System average rates of U.S. investor-owned electric utilities : a statistical benchmark study

Abstract

Using multiple regression methods, we have undertaken a statistical "benchmark" study comparing system average electricity rates charged by three California utilities with 96 other US utilities over the 1984-93 time period. Although system average electricity rates are much higher in California than for the national average, we conclude that use of such unadjusted prices provides no meaningful information on how one evaluates the performance of utility management. Rather, we find that average electricity prices are affected to a large extent by a number of factors outside direct and immediate management control, such as local costs of doing business, the availability of low-cost generation sources (e.g., hydro and coal), customer and service territory characteristics such as customer density, use per customer, and a number of regulatory and environmental factors. Once one controls for these various factors, the remaining impact of utility management on system average rates is rather modest, and for the California utilities the impact of utility management (relative to the national average) is insignificantly different from zero. This finding of no difference in prices, holding constant the effects of factors outside of California utilities' control, is robust, being sustained in a large number of alternative models and estimation methods.