NEWS ANALYSIS: By-election results predict how May 29 vote will go

With the 2024 election less than two weeks away, everyone is looking for clues about the results.

By-elections have historically revealed trends that are later reflected in general election results: the growth of the FF+ at the DA’s expense in the 2019 national election was foreshadowed by a series of by-elections that showed the FF+ gaining as the DA lost votes.

Could this current round of by-elections, measured from the 2021 local election, hold similar clues? It has been two-and-a-half years since that election and there have been by-elections in 218 wards, or about 5% of the total: 138 ANC wards, 51 DA, 15 IFP, three GOOD wards, two EFF wards, two independent wards, and seven other small parties each defended one ward.

By-election winners and losers

The Sankey diagram below shows the flow of ward seats from incumbent parties after 2021 to the winning parties in the by-elections. Hover over a single path to see the number of wards moving from one party to another.

Most of the wards were retained by incumbent parties. The ANC won 105 of the 138 wards it defended, the DA won 44 of its 51 wards, the IFP kept 13 of 15 wards. Some parties fared poorly: GOOD lost all three of its wards and the two independents, the Karoo Gemeenskap Party (KGP), Al Jama-ah, the PA and the FF+ all lost the ward they were defending.

The EFF also lost one of the two wards it defended but it more than made up for this by taking 10 wards off the ANC. The PA ended up six wards ahead by taking three wards from the ANC, three from the DA and one from GOOD.

The ANC was the biggest loser overall ending the period with 27 fewer wards. In addition to the 10 seats it lost to the EFF and three to the PA, it lost 15 wards to the IFP. The DA had a net loss of just one seat, paring its losses to the ANC and PA with gains from both parties, GOOD, and from the KGP. 

The biggest winner was unambiguously the IFP with a net gain of 15 seats. Most of these came from the ANC but the party also took wards from an independent and from Al Jama-ah. 

Measuring trends beyond ward seat movements

Only two out of five wards changed hands, so focusing on these wards does not tell the whole story. In 61 wards — almost 30% of the total — the winning party received less than 50% of the vote. In 23 wards the winning party received under 40%. Wards are increasingly competitive, particularly in KwaZulu-Natal and the Western Cape. 

Another trend to watch is the low turnout in DA safe wards. The scatter plot below shows the relationship between turnout in the 2021 election and the respective by-elections. The lower each point on the graph is vertically, the lower the by-election turnout. (Hover over a single dot for more information or use the dropdown in the top left to view individual parties.)

Turnout in ANC wards was below 30% in just nine cases, or in 6.5% of the party’s wards. For the DA it is 21 out of 51 wards, or about 40% of the party’s wards. This could indicate very soft support for the party, particularly in Cape Town and the three Gauteng metros where the party’s base has traditionally been the strongest.

Other trends and conclusions

The by-election results have also picked up the explosive arrival of the MK party in KwaZulu-Natal and the recent plateau of the party’s momentum.

By-election results in KwaZulu-Natal and Mpumalanga earlier in 2024 showed the party enjoying support of more than 20% in wards it contested while recent results have been less impressive.

The growth of the PA in the Western Cape and Gauteng is also reflected in recent polls that show the party growing its national share.

By-election results remain a useful indicator of voter sentiment. If the DA’s results on May 29 are softer than the current polls suggest, and if the horse races become closer across the country’s wards and voting districts, the clues will have been present over the last two years. 

• Berkowitz is the Director of The Third Republic