The case of partial agreements is most often analyzed in recent empirical literature, for two reasons. First, as shown in Section 10.2.1, theory suggests that a partial coalition forms at the equilibrium. Hence, the climate problem is neither a “tragedy of commons” nor a situation in which there are clear incentives to co-operative emission control. Second, the history of international environmental negotiations is a history of partial agreements that are slowly broadened as more and more countries decide to join the group of signatories. In the case of climate, in particular, the Kyoto agreement can be seen as a first partial climate agreement. Therefore, many papers have dealt with the costs and benefits of the Kyoto agreement and with the possible strategies to increase the number of countries that commit themselves to emission control targets (see the papers gathered in OECD (1998), and in Carraro (1999b, 2000); see also Burniaux (1998), Capros (1998), Ellerman et al. (1998), Grubb and Vrolijk (1998), Holtsmark (1998), Manne and Richels (1998), Mensbrugghe (1998), Carraro, 1999c), Nordhaus and Boyer (1999), and the surveys by Metz (1999) and Convery (1999)).

Two remarks are important. First, even if most recent analyses deal with the Kyoto agreement, there are studies that try to compute the optimal coalition structures, in terms of both participation and targets, independently of the decisions taken in Kyoto (a recent attempt is in Nordhaus and Boyer (1999)). Usually the conclusion derived from these papers is that Kyoto is neither economically nor environmentally optimal. However, the notion of optimality is not very useful when analyzing coalition formation. Indeed, what matters is the notion of the stability of a coalition. This identifies which countries have an incentive to join the coalition (sign the agreement) for different membership and institutional rules, baseline scenarios, abatement costs (and therefore climate policies, including the degree of adoption of Kyoto mechanisms), and environmental benefits (and therefore impacts, adaptation costs, etc.).

Second, the Kyoto agreement can theoretically be interpreted as a partial (Carraro, 1998) or as a global agreement (Chander et al., 1999). It is interpreted as a global agreement when all countries are seen as committed to emission targets. Those in Annex B are committed to emission targets with respect to 1990, the other ones are “committed” to emissions levels that evolve as in the baseline scenario. This second interpretation is nothing more than a “technical” interpretation, which is useful to show that:

• optimal emissions targets are not necessary because the same optimal outcome can be achieved through an international, unrestricted emissions-trading scheme among all countries (Chander et al., 1999); and
  
• the resultant outcome can be profitable to all countries if an appropriate economic and technological transfer scheme is adopted (Markusen 1975; Chander and Tulkens, 1995, 1997; Germain et al., 1997).

As a consequence, even a “partial”, suboptimal agreement like Kyoto can be transformed into a “global” optimal agreement (see Section 10.3.5).

Away from this ideal world of perfectly competitive and international market mechanisms, are the analyses of coalitions that, like the coalition formed by Annex I countries of the UNFCCC or Annex B countries of the Kyoto Protocol, are partial (formed by a subgroup of the negotiating countries). In this context, four questions need to be answered:

1. Are these partial coalitions effective?
   
2. Are they too costly for the signatory countries?
   
3. Can partial coalitions be enlarged by providing incentives for other countries to join? and
   
4. Is there a distribution of emission targets and/or of abatement costs such as to increase the size of partial coalitions and hence the effectiveness of a climate agreement?

(a) The answer to the first question depends on two main factors: the baseline scenario and the degree of leakage. If the baseline scenario is very ambitious and leakage is high, then countries find it difficult to undertake large emissions reductions (decreasing returns of scale in emission abatement are usually assumed), and also their effort is offset by the leakage effect (the increased emissions by free-riding countries). Hence, a partial coalition is effective whenever there is no or little leakage, high pollution levels characterize the baseline scenario, and signatory countries contribute a large share of the total emissions.

(b) For the second question, many studies try to assess the cost for Annex I countries of achieving given emissions targets under alternative policy options. These policy options include:

• the timing of the mitigation responses (see the special issue of Energy Economics edited by Carraro and Hourcade (1999));
• the degree of adoption of the Kyoto mechanisms and their features, such as banking (see the papers in OECD (1998) and Carraro (1999d));
  
• the role of complementary industrial policies, mainly designed to foster innovation (see Nordhaus, 1997; Schneider and Goulder, 1997; Kopp et al., 1998; Buonnano et al., 1999); and
  
• the effects of uncertainty about climate impacts or abatement costs (Carraro and Hourcade, 1999).

The main result can be summarized as follows. Despite their high variability, all the studies show that the Kyoto mechanisms sensibly reduce the costs of compliance, whatever the coalition structure. Hence, emissions trading, and more generally the application of the Kyoto mechanisms, can reduce overall mitigation costs without reducing the effectiveness of the climate policy. Chapters 6 and 8 give an extensive overview of relevant studies¹²

If assuming an even broader type of flexibility than incorporated in the Kyoto mechanisms (banking and borrowing and international emissions trading (IET) among all countries) then compliance costs are further lowered. This result is shown in Bosello and Roson (1999) for banking, Westskog (1999) for banking and borrowing, Manne and Richels (1999a, 1999b), McKibbin et al. (1998), and many others for IET among all countries. If in addition the incentives to innovation provided by the Kyoto mechanisms are taken into account, then compliance costs are even lower (Buonnano et al. 1999).

Continues on next page

IPCC Homepage