Applications of the stochastic frontier approach in Energy Economics

Abstract

In this paper we discuss some of the issues that energy researchers and regulators could address in two different areas of the application of stochastic frontier analysis (SFA) in Energy Economics: i) the estimation of productive or cost efficiency of electricity and gas distribution networks, and ii) the estimation of energy demand frontier models. Our review examines common problems that often appear using standard frontier models, such as the selection of input (cost) and output variables, the integration of quality incentives in cost benchmarking, the impact of distributed energy resources, or the effect of both unobserved heterogeneity and observed (e.g. environmental, socioeconomic, etc.) variables on firms’ cost or energy consumption. We also point out that the SFA can be used not only to measure the level of efficiency in the use of energy, but also “rebound effects” associated with improvements in energy efficiency.