Energy consumption and output: Evidence from a panel of 14 oil-exporting countries

Abstract

We examine the long-run relation and short-run dynamics between energy consumption and output in a panel of 14 oil-exporting countries over 1980–2007. Panel unit root tests, which account for common cross-sectional factors, fail to reject non-stationarity in both variables. Thus, we explore their long-run relation and short-run dynamics using three alternative panel estimation techniques — dynamic fixed effect, pooled and mean-group estimators before and after accounting for common cross-sectional factors. These estimators allow for various degrees of heterogeneity in long-run parameters and short-run dynamics. The results based on the mean group estimator with common correlated effects suggest (a) a stable relation between energy consumption and output; (b) bi-directional causality in both long- and short-run; and (c) the robustness of the long-run causality results to the inclusion of additional variables. As such, environmental policies designed to curtail energy may have significant long-run ramifications for economic growth, and policies designed to promote economic growth may have adverse environmental consequences.