At the IEC’s Results Operation Centre on Thursday, pre-election public opinion polls started converging with election results. These polls were one of the few approximately and presumedly accurate pointers to a post-election future that South Africans had had in the last few months. Yet, there was no certainty that their predictions would survive the test of election time.
Public opinion polls internationally are known to be useful, albeit often imperfect, projections of future. They often do not give full or sufficiently accurate indications of the electoral choices that are to be made in events such as elections and referenda. Some of the most notorious cases include miscalculations in United States presidential elections and public opinion polls at the time of the United Kingdom’s Brexit decision.
In South Africa in the last six months, a range of public opinion polls concerning likely party choices in the then pending (and now just completed) national elections started emerging. Some were of sound methodologies, others were dubious. This brief analysis refers to the former. These sound polls predicted generally that the ANC would slip below 50 percent of the national vote (even well below it). Many citizens and party agents chose to question the polls. It was wondered whether the pollsters had collectively lost the plot.
There was a significant amount of convergence between the best of these polls. The surveys by Ipsos, Markdata and Social Research Foundation recorded support levels for the ANC of forty percent or just above. While stressing that surveys can never be more than snapshots in time, Ipsos anticipated 40 percent for the ANC, Social Research Foundation (SRF) 42 percent, and Markdata 43 percent.
The ANC publicly countered the data on its apparently declining support by rolling out big guns like former ANC presidents and deputy presidents. As campaigners, they had to instill enthusiasm and mass ANC affirmation, or at least create the perception that such sentiments were taking hold. In the final months of the campaign, South Africans were assured repeatedly that the ANC was running for nothing less than an outright majority. In the weeks preceding Election Day, the ANC diffused the publicity that its mass outreach and door-to-door campaigning has been bearing fruit. The 94,000-strong final Siyanqoba rally was presented as evidence of unparalleled, undented ANC support.
By all present indications, these actions were too little and too late. As the final 2024 iterations of the Ipsos, Markdata and SRF polls also anticipated, South Africa’s voters were looking for an alternative. Many of them chose a party that would demonstrate their anger with the ANC, and express alienation from and disappointment with the former liberation movement. The MK party would fit this bill for many. The precise story on their motivations will be told with the wisdom of full hindsight.
In the interim, and as we get increasing clarity on the exact voter choices, the mid-day Thursday Centre for Scientific and Industrial Research (CSIR) result projections were bearing out the pollsters’ predictions and confirmed that the pollsters’ calculations had been giving reasonably definitive outlines of the political future that is unfolding courtesy of the results from Election 2024.
Professor Susan Booysen is a political analyst, Director of research at MISTRA, and visiting professor at the Wits School of Governance
