This analysis (below) of provincial party support is based on the findings of the moderate voter turnout scenario, i.e. ALL eligible voters will not vote, but currently about 77% of eligible voters are registered to vote or indicated that they are likely to vote or desire to vote. The views of this group as expressed in October - November 2013 are summarised below.
Ipsos South Africa undertakes a Pulse of the People study every six months. A randomly selected sample of South Africans of voting age (18+) are asked which party they would support on the provincial level if there were an election the next day.1 Respondents then fill in their own choices on a "ballot paper" - making this a secret vote.2
Provincial Party support amongst voters - in a moderate voter turnout scenario
The ANC has a clear majority in six of the nine provinces
The DA should keep its majority and control in the Western Cape
Both Gauteng and the Northern Cape will be heavily contested between the ANC and the DA (and other opposition parties who might hold the key to power sharing, especially in Gauteng)
The EFF can be the official opposition after the election in both Limpopo and North West
In KwaZulu-Natal, more than one in every ten (11%) refused to answer the question
Uncertainty is the highest in the Free State, with almost one in ten (9,5%) saying that they do not know which party to vote for.
Technical detail:
A total of 3564 personal face-to-face interviews were conducted with randomly selected adult South Africans. The interviews were done in the homes and home languages of respondents. Trained quantitative fieldworkers from all population groups were responsible for the interviewing, which took place from 22 October to 21 November 2013. Interviews were done all over the country, from metropolitan areas to deep rural areas. This methodology ensured that the results are representative of the views of the universe and that findings can be weighted and projected to the universe - i.e. South Africans 18 years and older.
Interviews were done using CAPI (Computer Assisted Personal Interviewing) and all results were collated and analysed in an aggregate format to protect the identity and confidentiality of respondents.
All sample surveys are subject to a margin of error, determined by sample size, sampling methodology and response rate. The sample error for the sample as a whole at a 95% confidence level is a maximum of 1,67%. When analysing the results for smaller provinces (i.t.o. population size) separately, the margin of error will be higher for those provinces.
About Ipsos: The Home of Researchers
Ipsos is an innovative, entrepreneurial, client-focused organisation, providing research services to clients on a global basis. We set ourselves high standards and aim to work collaboratively in partnership with our teams in order to service our clients most effectively.
Ipsos is proud to be the only global market research company that is still controlled and operated by researchers. We aim to remain the natural home for intellectually curious and passionate researchers.
Our goal is simple: to be our clients' preferred research partners in our areas of specialisation, methodologies and processes. We want our clients to be proud and pleased to work with us - and we want each one of us to be proud and pleased to offer our clients high quality standards, efficiency and intelligence.
1The question wording is: "If there were national elections tomorrow, which political party or organisation would you vote for? Please indicate your choice of party on a national level as well as on a provincial level."
2It is also important to keep in mind that this latest poll was undertaken before the death of ex-president Nelson Mandela and that the next poll planned for early 2014 will yield more contemporary results - thus this poll is for information and planning purposes and not intended to predict the results of the coming election. Polls are valid for the time when the fieldwork is done and should always be evaluated against this background and within the margin of error.