Dawie Scholtz | 2022 by-elections: ANC support in free fall


The ANC’s vote share is collapsing with black voters in and outside of KZN, writes the author.

The ANC’s vote share is collapsing with black voters in and outside of KZN, writes the author.

Alet Pretorius

Election analyst Dawie Scholtz reflects on the 89 ward elections that took place in 2022 to see what they reveal about the country’s electoral trajectory. He writes that ANC support appears to be in free fall and it is an open question as to whether the newly-elected leadership will be able to do anything in time to turn the trend.  

The old saying goes that two things are certain in life: Death and taxes. After assessing the by-elections that took place across South Africa during 2022, I’d say three things are certain: Death, taxes and electoral change. 

Eighty-nine “analytically-useful” ward elections took place in 2022, encompassing 170 588 voters. In my view, these by-elections act as a non-random sample of the electorate that provides clues to the country’s electoral trajectory. Below I’ve split the sample by demographic group and explain the emergent insights in detail. In summary, the key messages are:

  • The ANC’s vote share is collapsing with black voters in and outside of KZN, but it is especially pronounced in the northern provinces outside of KwaZulu-Natal. The Inkatha Freedom Party is growing its support in KZN, and the EFF is growing rapidly in the northern provinces (at the expense of the ANC).
  • The DA is recovering with coloured voters (it had enormous losses with coloured voters in 2021), but not yet fully up to the level of support it enjoyed with coloured voters in 2019. The Patriotic Alliance is growing at the expense of the DA in this part of the electorate.
  • There is an intensive tug-of-war between the DA and Freedom Front Plus for white Afrikaans voters. The effect is uneven, but on aggregate, the FF+ is growing steadily beyond the level it achieved in 2019, at the expense of the DA.
  • All of these trends are not counting in the effect of ActionSA, since ActionSA did not contest many by-elections in 2022. ActionSA had a significant impact on voting patterns where it contested in Gauteng last year (taking from both the DA and ANC), but it remains largely untested outside of Gauteng, with significant uncertainty about its potential impact.

Red flag for ANC

Here we have a large sample of 52 wards, with 85 000 voters voting in by-elections in 2022. Across all of these wards, the voting pattern was as follows: 

It should be noted that this analysis includes only the ANC and EFF since they are the only two parties that contested all of the wards in question. ActionSA, the DA, the UDM and others contested in some areas, but not significantly enough to be included in the analysis. 

On aggregate, these numbers should serve as an enormous red flag for the ANC. This is the core ANC electorate. Anything under 70% in this part of the electorate signals a big risk of being sub-50% nationally. At 53% in this sample, reflecting a 14% drop since just last year, the ANC’s national majority is at enormous risk. If the pattern were to continue into 2024, and at scale, the loss of the ANC’s national majority would be an absolute certainty. 

There is one interesting undercurrent to point out, which is that the ANC appears much more resilient in the Eastern Cape sub-sample than in the northern provinces. Below are aggregated results from nine overwhelmingly black Eastern Cape wards, encompassing 23 000 votes, in by-elections this year.

The ANC slide here is much slower, and it is increasingly looking like the Eastern Cape will be critical to the ANC’s attempt at defending its national majority in 2024. High Eastern Cape turnout, with high ANC margins, will be non-negotiable.

A final note on the sample is that there are zero wards included from Limpopo and the Free State since there were no relevant by-elections there this year. Limpopo is especially interesting and important since it has traditionally been, like the Eastern Cape, a bastion for the ANC. Limpopo by-elections next year will be particularly interesting and important to observe as this is an important uncertainty at this point.  

Surge for IFP

Here our sample includes 16 wards from across KZN, including 42 000 votes from by-elections this year. We also added the IFP in since, in addition to the ANC and the EFF, the IFP contested every ward. The results across these wards over time are:

Once again, shocking numbers for the ANC and a significant surge for the IFP (note: this excludes wards that are not analytically useful, for example eThekwini ward 99 where the DA stood back and endorsed the IFP). The EFF would also not be delighted with these figures.

If this trend were to continue and be replicated across KZN in 2024, it would similarly be an absolute certainty that KZN would be governed by a coalition of some kind. 

The DA’s FF+ problem

Here we have a sample of seven wards where both the DA and FF+ contested, with 17 000 votes cast in 2022 by-elections. 

The DA’s FF+ problem has not gone away, and in fact, it is arguably growing. The DA appears to be doing slightly better than the 2019 level (which was disastrous), but it is nowhere near the 80%+ levels it had historically achieved before 2019. A technical side note is that there is a small contingent of ANC and EFF voters in these wards that usually do not turn out in by-elections. But once they do, both the DA and FF+ % will come down slightly, implying the DA may actually be close to or even below the 2019 level in these wards. 

It should be noted that the DA has done very well this year in white English-speaking areas. In the four overwhelmingly white English-speaking wards contested this year (15 000 by-election votes), the DA did well:

However, ActionSA did not contest these, and as such they can be considered largely uncontested and therefore less meaningful. Based on the 2021 election results, where ActionSA was the second-largest party with white, English-speaking voters (winning 10-20% of these voters where it contested), it can reasonably be expected that the DA’s primary competition in this part of the electorate in 2024 will be ActionSA. We would need to see some DA-ActionSA by-elections in these kinds of wards to get a clearer view of the trajectory. 

Homework for the DA

This is a very interesting portion of the electorate at this stage. In 2021, we saw a dramatic fragmentation of the coloured vote, away from the DA and to many smaller parties. The PA also emerged as a significant player in some parts of the country, outright winning in many coloured communities in the northern part of the country. 

Here we have a sample of seven relevant wards, encompassing roughly 11 000 votes this year. The pattern emerging across these wards is as follows:

As this sample shows, in line with the trends, we saw nationally in 2021, the DA had a terrible election with coloured voters in 2021, but has recovered well in the by-elections in these wards this year. However, it should be noted that the PA has also done well and most likely continues to be the primary competitor to the DA in this part of the electorate. For the DA’s 2024 effort, it will be critical to get back to (or try to exceed) the 2019 level with these voters. They do not appear to be there yet.

The coloured electorate also appears to be increasingly clustered into four sub-clusters with different patterns emerging: Coloured voters in the northern provinces (where the PA is strongest), coloured voters in the Northern Cape (where there is a tug of war between the DA and the ANC), coloured voters in the rural Eastern and Western Cape (where the DA is strongest, but where the PA appears to be growing) and coloured voters in the urban Eastern and Western Cape where the DA is historically very strong but from where we have very little recent data, (presumably, the DA remains very strong in these areas but with a threat of PA inroads). A very interesting and dynamic situation!

For the record, we don’t have enough data from overwhelmingly Indian areas to be able to conduct a similar analysis. 

Given all of the above, a few concluding thoughts on each of the top six parties:

  • ANC: Support appears to be in free fall and it is an open question as to whether the newly-elected leadership will be able to do anything in time to turn the trend. In my view, the change that is taking place is being driven by weak economic and governance outcomes – not something that can be changed quickly.
  • DA: The DA appears to be recovering with coloured voters and is largely stable with white voters (the FF+ problem has not gone away). The party’s only significant growth avenue remains with black voters, where it remains unclear how the party is performing given that it isn’t contesting many by-elections in these areas. The results in Ditsobotla were not promising in this regard, but in a few others (Nyanga, Tarkastad) the numbers were better. The party’s performance with black voters will be the key thing to watch in 2023 and 2024, especially once ActionSA also starts contesting everywhere. 
  • EFF: The EFF appears to be growing in the northern provinces, but not yet so much in KZN, the Eastern Cape and the Western Cape. The EFF’s potential appears higher than previously anticipated, so we need to track this closely. 
  • IFP: The IFP is surging in KZN and will be delighted with the outcome of the ANC’s elective conference (which once again left the KZN ANC out in the cold). There’s clearly significant potential for the IFP to grow in KZN in 2024, and given the size of KZN, this could have national implications. 
  • ActionSA: ActionSA remains largely untested outside of Gauteng. It had good results this year in the two key by-elections it contested in Gauteng (growing further in both Soweto and a suburban area of Tshwane). However it had relatively weaker results in two by-elections it contested in the Eastern Cape. Probably the single biggest question for 2024 is whether ActionSA will be able to replicate the disruptive effect it had in the Gauteng 2021 local government elections (taking significantly from both the ANC and the DA) in other provinces outside of Gauteng. The party appears to have spent most of the year building structures and attempting to increase name recognition. Time will tell if it pays off. 
  • FF+: The FF+ appears to be at least steady and potentially growing on aggregate in the white Afrikaans-speaking electorate. The trajectory of this trend will be an important variable in the DA’s overall showing in 2024. There is perhaps an interesting, very preliminary question forming on white Afrikaans voters in the Western Cape and the extent to which the “Cape Independence” message is helping the FF+ disproportionately with that sub-pocket of the electorate. At this stage just an interesting question for observation based on only a very small sample of data.