Considering nonlinearity and structural changes in the convergence of clean energy consumption: the case of OECD countries

Purpose In this study, we aim to test the stochastic convergence of per capita clean energy use in 30 OECD (Organization for Economic Co-operation and Development) countries for the period of 1965-2017. Design/methodology/approach This study employed both linear and nonlinear panel unit root tests, and unlike other studies, this study allowed fractional values in addition to integer values for frequencies in the Fourier functions. Integer values of frequency indicate temporary breaks, while fractional values show permanent breaks. Findings The results of the linear panel unit root test indicate that clean energy use does not converge to group average for almost all OECD countries. However, the results of nonlinear panel unit root tests provide evidence that the stochastic convergence hypothesis of clean energy consumption cannot be rejected for most countries. This study does not find any evidence for stochastic convergence of clean energy use in Australia, Canada, Denmark, Ireland, Norway or Sweden. Therefore, the policies regarding clean energy are mandatory in these countries due to their effectiveness. This study also reveals that there are permanent structural breaks in the convergence process of clean energy consumption in approximately half of OECD countries. Originality/value This study considers temporary and permanent smooth structural shifts in addition to nonlinearity when testing the stationarity of clean energy consumption in a countryirelative to the group average. This new method eliminates deficiencies of the previous panel data techniques. Thus, it provides more reliable results compared to existing literature.