8.3.2.1 Impact of Emissions Trading

All of the above studies considered various forms of emissions trading for Annex I economies. It was universally found that most non-Annex I economies that suffered welfare losses under uniform independent abatement also suffer smaller welfare losses under emissions trading. This is also the case in all of the studies for which results on movements in the terms of trade are published (Coppel and Lee 1995; Harrison and Rutherford, 1999).

Why are overall welfare losses to non-Annex I regions reduced by emissions trading? A key point is that because the marginal and average cost of abatement for the aggregate Annex I is lower under emissions trading than under uniform abatement, a higher GDP is achieved for a given reduction in emissions. This means that the reduction in emissions is achieved through a heavier reliance on substitution relative to output reduction (substitution involves the substitution of less emission-intensive for more emission-intensive Annex I produced inputs). The heavier reliance on substitution means that there is a less severe decline in fossil fuel prices and a lower increase in the price of manufactured goods that are fossil-fuel intensive. There is also less increase in non-Annex I exports of fossil-fuel intensive manufactured goods to Annex I regions under emissions trading than independent abatement. However, these increased exports divert resources from activities in which the original non-Annex I comparative advantage was higher and the overall result is less beneficial to most non-Annex I economies.

Some non-Annex I economies that experience welfare gains under independent abatement also experience smaller gains under emissions trading; however, the aggregate effect of emissions trading is found to be positive for non-Annex I economies: those that suffer welfare losses under independent abatement suffer smaller losses under emissions trading.

To summarize, despite a number of identifiable numerical discrepancies, there is agreement that the mixed pattern of gains and losses under the Kyoto targets results in a more positive aggregate outcome than under the assumed and more stringent pre-Kyoto targets. Similarities in the regions that are identified as gainers and losers are also quite marked. Oil-importing economies that rely on energy-intensive exports are gainers (and more so if the exports’ carbon intensity is low), economies that rely on oil exports experience losses, and the results are more unstable for economies between these two extremes.

8.3.2.2 Effects of Emission Leakage on Global Emissions Pathways

As discussed above, a reduction in Annex I emissions tends to increase non-Annex I emissions, reducing the environmental effectiveness of Annex I abatement. Emissions leakage is measured as the increase in non-Annex I emissions divided by the reduction in Annex I emissions.

A number of multiregional models have been used to estimate carbon leakage rates (Martin et al. 1992; Pezzey 1992; Oliveira-Martins et al. 1992; Manne and Oliveira-Martins, 1994; Edmonds et al., 1995; Golombek et al., 1995; Jacoby et al. 1997; Brown et al. 1999). In SAR (IPCC, 1996, p. 425) a high variance in estimates of emission leakage rates was noted; they ranged from close to zero (Martin et al. (1992) using the GREEN model) to 70% (Pezzey (1992) using the Whalley–Wigle model). In subsequent years, some reduction in this variance has occurred, in the range 5%–20%. This may in part arise from the development of a number of new models based on reasonably similar assumptions and data sources, and does not necessarily reflect more widespread agreement about appropriate behavioural assumptions. However, because emission leakage is an increasing function of the stringency of the abatement strategy, this may also be because carbon leakage is a less serious problem under the Kyoto targets than under the targets considered previously.

Technically, there is a clear correlation between the sign and magnitude of spillover effects analyzed above and the magnitude of carbon leakage. It is important, however, to recognize those parameters that have a critical influence on results:

• The assumed degree of substitutability between imports and domestic production. This is why models based on the Armington assumption that imports and domestic production are imperfect substitutes produces lower estimates of emission leakage than models based on the assumption of perfect substitutability.
  
• The ease of substitution among technologies with different emissions intensities in the electricity and the iron and steel industries in Annex I regions.
  
• The assumed degree of competitiveness in the world oil market; this issue is considered in Section 8.3.2.3.
  
• The existence of an international carbon-trading system: for a given abatement strategy, emission leakage is lower under emissions trading than under independent abatement. This conclusion flows logically from the discussion above on movements in terms of trade. Greater Annex I output reduction under independent abatement stimulates greater emission-intensive production in non-Annex I regions, through both higher prices for emission intensive products and lower prices for fossil fuels. Support for the above conclusions on the impact of emissions trading is found in ABARE-DFAT (1995), Brown et al. (1997b), Hinchy et al. (1998), Brown et al. (1999), McKibbin et al. (1999), Kainuma et al. (1999), and Bernstein et al. (1999).

IPCC Homepage