Essays on energy technology learning and voluntary carbon offsets

Abstract: This thesis consists of an introductory part and three self-contained papers, all related to the issue of promoting renewable energy sources. Paper I presents the results from a hypothetical market experiment of Swedish students' stated willingness to purchase emission allowances for carbon dioxide within the European emissions trading system. Methodologically we draw heavily on recent developments in the literature on integrating norm-motivated behaviour into neoclassical consumer theory, and assume that individuals have a preference for keeping a self-image as a responsible and thus norm-compliant person. The results indicate that students' willingness to purchase emission allowances is determined by both price and the presence of norms; people who feel personally responsible for contributing to reduced climate damages are also the ones who appear more inclined to purchase emission allowances. The empirical findings are also consistent with the notion that perceptions about others' stated willingness to purchase emission allowances affect individual norms and ultimately expressed behaviour. Norms are also largely activated by problem awareness and the individual's perception of her ability to contribute to solving the problem. In paper II we conduct a metaanalysis of wind power learning rates, thus permitting an assessment of some of the most important model specification and data issues that influence the estimated learning coefficients. The econometric analysis in this paper relies on over 100 learning rate estimates presented in 35 studies, all conducted during the time period 1995-2010. The empirical results indicate that the choice of the geographical domain of learning, and thus implicitly of the assumed presence of learning spillovers, is an important determinant of wind power learning rates. We also find that the use of extended learning curve concepts, thus integrating most notably public R&D effects into the analysis, tends to result in lower learning rates than those generated by so-called single-factor learning curve studies.Finally, in paper III a critical analysis of the choice of model specification in learning curve analyses of wind power costs is presented. Special attention is paid to the question of the choice of national or global learning (cumulative capacity), and the inclusion of other variables such as R&D, scale effects and the inclusion of a time trend. To illustrate the importance of these methodological choices, a data set of pooled annual time series data over five European countries - Denmark (1986-1999), Germany (1990-1999), Spain (1990-1999), Sweden (1991-2002) and the United Kingdom (1991-2000) - is used to compare the results from different types of model specifications. The empirical results support the notion that the estimates of learning-by-doing rates may differ across different model specifications. In our data set the presence of global learning for wind power appears more important than that of national learning, but the estimates of the (global) learning rate are only marginally influenced by the introduction of R&D and scale effects. The results also show, though, that the impact of cumulative capacity on wind power costs appears to be very sensitive to the inclusion of a time trend in the traditional learning curve model.