Efficient price and capacity choices under uncertain demand: An empirical analysis

Traditional economic analyses of the peak-load problem typically assume an unrealistic degree of regularity in demand during well-defined peak and off-peak periods. This issue is addressed through a comprehensive statistical model that separates demand into its systematic and stochastic components. This model is combined with a traditional economic model and applied to local telephone service, leading to substantive conclusions relevant for managerial decisions as well as further research, among them:

? Neglecting the systematicand stochastic structure of demand may lead to inefficient tariffs. Efficient measured service structures typically price individual callsbelow incremental capacity cost.

? Industry wide capacity decision rules that are exclusively driven by blockage probability targets during narrowly defined time periods may be economically inefficient.

? For telephone service, spot pricing, which sets high prices during periods ofactual congestion, has the potential to be considerably more efficient than traditional tariffs that set high prices during periods ofexpected congestion.