84% Say No to Vote-Buying, But 62% Went Hungry Last Month: What Osun's Own Survey Reveals
Ask an Osun voter on the eve of Saturday's governorship election whether they would sell their vote, and the overwhelming answer — 84.39% of them, according to a new survey — is no.
Ask the same household whether it had enough to eat last month, and more than six in ten say they did not.
Those two numbers come from the same document: the 2026 Osun Elections Survey, a joint project of Iterative Learning for Development (ILD) and the Kimpact Development Initiative (KDI), with fieldwork completed by July 31 and results published on August 14 — the day before Osun went to the polls.
Surveys of this kind are not a routine feature of Nigerian state elections. That a dataset like this exists at all, days before Osun voted, gives readers a rare chance to check the mood against the numbers rather than against a politician's claim of momentum.
A survey built for a tense Saturday
The topline finding was reassuring on its face. 84.39% of registered-voter respondents told ILD and KDI's researchers they would not accept money in exchange for their vote, per the survey as reported by The Nation.
That is a strong, specific number, not a vague mood reading. It suggests that whatever inducements were on offer around Osun's polling units this week, most voters who spoke to the pollsters said, on record, that they would not be bought.
It is also a rare thing in Nigerian election coverage: a systematic, dataset-based attempt to measure what voters say they will do, rather than relying only on anecdote from the campaign trail or from politicians' own claims.
Principle versus the pantry
The same survey, though, found that 62.53% of respondent households reported not having had enough food in the previous month.
Sit those two figures side by side and a harder question appears. Rejecting vote-buying in a survey response is one thing. Doing so at a polling unit, when a household has been going hungry, is another — and the survey does not pretend the two pressures cancel out.
It is exactly the kind of gap that matters once the ballots are actually being counted: whether a stated principle survives contact with an empty stomach at a polling unit, when someone is waving cash at the gate.
84.39% of Osun voters surveyed said they would not sell their vote. 62.53% of their households said they had not had enough food the month before the election.
High turnout intent, real fear
The same respondents were remarkably ready to vote. 93.22% of PVC-holding respondents said they intended to cast a ballot.
But readiness was not the same as ease. 27.56% of respondents said they were fearful or worried about the election itself — more than one in four.
Trust in the institutions meant to referee the process had also slipped. Measured against the same pollsters' earlier survey round in Ekiti State, confidence in INEC fell from 80.69% to 71.08%, and confidence in the police fell further still, from 80.52% to 61.37%.
A ten-point drop in confidence toward the police, and a nine-point drop toward INEC, are not small movements for an election survey to record in the space of one electoral cycle. Both institutions are the ones voters are being asked to trust with the physical safety of their ballot and their person on election day itself.
None of this means the election itself was compromised. It means that a meaningful share of the people who turned out on Saturday did so while trusting the referees a little less than the survey's own baseline a year ago.
What the model says the finish line looks like
Beyond attitudes, the survey's projection model made a specific numerical call: it expected the eventual winner to clear at least 55% of the vote, with the runner-up held below 35%.
That would be a wider margin than the race that produced the last Osun government. The 2022 election was decided by a much tighter split, roughly 52% to 48%.
A projection is not a result, and the survey's model is not a count of actual ballots. But if the pollsters' numbers hold, whatever margin INEC eventually announces will say something about whether this year's contest was genuinely as close as the campaign trail suggested.
A cleaner, wider margin also carries a practical implication: the closer a race, the more each contested unit matters to the final tally. If the model's prediction holds, the arithmetic itself may leave less room this cycle for disputes over narrow, close-fought results.
A wide margin and a narrow one are not just different numbers — they change how contestable the outcome feels to the people who lose. Whether Saturday's actual result mirrors the survey's wide-margin prediction, or lands closer to 2022's tighter finish, will shape how the announcement itself is received.
The threat that wasn't in the survey
Not every pressure on Osun voters showed up in a polling questionnaire. Separately, on August 13, Sahara Reporters reported that a woman identified as "Arugba" had threatened, in a viral clip, to have voters' fingers cut if they did not vote for the APC.
It is a single, viral claim, reported by one outlet — not a police statement, and not part of the structured survey data above. But it is exactly the kind of concrete moment that a bare percentage like "27.56% fearful" is trying to capture in numeric form.
Fear, unlike a vote-buying offer, rarely shows up in tidy percentages until researchers go looking for it. That the survey's own number — more than one in four respondents fearful or worried — sits so close to a real, reported threat is the clearest evidence in this dataset that the worry was not abstract.
Taken together, the survey and the Arugba report describe the same Saturday from two different angles — one in percentages, one in a single viral clip — but they point at the same underlying picture: an electorate voting under more strain than a simple turnout figure would ever show.
Numbers on a spreadsheet said most of Osun would not sell its vote. What the same numbers do not settle is how many of those voters cast their ballot hungry, anxious, or with a stranger's threat still ringing in their ears.
