What follows is an attempt to sketch in general terms how many votes the ANC would need in Gauteng to secure 51% in the province on May 8, and what some of the considerable obstacles in the way of it achieving that are.
Here are the key background statistics from the last election:
- 2014 All Gauteng registered voters: 6,063,739
- 2014 Total Gauteng votes cast [provincial ballot]: 4,424,424
- 2014 Gauteng turnout [provincial ballot]: 72.965%
- 2014 ANC votes [provincial ballot]: 2,348,564 for 53.59%
As of March 19, the Electoral Commission of SA (IEC) had a total of 6,379,311 voters registered in Gauteng. That number will probably change a small bit, with late registrations, but not by much. Using it we can generate a number of turnout scenarios and see how many votes, for each scenario, the ANC would need to get to 51%.
The first thing one needs to do is work out what kind of turnout scenarios are likely. Here is the turnout percentage, on the Gauteng provincial ballot, for the past four provincial elections:
- 1999: 89.15%
- 2004: 74.23%
- 2009: 75.60%
- 2014: 72.97%
It would seem, then, that the general trend is a decline in turnout over time. So, a realistic range of turnout scenarios for 2019 would seem to be something like: 73%, 72%, 71%, 70%, 69%, 68% and 67%. Those may not be perfect, but they cover the likely terrain quite well.
Let’s use those seven scenarios and see what they do to the ANC’s numbers.
73%: High turnout
- A 73% turnout scenario would mean that approximately 4,656,897 valid votes would be cast in the province.
- In that scenario, approximately 2,375,017 votes would give the ANC 51%.
- For the ANC to reach that, it would need approximately 26,453 additional votes, compared to what it secured on the provincial ballot in 2014 (2,348,564 votes)
72%: Relatively high turnout
- A 72% turnout scenario would mean that approximately 4,593,103 valid votes would be cast in the province.
- In that scenario, approximately 2,342,482 votes would give the ANC 51%.
- For the ANC to reach that, it would need approximately 6,082 additional votes, compared to what it secured on the provincial ballot in 2014 (2,348,564 votes).
71%: Above-average turnout
- A 71% turnout scenario would mean that approximately 4,529,310 valid votes would be cast in the province.
- In that scenario, approximately 2,309,948 votes would give the ANC 51%.
- For the ANC to reach that, it could afford to lose approximately 38,716 votes, compared to what it secured on the provincial ballot in 2014 (2,348,564 votes).
70%: Average turnout
- A 70% turnout scenario would mean that approximately 4,465,518 valid votes would be cast in the province.
- In that scenario, approximately 2,277,415 votes would give the ANC 51%.
- For the ANC to reach that, it could afford to lose approximately 71,149 votes, compared to what it secured on the provincial ballot in 2014 (2,348,564 votes).
69%: Below-average turnout
- A 69% turnout scenario would mean that approximately 4,401,724 valid votes would be cast in the province.
- In that scenario, approximately 2,224,879 votes would give the ANC 51%.
- For the ANC to reach that, it could afford to lose approximately 123,685 votes, compared to what it secured on the provincial ballot in 2014 (2,348,564 votes).
68%: Relatively low turnout
- A 68% turnout scenario would mean that approximately 4,337,931 votes would be cast in the province.
- In that scenario, approximately 2,212,345 votes would give the ANC 51%.
- For the ANC to reach that, it could afford to lose approximately 136,219 votes, compared to what it secured on the provincial ballot in 2014 (2,348,564)
67%: Low turnout
- A 67% turnout scenario would mean that approximately 4,274,138 votes would be cast in the province.
- In that scenario, approximately 2,179,810 votes would give the ANC 51%.
- For the ANC to reach that, it could afford to lose approximately 168,754 votes, compared to what it secured on the provincial ballot in 2014 (2,348,564)
So, what do these numbers tells us? To better understand their implications, it is helpful to set out the ANC’s performance on the Gauteng provincial ballot over the last four elections:
- 1999: 2,485,064 votes for 67.85% [All registered Gauteng voters: 4,154,087]
- 2004: 2,331,121 votes for 68.40% [All registered Gauteng voters: 4,650,594]
- 2009: 2,662,013 votes for 64.04% [All registered Gauteng voters: 5,555,159]
- 2014: 2,348,564 votes for 53.59% [All registered Gauteng voters: 6,063,739 ]
The ANC’s problem in Gauteng, in a nutshell, is that for two decades it has failed to meaningfully grow its share of the electoral pool. As the pool has grown, the ANC vote share has stayed relatively stable, and thus, declined in the bigger picture.
Put another way, between 1999 and 2014, the ANC vote share has fluctuated between 2.6-million votes and 2.3-million votes. But, over the same period, the total number of registered voters in Gauteng grew by 1.9-million.
Here is a good way of illustrating the ANC’s problem: if you divide the party’s total votes for each of the last four elections by the total number of registered voters, you get the following ANC percentages:
- 1999: 59.8%
- 2004: 50.1%
- 2009: 47.9%
- 2014: 38.7%
Now, compare that to the DA’s performance in Gauteng, over the same past four provincial elections when you make the same calculations (DA voters as a percentage of total registered population):
- 1999: 15.8% [Actual number of votes: 658,231 for 17.97%]
- 2004: 15.2% [Actual number of votes: 708,081 for 20.78%]
- 2009: 16.4% [Actual number of votes: 908,616 for 21.86%]
- 2014: 22.2% [Actual number of votes: 1,349,001 for 30.78%]
This is the difference between a growing party and a declining party, over time. The ANC’s decline was masked by its final percentage. In 2004, it appeared relatively stable and its percentage actually increased marginally (from 67% to 68%) but behind the scenes its real decline had already started: when you compare its vote share to the total registered population, it had actually shrunk significantly in 2004, from 59.8% to 50.1%. And that trend has continued ever since.
The DA, however, has never lost ground. Its growth has seen it proportionally match the growing total registered voting pool in Gauteng from 1999 to 2009 and then, in 2014, to outstrip it.
And that is the ANC’s real problem: for 20 years it has failed to register new ANC voters in the province in significant numbers and, with time, that trend has caught up with it.
And that is the ANC’s real problem: for 20 years it has failed to register new ANC voters in the province in significant numbers and, with time, that trend has caught up with it. This failure means it now faces a monumental battle just to match its 2014 performance, come May 8.
Not only does it have to get it all its supporters to vote, even if it does, its unchanging share of the electoral pool means in a best-case scenario it is balancing on a 50% knife edge. And if the ANC vote is in anyway affected by apathy, it is going to experience a serious and exponential decline, because none of the opposition parties in Gauteng — the DA and the EFF in particular — seems to have the same kind of historical problem.
In the seven turnout scenarios, the ANC has some small breathing room. On the lowest turnout scenario (67%) it can afford to lose as much as 168,754 votes and still scrape a 51% majority. But between 2009 and 2014, it lost 313,449 votes. And that is the number that will really be worrying it. If anything like that happens again, the ANC will finish in the low forties. And there is little evidence to suggest a reversal.
(It is worth bearing in mind too, the threat that is the EFF. Over the past five years, unlike the DA, it has been able to eat directly into the ANC’s support base. The DA does that a bit, but grows more by attrition, registering new voters and on the back of differential turnout. If the EFF does take a chunk of potential ANC support in Gauteng, as it did in 2014 — and all indications are that it is set to grow further — it will only exacerbate the ANC’s problems.)
The number of new registrations in Gauteng this election is a relative pittance: just 315,572 new voters from 2014. By comparison, between 2009 and 2014, the IEC registered 508,580 new voters in the province, and between 2004 and 2009, a whopping 904,565 new voters. So that certainly does not suggest an insatiable appetite among the Gauteng electorate to get out there and vote this time. And you can be sure, even if marginally, the DA and the EFF out-registered the ANC among those 315,572 new voters.
The higher the turnout on the day, the more likely it is the ANC has succeeded and reinvigorated its supporters; the lower the turnout, the bigger the ANC’s problems are going to be. All the evidence points towards the latter being the case, and all of this is before the now very real advent of six weeks of load-shedding. If you support the ANC, and you are in Gauteng, things aren’t looking good at all.
• Van Onselen is the head of politics and governance at the South African Institute of Race Relations.
Would you like to comment on this article?
Sign up (it's quick and free) or sign in now.
Please read our Comment Policy before commenting.