Methodological comparison among radial, non-radial and intermediate approaches for DEA environmental assessment

This study compares three DEA approaches for environmental assessment, all of which are designed to examine the level of simultaneous achievement on economic prosperity and environmental protection, so measuring the degree of sustainability development. DEA, standing for Data Envelopment Analysis, has been widely applied for performance assessment in the past five decades. A new type of applications is referred to as "DEA environmental assessment" and it measures the performance of many organizations that utilize inputs to produce not only desirable outputs (e.g., electricity) but also undesirable outputs (e.g., CO2). In the previous studies, DEA-based performance evaluation for environment assessment is methodologically classified into radial or non-radial category. Recently, a new "intermediate" approach, analytically locating between the radial and non-radial measures, has been proposed as the third alternative. A use of the intermediate approach has several unique features, all of which cannot be found in the radial and non-radial ones. The new approach measures the degree of unified inefficiency on each production factor and determines the level of total unified inefficiency from the average of the sum of these inefficiency scores. This study discusses the analytical features by comparing the intermediate approach with the radial and non-radial ones. The methodological comparison attempts to convey a message that DEA is indeed an important methodology, but not perfect. Rather, it is an approximation approach to examine the performance of various organizations. Many DEA applications on energy and environment often suffer from a methodological bias, implying that different approaches produce different empirical results. Thus, in guiding a large policy issue such as the global warming and climate change, it is necessary for us to compare several different approaches (e.g., models and concepts) to derive a reliable empirical suggestion. The importance of such a message is applicable to not only DEA but also the other types of empirical research in natural and social sciences. Therefore, this study discusses the methodological bias issue from the practicality of the three DEA approaches in assessing various concerns on energy and environment. (C) 2017 Elsevier B.V. All rights reserved.