Multi-Regional Input-Output matrices for carbon footprints: strengths and weaknesses

The carbon footprint measures the greenhouse gas (GHG) emissions induced by a country's final demand. Although there is no internationally standardized method or data to perform the calculation, the use of International Input-Output Tables (MRIOs) is widespread in the literature. An MRIO traces all the flows of goods and services between industries in different regions of the world. It thus makes it possible to reconstruct the production process of goods and services, and therefore to measure the GHGs emitted at each stage. In this paper, we compare different MRIOs and GHG emission databases and the resulting carbon footprints in order to assess the reliability and robustness of the calculation method according to the sources used. We establish that the carbon footprint of France can vary up to 20% depending on the MRIO used. These differences are mainly due to the method used to break down imports by country of origin when constructing the MRIO. We show that the carbon footprint calculated with a MRIO that includes the details of France's main trading partners or that groups countries into economically homogeneous zones would be sufficient to estimate the carbon footprint adequately. On the other hand, aggregating industries into large sectors of activity can lead to a 15% change in the footprint. Using bootstrap methods, we establish that the variability of the coefficients, from one MRIO to another, leads to as much variability in the calculation of the carbon footprint as a Gaussian noise on the source data of the order of 20 to 30%.