Increasing social concerns over the environmental externalities associated with business activities are pushing firms to identify activities that create economic value with less environmental impact and to become more eco-efficient. Over the past two decades, researchers have increasingly used frontier efficiency models to evaluate productive efficiency in the presence of undesirable outputs, such as greenhouse gas emissions. In this paper, we identify critical flaws of existing frontier models and show that under these models eco-inefficient firms can be identified as eco-efficient. We develop a new eco-inefficiency frontier model that rectifies these problems. Our model allows us to calculate, for each firm, an eco-inefficiency score and improvements in outputs necessary to attain eco-efficiency. We demonstrate, through a Monte-Carlo experiment that our eco-inefficiency model provides a more reliable measurement of corporate eco-inefficiency than the existing frontier models. In the simulation experiment we develop a production function of multiple desirable and undesirable outputs that extends the classical Cob-Douglas function of a single output. The multi-output production function allows for greater flexibility in the simulation analysis of frontier models.

Increasing social concerns over the environmental externalities associated with business activities are pushing firms to identify activities that create economic value with less environmental impact and to become more eco-efficient. However, this task has proven challenging because there is no systematic methodology to integrate undesirable outputs, such as emissions, in the calculation of economic or productive efficiency. In this paper, we develop a methodology based on the nonparametric frontier approach to measure corporate eco-efficiency, and to compare it to productive efficiency. Our eco-efficiency model rectifies several problems encountered in existing approaches. Our methodology allows us to calculate, for each firm, the reduction in emissions necessary to attain eco-efficiency. In addition, our methodology measures changes in efficiency attributed to undesirable outputs. We apply our methodology to data from 84 U.S. electric utilities in 2007. Our analysis demonstrates how incorporating undesirable outputs in the measurement of efficiency can impact the distance of the firm to the best industry practice. We describe future research directions and potential applications of the methodology for managers and policymakers.