Survey - State of parties and health of democracy

Lawrence Schlemmer analyses the results of the most recent political opinion poll.

At the moment there is, or should be, great uncertainty about the relative strength of our different political parties. Since 1999 the ANC has managed to surround itself in controversy. Stances and policy on HIV/Aids, the hugely expensive arms deal and the muscular manipulation of the parliamentary enquiry into associated corruption, its highly ambiguous reactions to land seizures and the dubious elections in Zimbabwe and the near alienation of former President Mandela have all dominated the media. Critically high unemployment and associated poverty have fuelled ongoing disputes within its alliance and supporters in civil society. These and other problems, coupled with mid-term political disinterest would all be expected to have severely tested the support levels of our governing party.

Opposition parties have also been in the wars. After a high profile political marriage between the DP and the NNP that injected new hope into opposition politics, the even higher key divorce, and the NNP’s controversial coalition with the ANC, caused dismay in opposition ranks, to say the least. Resurgent tensions between the IFP and the Zulu monarch and ominous signs of renewed violence in KwaZulu-Natal, have struck at the morale of IFP supporters.

If our multi-party democracy is working as democracies should, there should be a large injury toll among the teams on the political playing field and many fans would be heading for the exits. Does our democracy work in this way? Evidence from a recent MarkData survey commissioned by the Helen Suzman Foundation offers answers.

This representative nation-wide survey of political perceptions conducted during March/April of this year allows comparisons with the results of earlier regular MarkData Omnibus surveys of identical design which have very closely predicted election outcomes in 1994 and 1999. Based on stratified probability samples of roughly 2 200 voting age adults of all races, they cover all areas from deep rural and traditional areas to cities and towns, including shack areas, migrant worker hostels and backyard accommodation. Face-to-face interviews are conducted by experienced fieldworkers in the home languages of respondents.

Trends in support for political parties: 1999-2002

The following results are based on questions on the party that respondents would support in a hypothetical forthcoming general election. The surveys compared were conducted prior to the 1999 general election, in July 2001 after the forming of the Democratic Alliance, and in March, 2002, after the dissolution of that alliance and the decision of the New National Party to co-operate with the ruling ANC.

The results in Table 1 are the responses of all adults, including those who would not vote or who do not make a choice of party. Because of the large number of uncertain voters at this point it would be statistically hazardous to predict an election outcome in the conventional way by recalculating party strength with the uncertain responses excluded. Although the results therefore do not estimate an election outcome they are a valid indication of relative party support among committed voters.

The 1999 survey was conducted a full three months before the election and some party support shifted in the intervening period. But the 1999 survey results, with uncertain and non-voters excluded, were nonetheless very closely in line with the election outcome.

What one notes is indeed a fall-off in support for the ANC among the adult public at large, but it is surprisingly slight and hardly reflects all the headline-making controversy. Also significant is the rise and fall of the “Democratic Alliance”. The NNP in its new company has attracted slightly over half of its 1999 support back. Among both coloured people and Indians the ANC has gained support as a result of the new divisions in opposition politics.

The DA/DP combination is now back to the level of the DP in 1999 — a setback, but it is the only one of the major parties to maintain support levels between 1999 and 2002. The IFP, after growing in 2001, may be marginally weaker in terms of openly acknowledged support. With the IFP, however, our polls have always found that the less-confident rural traditional voters in a tense province are reticent about disclosing their sympathies and are particularly prone to mid-term political disinterest.

The most noticeable “gain” has been a growth of uncertainty and disillusionment. The proportion of adults who are undecided, apathetic or disinterested in voting has grown by over 20% since 2001 and the split in the opposition alliance. Among whites, the largest minority, this disinterest and indecision has reached nearly 40%.

The political withdrawal by minorities in particular is not necessarily permanent — much of this lost voter interest will be re-stimulated as elections draw near. When this happens, how will the alienated voters split? After inspecting the pattern of responses across all survey questions of the people who make no choice, it is quite clear that if they are attracted back into the active electorate, a majority will tend to favour strong opposition. For example, the undecided and non-voting margin as a whole (all races) is 45% less positive than average about the performance of the President, 44% less likely than average to have voted ANC in 1999, 30% less positive than average about the co-operation between the NNP and the ANC and it is also more likely than average to favour a critical press.

At the same time it is very possible that if there are high levels of dissatisfaction among government supporters, their voting turnout will be lower than in 1999. Levels of dissatisfaction are quite critically high among ANC supporters at present. In the survey three questions were asked that were relevant to political performance, the results of which are noted in Table 2.

These results hardly point to a triumph of achievement for the ANC — broadly only half of its support is satisfied, one way or another.

It is impossible to quantify these indications of the prospects for party gains or losses because one would have to know the eventual voter turnout to do so. Suffice it to say that the ANC support estimated in the recent survey is probably a maximum, and the official opposition support a minimum that is very likely to be exceeded. This points to the ANC majority in a future election dropping towards 60% and the share of the official opposition, unless the damage from the Harksen saga is large, rising above 15%. After near eclipse, the NNP will be bolstered by its coalition in the Western Cape and stabilise, vying with the IFP for the third spot.

The quality of our democracy

These will be significant shifts, particularly for provincial politics, but in the larger scheme of things they would hardly be dramatic. We have an ossified floating vote — the life blood of a viable and accountable democracy. At the core of this weakness is a tradition of “solidarity” voting and solidarity politics — the kind of politics that the old National party exploited during apartheid, which has returned in full measure. The kind of politics that the Premier of Limpopo extolled when he accused Dr. Malegepuru Makgoba of “betraying his race” because he had been critical of the President’s views on HIV/Aids.

The survey offers some compelling evidence of ossified democratic choice. In addition to the weak correlation between voter satisfaction and majority party support implicit in the results given above, more direct evidence is seen in the levels of agreement or disagreement with the statement recorded in Table 3.

This denial of own views and interests in favour of “team solidarity” is difficult for people to admit to and therefore the real levels may be even higher than the figures suggest. Given our history, this mindset is understandable but it hardly contributes to democratic accountability.

Solidarity voting often makes it possible for governments and their supporters to occupy different worlds. Consider the lack of convergence that the following survey results reveal. No fewer than 68% of ANC supporters disagree with the seizure of white farms by political militants in Zimbabwe, while the government equivocated for weeks. Some 62% of ANC supporters oppose the arms deal and among them many think that it should be cancelled. Around 64% either did not understand the government’s viewpoints on HIV/Aids or disagreed with them.

When voters are asked who benefits most from government policies, the dominantly poor ANC supporters offered the breakdown documented in Table 4.

Summing up

Interpreting the various responses in the survey as they inter-relate, it seems as if the governing party at this stage can count on up to 60% retention of its support even when its voters are dissatisfied or feel neglected. Among African voters the retention of support would be higher at around 75%. In other words the ANC is so cushioned by its solidarity vote that it really need not worry intensely about satisfying the majority of its supporters. It does in fact worry about this a great deal but it is never in a power-threatening situation. So much for the assumption that multi-party democracy always keeps governments on their toes. Clearly political culture counts as much as constitutional structures.

The ANC’s majority is not at issue. It would be surprising indeed if South Africa’s leading liberation party did not enjoy clear majority support in the population. Whether it, or any conceivable political party, however, deserves the current levels of more or less automatic majority party support is an issue that strikes at the heart of effective democracy. One might expect this lack of political discernment in a country with a homogeneous mass of poor and dependent people, but in a fairly well-developed and culturally and socio-economically diverse society like South Africa, such comprehensive and persistent majority dominance is an anomaly.

This conclusion is not to suggest that the ANC should try to earn its super-majority by pandering to all grass-roots sentiment. This is usually a short term and self defeating strategy, particularly in the modern global economy. Effective and accountable governance requires encouraging a synergy that allows both more advantaged and less advantaged people to accommodate and support each other’s interests, to the greater good of everyone. So far this kind of synergy has not yet emerged, and one of the main reasons is that the government has such a comfortable majority that it does not have to work hard to maximise its support across all interest groups. If it had to do so, it would have to make a much greater effort to convince diverse interest groups to respect and accommodate each other’s interests — the hallmark of the stable pluralist democracies of, for example, most of the OECD countries.

One of the few sure measures available to democracy in persuading majority parties to make that effort is effective opposition, the legitimacy of which should be as sacrosanct as popular majorities.