R 300732Z MAR 09
FM AMCONSUL CAPE TOWN
TO SECSTATE WASHDC 3028
INFO SOUTHERN AF DEVELOPMENT COLLECTIVE
AMEMBASSY ABUJA

UNCLAS CAPE TOWN 000078 
 
 
E.O. 12958: N/A 
TAGS: SF PGOV KDEM
SUBJECT: WHAT INFLUENCES THE SOUTH AFRICAN VOTER? 
 
¶1. Summary: When South Africans go to the polls on April 22, 
they'll be weighing a host of factors that include their 
personal circumstances, social networks, and perceptions of 
government performance in order to help them make a decision. 
 Cape Town Poloff and Econoff on March 19 attended two 
seminars that helped explain these factors, examining trends 
since the 1994 democratic transition, and determine which 
were most important, both nationally and in Western Cape 
specifically.  Unsurprisingly, race emerged as a key factor 
in determining a party's support base in the three previous 
democratic elections, albeit in an indirect fashion by which 
race influences personal and social networks that influence 
party support.  Age, economic status, government performance, 
and press coverage all have proven less influential than race 
in voting patterns. 
 
¶2. That said, all experts noted that this election could very 
well display very different characteristics than the past 
three polls.  Voter registration is higher than in 2004, and 
enthusiasm over this poll suggests that turnout also will be 
higher than the last poll.  Many new registrants are under 30 
-- a group about which there is no good data on political 
perceptions or likely voting trends -- while the number of 
independent black voters has been increasing.  While race 
most likely will still be a key driver of political choice, 
the experts acknowledged that factors such as class, 
religion, ethnicity, and gender have not been sufficiently 
researched to gauge their impacts.  Lastly, the emergence of 
the Congress of the People (COPE) opposition party could 
provide a political home to the growing number of voters 
disenchanted with the ruling African National Congress (ANC) 
and the post-apartheid political system, although it's less 
clear, given the affinity many black voters have for the ANC, 
to what extent the party's lack of an effective organization 
will handicap it. 
 
¶3. Cape Town officers attended a seminar sponsored by the 
Institute for Security Studies Corruption and Governance 
Center on voter trends in South Africa since 1994.  The 
speakers were ISS researcher Collette Schulz-Herzenberg, who 
presented a detailed quantitative study of voter trends, and 
University of Cape Town professort Cherrel Africa, who 
presented on the role of media in influencing voter behavior. 
 Econoff that evening went to another session hosted by the 
Institute for a Democratic Alternative in South Africa 
(IDASA) that focused more specifically on voter trends and 
influences in Western Cape province.  Presenters were author 
Christi van der Westhuizen, University of Stellenbosch 
professor Amanda Gouws, and University of Cape Town lecturer 
Zweli Jolobe. 
 
-------------------------------- 
A LOOK AT THE NUMBERS SINCE 1994 
-------------------------------- 
 
¶4. Before looking at the factors, it's worth first examining 
the three election results since 1994, which have shown 
growing support for the ANC amid a steep decline in voter 
participation.  ANC support has grown from 63 percent in 1994 
to 66 percent in 1999 and 69 percent in 2004.  However, this 
has been mirrored by a sharp decline in turnout in the past 
two polls, decreasing from 89 percent turnout in 1999 to 77 
percent in 2004.  On top of that, the percentage of eligible 
voters who even registered to vote declined from 81 percent 
in 1999 to 74 percent in 2004.  So while ANC support in the 
2004 appeared quite strong at 69 percent, the party's support 
represented just 39 percent of the total voting age 
population.  That said, the combined opposition could only 
Qpopulation.  That said, the combined opposition could only 
muster 17 percent of the voting age population; the other 44 
percent stayed home. 
 
¶5. Schulz-Herzenberg in particular chalked the decline up to 
voters decreasingly identifying with the political parties on 
offer.  ANC partisans are still tremendously loyal to the 
ruling party, but whether due to dissatisfaction with the 
party's performance, alienation with the new party 
leadership, declining identification with the party's 
liberation struggle past, or other reasons, the ANC's core 
base of support is declining.  These floating voters, 
however, have not seen any opposition parties as worthy of 
their votes; in a 2004 study, the Democratic Alliance was the 
only political party viewed by more than 30 percent of voters 
as inclusive of all South Africans, as opposed to the ANC's 
70 percent. 
 
--------------------------------------------- - 
RACE A KEY, ALBEIT COMPLEX, DRIVER OF BEHAVIOR 
--------------------------------------------- - 
 
¶6. Strictly speaking, race is a significant factor in 
determining how South Africans vote, but Schulz-Herzenberg 
noted that its influence is more subtle than one might think. 
 She noted, for example, that South Africans do not practice 
straight "racial census" voting by which race is the only 
motivating factor.  Rather, racial cues factor into such 
things as perceptions of inclusiveness and social networks. 
Black voters, for example, have generally perceived the 
Democratic Alliance (despite its best efforts) as a "white" 
party, the Independent Democrats as "colored," and the 
Inkatha Freedom Party and United Democratic Movement as only 
for Zulus and Xhosas respectively. 
 
¶7. Schulz-Herzenberg focused extensively on the issue of how 
important personal social networks influenced voter 
behaviors.  Social networks in South Africa tend to be very 
racially and linguistically homogenous, and it can be very 
difficult here for new parties to enter into the social 
political dialogue.  The tendency toward reinforced 
groupthink is strong, and one that has reinforced ANC 
dominance in the past several elections.  Those voters who 
tend not to agree with the group consensus tend to not vote 
rather than voting for -- or much less campaigning for -- 
another party. 
 
¶8. Schulz-Herzenberg said her research -- which utilized 
survey data from nine national polls since 1994 -- suggested 
that government performance has heretofore had little 
influence on voter choices.  Cherrel Africa's presentation on 
media influence in campaigns suggests that coverage of 
election campaigns -- both by the press and party publicity 
-- also have relatively little effect on voters.  Her studies 
found that media outlets of all types tend to offer generally 
unbiased coverage of all major parties, and this objective 
coverage has not tended to influence voters one way or the 
other.  Parties' generally banal campaigns also have had very 
little impact; Africa bemoaned the lack of creativity shown 
by parties of all stripes. 
 
----------------------------------------- 
ELECTION 2009 COULD PROVE QUITE DIFFERENT 
----------------------------------------- 
 
¶9. When looking at the April 22 poll, all of the analysts at 
both seminars noted that electoral landscape has changed 
since 2004 -- and particularly since the formation of COPE 
last November.  First and foremost, the trend toward voter 
apathy appears as if it will be interrupted in this poll. 
The number of eligible voters registering has increased to 77 
percent, from the 74 percent in 2004,  Meanwhile, the 
Independent Electoral Commission and others are expecting a 
jump in turnout to back over the 80 percent mark.  Also of 
note is that 27 percent of eligible voters are under 30; 73 
percent of voters in the last two registration periods (both 
of which happened after COPE's formation) were under 30. 
 
¶10. There is little data suggesting how these young voters -- 
two-thirds of whom are black -- will vote.  Apartheid is a 
distant memory for these voters, if a personal living memory 
at all, and these voters are likely to have far more nuanced 
conceptions of identity -- taking into account issues such as 
class, for instance -- than their parents.  Their perceptions 
of government performance, or its influence on their vote, 
are unclear, but are less likely to be influenced by a sense 
of historical loyalty to the ANC (or other parties) than 
older generations. 
 
¶11. While the analysts acknowledged that the ANC's split and 
the formation of COPE certainly appear to have captured 
voters' attentions, they were loath to try to pinpoint how 
such enthusiasm will translate at the polls.  Jolobe 
emphasized that ANC success at the polls has long stemmed 
from its impeccable organization; COPE has not shown similar 
organizational capacity.  Schulz-Herzenberg said that COPE 
certainly has the potential to be an alternative political 
home for independent voters, but that they also would have to 
Qhome for independent voters, but that they also would have to 
show voters that they are a viable alternative -- COPE's 
problems with campaign organization and fundraising are not 
helping it in that sense. 
 
¶12. Also noted was the paucity of information on how other 
factors could impact this poll.  Van der Westhuizen noted 
that little study has been made of gender and politics in 
South Africa, particularly given a majority of registered 
voters are female.  She noted that what little research has 
been done has found little to suggest gender has much of a 
distinct role in voting patterns, although she opined that 
Jacob Zuma's rape allegations and "anti-woman" statements by 
ANC leaders Julius Malema and Tokyo Sexwale could damage the 
ANC among women voters.  (Note: A recent survey of popular 
attitudes towards the traditional African practice of 
polygamy revealed that 85 percent of black women, more than 
any other group, oppose this practice and expressed the view 
that personal values of politicians influenced these women's 
judgements and potentially their votes.  Pollsters 
self-consciously used Jacob Zuma's polygamy as a point of 
departure in this survey, revealing popular sentiment 
regarding this practice among political leaders.  End Note) 
Others noted that there is a dearth of research on the 
influence of religious and traditional leaders on voters as 
well. 
 
¶13. All of the experts at these seminars raised compelling 
points about how this election might play out, although they 
readily acknowledged that no one felt comfortable making 
predictions on the outcome, at least nationally.  The IDASA 
panelists did, however, feel quite strongly that the ANC 
would lose Western Cape.  Gouws noted than an October 2008 
Markinor poll -- before COPE's formation -- gave the 
Democratic Alliance 42 percent support and the ANC just 26 
percent.  She believes COPE's emergence will largely pull 
from ANC support, effectively shutting the ANC out in the 
province. 
 
------- 
COMMENT 
------- 
 
¶14. If anything, the experts' presentions show how difficult 
it is to accurately project how this election will turn out. 
Unlike more mature democracies such as in the United States 
and Europe, pre-election polling in South Africa is spotty 
and reliant on small samples, as good polling is too 
expensive for parties or media outlets.  Many of the 
assumptions posited by the analysts -- the role of race, for 
example -- are likely to hold true for this poll, but 
observers are unlikely to understand how or if a realignment 
in voter identification is in the offing until after the 
poll. 
 
 
MAYBERRY