“Recent South African opinion polls indicate that support for established political parties is changing ahead of the November 2026 local government elections, while voters’ assessments of the Government of National Unity, municipal performance and political leadership reveal a complicated picture of public confidence. Research published by The Common Sense and the Social Research Foundation, alongside reporting by eNCA and Daily Maverick, suggests that fragmented voter preferences and closely contested municipalities could make coalition negotiations and local service delivery central issues after the election.”
South African Voter Sentiment Shifts as Public Confidence, Party Support and Coalition Politics Face a Crucial Test
JOHANNESBURG, 16 October 2026 — South Africa is approaching a significant test of public opinion as political parties prepare for the November local government elections. Recent polling points to changing party favourability, divided voter preferences and continuing questions about municipal performance. The findings suggest that the election will involve more than competition between established political organisations: it will also test whether political leaders can persuade citizens that their votes will translate into accountable government and improvements in everyday life.
Research published in recent months by The Common Sense, the Social Research Foundation (SRF), Ipsos and other organisations provides insight into the political mood. Although the surveys differ in methodology, timing and projections, they indicate a political environment in which party loyalties are changing and several municipalities could face difficult coalition negotiations.
Public opinion reveals a complicated political picture
One notable finding comes from a September 2026 report by The Common Sense and the SRF. Their polling series indicated that the share of likely voters expressing a favourable view of the African National Congress (ANC) had declined from 41% to 30% across five survey rounds since February 2025. Over the same period, the share saying that the Government of National Unity was performing well remained within a relatively narrow range around half of respondents, never falling below 48%.
These findings highlight an important distinction between attitudes towards individual political parties and evaluations of a coalition government. The ANC is the largest party in the GNU, but its declining favourability does not automatically mean that citizens assess the coalition in the same way.
For voters, government performance may involve several considerations, including economic conditions, public services, political cooperation and the extent to which leaders deliver on their commitments. A coalition can receive a different public assessment from that of any individual party because citizens may judge the government’s collective actions separately from their opinions about its members.
However, survey results should not be interpreted as proof that the public has reached a single, settled position. Respondents have different priorities, and favourable opinions do not necessarily translate into voting intentions. The polling provides a snapshot of sentiment at particular times, rather than a definitive account of every South African’s political views.
Changing party support complicates election expectations
Reporting by Daily Maverick in September described an increasingly fragmented political landscape, with voter support distributed across the ANC, Democratic Alliance (DA), uMkhonto weSizwe Party (MK), Economic Freedom Fighters (EFF) and smaller political organisations. The publication highlighted differences between polling organisations and cautioned against assuming that declining ANC support would necessarily produce an equivalent increase for one opposition party.
This distinction is important because political dissatisfaction can produce several different responses. Some voters may support an established opposition party, while others may prefer a smaller organisation that they believe better represents their interests. Some may remain undecided, and others may abstain from voting altogether.
Consequently, a reduction in support for one party does not automatically identify which alternative will benefit. The outcome depends on voter choices across constituencies, participation rates and the electoral system.
The findings also illustrate why national polling figures must be interpreted carefully when assessing municipal elections. Local government contests are shaped by neighbourhood concerns, candidates, ward-level competition and the performance of individual councils. National political attitudes can influence these decisions, but they cannot fully explain every local result.
Municipal performance remains central to public confidence
Local government is where many citizens experience the practical consequences of political decisions. Water supply, electricity infrastructure, refuse collection, roads, housing administration and municipal finances influence how residents evaluate their councils.
The eNCA and MarkData polling discussion published in October illustrated the connection between voter sentiment and concerns about municipal administration. In Ekurhuleni, the survey measured DA support at 31.2% and ANC support at 29.4%. In Nelson Mandela Bay, the corresponding figures were 37.7% and 37.4%. These were survey estimates, not election results, and the narrow difference in Nelson Mandela Bay should not be treated as a statistically decisive lead.
The same reporting combined negative responses about municipal management and found that 58.7% of respondents in Ekurhuleni and 65.4% in Nelson Mandela Bay regarded their municipalities as badly run. Such answers measure perceptions, rather than independently verifying the condition of every service or the performance of every municipal department.
Nevertheless, public perceptions matter in democratic accountability. Residents who believe that local government is failing may demand new leadership, stronger oversight or changes in how municipal resources are managed. Political parties must therefore explain how their proposals would address residents’ concerns and how their progress could be measured.
Service delivery is not determined by elections alone. Municipalities also depend on administrative capacity, financial management, infrastructure maintenance and the ability of councillors to approve and implement budgets. Even a party that receives strong electoral support must confront these practical responsibilities once it takes office.
Coalition negotiations could become a defining challenge
Fragmented voter preferences create the possibility of councils in which no party secures an outright majority. In such circumstances, parties must negotiate agreements to elect municipal leaders, approve budgets and sustain administrations.
The eNCA analysis argues that a close election is not necessarily the same as a stable government. A party can finish first without having sufficient council support to govern alone. The resulting negotiations may determine whether a municipality can make decisions consistently and maintain effective oversight.
Coalitions are not inherently ineffective. Parties with different platforms can cooperate around shared priorities, provided they establish clear responsibilities and workable processes for resolving disagreements. Their success depends partly on whether agreements set out common objectives and allow the public to assess progress.
Problems arise when political disagreements repeatedly prevent decisions or when governing arrangements change without clear explanations to residents. Uncertainty can complicate budget approvals, planning and the continuity of municipal programmes.
Voters therefore have reason to examine not only campaign promises but also parties’ approaches to coalition governance. Their public commitments on financial controls, service delivery and accountability may be as important as their stated electoral ambitions.
Polling differences require careful interpretation
The growing prominence of opinion polls has also raised questions about how survey findings are collected and communicated. Different research organisations may interview different populations, use different fieldwork dates, apply different turnout models or ask questions in different ways.
The Common Sense and SRF reported that their second-quarter survey interviewed 2,214 registered voters between 8 and 31 July 2026 using telephone interviews. Its national results carried a reported margin of error of 2.1 percentage points at a 95% confidence level. Metro samples were smaller and carried a higher reported margin of error.
These methodological details matter because a poll is an estimate, not a count of ballots. The margin of error also does not account for every possible source of error, including non-response, question wording or differences between stated intentions and actual turnout.
Comparisons between polls should therefore consider dates, sample sizes, respondent selection, weighting and the precise question asked. Results measuring party favourability should not be confused with results measuring voting intentions or satisfaction with government.
Readers should also distinguish a pollster’s published findings from a political party’s interpretation of those findings. Parties naturally emphasise results that support their campaign messages, but independent scrutiny requires looking at the underlying data and methodology.
Voter participation remains an important uncertainty
Even accurate measurements of voter preferences cannot guarantee an election outcome if turnout differs from expectations. Citizens may express interest in politics without voting, while others may make their final decision close to election day.
Daily Maverick’s September reporting noted limited enthusiasm for politics among some respondents in the polling it examined and emphasised the importance of turnout assumptions when interpreting projections.
Participation can be influenced by whether voters believe their choices matter, whether candidates address local concerns and whether people can access the voting process. Political parties and electoral authorities consequently have an interest in ensuring that voters understand the process and can participate freely.
Low participation can also complicate claims that a particular party has received a broad public mandate. Election results establish the distribution of votes cast, but they do not mean every eligible citizen supported the winning party or its entire programme.
For this reason, participation figures and final results will be essential context when assessing the election’s political implications.
Economic and social concerns shape political attitudes
Public opinion cannot be separated from the conditions in which people live. Employment opportunities, household expenses, housing, infrastructure and access to reliable public services can influence how citizens assess political leadership.
The Common Sense listed another October polling report examining whether voters placed jobs among their leading priorities. This reinforces the importance of distinguishing between political messaging and the issues respondents say matter to them.
However, it would be premature to attribute every change in party support to a single economic or social cause without evidence directly establishing that relationship. Voters can evaluate parties through multiple experiences, and similar economic circumstances may produce different political responses.
Parties seeking public confidence must explain how their policies would address identified problems, how proposed measures would be funded and what results residents could reasonably expect. Specific commitments and transparent reporting provide voters with more useful information than broad promises alone.
Media literacy is essential as election day approaches
As campaigns intensify, the circulation of polls, political commentary and campaign claims is likely to increase. Citizens can evaluate these materials more effectively by checking the original publication date, identifying who commissioned the research and reading the methodology rather than relying exclusively on headlines.
It is also important to distinguish survey findings from interpretations. A measured decline in party favourability is a research result; the explanation for that decline is a separate question requiring additional evidence.
Similarly, an estimated lead does not establish that an election has been won, particularly when the difference between parties is small or the research carries a substantial margin of error. Responsible reporting should make these limitations clear.
Access to reliable information allows voters to scrutinise competing claims without accepting any party’s preferred interpretation automatically. It also helps prevent public debate from becoming dominated by misleading comparisons or unsupported conclusions.
Conclusion: Public confidence will be tested through accountability
South Africa’s latest available public-opinion research presents a political landscape marked by changing party assessments, divided voter preferences and uncertainty about the composition of future municipal councils. The findings on the GNU and individual parties suggest that citizens may evaluate coalition performance differently from their opinions of its constituent organisations. Meanwhile, metro polling demonstrates how closely contested elections and concerns about municipal management intersect.
These developments do not establish a predetermined election outcome. Rather, they identify questions that voters, political parties and journalists should examine as November approaches: whether municipal leaders can deliver services, whether coalitions can govern consistently, whether public finances are managed transparently and whether elected representatives remain accountable to residents.
Ultimately, the significance of public opinion lies not simply in which party leads a poll, but in what citizens expect from those seeking office and how they judge performance after election day. The forthcoming local government elections will provide an opportunity for voters to express their preferences, while the subsequent work of municipal councils will test whether political commitments can be translated into practical results.
