Electricity externality studies : do the numbers make sense?

Sammanfattning: During the 1980s and 1990s a number of studies have addressed the problem of valuing the externalities arising from electricity production. The main objective of these studies has been to guide decision-making with respect to future fuel choices. However, the results of the studies have been ambiguous, especially when one looks at the size of externality estimates. These sometimes differ with several orders of magnitude, and in this sense previous studies provide poor guidance for policy makers. Hence, the purpose of this thesis is twofold; first to critically survey existing electricity externality studies (especially in the context of comparability and policy relevance), and more specifically to investigate whether there are any systematical explanations to the apparent discrepancy of externality estimates. The thesis starts by reviewing the theoretical and practical approaches to the valuation of externalities. After that, five representative and influential studies are critically reviewed and compared, this both to illustrate how externalities may be appraised and the problems arising when doing this. In particular it attempts to highlight important differences among studies. The review shows that the comparability of studies is generally limited because studies differ in scope and hence that it is hard to draw reliable conclusions. Based on the review and acknowledging that part, but not all, of the disparity in externality estimates may be attributed to the specific location of the plant in question, the following main hypotheses are formed and qualitatively examined; first the methodological approach matters, i.e., the choice of method affects the size of externality estimates; further that differences in thoroughness and basic assumptions among studies may provide parts of the explanation. The qualitative analysis points to the fact that the methodological choice may affect estimates, that the thoroughness of approach in the studies affects the magnitude of estimates, and also that the basic assumptions of the studies may influence results considerably. Quantitative analysis, relying on the statistical technique ANOTA (ANalysis Of TAbles), is also attempted. The ANOTA-analysis indicates that the choice of methodological approach matters. There also seems to be statistically significant differences among fuels, i.e., that the renewables (e.g., solar and wind power) seem to, on average, cause lower impacts than the fossil fuels. The analysis could however not find any (statistical) support for the most common explanation to the disparity of estimates, i.e., that the majority of the differences among studies are due to plant location. Overall, it is concluded that, the comparison of electricity externality studies is made harder by the fact that the scope and methodological approach differ among studies, thus reducing the studies' relevance to policy.