Theresa May confirms to exit as PM on June 7
24 May 2019 – 15:42 | No Comment

After the UK Parliament rejected her Brexit plans for the third time, UK Prime Minister Theresa May has decided to step down as leader of the Conservative Party.
She announced her departure after talks with Graham …

Read the full story »

Energy & Environment

Circular Economy

Climate Change


Home » African Elections, Democracy & Governance, Elections and Governance, electoral, Focus, International, Race

Australia and South Africa: The Major Social Factors Influencing Voting Patterns

Submitted by on 05 Mar 2014 – 10:56

Nawid Zaher

Nawid Zaher

By Nawid Zaher, Editorial Researcher, Government Gazette

There are many social factors that influence voting patterns in federal elections. However, often there will be one dominant factor that can explain key patterns. Australia and South Africa are no exception to this theory and analysis of voting data reveals one or more major social factors that appear to correlate with the voting patterns of the electorate. In Australia, level of income and wealth are the major social factors that appear to be dictating the voting patterns of certain electorates. In South Africa, race has been the major social factor that appears to be dictating the voting of the nation – ‘so much so that many have described South African elections as a racial census’ (Ferree 2006, p. 803). This article will examine the relationship of these major social factors and their impact on voting in Australia and South Africa. For the purposes of Australia, historical data representing voting in federal elections is utilised to demonstrate a long and powerful history of voting in wealthy electorates. For the purposes of South Africa, the relationship between race and class is discussed to highlight the significance of the relationship and the consequential impact of underlying social classes on voting patterns. Race census theory is established to offer a framework for understanding the political and voting landscape of the nation. Overall, understanding these social characteristics can assist in predicting voting behaviour and this is a reflection of the impact such factors can have on the electorate.


There are 150 electorates in Australia and analysis of voting data from these electorates produces an obvious trend. Information from the Australian Electoral Commission (AEC) indicates a number of electorates have never been held by certain political parties. Thus, a stranglehold exists from one party as the electorate has voted in support of them for over a century of federal elections. This paper will identify similarities and patterns within these electorates to determine the major social factor influencing such voting patterns.

The AEC (2013) lists Wentworth, North Sydney, Curtin, Warringah and Bradfield as five federal electorates that have never been held by a non-conservative party since the first federal election in 1904. This is a powerful statistic considering four of the aforementioned five are in the vicinity of one another in the dominant and influential districts of Sydney (the major and most populous city of Australia). In another publication, the Sydney Morning Herald (2013) states the top five electorates in Australia by median family income per week as Wentworth, North Sydney, Curtin, Warringah and Bradfield.

These are the same electorates that have indeed never been held by the Labor party and always held by conservatives (AEC 2013). Furthermore, to reinforce this relationship between wealth and voting, two out of the next five electorates from the Sydney Morning Herald’s (2013) list of electorates by median family income per week have never been held by the Labor party either. A significant voting pattern has been created by this data given the longevity of the voting tendencies of the citizens of these seats.

The major social factor in these key electorates is level of income and therefore wealth. This is obviously a fundamental feature and the historical data available supports this concept. It is difficult to argue that this is a coincidence when the pattern is consistent across a number of electorates and for the entirety of the electorate’s existence. Deeper analysis of the electorate supports the hypothesis that wealthier electorates vote for conservative political parties (the Liberal party in Australia) and this is a significant voting pattern of Australia.

The electorate of Wentworth, which records the highest median family income of any of the 150 electorates in the country, contains some of Sydney’s richest and most established suburbs.  This is reflected by the fact it has the highest proportion of high income families (Green, 2013). The electorates of North Sydney and Warringah are in second and fourth place respectively according to highest median family income and both endure comparable characteristics. A feature of Warringah is that it records the second highest figure in relation to monthly mortgage repayments and weekly rent (Green, 2013); reflective of the fact it is indeed an expensive place to reside. Curtin in third place, which is the only electorate in the top five that is not within the vicinity of Sydney, is a mining electorate that benefits from the economic prosperity that Australia’s natural resources offer. Rounding out the top five is the electorate of Bradfield. The majority of voters in Bradfield are affluent and well educated. ‘Bradfield contains a large proportion of families in big houses that they own and the highest proportion of residents attending school’ (Green, 2013). None of the aforementioned five electorates have ever been held by a non-conservative political party.

Understandably, citizens within these electorates who wish to cast a vote for a non-conservative party are an anomaly and are likely to feel that their vote will have little to no influence on the outcome of the election. This has the potential to breed the mentality of following the herd and create an environment of self-prophecy where the electorate continues to strengthen as a conservative stranglehold. This predictability in voting patterns blunts the potency of an individual’s vote in an election where he or she chooses to go against the trend.

Thus there is an overwhelming voting pattern in Australia that affluent, highly educated and white-collar electorates appear to be regions where conservatives such as the Liberal Party will hold perpetual power. This relationship between social factors such as wealth and voting patterns in Australia is consistent and highly correlated. In these significant electorates where the rich hold strong influence, the impact of your vote is superfluous if it is not for the conservative party. As a result, appreciating voting patterns is crucial for policy-makers as ‘better understanding as to what makes individuals support a left-wing or right-wing party can provide new insights into how each party’s policies affect different groups in the population’ (Leigh 2004, p. 30).

However, in South Africa it appears voters are not voting ‘left or right’ and instead are voting ‘black or white’.

South Africa

A number of theories exist to explain voting patterns within the political landscape of post-Apartheid South Africa, however the leading one is the race census theory. In South Africa, ‘race has been an overwhelming predictor of voting behaviour for most of the South African electorate’ (Ferree 2006, p. 803). Scholars, who assert that race has been a major influence in the results of the four federal elections since the end of Apartheid, have largely supported this view. Understanding racial identity is thus central to understanding the relationship between the electorate and voting patterns. It is racial identity that drives voting patterns and these can be divided into four main categories according to Taylor (1996), ‘Whites’, ‘Coloureds’, ‘Indians’ and ‘Blacks’.

Race census theory is a school of thought that advances the view that voting patterns in South Africa resemble a prism of racialised politics because racial and ethnic political identities dominate the voting landscape. Therefore, blacks vote for ‘blacks parties’ and whites for ‘whites parties’ (Habib & Naidu 2006, p. 82). Historically, these have been the African National Congress (ANC) and the National Party (NP) respectively. In analysing the voting statistics from the first post-apartheid election in regards to racial categories; such a correlation does appear to exist between racial groups and political parties:

Racial Category

Population (%)

 (%) Votes for ANC






Racial Category

Population (%)

(%) Votes for NP






The above tables have been constructed from statistics in Taylor’s (1996) ‘Race and the transition to democracy in South Africa’. It reports that 75% of the population constituted African and 13% of the population were labelled white and that this corresponds to statistics, which report that the African National Congress received 63% of the vote and the National Party received 20% of the vote. The differences in this situation have been accounted for by the presence of other parties and in particular by the voting behaviour of those categorised Coloured and Indian, who are constituted as being 9% and 3% of the population respectively (Taylor 1996). These statistics from the first election post-apartheid reflect a rather strong correlation between race and voting patterns.

 Race and Class

Aside from race alone, another important relationship to consider for the purposes of studying voting patterns in South Africa is that of race and class. Cohen (2012) summarises a report from Statistics South Africa by reporting the fact that even though incomes for black households have increased an average of almost 170 percent over the decade between 2002 – 2012, their annual earnings are still only a sixth of that for whites. This issue is severely compounded by the fact that ‘about 80 percent of South Africa’s 51.8 million population is black’ (Cohen, 2012). This assists in advancing the racial census theory in South Africa as race continues to politically and economically segregate and influence citizens’ voting patterns. Such statistics are reflective of an environment where race is correlated to serious inequality in social classes.

Thus, when captured by racial categories, South African voting patterns appear to follow a predictable pattern. Indeed, as previously mentioned, it is dominated by the voting landscape of blacks voting for ‘blacks parties’ and whites for ‘whites parties’. Furthermore, as a result of the serious social and economic inequality issues confronting South Africa, consistent voting patterns in regards to social classes emerge as the statistics indicate strong divides between social classes. These classes (measured by level of income) ultimately offer an overlap between the voting patterns in Australia and South Africa. Australia has retained a concentrated number of elite wealthy electorates where the result is consistently for the Liberal party whilst South Africa’s wealthier social class (‘whites’) vote for the National Party. The underlying pattern in both countries is that poorer social classes vote for one party and the wealthier ones elect to vote for the other major political party and that this is the key driver for voting behaviour.