Keywords: performance evaluation, three–stage DEA model, SFA, stochastic frontier analysis, environmental effects, regional energy efficiency, China, data envelopment analysis, modelling, environmental variables
China's regional energy efficiency: results based on three–stage DEA model
Traditional data envelopment analysis (DEA) models ignore the influence of environmental variables and statistical noise, which may result in biased efficiency estimates. To solve this problem, three–stage DEA models have been proposed and widely applied in many areas. This study evaluates China's regional energy efficiency by using a three–stage DEA model based on the statistical data of 2010 and discusses the divergence of three different efficiency assessment methods. The empirical results show that environmental factors indeed influence the regional energy efficiency performance. After adjusting the environmental variables, the national energy efficiency average estimated by the three–stage DEA model decreased significantly relative to the estimate of the traditional DEA model, but the environmental influence in different regions varies due to diverse features. Some regional averages were overestimated by using the traditional DEA model, while some regional averages were underestimated. The three–stage DEA model is able to reflect the true efficiency by eliminating environmental effects compared with other methods.