's what I would look for.Alec Hogg (36:37) But in the results themselves... And I'm sorry to hear that, Gareth, eight weeks without water. I'm surprised all of Joburg isn't out there demanding a vote already. But in the results themselves, are there any particular areas you're going to be focusing on?Gareth van Onselen (36:57) No, not really. I'm really interested to see... Look, the big takeaway from this election is the extent of the ANC's decline. A lot of that can be explained by MK, the effect of which we've already seen in 2024. And so it's kind of built into thinking that the same chunk will be taken out of the ANC again. But the polls suggest it's not just going to be the MK chunk that goes; there's going to be additional decline in the ANC's vote beside that.And if it drops to around 30%, that's its biggest decline ever. And if you're interested in a fundamental change in South Africa's political landscape, you need to look at how that decline manifests in terms of particular ANC strongholds and VDs. There are strongholds in South Africa, a great many of them, in many metros and particularly in rural areas, where it doesn't matter to what extent the ANC declines nationally or provincially, or even within a metro; those strongholds remain impregnable. And if some of those crack for the first time since 1994, that will for me represent a fundamental change in the nature of the landscape. It has never happened before. If it happens for the first time, it means a core element of the ANC's support base is fractured. And I think that's very telling. But we'll have to wait until after the election to see.Alec Hogg (38:25) Which are the strongholds?Gareth van Onselen (38:28) Well, you could literally list them all. There are, you know, 20,000 VDs in South Africa. I would say at least five or six thousand of them fit that description. They're scattered all over the place. There are a lot in the Eastern Cape. There were traditionally a lot in rural provinces like Limpopo and Mpumalanga, and in the North West. But in the metros, there are particularly township areas, somewhere like Soweto, where you saw the beginnings of a crack in 2021, although they were in more marginal wards. But there are a number of them. And it's actually a great test for any political analysis. You could literally list all the VDs that the ANC has won with a majority since 1994 and see which of them crack. And they are in places like Soweto when it comes to urban centres.Alec Hogg (39:22) So when Herman Mashaba says he owns Soweto, he's going to win Soweto, do you, A, take him seriously, or B, if he's right, take a bigger message from it?Gareth van Onselen (39:33) Well, you can't decline eight points in Johannesburg and win Soweto. Those two things are not possible. You couldn't even win Soweto with 18 points. So even if he replicates 2021, he can't win Soweto. So that's just fantasy talk. It is a relatively strong area for ActionSA. I'm not saying it's devoid of support there, but its ability to actually win wards is very remote as things stand, or at least if the polls are to be believed. So no, he's not the king of Soweto.Soweto is becoming quite fluid, but it's not actually the most interesting area in Johannesburg. The most interesting area is the city centre. That area is dominated by MK and the EFF. If you look at just the VDs that make up the city centre and its surrounds, I think I did an analysis of them... I'm going to get the numbers wrong. Let's say 30 VDs... or no, no, about 80 VDs that comprise the city centre. Within those 80 VDs, there are about eight political parties that have won VDs. And that is what South Africa would look like if it had a healthy, diverse democracy: this little epicentre. It's a fundamentally interesting, fluid series of VDs. It is dominated by MK and the EFF, but there is clearly a market there for all parties. They're all represented. And how that turns out on election day is going to be fascinating to look at.That trend is seeping into Soweto, but Soweto is almost an election behind what's happening in the city centre. ActionSA has some strong wards in the city centre, and that could be an area in which it makes an impact. It's smaller and more limited. So there are all of these microcosms within Joburg, which make it so fascinating, which have a life of their own and a dynamic of their own, and parties do better or worse in each of them to some extent or another.Alec Hogg (41:37) Fascinating. VDs meaning voting districts.Gareth van Onselen (41:40) Voting districts, yeah. They're the basic level of electoral outcomes. Every ward has a series of voting districts, and a series of wards comprise a municipality.Alec Hogg (41:51) So we've got fantastic data. We fortunately have people like you who analyse it for us.Gareth van Onselen (41:57) Well, look, you're going to, as the youths say, trigger me now. I don't think we have fantastic data. I think the state of South African data is dire. It's in a terrible condition, poorly kept, records are incomplete and it's very difficult to access. It's badly formatted, and it is an absolute nightmare if you are interested in data analysis. My dream job is to be made the Statistician-General and to restructure South African statistics. There are pockets of excellence; some things are better. But as a general rule, it's just not good at all.Alec Hogg (42:36) So I should have rephrased that. We have the potential for fantastic data, because we have all of these voting districts, but unfortunately what's coming out of them at the moment is not that good.Gareth van Onselen (42:48) Yes. Well, it's accessibility that's the problem. The data is difficult to get, it is badly formatted and it's not coordinated. So different data sets across different departments and different institutions don't speak to each other. There are three fundamental institutions that provide electoral data. There's the IEC, which obviously gives you election results. There's the Municipal Demarcation Board, which is responsible for boundaries. And there is Stats SA, which is responsible for demographics.They all use different measuring points. They don't necessarily speak to each other in the right kind of way. It leads to a whole lot of conflict. Some of them have better data than others. Most of the data is incomplete or difficult to find. You need to unify and coordinate all that stuff to be able to generate proper insights. And I think it's a problem of institutions not talking to each other.Alec Hogg (43:45) Well, Gareth, who knows? Maybe one day you will have your dream job there. But for the moment, thank you for the insights that you've given us, with less than a week to go to South Africa's local government elections. There's going to be a lot of attention on the 5th of November and thereafter, to try and analyse and understand the results. And then, as Gareth explained to us, another two weeks of coalition discussions. It's going to keep the news cycle turning, and rather interesting for those of us who work in it. Gareth van Onselen is the chief executive of Victory Research. I'm Alec Hogg with your BizNews Edge..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.