Investigation of Macroeconomic Variable on Unemployment Rate in West African Countries using Panel Modelling Techniques in the Presence of Collinearity

Abstract

This study investigates the relationship between unemployment, inflation, and population growth in a panel of 15 West African countries using panel regression models. This consistency reflects the general trend of population growth in West African countries over the period (1990–2023). While the dataset comprises 450 observations spanning 33 periods. Descriptive statistics reveal a mean unemployment rate of 1.44, with inflation and population growth exhibiting minimal correlation with unemployment. Regression analysis indicates a weak explanatory power (R² = 0.0164), with inflation showing an insignificant positive effect on unemployment, while population has a statistically significant negative effect. Diagnostic tests reveal heteroskedasticity and omitted variable bias, suggesting potential model specification issues. Fixed-effects and random-effects models yield similar results, with the Hausman test favoring the random-effects model (p = 0.7249). The findings imply that macroeconomic variables alone may not sufficiently explain unemployment trends in the region, necessitating further exploration of structural and policy-related factors.