Cost efficiency analysis based on the DEA and StoNED models: Case of Norwegian electricity distribution companies

Abstract

Our paper applies data envelopment analysis (DEA) and stochastic non-parametric envelopment of data (StoNED) to measure cost efficiency of electricity distribution companies. The data cover 123 Norwegian electricity distribution companies during 2004-2010, and the performance of these companies is compared across the two models with and without environmental variables, i.e., variables that account for local conditions that affect the companies' costs. The results indicate that the cost efficiency estimates with the StoNED approach are much higher than with the DEA method when we do not consider environmental variables. It shows that the choice of estimation methods is important with respect to the estimated impact of environmental variables on the performance. In addition, the inclusion of the environmental variables has considerable effect on the classification of companies with respect to local returns to scale.