We explore the long-run impact of the Kyoto Protocol commitments to limit greenhouse gas emissions under various assumptions about the international spillover arising from actions led by the industrialised countries. International spillover comprises many complex processes including substitution due to price effects, diffusion of technology innovations, and policy and political spillovers. We represent these in terms of their aggregate impact on emission intensities over the next century. Limiting industrialised country emissions alone has limited environmental benefit if there is no international spillover; in our base case atmospheric concentrations by the end of the century rise to 730 ppm. However, this implies a large divergence of emission intensities, contrary to both empirical long term aggregate trends, and to identifiable influences towards convergence associated with economic globalisation. In contrast, if spillover leads to convergence of emission intensities by 2100, atmospheric concentrations are kept to below 560 ppm and are close to stabilising. We argue that zero or negative international spillover, as assumed in many analyses, is not credible; we estimate the most likely range for the international spillover parameter in our model to be 0.5-1.0. For our base scenario this would imply a mean global average temperature change from pre-industrial levels by 2100 of 2.7-3.4 degreesC instead of 4.2 degreesC, and rising at only 0.15-0.29 degreesC/decade instead of 0.45 degreesC/decade. Long-run sea-level rise is greatly curtailed. The regional benefits to the industrialised countries are also magnified because of the spillover to developing county emissions. Although the aggregate degree of spillover is uncertain, the available evidence suggests that it will be important and environmentally beneficial in aggregate. Spillover will help to spread the global effectiveness of the Kyoto first period and subsequent commitments, and deserves much further scrutiny.
