Crossing non-parametric and parametric techniques for measuring the efficiency: Evidence from 65 European electricity Distribution System Operators

Abstract

Benchmarking techniques have been one of the main tools used by National Regulatory Authorities (NRA) to provide reliable information to define the efficiency targets for regulated undertakings. This paper aims to define the operational efficiency of the Portuguese mainland electricity Distribution System Operator (DSO), a monopoly incumbent with more than six million customers. Given the monopolistic nature of the electricity distribution activity in Portugal, the analysis was supported on a wide sample of 65 European DSO that operate in 16 European countries. This geographic spread required the use of different data harmonization procedures for an adequate use of different methodologies, namely Stochastic Frontier Analysis (SFA), Data Envelopment Analysis (DEA) and Total Productivity Indices (Malmquist), in order to guarantee robust results for the efficiency assessment. The study combines two outputs (number of supply points and network extension) and two inputs (operating costs and total costs) in different models. The results obtained are relatively positive for Portuguese DSO, placing the mainland DSO at efficiency levels between the 1st quartile and the efficiency frontier. There is a high correlation between DEA and SFA predicted efficiency scores, which supports the robustness of the analysis.