If we are to believe the opinion polls, the African National Congress will sweep to power with a two-thirds — or even a 70% — majority on April 14.
How reliable are these surveys and what impact will these polls have on the undecided voter? Will they galvanise opposition supporters to cast their ballots and try and make sure that their parties are strengthened in the poll? Or will they result in ANC supporters not bothering to head to the polling station on the day because the organisation has the election in the bag?
Political surveys are there to inform the public about what other people are thinking, says Mari Harris, a director of market research agency, Markinor.
However, she cautioned that for a poll to be credible, the methodology used in collecting the information must be sound, the questions posed to interviewees must be unbiased and the sampling must be random. Importantly the interpretation of the interviewee’s responses must be above-board, not influenced by a political agenda. She also warned that a poll’s results only represent a ”snapshot” of what people are thinking at a particular time, pointing out that a lot could change between the time the survey is held and the day of the election.
The Human Sciences Research Council’s Dr Udesh Pillay, who describes opinion polls as a reliable indicator of voter intention, says even simple things can throw out predictions.
”Some people may wake up feeling totally apathetic and decide not to vote. Some may decide that the voting station is actually too far and decide not to vote because they don’t want to incur transport costs. The weather can also be a factor. When it rains, for example, a person with a car may still drive to the voting station whereas it is not so given for a person without a car,” he said.
Generally, political analysts regard surveys as quite accurate indicators of election results. In April 1999, an SABC/Markinor poll predicted that the ANC would secure 66% of the vote; the final result was 66,35%. The New National Party prediction in that survey was 6%, and it ended up with 6,87% of the votes.
On the other hand, the poll predicted the Democratic Party /Democratic Alliance would only get 7%, and it walked away with 9,56%; while Inkatha Freedom Party support was forecast to be 4% — it got 8,58%. These differences point to the importance of analysts understanding the peculiar dynamics around each party when they interpret the numbers produced by opinion polls.
IFP leader Mangosuthu Buthelezi has pointed out that the opinion polls that predict the ANC will win KwaZulu-Natal are totally unreliable. He correctly insists that in almost every past election IFP support has been underrated — partly because of the difficulties pollsters have in reaching his supporters in the rural areas of KwaZulu-Natal.
Despite these problems, political parties take opinion polls very seriously and their results definitely influence individual voters, says research consultant and University of Port Elizabeth politics professor Susan Booysen.
”There is a phenomenon of ‘bandwagon effect’ — some voters want to be associated with winners and would therefore vote for a popular party. That is why all political parties want to project themselves as strong and influential. It is almost like a gang culture where, if you want to survive, you go with the strongest gang.
”But some voters just want to know that their party will stand up for them. Opinion polls can definitely have a domino effect of further weakening the smaller parties,” she said. But, she adds, sometimes the poll may generate a sympathy vote for the weaker parties.
Western Cape
According to the latest SABC-Markinor Opinion Poll, the ANC
will win with 32% and its coalition partner NNP 13,3%. The DA will notch up a competitive 24,4% and other parties 7,9%. Undecided voters will amount to 7,9%.
KwaZulu-Natal
The ANC appears to be unassailable with 48,7%; partner NNP 1,4%. The IFP/DA coalition will receive 31% at best; IFP, 24,9%, loses control of KwaZulu-Natal; DA, 6,3%. But previous polls underestimated both the IFP
and the DA.
Limpopo
ANC 87,8%; DA 3%.
Northern Cape
ANC 69,6%,; DA 10%; NNP 1,6%.
Eastern Cape
UDM and DA in a tussle for the job of official opposition: UDM 5,4%, DA 4,5%. ANC 76,2%.
Gauteng
Decreased majority for ANC, 64,6%. DA 12,9%; IFP 1%; NNP 3,5%.
Mpumalanga
ANC 72,7%; DA 6,5%; UDM 3,7%.
North West
ANC 78,4%; DA 5%. Survey does not mention United Christian Democratic Party, the current opposition.
Free State
ANC 75,5%; DA 8,9%; ACDP will put up its strongest provincial showing with 4,5%.
National
ANC 68,6%; DA 9,2%; IFP 4,5%; NNP 3,2%; UDM 1,5%; ACDP 1,2%; Independent Democrats 1%. Other parties will collectively win 2,6%, but each will score less than 1%.
Note from Markinor/SABC.
Fieldwork was undertaken from January 29 to February 20 this year, and personal, in-home interviews were conducted with 3 500 randomly chosen respondents from all walks of life. Results are representative in terms of province, gender, race, age, working status and other demographic features. All sample surveys are subject to statistical error and the results must be evaluated accordingly. Depending on the response rate, the sample error for the poll as a whole is between 0,72% and 1,66%.