The present thesis studies climate policy design. It focuses on decarbonising energy systems thanks to combinations of policy instruments rather than stand-alone carbon pricing. Indeed, virtually eliminating carbon dioxide (CO2) emissions from energy by 2050 was found to require an extremely high carbon price by several authors, including ourselves - much higher than seems politically feasible. Thanks to supporting measures such as, e.g. subsidies and prescriptions, instrument mixes hold the promise of reaching this goal with a much lower carbon price. The dissertation assesses the performance of various instrument mixes with a diversity of methods.
We begin with a review of the international literature on packages of climate policy instruments. On this basis, we discuss the theoretical underpinnings for resorting to a plurality of policy measures and propose candidate policy tools adapted to specific decarbonisation levers and barriers in each sector. We find that shortcomings of stand-alone instruments may often be alleviated by thought-through combinations of measures, although trade-offs are inevitable. Desirable instrument mixes are found to vary between sectors, and typically include some level of carbon pricing.
One reason why simulation models featuring economic feedback effects require an extremely high carbon price to induce deep decarbonisation is that they employ constant elasticity of substitution (CES) functions to represent technological diffusion. Hence, we develop an alternative form based on the logistic distribution. After investigating its properties analytically, we implement the novel specification in key energy nests of our computable general equilibrium (CGE) framework. We show that model convergence is improved under the logistic form and that the required carbon price to induce deep decarbonisation is significantly reduced compared to the standard CES version.
The augmented CGE framework with logistic demand functions allows us to simulate decarbonisation pathways compatible with Switzerland's 2050 emissions target. Further, we introduce energy-specific capital with explicit investment vintages in key sectors of the model such as to represent novel policy measures on top of carbon pricing. Four strategically designed instrument mixes are employed as case studies to highlight trade-offs implied by differentiated instrument stringencies. We find that implementing additional measures reduces the carbon price needed to reach a set target, but that static and dynamic trade-offs remain between various macroeconomic indicators and the breakdown of non-abated emissions.
One motive for implementing instrument mixes instead of stand-alone carbon pricing is greater acceptability. Therefore, we administered a survey to a representative sample of Swiss residents to elicit social preferences for climate policy design toward net zero in Switzerland. An innovative feature of the survey is that it let respondents choose stringency levels of their favourite policy instruments via an interactive calculator. We find evidence of differentiated preference structures across sectors and an inclination towards mixes rather than stand-alone measures. Our regression models show that the language region, the number of cars in the household as well as the perceived danger of climate change are key determinants of preferences.