How large and uncertain are costs of 2030 GHG emissions reduction target for the European countries? Sensitivity analysis in a global CGE model

In the paper we address the problem of parameters uncertainty of computable general equilibrium (CGE) simulation results concerning the economic effects of climate policy actions. Large scale CGE models utilize extensive, detailed databases on the structure of the economies (industry-specific technologies, international trade patterns etc.). At the same time, the behaviour of the economic system modelled in the CGE framework is largely driven by assumptions rooted in theory, with relatively little empirical content. It is therefore crucial to understand how assumptions affect outcomes of policy experiments. We employ a static global CGE model PLACE, representing 35 regions and 20 industries, with a focus on representing links between economic activities, energy use and CO2 emissions. Applying systematic sensitivity analysis based on Stroud's (1957) Gaussian quadratures approach we test how variation in elasticity parameters (values of which are subject to substantial uncertainty) affects economic assessment of emission reduction policies. Using as our workhorse simulation scenario the imposition of the European Commission's 40% greenhouse gas (GHG) emission reduction target (with respect to 1990) we find that the uncertainty of model simulation results driven by the uncertainty in assumed elasticities values is quite remarkable indicating that presenting only mean simulation results from CGE models is not sufficient.