To the casual observer, the outcome of South Africa’s 2024 national and provincial elections, which herald a new era of national coalition governance, may appear to represent new electoral trends. But that is not the case.
While the shift in voting patterns is certainly the largest the country has seen since democratic elections began, voter fluidity – people shifting their votes from one election to the next – has been a growing trend among the South African electorate since at least 2016.
The 2024 general election is a watershed moment in South Africa’s democracy. For the first time in 30 years, the African National Congress (ANC) failed to win the majority of votes needed to form a government, securing only 40.2% of the vote.
The opposition Democratic Alliance won 21.8%, roughly the same as in 2019 (20.8%). The Economic Freedom Fighters’ vote share declined slightly from 10.8% in 2019 to 9.5% in 2024.
The disrupters of this election, the new uMkhonto we Sizwe Party (MK Party), gained 14.6% of the vote nationally. The party secured the largest share of the vote in the province of KwaZulu-Natal (45.9%), just short of an outright majority.
As sociologists, we have been researching voter choices over the last decade, by conducting exit polls and post-election surveys. We’ve looked at who votes for what parties, who chooses to abstain and how people explain these choices. This data has given us insight into past patterns and fluctuations.
We are conducting further research into the 2024 elections, but our past research already provides us with insight into which layers of the electorate are party loyal, the extent to which people switch their vote between parties, and who chooses not to vote at all. These past findings help to shed light on two important trends that were evident in the 2024 elections.
Read more: Uncertain times for South Africa's foreign policy as country heads for coalition government
First, greater voter abstention as 41.6% of the registered voter population chose not to vote. This is an increase of 7.6 percentage points from the 2019 national elections when 34.0% chose not to participate, despite being registered to do so.
Second, more voters turned away from the ANC as its support declined from 57.5% in 2019 to 40.2% in 2024.
What we did
In 2021, in the weeks following the local government elections, we conducted a post-election telephone survey with 3,905 registered voters living in five metropolitan municipalities: Johannesburg, Tshwane (which includes Pretoria), Ethekwini (which includes Durban), Cape Town and Nelson Mandela Bay (which includes Gqeberha). While this does not give us a nationally representative sample, it does provide a representative sample of important urban concentration points of the population.
The survey asked people about their voting histories across the last three elections: the 2016 local government elections, the 2019 national and provincial elections and the 2021 local government elections.
We were interested in the extent to which “race” played a role in shaping the vote choices of the South African electorate because of the way in which race continues to shape important elements of life post-apartheid.
Racialised fluidity
To understand how race may or may not shape voting choices we use the term “racialised fluidity”. We use the term “racialised” to highlight that, at the aggregate level, Black African, Coloured, Indian and white people tend to vote in distinct ways. However, we also find that many individuals act differently from one election to the next, hence “fluidity”.
From our 2021 research, we found that 76% of Black Africans in our sample had voted for the ANC at least once between 2016 and 2021. This contrasts sharply with other population groups: only 27% of Indian respondents, 22% of Coloured respondents and 7% of white respondents voted for the ANC at any time during the same period.
In contrast, we found that 93% of white respondents, 78% of Coloured respondents and 73% of Indian respondents had voted for the opposition Democratic Alliance at least once between 2016 and 2021.
We found that white voters were the most loyal to a party in the South African electorate: 52% of white respondents voted for the Democratic Alliance in all three elections. Smaller proportions of Coloured and Indian respondents demonstrated loyalty to the Democratic Alliance, 25% and 13% respectively.
Read more: South Africa’s election results present 3 options for government: all are fraught with danger
In comparison, only 32% of Black African respondents were loyal to the ANC across the same time period. A quarter (25%) of Black African respondents had never voted for the ANC between 2016 and 2021. This group was more likely to be younger, more educated and more affluent than those who had voted for the ANC.
While three quarters (76%) of Black African respondents voted for the ANC at some point between 2016 and 2021, this did not mean that they were always loyal ANC voters: 44% moved between voting for the ANC, abstaining or voting for another party.
In 2021, over a quarter (27%) of Black African participants who had previously voted for the ANC chose to abstain, rather than vote again for the ANC, while 8% left the ANC for an opposition party. We found that those who chose to abstain after voting for the ANC were more likely to be younger voters.
Looking across all racial groups, we found that 50.3% of our respondents had fluctuated between voting for the ANC, opposition parties or abstaining. Voter fluidity is not a new phenomenon.
These findings help us to interpret some of what has happened in the 2024 elections.
Two things appear to have happened. First, fewer people turned out to vote. Of the likely 42 million people eligible to vote, only 27.8 million registered and only 16.3 million voted. Turnout among the eligible voting age population was only 38.8%. This is lower than it was estimated to be in 2019 when the turnout among the eligible voting age population was 49.0%. For a possibly sizeable section of the electorate it was likely that they preferred not to vote at all than vote for a party in opposition to the ANC.
Second, some former ANC voters are likely to have switched their vote to an opposition party. We will need to wait for the data from the 2024 survey to confirm this and to understand where voters shifted their votes to. However, given that 69.0% of all the votes cast for the uMkhonto we Sizwe Party were made by voters from KwaZulu-Natal, an areas in which the ANC had a majority, it is likely that many voters moved from the ANC to uMkhonto we Sizwe.
Future fluidity
The 2024 national election results seem to suggest that voter fluidity may have increased with voters having a wider selection of opposition parties to choose from than ever before.
Whether uMkhonto we Sizwe can retain its electoral gains remains to be seen. Other opposition parties, such as the Congress of the People and Action SA, have shown that electoral gains can be difficult to build and sustain.
Of concern, though, are the increasing sections of the electorate who chose to exit the process. However, our research shows that those who choose to abstain in one election often return to the ballot box. So it is possible for political parties to win voters back – if they respond to the needs and desires of the South African people.