On the local government elections

When co-operative Governance Minister Sicelo Shiceka reportedly endorsed a Gauteng African National Congress (ANC) resolution proposing that all three spheres of government hold elections at the same time, he was criticised by many observers who saw this move as a cynical ploy by the ANC to try and prevent electoral losses at the local government level. Due to the disastrous performance of many ANC-controlled municipalities over the past 5 years (more potholes, disintegrating sewerage works, tardy or absent municipal services, lack of housing delivery), it is widely suspected that the ANC will lose considerable support when local government elections are held next year.

If all the elections were held at the same time, so the argument goes, the ANC would not be punished for its poor performance at local government level because the vast majority of voters would focus on national issues (apartheid=bad, transformation=good) and would vote for the ANC, the party with whom they closely identify – regardless of whether their local councillor had been a lazy and corrupt skelm or a wonderful community activist who really made a difference to the community he or she served.

It is said that in the absence of a national election campaign, the ANC would suffer huge losses because even if its disaffected voters could not bring themselves to vote for the DA or another opposition party, the traditional ANC voters would stay at home in disproportionately large numbers. This would boost the electoral fortunes of the DA, whose energised and angry voters (eager to fight b(l)ack(s)”) would stream to the polls to punish the ANC for the failure of ANC controlled local councils (and for the cheek of being black and in charge). More people might not vote for the DA but less would vote for the ANC, giving the DA a much larger percentage of the votes.

Moreover, as Steven Friedman argued at the time, it is feared that holding local government elections at the same time as those for national and provincial government could reduce these elections to a hollow ritual. Because national elections attract far more voter attention than their local equivalents and because voters in South Africa (still) have strong loyalty to national parties (most notably the ANC), local elections held at the same time as the national ballot would do little more than duplicate national results. Local issues will not be discussed and the local link between councillors and the electorate will be further weakened.

In theory, these arguments make a lot of sense. The only problem is that the election campaign and results of the 2006 local government elections suggest that in practice it might make little difference to the outcome of the local government election whether it was held at the same time as national elections or at a different time. It would also not make much difference to what issues get discussed during the campaign.

It is true that during the 2000 local government elections far more traditionally DA voters than ANC voters went to the polls and the DA obtained 22% of the vote versus the 59% for the ANC. The previous year the ANC had won 66% of the vote while the forerunner of the DA only received 9.5% of the vote in the national election.

The DA did not receive substantially more votes in the local government poll in 2000 than in the national election the previous year, but there was a dramatic drop in the amount of voters who voted for the ANC, suggesting that traditional ANC voters disproportionately stayed away from the polls. (This happened after the DA and the New National Party had their shotgun marriage, which must at least partly account for this huge jump in support for the DA in the 2000 local government election.)

But in 2006, this effect was not repeated. In that local government election the ANC won 64% of the vote and the DA 16%. Three years later at elections for national and provincial legislatures the ANC had won 66% and the DA 16.5% of the vote respectively. The overall results for the local government poll in 2006 and the national poll in 2009 was therefore almost identical. Both traditional ANC and traditional DA voters had stayed at home in droves. (The slight increase in ANC support in 2009 could possibly be attributed at least partly to the Zuma effect, which meant that the ANC had increased the percentage of its support amongst voters in KwaZulu-Natal significantly.)

There is therefore little evidence to suggest, as some pundits have done, that during local government elections traditional ANC voters stay away from the polls in disproportionate numbers to punish the ANC governed municipalities for poor or non-existent service delivery. Far fewer voters voted in the last local government elections (only about 50%) than in the last general election (77%) but the traditional voters of the ANC and the DA stayed away in equal proportions.

I suspect there are at least two reasons for this. First, even during local government elections, candidates representing political parties do not fight localised campaigns about the particular issues of concern to the voters of a specific ward. Election campaigns (both for national and for local government elections) are tightly controlled by centralised campaign managers. Each party usually  has a central message that is sold to the voters with very little adjustment to the local circumstances of a particular ward or city council.

Second, because of the extremely close identification of voters with their party of choice (whether it is the ANC or the DA), voters still vote for the party and not for the candidate or the issues during local government elections. The ANC or the DA could nominate a wife-beating, child-molesting, tenderpreneur and he or she will probably get more or less the same amount of votes as any other candidate which the party might have chosen to represent it. This means parties have very little incentive to put up credible community based candidates and so the possibility for local charlatans and money hungry incompetents to be nominated as ward councillors by a party are rather high.

(During by-elections, this effect is less severe as a by-election is really fought more on local than national issues and in the absence of a national election campaign with a national message and strategy the quality of the candidate can therefore make a marginal difference during by-elections. This is why the ANC probably won the Bredasdorp by-election earlier this year – bucking the trend of dramatic increased support for the DA in the Western Cape – because its candidate was a fisherman whom Gwede Mantashe famously said he could not understand but who could be understood by local voters.)

What can one learn from this?

I suspect the ANC and the DA will take different lessons from the tentative insights provided above. For the ANC, it would probably be a good idea next year to fight a national campaign with a strong central message (“give us more time”; “the DA is racist”) – just as if it was fighting a national general election. In the Western Cape, where the ANC is a minority party, it would probably help the ANC to fight very localised campaigns addressing very particular issues that resonate with potential voters (by, say, highlighting the open toilet debacle in Cape Town.)

I suspect the DA would be better off fighting very localised campaigns in which its candidates in each municipality identify the hot button issues for its potential voters (especially issues that would galvanise voters to go to the polls). The DA may want to stay away from campaigning on a national platform (stop Zuma! stop Malema!) and may want its candidates to address local concerns (the water is not being purified; the ANC councillors are stealing our money; the sewerage is not being cleaned).

The problem is that localised campaigns are more expensive and more difficult to run. And with the DA there is always the danger of “rogue” candidates displaying racist tendencies (as Julius Malema might say) or making gaffes if they are not strictly supervised by Helen Zille and her team. This the DA cannot afford because the ANC will exploit such gaffes to rustle up votes and motivate traditional ANC voters to go to the polls to vote against the “racist-apartheid-loving” DA.

It might be that things have changed since 1996. If the percentage of votes cast for the DA shows a significant rise, it might give an indication that ANC voters have finally begun to lose faith in their party, either by staying at home in larger numbers than DA voters or by switching votes. But if the DA receives less than 20% of the vote nationally, it would mean nothing much has changed in the electoral landscape.

In any event, on the available evidence it is far from clear that synchronising the elections for all three spheres of government will assist the ANC and be detrimental to the DA.

BACK TO TOP