Reducing the variation of environmental footprint estimates based on multiregional input-output databases

Footprint estimates from various MRIO databases sometimes show significant variation. As a result, conclusions about the absolute levels or trends of a footprint may be inconsistent. The sources of these variations are attributable to three phases in the footprint calculations: differences in data preparation, MRIO database construction and footprint calculation. Given that these steps include the work of the statistical community as well as academia, efforts from both disciplines will be required to help convergence in estimates.

Reducing the variation in footprint estimates is a major challenge. This paper aims to contribute to this convergence in three ways. First, this paper provides the first overview of footprint work at statistical offices, government agencies and international organisations. Second, a detailed analysis of the sources of the variation in estimates is provided. Thirdly, strategies are discussed that might help to reduce variation between footprint estimates.