The Social Research Foundation (SRF) recently published, in The Common Sense, a national opinion poll placing the DA on 28% national support, the poll having a 2-point margin of error. Internal DA numbers paint a similar picture of the party in the high 20s.
Given that the SRF’s survey in late 2025 put the DA on 32%, and that the IRR’s polling in April/May 2025 had the DA on 30%, the picture emerges of a DA now broadly able to claim the votes of 3 out of every 10 voters. This should be of some encouragement to the party and its supporters. In the aftermath of its disastrous 2019 national election, where it lost vote share for the first time since its formation, the DA was put on death watch. Opponents and commentators purred with satisfaction over the party’s imminent demise.
In October 2019, when a period of intense drama shook the party, with Mmusi Maimane departing, Helen Zille returning to the party’s top leadership, and John Steenhuisen taking the reins, the DA was at 16% national support. The recovery in six years to stabilise around double its 2019 low is worth appreciating from an objective standpoint as a formidable political accomplishment. It was not a given that the party would, or even could, recover.
So, a DA around 30% should give the party some satisfaction. Even so, there are further data points in SRF polling over the last few months that complicate the picture: points that should both encourage the DA and, more importantly, give it an urgent wake-up call: that it needs to change a few things about how it seeks electoral support.
To start with the encouraging stuff, the DA has a decent favourability score, with 42% of likely voters viewing the party either very or somewhat favourably. The party also performs best on a range of critical competence measurements. Among likely voters, 42% said the DA was the best party on service delivery, 41% on clean governance, 36% on working in coalitions, 35% on good leadership, 32% on accountability, and 32% on good policies. Those are positive pluralities in each case, placing the DA ahead of all competition in the field.
And yet the wake-up call these numbers should give the DA is that it is under-performing its own favourability by a double-digit margin. A party that has a 10-point gap between its favourability and support levels is doing something wrong.
Defining issue
In the case of the DA, this is likely the defining issue that will either frustrate the party’s ambition to become South Africa’s largest party or, if resolved, pave the way for it, allowing the DA to establish itself as a party with support of 35% or, in the most optimistic estimate for the party, reaching 40%.
That is what makes the DA’s present position so interesting. Over the 26 years of its existence, the DA has built an impressive brand. The message that it is a party that delivers has clearly been successful. The numbers prove it. The party is seen by a consistent plurality of registered and likely voters as the most competent party in the country. And yet, it is not the strongest party in the country, suffering from a seeming perpetual inability to rise to this level, despite an increasingly weak ANC.
It is important to understand why this is the case. I have seen many a comment, from professionals and lay observers alike, that this discrepancy between competence and support for the DA exposes the unreliability of opinion polling as a means of understanding politics in South Africa. No party, the thinking goes, can essentially underperform itself. The numbers must be wrong. Or the whole exercise of opinion polling is a waste of time.
However, this is a conclusion based on the erroneous assumption that competence and support are, electorally speaking, one and the same thing. That assumption collides awkwardly with the research, from around the world, on the idiosyncratic workings of the mind of the voter. Like consumers, voters are too often considered to be entirely rational actors. This is something of an outdated artefact of rationalist insistence.
Sport
The assumed equivalence of a 1:1 relationship between competence and electoral support betrays a misunderstanding of the reality that party competence is only one of a range of assessment sources voters use, consciously or subconsciously, with ballot paper in hand. One can take a step back and place the issue of support psychology in a different context familiar in South Africa: sport.
Teams, from soccer to rugby and everything else, do not attract support clinically merely from their win record or ability. Brilliant teams are disliked by many, whilst struggling teams are clung to loyally by others. The human brain is not a dispassionate calculator, but a more complex decision-making machine, capable of remarkable feats of objectivity, but, of course, intensely personal, subjective, and experiential.
Seen in that light, the first point to grasp, in assessing the DA’s predicament of underperforming its own reputation, is that favourability is not the same thing as political support. A voter can think well of the DA, approve of its role in government, and even concede, as many do, that it is better able to govern than its rivals, without feeling that it is their party. The SRF numbers show precisely this kind of gap. The DA is attracting respect, in some cases even begrudged, beyond its electoral footprint. It is failing to convert that respect into identification, emotional and societal permission, and endorsement on the same scale.
This DA failure becomes clearer when one looks at how the favourability-support gap appears across different parts of the electorate where the DA should want to accomplish growth. Among black possible voters, the DA is on 15% support, yet its favourability with these voters is again almost 10 points higher at 24%.
Among township residents, it is on 16% support, but its favourability similarly rises to 24%. In KwaZulu-Natal, a province rapidly detaching itself from the usual patterns and expectations of South African politics, the DA also sits on 16%, while its favourability is 26%. In the Eastern Cape, neighbouring province to the DA’s Western Cape fortress, it is on 14%, but its favourability is 30%. Among isiZulu voters, the single biggest language bloc in the country, the DA is particularly weak in terms of support at 5%, yet its favourability is four times that, at 20%. Even among coloured voters, where the party is in a decent position with above 50% support, favourability still runs higher still at 66%. Only amongst white voters and Indian voters do we see matching favourability and support levels.
Not willing to support
These are hardly the numbers of a party in crisis, one that voters dismiss, or one in need of more efforts to convince people of DA-governing competence. Instead, they are the numbers of a party that many voters are willing to credit as able, but, given the topline party-support figures, are not willing to support.
That distinction matters because it tells us where the DA’s problem is located. And, almost more importantly, where it is not. The problem is not that voters have failed to acknowledge the party’s competence. Many clearly have, even those not supportive of the party. The problem is that competence, by itself, is no longer doing enough political work for the party to grow, come election time.
The DA’s well-established default message, that where it governs, it governs better, was indispensable in establishing its modern brand. It gave many uncommitted voters in the immediate post-1994 democratic contests a practical reason to take the party seriously and vote for it out of pragmatism, and it more recently helped pull the DA out of the slump in which it found itself after the 2019 election decline.
But the very success of that message now creates a new challenge, if the party is to grow. Once a party is already widely recognised as more competent, simply repeating that claim does not bring in more voters. At some point, it can begin to produce diminishing returns. In the DA’s case, it may, in fact, be starting to produce negative returns.
That is because the DA’s usual track-record message no longer lands in a neutral emotional environment. For many voters outside the DA’s established base, the party carries an image of ability, but also of distance. The acknowledgement of an ability to govern better than other parties too easily runs into the difficulty of the party’s ‘perception’ weakness: that it is not a party where black South Africans can or ought to feel at home. The track-record message answers the question of who is more efficient, but it does not answer the more emotive question of whether the party cares about “people like me”. Here, the DA is very clearly struggling.
The SRF data gives us several ways to see this. One is through the party’s negative attributes. Among likely voters overall, 32% associate the DA most with the label “racist”, while 17% associate it most with being “anti-poor”. Among black voters, the “racist” figure rises sharply to 45%. Among township residents, it is 43%. In KwaZulu-Natal, 42% attach that label most to the DA, and among isiZulu speakers the figure reaches 48%.
These numbers should not be read lazily or melodramatically. They do not mean that nearly half of all black or isiZulu-speaking voters are obsessed with the DA being a “white party” or that they are all beyond reach.
No internal contradiction
Even so, the fact that a party can be acknowledged as competent whilst being considered racist contains no internal contradiction. Racists can perhaps be competent and vice versa. So, there is no inherent issue there. But when a party like the DA faces this perception from groups of voters that it needs to convert into supporters, a message like “the DA difference” becomes self-defeating.
Given the DA’s political base, in terms of governance record and support, the DA’s boast of particularly Cape Town and Western Cape competence serves as a reminder of both the party’s competence and, damagingly, its distance, its elsewhere focus, removed from the majority of voters who struggle, for reasons of identity, geography, language, race, and other forms of association to see the governing successes of the DA in Cape Town and the Western Cape as to the benefit of “people like me”.
This is why the track-record message is now less straightforward than it once was. In a constituency already comfortable with the DA, it reassures to an extent. In a constituency far from Cape Town, in more ways than simply distance, but identity, and where the party is still viewed as distant, socially outside, or “for others”, it reinforces the distance and the sense of non-ownership many voters have towards the DA. A voter may hear: yes, the DA governs better, but it governs better for Cape Town, for suburbs, for minorities, for the middle class, for people unlike me.
It does not help the DA that its heartland of the Western Cape is the province that, in superficial shortcut of identity association, least resembles those black voters whom the DA needs in order to grow. Whether this perception of the DA being ‘for others’ and not for the majority of black voters is fair is, politically, beside the point. The point is that it exists, and the polling shows that it is strong enough to suppress conversion from perception of competence and favourability to support.
This is also not to say that the black voters who do not vote for the DA do so simply out of racist bias. More accurately understood, it is the data-based reality that many black voters see the DA achieve success for communities and people visibly different from their own communities and lives. This is true even for those who view positively the ability of the DA in office.
Pre-existing scepticism
The problem with the DA’s reliance on a binary of track records of competence is that every time the DA directs attention to its especially Western Cape successes, it reinforces a pre-existing scepticism towards the party: that it is competent, but competent for other people. People who are not, for the majority of South Africans, “like me”. It is a party that, to coin a phrase, gets things done, but for people over there, who live without “my” particular experience, struggles, fears, and ambitions.
That is why the notion of “lived experience”, however easily parodied, cannot simply be wished away.
One may reject the phrase, but that rejection cannot alter the reality that lived experience, often filtered through complex rational, emotional, and psychological processes to shape political conclusions, is key to political messaging. The ability to share with a potential supporter, whether politically or commercially, the tangibility of a free sample to affect momentarily their lived experience is a universally established part of marketing. That free sample nudges the receiver’s lived experience into a new direction.
Describing the incredible taste of this new morsel through advertisements and descriptions and depictions is worth only a shred of the impact of a genuine lived experience of it. Experience beats explanation every day of the week. What is good for commerce is good for politics.
And in terms of a free sample, an illustration of impact that ordinary voters could taste and feel, there is a recent notable example that showed the DA’s problem is not that ordinary voters are incapable of rewarding competence with reconsidered political loyalties. On the contrary, they do respond when competence is made immediate, tangible, and clearly aligned with their material interests, rather than with more remote coastal competence.
Stance with impact
The DA saw enormous benefit from the VAT showdown, when it took a stand against the ANC’s attempt to increase VAT. That mattered, because VAT is one of the few fiscal issues with truly universal cut-through. It is not abstract. It is felt in the cost of living, in the till slip, in the pressure on stretched households. The result of the DA taking a stance with impact beyond its traditional base was striking.
The DA’s support among black voters rose from 5% in September/October 2024 to 18% in April/May 2025. That was not simply a moment in which the party appeared distantly competent. It was a moment in which the DA appeared to be concretely, and in the lived experience of even non-DA voters including those who occupy the 10-point space between the DA’s support and its favourability, on the side of ordinary people against a policy that would immediately make life more expensive.
It was not only that the DA was on the right side of a crucial bread-and-butter issue, but that the issue itself was not merely to the benefit of people far off, but tangible, and universally part of millions of lived experiences, enough to cut through the party’s normal distance problem. The DA’s VAT victory, perhaps for the first time in the party’s history, brought its widely perceived ability to bear on the lives of all South Africans.
That lesson has an internal corollary, and it is an uncomfortable one. There is a dangerous habit of thought inside the DA, from top to bottom, and among many of its supporters, that should be treated with immense caution, and possibly deliberately eradicated. It is the familiar sentiment that voters, especially non-DA voters, “get the government they vote for”.
It is a line heard most often when ANC failures affect areas where the electorate rejected the DA. In one sense, this sentiment is generally true. And the frustration behind it is entirely understandable at a human level. A party with a reputation for competence, and with numbers acknowledging it, like these SRF ones, will naturally feel aggrieved when voters still do not reward it accordingly. But that frustration is both useless and dangerous.
Dangerous
It is useless because, while voters do indeed get the government they vote for, that observation explains nothing about how a party that wants to govern more broadly should behave. And it is dangerous because it encourages the DA to retreat into a posture of injured and distant superiority. It allows the party to say, in effect, that if people do not choose DA competence, they do not deserve the competence that the party bestows on “other” places. That may feel emotionally satisfying to frustrated supporters. Rationally, it is fair. Politically, it is poison.
Why? Because it adds more substance to the very perception at the heart of the DA’s problem: acknowledged competence, but the fruits of which are not meant for “people like me”.
A politics of hurt disdain only deepens the impression that the DA’s good government is something offered on moral probation, rather than a public good to which all citizens are equally entitled. In a very real sense, the DA must ask itself whether it wants merely to be right, or whether it wants to win.
If it wants to win, it needs to move away from this entirely fair but petulant posture. It cannot behave as though those who do not yet vote DA are simply beyond its sympathy. A party seeking to grow cannot afford to speak as though its best qualities are reserved only for those enlightened enough already to support it, especially not when this is already a brake on the party’s growth ability.
This is why the SRF’s ideological findings are, in addition to political and perception findings, so important for the DA. They suggest that the electorate is more open to a number of the DA’s core policy instincts than the DA’s own support levels would suggest. Overall, 68% of respondents support directing upliftment by poverty and disadvantage rather than race, while 75% support public procurement based on value rather than race.
Need-based upliftment
Among likely voters, those figures rise to 72% and 77%. Among black respondents, support still stands at 60% for need-based upliftment and 70% for value-based procurement. Among township residents, it is 65% and 69%. Among ANC voters, 62% support need-based upliftment and 71% support value-based procurement. In the Eastern Cape, both figures are 79%. Even among MK voters, support stands at 52% for need-based upliftment and 66% for value-based procurement.
These are not small details. They matter because they suggest that the DA does not need to retreat from non-racialism, merit, or clean procurement in order to grow.
The public is already considerably more receptive to those ideas than many in the political class assume. The problem is not primarily one of principle, but of translation. Too often, the DA communicates sensible ideas in a way that focuses attention away from rather than directly to the lived experience of many South Africans who regard it as competent and might be converted into DA voters, but aren’t yet at that point.
That point becomes especially important when one looks at the blocs from which the DA would need to draw, if it wants to move materially beyond 30%.
Start with black working-class and township voters. This is the largest and hardest task, but also the most important. The numbers here are low enough to caution against fantasy, yet strong enough to show a foothold. More importantly, on practical local-government questions the DA is already competitive in these environments. Among township residents, 29% say the DA is the best party on service delivery, ahead of the ANC’s 21%. On clean governance, the DA is on 24%, ahead of the ANC’s 19%. That is not what an irrelevant party looks like. It is what a party with partial credibility and incomplete legitimacy looks like. The route forward here is not what too easily comes across as self-congratulation for far-off achievements. It is making competence feel credibly part of a promise of citizen and resident service and protection.
Practical governing actor
The second major route lies among softer ANC voters. Here too the numbers suggest more openness than conventional political commentary often allows. 14% of ANC voters are favourable towards the DA. 57% say the DA is performing well in the GNU. 73% agree that if the ANC and DA can form a majority in a municipality, they should do so. That does not mean ANC voters are about to become DA loyalists en masse. But it does mean that a meaningful minority is already capable of seeing the DA as a practical governing actor.
The opening here is not to humiliate ANC voters or demand that they repudiate their political history. It is to invite practical judgment, especially in local government. The message cannot be “you were fools to support the ANC”. It has to be “you have the right to insist on a municipality that works”.
Among coloured voters, the DA already holds a dominant position, but even here it under-converts its goodwill. The party’s support is 55%, while its favourability is 66%. The obstacle in this bloc appears to be less about accusations of racism than about tone and class coding. Only 10% of coloured respondents associate the DA most with “racist”, but 27% associate it most with being “anti-poor”.
Among PA voters, that figure rises to 47%. That tells us something important. In this part of the electorate, the DA is not chiefly losing on principle. It is losing where it sounds aloof, too suburban, or insufficiently on the side of working households under pressure. A party can have the right policy and still fail to sound as if it understands the life of the person hearing it.
KwaZulu-Natal and the broader isiZulu-speaking electorate present another kind of challenge. The DA is weak there, but once again favourability runs ahead of support.
The problem is not that there is no opening at all. The problem is that the opening sits inside a field crowded by identity, history, and emotional politics.
Comparable advantage
Here the DA’s best self is unlikely to be an imitation of the loudest actors in the field. Its comparative advantage is seriousness, order, and restoration. In a province worn down by spectacle, factionalism, and municipal collapse, the party’s route lies in becoming the calm, no-nonsense force of repair. But it needs to speak in a way that sounds local and grounded, not imported and triumphalist.
The Eastern Cape may be one of the most important medium-term opportunities, precisely because the party’s support there is so much lower than its favourability. At 14% support and 30% favourability, the province offers clear headroom. The issue mix there is also instructive: crime (21%), roads (16%), jobs (15%), drugs (12%), and housing (10%) dominate. This is an electorate not looking for ideological performance art. It is looking for repair. That should shape the DA’s presentation. In the Eastern Cape, the party should sound less like a national lecture and more like the party of fixing roads, restoring standards, repairing towns, and ensuring public money goes to the public rather than to patronage.
The same broad point applies to poorer households and job seekers across the country. The DA will not become a substantially bigger party by merely becoming even stronger among the affluent and the formally secure. Yet here too, the data suggests that support lags sentiment. Among households earning R2,000 to R3,000 per month, the DA sits on 21%, but its favourability is 35%. Among those looking for work, it is on 16%, with favourability of 27%.
The lesson is not that poorer voters are ideologically unavailable to the DA. It is that the party must learn to translate reform into household arithmetic. Clean government must mean lower waste and more reliable services. Better procurement must mean fewer inflated costs. A pro-work economy must mean fewer barriers to trading, earning, and getting ahead.
Jolted
The case that emerges from the SRF data is therefore quite clear. The DA should be encouraged by its recovery to around 30%, by its 42% favourability, and by its lead on competence measures. It has rebuilt itself from crisis. But it should also be jolted by the fact that it is now underperforming its own political assets.
A party that is seen as the most competent in the country, yet remains stuck well below its own favourability ceiling, cannot simply tell itself that it must say the same thing louder and more, especially not as an election campaign of intense competition for the metro battlegrounds is already heating up.
The DA’s route forward lies not in ideological compromise or pandering, at which it is historically quite bad, but in changing the emotional centre of its message. The DA needs to move from saying “we are better” to showing, concretely and repeatedly, how government on the side of ordinary people makes life safer, fairer, and less expensive. Its track record should remain key, but as proof of a new formula, not as self-satisfied self-description. Or worse, self-congratulation.
In that sense, the DA’s track record must be repurposed to be less about its achievements and more about its bursting eagerness to be the servant of more people. Good government needs to be presented as affecting lived experience: water that runs, roads that work, fewer insiders jumping the queue, safer streets, and jobs no longer blocked by dysfunction. Governmental competence cannot be framed in clinical terms as part of a party’s CV or application for political power. Instead, it must be brought home as the basis for emotional buy-in, not mere cerebral acknowledgement.
A flavour of humility is needed here: the DA cannot continue saying “look at what we as a party have done”. Instead, it must be willing to share the bulk of the success of the places where it governs with the people of those places.
Repackage its message
Instead of saying the DA made Cape Town great, the party must fully repackage its message to proclaim in no uncertain terms that the DA stood in a supporting role to the people of Cape Town, and that the formula for success is not “people supporting the DA = governing excellence”, but that “the ingenuity and hopes of people like me, supported in a posture of service by the DA = a better life for people like me”. It is the difference between a DA that appears willing to serve “people over there” and a DA that is committed to help “people like me”.
The SRF polling, understood holistically, offers the DA both encouragement and warning. The positives are obvious: the party has recovered from a near-death period to become a stable 30% force with a powerful reputation for competence. It is well-poised to entrench its position as an anchor of moderate, common-ground coalitions. But the urgent warning is perhaps even more obvious: that reputation is no longer enough on its own.
Until the DA learns how to convert objectively granted respect into subjectively felt association rooted in compassion for “people like me”, this association into reliable votes, and these votes into sustainable amplifiers of on-the-ground legitimacy and ownership, it will continue to receive political credit it is incapable of fully benefiting from. It will continue undermining its own mines.
As long as the DA believes it can win through stubborn repetition of its credentials and achievements, pointlessly comparing its competence against that of its rivals, its leaders and supporters can pontificate till they are blue in the faces; the DA will not make a difference in getting more votes for itself.
[Image: https://www.facebook.com/photo.php?fbid=1280537444264538&set=pb.100069248828030.-2207520000&type=3]
If you like what you have just read, support the Daily Friend