Paul Berkowitz | 23 August 2016
The City of Cape Town metro was once a fiercely contested battleground. Control of the metro changed hands a few times between 2000 and 2007, with various ANC-led and DA-led coalitions holding power at different points.
From 2007 onwards the DA has cemented its majority in Cape Town, from leading a stable coalition in 2007 (with the ID) to winning an outright majority in the metro in the 2011 elections. The party's share of the vote grew from 61 percent in 2011 to 67 percent in 2016, and the DA won exactly two-thirds of the council seats in the latest municipal elections (154 seats out of 231).
The DA's increased majority has been due partly to the successful campaign that the party ran in 2016, but also party due to the ANC's poor performance in the metro and the Western Cape province in general. The data show how the ANC's support has declined in absolute terms (i.e. number of votes), particularly in wards which were previously the party's strongholds.
The maps and graphs below will confirm the decline of the ANC in the City of Cape Town metro, and they will also illustrate some observed trends of the smaller parties in the metros.
Voting in Cape Town by the numbers
The table below shows the general voting patterns in the metro, comparing registered voters and voter turnout in the 2011 and 2016 municipal elections:
The number of registered voters in the Western Cape province increased from 2.7-million in 2011 to 3.1-million in 2016, while registered voters in the metro increased over the same period from 1.7-million to 2.0-million.
Voter turnout in Cape Town rose from 1.1-million in 2011 to 1.3-million in 2016. While national voter turnout rose in percentage terms over the period (from 57.5 million to 58.1 million), it fell in both the Western Cape province and the Cape Town metro. The changes in turnout are small in all cases - less than a percentage point nationally, provincially and at the metro level - and are not statistically significant.
The average number of registered voters per ward is just over 17 000. Most wards have registered voter numbers which are within a 15 percent deviation of the average, or within a range of around 14 500 to 19 500 registered voters. There are seven wards with more than 19 500 voters and nine wards with fewer than 14 500 voters.
The map below shows the number of registered voters per ward. Much of Khayelitsha and Langa have low numbers of registered voters. Two contributing factors could be demographics (higher numbers of under-30s) and the relative difficulties in registering (transport, free time, etc).
Average voter turnout in the metro was just over 64 percent. Turnout at the ward level ranged from 49 percent all the way up to 82 percent. The map below shows turnout levels by ward:
It can be seen from the map that Khayelitsha experienced both low voter registration numbers and low levels of voter turnout.
The DA won 81 of the 116 wards and won them convincingly - with more than 80 percent in 54 wards and more than 65 percent in 75 wards.
The party enjoyed similar success in the 2011 elections, but made gains in 2016 by winning Ward 43 convincingly. In 2011 the party only polled 11 percent in what was Ward 80. In 2016 the DA won the ward with 79 percent. The party also made gains in the south of Khayelitsha, but this may have been due in part to the redrawing of ward boundaries.
The ANC performed badly in the metro. The party only won 35 out of 116 wards and only won with more than 80 percent of the vote in seven of these. The ANC won with more than 65 percent in 28 wards.
Comparing results at the township level
A disaggregated analysis of voting at the township level, with a focus on the absolute number of votes won, illustrates how poorly the ANC did among its traditional voter base. It also demonstrates the gains that the DA made in the same areas.
A direct comparison between the 2011 and 2016 results is difficult due to the changing of ward boundaries, but there are clearly trends at play. The calculations may not be precise due to boundary changes. The redrawing of ward boundaries has made it too difficult to compare results in Gugulethu, but comparisons are possible for Khayelitsha and Langa:
It can be seen that, even though the number of registered voters increased between 2011 and 2016, the number of votes cast fell over the same period. In Khayelitsha in particular, the fall in voter turnout was pronounced, going from 74 percent in 2011 to just 56 percent in 2016.
The DA increased its share of the vote from 5 percent to 13 percent in Khayelitsha and from 3 percent to 6 percent in Langa. The ANC lost a significant share of the vote in Langa and also went from over 109,000 votes in Khayelitsha to under 84 000 votes.
The ANC was hit by a combination of lower turnout from its core base and increased competition by the DA and other parties - in particular the EFF.
Part of the reason for the ANC's low victory margins is the EFF factor. Although the party didn't win more than 15 percent in any ward, it did best in ANC strongholds.
The party polled highest in parts of Khayelitsha and Dunoon. Its levels of support don't currently threaten the ANC's control of wards in these areas but they will certainly reduce the party's PR seat haul.
Other identifiable trends among the smaller parties
The DA, ANC and EFF shared over 94 percent of the vote. The fortunes of the smaller parties have no bearing on the control of the metro council. Still, there are trends of interest that may inform some of the broader debates around the 2016 elections.
The ANC is losing votes on a number of fronts. Apart from the apathy of its erstwhile supporters the DA and EFF account for most, but not all, of the party's losses at the polls. Smaller parties such as the PAC and the AIC have also hurt the ANC.
The PAC didn't poll higher than 4 percent in any ward, but all of the party's gains have been in and around Khayelitsha:
The AIC has also taken votes off the ANC in the same areas. Interestingly, the distribution of the AIC's votes busts the myth that the AIC is only winning votes from nearsighted and confused ANC supporters - if this was the case then the AIC's votes should be even distributed across all wards and not concentrated in areas like the south of Khayelitsha: