Stochastic frontier panel data modelling for regulatory benchmarking: An application to the Turkish electricity distribution sector

This paper considers stochastic frontier panel data methods for regulatory benchmarking that allow for both latent heterogeneity and inefficiency, encapsulating the regulatory dilemma in comparative efficiency analysis for incentive regulation. It applies a distance function model with appropriate concavity properties for econometric estimation to a panel of electricity distribution utilities in Turkey since, in preparation for accession to the European Union, Turkey is required to align the sector with incentive benchmarking regulation that characterizes the European Union. The results confirm the importance of allowing simultaneously for heterogeneity and inefficiency and emphasize the need for specific time-invariant heterogeneity information, such as geographical data, on regulated companies in different regions.