Disaggregated Energy Use and GDP Growth Nexus Revisited: A Comparison Between Thoughtful and Naïve Bayesian Estimations

To achieve the higher degree of preciseness in statistical modeling, Bayesian estimation is considered a viable alternative to frequentist estimation. In Bayesian small sample research, the more information is added to a correctly-specified prior distribution, the more accurate the posterior result becomes. But, it is argued that in Bayesian modeling based on large sample sizes, with more data included, the preciseness of posterior estimates increases. To prove this argument, we conduct a case study on the relationship between disaggregated energy consumption and GDP growth in selected developed countries. Using the standard deviations and Monte Carlo standard errors, we obtained the conclusion that thoughtful Bayesian estimation outperforms frequentist and even naïve Bayesian estimation. We strongly recommend using a large sample even in thoughtful Bayesian estimation.