SECTOR DISAGGREGATION, STRUCTURAL EFFECT AND INDUSTRIAL ENERGY USE - AN APPROACH TO ANALYZE THE INTERRELATIONSHIPS

In the analysis of industrial energy use, the energy impact of structural change has often been estimated at a certain level of sector disaggregation using one of a large number of decomposition methods available. Such an estimate has two limitations. First, it gives the ''net'' energy impact which is a residual value after cancellation of the positive and negative components of the structural effect. The cancellation conceals information on structural change useful in energy demand analysis and forecasting. Second, such an estimate is highly specific since it is valid only at the level of sector disaggregation considered. We discuss these two limitations and present some possible solutions. We describe the concept of cancellation and propose two indicators to measure it. We then introduce the decomposition tree approach which is a procedure for studying the influence of level of sector disaggregation in a systematic manner. The approach can be easily adopted in many decomposition methods and has several other useful applications.