Estimation of cost-effectiveness of the Finnish electricity distribution utilities

This paper examines the cost-effectiveness of Finnish electricity distribution utilities. We estimate several panel data stochastic frontier specifications using both Cobb-Douglas and Translog model specifications. The conventional models are extended in order to model observed heterogeneity explicitly in the cost frontier models. The true fixed effects model has been used as a representative of the models which account for unobserved heterogeneity and extended conventional random effect models have been used in analysing the impact of observed heterogeneity. A comparison between the conventional random effects model and models where heterogeneity component is entered either into the mean or into the variance of the inefficiency term shows that relative efficiency scores diminish when heterogeneity is added to the analysis. The true fixed effects model on the other hand gives clearly smaller inefficiency scores than random effects models. In the paper we also show that the relative inefficiency scores and rankings are not sensitive to the cost function specification. Our analysis points out the importance of the efficient use of the existing distribution network. The economies of scale results suggest that firms could reduce their operating costs by using networks more efficiently. According to our results average size firms which have high load factors are the most efficient ones. All firms have unused capacities so that they can improve cost-effectiveness rather by increasing the average distributed volumes than by mergers.