(Strategic position as on the 18th November 2026)
I have striven not to laugh at human actions, not to weep at them, not to hate them, but to understand them.
Baruch Spinoza
Why can’t you understand, it’s so logical, so simple?!
Most of us
SNAPSHOT
The DA has just won the biggest share of the vote in Johannesburg, Tshwane and Ekurhuleni, and it may be about to govern none of them. That is not a paradox. It is what happens when a party wins an argument about competence and loses a conflict about identity.
The numbers tell some of the story. Nationally, the ANC fell from 45.6% in 2021 to 30.8% on 4 November; the DA rose from 21.7% to 24.6%. In Johannesburg the DA took 34.8% and 91 of 270 seats. In Tshwane it took 34.4% and 74 of 214. It is the largest party in all three Gauteng metros. And yet, as I write, the ANC has reportedly offered ActionSA the mayoralties of Johannesburg and Tshwane while keeping Ekurhuleni for itself, and Herman Mashaba's party appears closer to that deal than to any arrangement with the DA. Councils must be constituted within days.
This analysis assumes the freeze-out happens. It is written for a party I respect, whose governing record in the Western Cape and Cape Town I value, and whose arguments about competence and clean government I largely share. But respect also means that, as a South African voter, I hold them to high standards. The DA is very good at governing and still poor at conflict, and above all at the particular kind of conflict that decides who governs in South Africa: identity conflict. If it treats the next few weeks as a negotiation setback rather than a strategic diagnosis, it will win more votes in 2029 and still be outmanoeuvred afterwards.
What follows covers my assessment, often stated in my public written work, how to read what is happening, the identity dynamics the party keeps misreading, whether to stay in the Government of National Unity, how to conduct opposition in the metros, and a summary of a few specific conflict strategies they can consider at this stage.
Read the conflict correctly: this is not just about Gauteng
The first discipline of conflict analysis is to describe the conflict the other side thinks it is in, not the one you would prefer it to be in. The DA tends to describe coalition-making as a rational exercise: the largest party, the best record, the most detailed coalition framework, therefore the mayoral chain. Its opponents are playing a different game.
Three things are happening at once.
First, the ANC is using the metros to fight a national war: Reporting on the ANC's national executive committee suggests that many members blame the DA partnership for the party's collapse at the polls, that ActionSA is the first preference and the DA a last resort, and that the municipal talks are being treated as a testing ground for a wider reassessment of the GNU. One insider reportedly said the ANC was "failing in explaining to the black population our reasons for working with the DA." Another worried that a DA mayor in Johannesburg would use the office to put years of ANC decay on display. Read together, this is not hostility to DA policy. It is fear of a DA showcase, and of what being seen next to the DA does to the ANC's standing with its own base.
Second, ActionSA is behaving as a kingmaker with a grievance, not a kingmaker with a price:
Herman Mashaba's conditions were specific and personal: Helen Zille out of the Johannesburg race, Nasiphi Moya kept as mayor of Tshwane. He has called the DA a racist party and, before the election, called Zille racist outright. The rupture goes back to 2019, when Mashaba resigned from the DA after Mmusi Maimane was pushed out, saying he could not reconcile himself with people who believed race was irrelevant. The DA keeps reading this as ego. It is ego, but it is also something more durable: a founding story of exclusion. Grievances of that kind cannot be bought with an MMC post. This is probably the clearest, most visible instance of how the DA misreads identity conflict (again, not the same as identity politics).
Third, the numbers make the DA dependent in a way its rhetoric does not reflect:
DA plus ActionSA gives 134 seats in Johannesburg, two short of the 136 needed; 122 in Tshwane, comfortably over 108; and 106 in Ekurhuleni, seven short of 113. ANC, ActionSA, MK and EFF together reach 138, 128 and 109 respectively. In other words, every road to power in Gauteng runs through ActionSA. The party that said voters should choose between chaos and a DA mandate did not get a mandate, and must now negotiate with people it has spent a campaign describing as the chaos. The analytical conclusion is uncomfortable. The DA is not being frozen out despite winning. It is being frozen out because it won, and because a DA win in Johannesburg is read by its rivals as an identity threat, not a policy threat. A party that understands that will respond very differently from one that thinks it has merely been outbid. We are not seeing this.
Identity conflict, not identity politics: the DA's blind spot
For the past two years I have argued that the DA's central strategic weakness is not its policies, its candidates or even its communications. It is that the party does not understand identity conflict and therefore cannot integrate it into its campaign and coalition strategy. The Gauteng negotiations are the clearest demonstration yet. The distinction matters, so let me be precise, if my earlier work has not made the point clear enough:
Identity politics is a mobilization strategy: appealing to voters on the basis of race, ethnicity, language or group grievance. The DA rightly rejects it as a governing philosophy, and its commitment to non-racialism is one of its genuine assets. This form of voter manipulation is easy to read, easy to manage.
Identity conflict is something else. It is a conflict in which what is at stake is not only resources or policies, but people's sense of who they are, whether they are respected, whether their history is acknowledged, and whether they belong to the future being proposed. It exists whether or not anyone mobilises it. You cannot opt out of it by declaring yourself non-racial, any more than a doctor can opt out of an infection by declaring herself against disease.
The conflict-resolution literature has understood this for decades. John Burton argued that conflicts rooted in non-negotiable human needs, such as identity, recognition and security, cannot be settled by bargaining over interests. Jay Rothman showed that identity-based conflicts escalate when parties treat them as resource disputes, because the other side experiences the bargaining itself as a denial of what really matters to them. Herbert Kelman's work in the Middle East made the same point from another angle: parties will not make concessions that feel like a betrayal of who they are, however rational the deal.
My earlier work sets out the crucial, pioneering work done by Jonathan Haidt, Jay van Bavel and others in establishing the levers and pulleys of this essential modern conflict lens. This is precisely the DA's recurring error. It approaches conflicts that are about dignity, recognition and belonging as though they were about delivery, competence and portfolios. It offers a better-run city to people whose objection is not to well-run cities, but to the feeling of being lectured, rescued or replaced. It treats the question "who are you to fix us?" as a question about qualifications. It is not. It is a question about relationship, about the past and the future as experienced.
Case studies show how, if we get these conflict dynamics wrong, we not just fail to convince people, but we actually entrench them in their previously held views, despite our potentially objective arguments, and despite their continued refusal to be persuaded possibly being to their disadvantage. Think about that, and look at what we are seeing happening in real time.
The consequences are visible in the polling. A Social Research Foundation survey earlier this year found that 80% of black Johannesburg voters believed conditions where they live would improve under the DA, but only 25% intended to vote for it. That 55-point gap is the single most important number in South African opposition politics. It means the DA has largely won the competence argument and is losing the identity conflict. More competence messaging cannot close that gap. Only a different relationship can.
The Zille candidacy illustrates the problem without being reducible to it. Nobody serious doubts her administrative ability. But a party that read identity conflict well would have asked a second question alongside "who can fix Johannesburg?": namely, "what story will this candidacy tell, and to whom?" A white former premier arriving from Cape Town to rescue a majority-black city activates a deep and painful frame, whatever her intentions. Critics did not need to invent it. Days after the vote, one of them, casting doubt on an ANC-DA deal for the mayoralty, simply named it, mocking the idea of a "messiah" travelling from Cape Town to fix the city. Mashaba weaponized it.
The DA's response, that critics call it arrogance whenever the party refuses to drop its principles, is true as far as it goes. It is also exactly the response that confirms the frame. The lesson is not that the DA should choose candidates by race. It is that every choice a party makes in a divided society is also an identity signal, and the party must read that signal as carefully as it reads a budget. Persuading people takes more than objective facts. To change the popular meme – feelings don’t care too much about your facts.
The undertone of racial defensiveness: how to read it, how not to answer it
Anyone listening carefully to the past fortnight will have heard an undertone in some opponents' public statements that is not quite racial attack and not quite policy disagreement. It is closer to racial defensiveness: the sense that a DA victory in Johannesburg would be a judgment on black political leadership as such, and must therefore be resisted as a matter of collective self-respect. You can hear it in the reported ANC worry about explaining the DA relationship to its black constituency. You can hear it in Mashaba's "racist party" language. You can hear it in the Conference of the Left, the MK Party, EFF, SACP and allies, declaring that it will not back any ANC coalition that includes the DA.
And it was audible during the campaign itself, when a remark reported from the trail framed the election as handing power to a white person because black people had failed. The DA did not say that. But it is the kind of sentence that attaches itself to a campaign whose central message is that someone else has failed and you will fix it. The DA's instinct, or at least many of its public supporters, is to treat all of this as bad faith: race-baiting by parties that have run out of arguments. Some of it is exactly that. But a conflict analysis has to separate two things that look alike.
Racial mobilization is a deliberate strategy by elites to keep voters loyal by invoking group threat. It should be named and resisted.
Racial defensiveness is a real and widely shared emotional response, in this case among black South Africans who may be furious at the ANC, may be voting against it, and still feel wounded by a narrative in which the cure for perceived black misgovernment is perceived white competence. Elites exploit it, but they did not create it. It comes from history, from the daily experience of being judged, and from a post-1994 settlement in which political power was the one domain where black South Africans held undisputed ownership.
The mistake the DA makes is to answer the second as though it were the first. When the party responds to defensiveness with accusations of race-baiting, it tells the defensive voter: your feelings are a trick being played on you. That is experienced as one more humiliation, and it hands the mobilisers exactly what they need. What would a skilled response look like?
Again, a lot of my earlier work dealt with that, but the Gauteng metro battle has really sharpened the focus on many of the underlying conflicts. There is a very specific way in which identity conflict persuasion works, and that will require far more space than what we have here, but a few simplified points would be:
Say plainly that the history behind the reaction is real, and that the DA understands why a white-led rescue narrative is painful, even to people who want the ANC gone. Acknowledgement is not concession. In identity conflicts it is the price of being heard at all. The harm is done with this election, but for future campaigns, this should be the key to growth and stability.
Respond to Mashaba the strategist firmly, but never dismiss the constituency his language speaks to. Never answer "racist" with "racist". Helen Zille and Geordin Hill-Lewis have been generally good with this throughout the campaign, but some of the DA’s senior people have been particularly counterproductive with public comments at crucial times.
Voters can tell the difference between a leader who owns the message and one deployed to defend it. The DA's black public representatives in Gauteng should be the primary voice on these questions, and should be trusted to say things the federal leadership would phrase differently. This could have been managed much better.
The most damaging implied message of the campaign was that one must choose between dignity and working services. The DA's message should be that working services are a form of dignity, and that the party's project is to restore the promise of 1994, not to rescue the city from it. None of this requires abandoning non-racialism. It requires recognising that non-racialism is a destination, and that in a society this wounded, the road to it runs through acknowledgement of race, not around it. That was, after all, the very argument on which Mashaba left the party in 2019.
"Only the DA can fix this": why the narrative polarises
The freeze-out debate has once again split South Africans along familiar lines. Many white, coloured, Indian and middle-class black voters see a straightforward injustice: the largest party, with the best record, locked out by a coalition of losers. Many black voters, including many who did not vote ANC, are uncomfortable with that reading. They are not defending the ANC. They are recoiling from the claim underneath it: that only the DA can fix this.
That claim does three things the DA does not intend.
It makes the DA's victory someone else's defeat as people, not only as parties
"Only we can fix it" implies that everyone else, including the voters who backed ActionSA, the EFF, MK or the ANC, has chosen chaos. Zille put it almost exactly that way before the election, asking whether voters wanted chaos or a DA mandate. That is a fine mobilizing line for the DA's base. To everyone else it says: your vote was a mistake.
It turns a plural electorate into a binary
The 2026 result was not a DA landslide. It was fragmentation. Voters spread their support across the DA, ActionSA, MK, the EFF, the ANC and others. A narrative of exclusive salvation misreads that result, and it makes the DA look as if it is trying to win a conflict it has not yet won at the ballot box. It publicly dismissed parties and individuals across those identity lines that it now required.
It locks the DA into a saviour role that coalition partners, and large parts of the electorate, cannot accept
No party joins a coalition to be rescued. If the DA's brand is that it alone can govern, every potential partner faces a choice between subordination and resistance. ActionSA chose resistance, and the ANC made it easy for them. The alternative is not false modesty. It is a shift from exclusive competence to shared ownership. Compare two sentences:
The second makes the same claim about capability, but it makes the city's renewal a project voters and partners can join, not a verdict they must submit to. It allows a black voter in Soweto who backed ActionSA to feel included in the story rather than indicted by it. Skilful storytelling is a massive driver of identity conflict persuasion, and the DA’s campaign has been a part of its current results.
There is also a hard electoral point here. The 55-point trust-to-vote gap among black Johannesburg voters will not close through more evidence of competence; it is already priced in. It will close when voting DA no longer feels like a statement about one's own community's failure. Language of exclusive salvation widens that gap every time it is used.
Should I stay or should I go? The GNU quo vadis?
Geordin Hill-Lewis has already said that if the DA is iced out locally, members will inevitably ask whether it should stay in the national government, and that it will be a tense time. He is right. The emotional case for walking will be strong: why prop up a partner that has just conspired to lock you out of the cities you won? My advice is clear, in line with what I said last year during an earlier crossroads moment: stay, but stop staying on the same terms. The reasoning matters more than the conclusion, so let me set it out.
Again, the November local elections have provided us with clarity that uncontrolled conflict escalation often brings.
The case for walking
It deserves a fair hearing. Walking would give the DA clarity going into 2029: it would no longer be tied to anticipated ANC failures, especially on crime, the economy and corruption. It would end the corrosive cycle of threats and retreats. And it would signal that the party will not accept being treated as a junior partner nationally while being treated as a pariah locally. These are real arguments.
Why walking would be a strategic error
1. The ANC wants you to walk
Reports this month suggest the ANC's NEC is leaning towards calling for the DA's removal from the GNU, and earlier in the year sources indicated the ANC would welcome a break-up provided the DA triggered it. The ANC's problem is that expelling the second-largest party, which brought markets, investor confidence and reform momentum into government, is costly and unpopular with a large part of the electorate. If the DA walks, the ANC escapes that cost and gets a story: we tried, they left. In conflict terms, never solve your opponent's problem for them.
2. Walking surrenders the levers that matter for 2029
The DA's case to the country rests on demonstrating national governing competence. Its ministers are its strongest exhibit. Leaving gives that up for a gesture. Until, and if, they start working identity conflict levers successfully, this is their engine room.
3. Walking opens the door to the alternative coalition
An ANC-led government relying on MK and the EFF would hold a parliamentary majority. Gayton McKenzie has already said, in effect, that the ANC can remind the DA of that option. The DA should not be the party that makes it happen.
4. Threat inflation has already cost credibility
The DA has threatened to leave the GNU repeatedly since June 2024, over Cabinet posts, the BELA Act, the Expropriation Act, the budget and the National Dialogue. Each threat not carried out has lowered the value of the next. Another walk-out threat now will be heard as noise; an actual walk-out will be heard as a tantrum after losing Gauteng. Neither helps. In identity conflict that is costly, and unnecessary.
5. The metro freeze-out is a local identity conflict; the GNU is a national interest coalition set in a larger identity conflict
They are different conflicts, and it is a classic escalation error to merge them. Linking them tells voters the DA's participation in national government is conditional on getting mayoral chains, which is exactly the self-interested reading opponents want.
How to stay, but on different terms
Staying must not mean staying quiet. The DA should:
State publicly that the DA remains committed to the GNU's statement of intent, will continue to serve, and will leave only if expelled or if the agreement's core principles are breached. If the ANC wants the DA out, it must say so and carry the cost.
A short, public list of specific breaches that would end participation, such as the rule of law, fiscal stability or the independence of key institutions, is worth more than another season of brinkmanship.
Put the metro conduct on the record as a breach of the spirit of partnership through the clearing-house process, in writing, without public theatre. It builds a record if an exit is ever needed.
Say in clear terms: we will not punish South Africans nationally for what happened in Gauteng. That sentence alone takes the moral high ground and denies the ANC its escalation story. It is also a big first step in the identity conflict work that needs to be done.
Opposition in the metros as a governing strategy
Hill-Lewis has said the DA is prepared to sit on the opposition benches in Johannesburg, Tshwane and Ekurhuleni if it must. Good. But how it sits there will decide whether 2026 becomes the prelude to a 2029 breakthrough in Gauteng or another chapter in the revolving door. The danger is that the DA plays the role its opponents have already written for it. The Freedom Front Plus has accused the DA of wanting Johannesburg to collapse so it can return as saviour. ActionSA has made similar arguments. If the DA's opposition looks like waiting for failure, that accusation will stick, and the saviour frame will harden further.
The better model is opposition as visible stewardship:
The point is strategic, not only ethical. If an ANC-ActionSA government fails, voters must remember the DA as the party that tried to help it succeed. If it succeeds, the DA must be visibly part of that success. Either way, constructive, skilled opposition wins.
Summary: conflict strategies if the freeze-out happensImmediate (the first 72 hours)
Congratulate the incoming mayors by name, including Nasiphi Moya and whoever takes Johannesburg. Grace in defeat is the cheapest and most powerful identity signal available.
It was a lawful coalition of elected councillors. Calling it a theft of the voters' will alienates the very voters who chose those parties. This goes for senior DA public figures in particular.
Something like: we respect the outcome, we will hold the new government to account, and we will support every decision that helps residents. Repeat it; do not freelance.
Say explicitly that national government will not be held hostage to local disputes. Short term (the next three months)
Stay, publish a brief list of genuine red lines, and use formal dispute mechanisms rather than press conferences.
Medium term (to 2029)
Replace "only the DA can" with shared-ownership framing: renewal as a project others can join.
Develop a clear public language for engaging race, history and dignity without abandoning non-racialism. Train leaders in it.
Be firm with Mashaba the strategist; be respectful towards the voters and councillors who backed ActionSA.
Build working ties at council level with ActionSA, smaller parties and independents now, so that 2029 negotiations start from trust rather than from grievance.
Ask not only who can govern, but what story the choice will tell and to whom.
Bring conflict and negotiation specialists into the party's coalition and strategy units, with a specific mandate on identity-based conflict, not just bargaining and portfolio allocation.
Closing: the conflict the DA must learn to win
The DA has spent two decades proving that it can govern. On 4 November it proved that a growing number of South Africans, across racial lines, believe it. That is a historic achievement, and the party should take quiet pride in it. But South African politics is not decided only by who can govern. It is decided by who people feel they belong with, who acknowledges their history, and who they can imagine sharing power with. Those are the terms of identity conflict, and on those terms the DA keeps losing even as it wins.
Deny it, or argue about it as much as necessary, but this is the invisible wall right here, the place from which they reach a higher level, or start stagnating. This was a very visible, expensive national lesson in identity conflict. The freeze-out, if it comes, is painful. It is also useful. It shows, more starkly than any poll, that competence is necessary but not sufficient, and that a party which wants to govern Gauteng must learn to read dignity, recognition and belonging with the same rigour it brings to a budget.
This is not the place the DA and its voters would have expected itself to be, but maybe it forces a reckoning long in the making. The DA does not need to become a different party. It needs to become a better political conflict actor. If it does, 2026 will be remembered not as the year it was locked out, but as the year it learned how to be let in.
RESOURCES AND RECOMMENDED READING
Full references, further reading material, courses, coaching, study material, workshops / seminars, mediation and representation are available on request.
Andre Vlok can be contacted at andre@conflict1.co.za for any further information.
(c) Andre Vlok
November 2026