This article covers five assumptions about tomorrow’s election and exposes them to immense scrutiny to see why they might be wrong.

The DA will win in Cape Town

The average of polls to date places support for the Democratic Alliance (DA) in the city at around 53%, and the assumption, therefore, is that it wins comfortably.

Stress-test that call, and some real doubt can be conjured up. Consider that the DA won 67% of the city in 2016 and 58% in 2021. That was a nine-point drop. Consider further that a wandering albatross reports that the DA has polled below 50% in the city in the run-up to this election. Consider also that there is a lot of complacency in the Cape Town middle-class base. The idea that the DA might ever lose the city is often met with great surprise. The DA has also taxed its core support base heavily, and grumbles are heard now and again. The Freedom Front Plus (FF+) is also showing signs of growth in the broader Cape, especially as doubts about the DA’s ideological bona fides rise among more conservative voters. Perhaps most important is that coloured nationalism is on the rise in the broader Cape. Elections are no longer about the DA and the African National Congress (ANC) competing for the coloured vote. Coloured parties are now competing for that vote too. A surprise result could therefore easily materialise where the DA falls a point or two short of 50% and needs the FF+ to bail it out.

The DA won't win Johannesburg and Pretoria Outright

The sum of all polls, as we have calculated them, places the DA at 37% in Tshwane and 36% in Johannesburg, hence the cities seem deadlocked.

Try to doubt that, and a good place to start is with some maths. Assume a margin of error of about 3% on those numbers. Give the DA the maximum benefit of that, and its average numbers could be rounded up to about 40% in the two cities. Next, work in the relatively small number of voters, even in an election like that of Johannesburg, where just over one million people vote. A 50 000 to 70 000 turnout swing for the DA can shift the overall result by four or five or six points. Now that 40% can be stretched to 45%+. Consider now that the DA won 43% of Tshwane in 2016 and 38% in Johannesburg. Ten years later, for all the ANC’s failures, can the DA really be flat in both cities? Its vote shares in the two cities fell sharply to 32% and 26% in 2021 because of the rise of small parties. But these may turn out to have frustrated voters, giving the DA another point or two or three. Consider also that the Social Research Foundation (SRF), which called the 2024 result to within a percentage point, did so via a 30-day tracking poll with a rolling sample of voters. In other words, it polled daily for the 30 days up to the election. It has not done so this time for any of the cities and hence has made no forecast of the result. The surprise result of a DA win in Johannesburg and Pretoria gets conjured out of all that reasoning. Even Gqeberha becomes interesting when exposed to this same set of tests.

Durban will be deadlocked

In Durban, the uMkhonto weSizwe Party (MKP) is averaging 42%, followed by the DA, ANC, and Inkatha Freedom Party (IFP). But for the turnout maths reasons cited above and the lack of a tracking poll, that 42% could become a surprise 50%+. Jacob Zuma’s bout of ill health may be the catalyst here. When Prince Buthelezi fell ill and died, the SRF picked up a near ten-point lift in IFP support, a homage of sorts to the man. Mr Zuma’s ill health, and his brave appearance on stage this weekend despite that, may have a similar effect.

The post-election talks will be chaotic

In addition to the major centres, an assumption around this election is that in more than 100 towns and cities, no party will win a majority, translating into 100 separate coalition negotiations. And that the result will be a big mess.

In Gauteng, the most important theatre, the regional ANC has adopted its own rogue version of the Government of National Unity (GNU), rejecting the DA/ANC pact for an Economic Freedom Fighters (EFF)/MKP/ANC pact. The national ANC has to date allowed that to persist, including for the reason that the split serves its purposes, as it ends up as top dog in both GNUs and assumes it may remain that regardless of which GNU ultimately comes to the fore. But its pending losses in Gauteng are going to be so severe that, in both regional and national corners, the sense is emerging that it really might be better to see Gauteng follow the national pact. That, of course, will require the ANC central leadership to make some firm decisions and to overrule the region.

The region, of course, has a strong incentive to reject that to retain the advantages of patronage that arise from the EFF/MKP/ANC pact, advantages that would be greatly diluted if the DA came to power. But should the ANC decide to make a surprisingly firm decision and countermand the region, then the national GNU pact may extend to Gauteng.

The ANC is continuing to weaken  

On the average of all polls this year, the ANC stands on about 34%, which is way down from the 54% it secured in the 2016 local elections and the 46% in 2021, suggesting that its post-2024 support continues to dive.

How can this be tested to breaking point to deliver a conclusion other than that the ANC is indeed crashing? Give it some margin of error and maybe you can lift that 34% number to closer to 40%. Consider the extent to which this low figure arises from frustration at Cyril Ramaphosa, but that he will likely be removed as ANC leader in the next 18 months (yesterday this newspaper showed how twice as many South Africans believe their lives will improve if he goes relative to those who think they will be worse off). Mr Ramaphosa is loathed in KwaZulu-Natal, where his party has polled at just over 10% at times. Now factor in the effect of a better ANC leader, if one materialises, and consider that for much of the past two years, with the immediate exception of the run-up to this election, the ANC was polling at around 40%: on a par with its 2024 national election number. Lift its KwaZulu-Natal vote count again, say in the event that the MKP fades post this election, and the ANC may prove surprisingly resilient.

Conclusion

At the moment, these are all minority cases. Back the data is the rule and the data say the DA should squeeze out a win in Cape Town, fall short of 50% in Gauteng, and the MKP should be the biggest party in Durban, there will be chaotic coalition talks broadly, and an ANC-DA coalition will only happen in Gauteng if the ANC countermands its regions and then agrees to reform pacts the DA is confident may hold. But in any volatile emerging market, even as you follow the rule, remain nimble enough to turn on a dime and retain enough capacity for doubt, so that results, when they go against the data, surprise and don’t shock.