The Dilemma of Economic Versus Statistical Models of Energy (and Some Results of Forecasting Monthly Peak Electricity Demand Using a Transfer Function Model)

Abstract

The recent surge of interest among energy planners in economic models for predicting the energy outlook has coincided with a growing sense of disillusionment among many practicing econometricians about the forecasting performance of economic models (see, for example, Stekler, 1968). Many economists argue that the problem with economic models lies in the economic theories behind them. These theories analyze the impact of policy changes on the assumption that the structure will not change, when in fact what may happen is that the structure itself, and not just the variables of interest, may change as policy changes. What is needed is a theory that predicts how the structure will change in response to such policy changes.