Dynamic Super Efficiency DEA Model And Its Application In Energy Efficiency

The Data Envelopment Analysis(DEA)method is an analysis method that is gradually developed based on the idea of "relative efficiency evaluation".Traditional DEA models not only fail to handle negative data and rank decision units(DMUs)completely,but also observe the production process of DMUs from a static perspective and ignore the impact of intermediate variables on efficiency during two consecutive periods.The integrated SBM super-efficiency model proposed based on the Slacks-based Measure(SBM)model can provide the complete rankings of DMUs,and the model is non-radial,which can measure the superefficiency of DMUs in both the inputs and outputs.However,this model may be infeasible when dealing with negative and zero data,and cannot dynamically evaluate the efficiencies of DMUs.Based on the definition of this integrated model,this paper has carried out a series of researches on dynamic super-efficiency models that can handle negative data.The main contributions are as follows:First,by introducing a range vector,a new integrated additive SBM super-efficiency model capable of processing negative data is proposed.Secondly,the model is extended into the dynamic structure,and the intermediate variables are used to connect two consecutive periods to build a dynamic super-efficiency model,which allows the model not only to completely rank DMUs,but also to handle negative and zero data.Finally,the efficiency scores are classified according to whether the weight coefficient is given in advance and the factors affecting the increase of free intermediate variables.Numerical examples show that all the new models proposed in this paper can deal with negative data,and also can completely rank the DMU.In order to highlight the practicability of the new model and evaluate China's inter-provincial energy efficiency from a dynamic perspective,this article uses coal consumption,capital,and labor as inputs,and environmental pollution emissions and economic growth as undesired outputs and outputs,respectively.Capital formation as a free intermediate variable,the proposed dynamic super-efficiency model is used to measure the energy efficiency of 30 provinces in China,and a comparative analysis is made of the energy efficiencies of the three major regions of China in order to give suggestions for improving energy efficiency.The research in this paper combines dynamic structure and super-efficiency models for the first time.It not only can dynamically evaluate energy efficiency,but also can handle the efficiency evaluation of other industries containing negative and zero data,which has broadened the application scope of existing superefficiency models.