• Monday, June 24, 2024
businessday logo


2023: The fetish and fallacies of election opinion polls in Nigeria (I)

2023 election, the electoral institutions and the rest of us

Polls provide the best direct source of information about public opinion. Opinion polls in Nigeria, as in other societies, help to define and understand electoral patterns and electoral landscapes that reveal holistic evidential data on political participation, ideological alignment of voters and trust in basic democratic values. Thus, opinion polls tend to offer vital logical insights into voting and voter preferences.

Nigeria’s 2023 presidential election has seen its fair share of polling surveys so far, including those from reputable international organisations that have been romanticised by supporters of favoured candidates and denounced by enthusiasts of other parties.

On September 28, 2022, CNN and Bloomberg reported on a presidential poll conducted by Premise, which reported, among other things, that: 72 percent of decided voters name Peter Obi as their preferred choice for president, followed by Bola Tinubu (16 percent) and Atiku Abubakar (9 percent). Obi also led among undecided voters (45 percent).

Another poll conducted by We2Geda Foundation puts Obi, ahead of other contestants, garnering 51 percent of registered voters’ support. ANAP Foundation also recently released the result of an opinion poll conducted for it by NOI Polls on the forthcoming presidential election. The poll showed Obi of the Labour Party (LP) holding a significant lead over his nearest rival.
BusinessDay, in September, reported that an Economist Intelligence Unit (EIU) report predicted that Tinubu, the presidential candidate of the ruling All Progressives Congress (APC) for the 2023 general election, would likely win the contest.

Although the media’s obsession with political forecasting has shifted to electoral prophecy, noted the Mumbai-based Economics and Political Weekly, “psephology continues to provide the best telescopic view of elections based on the feedback of citizens. The ascertainment of subaltern opinion by surveys not only broadens the contours of understanding electoral democracy, but also provides an empirical alternative to the elitist viewpoint of competitive politics” in Nigeria.

The current fetish with and idolisation of election opinion polls must be tempered with heavy doses of realism and knowledge. While polling survey is scientific, it is not an exact science, which does not require an understanding of human psychology. It is a social science; the study of human behaviour. And the main reason social science is difficult to execute rigorously is that it is rarely possible to engage in double-blind, controlled experiments, for both ethical and practical reasons. By their nature, therefore, social sciences are more descriptive than experimental and are based on the premise of ceteris paribus.

Traditional polling methods using random digit dial phone interviews, opt-in samples of online surveys and interactive voice response are failing to predict election outcomes across the world.

The victory of Donald Trump in the 2016 US presidential election came as a shock to many, as none of the pollsters and political journalists and pundits, including those in Trump’s campaign, could predict this victory.

Read also: 7 in 10 Nigerians proud of the country – Poll

Zhenkun Zhou et al, in an October 2021 research report in the Journal of Big Data, identified the reasons for the failure of polling surveys to predict election outcomes. Firstly, they noted, “The percentage of responses to traditionally conducted surveys has decreased and it is becoming increasingly difficult to get people’s opinion. Response rates in telephone polls with live interviewers continue to decline, and it has reached a 6 percent lower limit recently. Response rates could be even lower for other methodologies, like internet polling or interactive voice response.”

Compounded with declining response rates, they noted, “is the concomitant problem of misrepresentation of the survey samples. That is, the sample surveyed by pollsters does not represent the demographic distributions of the general population”.

This problem, they posit, “is ameliorated by re-weighting the surveys sample according to the general demographics of the population in a process called sample-balancing or raking”.

The researchers further found that, in countries where voting is not obligatory, re-weighting a sample to the general population demographics (obtained from Census Bureau) could fail since the general population does not match necessarily the demographics of the voter turnout. Thus, they warned, “it is important to predict which demographic groups will turn out at the voting station. For instance, if an underrepresented group in the polls, “the hidden vote”, decides to vote on election date, the re-weighting fails leading to inaccurate results”.

The issue, they assert, is believed to be one of the major reasons for the generalised failure of pollsters to predict the triumph of Trump in the 2016 US presidential election, where groups generally defined as “white voters without college degree” mostly voted for Trump but were under-sampled by pollsters. “Even with this historical information at hand, which supposedly allowed pollsters to resample their surveys more carefully, pollsters again under-predicted the support for Trump in the subsequent 2020 presidential election in some states or under-predicted the voter turnout supporting Biden in newly created battleground states. The inability to accurately predict the voter turnout to deal with sampling misrepresentation might render the pollsters obsolete”.