Politics

Which poll comes closest to Johannesburg's election result?

Three published polls give Johannesburg numbers, and they disagree. Ipsos (interviews in June and July, published 4 September) had the ANC on 31% and the DA on 28%. The Social Research Foundation (fieldwork 18 September to 12 October) had the DA on 36% and the ANC on 22%. eNCA and MarkData (interviews in August and September, published 17 October) had the DA on 35.0% and the ANC on 21.8%. On 4 November, Joburg's voters decide which poll got it right.

1 person has guessedClosed 3 November 2026

This market has closed. Here is where the country landed.

What the world believes

Ipsos100%
Social Research Foundation0%
eNCA / MarkData0%

How this resolves

Closes: 23:59 SAST on 3 November 2026
Resolves: When the IEC publishes the City of Johannesburg's results, usually within a week of 4 November.
Source: Electoral Commission of South Africa (IEC) official results
Rules: For each poll, add two gaps: the difference between its ANC figure and the ANC's actual share, and the difference between its DA figure and the DA's actual share. The poll with the smallest total wins. Actual shares are each party's percentage of valid proportional representation (PR) ballot votes in the City of Johannesburg, as published by the IEC. The poll figures are fixed as follows: Ipsos, ANC 31% and DA 28%. Social Research Foundation (fieldwork 18 September to 12 October 2026), ANC 22% and DA 36%. eNCA and MarkData, ANC 21.8% and DA 35.0%. Later or earlier polls by the same organisations do not count. If two polls have exactly the same total, the one with the smaller ANC gap wins.
If the source is down: If the IEC results site is unavailable, use the IEC's official Johannesburg PR results as reported by at least two reputable South African news organisations.
If plans change: If the election is postponed, the market waits for the rescheduled vote. Void if no Johannesburg PR result is published by 31 March 2027.

Report a problem with this market