Controlling market power and price spikes in electricity networks: Demand-side bidding

Interdisciplinary Center for Economic Science, George Mason University, 4400 University Drive, MSN 1B2, Fairfax, VA 22030

Contributed by Vernon L. Smith

Abstract

In this article we report an experiment that examines how demand-side bidding can discipline generators in a market for electric power. First we develop a treatment without demand-side bidding; two large firms are allocated baseload and intermediate cost generators such that either firm might unilaterally withhold the capacity of its intermediate cost generators from the market to benefit from the supracompetitive prices that would result from only selling its baseload units. In a converse treatment, ownership of some of the intermediate cost generators is transferred from each of these firms to two other firms such that no one firm could unilaterally restrict output to spawn supracompetitive prices. Having established a well controlled data set with price spikes paralleling those observed in the naturally occurring economy, we also extend the design to include demand-side bidding. We find that demand-side bidding completely neutralizes the exercise of market power and eliminates price spikes even in the presence of structural market power.

electric power deregulation|experimental economics

Footnotes

↵ * To whom correspondence should be addressed. E-mail: bwilson3{at}gmu.edu.
↵ † However, there are many power systems that are essentially radial: e.g. Australia, New Zealand, and the United Kingdom. The latter is similar to the network we study here, with London as the main demand center to which power is transmitted from large supply sources to the north and a smaller source to the south.
↵ ‡ The firm that does not raise its offer realizes a profit of 944 [(166 − 20) × 4 units + (166 − 76) × 4 units].
↵ § The Cournot equilibrium involves any combination of S1 and S2 outputs such that total output is 18, all baseload capacity is included in the equilibrium, and the price is 166. This concept of organizing behavior can be rejected if S1 and S2 offer quantities such that the aggregate quantity exceeds 18.
↵ ¶ It should be noted that a seller can submit any offer up to 96 for intermediate cost generators and hence set the price above the marginal cost of 76, but this action does not reduce efficiency. All the gains from trade are still realized with this deviation from the strict Bertrandesqe competitive equilibrium.
↵ ‖ For brevity, the results are presented exclusively in terms of price outcomes. Results for efficiency parallel our price observations.