DEA-based Research On Indicator Selection And Environmental Performance Measurement

As a useful tool of performance evaluation, data envelopment analysis (DEA) has gained impressive growth both in theoretical development and applications. Since DEA analysis is based on the data of evaluation indicator, selecting appropriate input and output variables is one of the crucial considerations that the managers must keep in mind. Unfortunately, there has been inadequate attention to this issue, and many problems exist in present methods for selecting variables, such as the choice means, correlation analysis on variables, and the classifications of input versus output status. A better way which can overcome these problems is highly necessary.Environmental issues along with the rapid development of economy are of global importance and environmental efficiency evaluation has attracted more and more research interests in recent years. However, the traditional DEA assumptions on efficiency and efficiency improvement can’t be applied when we consider the undesirable outputs, such as CO2and SO2. Thus, one should consider the special properties of the undesirable outputs and modify the existing models to evaluate the environmental efficiency accurately.According to these two points, this paper studies indicator selection and environmental performance measurement when undesirable outputs involved based on DEA theory.This paper involves eight chapters, and the essence is summarized as follows.Chapter1firstly summarizes some important conceptions, the basic models and applications of DEA. Then it introduces the concept of environmental efficiency and many different techniques for measuring environmental performance. Finally, the contents and importance of this paper are pointed out.Chapter2studies inputs and outputs selection for efficiency measurement. A new approach based on Principal Component Analysis (PCA) on efficiency improvement is proposed in order to improve the disadvantage of the existing method. Through this proposed method, the performance of non-efficient decision making units can be improved by projecting input and output vectors to the efficient frontier. Next, based on the conception of cash value added, this chapter proposes another method for choosing DEA variables, with which the uncertainty of the subjective selection method can be avoided. The new method takes advantage of financial statements and selects those items which have important influence on cash flow as evaluation indicators. The results of DEA analysis can help managers to create more value for shareholders. The application shows that the new method can make decisions as to which variables to include as inputs and which ones as outputs at the same time, illustrates its usefulness and rationality.Chapter3discusses the disposal of undesirable outputs, the construction of new production possibility set and some environmental efficiency DEA models. Fare et al.(1989) firstly adopted DEA approach to research the environmental efficiency measurement. Since then, various methods for dealing with undesirable outputs have been discussed in the framework of DEA, and are applied in a variety of fields, including paper mill, airport, power plants, and agriculture etc.Chapter4proposes a slacks-based environmental efficiency index based on DEA. Different from those former methods, new model is a non-oriented approach, aiming at maximizing the slacks of all inputs and outputs, and thus providing an aggregated index on production process within the context of environmental protection. An empirical application to industry sector of China presents that about two-thirds of provinces of China are inefficient. Moreover, Spearman test demonstrates that SO2and solid waste have more impacts on industrial aggregated efficiency and should given more attention in practice. Sensitivity analysis further shows the robustness of the proposed model in terms of ranking.Chapter5builds a slacks-based but different from SBM DEA model to measure environmental efficiency, in which a special variable, namely non-separable variable, is involved. A numerical example of power industry in30Chinese provinces illustrates the proposed approach. The results show that any change in non-separable input involves both desirable and undesirable outputs, and the environmental protection in power industry can be further improved.Chapter6takes the environmental capacity into consideration to explore the effect of emission quota on environmental performance. Compare with the model in Chapter5, the relationship between the whole efficiency and the efficiency of each decision making unit can be ontained. The empirical research on the environmental efficiency of power industry shows that the current emission quota in our country is far more than the ideal value and the environmental protection in power industry can be further improved, which are also helpful to policy making.Chapter7firstly introduces some means on environmental treatment, illustrates the efforts of our country on environmental protection. Then, it analysis the environmental efficiency changes in the past ten years, by combining the macro-level policy with micro-level efficiency. The results not only verify the China’s achievements in environmental protection but also validate the mutual promoting relationship between environmental treatment and economic development.Chapter8summarizes the main research work in this paper, and points out some possible directions for future study and improvement.