In the context of the 96/92 European Directive, European countries are reforming the institutional structure of their electric power industry. Great Britain is one of the most advanced country regarding the implementation of such reforms and is currently involved in a complete review of the trading arrangements of its already deregulated electric power industry. The Review of Electricity Trading Arrangements (RETA) driven both by the government and the regulator encourages different changes of the current trading arrangements, namely the English Pool, in order to improve its efficiency. Such changes are detailed in a document published by the regulator in October 1999. It details the New Electricity Trading Arrangements (NETA) for the England and Wales (E&W) electric power industry that will be implemented in autumn 2000.

The subject of the present submission is the replacement of the Electricity Pool of England and Wales by the New Electricity Trading Arrangements. The research focuses on its effects on market prices and physical dispatch efficiency. Clearly, there is no concluding analysis to date that shows whether the NETA will improve the efficiency of the electric market or not. This is due both to the radical nature of the modifications - which do not allow any 'term to term' comparisons - and to their multidimensional nature. Among these modifications are the ones concerning the pricing processes and the physical and economical co-ordination mechanisms. The Electricity Pool typically works as a centralised uniform multi-unit offer auction, where the System Operator organises both the pricing of electricity for the next 24 hours and the physical dispatch. To the contrary The NETA consists of a succession of continuous markets (futures, forward and spot markets) down to four hours ahead of real time, followed by a continuous balancing mechanism where only there the System Operator operates.

Some papers have already addressed critics to the governments' proposals (Newbery, 1998; Green, 1999; Bower, Bunn 1999; Wolfram, 1999). The most severe of these critics are concerning the switching from a uniform-price auction on a compulsory centralised market by a discriminatory auction on both a spot market and a balancing mechanism. At the best, Green and McDaniel (1999) focusing on the balancing market provide evidence of the revenue equivalence between the two auctions. At the worth, Bower and Bunn (1999) have conducted simulations showing that prices should be higher and market power amplified in the bilateral market situation than in the Pool-based situation, all the more if producers get paid their own bids instead of the marginal price. Wolfram (1999), Bower and Bunn (1999) and Henney (1999) all argue that without properly addressing the remaining market power problem affecting industry structures, it is unlikely that changes will promote an efficient electricity market.

Our contribution aim at using the experimental economics method in order to compare the new and the old auctions in the core of the problem: the switching from a centralised market with uniform-price to a decentralised discriminatory auction. Experimental method has already made major contributions to using the market to deregulate the electrical sector, in particular in the United States (Denton et al. 1998, Plott 1997). We intend to compare the two market institutions, both in two different market structures (competitive and impure duopoly). Therefore, we shall compare four different treatments, with real subjects ceteris paribus.

Our results will allow to measure the efficiency of the four different market treatments, measured by prices, global surplus, and equilibration properties (speed and convergence) in real live experiment. We will show to what extent, depending on the institutional and structural market context, real strategic behaviours of subjects both producing and distributing companies might lead to efficiency.

Our first intermediate results confirm the theoretical predictions. Indeed, our first set of experiments already shows the inefficiency of both market institutions, even with a structure a competitive market structure. We should extend such experiment in the case of market power problem.

Keywords: Electricity, Market design, Auctions, Experimental Economics

JEL Codes: C92, D44, L94

Co-authors Carine STAROPOLI, Dominique FINON, Bernard RUFFIEUX, Stephane Robin