|
Abstract
|
Background: The key role of energy as one of the important production factors in the economic growth and development of countries, the limitation of energy sources and their ending, as well as issues related to environmental pollution, highlights the issue of identifying and studying factors affecting the intensity of energy. Past studies in this regard have been made without regard to the model uncertainty problem.
Aim: The aim of this research is to identification and investigation of effective factors of energy intensity in Iran Provinces under the assumption of model uncertainty.
Methodology: This study has been done on 30 Iranian provinces over the period 2008-2014. A Bayesian non-standard econometric method (Bayesian averaging and Weighted averaging least squares methods) and information criterions have been used to select the model. The estimation of the selected model is also done using Bayesian standard econometric method (Bayesian panel method).
Finding: By estimating of over 8 million regressions and Bayesian averaging of the coefficients, among the 24 variables affecting the energy intensity (based on theoretical bases and e export production ratio mpirical studies), 8 variables of the share of the services sector from production, the share of oil and oil products from energy consumption, the export production ratio, per capita income, the capital labor ratio, energy price, population growth rate and number of warm months of the year were identified as the most important factors affecting energy intensity in Iran provinces, and other variables were ranked based on their impact on energy intensity. Based on Bayesian Information Criterion, a model including the best subset of the most important variables was selected as the optimal model.
Conclusion: The results of estimating the optimal model using the Bayesian panel method and using the Monte Carlo simulation showed that the export production ratio, population growth rate and the number of warm months of
|