Three recent polls of party-political support are confusing and contradictory. With tectonic shifts in South African politics possibly under way, polls are in demand to see how we stack up and predict our possible direction of travel.
Polls can be spot on, but also far out. Like weather forecasting, polling gets more accurate the closer to the event. But we are still two years away from the election, and in politics that is a very long time.
ANC support is diminishing, but falls of the scale shown by a poll leaked to Sunday newspaper Rapport and another conducted by market research firm IPSOS are very steep. The poll results in Rapport say ANC support is only at 38 percent. That is nine percentage points below the 47.5 percent the party achieved in the local government elections held less than a year ago. IPSOS polls ANC support at 42 percent, which is also a sharp drop on last year.
A poll conducted last month by the Social Research Foundation (SRF), a recently established policy think tank, shows ANC support at nearly 47 percent. And if the assumption is made of a turnout short of 56 percent, the ANC would obtain 52 percent of the vote. That could mean the ANC loss of support is not happening as fast as the poll leaked to Rapport and the IPSOS poll have found
And to add further confusion, the poll leaked to Rapport measures DA support at 27 percent. That is seven percentage points above what the party managed to get last year. The rise in DA support polled is mainly due to defections from the ANC, the Rapport poll indicates. But the IPSOS poll says DA support is only at 11 percent and the party cannot lure the ANC defectors. An SRF poll measured DA support at 25 percent, the result of a recovery.
The IPSOS poll points to former ANC voters tending not to vote rather than switch to an opposition party. All polls see the EFF as being pretty much stuck in its current position of a little over ten percent.
How does one go about sorting through these mixed signals about our political future?
The poll by Rapport and that by IPSOS show that fundamental, albeit different, sorts of change in the political landscape are not far away. But if the SRF poll is considered, change will be slower than suggested by the Rapport and Ispos polls. Given the margin of error, the issue of a coalition government in 2024 is still in the air if the results are close to those in SRF poll.
All polls should carry user warnings. While often seen as predictions of election results, they are instead snapshots.
It is unwise to view a snapshot as a definitive sign that the ANC will be out in 2024. There is no reason the ANC might not enjoy a temporary and partial recovery. The ANC tends to gain a boost just before elections. It will come up with a recovery plan with lures of bigger grants and turning around Eskom. It has recently spoken about a wealth tax to fund a Basic Income Grant and has been talking about foreigners being the cause of South Africa’s problems.
And we certainly need to know what methods a poll uses to assess its use.
Rapport describes the poll it reported on as “confidential” and says nothing about its source, the methods used, and who was polled. With the nine percentage point loss in ANC support measured by the poll, it might have been conducted in certain urban areas where the DA is strong. With these great unknowns, the credibility of the message that the ANC stands at 38 percent and that it is bleeding support to the DA is greatly diminished.
In the UK it is a violation of the country’s press code to run the results of a poll in the media without an explanation of the methods that were used.
IPSOS fully declares its methods. Some 3 600 South Africans who were 15 years old or older were randomly selected, and the margin of error is between 0.5 and 1.67 percent on a 95 percent confidence interval. This was an omnibus poll that asked all sorts of questions and was not only aimed at determining political party support alone. IPSOS interviewed people in person. That has its uses, but it is difficult to get into the protected townhouse complexes of the middle class and the deep rural areas of KwaZulu-Natal where Inkatha finds support. IPSOS does not calculate its party support on the basis of all those who are likely to vote, but rather as a share of its total sample, which also includes those who refused to answer and are not registered
Omnibus polls are viewed as controversial by some political pollsters who say they are too broad, with possible problems with accuracy arising. IPSOS says there is often a correlation between the questions that are posed, and it has a proven track record of accuracy. The company says it has written a detailed reply to the DA dismissing the party’s claim of IPSOS bias in its polling.
And some political pollsters also say that having registered and unregistered voters in the sample is problematic. After all, only registered voters can vote. IPSOS says that it filtered out those in its sample who were under 18, and hence too young to register. With the election two years away it says there is still time to register, and its political polls closer to the election will only include the registered.
The SRF interviewed 3 204 randomly selected people over the phone and it has a margin of error similar to that of the IPSOS poll. Only registered voters were polled by the SRF. The SRF is a small privately funded policy think tank that commissions polls and releases policy papers. The former Institute of Race Relations CEO, Frans Cronje is a director.
Cronje says one key message from the poll by the SRF is that: “We should not be too fast to call the end of the ANC. It is dying, but not yet dead.”
There are also a lot of factors unknown to us now that will emerge as we get closer to the election. “There will be two or three political parties launched before the election in two years’ time, and these will be parties positioned to the centre right of the ANC. We do not know the impact they will have on the ANC support base. So we are not really all that sure what will happen in the 2024 election,” says Cronje.
Key will be where former ANC supporters go. A Soweto focus group conducted by the SRF provides some insight into their thinking. Leaving the ANC has been similar to a divorce in the trauma it has caused them. They are disillusioned with politics politicians . They find that their old party still tugs their heart strings. They say they have respect for some opposition parties, but are worried that they will turn out like the ANC and be full of talk and no action on the key issues like jobs and unemployment. They also don’t like constant bickering among the opposition.
So, will they give the ANC one last vote in 2024?
If that happens on a large scale, there will be no opposition coalition that takes over. We will then have to wait until 2029 for change. In theory, says Cronje, that may not be a bad thing as it will give the current opposition five more years to prepare for government.
“It is best to make sure that opposition is ready for government before it beats the ANC.”
The views of the writer are not necessarily the views of the Daily Friend or the IRR
If you like what you have just read, support the Daily Friend