Illustrative image | President Cyril Ramaphosa and DA leader John Steenhuisen. (Photos: Gallo Images / Jeffrey Abrahams | Gallo Images / OJ Koloti | Rawpixel) 
As we come closer to the anniversary of last year’s elections that saw the ANC falling to just 40% of the vote, there is a flurry of public opinion polling.
Last week, the Brenthurst Foundation reported that a large majority of people support the national coalition that currently includes 10 parties, including the ANC and the DA.
The Social Research Foundation has recently found that while more people still support the ANC rather than the DA, this would change if President Cyril Ramaphosa were removed in a confidence vote.
Now, the Institute of Race Relations (IRR) is reporting that it has found the DA wins the support for 30.3% of those polled, while the ANC wins 29.7%.
Of course, it is well known that polling can be political. For years, political groups have manipulated questions, numbers and all sorts of other data to try to suggest there is support for them, or their cause.
That is why there is so much debate about them.
It may be important to fully understand what can be trusted and what cannot, which is why this guide by Professor David Everatt is so important.
How this poll was conducted also matters.
As the IRR explains it:
“A total of 807 respondents participated in the poll, comprising a diverse demographic cross-section. The results have a margin of error of ±4% at a 95% confidence level, indicating that the findings are highly reliable and represent public opinion within this range. Data were collected using Computer-Assisted Telephonic Interviews (CATI), a reliable method that ensures consistency in questionnaire administration and minimises interviewer bias. The survey was limited to registered voters, ensuring the data reflected the electorate’s views.”
There is a lot to consider with this.
First, it was not door-to-door, which can make polling more reliable but is a lot more expensive. Also, it means these are the views of those who were phoned, answered, and agreed to talk.
Perhaps the biggest variable is that it only considers registered voters. The number of people who are eligible to vote but are not registered to vote has been growing strongly in recent years.
A poll like this might exclude those who could be encouraged to vote by events, or by a new party.
At the same time, it is probably wasteful and misleading to include the views of people who have not registered to vote, as they are very unlikely to do so in the next elections.
Read more: Next steps: ANC’s plan to navigate Budget impasse and restore coalition trust
The economics of politics
In its commentary on the polls, the IRR explains that the timing of its fieldwork (when people were actually phoned) is critically important, because it was in the days immediately after the current fight began over the Budget, and particularly VAT.
In other words, the main issue of the moment was VAT, and the ANC’s stated intention (at the time), to raise it by two percentage points.
In the view of the IRR, this might indicate a shift in our politics, away from race (and the long-running dynamic of the ANC vs the DA) and into the domain of economics and economic growth.
Certainly, it could be argued that this should be the real debate in our politics. The fact that so many young people are unemployed and hold no hope of getting a job is the biggest long-term problem we face.
But it is of course difficult to know what the real motivations of respondents to the poll were.
Media reportage of that period was indeed dominated by the discussion around VAT. And it is also true that it would make everyone poorer (and feel poorer).
But it is also not certain what impact the media really has.
A very large number of people in South Africa will have access only to radio news, and the impact of the English-speaking media can be overstated.
That said, it is also true that in almost every part of the country, almost all South Africans went from having never worn a face mask in public to automatically wearing one in the space of around 10 days.
This was surely because of the media. No other institution can reach so many people so quickly.
It should also be remembered that it is not only the DA that opposed the VAT increase. Even before the Budget, Cosatu issued a statement saying there should be no VAT hike. The SACP, too, has said publicly this should not happen (and the SACP is now going to contest elections independently of the ANC).
Now, as the debate has moved on, virtually all parties oppose the VAT hike. As News24’s Carol Paton has reported, it now appears even the ANC is giving up on the idea.
If the VAT debate was a factor in the outcome of this poll, it again raises the question: why did Finance Minister Enoch Godongwana think it would fly? He appeared to have been curiously disengaged before his first attempt to present the Budget. It must have been obvious to some of those around him that a two-percentage-point hike would be opposed.
At the same time, this poll does raise other questions.
A DA breakthrough?
In the IRR’s commentary, it suggests that around 18% of black voters now support the DA. In terms of black voters, the ANC still receives the highest number, followed by MK, then the DA, and then the EFF.
It would appear that it is in this group of voters where the DA’s increase in support has come from. Intriguingly, this appears to almost go against what has appeared to be DA policy over the past few years.
While it made a deliberate choice to attract more black voters while it was led by Mmusi Maimane, it seemed that was not the main aim during the John Steenhuisen era.
While its members are incredibly diverse, and possibly a majority are black, its leadership structure is overwhelmingly white (these are people who are elected into their positions by party members).
And, during elections, as has been pointed out before, it seemed the DA had decided to concentrate on the minority vote while the parties that received support from black voters continued to splinter.
In keeping the white and minority vote together, the DA strategy seemed to be to watch the fracturing of our politics and emerge, perhaps, as the biggest party in Parliament.
Read more: Budget bust-up forces ANC to rethink GNU rules, Zille says forget it
All of that said, perhaps the DA will one day look at this as a breakthrough moment, when at last it was able to consider beating the ANC.
For the ANC, this might be a wake-up call, a reminder that it seems as if its leaders are not focusing on the bigger picture.
Instead, it can sometimes appear as if some people are focusing on their internal ambitions rather than saving the party.
This poll might also show that the space in what could be called the “moderate middle” of our politics is about to become more contested. In other words, the DA and the ANC, as the two most diverse parties in the country, will be contesting for the same voters in a very interesting way.
In the immediate short term, this might well play into the DA’s hands in the current dispute over the Budget.
It might give them confidence and even weaken the hand of the ANC at a crucial time.
Momentary snapshot
However, it is wise not to get too excited.
As the IRR has said in its commentary (and other polling institutions have said in theirs in the past), a poll does not predict the outcome of an election.
Even a perfect poll is merely a snapshot of political preference at one moment in time. And, 807 respondents cannot really give a proper snapshot of the entire country.
However, until the local elections next year (or in early 2027), they are as good as we’ll get.
Considering that our politics is beginning to change more quickly than ever before, the information that polling gives us will become more and more important. DM