Afrobarometer Briefing Paper No. 108  
  
Party Identification in South Africa:  
Profiles for the ANC and the DA 
  
by Paul Graham 
 
December 2012 
 
 
  
Who are the African National Congress (ANC) and the Democratic Alliance (DA)? Together with March 2012 
party membership figures and election results, there is one additional reliable measure² party 
identification. The Afrobarometer study aso yks tou feehe ql closuese to ati polon:iti caDl party?´ 
and, if yesWhi, ch one?Six   in 10 South Africans do feel close to a political party. This figure has  
remained more or less constant since the 2008 Round 4 Afrobarometer study. However, for the first 
time, one in ten people feel close to the DA, providing us with a large enough sample (n=229) to look 
at the details of those who identify with the party, alongside similar information on the ANC 
(n=1064).  
This brief provides some profile information on the supporters of each party, as the DA emerges from 
its Congress with new leadership (although with the key figures still in place), and as the ANC moves 
into its electoral conference, to be held in Mangaung, the city which was the birthplace of the ANC 
100 years ago. 
 
The Survey 
The Afrobarometer is a comparative series of public attitude surveys, covering up to 35 African 
countries in Round 5 (2011-2013). It measures public attitudes on democracy and its alternatives, 
evaluations of the quality of governance and economic performance. In addition, the survey assesses 
the views of the electorate on critical political issues in the surveyed countries. The Afrobarometser ¶
main goal is to produce scientifically reliable data on public opinion in Africa while strengthening 
institutional capacities for survey research, and research findings to inform policy and practice. The 
Afrobarometer also provides comparisons over time, as four rounds of surveys have been held from 
1999 to 2008 and Round 5 is currently underway. 
During Round 5, Afrobarometer surveys will be conducted in up to 35 countries in Africa, using a 
common survey instrument and methodology. The findings reported here draw from the most recent 
survey in South Africa conducted between October and November 2011. This is the sixth 
Afrobarometer survey conducted in South Africa (others were conducted in 2000, 2002, 2004, 2006 
and 2008). The survey was based on a nationally representative random sample of 2399 adult South 
Africans drawn from all nine provinces of the country. The findings reported here have a margin of 
sampling error of +/- 2 percent at a 95 percent confidence level. Fieldwork was conducted by Citizens 
Surveys Ltd., with technical support from Idasa.  
Party Demographic Profiles  
Those who identify themselves closely with a party are the potential cadres on which the party will 
rely for organising branches, getting out the vote and proselytising for their cause. At the end of 2011, 
44% of respondents identified themselves with the ANC, and 10% with the DA, the latter up from just 
1 
 




--------------------------------------- 2

4% in 2008. Other parties, although they are represented in one or other legislature, have very few 
1
supporters willing to identify themselves (4% for all other parties combined) and find that they are 
therefore always struggling to increase their size and reach. 
For both the ANC and the DA, the numbers who self-identify as being affiliated with the party are 
large enough that we can explore in more detail who these loyalists are. We can be more confident in 
the profile of ANC supporters because there were many more of them in the survey sample (margin of 
error of +/-3% at a 95% confidence level); but DA numbers are sufficiently high that we can also 
provide tentative descriptions, with a somewhat higher margin of error of +/-6.5% at a 95 percent 
confidence level. This briefing paper will focus on the demographic attributes of supporters of each 
party. More analysis could also be done to explore levels of political participation or the attitudes 
espoused by supporters of each party, and the issues which they prioritise. But this research will be 
saved for a later briefing paper. For each of the two parties, the sections below provide the profile of 
the party¶s supporters. The numbers are not comparative levels ofA  support between ANC and D
(e.g., what proportion of residents in Eastern Cape support the DA, and how many support the ANC), 
but rather, we ask what proportion ofi s dthe DArawn from¶s supp Eastern Cortap e, and what 
proportion of the ANC¶s support is bas²ed tin otherher weords, how dependent is each on Eastern 
Cape, or any other province or demographic group.  The numbers shown for each party will in general 
sum to 100%. 
In order to provide these pictures of the two largest parties, we approach each of them separately² so 
we are not saying that because the ANC have, for example, 14% of their supporters in the Eastern 
Cape and the DA have 13% of theirs in the same province that they are therefore in a neck and neck 
battle for the province. The ANC obviously have many more supporters than the DA. What it does 
indicate is that the DA has a cadre of people in that province on whom they can draw, whereas they 
might face an uphill organisational battle in the North West where only 1% of their party identifiers 
are resident. The ANC on the other hand, has a similar problem, although not of the same magnitude, 
in the Western Cape, where only 7% of their public supporters live. 
The DA finds the majority of its strong supporters in three provinces² Gauteng, the Western Cape 
and the Eastern Cape, which make up 80% of their profile. ANC supporters are more evenly spread 
throughout the country, although Gauteng and Kwazulu Natal make up fully 49% of the party¶s 
profile. 
The results of this brief analysis suggest that the DA is a fundamentally urban party, with 88% of their 
party identifiers living in one or other town or city. The ANC too is a majority urban party at the 
moment² with 55% of their party identifiers living in such circumstances² but 45% of their close 
identifiers do live in rural areas.  
                                                           
1
 39% of South Africans do not identify closely with a political party. In preparing this brief, we use the 
numbers who do identify as a percentage of the entire population. 3% of respondents identify with a party but 
will not say which party. 
2 
 



--------------------------------------- 3

   
Table 1: Distribution of Party Support Across Provinces 
 
  
   
 
  
 
Gauteng
Free StateLimpopo
Eastern CapeMpumalangaNorth WestWestern Cape
Kwazulu NatalNorthern Cape
 ANC 14 7 21 18 14 10 6 3 7 
DA 13 4 35 6 4 4 1 2 32 
Census 2011 
Population 
Distribution 13 5 24 20 10 8 7 2 11 
 
While the Afrobarometer, like many other studies in South Africa, continues to find apartheid racial 
categories convenient as proxies for reporting on a range of differences, it also allows people to self-
identify in other ways. One of these is by primary language. People who feel close to the ANC and the 
DA have quite different language profiles.  Fifty-six percent of the DA profile (Figure 1) speak 
Afrikaans, and nearly one third (35%) are English speakers. A smattering choose Xhosa, Northern 
Sotho/Pedi, Sesotho, Venda and Zulu. For the ANC, in contrast, supporters speak all official 
languages, with Zulu and Xhosa in the majority with 24% and 23% each, followed by SePedi with 
14% (Figure 2).  
Figure 1: First Language of DA Supporters 
2 
2 
2 
2 
35 English
Afrikaans
Zulu
Sepedi
Sesotho
All others
56 
 
3 
 



--------------------------------------- 4

Figure 2: First Language for ANC Supporters 
20 23 
Xhosa
Sepedi
Sesotho
Setswana
14 Zulu
24 All others
9 
9 
 
 
The party profiles differ with respect to age as well (Figure 3). ANC supporters are more youthful, 
while 50% of the DA support comes from people between the ages of 36 and 55. Retirees provide five 
to six percent of the support for both parties.  
Figure 3: Age Profile of DA and ANC Supporters 
60
50
40
30ANC
DA
20
10
0
 18-2526-3536-5556-6566-
There are also differences in the educational levels of party identifiers, As Figure 4 demonstrates, the 
majority of public supporters of both parties attained a secondary education. But a significant number 
(29%) of DA support comes from those who have gone on to post matric studies, while only seven 
percent of ANC supporters have done likewise. At the other end of the scale, the DA attracts fewer 
people whose education stopped after primary school, whereas the ANC has 20% of its support from 
those adults. 
  
4 
 



--------------------------------------- 5

Figure 4: Educational Level of DA and ANC Supporters 
80
70
60
50
40ANC
30DA
20
10
0
No formalInformalPrimary Secondary TertiaryPost
 educationeducationGraduate
Finally, the Afrobarometer asks a series of livelihood and access to assets questions in order to assist 
us in developing an understanding of poverty in South Africa. The focus of the survey is not primarily 
demographic and it does not ask for income levels. However, amongst the questions is one that gets to 
the heart of poverty. While both rich and poor in South Africa experience shortages of water and 
electricity² for obviously different reasons² we also ask about whether our respondents have gone 
without food in the 12 months before the survey (Figure 5). The DA profile reveals that 83% of its 
supporters have never gone without food. But the DA does have supporters who struggle: 11% have 
gone without once or twice, three percent several times, and two percent many times. The ANC, on 
the other hand, also draws a majority of its support from those who are never without² 59%. But they 
have far more supporters who struggle economically: 18% have gone without once or twice, 14% 
several times, 7% many times, and two percent who are always short of food. 
Figure 5: Poverty Profile of Party Support 
90
80
70
60
50
ANC
40
30DA
20
10
0
NeverJust onceSeveralManyAlways Don't know
or twicetimestimes
 
In the past year, how often, if ever, have you or anyone in your family gone without enough food to eat? 
5 
 



--------------------------------------- 6

Conclusion 
South Africa is not a two party country. Its electoral system and constitution encourage a multi-party 
environment and there are over 140 parties registered with the Independent Electoral Commission. 
But the majority of these parties are small; some parties inhabit value, geographic or ethnic niches. 
But two parties contend directly for voters across the country, and two parties have national 
aspirations. The ANC dominates the landscape and the various sites of government. The DA argues 
that it is a government-in-waiting, and indeed it already governs at the provincial and municipal level. 
While both parties have established themselves as broad churches and espouse programmes which 
voters often cannot easily differentiate on strictly ideological grounds, they clearly differ in the profile 
of their closest supporters. Because the absolute number of DA supporters in our sample is still 
relatively small, our conclusions on this party have to remain tentative, although indicative. With the 
ANC profile we are confident of the findings. Each of these parties will have to decide whether these 
profiles are indicative of what they are trying to become. But if they wish to change, change will be 
constrained by the ability of their key supporters, those who feel close to the party, to transcend the 
boundaries of their race, geography, language, age and class. 
This Briefing Paper was prepared by Paul Graham, Executive Director of Idasa. 
 
The Afrobarometer is produced collaboratively by social scientists from more than 30 African countries. 
Coordination is provided by the Center for Democratic Development (CDD-Ghana), the Institute for Democracy 
in South Africa (Idasa), the Institute for Development Studies (IDS), University of Nairobi, and the Institute for 
Empirical Research in Political Economy (IREEP) in Benin. We gratefully acknowledge generous support from 
the UK¶s Department for International Development (DfID), the Mo Ibrahim Foundation, the Swedish 
International Development Agency (SIDA), and the United States Agency for International Development 
(USAID) for support on Afrobarometer Round 5. For more information and further requests for analysis please 
visit Afrobarometer website:  www.afrobarometer.org or contact Anyway Chingwete on +27 12 392 0500 or 

 This email address is being protected from spambots. You need JavaScript enabled to view it.
  
 
 
6