The study seeks to model economic growth in
Africa using the Bayesian Model Averaging technique.
The dataset for the study is a pooled dataset spanning
through 2010 to 2021 yearly for seven variables namely:
Economic growth (GDP), inflation, unemployment,
government consumption, food production,exchange rate
and trade openness in 24 Africa economies.
The Bayesian Model Averaging technique is adopted
having the capacity to extract the posterior inclusion
probability under different model prior associated with g-
priors attributed to different Bayesian Model Sampling
Findings from the study shows that, in modelling
economic growth in Africa, a BMA with Uniform model
prior with EBL gprior is most plausible while on the basis
of sub-regions, the most plausible BMA model is Uniform
model prior with Hyper gprior having accounted for the
highest Posterior Model Probability correlation value.
Bayesian models with other variants of g-prior
should be explored for better detection of the true
determinants of economic growth among feasible
identified factors under consideration.
Keywords : Economic growth, g-prior, Posterior Model Prior, Posterior Inclusion Probability, BMA.