Abstract
Public concern about energy-related environmental damage has put pressure on governments to enhance abatement policies. This has in turn led to increasing demands for information about and analysis of expected and optimal levels of environmental quality, and the means of influencing these levels. Energy demand modelling can act as a valuable tool in the production of this information, particularly where it relates to future levels of environmental quality and to how policy intervention can alter constraints facing consumers in order to achieve desired levels. The thesis focuses on two features of energy demand analysis. First, major developments in dynamic econometrics, and in particular the cointegration approach, have enabled energy demand modellers to incorporate long run equilibrating relationships between energy use and its main determinants, such as economic activity and relative prices, within a dynamic framework. Second, the existence of a diverse market for energy both in terms of the fuels used and the users suggests that energy demand should be analysed at a disaggregated level. The purpose of this thesis is to examine whether disaggregated dynamic energy demand modelling can provide additional valuable information for the formulation of environmental policy. The core of the thesis takes the form of six separate papers: three review papers link dynamic energy demand modelling with environmental policy and three case study papers use disaggregated dynamic energy demand modelling to examine aspects of future United Kingdom energy-related environmental quality and how it can be altered through changing the constraints facing consumers. While theory and traditional econometrics have provided useful information about energy consumption behaviour, the introduction of the cointegration approach and the error correction model are enabling economists to estimate more reliably the long run relationships between energy demand and its main determinants, and the gradual adjustment of consumers towards equilibrium consumption levels after a disequilibrating disturbance. This new information is giving a greater understanding of how to achieve desired future environmental quality levels. The heterogeneous nature of energy use, both in terms of the fuels used and their users, indicates that the accuracy of elasticity estimates and, thus, the quality of information they can provide for environmental policy might be improved through the disaggregation of dynamic energy demand modelling. Employing the cointegration approach and sectoral and fuel specific error correction models to generate estimates of income, three separate papers provide evidence in support of this argument. Individually, these papers provide information about sectoral and fuel specific elasticities, about how residential users alter their behaviour as economic constraints change and about how certain car users alter their fuel consumption patterns when provided with adverse environmental publicity about fuels. Together, these papers suggest that there exists considerable variation in income and real price elasticities between sectors and fuels, and that using the estimates generated by such disaggregated dynamic models will provide greater accuracy than aggregated and static models. These variations are of particular importance to environmental policy makers because of the difference in environmental impact associated with different fuels. The thesis, therefore, finds that, by generating more detailed and reliable elasticity estimates, disaggregated dynamic energy demand models provide new and valuable insights for the formulation of environmental policy. It concludes that disaggregated dynamic energy demand modelling will form an increasing share of the models used for environmental forecasts and policy analysis - although the overall interest in this area of research may shift as levels of certain energy-related environmental indicators improve (eg carbon dioxide, sulphur dioxide and nitrogen oxides) and others worsen (eg volatile organic compounds and PM10). The thesis suggests that considerable refinements in energy demand modelling can be made, both in the methods used and in the focus of empirical studies, which will lead to further improvements in the formulation of environmental policy.