Returns to scale and damages to scale on U.S. fossil fuel power plants: Radial and non-radial approaches for DEA environmental assessment

Abstract

This study proposes a new use of Data Envelopment Analysis (DEA) for environmental assessment, paying attention to a methodological bias (i.e., different methods produce different results). DEA is analytically classified into radial and non-radial approaches. The radial approach determines the level of unified (operational and environmental) efficiency by measuring an inefficiency score. The score indicates the distance of a projection from an observed performance to an efficiency frontier. In contrast, the non-radial approach determines the level of unified efficiency by measuring a total amount of slacks because each slack indicates inefficiency. To avoid the methodological bias, this study considers the use of radial and non-radial approaches. In this study, we consider a production process where all organizations produce not only desirable (good) but also undesirable (bad) outputs as a result of their business operations. To unify the two types of outputs, this study discusses the concept of disposability, which is separated into natural and managerial disposability. The natural disposability indicates negative adaptation to a regulation change on undesirable outputs. In contrast, the managerial disposability indicates corporate strategy by which a firm considers the regulation change as a new business opportunity. A firm attempts to improve its unified efficiency by utilizing new environmental technology and/or new management. This type of strategy indicates positive adaptation to the regulation change. Under the two disposability concepts, this study discusses how to measure RTS (Returns to Scale) under natural disposability and DTS (Damages to Scale: corresponding to RTS on undesirable outputs) under managerial disposability. The two scale measures are analytically discussed by the proposed radial and non-radial approaches with SCSCs or without SCSCs, where SCSCs indicate strong complementary slackness conditions. An illustrative example on U.S. fossil fuel power plants indicates a policy implication that they need to introduce new technology for environmental protection. This study also discusses a necessity of examining a methodological bias in energy studies. Empirical findings identified in this study document the practicality of the proposed approaches to measure RTS/DTS.