Abstract

Providing carbon footprints of foreign direct investment (FDI) is one of the key recommendation of the G20 Data Gaps Initiative (DGI-3). To do so, the IMF suggests a leading methodology to capture the carbon footprint of inward FDI flows using macro data and splitting aggregate carbon emissions (from IEA data) into FDI breakdowns using multi regional Input-Output tables (from OECD data). In parallel, the authors have worked on an alternative methodology to estimate FDI stocks’ carbon footprint using micro data (see forthcoming IFC bulletin, Genre et al. (2024)). We use carbon footprints of listed companies, either directly reported by the companies themselves or estimated by private providers such as Institutional Shareholder’s Services. This data is yearly and available from 2012 to 2022 for up to 30 000 companies. We then suggest a model to allocate GHG emissions to various entities of the same group or to estimate it when too few information is available about the group. It allows us to compute FDI stocks’ carbon footprint from which we can infer FDI flows’ carbon footprint. In this paper, we compare both methods in order to provide robustness evidence to the IMF approach. We first compare both concepts, listing pros and cons of both approaches. While the IMF method uses macro data, our complementary methodology relies on granular data. The IMF method focuses on FDI flows while our method consider stocks. The IMF methodology estimates FDI stocks’ carbon footprint for Scope 1 only, while our methodology is able to deal with all scopes. We then discuss the missing links to applying both methods and identify the data needed to fill in the gaps. Eventually, we compile results with both methods at different granularity levels for both geography and activity sector information and try to draw conclusions of recent trends and levels.