Energy intensity and its components in Iran: Determinants and trends

Iran energy intensity is among the highest in the world, which has been increasing during the recent decades, originating from the channels related to the efficiency and structural (scale) changes. The present study aimed to investigate the changes in the aggregate energy intensity (EI) and its components including the Efficiency (EE) and Structural Change (SC) indices as well as their driving forces based on a regression analysis. Furthermore, the multilayer perceptron and wavelet-based neural networks (WNN) were proposed to evaluate the ability of regression models in forecasting energy intensity and its components. The results suggested a non-linear relationship between the energy intensity indices and income as well as the capital-output ratio. However, no significant role was observed for trade and energy price index. Further, the turning points measured for income and the capital-labor ratio indicated that the effect of income on energy intensity is increasing while the capital-energy is turning out to be as the substitutes. Furthermore, urbanization could significantly reduce energy intensity. The forecast results indicated that energy intensity and its components may be predicted with a prediction error of less than 0.35%. Finally, EE and SC can be predicted more accurately based on the ANN-based models while the regression models can predict the EI index more precisely.