Applying input-output analysis to scenario analysis of ecological footprints

Application of input-output analysis to ecological footprints (EFs) is shifting from an ex-post static calculation toward an ex-ante scenario analysis for enhancing the policy relevance of EF analysis. This change in application prompts two issues requiring careful examination: (1) what is measured by Leontief inverses or extended environmental Leontief inverses, and (2) whether a sector's land multiplier (or compositions of land multiplier) can appropriately reflect the effect of delivering one unit of the sector's output to final demand on the required area(s) of production lands used by the sector itself and by other producing sectors whose products are contributed directly and indirectly to its production. The underlying message of these two questions is whether the assumed linear marginal relationship between a sector's output and its intermediate inputs (input-output coefficients)--a critical assumption made by W. Leontief to transform a transactions table from an accounting framework into the input-output model--can be extended to assume that the marginal relationship between a sector's output and the area of production land it uses for generating output is fixed (land-output marginal coefficient equal to average coefficient). By reviewing the literature on input-output analysis and its application to environmental issues and by theoretically and empirically examining the relation between sector output and land appropriation, this study advises against the use of land multipliers or their compositions in the EF scenario analysis. To apply the input-output model to EF scenario analysis for enhancing policy relevance with due attention to the relationship between sector output and land appropriation, this study suggests a two-stage EF calculation procedure. In the first stage, the input-output application estimates only the required raw materials (or generated pollutants) for meeting a given consumption pattern, which is the objective of environmental input-output analysis; in the second stage, the estimated amount of raw materials or pollutants is converted into land/water area, and a choice of conversion methods is suggested according to the research questions and the availability of conversion methods and required data.