Where the parties stand

Lawrence Schlemmer compares the results of three national opinion surveys since 1994.
Looked at broadly, our three MarkData surveys of party choices since 1994 suggest the entrenchment of a pattern of party imbalance (Table 1). After a fall-off of initial liberation euphoria, the ANC has settled into a stable super-dominant position among voters at a level just short of a two-thirds majority. Many wonder whether South Africa is heading for another one-party dominant state with demoralised opposition parties and voters unresponsive to the issues on which effective governance and party choice should be based in a pluralist democracy.


 Table 1: Voter choice in 3 national surveys (%)
   Feb 94  Oct 96   Oct 97
 ANC/SACP  69.4  64.4 64.2
 PAC  1.3  2.8  3.5
 NP  16.9  16.0  11.7
 DP  1.1 2.2  2.1
 IFP  6.7  10.3  9.4
 CP  1.0  1.0  1.8
 UDM  NA  NA  4.3
 ACDP  0.5  0.7  0.4
 Other  1.5  0.3  0.7
NB The results of a pre-election survey instead of the actual election results are used for 1994. Don’t knows and persons who will not vote are excluded.


The action, as it were, has been among the minority parties. The National Party, as is now common knowledge, has lost a major part of its support. The Pan Africanist Congress, the Democratic Party, the Freedom Front and the Conservative Party have all increased their support, the PAC more than the others. The Inkatha Freedom Party has lost support very slightly, particularly among non-African voters. The surprise is the United Democratic Movement, the party of Bantu Holomisa and Roelf Meyer, which has grown very rapidly over the past months. This may be a short-run success based on the novelty of two such unlikely bedfellows or it could be the early growth of a party with the capacity to introduce significant liquidity into the political market.

Is the dominant position of the ANC as secure as it appears to be? Perhaps not. First, ANC support as a second choice with voters has fallen from 9 per cent in 1996 to 3 per cent in 1997. This means that the party is losing some of the fringe of potential support that surrounds its primary support base.


Second, in the 1997 survey we asked voters how close or distant they felt towards all the parties, in order to determine the extent of sympathy for a party, and found a sharp reduction of intense sympathy for the ANC since 1994. The proportions of all voters who felt “very close” to the ANC fell from 47 per cent in 1994 to 35 per cent in 1997.

The results also showed that the ANC is the only party that gets more first choices than sympathy. All other parties get more sympathy than first choices (Table 2). These findings suggest that the ANC, the IFP and the CP have achieved, or nearly achieved, their highest potential support levels. Parties with significant zones of voter sympathy and, therefore, growth potential are the PAC in particular, but also the UDM and the DP. The wider sympathy zone of the NP may be more a fading reflection of its past strength than a sign of future vigour.


Table 2: October 1997: First choice of party (rounded %) and feeling “close” or “very close” to parties (%)
    First choice
 of party
  Feel close/
very close
 ANC 65 57
 PAC 4 24
 NP 12 15
 DP 2 7
 IFP 9 10
 FF 2 4
CP  2 3
 UDM 4 11


The fact that the ANC has more voter support than sympathy indicates that it is unlikely to achieve a two-thirds majority in next year’s national election. The excess of the ANC’s current support over sympathy is due to a variety of factors. First, as the governing party it gets a measure of nominal support from those people who like to be correctly positioned in relation to power, influence and political opportunity - the zone of patronage. Second, its sheer size creates a bandwagon that attracts support of a nominal kind. Third, it gets much more publicity than any other party, largely due to its executive functions and its historical significance.


Finally, factors such as the constraints identified in our earlier research may still exist and benefit the ANC in its areas of dominance. A key finding in the 1994 election was a disturbingly high level of constraint on free choice among all voters. Aside from evidence of some direct intimidation, the more pervasive constraint was a general pressure to conform to majority choices within the stronghold-areas of particular parties, most notably in white right-wing areas and in the mass, low-cost residential or rural areas dominated by either the ANC or the IFP. A mid-term survey carried out in October 1996 found that such constraints still existed to an unhappy degree.


The racial images of parties in South African politics and the way this relates to the history of apartheid are important influences on voting choices. We repeated an item from the 1996 survey: “No matter how good their policies may be, I will never feel able to support parties that used to be supported by whites before 1994.” (Obviously, this item could not be fielded among whites). Some 64 per cent of African voters agreed with the statement, roughly the same level as we found in 1996 (65 per cent). Among ANC supporters the proportion is 69 per cent (67 per cent in 1996), and among IFP supporters it is 64 per cent (67 per cent in 1996).


This suggests very little change over the past year in the African electorate as a whole, but there have been some marked regional shifts. Whereas in 1996 only 39 per cent of voters in the Eastern Cape endorsed this statement, by late 1997 it had risen to 72 per cent. In the Northern Cape it rose from 29 per cent in 1996 to 45 per cent in 1997. Among Swazi-speaking voters the rise has been from 39 per cent to 71 per cent.


These shifts indicate that the old-order parliamentary parties, the NP and the DP in particular, are substantially “off-limits” for African voters. The causes of this are probably complex, but the huge publicity given to past events by the Truth and Reconciliation Commission must be a major factor. Unpopularity of the older parliamentary parties is greatest in provinces in which African-controlled parties are super-dominant, such as the Eastern Cape (72 per cent endorsement), Mpumulanga (76 per cent), North West (82 per cent) and KwaZulu-Natal (64 per cent). It has least purchase in provinces in which former parliamentary parties have more weight, such as Gauteng (52 per cent), Northern Cape (45 per cent) and Western Cape (44 per cent). Clearly, the notion of the “illegitimacy” of the NP and the DP is more difficult to promote in provinces in which the old parliamentary parties have a muscular presence.


But the exclusion of alternative parties from the range of choice for voters could be based on current racial and ethnic labelling without it necessarily being a reference to the apartheid past. We asked voters to choose between the two following statements: “I support my political party because it represents people of my group” or “I support a political party because of what the party wants to achieve”. Some 34 per cent of all African voters endorsed the group loyalty statement. The level for Afrikaners was 35 per cent, for coloured people 36 per cent and for Indians 38 per cent. Whites as a whole were slightly less con-cerned with group-based politics, endorsing the item at a level of 29 per cent. And since any admission of group loyalty in politics is not exactly fashion-able in the new South Africa, one must assume that the level of endorsement is an underestimate. The fact that at least one-third of African voters and many non-African voters are prepared to endorse this alternative rather than one that refers to the more obvious policies and goals of a party, shows just how powerful racial and ethnic symbolism remains in South African politics.

The 1996 survey showed what R.W. Johnson referred to as an “undiscriminating racial solidarity created by apartheid”, and the 1997 results are even less encouraging. Aside from the evidence just mentioned, only some 3 per cent of whites, for example, chose the ANC as either their first or second choices (down from 6 per cent in 1996), and only 6 per cent of Africans chose the NP as first or second choice (down from 13 per cent in 1996). Among Africans only 1 per cent chose the DP in first or second position, despite that party’s attempts over the years to promote the politics of equal opportunity. Racial polarisation is deepening, and the so-called rainbow nation is a notion that lives only within a small non-racial intelligentsia and among a quarter to a third of coloured and Indian voters. The challenge posed by the UDM will be the acid test of the strength of racial solidarity. So far it has made a promising early start, but the odds against it are enormous.

The racial solidarity aspect of South African politics is also latent in the replies to another item repeated from our earlier polls: “I will support and stand by my political party and its leaders even if I disagree with many of its policies and actions.” Among all voters, some 53 per cent of voters endorsed this mindless “house spirit”; 33 per cent among Indians, 36 per cent among whites, 49 per cent among coloured people and 59 per cent among Africans. Among Africans, the level of endorsement has increased by 11 percentage points since 1996. Even among those African voters with post-matric qualifications, 58 per cent display this automatic commitment. If a sophisticated voter group can respond in this way, it is not surprising that 47 per cent of the most alienated group, the unemployed poor, should respond in the same way, despite their discontent.

IFP and ANC supporters, as in 1996, are very similar in their responses on this item, with 66 per cent and 63 per cent of the two categories displaying an undiscriminating commitment to their parties. The least prone to this were white-collar voters (42 per cent), Shangaan-speakers (50 per cent), Zulus (52 per cent), Swazis (35 per cent) and that very small category of African niche voters, the English-speaking Africans (13 per cent). The Eastern Cape and Northern Cape voters, who seemed to deviate from the norm in the 1996 survey, have been brought back into line, and currently are quite as inclined as African voters in any other province to endorse their party, right or wrong.

Some of this solidarity, however, may be phoney and could be the result of fear and caution induced by local political dominance. Hence we also repeated the question “How difficult or easy would it be for people like yourself to live in your present neighbourhood if your political views were different from most other people?” The results for 1996 and 1997 are virtually identical: 64 per cent and 63 per cent of African voters answered that it would be “difficult” or “impossible”. The results for the main African parties, the ANC and the IFP were, as before, indistinguishable.

The proportions of people in some categories and regions who replied that this kind of political dissent would be impossible are very high. Among PAC supporters, for example, 23 per cent felt that it would be impossible, which may partly explain why the support for this party lags so badly behind its sympathy levels. Among the African intelligentsia — those with post-matric qualifications — 22 per cent felt that it would be impossible, showing once again that humble clerks are more at liberty to make up their own minds than more visible and vulnerable graduates. In the Northern Cape, the Eastern Cape, the Free State and North West province, between one quarter and one third felt that political dissent would be impossible. In these provinces well over 80 per cent of African voters considered that differing politically from one’s neighbours would be either impossible or difficult.

Here again we see that political dominance has a tendency to reinforce itself. Provinces in which a greater degree of balance exists between parties, KwaZulu-Natal, the Western Cape and Gauteng, fewer people felt constrained in their choice. One-party dominance breeds a culture of political intolerance that can restrict freedom without any formal infringements of civil liberty.

These patterns of response are mirrored in the responses on the issues of multiparty competition and opposition in a democracy. Some 50 per cent of African voters supported the idea of fewer political parties in a democracy in order to promote unity, compared to 34 per cent who favoured many parties competing in order to encourage government to be accountable. African supporters of the NP, the PAC and the UDM were the exceptions to this anti-pluralist view, perhaps for opportunistic reasons. Only 36 per cent of African voters expressed support for a strong and critical opposition compared with 62 per cent who wanted either no opposition or a “co-operative” opposition. Here African PAC, DP, UDM and to some extent IFP supporters stand out, with majorities favouring tough opposition.

Some 54 per cent of all African voters feel that an “extremely dominant” position for the ANC in South African politics will be a “very good” or “reasonably good thing” for the country. One must obviously forgive this sentiment among ANC supporters, but much more sobering is the proportions of African opposition party supporters who felt that a super-dominant ANC will be good for South Africa (Table 3). On the basis of these results, we are inclined to conclude that the consolidation of South Africa’s democracy, at the level of voter perceptions and their political culture, is progressing painfully slowly, if at all. There is little evidence in this survey of the growth of attitudes that can underpin real multiparty competition in the country.


Table 3: African opposition party supporters who believe a super-dominant ANC will be good for South Africa (%, Oct 97) 
 PAC supporters 41
 NP supporters 43
 UDM supporters 45
 Non-voters 32
 Uncertain voters 29
 Coloured voters 18
Indian voters 19
 White voters 5


Multiparty competition will obviously continue, but as before it will be based very substantially on ethnic and racial loyalty rather than on a free-ranging popular debate on political principles, objectives and policies in the public sphere. This is despite the fact that our earlier results in 1996, and other more recent findings, showed that majorities of people of all races and classes have a great deal to debate, because levels of discontent with quality of government performance are high and rising.

We have been criticised before for drawing this conclusion. One critic has argued that the fact that voters in different race groups choose different parties does not mean that their motivations are based on racial considerations. His submission was that the interests of different groups differ objectively and that their choices are determined by their interests and by the information that they are exposed to. This explanation has to be considered, but in the 1997 survey we have demonstrated that the voters themselves consciously express loyalties based on considerations other than interest-based politics.

Our results so far suggest that political choice in South Africa is not progressing in a way that will strengthen a floating vote — an essential ingredient of accountability and good government in a democracy. If anything, our most recent results suggest that the trend is in the opposite direction. Even if ANC support drops in 1999 to about 60 per cent because of dissatisfaction over socio-economic reform, the opposition parties will struggle to gain ground. All this points to a kind of stasis in our politics which, if not a type of one-party dominance, comes fairly close to it. We hope that these conclusions are wrong and will be monitoring voter attitudes in the run up to the 1999 national elections with keen anticipation.