Support for political parties
6 February 2024
In South Africa, the 2024 National and Provincial elections are eagerly awaited by many. A record number of political parties are already registered with the Independent Electoral Commission (IEC) and voters will be spoiled for choice. Yet, a third (35%) of South Africans who are registered to vote say that there is NO political party that represents their views and there is much uncertainty about the future and apprehension about the possible outcomes of the elections.
With President Cyril Ramaphosa expected to announce the election date during Thursday's Parliament opening ceremony, it is prudent to assess the current level of support for political parties before the onset of fully-fledged election campaigns.
Questions linger regarding the potential impact of voter turnout on the perceived legitimacy of the elections. To address this, the IEC conducted another special voter registration weekend on 3 and 4 February. Political parties, politicians, and analysts keenly observe the additions to the voters' roll following this initiative. Presently, approximately 42.3 million South Africans aged 18 years or older are eligible to cast their vote. However, as of Monday, February 5, 2024, the IEC website indicates that only 27,173,114 South Africans are registered to vote. Thus, almost 64% of those eligible to vote are registered to vote, leaving about a third of adults who have chosen not to exercise their democratic right to vote by registering.
Support for political parties
Ipsos conducted its latest representative, face-to-face poll in South Africa from 23 October to 1 December, 2023. One of the key questions posed was: "If there were National [or Provincial] elections tomorrow, which political party would you vote for?"
In this Pulse of the People™ series respondents are requested to select their choice(s) on an imitation ballot paper on a computer, before handing the computer back to the interviewer. This approach replicates a real ballot paper process and ensures the confidentiality of the vote. Ipsos incorporates an extensive array of possible party choices in the questionnaire; nonetheless, not all parties registered with the IEC are featured on the Ipsos list. Inclusion is determined primarily by past electoral performance rather than media presence or the prominence of party leadership. However, respondents have the option to write in the name of any political party they wish to support, even if it's not listed.
Fieldwork for this poll was conducted prior to the establishment of the uMkhonto we Sizwe party by ex-president Jacob Zuma on 16 December 2023.
The graph below shows the party choices of South Africans on the national ballot paper. Results for those who have indicated that they are registered to vote are given:
In the data depicted, the EFF performs marginally better than the DA, potentially positioning itself as the “official opposition”. However, these results should not be taken at face value as the figures include 10.1% (one in every ten) of the registered electorate who have not aligned themselves with a particular political party, indicated as "will not vote", "refuse to answer", or "don't know". These responses are all valid answers to the question asked, and delving deeper into this data can provide valuable insights to be able to analyse the “undecided vote”. Simply disregarding these responses and reallocating the one in ten uncertain, or rather unwilling, voters to other political parties would be comparable to disregarding people's democratic right to make their own choices.
It's crucial to acknowledge that individuals who abstain from expressing a preference might not align with the majority of the electorate on voting day. Consequently, additional analysis is necessary. The questionnaire also delves into matters of party choice, preference, or affiliation through other inquiries. These findings aid in discerning the party selection of individuals who haven't explicitly indicated their choice on the ballot paper. These results are representative of the party choices of the registered electorate if an election were held the day after the interview.
The additional analysis of the data reveals:
- The support for the ANC is well below 50%.
- The DA is moving back into the position of official opposition.
- The EFF is still challenging this position.
- The IFP support on a national level is in the region of 5% - but mainly coming from KwaZulu-Natal where the party is a strong contender.
- The Action SA support is mainly located in Gauteng.
Looking at a different analysis of these results, the Multi-party Charter for South Africa looks to draw about 33% of the vote at this stage.
It's critical to emphasise that these results DO NOT CONSTITUTE AN ELECTION PREDICTION. Credible election predictions can only be made in the final stages leading up to the election. Various factors, including campaign dynamics, political and economic changes, as well as the state of essential services like electricity and water provision, can influence the election outcome in the final months. It is prudent to monitor all these developments closely before drawing any conclusions about the final election outcome.
The pivotal factor in this election hinges on voter turnout on the day of the election. Several analyses indicate that anticipating a high voter turnout may not be realistic, largely attributed to widespread despondency regarding the country's situation and low levels of trust in politicians and political parties. Conversely, a low voter turnout could potentially benefit the ANC, potentially elevating ANC support close to the halfway (50%) mark. This scenario relies on variable voter turnout across metropolitan and rural areas. A higher turnout in rural areas would favour the ANC, given its predominantly rural support base.
According to Ipsos’ current modelling, a low voter turnout will mean that in the region of 39% to 41% of the registered electorate turns out to vote; a medium turnout that between 55% to 57% of the registered electorate turns out to vote; and a high voter turnout that between 69% to 71% of the electorate turns out to vote. (These proportions have increased from the previous calculation of possible turnout in October 2023).
Calculating support for political parties according to the turnout scenario’s is possible:
This suggests that the possibility of a national-level coalition is conceivable, though not guaranteed. In the event of such an election outcome, the ANC would only require a party with about 4% to 6% national support as a coalition partner to establish a national government. A coalition involving just two parties could streamline negotiations and potentially enhance the coalition's effectiveness. However, it's essential to stress that this is purely speculative, and numerous further studies will be conducted in the lead-up to the 2024 election.
Follow Ipsos in South Africa for the latest updates on the 2024 election polling. Our next Ipsos press release will delve into the Trust/Mistrust in political leaders and parties.
About the study
3,600 face-to-face interviews were conducted by trained Ipsos interviewers in the homes and home languages of respondents. Interviews were conducted from 23 October 2023 to 1 December 2023. A scientific process of multi-stage stratified random selection distributed interviews in all areas of the county, including deep rural areas. This methodology ensured that the results are representative of the views of the universe (South Africans eligible to vote) and that findings can be weighted and projected to the universe. Trained quantitative fieldworkers from all population groups were responsible for the interviewing and CAPI (Computer-Assisted Personal Interviewing) was used. All results were collated and analysed in an aggregate format to protect the identity and confidentiality of respondents. Using a ballot paper like that used in an election, respondents had to “vote for” their choice of political party. The question specified that they needed to consider their choice as if the election were happening the next day. All results were aggregated and then analysed as described. All sample surveys are subject to a margin of error, determined by sample size, sampling methodology and response rate. The sample error for this sample at a 95% confidence level is a maximum of plus or minus 1.8%.
