Voting and the fear factor

Our latest opinion poll found that many black South Africans are afraid to vote as they wish.

The upheaval at the Independent Electoral Commission and the continuing difficulties over voter registration that Patrick Laurence explores both undermine confidence that the coming election will be free and fair. But the Helen Suzman Foundation’s most recent opinion survey suggests a more direct threat: the fear engendered among voters by political intolerance in their communities.

Twenty-two per cent of the African voters interviewed say that they would find it impossible or dangerous to disagree politically with politicians. Slightly less, 21 per cent, feel it would be impossible or dangerous to disagree with street and area committees, 20 per cent say the same about traditional chiefs and 17 per cent about civics. Since people are unlikely to say that "it is impossible or dangerous" to disagree politically with a particular group unless they have in mind concrete instances where people have been beaten up or killed for expressing such disagreement, these figures show that South Africans are still subject to many undemocratic pressures.

The situation has improved slightly over the past four years, but nowhere can one feel that African voters are truly able to act freely. Before the 1994 election many people thought that it would be either difficult or impossible to live in their communities if they held political views at variance with those of their neighbours. Election studies that I conducted with Professor Lawrence Schlemmer at the time found over and over again that substantial numbers of South Africans believed that their ballot was not secret, that they tended to live in areas dominated by only one party, and that people not supporting that dominant party often felt worried or frightened. A mid-term survey by the Helen Suzman Foundation in October 1996 (reported in Focus 6 and 7) found many voters still reported feeling such pressures even though election fever had long since subsided.

The Foundation’s most recent opinion poll conducted in October 1998 by MarkData began by asking voters whether it was normally possible for someone to live in their community without anyone knowing which party he or she supported and voted for. We then compared the results with the answers to the same question posed in February 1994. The results are set out in broken down by race and region.
Among Whites, and especially among Asians, there has been a sharp increase in confidence in the confidentiality of the voting process. Many Asians approached the 1994 election in a state of great anxiety and this was doubtless why 40 per cent (an exceptionally high figure) then believed their vote was not secret. Happily those anxieties have diminished: Asians are now the group least worried about voting secrecy. African voters, however, remain as likely as in 1994 to believe that others will know their party preference.

The most striking change is the sharp increase in the numbers of Coloureds who believe that it is impossible to keep their political convictions hidden from their community. This increase reflects the heightened competition for Coloured votes that has followed the 1994 battle between the African National Congress and the National Party in the Western Cape and the ANC’s intense efforts to reverse that verdict ever since. There is a class aspect to this finding: working-class Coloureds are more likely than middle-class Coloureds to feel that their political affiliations could not be kept secret. Given the New National Party’s (NNP) strength in working-class areas and the ANC’s stronger hold among middle-class Coloureds, it is not surprising that 39 per cent of NNP supporters said their views would be known compared to 29 per cent of ANC supporters. Regionally, this effect was clear in the sharp increase in the number of Western and Northern Cape voters who believe their party preference would be known within their community. There has also been a sharp, and unexpected, increase in the number of such voters in the Free State.

Compare the numbers seen in 1994 and 1998 by party support. Supporters of the United Democratic Movement are the most likely to believe that their communities will know how they vote. The number of Pan African Congress (PAC) and Inkatha Freedom Party (IFP) voters believing this has also risen, perhaps because supporters of these two parties tend to live among otherwise ANC-voting communities who are keenly aware of their political rivals. On the other hand, the number of Freedom Front (FF) and other white right-wing voters feeling pressure from within their communities has fallen since 1994.

As in 1994 we asked voters whether there are parties or groups controlling the area in which they live who try to make sure that they vote in particular way. Among White and Asian voters there has been a welcome decline in those believing this but among Coloured and African voters the number answering positively has more than doubled. By now, with the experience of both national and local elections, voters are probably thoroughly aware of the dominant party in their area. In most parts of the country awareness of territorial domination by political parties has risen to levels previously only seen in KwaZulu-Natal. The exception is the Eastern Cape where far fewer voters than before feel that their area is controlled by a particular political party. This no doubt reflects the political competition inaugurated there by the arrival in strength of Bantu Holomisa’s UDM.

In October 1996 we asked voters how easy it would be to live in their community if their political views differed from those of other people living around them. No less than 46 per cent said that it would be difficult and a further 9 per cent said it would be impossible. We asked the same question in October 1998 and a clear improvement is visible.

The trend is clear, too, when we looked at some of the groups who are most affected. In 1996 no less than 83 per cent of IFP voters said that it would be difficult or impossible for them to live in a community if their political views differed from the majority, but in our latest survey this figure has fallen to 60 per cent. For uncommitted voters the proportion fell from 70 per cent to 49 per cent. But majorities in KwaZulu-Natal, the Western Cape and Northern Cape still believe that it would be difficult or impossible to live in that situation. Rural Coloureds, particularly ANC voters, are most likely to feel these community pressures, while among Africans such pressures peak among hostel dwellers and squatters. Although IFP supporters — and African voters in KwaZulu-Natal in general — feel such pressures most, they also remain very high in Gauteng. In the country as a whole 49 per cent of African still think they could not live in a community where their political preference differed from the majority.

In the run up to the 1994 election Lawrence Schlemmer and I also looked at how far groups such as churches, trade unions, traditional chiefs, street and area committees influenced voter choice. We found that most of these groups had a significant influence, especially on African and Coloured voters. The picture was not so much that any particular group had overwhelming influence but more that both coloured and African voters lived within a nexus of extremely dense community networks and pressures. White and Indian voters, on the other hand, were not subject to such intense pressures. The Foundation’s latest survey again looked at the influence of various groups.

We asked voters first how easy or difficult it is to disagree politically with their family members and with their friends and colleagues. In most countries these intimate groups have the greatest influence on individual voting; in particular, voters tend to inherit the political views of their parents. In South Africa these influences exist but they are less powerful than those exerted by more external forces.

No white English-speakers, only 8 per cent of white Afrikaans-speakers and low proportions among Asians and Coloureds say that it would be difficult to disagree with their family members. Among Africans 17 per cent say that it would be difficult to disagree with family members and 4 per cent think it would be impossible or dangerous. Friends and colleagues exert more influence among all groups, with coloureds and Asians again more likely to be influenced by them than other groups.

When we asked how difficult it would be for voters to disagree with trade unions, the figures leap for all races. The figures are almost as high for coloureds as for Africans especially among NNP-supporting coloureds, with 30 per cent saying it would be difficult to disagree politically with the trade unions and another 11 per cent that it would be impossible/dangerous to do so. Among Africans only the figures are even higher and are sharply differentiated by the type of community in which voters live. Half of all squatters and hostel dwellers, who live in the most densely populated surroundings, say that it would be difficult or even impossible/dangerous to disagree with the trade unions. The fact that no less than 30 per cent of hostel dwellers think it would be impossible or dangerous to disagree with the unions is a particularly alarming figure.

The levels of constraint seem even higher when we examine the influence of the civics with almost half of all African voters saying it would be difficult or impossible/dangerous to disagree with them. Table 1 breaks these statistics down for African voters only by party support. Levels of perceived duress are particularly high among IFP supporters, followed by UDM voters. The somewhat lower figures for ANC supporters may be explained by the fact that the civics are generally exerting a pro-ANC influence, accordingly those who spontaneously support the ANC may feel less pressure than those who do not.

Regionally, we found that although pressures on Africans are predictably high in KwaZulu-Natal, the highest figures were recorded in Gauteng and the Eastern Cape. Clearly, the pressures on voters in many of the big reef townships are still great, while the battle between civics and traditional leaders in the Eastern Cape has doubtless taken its toll.

When we asked the same question about traditional chiefs, no less than 53 per cent of all African voters said it would be difficult or impossible/dangerous to disagree with them politically. The figures are highest among rural dwellers and people living in squatter camps among IFP supporters and in KwaZulu-Natal, but once again the figures for Eastern Cape and Gauteng are extremely high.

We then asked the same question about street and area committees. Among Whites and Asians the figures are higher than one might have expected, no doubt reflecting the influence of neighbourhood associations of one kind or another. Once again, however, it is among coloured and African voters that the figures are highest. When the figures are broken down for Africans only we see that the figures peak in KwaZulu-Natal, the Eastern Cape and Gauteng but significant constraints are evident in Mpumalanga, Northern Province and the North West.

Then we asked voters how easy or difficult it is for them to disagree politically with politicians. To some extent this question may be measuring ordinary party loyalty, with voters acknowledging that once their political leaders have taken up a position they will find it hard to adopt independent positions themselves. Nonetheless, the overall figures are quite striking.

Among Whites DP voters feel much more able to disagree politically with their leaders than either the NNP or Freedom Front voters, while among coloureds a majority of NNP voters think it would be difficult or impossible/dangerous to disagree with political leaders. Among Africans 58 per cent said they would find it difficult or impossible/dangerous to disagree politically with politicians, with quite remarkable figures being seen among IFP supporters: 82 per cent of them say they would find it difficult or even impossible or dangerous to disagree with politicians.

Finally, we amalgamated all these figures about the influences that African voters are subject to and expressed them in terms of the type of community in which they lived. This shows the hierarchy of influences at work among African voters living in different circumstances. Among squatters and hostel dwellers the influence of trade unions, civics, street/area committees, politicians and even chiefs is far higher than among rural and township dwellers. The latter are more influenced by family, friends and colleagues. Overall Table 1 suggests that at least a fifth of the African electorate is experiencing undemocratic pressures. In practice the figure is probably higher since groups such as trade unions, civics etc do not wholly overlap. Probably between a quarter and a third of the African electorate does not feel free to vote as it wishes.