Harmonizing the Nordic Regulation of Electricity Distribution

Abstract

Regulators for electricity network infrastructure, such as electricity distribution system operations (DSOs) face some particular challenges in the Nordic countries. Due to institutional, economic, and historical reasons the DSOs in the Nordic area are relatively numerous and heterogeneous in terms of ownership structure, size, and operating conditions. Since the deregulation in 1994–1999, the national regulators have independently devised regulation mechanisms that address the heterogeneity through econometric or engineering cost models as a basis for high-powered regimes. The frontier analysis models (such as data envelopment analysis in, e.g., Norway and Finland) are particularly useful here, given their incentive properties and cautious estimation of the production set. However, the total information rents in yardstick regimes and the bias in the frontier estimation are related to the number of observations (firms), which undermine their future application in the Nordic area under increasing interregional concentration. This paper develops a proposal for an alternative model, the revenue yardstick model, that can be applied across the national regulations and permit frontier estimations on final user cost rather than cost estimates, sensitive to, e.g., capital cost estimates, periodization, and allocation keys. The core of the model is a dynamic frontier yardstick model such as Agrell et al. (2005), but here applied only to strictly exogenous conditions, the output dimensions and the claimed revenues of the DSO. An equilibrium is implemented using asymmetric penalties for positive and negative deviations from the ex post frontier revenue, the yardstick, using the classic superefficiency model in analogy with Shleifer (1985). The model is particularly aimed at an international (interregional) application as it may embed national differences in regulation without jeopardizing the long-term sustainability of the model.

The constructive comments from two anonymous referees on a previous version of the manuscript are acknowledged. Support from Nordenergi for the NEMESYS project and contributions by Sumicsid AB, EC Group AS, Gaia Group OY, SKM Energy Consulting AS and RR Institute of Applied Economics as project partners are acknowledged. The paper represents only the opinion of the authors that remain responsible for any errors in the presentation.