Structural Change and the Environment

A Case Study of China's Production Recipe and Carbon Dioxide Emissions


  • Jun Lan,

  • Manfred Lenzen,

  • Erik Dietzenbacher,

  • Daniel Moran,

  • Keiichiro Kanemoto,

  • Joy Murray,

  • Arne Geschke

Jun Lan, University of Sydney, Integrated Sustainability Analysis, School of Physics A28, University of Sydney, Sydney, New South Wales 2006, Australia. E-mail:


We use the input-output tables in constant prices extended with carbon dioxide (CO2) emissions for examining the development of China, a country undergoing rapid growth. We undertake this empirical analysis in terms of a new and therefore rarely applied methodology: instead of average coefficients characterizing the average (old) technology operating throughout a particular reporting year, we calculate marginal coefficients—in monetary and CO2 terms—that capture the additional (new) technology installed after that year. Marginal coefficients are increasingly recommended in the literature for applications such as consequential life cycle assessment, where they are supposed to lead to more realistic results, especially in prospective analyses. Our work provides a first, broad overview about the magnitude and distribution of these coefficients across recent years in China's rapidly growing economy for which marginal coefficients could be expected to differ greatly from average coefficients. We find that (1) marginal coefficients can differ substantially from average coefficients, thus lending support to the need expressed in the literature for coining consequential life cycle assessment (LCA) and similar prospective assessment in marginal rather than average terms; (2) marginal CO2 emissions coefficients differ more from their average counterparts than marginal monetary coefficients, showing that for China, within-sector technological solutions to emissions abatement have played a more important role than the reorganization of supply structures; and (3) there exists considerable scatter and variation of marginal coefficients across years, which to a certain extent precludes the identification of clear temporal and sectoral trends.