Statistical Modeling of Electoral Outcomes: Assessing the Impact of Socioeconomic and Demographic Variables on Voting Behavior

Abstract

The present study explores the socioeconomic and demographic forces behind the voting choice and election outcomes in Nigeria's 2015, 2019, and 2023 presidential elections based on a mixed-methods approach informed by quantitative studies. The considered electoral results and survey responses from 1,200 respondents drawn from the six  geopolitical zones in Nigeria as well as the engagement of  some statistics including  Ordinary Least Squares (OLS), Multinomial Logistic Regression, and Multilevel Regression with Post-Stratification (MRP) models. Findings show that poverty level, youth population, education level, ethnic identity, and election violence have a significant effect on voting behavior. The average national voter turnout in the three elections was 36.5%, with the lowest recorded in 2023 at 27.1%. Statistical analysis shows that a 1% rise in the poverty rate corresponds to a fall in turnout of 0.42% (p < 0.05), while a 1% rise in youth support boosts turnout by 0.38% (p < 0.01). Ethnic identification remains the strongest predictor of party allegiance, with an ethnic identification Voting Intention correlation of 0.71. The 2023 statistics indicate growing digital engagement, with 64% of the young respondents naming social media as their primary source of political mobilization. Prevailing trend in voters turn out seems to favours higher digital and civil participation and awareness in the south while  lower turnout in the north is probable due to poverty and insecurity.  The study discovers that economic disparity, demographic composition, and identity politics continue to shape Nigeria's democracy, with youth digital mobilization emerging as a transforming force.