DUAL METHODOLOGICAL APPROACH OF RANDOM FOREST REGRESSOR AND ORDINARY LEAST SQUARES IN ASSESSING THE IMPACT OF LABOUR FORCE ON GROSS DOMESTIC PRODUCT (GDP) IN NIGERIA.

Abstract

This study examines the impact of total labour force on Gross Domestic Product (GDP) in
Nigeria using a comparative methodological framework that integrates Ordinary Least Squares
(OLS) regression and Random Forest regression. While classical economic theory assumes a
linear relationship between labour and output, emerging data science techniques suggest that
macroeconomic relationships may exhibit nonlinear dynamics. The objective of this study is to
determine whether the labour to GDP relationship in Nigeria is predominantly linear or
potentially nonlinear. Annual time series data were analyzed using OLS to estimate the linear
effect of labour force on GDP, followed by Random Forest regression to capture possible
nonlinear structures and improve predictive performance. The OLS results reveal a positive and
statistically significant relationship between labour force and GDP, explaining approximately
79.5% of the variation in economic output. However, the Random Forest model outperformed
OLS, explaining 87.3% of GDP variation and significantly reducing prediction errors (RMSE,
MSE, and MAE). Based on comparative model performance, the findings suggest that although
labour force significantly influences GDP, the relationship is not strictly linear but potentially
nonlinear and structurally complex. The study concludes that integrating machine learning
techniques with traditional econometric methods enhances macroeconomic analysis and policy
interpretation. Policy implications emphasize labour productivity, human capital development,
and data-driven economic planning. This research contributes to the growing literature bridging
applied statistics, machine learning, and economic development analysis in developing
economics.
Keywords: Labour force, GDP, Regression, OLS, Random forest, Machine learning