Last Friday morning I woke to the usual neighbourhood WhatsApp group communiqués in Joburg’s Ward 87. Water pressure collapsing, supply might disappear altogether, no firm restoration time and City Power investigating yet another outage.
The timing was almost comic. The previous day Cyril Ramaphosa had launched the revised white paper on local government and observed, sententiously, that “trust will not be restored by speeches”. It would be restored when faults were fixed and officials held accountable. By breakfast his speech was curdling like the milk in my fridge.
The political question of why the ANC is losing support has largely argued itself. The government’s own white paper describes a municipal system damaged by patronage, politicised appointments, party interference, criminality, unstable coalitions and impunity. Quite a self-indictment.
Jason Lightfoot of Futuregrowth argues, as I have before, that the white paper’s dullest reforms may be the most useful: ring-fenced water and electricity revenues, separation of service authority from provider, and compulsory licensing. They go straight to bankability. Private capital needs predictable cash flows, credible regulation and risk sitting with institutions capable of carrying it, not revenues disappearing into a municipal slush fund or rules changing with the next coalition deal. It also means that if the DA leads Joburg, there’s a fighting chance; under this model, the private sector will get involved to help fix the mess.
Against this backdrop, you might expect the DA to be sweeping up the political debris. Yet the polls suggest otherwise.
Political analysts Dawie Scholtz and Gareth van Onselen recently had a revealing disagreement on X about what constitutes electoral “change”. Van Onselen points, fairly, to potentially extraordinary outcomes. An ANC collapse of historic proportions, the DA possibly leading Joburg and Ekurhuleni, the ANC below 50% in every metro and the national gap between the two largest parties narrowing substantially.
Scholtz’s response is more unsettling. If the underlying voting preferences of demographic groups remain largely where they were in 2024, and the apparent DA advance comes mainly because its supporters turn out at higher rates while disillusioned ANC voters stay home, then the electoral outcome has changed but the political structure underneath it has not.
That’s the problem, isn’t it?
The numbers make the point brutally. According to a Social Research Foundation (SRF) poll, in March the DA polled at 12% among black voters; by July it was 13%. Over the same period ANC support among black voters had fallen to 36%, while MK had climbed to 24% and the EFF stood at 12%. The black electorate is plainly fragmenting, but the DA isn’t capitalising on this.
There’s even a useful denominator trap here. The DA can correctly point out that black South Africans constitute the largest single racial component of its own support base (36% in July’s SRF polling) but that is not the same as saying it commands broad support among the black electorate.
What makes this more intriguing is that the evidence doesn’t suggest some impenetrable wall of black hostility towards the DA. Earlier SRF polling found that 61% of black respondents thought the party was performing well in the GNU. In October, 69% supported ANC-DA municipal coalitions where the combination could provide stable government. A majority of black voters have even supported the DA’s proposal to recast empowerment around disadvantage rather than race, while 51% of black respondents in late 2025 supported Helen Zille standing for Joburg mayor.
‘Racialised fluidity’
Therein lies the puzzle. Black voters can approve of the DA, work with the DA, endorse policies associated with the DA and sometimes prefer DA politicians to ANC incumbents, yet approval does not reliably become identification, and identification is what ultimately fills a ballot paper.
Political scientists Marcel Paret and Carin Runciman have a useful description for South Africa’s electorate: “racialised fluidity”. Their research found considerable movement between parties and abstention — roughly half their metropolitan sample changed position across three elections — even while race remained strongly correlated with party choice.
Michael Braun’s work on declining ANC attachment supplies the other half of the explanation. Many ANC voters no longer display the old partisan loyalty, but retain a thinner attachment built from history, habit, material experience and, crucially, the absence of an alternative they trust sufficiently to embrace.
That last point should bother the DA considerably more than another argument about messaging.
For two decades the DA has wrestled with this idea that it battles to break out of the suburbs and into the townships. What remains unresolved is the old question of whether a liberal party whose strongest inheritance remains rooted in minority constituencies can convince millions of black South Africans that they are not customers for a better-run state but co-owners of the political movement offering it.
If the opinion polls are right, November 4 may produce dramatic changes in councils while reproducing much of the country’s underlying political sociology almost perfectly.
The remarkable thing is that the ANC’s structural collapse hasn’t yet produced a commensurate expansion of the DA. Instead, the political centre is being vacated faster than the principal opposition can occupy it. That’s not merely an ANC problem or a DA problem. It is the unresolved problem of the post-1994 political order.
