Macro-economic impacts of energy saving policies in Denmark

Abstract

Many different mechanisms and instruments are relevant to consider with the aim of reducing the energy system’s dependence on fossil fuels in order to reduce the harmful emissions of greenhouse gases. In this PhD, I examine some of these mechanisms in order to evaluate if they work, as well as quantify how well they work. This is done within the context of Danish data, but the methods presented can easily be implemented using data from other countries. In addition, the historical progress in reducing harmful climate-related emissions is evaluated for the Scandinavian countries, Denmark, Norway and Sweden. Based on an initial hypothesis that households and industries react differently to different mechanisms, households and industries are analyzed separately. In this respect, a non-monetary transparency policy is evaluated for households. Specifically, the introduction of the new EU energy label in 2010 has been examined in order to evaluate added uptake of the more efficient household appliances. For industries, a mapping is presented describing which industries react to changes in the electricity price by using energy more or less efficiently. This is of interest as it is expected that a less polluting energy system depends to a higher degree on electricity, which will potentially cause the electricity price to fluctuate in the transitioning phase. Such a mapping should, especially, be of interest to policy makers, as policies may be targeted to specific industries in order to be more efficiently implemented. Regarding the new EU energy label, it is found that the sales of high efficiency appliances rose by 55% at the announcement date of the new energy label and rose additionally by 42% at the implementation date. In addition, it is found that low efficiency appliances were only affected around the implementation date, when sales decreased by 45%. For the mapping of industries, Danish industry is disaggregated and 99 (sub) industries that are analyzed individually. It is found that 15 industries use electricity more (less) efficiently when electricity becomes more (less) expensive. For 18 industries, it is found that there are efficiency improvements for both electricity and other energy (an aggregate of other energy sources than electricity) when electricity prices increase. For (only) 4 industries it is found that an increase in the electricity price results in improvements in how efficiently electricity is used while it also results in other energy being used less efficiently, thus, indicating substitution from electricity to other energy sources. In addition, the text of the thesis presents a discussion of how estimates from the individually analyzed industries may themselves yield interesting results. In a Scandinavian context, decoupling of emissions and income is analyzed. The thesis adds to the current understanding of decoupling by considering both long-run and short/medium-run decoupling. It is found that, for all three countries income has risen faster than emissions in the analyzed period (1965-2018). However, in the shorter run income and emissions follow the same movements in all three countries. That is, while the long-run deterministic trends in income and emissions differ, the shorter run stochastic movements are the same. This may be interpreted as income and emissions still being linked but that the three countries have become more efficient in that they produce more units of output pr unit of emissions now than they did earlier. In general, this thesis implements econometric time series methods, including coinetegration and automatic model selection algorithms, to analyze energy data. The implementation of these methods in this framework is novel and, thus, the methodological approach may easily be implemented on other data or tweaked marginally to obtain other novel results, as is discussed.