Key points are not available for this paper at this time.
Supply chains become increasingly globalized. Multi-regional input-output databases contain all the information to assess impacts along the value chain, but standard calculation routines to track the impacts of any sector along the global upstream and downstream value chain are missing. Mapping the impacts of materials has been a particular challenge owing to difficulties with double-counting. This is attributed to the strong intertwining of the material supply chain meaning that different materials occur in the supply chains of other materials. Here, we present a new method which can be applied to any MRIO system to track the impacts of any sector or region without double-counting upstream and downstream the global value chain. We apply this approach to EXIOBASE3 and implement a cutting-edge set of regionalized environmental impact categories and socio-economic indicators. Applied to global material production, our method shows that the issue of double-counting (prevented in this study) would overestimate global impacts of materials by up to 30%. In contrast, assessing only the direct impacts would lead to an underestimation by ~20%. Our evaluation further reveals that 25-35% of global material-related impacts are embodied in trade among ten world regions. Thereby, we identify the major international trade relations of key materials and found a clear trend of industrialized nations causing impacts in less developed economies. It was further revealed that during 1995-2011, the share of materials in total global climate change impacts has remained almost constant at ~50%, but total impacts have significantly increased for minerals and fossils. Our results demonstrate the importance for improved environmental policy strategies that target several stages of the global value chain. The methodology is provided as Matlab tool and can be applied to any material, industrial sector and region to track the related impacts upstream and downstream the global value chain.
Cabernard et al. (Tue,) studied this question.