Considering environmental costs of greenhouse gas emissions for setting a CO2 tax: A review

Triggered by the Fridays for Future movement, policy makers increasingly discuss the introduction of a CO2 tax. Determining an adequate CO2 tax level is complicated for political reasons, but even more so for scientific, economic, and ethical reasons. Through a literature review and subsequent analyses, this study explores the influence of methodological choices on greenhouse gas (GHG) environmental costs, focussing on the European context. Two main approaches exist, relying either on marginal damage costs (MDCs; 17 studies considered) or on marginal abatement costs (MACs; five studies considered). Integrated Assessment Models (IAMs) quantify the utility losses for individuals, expressed as MDCs. Only seven of the 17 studies examined policy-relevant GHG MDCs applicable to the European context. Current MDCs are either limited in terms of impacts considered (and those that are included are based on dated cause-effect relationships) or rely on Gross Domestic Product (GDP)-based impact functions that disregard, or arbitrarily include, disutility from impacts on non-marketed goods and services, including due to catastrophic events. Therefore, this paper identifies development needs of the IAMs. MACs, in turn, correspond to politically consistent costs to achieve politically set goals, if existing, with no immediate link to the impacts avoided. Depending on the methodological choices made, MDCs and MACs may well span two orders of magnitude. Consensus exists that discounting be conducted at low (social) rates. Primarily moral arguments support the use of equity weighting. Because there are impacts specific to CO2 and CH4 and unrelated to radiative forcing, and because measures differ between GHGs, using global warming potentials (GWPs) for deriving non-CO2 environmental costs from those for CO2 is only a fall-back solution. For a better policy support, MDCs that comprehensively cover climate change impacts and show a high degree of detail should be provided. (C) 2020 Elsevier B.V. All rights reserved.