Geography and voting: the growth of urban opposition in South Africa two decades after democratisation

Abstract

The results of the 2016 local government elections in South Africa provided a new opportunity to assess voting trends and to explore the synergies of demographics, geography and voting behaviour. Pre-election protests and polls suggested haemorrhaging of support levels for the ruling African National Congress (ANC). Increased support was thus predicted for opposition parties on the political right and left of the ANC, especially for the Democratic Alliance (DA) and the Economic Freedom Front, respectively. These predictions were realised when support for the ANC declined from 62 to 54% nationally, and the ANC received less than half of votes cast in five of the country’s eight metropolitan municipalities. This paper interrogates the election outcomes in the municipal wards of the two largest metropoles, Johannesburg and Cape Town, to determine dynamics of the neighbourhood level changes in voting patterns.

Keywords:

• Election
• opposition
• urban
• ward
• demographics

Acknowledgements

I am grateful for the maps produced by S’bo Zama and the insightful suggestions of two anonymous referees.

Notes

1. Maimane was one of several young black politicians who achieved prominence in the DA prior to Zille’s resignation as leader. Others included Lindiwe Mazibuko, Solly Msimanga, Khume Ramulifho, Makashule Gana and Mbali Ntuli (Jolobe, 2014).

2. http://www.tradingeconomics.com/south-africa/unemployment-rate.

3. Runciman et al. (2016) contend that the protests reported in the media underestimate the total number, which is likely to be closer to the number of police-reported protests captured in the SAPS IRIS database.

4. The NFP had splintered from the IFP in 2011 (Piper, 2014) and had received 1.57% of votes cast nationally (6.43% in KwaZulu-Natal) in the 2014 national elections, in comparison with the IFP’s 2.4% (10.17% in KwaZulu-Natal).

5. The number of wards increased from 130 to 135 in Johannesburg and 111 to 116 in Cape Town. The new wards were in areas in high population growth areas in Johannesburg’s north-west (Bloubosrand, Sundowner), south-west (Naledi) and north-east (Ivory Park, Vorna Valley); and in Cape Town’s north-west (Killarney), south-east (Mfuleni, Lentegeur), north-east (Durbanville Extension) and CBD (City Bowl).

6. This trend was also identified in some Johannesburg and Tshwane townships by Dufour and Calland (2016).

7. A national survey comprising 2398 respondents which was part of the Comparative National Elections Study (CNEP).

8. Owing to the increase in 2016 of five new wards to each of the two cities, the 2011 census data from adjacent or demographically similar wards were extrapolated to the newly created wards.

9. The ecological fallacy is an error in logic, whereby data pertaining to a group are assumed to be imputable as information about an individual in that group. In this context, an ecological fallacy would be incurred in the assumption that all white voters support the DA, or all that voters from low-income households are ANC supporters.