Local govt elections: What do the polls really show?

RW Johnson analyses the Ipsos/eNCA poll predictions, and the implications for the ANC, DA and EFF

Thinking about the Local Elections

Currently the only publicly available data on the August 3 municipal elections is the series of Ipsos (formerly Markinor) polls being published each week by eNCA. These are very amateurishly reported. Figures are given for ANC, DA and EFF plus Don't Knows. When you add these up you realise that Other Parties have, for no reason, simply been left out.

Yet these - UDM, Cope, IFP, FF+, Minority Front, NFP, United Front, PAC, Azapo, ACDP, Muslim League and Independents - play a crucial supporting role. It should not be forgotten that the first DA administration in Cape Town relied on a DA with only 42% - with the rest constituted from these smaller parties.

So let us laboriously reconstruct the Ipsos polls, allowing for these Others. Then, omit the Don't Knows altogether as almost certain non-voters. Then recalculate the results out of 100% and this gives you


ANC          41.82
DA            36.11
EFF          11.39
Others.      10.68


ANC       31.15
DA         46.66
EFF         14.3
Others     7.89


ANC        25.27
DA         49.86
EFF        11.47
Others     13.4

Apart from all the usual caveats about opinion polls, it should be noted that these results are blithely reported without any indication of whether or how Ipsos has sought establish if the respondents are really going to vote. On the face of it all the above results are reported as if 85-90% of the electorate will vote when we know that 60% would be a respectable turnout in municipal elections. Really what is necessary is for pollsters to press electors very hard to find out whether they are absolutely certain to vote, very likely to vote, quite likely, probably will vote, probably won't vote and so on.

Having done that one should eliminate all except those who say they are absolutely certain to vote - experience shows that virtually all the others won't vote when it comes right down to it. Indeed, the problem is that when asked about their voting intention respondents are perfectly aware that the "right" answer, showing that they are good citizens, is to say they will vote - and South Africans seem very prone to saying this even when they actually have very little inclination to vote.

Moreover, for someone to insist that they won't vote is to mark themselves out as a fed up dissident, something that many people are shy of advertising. This caveat - that far fewer people will actually vote than suggested by the figures above - must be borne in mind in all that follows.

As may be seen, the figures everywhere show a considerable collapse in the ANC vote. Indeed, in Tshwane and Nelson Mandela (i.e. Port Elizabeth) the EFF is almost half the size of the ANC and in Joburg just under a third of the ANC's size. This contrasts sharply with the national picture at the 2014 parliamentary elections when the EFF were only one tenth the size of the ANC vote.

However, the EFF picture is uneven. In 2014 the party scored just over 10% of the vote in Gauteng, so it has not actually advanced very much from that level in Joburg - though it has in Tshwane, perhaps as a result of the bitter ANC faction fighting there. It looks, in other words, as if some disgruntled ANC voters have peeled off towards the EFF in response to their faction losing out. On the other hand, the EFF is doing well to pull over 11% of the vote in PE - where it did far less well in 2014.

There is some suggestion here that the EFF is now leveling up to a presence of 10% or a little more in all the country's big city environments. It should be pointed out that this is not as high a level as has often been predicted. Earlier in the year a DA poll suggested that the EFF had just about doubled its support since 2014, but if that were the case one would expect to find the party now at 15-20% levels in Gauteng - which is not the case. Indeed it looks rather more as if EFF support has hit a glass ceiling. It must be remembered that the EFF polls dramatically better in urban environments and far less well anywhere in the countryside.

The DA surge is impressive though nowhere has it hit the 50% mark. Its gains appear to have come straight from the ANC. The situation is tightest in Joburg where the ANC vote is holding up far better, but it's collapse is vertiginous in both the other two cities. The ANC has pinned great hopes on the possibility that Danie Jordaan will be able to bring some soccer magic to stop the rot in PE but, on these figures, he has signally failed to do so while Thoko Didiza's triumph in Tshwane's factional struggles now looks like an entirely Pyrrhic victory.

There has been a great deal of speculation about the likelihood of DA-EFF coalitions but it must be noted that in both Tshwane and PE the DA looks as if it will only need a deal with some of the minor parties and will not need to embrace the EFF. The real poser is Joburg where there are a host of different coalition possibilities. For the DA to capture the mayoralty it would need a deal with both the EFF and the other minor parties there, but equally, the ANC could cling on to power just by offering a deal to the minor parties and not have to deal with the EFF. This would, at this point, seem the most likely result there.

Why should the EFF or the minor parties prefer a deal with the DA rather than the ANC ? This is a point too often left unexplained. What has to be understood is that wherever the ANC has won power it follows, as night follows day, that it packs the administration of that town or city with its own deployees - in any case it has to satisfy its supporters' hunger for these jobs. The result is that if, for example, the DA in Joburg attempted a Grand Coalition with the ANC, it would quickly find that it had no real power because the entire municipal civil service would continue to take its orders from the ANC and there would be no effective power-sharing at all.

But if, for example, the minor parties gang up with the DA to take over a town, not only can they hope to achieve real power but they too will get to reward some of their supporters with jobs in the town's administration as ANC cadres there are winkled out. The only exception to this is where the ANC is dominant and can achieve a majority with the help of a handful of minor parties.

Typically what happens in such a situation is that ANC cadres remain in place in the administration but minor party politicians are rewarded with patronage, jobs or a few tenders. These boil down to rewards for a few individuals without any policy change. One may take it as given, however, that neither the DA nor the EFF would be content with such crumbs from the table.

There is a final twist, well understood by both the big parties. When the DA first took over Cape Town in 2006 it did so thanks to a coalition with a host of minor parties. This enabled it to begin the cleaning of the Augean Stables and many leading ANC cadres lost their jobs in the city's administration as they were "performance managed" out after it had become clear that they were ignoring orders from the new council majority.

Naturally, each such departure was accompanied by furious accusations of DA racism but these were simply brushed aside, for everyone knew these were the obligatory explosions which could be expected when both political influence and well-paid jobs were on the line. The cumulative effect of this over five years was to greatly further weaken the ANC in the city and, indeed, it was not long afterwards that the Cape Town ANC went bankrupt. And at the subsequent election the DA won an overall majority without any further need for a coalition with the minor parties.

That is to say, when one of the major parties cobbles together such a coalition deal it hopes to use this foot in the door to obtain complete control a few years down the line. This is one more reason why the coming local elections on August 3 are of such importance to the country.

The polls above suggest that Joburg will be the marginal case. This in turn implies that Ekurhuleni will be safe for the ANC, let alone Durban. Of the other major urban centres the ANC should hang on without difficulty in Mangaung and Pietermartizburg but it has to be worried about East London if the collapse of its support in PE is mirrored across the Eastern Cape.

The decline in the ANC vote should see the DA hang on in Midvaal but otherwise the key point to watch will be whether the DA's gradual progression northwards from Cape Town continues. The first question is whether some of the remaining ANC-ruled towns in the Western Cape will go, with attention fixed particularly on the fate of Truman Prince in Beaufort West - now the only town in the Western Cape with an overall ANC majority. But the larger question is whether the DA surge will now begin to spread into the Northern Cape.

Thus far this has seemed a far tougher nut to crack. A great deal will depend on whether the EFF is able to poll as well in these smaller towns as in the big cities. If it were to achieve that - and its votes come overwhelmingly from the ANC - the result could be to overthrow ANC majorities or pluralities in many towns, though of course the EFF is not in sight of winning even one municipality and will probably not even enter coalitions in more than a handful of cases.

It goes without saying that none of these trends is likely to be mirrored in the countryside where ANC control will be guaranteed by the chiefs in classic Bantustan fashion. The present indications are that we are sliding gradually towards the situation that Zimbabwe found itself after 20 years of ZANU-PF control, where the Opposition had gained control of the big cities (Harare and Bulawayo), where power was eagerly contested in many smaller towns, but where ZANU-PF ruled undisturbed in most rural areas.

The big difference, of course, is that South Africa is a far more urban country. Under the Afrikaner Nationalists the Opposition remained strongest in the big cities and, in effect, the platteland ruled the country but urbanization has moved on apace since then, making this a harder and harder trick to achieve.

RW Johnson