ANC looks set to win 63.3% of vote

THE African National Congress (ANC) will win 63.3% of the vote and the Democratic Alliance (DA) 22%‚ according to projections by the Council for Scientific and Industrial Research (CSIR)‚ which its statisticians say have typically been within 0.5% of the final result.

The CSIR’s model‚ which makes projections based on voting patterns from previous elections‚ is acknowledged for its accuracy by political parties. The projections take the surprise out of the final result‚ which the Electoral Commission of South Africa (IEC) says will be available by Friday or Saturday.

CSIR analyst Zaid Kimmie said that its projection was based on 10% of the vote‚ which in previous elections had been sufficient to project the final results with 0.5% accuracy.

Other projections for political parties in the national vote include 4.8% for the Economic Freedom Fighters (EFF)‚ 2.4% for the Inkatha Freedom Party‚ 1.6% for the National Freedom Party and 1.1% for the United Democratic Movement. The Freedom Front Plus gets 0.9% with all other parties falling below this threshold.

The predicted voter turnout is 72%‚ three percentage points lower than in the last national election in 2009.

In Gauteng‚ the ANC is projected to win 56.3% of the provincial vote and the DA 29.1%.

In the Western Cape‚ the DA wins with 57.2% and the ANC 34.2%.

The best showing of the EFF is in the North West where it is projected to get 10.7% followed by Limpopo‚ where it is expected to get 10.1% of the vote.

However‚ it is only likely to win official opposition status in Limpopo as the projections have it third behind the DA in the North West.

In the Eastern Cape‚ where the DA had been hopeful about making inroads‚ the ANC is anticipated to get 71.4% and the DA 16.6%. Compared to the 2009 provincial results‚ the ANC drops eight percentage points‚ which are picked up by the DA

BDlive 2014

One thought on “ANC looks set to win 63.3% of vote

  • May 8, 2014 at 3:38 pm

    Some quite big chunks of urban areas including around universities are yet to be counted. I wonder how the CSIR can model totally new parties with this sort of variability. For example, Agang has done a lot better in the vicinity of Rhodes University than generally – maybe not enough to make a significant difference, but this is the sort of thing a predictive model using previous elections may not get quote right.


Leave a Reply