A re-examination of stationarity of energy consumption: Evidence from new unit root tests

In this paper, we re-examine the stochastic behaviour of per capita total primary energy consumption across 178 countries around the world. For this purpose, in addition to conventional unit root tests, we apply newly developed tests that allow for nonlinear adjustments and structural breaks in the data generating process. The results of the unit root tests show that allowing for structural breaks and nonlinearity result in more frequent rejection of the null hypothesis of unit root, suggesting that most of the series under consideration follow a stationary process. These findings imply that both energy economists and policy makers must be careful and take account of possible nonlinearities and structural breaks in their analyses of energy consumption. In addition, the results imply that government interventions into the energy markets in most countries have no effect in the long run, although the short-run effects of the interventions depend on the size and the direction of interventions.

► Linear models are not suitable to examine the dynamics of energy consumption.
► Nonlinear econometric models seem to be more appropriate to model energy consumption.
► Shocks have transitory effects on energy consumption in most of the countries.
► The effects of shocks to energy consumption depend on the sign and size of the shocks.