Reports

This report is the second in a series tracking the political race in America ahead of that country's November 2024 presidential election. The data is drawn from a wide cross-section of polling organisations. The methodology applied to the data in this report was to award all states that polled within the margin of error in favour of Mr Trump to Mr Biden. All states that polled in favour of Mr Biden were awarded to Mr Biden regardless of margin of error. Only states that polled in favour of Mr Trump to a degree beyond the margin of error were awarded to Mr Trump.

This report investigates the energy policy preferences of registered South African voters. The data in the report is drawn from a survey of 1 512 demographically and geographically representative registered voters conducted in March 2023. That survey had a margin of error of 5%.

This report investigates political party support in South Africa across different home language groups. It does so via two different charts. The first chart measures what share of people in a particular home language group vote for a particular political party. The second chart measures what share of the support of a particular political party comes from a particular home language group. The data in the report is drawn from a survey of 7 196 demographically and geographically representative registered voters conducted in April and May 2024. That survey had a margin of error of 2%.

This report investigates political party support in South Africa across different age groups. It does so via two different charts. The first chart measures what share of people in a particular age group vote for a particular political party. The second chart measures what share of the support of a particular political party comes from a particular age group. The data in the report is drawn from a survey of 7 196 demographically and geographically representative registered voters conducted in April and May 2024. That survey had a margin of error of 2%.

This report examines the support amongst ANC and DA voters for an ANC/DA coalition. The data in the report comes from 3 previous Foundation surveys conducted in July 2022, October 2023 and April 2024. These surveys had sample sizes ranging from 1 412 to 3 204 and margins of error that ranged from 1.7% to 5%. The data demonstrates that there is considerable support amongst ANC and DA voters for the parties to enter into coalition as a government of national unity.

This report is a projection that was produced by the Foundation of the likely final election result after voting had concluded and the vote count had commenced. In each of the 3 charts below the trendline of the Foundation's tracking poll in the month ahead of the election is shown to the point where it concluded together with a Foundation projection of the final result conducted after vote counting had got underway.

This report estimates the May 29 South African general election results. It does so via inferring from the Foundation's tracking poll the trajectory of each key political party in the week ahead of the election and applying to that trajectory the margin of error of the poll plus an additional 1 percentage point (the extra point is to account for the continuing volatility we see in the most recent political data). In the charts below three lines are produced for each political party. The blue line is the figure towards which that party was tracking in the final week ahead of the election. The black line is the likely upper extreme towards which the party was tracking once the poll's margin of error was applied. The orange line is the likely lower extreme towards which the party was tracking once the poll's margin of error was applied (on the 58% and 60% turnout model charts for the ANC and the DA you will see a fourth, green line which is the projection of the final result that the Foundation developed after voting concluded). Be warned that in the final days before an election it may be customary to see tracking poll lines flatten as political opinion crystalizes. This has not happened in the case of this election for the reason that ANC voter opinion is so fractured on the fringes of that party that a share of voters continue to ping pong between the ANC and the EFF, MK and the DA in the main. This fractured ANC support has been a key insight generated by the tracking poll but necessarily diminishes the predictive qualities of the poll. At the Foundation we continue to see the media misreporting polling as forecasting. We must stress that in this election particularly significant swings in voter sentiment, even beyond the margins of error, will make it very difficult for any entity to call the result with precision.

This report explores a potential route to an ANC majority in the South African elections that will take place later this week. The case for an ANC majority rests on four insights. The first is that in both 2022 and 2023 the Foundation polled ANC support at a relatively stable near 50% mark – and on occasions even slightly above that. It was only the 2024 introduction of the MK party into the election that changed this driving ANC support into the mid- and even lower-40 percentiles (as MK support rose to just over 10%). The second is that after hitting a low point of near 40% roughly six weeks ago ANC support lifted steadily to break through the 45% mark about 10 days ago as the ANC campaign machine got into full swing (a long established phenomenon of South African elections). ANC support was subsequently driven downwards again by the joint events of the government’s NHI announcement and the Constitutional Court judgment that Mr Zuma was unfit to stand for parliament (an event that drove support for Mr Zuma’s MK party sharply upwards). The ANC has since held its final (successful) rally in Johannesburg and also commandeered the state broadcaster to address the nation and both of these actions should lift its support levels by an extent that the Foundation’s tracking poll may not fully digest ahead of the vote. The third is that the MK party is a new phenomenon and that its growth is carried by sentiment, which may be fickle, whilst the KwaZulu-Natal province, which serves as its home base, is difficult to poll. Trendlines of the past week have continued to put MK party support levels at just upwards of 10%. Whether MK will be able to turn out the vote to secure something approximating that number is a key uncertainty of this election around which the question of a surprise ANC majority pivots. The fourth is that trendlines for the DA and EFF at upwards of 20% and near 10% respectively are in line with their established numbers of several past elections. A minority call therefore, of the Foundation, is that if MK support flounders on the day of the vote to something near 5% to 7% that will free up around 5% of the vote, the bulk of which should flow back to the ANC and may thereby be sufficient to drive its final number to just over 50%. Whilst it is not probable, it therefore remains plausible, that the ANC may secure a majority. In the charts below the Foundation has compared and contrasted ANC and MK support numbers through May to demonstrate how MK underperformance may deliver an ANC majority.

This report examines the volatility in the final week of the Foundation's South African election tracking polls. Towards the final days of a tracking poll it should be possible to see the trendlines flatten out as political opinion crystallizes immediately before an election. However, towards the late stages of this tracking poll that is not happening. The reason for this is that an unprecedented degree of voter uncertainty about which party to support, as the ANC fractures, is causing a small share of voters to move back and forth between the ANC, EFF, MK and also the DA. Two significant events occurred over the past 10 days which appear to have particularly animated this trend. The first was the signing of the draft NHI legislation into law on 15 May. The ANC greatly miscalculated here in assuming that the signing would be read as a statement of social solidarity. However, given that several million middle and aspirant middle-class people make use of private medical care both via insurance products they purchase as well as via out-of-pocket expenditure, the announcement was broadly interpreted as an attack on middle and aspirant middle-class standards of living. This is a highly registered constituency and one in which the ANC was already particularly vulnerable to shedding support. In addition, there is scant evidence that the promise of a nationalised public healthcare service animates voting constituencies in lower socio-economic strata. The second event was the constitutional court ruling against former President Jacob Zuma on 20 May. The strength of Mr Zuma's MK party arises chiefly from the perception of his persecution. Mr Zuma has positioned himself as a fallen hero who is being persecuted by a distant and aloof ANC that has long neglected and even forgotten about core bands of its erstwhile supporters. The finding of the court that he was unfit to stand for parliament provided new impetus for this perception. The combination of these two events drove ANC support down, bucking the trend of a clear upward drift in ANC support of the previous 3 weeks, as both MK and the DA saw their support levels lift.