Opinion poll: Coalitions and floor-crossing: How will the voters respond?

Lawrence Schlemmer finds that the political disarray caused by the floor-crossing spree has left many voters disillusioned.

The official opposition, the Democratic Alliance (DA), may well end up strengthened relative to its pre-alliance support as the Democratic Party (DP) in 1999. But the loss of much of its New National Party (NNP) wing to a coalition with the African National Congress (ANC), before and after the floor crossings by well over half the DA local councillors to rejoin the NNP, has been a harsh setback for the DA. The most serious effect for the official opposition is a significant loss of momentum in its steady growth over the past five years or more.

The events have wider implications, however. They compromise the democratic principle of primary accountability to voters and they have reinforced the image and the political structure of South Africa as a one-party dominant state. These effects are injuries to the collective ambitions of South Africans to live in and consolidate a world-class democracy.

It is not that coalitions are intrinsically undesirable. If a party’s supporters endorse the arrangement, if the parties involved can defend their constituency’s interests within the co-operation and if there is genuine give-and-take on policy issues, a coalition can broaden the policy base of governance and help to secure the vital interests of minorities. The problem with the ANC-NNP coalition is that the likelihood of these conditions being met seems rather remote. Not only did the ANC allow its earlier and crucially important coalition with the National Party (NP) under FW de Klerk to deteriorate into hostility, as has its coalition with the Inkatha Freedom Party (IFP) in KwaZulu-Natal, but it is clearly imposing its policies and priorities on its long-standing alliance partners to its left — the South African Communist Party (SACP) and the Congress of South African Trade Unions (Cosatu).

Is there any real basis for a balance of compromise in its new coalition with the NNP? There is more than a whiff of opportunism and greed for power in the present arrangement, from both sides.

Similarly, floor crossing is also not necessarily always undesirable. The main argument in favour of floor crossing is what the ANC and the NNP have termed “freedom of conscience”. They also argue that this fluidity in politics allows representatives to respond to changing conditions and opportunities in political life. The United Democratic Movement (UDM), the IFP and the DA, who have all at one time or another supported the principle, argue however that the representatives in a proportional list system are not free to cross the floor to other parties without proper consultation with the voters at large, because they are elected not as individuals but simply as units on a party list. With manifest justification they have condemned the latest spate of floor crossings as a cavalier disregard of voter choice.

Floor crossing has merits in a constituency system in which representatives are elected as individuals, but it is deeply and blatantly non-democratic in a list system. Many observers were surprised that the Constitutional Court overlooked this basic democratic principle of a popular mandate in its ruling, so much so that Steven Friedman, of the Centre for Policy Studies, in an SAFM interview, expressed concern about party-political bias among the Constitutional Court judges.

In responding to the floor crossing and the DA’s consequent loss of most of the local authorities that it has controlled since 2000, DA leader Tony Leon has predicted that opposition voters will see the ANC-NNP coalition for what it is and punish the NNP accordingly in the next elections in 2004. Furthermore, even after the recent defections, the DA can rebuild opposition with a broader team of party representatives and workers than it had in 1999. How realistic is his optimism?

In previous analyses in Focus, evidence from a March 2002 MarkData Omnibus poll (see State of parties and health of democracy, Focus 26, 2nd quarter 2002) suggested that the majority of voters who do not support the ANC want strong and principled opposition. This does not mean that they oppose coalitions (see the March poll and the June 1999 poll: How to use that huge majority, Focus 16, November 1999). But co-operation between parties is endorsed mainly in order to promote development or to provide political minorities with leverage, certainly not to augment the already dominant position of the ANC.

A more direct indication of likely voter reaction lies in current trends in support for political parties. Since the last Focus poll of party support in March 2002 results from a July/ August poll have become available. This preceded the actual floor-crossings, but by July/August the voters would have had time to contemplate the shape of the co-operation between the ANC and the NNP.

What are the indications?

In Table One we present the trend in party support since just before the June 1999 elections up to the most recent July/August poll this year1.

We note that DA support increased between 1999 and 2001 as it absorbed the NNP. It then fell back to its 1999 levels when the NNP leadership and some representatives defected to the ANC-NNP coalition. More recently it appears to have recovered and may still be rising. The NNP support, which had been all but eclipsed while the party was in the DA alliance, has certainly strengthened but seems to have peaked at around half the support that the DA has. On the face of it this would suggest that the DA has retained around half of the support that it gained while in alliance with the NNP.

But something else has also happened. The gyrations of the politicians have left very many voters confused or disillusioned. People not making a choice of party or indicating that they will not vote in a future election, have increased and their numbers may have stabilised at a damagingly high level. These uncertain or disillusioned voters, as indicated in Table Two, are particularly numerous where they are needed most — in KwaZulu-Natal, Western Cape and Gauteng — the provinces in which the ANC is not super-dominant.

This current exit of voters has a particularly distorting effect on the party standings in the Western Cape and KwaZulu-Natal. Some idea of what would happen if these voters stay away at election time is given for the two provinces in the August poll, in Tables Three a and b. In Tables Three a and b the respondents making no choice are excluded from the percentages to simulate an election outcome if they do not vote. Estimates are presented in detail to reflect very small party support.

In KwaZulu-Natal the NNP has recovered some support, as in the national results, but the DA retains a very clear advantage over the NNP. However the ANC gains hugely as a result of the disillusionment among mainly opposition party voters, the IFP included. With the Minority Front of Amichand Rajbansi, it may not even need the NNP to help it to secure a majority.

The Western Cape is different. Here the ANC has not strengthened since 2001 but the NNP seems to have recovered the bulk of its support that had moved into the DA alliance. If the July-August pattern holds, the ANC-NNP could garner nearly two-thirds of the vote in the province.

The reason for this strong recovery by the NNP in the Western Cape can be detected if one analyses the results according to population group. For reasons of brevity only the results for the ANC, the DA and the NNP are presented in Table Four.

In the abridged results in Table Three we note that while the NNP has lost ground since 1999 among all groups, it has remained fairly strong and has fallen least among so-called coloured people, mostly resident in the Cape. The DA has in fact gained among all groups except whites, whose support level has remained on a high plateau. The DA is larger than the NNP among all groups except among coloured people, which also helps to account for the NNP’s strength in the Western Cape.

Around August, therefore, indications were that many former NNP supporters are indeed likely to punish the party for its connivance with the ANC, and remain with the DA. But enough NNP supporters had bought the idea of being part of a governing coalition to suggest that the NNP’s strategy for survival can work. If the high levels of apathy and disillusionment among opposition supporters persist, the ANC (with the NNP where necessary) will control all provinces and the vast majority of local councils. From poll results in previous issues of Focus, however, it is clear that NNP supporters expect their party to be able to hold its own and defend their interests in the coalition. Whether the ANC will allow this to happen has yet to be seen. If so it will be a first for that party.

The challenge for the DA, on the other hand, is twofold. First it must make sure that it brings voters back from the cold of disillusionment, and, in its opposition it must test the quality of a coalition that was entered into by the ANC mainly to keep the DA out of any participation in government. The August results show that these are tough tasks. But if the DA can recover its lost momentum, they are achievable objectives.



The MarkData Omnibus polls are strictly comparable over time. They are based on personal interviews with 2,200 adults in stratified random samples covering all areas, both rural and urban, including all types of residential areas from traditional and commercial farming areas to houses, shacks, apartments and hostels in cities.