Voting preference in early democratic South Africa: Understanding the theory

27th September 2010 By: In On Africa IOA

Voting behaviour is recognised as one of the most influential study fields in South Africa. Voting preferences determine the voting behaviour of the voters and are dependent on various factors. The factors are divided according to long-, medium- and short term influences and are manifested, amongst others, as party identification, issue voting, outcome preference voting, retrospective voting, race (or class) census theory, as well as ideological orientation. It is, however, important to note that the line is often blurred between the different influences of voting preference and that the influences themselves often overlap. All the different influences of voting preference are not always evident in South Africa. This article will discuss and compare the main elements of voting preference in South Africa. It will make reference to the first democratic election in order to indicate the origin of voting preference influence that set the tone for the following elections in the country. Furthermore, this article will explain the importance of understanding the theory behind voting preference in order to overcome and avoid election-related conflict in South Africa.


Long-term influence


The most important long term influence is the underlying influence of race, ethnicity and class. An easy way to understand why citizens vote as they do is to compare their votes with observable voter characteristics. The variables that emerge are ethnicity, race and class. The correlation in ethnicity, race and class are measured in order to establish a logical reason for voting preferences.(2) Voting according to race is encapsulated in the race-census theory. This theory overlaps greatly with party identification, seeing that it is recognised as party identification on grounds of race. Party identification will, however, receive more attention later in this article. According to Horowitz(3), race and ethnicity are recognised as the most important factors that influence the individuals' political attitudes and behaviour. Owing to this, voting preferences are often build around race and ethnic identity.


With regards to the South African context, during the 1994 elections, voting preference was greatly formulated according to race preferences. It was especially noted amongst the black population as they, as a nation of race, supported the liberation movement against apartheid. This method of voting has, however, a heavy emotional and self-interest connotation seeing that the African National Congress (ANC) voters based their votes largely on emotions of excitement rather than to vote reasonably. Few supporters showed that they would vote according to issues and thus voting preference was based on symbols and the voters' concern over physical- and group security, as well as charisma.(4)


It is, however, important to note that this correlation between race/ethnicity and voting preference is rather seen as a mere description, seeing that people do not necessarily base their voting preference explicitly on their race/ethnic identity. In the same sense, race cannot be seen as the exclusionary factor that influences voting preference. Few voters recognise their parties as race/ ethnic exclusionary - even the voters with a strong race and ethnic identities do not necessarily identify with parties that only promote their own race/ethnicity.(5) Even though it did seem as though race/ethnicity was the basis of voting during the 1994 election, it was not the only influential factor.


The main reason why the 1994 election was seen as a race-driven election was owing to the parties' abilities to gain support from specific race/ethnic groups. Five parties emerged as ‘race parties' during this election; it was the ANC, Inkatha Freedom Party (IFP) and Pan Africanist Congress (PAC) as the ‘African parties' and the DP (Democratic Party) and FF (Freedom Front) as the ‘white parties'. Even though the 1994 elections seems as though it was voted according to the race-census theory, the underlying emotional driving force of the votes - mainly form ANC members - were recognised to be the binding force amongst the voters. This appearance seems to be a reflection of elections on the African continent.(6)


A close correlation exists between income and voting preference. This relationship is also witnessed between occupation class and voting preference. According to this, the 1994 elections did not merely happen according to the race census theory, but also according to a class census theory.(7) Seeing that a close correlation between income of the South African population and voting preference during the election existed, it is noted that the 1994 elections could just as well have been based on the class census theory.(8) This theory does not, however, act as an explanation of the political behaviour and it does not explain why people voted as they did during the election.(9)


There are, however, some shortages are noted. Even though many of the South African political parties are supported by a certain race/ethnic group, the entire race group does not necessarily support the political party exclusively. The IFP is recalled as an example. It is recognised as a Zulu-based party, yet a great number of the Zulu-population supports the ANC. In the same way, Afrikaans speaking South Africans do not orientate themselves towards a single party. During the 1994 election, their votes were dispersed among the NP (National Party), DP and the FF. Another interesting fact is that the parties that reinforced their race identities during the 1994 election (IFP, FF, PAC), fared much worse in the election than those parties that did not emphasise their racial identity.(10)


Another point of criticism is that the race census theory accepts race groups as homogenised groups. This is, however, a fallacy seeing that the coloured and Indian populations cannot be seen as homogenised race groups that support certain parties. Here, the class census theory should rather be applied to voting preference - the richer coloured and Indian populations supported the ANC during the 1994 election, whereas the poorer coloured and Indian populations supported the DP. Therefore a specific ‘coloured' or ‘Indian' vote do not exist in South Africa. This acts as the main point of evidence that race/ethnic identity is not seen as the primary motivation for voting preference in South Africa. This class difference does, however, have the potential to manifest as race division.(11)


Thus, the variables of race, ethnicity and class do not necessarily explain the voting behaviour of the South African population. Therefore, it is necessary to recognise the reason for this form of assembly. Membership to certain groups provides opportunities for the continuous interaction and perceptual reinforcement with other members in the same group. These group members can also act as appointment for interest. The group provides the members with a common political perspective. Thus, a correlation between race/ethnicity and vote is not necessarily the result of non-rational motivation. This is especially seen in a divided society like South Africa where it is inevitable that issues and interest are parallel to factors such as race and ethnicity. It can thus be said that race/ethnicity act as an informal shortcut that designate and signal voters.(12)


Medium-term influences


Amongst medium term influences, party identification, the left-right continuum and retrospective voting is recognised.(13) According to Jeremy Seekings(14), voters usually vote according to their social position. When people vote according to their social circumstance, their social characteristics determine their political preference. It seems that South African voters are voting according to their social characteristics, even when their individual opinions are different to their parties' policy positions. In this way, part identification plays an intermediating role between social factors and party preference. Except for the social aspect, party identification also has a psychological aspect as it shows toleration towards people's identification with their respective parties.


Party identification also acts as anchor that reinforces people with their voting preference. Therefore, people do not change their votes from election to election. It is however, important to note that party identification does not act as the only influence on voting preference. If it was, voting results would have merely reflected the balance between party identifications.(15) Party identification overlaps with various other influences, which will be discussed under the short term influences.


Seekings(16) recognises the possibility that party identification often occurs due to race and ethnic orientation. Here race and ethnicity, as well as class, seem to influence peoples' support for the different parties. Thus this method of voting is closely related to the race and class census theories.(17)


Furthermore, people identify with certain parties owing to their ideological orientation. This is expressed through the left-right continuum.(18) With regards to public opinion, there is a clear tendency amongst voters to tend towards the two extremities of the continuum. These extremities are expressed by liberalism to the left, and conservatism to the right.(19) It is important to note that even though the left-right continuum play an important role with regards to voting preference, this theory is not that relevant in South Africa and therefore cannot be regarded as one of the main sources of conflict which arises during election time. The reasons why it is difficult to explain voting preference according to this theory is owing to the fact that various South African political parties identify with more than one ideology and also owing to the parties in South Africa sharing their ideological orientation with regards to their respective policies.(20)

The third medium term influence is retrospective voting. This method of voting does not only include the evaluation of Government performance, but also the recognition of the presidential party. This is done through the evaluation of the president, where the evaluation of the economy and the level of economic well-being play an important role in the popularity of the president.(21) Voting according to issues as retrospective voting occurs owing to voters voting according to their evaluation of Government performance as witnessed in the past.(22) With this method of voting, there is a relationship between the voters' evaluation on the country's performance and which party to support. In general, for example, the ANC receives more support from those that feel that the country is in a better economic position as opposed to its economic position of the past. This influence of voting preference is, however, seen more amongst the white and coloured population where there is a strong relationship between party support and the voters' evaluation of the national circumstances.(23)


Short-term influences

Finally, the short term influence will be explained. These include issue voting, election campaigns as well as the candidates running for the different parties. Issue voting is seen as a relatively new theory as it developed during the 1960s. It emphasises the different parties' positions on important societal issues. This method of voting is closely related to retrospective voting and can therefore be seen as a rational and more accountable method of voting.(24) The role that policy issues play is, however, very uncertain within the South African context. When the American example is taken into account, it is clear that political issues, together with political ideologies, act as the determining factor of voting preference. One of the reasons for the growing popularity of issue voting is that issues and their respective (policy) decisions polarise towards the two extremities of the left-right continuum.(25) For this reason, more people are voting for the candidates that reflect their position on the issues best. This is in correlation with the political identification where people identify with political parties owing to the party's ideological orientation.(26)


With regards to South Africa, it seems as though issues do not play an important role in the influence of voting preference. Even though some similarities between people's feelings towards the issues and their voting preference do exist, it is not clear to what extent people are willing to drop other influences and vote only according to issues.(27) Here, party identification and other voting preferences still act as the main influences of voting preference. Furthermore, many of the South African voters do not have educated opinions on the formulation and execution of the policies and relevant issues. Thus they orientate themselves around policy issues, but these are usually influenced by other elements such as party identification.(28)


The second form of short term influences is campaigns. By the time that campaigns are launched, presupposed ideas of party orientation is already established. An election campaign does, however, provide the opportunity for political parties to convince smaller groups of people to vote for them. These groups of people consist of voters with poorly formed or no preferential procurement. Effective politico therefore does not only use campaigns as opportunities to make future-orientated issue-based promises, but also as opportunities to debate on the shortages of the past. In this way, they promote their own image and remind their voters of the mistakes of other parties in the past.(29)


The third short term influence is the candidates themselves. The candidates fall under this category because they change from election to election and also because the voters pay more attention to the candidates during the campaigning period. Effective candidates can increase votes and strengthen the support of the current supporters; whereas ineffective candidates can lose some of that support. South Africans are obliged to take their candidates into account through the use of the party list system and proportional representation system. Yet candidates do not play such an important role in voting preference amongst South Africans. It has been found that South Africans evaluate candidates in terms of their performance, rather than elements such as charisma. This indicates the voters' ability to make relatively sophisticated assessments over the capability of the candidates to fulfil their roles as party leaders.(30)


Conclusion


The different influences of voting preference can act as impetus for election related violence. It was during the 1994 elections that the potential for violence was recognised as a phenomenon that could be provoked by these different influences that induced voters' voting preference. The two opposing groups - the newly empowered owing to the acquisition of power; and the sometimes disgruntled, discouraged but also excited for change - lay the foundation for possible conflict during the first South African democratic elections. In order to avoid the election related violence, the need to understand the theory behind voting preference was realised. Factors including race, ethnicity and class as the long term influences; party identification, ideological orientation and retrospective voting as the medium term influences; and issue voting, campaigns and candidates as the short term influences, were recognised as issues to be aware of in order to manage possible conflict surrounding the election period.

 

Written by: Lize-Marie Smuts (1)


NOTES

(1) Contact Lize-Marie Smuts through Consultancy Africa Intelligence's Conflict & Terrorism Unit (conflict.terrorism@consultancyafrica.com).
(2) Mattes, R. 1995. The Election Book: Judgment and Choice in South Africa's 1994 Election. Cape Town: IDASA Public Information Centre.
(3) Seekings, J. 1997. "From the Ballot-Box to the Bookshelf: Studies of the 1994 General Election" in Journal of Contemporary African Studies. 15(2).
(4) Ibid.
(5) Ibid.
(6) Ibid.
(7) Ibid.
(8) Mattes, R. 1995. The Election Book: Judgment and Choice in South Africa's 1994 Election. Cape Town: IDASA Public Information Centre.
(9) Seekings, J. 1997. "From the Ballot-Box to the Bookshelf: Studies of the 1994 General Election" in Journal of Contemporary African Studies. 15(2).
(10) Mattes, R. 1995. The Election Book: Judgment and Choice in South Africa's 1994 Election. Cape Town: IDASA Public Information Centre.
(11) Habib, A. & Naidu, S. 1999. "Election '99: Was there an Indian and Coloured Vote?" in Politikon. 26(2).
(12) Mattes, R. 1995. The Election Book: Judgment and Choice in South Africa's 1994 Election. Cape Town: IDASA Public Information Centre.
(13) Ibid.
(14) Seekings, J. 1997. "From the Ballot-Box to the Bookshelf: Studies of the 1994 General Election" in Journal of Contemporary African Studies. 15(2).
(15) Erikson, R. & Tedin, K. 2005. American Public Opinion. New York: Person Longman.
(16) Seekings, J. 1997. "From the Ballot-Box to the Bookshelf: Studies of the 1994 General Election" in Journal of Contemporary African Studies. 15(2).
(17) Habib, A. & Naidu, S. 1999. "Election '99: Was there an Indian and Coloured Vote?" in Politikon. 26(2).
(18) Seekings, J. 1997. "From the Ballot-Box to the Bookshelf: Studies of the 1994 General Election" in Journal of Contemporary African Studies. 15(2).
(19) Heywood, A. 2002. Politics. New York: Palgrave MacMillan.
(20) Mattes, R. 1995. The Election Book: Judgment and Choice in South Africa's 1994 Election. Cape Town: IDASA Public Information Centre.
(21) Erikson, R. & Tedin, K. 2005. American Public Opinion. New York: Person Longman
(22) Seekings, J. 1997. "From the Ballot-Box to the Bookshelf: Studies of the 1994 General Election" in Journal of Contemporary African Studies. 15(2).
(23) Mattes, R. 1995. The Election Book: Judgment and Choice in South Africa's 1994 Election. Cape Town: IDASA Public Information Centre.
(24) Seekings, J. 1997. "From the Ballot-Box to the Bookshelf: Studies of the 1994 General Election" in Journal of Contemporary African Studies. 15(2).
(25) Erikson, R. & Tedin, K. 2005. American Public Opinion. New York: Person Longman.
(26) Ibid.
(27) Seekings, J. 1997. "From the Ballot-Box to the Bookshelf: Studies of the 1994 General Election" in Journal of Contemporary African Studies. 15(2).
(28) Mattes, R. 1995. The Election Book: Judgment and Choice in South Africa's 1994 Election. Cape Town: IDASA Public Information Centre.
(29) Ibid.
(30) Ibid.
CONSULTANCY AFRICA INTELLIGENCE

Voting preference in early democratic South Africa: Understanding the theory


Voting behaviour is recognised as one of the most influential study fields in South Africa. Voting preferences determine the voting behaviour of the voters and are dependent on various factors. The factors are divided according to long-, medium- and short term influences and are manifested, amongst others, as party identification, issue voting, outcome preference voting, retrospective voting, race (or class) census theory, as well as ideological orientation. It is, however, important to note that the line is often blurred between the different influences of voting preference and that the influences themselves often overlap. All the different influences of voting preference are not always evident in South Africa. This article will discuss and compare the main elements of voting preference in South Africa. It will make reference to the first democratic election in order to indicate the origin of voting preference influence that set the tone for the following elections in the country. Furthermore, this article will explain the importance of understanding the theory behind voting preference in order to overcome and avoid election-related conflict in South Africa.


Long-term influence


The most important long term influence is the underlying influence of race, ethnicity and class. An easy way to understand why citizens vote as they do is to compare their votes with observable voter characteristics. The variables that emerge are ethnicity, race and class. The correlation in ethnicity, race and class are measured in order to establish a logical reason for voting preferences.(2) Voting according to race is encapsulated in the race-census theory. This theory overlaps greatly with party identification, seeing that it is recognised as party identification on grounds of race. Party identification will, however, receive more attention later in this article. According to Horowitz(3), race and ethnicity are recognised as the most important factors that influence the individuals' political attitudes and behaviour. Owing to this, voting preferences are often build around race and ethnic identity.


With regards to the South African context, during the 1994 elections, voting preference was greatly formulated according to race preferences. It was especially noted amongst the black population as they, as a nation of race, supported the liberation movement against apartheid. This method of voting has, however, a heavy emotional and self-interest connotation seeing that the African National Congress (ANC) voters based their votes largely on emotions of excitement rather than to vote reasonably. Few supporters showed that they would vote according to issues and thus voting preference was based on symbols and the voters' concern over physical- and group security, as well as charisma.(4)


It is, however, important to note that this correlation between race/ethnicity and voting preference is rather seen as a mere description, seeing that people do not necessarily base their voting preference explicitly on their race/ethnic identity. In the same sense, race cannot be seen as the exclusionary factor that influences voting preference. Few voters recognise their parties as race/ ethnic exclusionary - even the voters with a strong race and ethnic identities do not necessarily identify with parties that only promote their own race/ethnicity.(5) Even though it did seem as though race/ethnicity was the basis of voting during the 1994 election, it was not the only influential factor.


The main reason why the 1994 election was seen as a race-driven election was owing to the parties' abilities to gain support from specific race/ethnic groups. Five parties emerged as ‘race parties' during this election; it was the ANC, Inkatha Freedom Party (IFP) and Pan Africanist Congress (PAC) as the ‘African parties' and the DP (Democratic Party) and FF (Freedom Front) as the ‘white parties'. Even though the 1994 elections seems as though it was voted according to the race-census theory, the underlying emotional driving force of the votes - mainly form ANC members - were recognised to be the binding force amongst the voters. This appearance seems to be a reflection of elections on the African continent.(6)


A close correlation exists between income and voting preference. This relationship is also witnessed between occupation class and voting preference. According to this, the 1994 elections did not merely happen according to the race census theory, but also according to a class census theory.(7) Seeing that a close correlation between income of the South African population and voting preference during the election existed, it is noted that the 1994 elections could just as well have been based on the class census theory.(8) This theory does not, however, act as an explanation of the political behaviour and it does not explain why people voted as they did during the election.(9)


There are, however, some shortages are noted. Even though many of the South African political parties are supported by a certain race/ethnic group, the entire race group does not necessarily support the political party exclusively. The IFP is recalled as an example. It is recognised as a Zulu-based party, yet a great number of the Zulu-population supports the ANC. In the same way, Afrikaans speaking South Africans do not orientate themselves towards a single party. During the 1994 election, their votes were dispersed among the NP (National Party), DP and the FF. Another interesting fact is that the parties that reinforced their race identities during the 1994 election (IFP, FF, PAC), fared much worse in the election than those parties that did not emphasise their racial identity.(10)


Another point of criticism is that the race census theory accepts race groups as homogenised groups. This is, however, a fallacy seeing that the coloured and Indian populations cannot be seen as homogenised race groups that support certain parties. Here, the class census theory should rather be applied to voting preference - the richer coloured and Indian populations supported the ANC during the 1994 election, whereas the poorer coloured and Indian populations supported the DP. Therefore a specific ‘coloured' or ‘Indian' vote do not exist in South Africa. This acts as the main point of evidence that race/ethnic identity is not seen as the primary motivation for voting preference in South Africa. This class difference does, however, have the potential to manifest as race division.(11)


Thus, the variables of race, ethnicity and class do not necessarily explain the voting behaviour of the South African population. Therefore, it is necessary to recognise the reason for this form of assembly. Membership to certain groups provides opportunities for the continuous interaction and perceptual reinforcement with other members in the same group. These group members can also act as appointment for interest. The group provides the members with a common political perspective. Thus, a correlation between race/ethnicity and vote is not necessarily the result of non-rational motivation. This is especially seen in a divided society like South Africa where it is inevitable that issues and interest are parallel to factors such as race and ethnicity. It can thus be said that race/ethnicity act as an informal shortcut that designate and signal voters.(12)


Medium-term influences


Amongst medium term influences, party identification, the left-right continuum and retrospective voting is recognised.(13) According to Jeremy Seekings(14), voters usually vote according to their social position. When people vote according to their social circumstance, their social characteristics determine their political preference. It seems that South African voters are voting according to their social characteristics, even when their individual opinions are different to their parties' policy positions. In this way, part identification plays an intermediating role between social factors and party preference. Except for the social aspect, party identification also has a psychological aspect as it shows toleration towards people's identification with their respective parties.


Party identification also acts as anchor that reinforces people with their voting preference. Therefore, people do not change their votes from election to election. It is however, important to note that party identification does not act as the only influence on voting preference. If it was, voting results would have merely reflected the balance between party identifications.(15) Party identification overlaps with various other influences, which will be discussed under the short term influences.
Seekings(16) recognises the possibility that party identification often occurs due to race and ethnic orientation. Here race and ethnicity, as well as class, seem to influence peoples' support for the different parties. Thus this method of voting is closely related to the race and class census theories.(17)


Furthermore, people identify with certain parties owing to their ideological orientation. This is expressed through the left-right continuum.(18) With regards to public opinion, there is a clear tendency amongst voters to tend towards the two extremities of the continuum. These extremities are expressed by liberalism to the left, and conservatism to the right.(19) It is important to note that even though the left-right continuum play an important role with regards to voting preference, this theory is not that relevant in South Africa and therefore cannot be regarded as one of the main sources of conflict which arises during election time. The reasons why it is difficult to explain voting preference according to this theory is owing to the fact that various South African political parties identify with more than one ideology and also owing to the parties in South Africa sharing their ideological orientation with regards to their respective policies.(20)

The third medium term influence is retrospective voting. This method of voting does not only include the evaluation of Government performance, but also the recognition of the presidential party. This is done through the evaluation of the president, where the evaluation of the economy and the level of economic well-being play an important role in the popularity of the president.(21) Voting according to issues as retrospective voting occurs owing to voters voting according to their evaluation of Government performance as witnessed in the past.(22) With this method of voting, there is a relationship between the voters' evaluation on the country's performance and which party to support. In general, for example, the ANC receives more support from those that feel that the country is in a better economic position as opposed to its economic position of the past. This influence of voting preference is, however, seen more amongst the white and coloured population where there is a strong relationship between party support and the voters' evaluation of the national circumstances.(23)


Short-term influences

Finally, the short term influence will be explained. These include issue voting, election campaigns as well as the candidates running for the different parties. Issue voting is seen as a relatively new theory as it developed during the 1960s. It emphasises the different parties' positions on important societal issues. This method of voting is closely related to retrospective voting and can therefore be seen as a rational and more accountable method of voting.(24) The role that policy issues play is, however, very uncertain within the South African context. When the American example is taken into account, it is clear that political issues, together with political ideologies, act as the determining factor of voting preference. One of the reasons for the growing popularity of issue voting is that issues and their respective (policy) decisions polarise towards the two extremities of the left-right continuum.(25) For this reason, more people are voting for the candidates that reflect their position on the issues best. This is in correlation with the political identification where people identify with political parties owing to the party's ideological orientation.(26)


With regards to South Africa, it seems as though issues do not play an important role in the influence of voting preference. Even though some similarities between people's feelings towards the issues and their voting preference do exist, it is not clear to what extent people are willing to drop other influences and vote only according to issues.(27) Here, party identification and other voting preferences still act as the main influences of voting preference. Furthermore, many of the South African voters do not have educated opinions on the formulation and execution of the policies and relevant issues. Thus they orientate themselves around policy issues, but these are usually influenced by other elements such as party identification.(28)


The second form of short term influences is campaigns. By the time that campaigns are launched, presupposed ideas of party orientation is already established. An election campaign does, however, provide the opportunity for political parties to convince smaller groups of people to vote for them. These groups of people consist of voters with poorly formed or no preferential procurement. Effective politico therefore does not only use campaigns as opportunities to make future-orientated issue-based promises, but also as opportunities to debate on the shortages of the past. In this way, they promote their own image and remind their voters of the mistakes of other parties in the past.(29)


The third short term influence is the candidates themselves. The candidates fall under this category because they change from election to election and also because the voters pay more attention to the candidates during the campaigning period. Effective candidates can increase votes and strengthen the support of the current supporters; whereas ineffective candidates can lose some of that support. South Africans are obliged to take their candidates into account through the use of the party list system and proportional representation system. Yet candidates do not play such an important role in voting preference amongst South Africans. It has been found that South Africans evaluate candidates in terms of their performance, rather than elements such as charisma. This indicates the voters' ability to make relatively sophisticated assessments over the capability of the candidates to fulfil their roles as party leaders.(30)


Conclusion


The different influences of voting preference can act as impetus for election related violence. It was during the 1994 elections that the potential for violence was recognised as a phenomenon that could be provoked by these different influences that induced voters' voting preference. The two opposing groups - the newly empowered owing to the acquisition of power; and the sometimes disgruntled, discouraged but also excited for change - lay the foundation for possible conflict during the first South African democratic elections. In order to avoid the election related violence, the need to understand the theory behind voting preference was realised. Factors including race, ethnicity and class as the long term influences; party identification, ideological orientation and retrospective voting as the medium term influences; and issue voting, campaigns and candidates as the short term influences, were recognised as issues to be aware of in order to manage possible conflict surrounding the election period.

Written by: Lize-Marie Smuts (1)

NOTES

(1) Contact Lize-Marie Smuts through Consultancy Africa Intelligence's Conflict & Terrorism Unit (conflict.terrorism@consultancyafrica.com).
(2) Mattes, R. 1995. The Election Book: Judgment and Choice in South Africa's 1994 Election. Cape Town: IDASA Public Information Centre.
(3) Seekings, J. 1997. "From the Ballot-Box to the Bookshelf: Studies of the 1994 General Election" in Journal of Contemporary African Studies. 15(2).
(4) Ibid.
(5) Ibid.
(6) Ibid.
(7) Ibid.
(8) Mattes, R. 1995. The Election Book: Judgment and Choice in South Africa's 1994 Election. Cape Town: IDASA Public Information Centre.
(9) Seekings, J. 1997. "From the Ballot-Box to the Bookshelf: Studies of the 1994 General Election" in Journal of Contemporary African Studies. 15(2).
(10) Mattes, R. 1995. The Election Book: Judgment and Choice in South Africa's 1994 Election. Cape Town: IDASA Public Information Centre.
(11) Habib, A. & Naidu, S. 1999. "Election '99: Was there an Indian and Coloured Vote?" in Politikon. 26(2).
(12) Mattes, R. 1995. The Election Book: Judgment and Choice in South Africa's 1994 Election. Cape Town: IDASA Public Information Centre.
(13) Ibid.
(14) Seekings, J. 1997. "From the Ballot-Box to the Bookshelf: Studies of the 1994 General Election" in Journal of Contemporary African Studies. 15(2).
(15) Erikson, R. & Tedin, K. 2005. American Public Opinion. New York: Person Longman.
(16) Seekings, J. 1997. "From the Ballot-Box to the Bookshelf: Studies of the 1994 General Election" in Journal of Contemporary African Studies. 15(2).
(17) Habib, A. & Naidu, S. 1999. "Election '99: Was there an Indian and Coloured Vote?" in Politikon. 26(2).
(18) Seekings, J. 1997. "From the Ballot-Box to the Bookshelf: Studies of the 1994 General Election" in Journal of Contemporary African Studies. 15(2).
(19) Heywood, A. 2002. Politics. New York: Palgrave MacMillan.
(20) Mattes, R. 1995. The Election Book: Judgment and Choice in South Africa's 1994 Election. Cape Town: IDASA Public Information Centre.
(21) Erikson, R. & Tedin, K. 2005. American Public Opinion. New York: Person Longman
(22) Seekings, J. 1997. "From the Ballot-Box to the Bookshelf: Studies of the 1994 General Election" in Journal of Contemporary African Studies. 15(2).
(23) Mattes, R. 1995. The Election Book: Judgment and Choice in South Africa's 1994 Election. Cape Town: IDASA Public Information Centre.
(24) Seekings, J. 1997. "From the Ballot-Box to the Bookshelf: Studies of the 1994 General Election" in Journal of Contemporary African Studies. 15(2).
(25) Erikson, R. & Tedin, K. 2005. American Public Opinion. New York: Person Longman.
(26) Ibid.
(27) Seekings, J. 1997. "From the Ballot-Box to the Bookshelf: Studies of the 1994 General Election" in Journal of Contemporary African Studies. 15(2).
(28) Mattes, R. 1995. The Election Book: Judgment and Choice in South Africa's 1994 Election. Cape Town: IDASA Public Information Centre.
(29) Ibid.
(30) Ibid.