Greenhouse abatement should be carried out where it is cheapest to do. In addition to cost efficiency in the domestic context, which in many ways has been the topic of the previous section on instruments, a least cost solution should therefore also be sought at an international level. If domestic efficiency involves equalising the marginal costs of abatement between the different sectors of an economy, international efficiency requires their equalisation between countries. It is well known that imposing a scheme of uniform percentage reductions across countries
The secondary benefits literature (see section 2) does deal with multiple externalities, but neglects the revenue raising aspect.
Table 5: Efficiency Gains from Emissions Trading
(2 percentage point reduction scenario)
Edmonds Reilly GREEN Manne Richels Tax ($/tC) GDP loss (%) Tax ($/tC) GDP loss (%) Tax ($/tC) welfare loss 2020 equal cut 283 1.9 149 1.9 325 - equal tax 238 1.6 106 1.0 308 - 2050 equal cut 680 3.7 230 2.6 448 - equal tax 498 3.3 182 1.9 374 - 2100 equal cut 1,304* 5.7* - - 242 8.0" equal tax 919* 5.1* - - 208 7.5" “ year 2095
^ discounted consumption losses through 2100, trillions of $ Source: OECD (1993).
would generally not lead to this outcome, since countries facing high abatement costs may be forced to undertake abatement which could more cheaply be achieved elsewhere (see Hoel, 1991a). Table 5 shows estimates of the cost difference between a uniform reduction agreement and efficient cooperation, achieved through a uniform carbon tax.
Just how an efficient outcome would have to look like is however unclear. The question has been analysed e.g. by Martin et al. (1992) using their GREEN model, and Kvemdokk (1993), using a reduced form abatement equation based on the
Manne and Richels model. A cost efficiency scenario - achieved through a global carbon tax - was also part of the OECD model comparison (see OECD, 1993). Barrett (1992c) has investigated the question for the EC. The results are rather ambiguous. GREEN and the Edmonds and Reilly model predict that, compared to equiproportionate cuts, a cost efficient scheme would shift the abatement burden away from rich and semi-rich countries to the developing world. Manne and Richels, and thus also Kvemdokk, come to exactly the opposite conclusion. The difference seems to hinge on the treatment of coal. While Martin et al. make out cheap fuel switching opportunities in coal-based economies such as China, Kvemdokk argues that for a poor country like China switching from cheap domestic coal to imported oil and gas will be relatively expensive.
Efficiency questions are also addressed in the more stylised game theoretic models by Tahvonen (1993), and Eyckmans et al. (1992)^°. These models are basically derivations of the cost-benefit models introduced in section 4, simplified with respect to the climate dynamics, but extended to include several world regions. This extension allows the calculation of the socially efficient abatement levels for each region, as well as the individually optimal levels in a situation without agreement, and therefore also the gains from cooperation, i.e. the increase in welfare by moving from one situation to the other. On a whole, the models confirm the scattered picture from the cost efficiency literature, although there seems to be some agreement that the gains from cooperation may be negative for regions like China and the former USSR.
This raises the question of side payments. Much as in the domestic case, the above efficiency criterion has to be complemented by equity considerations. In the international context the question is mainly about a just distribution of the abatement burden between the countries of the developed and the developing world. Analytically, distributional issues are often examined in the context of the allocation of tradeable emissions permits. This is mainly for analytical convenience, though.
as it allows for a straightforward separation between the equity and efficiency issues^\ The question could of course equally be treated in terms of e.g. the redistribution of tax revenues or proceeds from a permit auction, or indeed the size of transfer payments in general.
A variety of allocation criteria have been discussed in the literature, including a distribution in proportion to welfare, population, land area, previous greenhouse gas emissions, as well as several variants and combinations of these (see e.g. Rose, 1992; Barrett, 1992a; Kvemdokk, 1993; Grubb et al., 1992; Grubb, 1989). The choice between them is influenced by various criteria of social justice on the one hand, and by the necessities of realpolitik on the other (Grubb et al. , 1992). On the equity side there is Rose (1992; 1990), who has argued that countries will be willing to accept a fair distribution even if it is to their disadvantage, and who has put forward several ethical criteria to determine what "fairness" may mean. On the other side there is the more down to earth stand taken by e.g. Barrett (1992a, b) and Hoel (1991a), which is also implicit in the two greenhouse game papers mentioned above. These authors argue that in the absence of an international legislative body able to enforce an agreement, a treaty will only be signed if it is beneficial to all participants. The main distributional problem is thus not equity as such but finding an allocation rule which guarantees a beneficial outcome for a sufficient number of countries^^.
The ethical criteria put forward by Rose include, among others, horizontal equity (encompassing egalitarianism as its most extreme form), vertical equity (encompassing Rawlsian minimax), market justice (the free market is fair), ability
There are schemes which for equity reasons foresee different abatement targets for different groups of countries, such as stabilisation of emissions in developed countries only. Note however that these, much like uniform reduction schemes, are unlikely to fulfil the efficiency criterion.
^ Even if this is achieved, there may still be a problem about stability. That is, even if a certain agreement is generally preferred to a situation without, there is no guarantee that countries will not find it even more profitable not to join and to ffee-ride on the efforts of others. See section 6.2 below.
to pay, but also the Pareto rule and consensus building, which opens the door to the realpolitik view of Barrett. Although emphasising the importance of consensus building, ability to pay and vertical equity - and thus putting the ball into the comer of developed countries -, Rose is rather reluctant to identify a single preferred criterion. Other authors are less so. Grubb (1989) and Kvemdokk (1992) advocate an allocation according to population, pegged to a base period or restricted to adult population to avoid perverse incentives on population growth. Barrett (1992a) has put forward a "Kantian" mle, according to which countries have to abate at least as much as they would wish the others to do. Gmbb et al. (1992) see a mixed formula as the only feasible option. One view of the developing world has been made explicit in Agarwal and Narain (1991), who detect "environmental colonialism" in the debate, and who emphasise the importance of historical responsibilities. They further criticise the greenhouse gas emissions statistics of the World Resources Institute (see WRI, 1990) and propose that emissions accounts should be net of the carbon sinks found in a country. The lively discussion provoked by their contribution shows that the question of burden sharing is far from settled, and that the discussion has probably only just started. Inevitably the issue will have to be decided in the context of the actual negotiations towards a Climate Change Protocol.