Abstract
General elections are held every five years in South Africa. During the 12 to 24 hour period after the close of the voting booths, the expected final results are of huge interest to the electorate and politicians. In the past, the Council for Scientific and Industrial Research (CSIR) has developed an election forecasting model in order to provide the media and political analysts with forecasts of the final results during this period of peak interest. In formulating this model, which forecasts the election results as the results from voting districts (VDs) become available, some assumptions had to be made. In particular, assumptions were made about the clustering of previous voting patterns as well as the order in which VD results are released. This election forecasting model had been used successfully for a number of elections in the past and in these previous elections, with around 5%-10% of the results available, the predictions produced by the model were very close to the final outcome, particularly for the ANC, being the largest party. For the 2014 national election, however, the predictions, with close to 50% of the voting district results known (equivalent to an estimated 40% of the total votes), were still not accurate and varied by more than 1% for both the ANC and the EFF. This paper outlines a post-election analysis to determine the reasons for these discrepancies and how they relate directly to the model assumptions. The aim is to highlight how practical realities can affect the assumptions and consequently their impact on the forecasted results. Reference is made to previous election forecasts and the 2014 post-election analysis is presented.
Highlights
In most democratic countries, general elections are held every four or five years
5% of votes counted was still too little in the case of the Democratic Alliance (DA) to calculate a more accurate forecast, illustrating the impact of the model assumptions, i.e. assumption 2 which assumes that the initial sample of voting districts (VDs) results released would provide a fair representation of voter behaviour well distributed across the clusters did not hold true for the DA, mainly due to a large degree of bias in the initial 5% of VDs counted, which contained high DA votes and which had memberships concentrated in a few clusters, causing an over-estimation of the DA percentage prediction in those clusters
With only 5% of the votes counted in the 2014 elections, the prediction for the DA and Inkatha Freedom Party (IFP) were very close to the final outcome but this was not the case for the African National Congress (ANC) and the Economic Freedom Fighters (EFF)
Summary
General elections are held every four or five years. The outcome of elections causes great interest amongst both the electorate and politicians. During the most recent national election in 2014, with almost 50% of VDs declared (equivalent to about 40% of the votes counted), the national forecast still deviated fairly significantly from the final result for two of the political parties. In 2014, with 49.3% of VDs declared, the predicted national result forecasted for the ANC and the new Economic Freedom Fighters (EFF) party differed by 1.15% and 1.56% respectively from the actual final outcome. These predictions were considered to be unacceptable and contrary to experiences from previous elections. A short discussion on model assumptions is given and some conclusions follow
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
