## Kyoto Mechanisms

Here, we shall discuss each flexibility mechanism in the KP and examine the design features and incentive issues involved.

26.2.1 Emission trading (ET)

The idea of ET as a resolution to environmental problems has already been put in practice in several areas. For greenhouse-effect gases (GHG) emissions, the EU started the EU emission trade system (EU-ETS) in 2005. ET is a widely accepted concept in economics and has been well tested.

Let us introduce a simple model often utilized for this issue. There are n Annex B nations ratifying the KP. Their GHG emission levels (for the entire 5-year commitment period of the KP or annual) are denoted by xi while the assigned amount is : By the economic activity level associated with xi, country i enjoys benefits R (xt) and incurs costs Ci (xt) to achieve that emission level, and let us denote the net benefit by Bi (xi), which we assume is continuously differentiable, strictly increasing, and strictly concave. Note that, even though the KP requires that Z a = Z xt, xi could be more than ai because country i can purchase emission permits from the market of the amount xi - at at the going price p.

The efficient emission level (x*) given quota (at )n=1 is a solution to max Z Bi (xi) s.t. Z al = Z xi and characterized by mi (x*) = Mj (x*) for each i, j, with Z at = Z xi, where mi (x*) = B'(xi) is the marginal net benefit, which is usually referred to as the marginal emission reduction costs. We assume that the condition for the existence of the interior solution, i.e., x* > 0 is satisfied for all i.

In a competitive equilibrium, each country chooses xi so that the net benefit with market revenue, Bt (xt) + p(ai - xi), is maximized given a price p. Given an arbitrary p, the corresponding emission level xt (p)' s does not necessarily satisfy the market clearing condition Z at = Z x (p). When this condition is met, we call the price an equilibrium price p". It can be easily seen that p* = mt (x*) for each i.

This competitive analysis alone is classified as market analysis but there are many related issues where game theoretic analysis can be useful. Some of these issues are

(i) initial allocation of ai ' s

(ii) noncompetitive behavior

(iii) compliance (vi) dynamic issues

As for (i), one can see that the final payoff depends on the initial allocation of the quota and in fact, given the prospect of a continuation of Kyoto-style schemes, it would be a central issue in negotiation. We shall discuss it in the subsequent section. Emission trading works only if these quotas are firmly enforced, and so the issue could be viewed as one that precedes the emission trading. Issue (ii) is also an important issue as, once some agents start exerting market power, the efficiency property as outlined above no longer holds. People are also worried that Russia may become a single significant supplier. (One analysis accounting for such a possibility is Bernard et al. (2008) which combines computable general equilibrium (CGE) analysis and dynamic games.) Russia has been very slow in participating in ET, and its intention is not clear yet.

Issue (iii) is a fundamental issue for the entire mechanism. Emission trading is carried out over a fixed duration of time. Emissions are like other durable commodities that can be consumed or stored. But, unlike other commodities, there are entities that might have to purchase a certain amount of emission reduction units from the market (irrespective of its past trade position in emission trading) at the end of the period. Otherwise, they are deemed to be in a state of noncompliance. The immediate costs of noncompliance under the KP are a reduction in the emission cap in the next commitment period by an amount that is more than the actual excess emission in the current commitment period. (Actually, one is not sure whether the next commitment period exists so it is uncertain whether such a penalty is effective or not, and so there is a non-negligible probability that one gets away with this noncompliance penalty. Barrett (2005) views this loose compliance as the fatal defect of the KP. Of course, besides the "real" penalty, not delivering on its past promise harms the government's reputation, which is often of considerable importance in ensuring that nations abide by treaties in international relation theory.)

Besides, there is a lot of uncertainty (and rules yet to be determined) with regard to the dynamic aspect of the emission trading market, but an important issue in emission trading is the fact that only a few months after the end of the commitment period does the nation discover its own emission level and, here, issue (iv) becomes relevant. To accommodate this time lag, an additional 3-year time period is allowed. This creates an interesting end-period problem for this type of market. (In fact, no rule has been specified for this transition phase.) Ishii and Imai (2008) apply an extension of existing literature on execution strategy to the game situation, to this end-period problem. The number of applicable situations is limited and so there are only a few nations whose emissions may exceed the assigned amount together with a small number of speculators, and the emission quantity held by the third party is uncertain. The obtained result confirms the earlier assumption that speculators and the country under consideration wait until the last moment to execute their transactions.

26.2.2 Joint implementation (JI)

JI is a somewhat superfluous mechanism because there seems no need to award a special status to an emission-reducing project taking place inside Annex B nations because that reduction is a part of the emission reduction achieved within that nation and, to that effect, if the credit is obtained by the entities belonging to the other nation, it is nothing but one form of ET, provided that those entities are allowed to participate in ET. Therefore, except for the case where the national inventory system is discredited so that the nation is excluded from ET, JI has little to offer.

Noting this fact, the JI supervisory committee (JISC) offered two options to choose from for the proponents of a project: one without any special registration and monitoring system and one that parallels that of the CDM. In fact, for investing entities, there could be merit to choosing the second option, as they are assured of obtaining credits (called ERU (emission reduction units)) once the project is approved by the government of the host country, without further intervention.

Still, there arises several points of interest that might deserve further investigation. One is that of a possible switch from a non-Annex B nation to an Annex B nation and, concomitantly, a formerly CDM project would become a JI project. The Russian proposal at SB 25 and COP13 in 2007 turned such a possibility into a reality. Would such a shift create an incentive problem? Analysis of this problem would be potentially interesting and important. A related study of a non-Annex B nation's incentive given its future switch of status is given in Akita (2003).

Another problem would be the governmental incentive for engaging in JI concerning entities belonging to that nation. This is also a fruitful future research topic.

### 26.2.3 Clean development mechanism

The CDM is one innovative mechanism introduced into the KP, and it has already produced credits (called CER: certified emission reduction) before the first commitment period started. The UN announced that the amount of credits from the projects (including those in the pipeline) reached 2 billion tons (equivalent of CO2) which, according to some estimates, would be more than sufficient to fill the gap between the demand and supply of the emissions among Annex B nations ratifying the KP for the first commitment period. (According to the World Bank's 2008 report on market outlook, there is still a gap between demand and supply. See State and Trend of Carbon Markets issued annually.) This in turn implies that there would be no emission reduction for those nations if the marginal cost of the reduction is above those in non-Annex B nations. Thus, the overall environmental effect of the CDM is subtle, as the emission reduction in non-Annex B nations is used up by Annex B nations. The anticipated by-product, the technology transfer through the CDM, seems to have been unsuccessful (because of the prevalence of a unilateral CDM, which marks some distinction from JI). Partly because of this fact, the EB (executive board) of the CDM is rather rigid (by following the Marrakech Accord) in the approval of methodologies and credits in some instances.

26.2.3.1 Issues concerning baseline setting methods and additionality

The KP requires the CDM credits to be those emission reductions that are additional, i.e., those reductions occurring only because the project is undertaken because it is registered as a CDM project. Consequently, one must choose a baseline emission level so that the difference between the baseline and the actual emission level is the credited emission level. To determine the baseline, the applicant must select from the list of approved methodologies. For those familiar with counterfactual estimation in statistics, it should be well understood that reaching a reasonable estimate of the baseline is difficult. Therefore, skeptics have cast substantial doubt on this part of the mechanism. In fact, the methodology remains one of the major issues for the EB even now.

One issue concerns the basic principle of setting a baseline. The Marrakech Accord 2001 prepared three such approaches, which we may term (1) historical baseline, (2) industrial average baseline, and (3) technologically estimated baseline. Fischer (2005) evaluated these three approaches from an incentives viewpoint.

Apart from the choice of data source determining the baseline scenario, there is another issue concerning the baseline emission level. The easiest example is the case of a manufacturing firm undergoing an energy improvement project under the CDM. To produce an output level x0, one had to emit e0x0 before the project whereas, after the project, the emission level associated with an output level x0 is ex0 with e < e0. However, suppose the output level changes to x after the project. Thus, the emission level after the project is ex. Suppose simplistically, no technological progress was foreseen and so the historical baseline is appropriate. This implies that, to reach the baseline emission level, using the old emission coefficient e0 is appropriate. Still, the question remains whether the baseline emission level is based on the old output level or the current output level. We call the former method the ex ante baseline and the latter method the ex post baseline. (Other authors use names such as absolute vs relative, and level-based vs rate-based.) Note that the ex ante baseline emission level is deterministic at the start of the project whereas the ex post baseline emission level is uncertain as it depends on the actual output level.

Laurikka (2002) and Imai and Akita (2003) compare these two methods from a risk perspective. All these analyses fall into the category of individual decision making. Imai, Akita, and Niizawa (2008a) assume that the CDM firm is operating in an imperfectly competitive industry. Under such circumstances, one firm's decision to adopt a CDM project may well affect another firm's performance and, hence, incentive via the change in the relative cost advantages. In that paper, we compare not only the simplistic ex ante and ex post baselines but also other methods representing the industry average baseline and a conservative baseline, that is, an ex post baseline with an upper bound on the baseline emission level. In an analysis utilizing the Cournot competition game, we show that the impact of the incentive effect of baseline methods tends to be mitigated by the existence of other firms in the industry.

Consideration of the CDM by a firm interacting with other firms motivates one to study the incentive effect of the adoption of the CDM in a dynamic setting. In Imai, Akita, and Niizawa (2008b), the possibility of one firm adopting a CDM project may trigger adoptions by other firms. In general, this needs a special assumption of complementarity of their products and some sort of imperfection of knowledge. Given a related goods oligopoly model under several different settings, this possibility is established.

### 26.2.3.2 Bottom-up method

Most complaints regarding the current mechanism are about the management of the CDM. The basic issue is the role and functioning of the CDM-EB, which critics consider too centralized and slow, and too inconsistent over time with political views. However, at least some of these problems originate from the principle of the "bottom-up" method that was adopted in its inception. This method has a typical game theoretical problem and we shall briefly describe this below.

The bottom-up method implies that the UNFCCC office or the EB does not provide any prototype of methodologies usable for projects at the inception. Thus, the first project proposer has to draft up a methodology as well or it has to wait until somebody else develops a suitable methodology that is approved by the EB. To get a methodology approved is a costly procedure, apart from the writing of the project design document. Thus, the bottom-up method poses a typical collective-action problem of Olson. As this problem typically indicates, the initial pace of application for new methodology approval was slow and the coverage of approved methodologies remained quite incomplete for some time, so that solicitation for methodology applications was made in particular areas by the EB. A resolution for this would be some sort of coordination among the potential beneficiaries. One of the costs in those days was the cost of drafting a methodology where few templates existed and there is also uncertainty regarding whether the project can be registered, even though the methodology was approved. In contrast, the delay involved in waiting for others to submit a usable methodology is also costly, but other costs may be substantially reduced. Consequently, there could be incentives for all prospective proposers to wait, which would result in an all-wait equilibrium if the cost is too large for one entity to bear (cf. Fig. 26.1). If no one applies for approval of a new methodology, then nothing happens and if applying alone is profitable, or the agent has a motivation other than profits, then that agent may initiate. However, if there are two such agents, again a coordinating issue arises in the form of competition for the position of a free-rider.

One resolution for "nobody volunteers" or potential coordination failure would be a public coordination. In fact, there has been a request for the EB to provide a template. Furthermore, to help the EB process many requests, donations were made by beneficiary firms so that the EB can meet more frequently and employ more staff.

Fig. 26.1 Prisoners dilemma in bottom-up method (regardless of the opponent's, five choice of actions, each party finds it better to choose "wait").