The ANC's last chance

RW Johnson on ten key takeaways from Markdata's pre-election polling

Everyone who carries out an election survey is careful to say that the results are not a prediction, just a still photo of a moving picture. Despite that polls are invariably judged on how close their overall figures were to the final result. In the case of the last election Carol Paton summed up the conventional wisdom in Business Day (14 May): “In the end none of the pollsters got it right, and some did worse than others.”


She compared the polls done by David Everatt for the Gauteng ANC, those done by the Institute of Race Relations and by IPSOS. Her conclusion was that IPSOS was the best, with its results of ANC 61%, DA 19% and EFF 11%. Indeed, her column sounded very much like an endorsement of IPSOS: “IPSOS has by far the most resources in the market to conduct research. It has a larger sample than other pollsters (2,834) randomly drawn from all eligible voters...It conducts face-to-face interviews with people in their home language.”

I read Ms Paton’s article with something approaching stupefaction for I had conducted a survey for ENCA using the services of Markdata. Markdata’s methods were exactly the same as IPSOS but its sample was bigger (3,500) and its results were a lot closer to the final result. (The fact that Markdata was No.1 and IPSOS No.2 confirm that their methods are superior: cellphone interviews and using interviewers with a different first language than the respondent are not serious options in South Africa.)



Actual result













In fact, my experience from working with pollsters in Britain and France was that many of the survey firms regarded election work as a very mixed blessing. Normally they would have to come up with results such as 70% of housewives preferred soap powder X, and it really made no difference if the right figure was 68% or 72%: the client got the essential information right. But in an election if the firm came up with party A winning 51-49 against party B and the result was 51-49 the other way (i.e. results well within the margin of error), they would suffer denunciation as dunderheads and receive enormous bad publicity.

In fact the real value of a well done survey lies less in this tired old game of “how well did they predict the final result”, when they’re not actually making predictions, than in the deeper realities they sometimes reveal. In the case of 2019 I would select a number of themes:

1. The ANC vote is a lot less solid than before.

Only 33.5% of black voters said they’d vote ANC irrespective of who the leader was. One suspects that figure would have been 80%-90% in 1994. 27.3% of black voters said they would never vote ANC. That is much more than the black opposition parties can muster so there must be a number of terminally disaffected in that figure.

19.4% of black voters said they liked Ramaphosa enough to vote for the list irrespective of who else was on the list. This bears out Fikile Mbalula’s suggestion that without Ramaphosa the ANC would have been at 40%. It is likely that the ANC was looking at polls which showed that.

19.9% of black voters said they might have voted ANC but wouldn’t because there were too many corrupt people on their list. This suggests that such voters have become much more discriminating. Traditional ANC loyalties count for less. It also shows what a huge handicap the Zuma-Magashule faction is to the list.

2. Ramaphosa’s popularity is large but fragile.

Among black voters 50.7% thought he would carry out reforms and get rid of the corrupt elements (this was largely a matter of party loyalty: 67.1% of ANC voters thought this but very few among the non-ANC). 25.8% of black voters thought he was too weak to carry out such reforms and 23.6% thought he was just another politician making empty promises (that figure rose to 41.5% in KwaZulu-Natal, a hive of anti-Ramaphosa feeling). This suggests that Ramaphosa had better hurry with reform. Delay could see his popularity implode.

3. The ANC is losing credibility.

ANC propaganda against the DA has rested principally on the claim that if the DA gained power it would restore apartheid and abolish social grants. This was now believed by only 38.8% of black voters but disbelieved by 43.3%. (An extra 18% said they wouldn’t vote DA anyway, so this claim made no difference to them.) When we broke this down by party we found that 53.3% of EFF supporters disbelieved this claim, as did 89.2% in the Western Cape, 83.9% in the Northern Cape and 63.4% in the Eastern Cape. This means these three Cape provinces are potentially the most open to DA appeals if the DA can get its act together. Black men were particularly unlikely to believe this ANC propaganda.

4. The young are more disaffected.

Among the 18-24s only 28.2% believed the ANC propaganda about the DA, while 56.3% didn’t. And only 42% of this age group wanted to vote ANC. 34.6% of this age group said that their opinion of the EFF had improved during the election campaign, compared to 25.9% saying that of the DA. But the proportion of this age group saying they would vote ANC irrespective of who its leader was, was only 19.5% and two-thirds either thought Ramaphosa too weak to carry out reforms or that he was anyway just another promise-breaking politician. Large majorities of this age group thought the railways ought to be privatized and that towns and cities should be allowed to forsake Eskom and get their power from independent producers.

5. Voters have lost faith in Eskom.

Only 39.9% of all voters thought the government would manage to get Eskom back to what it once was. 30.2% thought the government might try to do this but would fail and 29.9% were for privatization right away, with EFF voters particularly in favour of this option. The largest group in all three major parties favoured the privatization of all SOEs, with EFF voters the most in favour. This is a resounding vote against the National Democratic Revolution.

6. Trust in most institutions is very low.

If we take those who said they had a lot or great deal of trust in an institution and then subtract those who said they had little or no confidence, the Constitution scored +10, judges +4, the Presidency +2 and the media +1. All the other institutions queried (chiefs, parliament, trade unions, the cabinet) had minus scores and easily bottom were the police with -22. The fact that even the most trusted institutions scored only +10 or less suggested that faith in the entire political system is now very low.

7. Unemployment, the overwhelming issue

Unemployment dwarfed all other issues and was the principal driver not just of the EFF (over 37% of whose voters were unemployed) but also of the fact that EFF voters generally and in large numbers favoured privatization and pro-business policies, flatly against the policies of their leaders. The hunger for jobs is far too great for such voters to be held back by the constraints of party discipline. 57% of all voters said that the ANC “doesn’t really care about the unemployed” and even among ANC voters 55% said that “the ANC may care but its policies just don’t work”.

When asked what the reasons for unemployment were 72% both of ANC and EFF voters said “there are too many foreigners in the country”, an opinion shared by 63% of DA voters. This is the nearest any issue comes to unanimity – although a majority of all groups were in favour of the immigration of skilled people. This issue has explosive potential, as we already know. It is staggering that the government has signed a free trade deal allowing the free movement of foreigners into South Africa. This is a bomb waiting to go off.

8. The dislike of chiefly rule.

The fact that many chiefs, especially in KwaZulu-Natal, have begun to charge rent for the land they distribute, thus turning themselves into a class of wealthy rentiers, caused Kgalema Motlanthe to criticise them as “tinpot dictators”. This produced chiefly unrest in KwaZulu-Natal and Ramaphosa hastily ran to King Zwelethini to withdraw Motlanthe’s report. Since then the ANC has renewed its support for further chiefly powers and rights, doubtless motivated by the need for the chiefs’ electoral support. In fact when we tested oponion on this issue we found that less than a fifth of voters supported the chiefs. 50% of ANC voters and 61.2% of EFF voters said that whoever lived on the land should own it – this despite the fact that neither party leadership favours the growth of an independent peasant-farmer class that this would produce.

9. The government, not apartheid, is blamed for inequality and poverty. (This was a multiple choice question so answers add up to over 100%.)

When asked who was responsible for the increase in poverty and inequality, 55% of ANC voters blamed the government and 29% the ANC. 16% blamed “rich whites” and 1% or less blamed “the whites in general, apartheid”. Given all the frantic attempts to blame “white privilege” and “300 years of white domination” for our present ills, this represents a large propaganda failure.

10. Disillusionment with politics.

Among all voters 27% said they had never been interested in politics and another 18% said they used to be interested but are much less so now, while 3% said Don’t Know. Only 52% said they were interested in politics. Half of all EFF voters said they weren’t interested while among ANC voters 40% said they had never been interested or were much less so now.

We also tested this by asking about attitudes to young people who “can’t be bothered to vote”. The answer that this was shocking given the sacrifices made to gain the vote – which would probably have been the overwhelming response in the 1990s – was now given by only 20% of voters. 28% of voters said they actively sympathized with those who didn’t vote because they themselves didn’t trust any politicians, while the remaining 46% were tolerant, saying their own view was different but abstainers had a right to their own views. What was shocking, in other words, was that so few thought abstention was shocking.

The overall picture then is one in which the ANC bloc is disaggregating, losing its unity and its credibility. More and more clearly the mass of black voters disagrees with its political elite on a whole range of issues, dramatizing the fact that South Africa is ruled by a small activist minority which is quite unrepresentative.

This may be why support is growing for electoral reform. Only 19% of voters want to keep our present electoral system while 35% want MPs to be responsible to their constituencies and 40% favour direct presidential election. Change towards either of the two latter alternatives would make it much more difficult for today’s unrepresentative elite to continue to prevail. Among younger voters, direct presidential election has close to 50% support.

The situation is tenuous. There is no doubt that the stress on “white privilege” is in part an attempt to deflect attention from the fact that almost all indices (health, education, crime, unemployment etc.) have deteriorated badly under ANC rule. Yet it hardly makes sense to blame “white privilege” for our present woes if the advent of black power has actually made things worse. What is striking is that this narrative is not working, and nor is the demand for “radical economic transformation”, rejected by 68% to 28% even by ANC voters, and by 76% to 22% among EFF voters.

In effect black voters are simply losing faith in governmental action, or talk about governmental action, to remedy their woes. It is simply no good talking about “the national democratic revolution”, “radical economic transformation” or “the developmental state” if in reality people are getting steadily poorer every year. You can hardly expect them not to notice that these slogans are contradicted by their ever-growing poverty.

Similarly, large majorities of black voters were quite happy to jettison affirmative action, BEE and expropriation without compensation if the result would be to bring more investment and thus jobs. This straightforward disavowal of so many African nationalist articles of faith suggests that no amount of ideology is proof against twenty-five years of rising unemployment and five years of steadily falling real incomes. Black South Africans are getting poorer and poorer every year and if the ANC maintains its present priorities, these processes are set to continue.

It would be folly to imagine that this can simply go on, unchecked, without a major popular reaction. It is already a striking situation where less than a fifth of the electorate are happy with the present electoral system and where large majorities believe the government simply “doesn’t care” about all the issues which are important at grass roots.

It was very striking that a number of ANC elders, including former President Kgalema Motlanthe, expressed the view that this election was “the ANC’s last chance”. Given the slow and cautious reactions of the Ramaphosa administration to date and the general preference for business-as-usual in government, symbolised, despite all the promises, by an executive which is almost as overblown as Zuma’s – it seems unlikely that Ramaphosa is seized with the sort of urgency that one would imagine a “last chance” would generate.

If this is indeed the case one should expect a rapid deterioration in Ramaphosa’s popularity, an end to Ramaphoria and, in that darkening mood, probably more xenophobic riots. This would doubtless complete the country’s transition to recession, junk status and end the Ramaphosa project before it had ever really got started. We must all hope for better but this is what a reading of the tea leaves suggests.

RW Johnson


If you enjoyed this article, and appreciate what Politicsweb does, please consider becoming a supporter here