In the buildup to the 2023 general elections, many opinion polls were conducted, projecting the likely outcome of the presidential election. But the results indicated that the opinion polls were largely wrong in their predictions. In this analysis, experts have advanced reasons why pre-election polls in Nigeria don’t conform with outcomes.
Many opinion polls were conducted ahead of the February 25 presidential election in the country.
Polls are a survey of public opinion from a particular sample group, and can be useful in informing politicians about the views of specific groups of the electorate
They help politicians understand what the public wants and tell who is popular at the ballot box.
But how accurate were the opinion polls in respect to the outcome of the February 25 presidential election?
Although victory was largely predicted the way of Peter Obi of the Labour Party (LP), the poll was eventually won by the candidate of the All Progressives Congress (APC), Asiwaju Bola Tinubu after results from the 176,846 polling stations were counted by the Independent National Electoral Commission (INEC).
Tinubu scored 8,794,726 votes, representing 37 per cent of the votes cast, while the main opposition candidates Atiku Abubakar of the Peoples Democratic Party (PDP) had 6,984,520 votes (29%) and Labour Party’s Peter Obi polled 6,101,533 votes (25%).
Rabiu Musa Kwankwaso of the New Nigeria Peoples Party (NNPP) came fourth with 1,496,687 votes.
In a series of polls commissioned by ANAP Foundation and conducted by NOI polls Limited (NOIPolls), Obi was predicted as the winner. In the poll published in September and December 2022, as well as February 2022, Obi was placed in the lead.
“Our third and final poll result in February 2023 reveals that Peter Obi of the Labour Party (LP) remains in the lead, with Asiwaju Bola Ahmed Tinubu of the APC and Atiku Abubakar of the PDP both trailing him. Rabiu Kwankwaso of the NNPP emerged as the lone outsider.
“Peter Obi leads, with 21 per cent of registered voters proposing to vote for him if the presidential election were to be conducted today, and 13 per cent proposing to vote for Asiwaju Bola Ahmed Tinubu, who fell in second.
“Alhaji Atiku Abubakar was third with 10 per cent and Rabiu Kwankwaso was a distant fourth with 3 per cent of voters proposing to vote for him,” said Atedo Peterside, the president and founder of the foundation in the poll report.
ANAP, which credited itself to have used the same methodology that rightly predicted the winners of the 2011, 2015 and 2019 presidential elections, however, noted that given the large pool of undecided voters and/or those who refused to disclose their preferred choice, Peter Obi’s 8 per cent-point lead at this stage is significant but not sufficient to separate him from a leading pack of candidates scoring 13 per cent, 10 per cent and 3 per cent.
Also, Bloomberg News polling 2,384 Nigerians from January 26 to February 4 via a smartphone app noted that two-thirds of respondents said they intended to vote for Obi.
“Of the 93 per cent of participants who said they had decided how to vote, 66 per cent named Obi as their preferred choice. Obi scored a slightly higher 72 per cent among decided respondents in an earlier premise poll was released by Bloomberg in September as the official election campaign kicked off,” it stated.
In the same vein, Nextier SPD poll released on February 5 indicated that Obi was leading (37%) ahead of Atiku (27%) and Tinubu (24%).
On February 7, a Lagos-based data company, Stears, announced that Obi would win with 41 per cent of the votes ahead of Tinubu (31%) and Atiku (20%) “as long as Nigerians followed through on their stated intent to vote.”
The result of the We2Geda Foundation poll on September 17, 2022, showed Obi ahead of other contenders with 51 per cent of the votes, Atiku had 25% and Tinubu 19%.
According to a poll conducted by POLAF, Atiku was predicted to emerge the winner in a close race. According to the organisers, the poll had the largest sampling size of 3,123,660 respondents across 165 local government areas contacted via telephone and were spoken to in English and any other native language they were most comfortable speaking in.
In the POLAF poll, Atiku secured 38 per cent to emerge as the preferred candidate, followed by Tinubu, who got 29 per cent, while Obi secured 27 per cent to take the third position. Kwakwanso was a distant fourth with only 5 per cent of the total votes.
However, the Economist Intelligence Unit (EIU), the research and analysis division of the Economist Group, predicted a win for Tinubu, basing its outcomes on the political dynamics in the country.
The EIU anchored its prediction on the internal wrangling in the PDP and Obi’s emergence, which would eat into PDP’s strongholds in the South East and South-South.
Also, a four-week opinion poll coordinated by FREDDAN Continental predicted a win for Tinubu. It stated that 37.2 per cent of 287,033 respondents preferred Tinubu, followed by Obi’s 32.1 per cent, while Atiku was selected by 30.7 per cent of total respondents.
The poll also suggested a clear win for Tinubu on the first ballot as against some other predictions that there may be a rerun.
Also, Enough is Enough (EiE) in its poll, had predicted a second round, noting that no candidate would win at the first ballot.
The co-founder and head of intelligence, Stears, Michael Famoroti, noted that the company rightly predicted the outcome of the 2023 polls.
He noted that while the basic Stears poll put Obi ahead with 27 per cent of the votes compared to 15 per cent for Tinubu and 12 per cent for Atiku, a first-of-its-kind electoral prediction model correctly projected that a low turnout scenario (29%) would be a victory for the APC candidate.
“Separately, we predicted an Peter Obi victory in a high-turnout scenario (roughly 85% turnout) and maintained that prediction, even under a turnout scenario of 45 per cent.
“The Stears prediction model was also accurate at state level. The basic poll results put Peter Obi marginally ahead of Bola Ahmed Tinubu in Benue State. But, once the model was applied to the poll results, the order flipped as we projected that Bola Ahmed Tinubu would win Benue by a single percentage point, exactly as it played out.
“The Stears prediction model accurately projected a critical APC win in Kwara State and an LP upset in Lagos State,” he added.
Why pollsters got polling wrong – Experts
Noting that polls are important in understanding public opinions, some experts opined that certain factors limit the relevance of the exercise in Nigeria.
The vice chancellor of the Federal University, Kashere, Gombe State, Professor Umar Pate, who underscored the importance of survey in both public and private practices, noted that the methodology adopted in a poll was as important as the outcome.
The revered media scholar, who also said it was important to identify the pollsters, their objectives, capacities to conduct a poll and the instruments used among others, added that some of the organisations conducting polls in Nigeria were in cahoots with politicians, thereby denting the credibility of the exercise.
The executive director, Development Specs Academy, Prof Okey Ikechukwu, on his part, noted that some of the polls lacked merit and were mere projections because they were built on faulty methodology and small sample size.
He also argued that there were some external factors beyond the control of the pollsters, which might make the outcome of the election to be different from the polls. He said such factors included intimidation, violence, suppression, manipulation and other hitches witnessed during the conduct of an election.
“Polls are relevant where several factors can be guaranteed, including the right sample size, right stratification of the demographics (their age, education, biases etc). Most importantly is the credibility of the pollsters,” the public communication and media professional noted.
An associate professor of Journalism and pioneer director of the Digital Media Research Centre (DMRC) of the Lagos State University, Dr Tunde Akanni, stressed that “in most cases, the pollsters are not independent, which is why the codes of conduct prepared for journalists covering elections highly recommend that in reporting the outcome of electoral polls, they must also report the profile of the organisers of polls so that their audience would know the biases or interest of the pollsters.
While noting that well conducted polls have the capacity to guide the public in making choices and empowering politicians to know where and how to intensify their campaigns, he counselled media consumers not to take results of polls “hook, line and sinker because of the biases of pollsters who might be commissioned by desperate politicians seeking to boost their profile ahead of an electron.”
The president of the Nigeria Political Science Association, Professor Hassan Salihu, said opinion polls were losing relevance in Nigeria because pollsters were biased.
“This is why you see that in the opinion polls conducted, people were just off the mark because they left substance and allowed sentiment to guide what they did. But like I said, where it is well conducted, it is a good measure of what should happen and what is happening. So it should be encouraged, but in the Nigerian context, it has been bastardised. People have inputted all kinds of sentiments into it, and that is why you see a wide gap between what has been predicted and what later happened,” he said.