An Application of Peak-Load Pricing

IN 1960 an article dealing, in general, with a theory of pricing for nonstorable products and, in particular, with the pricing of electricity appeared in this Journal.' It introduced at a theoretical level a novel approach to pricing pioneered by the French. Subsequently, other papers on the theory of peak-load pricing have appeared.2 All have been largely concerned with theory and there has been no empirical application in the United States. The purpose of this paper is to demonstrate the applicability of peak-load pricing to a typical U. S. electric utility. The paper will also isolate problems which will require further investigation if the theory is to be made generally applicable. In addition, it will be argued that existing rate structures do not adequately reflect resource costs and thus result in poor utilization of resources. The firm studied is considered to be typical of privately owned electric utilities in the United States. All cost data are for the year 1963, which was the latest year for which complete information was available when the project was undertaken. Before presenting the empirical study, the theory of peak-load pricing and events leading to its development will be reviewed. Treatment of the theory necessarily will be brief and those interested in a more complete development may refer to the original articles. In a classic article published in 1938, Hotelling proposed the use of marginal cost pricing for utilities.3 This system, he believed, would produce a welfare optimum in the Paretian sense. In the ensuing debate it was shown that marginal cost pricing in the case of decreasing-cost industries cannot be proven optimal without resorting to interpersonal comparisons. Since in decreasing-cost industries average cost is greater than marginal cost, pricing at marginal cost must produce a deficit. No feasible method of financing deficits while maintaining marginal conditions has been found. Moreover, a welfare choice between marginal cost pricing, with some practical method of financing resulting deficits, and any other operational pricing scheme is not permitted by the Paretian criterion. If each pricing method were represented by points on a two-dimensional utility map, neither would be on the * Assistant professor of business economics and management, Syracuse University.