South African voting patterns have always been related closely to demographic, economic and ethnic factors. The changes over the past few decade economically and demographically can be linked to the formation of voting blocks that are closely aligned to economic classes and regional demographic distributions. Using previous election results, Census 2011 socio-economic data and various statistical publications from the period 2011-2016, I ran an algorithm which models socio-economic changes at a very small geographic level since the Census of 2011. Changes such as education levels, personal income, racial distribution, linguistic distribution, voter turnout rates and population/age shifts were modeled. These changes were correlated to voting trends from the 2014 National Elections and then projected to mid 2016. Agglomerating the predicted party vote projections to the provincial and national level produced the following results:
ANC Nationally = 53.816 %
IFP Nationally = 2.121 %
DA Nationally = 30.072 %
EFF Nationally = 7.213 %
ANC Province of Western Cape = 22.914 %
IFP Province of Western Cape = 0.067 %
DA Province of Western Cape = 67.659 %
EFF Province of Western Cape = 2.991 %
ANC Province of Eastern Cape = 58.356 %
IFP Province of Eastern Cape = 0.048 %
DA Province of Eastern Cape = 24.383 %
EFF Province of Eastern Cape = 7.673 %
ANC Province of Northern Cape = 34.689 %
IFP Province of Northern Cape = 0.094 %
DA Province of Northern Cape = 53.039 %
EFF Province of Northern Cape = 4.528 %
ANC Province of Free State = 61.329 %
IFP Province of Free State = 0.424 %
DA Province of Free State = 24.143 %
EFF Province of Free State = 8.063 %
ANC Province of KwaZulu Natal = 55.172 %
IFP Province of KwaZulu Natal = 7.143 %
DA Province of KwaZulu Natal = 20.620 %
EFF Province of KwaZulu Natal = 7.865 %
ANC Province of North West = 62.318 %
IFP Province of North West = 0.238 %
DA Province of North West = 22.306 %
EFF Province of North West = 8.168 %
ANC Province of Gauteng = 53.291 %
IFP Province of Gauteng = 1.947 %
DA Province of Gauteng = 32.233 %
EFF Province of Gauteng = 7.049 %
ANC Province of Mpumalanga = 63.291 %
IFP Province of Mpumalanga = 2.304 %
DA Province of Mpumalanga = 21.198 %
EFF Province of Mpumalanga = 8.497 %
ANC Province of Limpopo = 69.269 %
IFP Province of Limpopo = 0.115 %
DA Province of Limpopo = 17.423 %
EFF Province of Limpopo = 9.094 %
When the results are announced later this month, I will post back on the accuracy of the prediction. Surely, the validity of my model's assumptions and other factors that may affect the mind of the South African voter will be revealed in this exercise. I also expect to find out how accurately the concept of big data can be used to forecast something as complex as an election process in a diverse socio-economic landscape such as South Africa.
ANC Nationally = 53.816 %
IFP Nationally = 2.121 %
DA Nationally = 30.072 %
EFF Nationally = 7.213 %
ANC Province of Western Cape = 22.914 %
IFP Province of Western Cape = 0.067 %
DA Province of Western Cape = 67.659 %
EFF Province of Western Cape = 2.991 %
ANC Province of Eastern Cape = 58.356 %
IFP Province of Eastern Cape = 0.048 %
DA Province of Eastern Cape = 24.383 %
EFF Province of Eastern Cape = 7.673 %
ANC Province of Northern Cape = 34.689 %
IFP Province of Northern Cape = 0.094 %
DA Province of Northern Cape = 53.039 %
EFF Province of Northern Cape = 4.528 %
ANC Province of Free State = 61.329 %
IFP Province of Free State = 0.424 %
DA Province of Free State = 24.143 %
EFF Province of Free State = 8.063 %
ANC Province of KwaZulu Natal = 55.172 %
IFP Province of KwaZulu Natal = 7.143 %
DA Province of KwaZulu Natal = 20.620 %
EFF Province of KwaZulu Natal = 7.865 %
ANC Province of North West = 62.318 %
IFP Province of North West = 0.238 %
DA Province of North West = 22.306 %
EFF Province of North West = 8.168 %
ANC Province of Gauteng = 53.291 %
IFP Province of Gauteng = 1.947 %
DA Province of Gauteng = 32.233 %
EFF Province of Gauteng = 7.049 %
ANC Province of Mpumalanga = 63.291 %
IFP Province of Mpumalanga = 2.304 %
DA Province of Mpumalanga = 21.198 %
EFF Province of Mpumalanga = 8.497 %
ANC Province of Limpopo = 69.269 %
IFP Province of Limpopo = 0.115 %
DA Province of Limpopo = 17.423 %
EFF Province of Limpopo = 9.094 %
When the results are announced later this month, I will post back on the accuracy of the prediction. Surely, the validity of my model's assumptions and other factors that may affect the mind of the South African voter will be revealed in this exercise. I also expect to find out how accurately the concept of big data can be used to forecast something as complex as an election process in a diverse socio-economic landscape such as South Africa.
