A Review of Energy Demand Models

Abstract

Energy demand modeling is one of the main challenging issues for energy experts. This study attempts to identify the gaps and differences in the energy demand modeling literature. More specifically it concentrates on the properties of two approaches: "single equation" and "demand system", which are commonly used in this area. Our review indicates that the single-equation approach, despite its simplicity for estimating parameters, suffers from severe problems, such as endogeneity and instability in parameters. However, the serious problem with this approach is that its specification has not been supported by the rigorous microeconomic theory. On the other hand, the energy demand system model is explicitly derived from the microeconomic theory that makes it possible to derive own- and cross-price elasticities of energy fuels and consider inter-fuel substitution. However, the major problem of this approach is that the demand elasticity estimates are more sensitive to selecting functional specifications. We investigate and compare the properties of the two common functional forms in energy demand systems namely, translog and linear logit model. Our results indicate that the linear logit model, under theoretical regularity conditions, can provide more convenient estimates of demand elasticities compared to the translog functional form.

Keywords

• demand elasticity
• inter-fuel substitution
• translog functional forms
• linear logit model
• dynamic models.