How Nigeria’s 2015 presidential election outcome was forecasted with geodemographics and public sentiment analytics

Abstract

In 2015, Nigeria held one of the most fiercely contested presidential elections in the nation’s recent democratic history. The outcome of the election was expected to exert significant influence on democratic practices on the African continent. The stiffness of the contest also meant that it was difficult to predict the likely winner of the election. This paper summarizes how an empirical approach was used to forecast the outcome of the election by modeling public sentiment data-set using a geodemographic framework. Results indicate that the main electorates that determined the outcome of the election were situated in thirteen battleground states. Additionally, results showed that two years before the presidential election, Goodluck Jonathan’s public approval ratings on corruption, insecurity, and the economy (the main drivers of the 2015 election) had nose-dived across many of the battleground states. This eventually contributed toward his loss.

Keywords:

• Geodemographics
• Nigeria
• public sentiment analysis
• presidential election
• geographic information science
• forecasting