Transaction costs and the clean development mechanism



The emissions trading provisions of the Kyoto Protocol and its clean development mechanism (CDM) are designed to permit greenhouse gas (GHG) emission reductions at the lowest cost globally. However, to ensure climate integrity, these reductions must pass through vigilant approval, monitoring and evaluation procedures that create additional transaction costs unrelated to the physical process of eliminating GHGs. Moreover, the CDM's additionality criterion creates constraints that magnify the influence of these transaction costs on project viability. If these costs are extreme, they could undermine the success of the CDM, and possibly of the Kyoto Protocol itself. This article describes the trading provisions of the treaty, creates a working definition of transaction costs, and discusses their effects. It then analyzes the process of creating a CDM project to identify the sources of transaction costs, illustrated by an example of a fuel substitution project in Ghana. The conditions for project profitability are analyzed and compared with recent GHG emission credit prices in Europe. The specific Ghanaian results are not generalizable to all CDM projects, but the model does suggest a template that can be used to analyze the effects of project and transaction costs in other contexts.

1. Introduction

Nearly two years after the Kyoto Protocol entered into force, the market for greenhouse gas (GHG) emissions has developed considerably. The Protocol, which was signed in 1997 and entered into force in February 2005, represents an international effort to address climate change threats by placing caps on GHG emissions in key countries. To achieve these reductions at the lowest cost, the Protocol includes an innovative GHG trading system, allowing the exchange of “tonnes of carbon dioxide equivalents (tCO2e)” to meet protocol obligations. This makes the Protocol one of the most ambitious pollution control systems ever undertaken: global in scope, covering multiple compounds, and combining national GHG accounting allocations with individual project-based accounts. Some countries have obligations to reduce GHG emissions; others do not. A concern raised by these challenges is that the treaty's trading provisions will require substantial expenses over and above the technical cost of reducing emissions in order to document and track these largely invisible compounds and create the confidence that investors and buyers need to trade in these goods. These additional expenses — transaction costs — could undermine the very processes designed to make compliance affordable and, ultimately, realistic.

Transaction costs arise at several points in the carbon trading process. Some happen as emission reductions are created; others affect the exchangeability of emission reductions that have been created or purchased elsewhere. This article begins by describing the Kyoto Protocol's GHG trading mechanisms, emphasizing the clean development mechanism (CDM). It then discusses the specific challenge posed by the Protocol's “additionality criterion”, provides a working definition of transaction costs, and analyzes the dynamics of CDM supply and demand. Finally, it identifies the points in the CDM process that generate key transaction costs and illustrates their effect on project success with an example from Ghana. The final conclusions focus on the effects of transaction costs and the future prospects for the CDM and the Kyoto Protocol in general.

2. Greenhouse gas markets, the Kyoto Protocol, and the flexibility mechanisms

The Kyoto Protocol establishes a framework for a worldwide market in greenhouse gas reductions using a globalized cap-and-trade approach. In brief, a set of countries — Annex B of the Kyoto Protocol,1 and corresponding roughly to “the industrialized countries”— agrees to “cap” their emissions of greenhouse gases at some level, specified in Annex B as a percentage reduction in emissions relative to a baseline year. The goal is that total greenhouse gas reductions from participating countries will remain at or below the sum of all caps in the Protocol. For Annex B countries, emission reduction commitments are typically between 5% and 8% of 1990 levels, although there are some exceptions. Although non-Annex B countries do not have present requirements, they are expected to start reducing their overall emissions intensity as they develop.

The Protocol permits countries to exchange greenhouse gas equivalents in a trading system in order to lower the average cost of emissions reductions. This creates incentives for countries that can achieve inexpensive reductions to go beyond their treaty requirements and sell “excess reductions” to countries that have higher costs. The specific rules (“modalities”) behind the exchanges are governed by three “flexibility mechanisms”— emissions trading (ET), joint implementation (JI), and the clean development mechanism (CDM) — defined in the Protocol. Although this article focuses specifically on the clean development mechanism, many of the issues associated with the trading system are more clearly seen when contemplating all three mechanisms together.

Emissions trading sets out the basic paradigm, allowing countries that have reduced emissions below their targets to sell excess reductions to countries that have not. This creates an incentive for countries that can reduce emissions cheaply to do so, creating less expensive options for other countries to meet their treaty obligations by purchasing excess reductions at a lower cost than can be achieved domestically. For example, Greece may be able to reduce emissions cheaply because its technological base can benefit greatly from upgrades. Germany, which may already employ upgraded technologies, may not be able to reduce emissions as cheaply and could therefore purchase Greece's excess reductions to meet its own targets. Provided Greece is able to meet its own reduction commitments, it stands to gain additional revenue from the trade.

Joint implementation takes the emissions trading concept one step further by allowing countries with emission reduction commitments2 to work jointly on a project that will reduce emissions in one Annex B country and divide credit for the reductions among themselves in some mutually agreeable formula. For example, Germany and Greece might agree to retrofit a Greek coal-fired electricity generator with German technology to improve efficiency, and thereby lower emissions. If the retrofit is expensive, Greece might not have chosen do it on its own, but Germany could find the arrangement attractive enough to transfer the technology, expertise, and/or financing in exchange for a share of resulting emission reductions. Whereas ET creates an incentive for Greece to be pro-active in reducing emissions further than its obligations require, JI creates an incentive for Germany to be pro-active in seeking out opportunities for its finance and technology to reduce emissions elsewhere.

The clean development mechanism can be thought of as the joint implementation mechanism extended to countries that do not presently have emission reduction commitments, although the CDM and JI are managed by administratively separate entities. The advantage for Annex I countries is that many of the lowest-cost GHG reduction opportunities lie in developing countries. However, without emission reduction requirements, non-Annex B countries have little or no incentive to reduce their own greenhouse gas emissions. Most, however, have an interest in technology transfer and in increasing foreign direct investment. The clean development mechanism thus permits Annex B countries to meet their emission requirements by investing in GHG emission reductions in non-Annex B countries, provided that the projects are approved by the host country government as meeting local sustainable development objectives and approved by the CDM executive board as representing real, measurable reductions of greenhouse gases beyond what could be expected to occur in the absence of a project.

3. The challenge of emissions accounting and the additionality criterion

Global emissions trading is made more complex by the fact that ET credits are apportioned on a national accounting basis, whereas JI and CDM create project-based emission reduction credits that can be owned by any juridical entity. The Protocol allocates each country a quantity of assigned amount units (AAUs) corresponding to its emission reduction commitment.3 Each AAU represents the ability to emit one tonne of CO2 or equivalent GHG and stay within treaty requirements. At the end of each accounting period, total observed emissions are subtracted from the number of AAUs in a country's account. Any unused AAUs are available for sale under ET rules and any deficits must be accounted for should there be fewer AAUs than total emissions. The sale and trade of AAUs is strictly an arrangement between national governments.4

If an Annex B country is host to one or more JI projects, its AAUs must be reduced by the number of JI emission reduction units (ERUs) produced in the same accounting period. This is necessary to reconcile national and project-based measurement systems and prevent double counting and is illustrated in Table 1. National emissions are estimated by a survey of all measurable emission sources within a territory. The survey's total will include the effects of any emission reductions that come from JI projects. Thus, when JI emission reductions are certified and turned into ERUs, the effects of those reductions have already reduced the total number of emissions observed, and so the number of AAUs allowed must be reduced accordingly. Essentially, JI projects transform a certain number of host country AAUs into an equivalent number of ERUs that are now transferable to other governments and private organizations in GHG markets.

Table 1. Accounting and compliance procedures
 Annual GHG Inventory Accounting
Calculation of AAUs Total National Emissions from all sources
 − AAUs (national account)
+ AAUs sold to other countries
+ ERUs (from hosted JI projects)
Sample CountryNational compliance Gap (or Surplus)
Baseline Year1000tCO2e 
Commitment−8% − ERUs Retained (from hosted JI projects)
AAUs920tCO2e / yr− CERs held
− Other Purchases (ERU, AAU, RMU, etc.)
 Fineable Gap (or saleable surplus)

Joint implementation and the clean development mechanism are both project-based accounting systems that must coexist with the national accounting system created by Kyoto. As a result, emission reductions from either mechanism must be measured against a baseline of “business as usual”, representing the best projection of the emissions quantity that would have taken place if JI or CDM were not available as a market for emission reductions. This criterion is called “additionality”, because it stipulates that project-based emission reductions must be “additional to” emission reductions that would have occurred if JI or the CDM were not available. The additionality criterion has important implications, addressed below.

For joint implementation, the baseline measurement demanded by additionality is important for appropriate pricing and fair distribution of benefits between the investor and the host country, but not necessarily for keeping track of total emissions, because any additional ERUs created by overestimating a baseline will be compensated by having an equivalent number of AAUs removed from circulation and vice versa. As a result, misquoting a baseline may cause ERUs to be cheaper or more expensive than they might otherwise be, and a larger number of ERUs could be distributed to implementation partners, but the total number of emissions would still be capped appropriately. Under JI, estimating an appropriate baseline is largely a policy problem for the host country, but it is not, strictly speaking, a challenge to emissions objectives.

In the clean development mechanism, which uses certified emission reduction (CER) units, creating an appropriate “business as usual” baseline scenario is substantially more important. The immediate concern is that, since CDM host countries have no current emission reduction commitments, both project developers and the host country designated national authority (DNA) have incentives to overestimate business-as-usual emissions in order to maximize the number of marketable CERs that can come from a project. This would increase project income and make it more attractive. Since no AAUs are allocated to these countries, there is no inbuilt balancing mechanism as there is with joint implementation. The CDM Executive Board does exercise some control by screening CDM projects and requiring that a CDM accredited designated operational entity (DOE) validate the methodology and baseline used for the project.

From a climate science perspective, baseline scenarios in the CDM are important for another reason. Once a baseline emission rate has been determined, project CERs represent the difference between the baseline and observed emissions. However, since CERs are ultimately used to permit additional GHG emissions elsewhere, the difference between observed emissions and the baseline scenario does not actually represent emissions eliminated from the atmosphere, merely emissions transferred to an Annex B country from a non-Annex B country. As a result, baseline emissions rates do not simply represent how much GHG would go into the atmosphere in a business-as-usual scenario, it also determines how much GHG will actually go into the atmosphere as a result of the project and whoever purchases the resulting CERs.5 This means that, in the CDM, an overstated baseline will have real effects on climate, but the same will not happen in JI. As a result, the Kyoto Protocol's climate integrity objective demands higher vigilance over the CDM, and this justifies some higher transaction costs.

The additionality criterion further complicates the CDM, because any project that reduces GHG emissions and generates a profit before adding CER revenue would be considered by definition a business-as-usual scenario and therefore ineligible for CERs.6 The overall result is that CDM projects must be projects for which the sale of CERs makes the difference between profitability and non-profitability. Figure 1 illustrates the effect with a hypothetical project requiring $100,000 in capital outlays. If the net present value (NPV) of non-CER project revenue is greater than $100,000 (first column), then the project is profitable, classified as business-as-usual, and not eligible for the CDM. If the total NPV of non-CER and CER revenue is less than $100,000 (last column), then the project is not economically viable, even with the sale of emission reductions, and will not be pursued without external subsidies. The only projects that are both eligible for the CDM and economically viable are projects where the difference between the upfront capital outlay and the NPV of non-CER revenue is less than the NPV of CER revenue. Graphically, the narrow acceptance region in Figure 1 is a region whose high value is set at the size of initial capital outlays and whose width is the NPV of CER revenue. If the NPV of non-CER revenue falls in this region, the project will be a viable CDM project.

Figure 1.

CDM eligibility and profitability assuming $100,000 in start-up costs: Effects on viable CDM projects.

A corollary is that the sale of CERs is the maximum obtainable profit from a CDM project. This means that small-scale projects are unlikely to be financially self sustaining, unless the reduction cost per CER is extremely low. A second corollary is that, if a CDM project involves additions to or retrofits of an existing operation and the additions have no other benefits, then the additions must cost less than the expected revenue from CERs. Together, both of these conditions suggest that CDM projects may have very narrow profit margins and be highly sensitive to fluctuations in the price of CERs. Large scale activities with high operating leverage and substantial reductions in GHG equivalents are likely to be the most profitable applications of the CDM under these conditions.

4. Defining transaction costs

Nearly all economic activities have transaction costs, but CDM transaction costs are especially important because the financial sustainability of CDM projects is so closely linked to the size of the CER revenue stream.7 Any factor that increases the cost of generating a CER or reduces its market price can have powerful effects on the number of CDM projects that are economically feasible and, by extension, how much GHG will be removed via the CDM.

It may be surprising that there is no universally accepted of even consensus definition of “transaction cost”. Most definitions in use are variations on “the costs associated with making an economic exchange”, the “cost of constructing and maintaining institutions [to implement and supervise exchanges]” (Furubotn and Richter, 1997), or sometimes just “the cost of organizing and conducting business activities” (Holden, 2004). North and Thomas (1973) break down transaction costs into three main categories: search, negotiation, and enforcement costs. Search costs are those related to locating an appropriate exchange counterparty; negotiation costs are the those involved in determining and securing mutually acceptable terms of trade; and enforcement costs are those related to enforcing the terms of the exchange, usually in the form of legal fees and taxes that support the judicial system or other conflict resolution mechanism.

Most definitions of transaction costs implicitly assume that the item being exchanged already exists and simply needs to be transferred from one owner to another. Assuming this, one can use a highly restrictive definition that simply includes the costs of transferring ownership and obtaining (or releasing) secure title to the item. However, the determination of what expenses should be included under transaction costs also depends on how one defines the beginning of a transaction. In a home sale, for example, real estate broker fees, title search fees, transfer taxes, advertising fees, and the like are clearly transaction costs associated with transferring ownership, but how does one categorize the cost of “making a house attractive for sale”? These costs would not normally be incurred without anticipating a transaction, and although one could argue that they are somehow “optional”, not taking them simply leads to a separate cost: accepting a lower price. Depending on where one defines the “start” of the transaction, “ready-for-sale” expenses could be viewed as asset improvements or as additional transaction costs.

With simple title transfers, greenhouse gas markets are facilitated by exchanges such as the Chicago Climate Exchange (CCX) and the European Climate Exchange (ECX), which reduce search costs substantially by providing a focal point for buyers and sellers to meet. Exchanges also lower negotiation costs by providing standardized contracts and historical price information to facilitate negotiation.8 Private over-the-counter exchanges of CERs or other GHG reductions are also possible, although the quantity of GHGs exchanged under this mechanism is difficult to estimate, because over-the-counter agreements are not public documents.9 Over-the-counter agreements for virtually all assets tend to have higher search and negotiation costs than exchanges, unless both parties have well established reputations and track records, but the presence of exchanges tends to ensure that over-the-counter exchanges only take place when the benefits are expected to outweigh additional transaction costs for both parties. In both the exchange and over-the-counter systems, enforcement costs are primarily associated with delivery risk. Existing, verified GHG reductions can be purchased on the spot market and have relatively low delivery risk. Some contracts specify forward or future delivery of GHG equivalents on a specific date, and these items are subject to much greater delivery risk.

In the clean development mechanism, CER title exchange costs are likely to become increasingly important as more CERs are exchanged, but there is presently greater concern over “pre-exchange” costs that increase the final cost of saleable CERs above the simple cost of removing GHGs from the atmosphere. Some observers might prefer to call these “costs of production” rather than “transaction costs”, but categorizing these expenses as transaction costs still makes sense because CERs are administratively created goods. Unlike a factory that owns its computer after assembling it from components, a CDM project may successfully remove GHGs from the atmosphere, yet cannot secure title to the corresponding CERs without engaging in further legal and administrative processes (e.g., verification). These are costs that producers face simply to obtain title to CERs they have themselves produced, and they can be thought of intuitively as “costs in the supply of CERs which cannot be attributed to the physical process of removing greenhouse gases from the atmosphere”. A simple test is suggested for determining whether specific expenses should be considered as transaction costs.

Simple test for determining if an item should be considered a CDM transaction cost:

If the project were undertaken as is, but without this expense, would it release additional greenhouse gas?

If yes: The expense is an operational cost contributing to the reduction of GHGs.

If no: The expense is a transaction cost required to obtain title to CERs.

In an earlier work on this topic, McKloskey et al. (2005a) attempt to define CDM transaction costs as “that part of a CER's price that cannot be attributed to the physical process of removing GHGs from the atmosphere”. However, this formulation neglects the fact that CER market prices are also affected by the level of demand, and therefore requires adjustment to separate the transaction cost effects from the demand effects.

Clean development mechanism transaction costs are:

Components in the price of a CER that cannot be attributed to either:

  • 1The physical process of removing GHGs from the atmosphere; or
  • 2The level (or changes in the level) of demand for CERs.

This supply and demand based formula is more challenging to apply in practice, because it requires estimating demand curves and their effect on observed prices. However, it has the advantage that it can be used with market prices, which are observable. The definition that focused only on transaction cost contribution to CER supply costs is easier to explain, but it requires access to supplier cost data that are not readily available.

5. The effects of transaction costs

Transaction costs are undesirable, as they consume resources that could be spent on productive activities elsewhere or to lower end prices. In terms of production possibilities, most economists take them, like taxes, to be a net drag on the economy.10 Ronald Coase's famous article (1960) points out that if transaction costs can be reduced to zero, decentralized market outcomes will be efficient or Pareto optimal. In the presence of transaction costs, however, market outcomes depend on the rules and the structure of surrounding institutions, which have the effect of giving some actors more influence11 over outcomes than others and shift market equilibria away from the original “optimum”. Moreover, these institutions themselves require resources to create and maintain, absorbing resources from other productive activities.

Nonetheless, it is important to emphasize that although transaction costs are real costs to producers, consumers, and economies, they are not all bad. As useful as markets are, they cannot function without institutions that — at minimum — enforce contracts, disseminate information, and resolve disputes. Certainly, few practitioners believe that a system of exchangeable credits, that represent the absence of mostly invisible GHGs in the atmosphere, could function without some credible monitoring and verification system, and this system is unlikely to be costless. The challenge is to keep these and other costs low enough that the system can achieve cost-effective reductions of GHGs while sustaining the institutions that are necessary to make the results believable.

Transaction costs affect the economics of the clean development mechanism in two ways: first, through effects on CER creation (i.e., supply) costs, and second, through effects on CER trading costs. At this writing, with most CDM projects in only their first or second year of operation and others waiting to be approved, the effect of transaction costs on CER creation receives more attention because it affects project developers’ immediate decisions as to whether or not to use the CDM. If transaction costs raise the cost of CER creation, fewer CDM projects will be attempted, resulting in fewer technologies being transferred to non-Annex B countries, less foreign investment achieved, fewer CDM-related GHGs being removed from the atmosphere, higher costs for GHG reductions elsewhere, less international confidence in the emissions trading system, and, by extension, less confidence in the effectiveness of the Kyoto Protocol. For these reasons, it is essential to keep transaction costs as low as feasible, while maintaining the integrity of the GHG credit exchange system.

Figure 2 illustrates the effect of transaction costs on the supply and demand for certified emission reductions, showing supply and demand curves for two markets in equilibrium. Figure 2A diagrams a hypothetical CER market assuming no transaction costs. Without transaction costs, the market achieves equilibrium at point E0, removing Q0 tonnes of GHG from the atmosphere and selling those GHG credits at a market price, P0.

Figure 2.

The effect of transaction costs on the supply and demand for CERs.

Figure 2B shows the same market, this time with transaction costs included. These costs would include the costs of registering a CDM project in accordance with CDM rules, paying verifiers to certify that CERs represent real GHG reductions, and additional costs related to legally receiving CER titles. Together these costs shift the supply curve upward and to the right: creating the same number of CERs will be more expensive with transaction costs included, or, conversely, fewer CERs will be produced for a given price. Together these effects shift the market equilibrium to ET, where QT tonnes of GHG are now removed from the atmosphere, and the market rises to PT.

Shifting from E0 to ET, the key observation is that transaction costs reduce the quantity of GHGs removed from the atmosphere, and raise the market price of CERs. The lower quantity of GHG removal reduces the effectiveness of the CDM as a tool for addressing climate change, and the higher prices reduce the effectiveness of the CDM as a mechanism for removing GHGs at least cost. Whether the total revenue transferred in the presence of transaction costs (PT × QT) is higher or lower than the revenue without costs (P× Q0), depends on the specific shapes of the supply and demand curves. For the project developer, however, transaction costs result in lower total profits, reducing the attractiveness of undertaking the project in the first place. If total revenue increases, they are offset by even higher rises in per-CER transaction costs.

6. Supply and demand dynamics under the CDM

Analysis of supply and demand within the CDM framework is more involved than supply and demand analysis for traditional commodities. As with traditional products, events can cause CER supply curves to rise or fall, as well as demand curves. An analysis of these events needs to consider the impact on not only CER producers (CDM project developers) and buyers (Annex B countries and some market intermediaries), but also on intermediary holders of CERs, as well as the CDM itself. Relevant effects include:

  • • How much GHG is being removed from the atmosphere via the CDM?
  • • How much technological transfer is the CDM stimulating?
  • • To what extent is the CDM proving to be a viable mechanism for addressing either climate change needs or non-Annex B country needs for technology transfer, foreign investment, and environmentally beneficial revenue?

Table 2 summarizes the changes that can take place and their effect on the CDM system. It breaks down the CDM system into CER producers, holders, and buyers, and summarizes the effects of changes in CER supply cost and demand. It also gives an overall assessment of effects on the CDM system as a whole. CER producers are actors who engage in removing GHG from the atmosphere and receive CERs for these reductions from the CDM Executive Board. CER holders are actors in possession of verified CERs who either plan to sell them or use them against their own excess emissions at some point in the future. CER buyers are those who plan to purchase CERs, either to offset excess GHG emissions or to hold them for resale at a future date. These classes are functional and not mutually exclusive. For example, most CER producers will also be CER holders in the period of time between receiving CERs and transferring them to other holders. Similarly, CER buyers are simultaneously CER holders if they intend to resell their CERs or if they plan to bank them for use against later emissions.

Table 2. Effects of supply and demand shifts on CDM participants
EventCER ProducersCER HoldersCER BuyersCDM System
  1. Notes:

  2. CER producer: Reduces GHG emissions and receives CERs from the CDM Executive Board per CDM.

  3. CER holder: In possession of valid CERs and plans to resell later or use them against GHG emissions.

  4. CER buyer: Intends to buy CERs for use against GHG emissions or for later resale.

CER supply costs increaseBadGoodBadBad
(e.g., transaction costs up)
CER Quantity: ↓ Price: ↑
CER supply costs decreaseGoodBadGoodGood
(e.g., new technologies)
CER Quantity: ↑ Price: ↓
CER demand increasesGoodGoodBadGood
(e.g., other GHG reductions hard)
CER Quantity: ↑ Price: ↑
CER demand decreasesBadBadGoodBad
(e.g., other GHG reductions easy)
CER Quantity: ↓ Price: ↓

Producers of CERs benefit from increases in CER demand and decreases in CER supply costs. Demand increases will result in a larger number CERs available for sale and a higher price per CER. Producer revenue increases substantially. Other things equal, supply cost reductions will result in a larger number of CERs available for sale, albeit at a lower price. In general, CER producers will benefit from lower supply costs, because price reduction represents fewer total expenses, not necessarily lower profitability.12 The reduction in prices also tends to increase the demand for CERs relative to other GHG reduction methods, which benefits CER producers.

Buyers and holders of CERs have opposed interests linked to the movement of CER prices. Buyers would like to see CER prices fall to keep purchasing costs down. Holders who intend to sell CERs would like to see CER prices rise, so that larger profits are realized at sale time. Holders who intend to use CERs to cover future excess emissions benefit somewhat from price rises, because it is less expensive to use CERs already purchased than it would be to buy additional CERs at the new market price. The advantages to these holders could be short lived, however, should they need to purchase new emission credits to replace the credits already held.

The CDM system as a whole benefits most when the quantity of CERs produced is high and the cost to produce them is low. This means that the system is removing large quantities of greenhouse gases at low cost, providing Annex B countries with the benefit of inexpensive emission reductions and non-Annex B countries with technology transfer as initially envisaged. Less beneficial but also good for the CDM is when a large quantity of CERs is produced, but the price is relatively high. This means that the system is successfully reducing GHG intensity in non-Annex B countries and providing investment and technology transfer, but not providing as cost-effective a solution as possible for the Annex B countries.13 Low CER quantities at low prices suggests that the system is not producing a large number of emission reductions and is not extremely useful because other GHG reduction strategies are available, whereas low CER quantities at very high prices suggests that the system is too expensive to use effectively and is not succeeding in reducing emissions substantially.

Table 2 shows that the health of the CDM system is most closely aligned with the interests of the CER producers. Ideally, CER producers would like to obtain high prices, and the CDM system would prefer to have the market clear at a lower price point, but both have an interest in having the system generate large quantities of saleable CERs, as long as doing so does not come at the cost of undermined buyer confidence in the validity of CERs. For this reason, some transaction costs are necessary, even if they reduce the total number of CERs generated and raise the market price.

7. Tracing transaction costs: the CDM process

CDM transaction costs will never be eliminated entirely because the key commodity — Certified Emission Reductions — is unusual and requires careful monitoring, verification, and evaluation before anyone would pay for it. Unlike crude oil, gold, silver, or contracts, a CER represents the absence of a quantity of greenhouse gases in the atmosphere. Not only does a CER represent the absence of greenhouse gases, but most of these gases are invisible even when they are present. This creates additional challenges to proving that they have been reduced. Finally, this absence is not even an absolute absence of GHGs; rather, it is the absence of gases relative to a theoretical baseline of a project that was, in many cases, not undertaken to begin with. A CER does not have the concrete reality check of an exchange for physicals (EFP) delivery that oil or gold have, since it is impossible to deliver an “absence” to a specific holder. If people and governments are expected to pay for CERs, there must be procedures and checks to prevent the many ways one could create meaningless CERs. Without these, CERs would cease to be credible units of greenhouse gas reductions and the effort would fall apart. However, these procedures necessarily create transaction costs.

CDM projects will pass through up to eight stages from initial conception to the issuance of marketable CERs. The stages are summarized in Box 1, diagrammed in Figure 3, and described in more detail below. Each step of the process fits the definition of a transaction cost: they are costs in the creation of a CER that do not contribute directly to the reduction of greenhouse gases, and which are not directly determined by changes in the demand for CERs.

Table Box 1: . Stages in CDM project development
1. Project concept and initial design.
2. Develop new methodology or use existing measurement methodology.
3. Prepare CDM project design document.
4. Have project validated by designated operational entity (DOE).
5. Have project approved by host country designated national authority (DNA).
6. Register project with CDM Executive Board
7. Monitor and verify emission reductions using an accredited DOE.
8. Receive certified emission reductions.
Figure 3.

Diagram of the CDM project development procedure
Source: Diagram taken from McCloskey (2005b).

The process begins with a project concept and initial design effort. This is one step that nearly any project would require. This stage identifies opportunities to use the CDM, the main participants, necessary technologies, the scale and magnitude of expenses, and estimated payoff streams and revenues. These calculations may be done by in-house staff, but usually involve hired consultants who are specialists in CDM technologies and legal requirements. Three key questions must be answered at this stage:

  • • Is the project technically feasible?
  • • If undertaken, will it be economically profitable without the sale of CERs? and
  • • If not, is it sufficiently close to breaking even that the sale of CERs would make it profitable?

As shown in Box 2, for a project to be viable under the clean development mechanism, the answers to these questions must be:

Figure Box 2:  .

Decision flowchart for pursuing a CDM project

  • • Yes, the project is technically feasible.
  • • No, the project would not make a profit without revenue from CERs, and
  • • Yes, the project is sufficiently close to breaking even that the sale of CERs would make it worth undertaking.14

At stage two, projects must decide on a methodology to use to determine the level of greenhouse gas reductions envisaged. In order to be CDM registered and receive CERs, projects must use a baseline and measurement methodology that has been approved by the CDM Executive Board in Bonn. The baseline and measurement methodology is the method that defines the standard and means by which project emissions and reductions will be measured. The baseline sets the “business as usual” scenario and projects the quantity of GHGs that would be released under that scenario. The higher the baseline emissions rate is set, the more CERs will be generated from the project, so project developers have an incentive to make baselines as high as possible. Since the baseline also determines the total emissions that will go into the atmosphere and the primary purpose of the Kyoto Protocol is to reduce total GHG emissions, the UNFCCC and CDM technical staff have an interest in preventing baseline overstatements.

The least expensive option at this stage is to use one of the approved existing methodologies listed on the CDM website of the UNFCCC. This would allow a project developer to proceed directly to the next step. However if existing methodologies are somehow not useful or appropriate for the project, project developers may propose their own methodology and submit it for Executive Board approval. This process is both time consuming and expensive, generally involving highly paid specialists and likely substantial rounds of travel and negotiation with UNFCCC officials for approval, but the costs are arguably necessary if a project has an innovative approach to reductions not covered by existing methodologies.

Once developers have designed a project and selected a baseline and measurement methodology, they must prepare a CDM project design document (PDD), which is an official UN CDM document that will be evaluated by at least three key entities before it can be approved and registered. In addition, the PDD requires one or more periods of open commentary for community stakeholders, who might be affected by project activities. Any substantial complaints must be addressed in a revised PDD, possibly involving reformulation, to be resubmitted. If a project has been well designed and implemented with appropriate stakeholder consultation, the process is not especially difficult, but it has the potential to be time consuming and delay project start-up for some time.

As mentioned above, once a PDD has been created, the project must be approved by at least three entities — a designated operation entity (DOE), a designated national authority (DNA), and the CDM Executive Board — before it can be registered with the CDM and receive CERs for emission reductions. The DOE and the DNA approvals come in any order, but both must be secured before Executive Board approval will be given, and some DNAs require a DOE's approval before they will give theirs. A DOE is an organization accredited by the CDM Executive Board as having the necessary legal, economic, and scientific expertise to determine whether a project's methodology is sound and whether its GHG reductions fit the requirements of the CDM. The purpose of the DOE is to verify the science behind greenhouse gas reduction and to ensure that GHG reductions represent “real, verifiable, emission reductions that would not have taken place in the absence of the CDM.”15

Designated national authorities exist to ensure that host countries retain sovereignty over the use of the CDM in their territory. Their main role in the approval process is to ensure that CDM projects are consistent with national sustainable development objectives, as stated in the Protocol, and, particularly, to ensure that CERs are not issued for activities contrary to government policy. The DNA's role as a gatekeeper for the host country can be important, since the projects approved can affect future definitions of “business as usual” and the baselines to which national projects are held accountable. If the DOE's function is to establish the scientific legitimacy of the project, the DNA's function is to establish a project's political legitimacy.

After relevant DOEs and DNAs have approved the project, the CDM Executive Board will register the project for the CDM and charge a registration (“adaptation”) fee, based on project scale and the quantity of CERs anticipated per year. The Executive Board has the authority to review any project in greater detail and demand modifications, but in many circumstances, they defer to the decisions of DOEs and DNAs. If national or international non-governmental organizations (NGOs) oppose the project and have failed to convince the local DNA to withhold approval, they may try to pressure the Executive Board to intervene at this stage. So far, the Executive Board has tended to listen to such challenges only to the extent that they address the scientific or economic basis for measuring GHG reductions and to the extent that host countries are perceived to have used insufficient consultative procedures with project stakeholders.

Once registered, a project is now able to receive CERs for greenhouse gas reductions, but these reductions must still be achieved and verified before CERs can be issued. Verification is achieved using the methodologies established in the PDD and using an accredited DOE to verify that the claimed reductions have been achieved. The services of a verifier, as with an auditor, must be paid for, as with any other service. A project can receive saleable CERs only after it is registered, underway, has achieved emission reductions as planned and had those reductions verified.

Each step in the process adds costs that do nothing to remove additional greenhouse gases, but that are essential to ensure that CERs on the market have both scientific credibility and are politically acceptable to project host countries. The expectation is that over time, the process will become more standardized, and the costs of individual steps will decline, creating a more efficient and streamlined approval process and therefore lower transaction costs.16 These costs can still be substantial, however. The next section presents a case study illustrating the effects of these costs on project feasibility.

8. Transaction costs in practice: an illustration from Ghana

The following case study helps to illustrate the effects of the CDM process on transaction costs. The case involves a proposed project to substitute liquefied petroleum gas (LPG) for wood and charcoal for home cooking.17 Although LPG is a fossil fuel, its heat can be channelled and controlled more efficiently for use in home cooking and heating than that of wood and charcoal. As a result, LPG can replace enough wood to reduce the total quantity of CO2 released per family per day, reduce the pressure on local forests, and provide technological and political benefits by connecting families to a higher quality energy infrastructure. Managing the project could be difficult, however, since creating the initial distribution infrastructure requires a substantial investment, and individual families must invest in new domestic heating hardware.

As part of a master's degree exercise, Alexander McCloskey (2005b) gathered data and prepared an analysis of the effect of each stage on project profitability. His estimates suggested that a well run project could reduce emissions by approximately 2700 tCO2e per year compared to current wood fuel emissions, which in this case forms an observable baseline. At a sale price of $20 per tonne of CO2 equivalent,18 such a project could count on revenues of approximately $54,000 per year from the sale of CERs.

Recall that in order to be eligible for CERs, the additionality criterion requires that CDM projects be non-viable in a business as usual scenario. Effectively, this means that the LPG project cannot make a profit without the revenue from CERs, and therefore, $54,000 represents the maximum expected annual profit achievable at $20 per tCO2e. In addition, assuming that project developers choose to avoid projects that lose money,19$54,000 per year is also the largest annual pre-CER deficit that developers will be willing to tolerate, unless they expect tCO2e prices to rise. Together, these criteria mean that a workable CDM project will have pre-CER cash flows no greater than zero (otherwise it is a “business as usual scenario”), and no less than negative $54,000 (in which case the project would not be profitable even if CER revenues are included).

McCloskey's analysis also looks at methodology and project preparation costs, estimating that initial project concept and ground testing would cost approximately $40,000 in research, consultancy, and pilot testing. Further development of the concept into a project design document would cost $35,000 in time, consultancy, and legal fees, as well as to cover community outreach and participatory workshops. DNA approval was expected to be relatively inexpensive at $5,000, since the DNA is a public agency supported by public funds. The DOE verification step was substantially more costly ($40,000), as DOEs require uniquely qualified professionals and perform their work on a fee-for-service basis. With these processes complete, McCloskey estimated CDM registration costs at $10,000, broken down into $5,000 for final preparation and $5,000 for the registration fee itself.20

Aside from document preparation, registration, and approval costs, McCloskey noted that existing methodologies did not fit requirements of the proposed project. To move forward, a new CDM-acceptable methodology would need to be developed, vetted, and approved by the CDM Executive Board to define a method for determining the business-as-usual baseline, monitoring decentralized consumption of LPG, ensuring that fuel switching occurs as planned, and accounting for potential sources of GHG leakage.21 His estimate was that development, testing, submission, approvals, revision, and registering could cost up to $200,000, by the time consultant fees, travel, pilot test data, DOE testing and validation are completed. This would be a substantial up-front cost, although if many projects could benefit from a similar methodology, it might be possible to find public or other support for these expenses.

McCloskey estimated that total up-front costs would be $330,000 if a new methodology were developed, but only $130,000 if the project could use an existing methodology. The project would then receive $54,000 per year in revenue from sales of 2700 tCO2e at a price of $20 per CER. After monitoring and verification costs of $8,000 per year provided by a DOE, the CDM mechanism ends up providing an estimated net cash flow of $46,000 per year. A summary of estimated expenses and revenue is shown in Box 3.

Table Box 3: . Summary of Ghana LPG transaction costs
Up-front costs
Summary costsDetail
  Project concept $40,000
Project documents $75,000Methodology development$200,000
Approvals and registration $55,000Project design document $35,000
Methodology development$200,000DOE validation $40,000
Total up-front costs$330,000DNA approval  $5,000
(Costs without methodology)$130,000CDM registration $10,000
Total up-front costs$330,000
(Costs without methodology)$130,000
Annual recurring costs and income
Monitoring and verification($8,000) per year
2700 tCO2e CERs$54,000 per year at $20/tCO2e
Net cash flow at $20/tC02e$46,000 per year

Should this project be undertaken? The answer depends on assumptions that go into the model: sunk start-up costs, annual monitoring costs, total project life, the number of CERs actually generated, and the sale price of a CER. The Ghana study helps to identify the steps that project developers must consider when deciding whether a CDM project will be viable. These steps illustrate transaction costs over and above the cost of project engineering, and the final decision depends on several project particulars.

Of the model assumptions, the most difficult may be to estimate the final sale price of a CER. This factor is critical, because CER revenue determines the project's maximum obtainable profit. The higher the expected CER revenue, the larger the range of acceptable project cash flows.

Figure 4 uses McCloskey's estimates of project costs and CER production to calculate project internal rates of return (IRRs) at different CER sale prices, assuming a five year project life (2008–2012, inclusive). Corporate finance theory says that projects provide net value to their owners if the IRR is greater than an investor's required discount rate for similarly risky projects, which for company managements is usually calculated as the company's average cost to raise capital or funds from all available sources.

Figure 4.

Ghana LPG example: Internal rates of return at different CER prices.

The figure gives IRRs as a function of CER sale price under two scenarios:

  • • First, a project with methodology development costs;
  • • Second, an identical project assuming that a suitable methodology already exists.

Unsurprisingly, the present value of the LPG project increases when the CER sale price rises: how fast it increases depends on the size of initial start-up costs. The net value of a project requiring a new methodology rises more slowly than a project that can use an existing methodology. Below a critical CER price — where the curves cross the x-axis — IRRs are negative, indicating that no project would be desirable without additional subsidies from other (presumably public) sources. If the IRR curve crosses the x-axis, a project becomes economically attractive when it rises above the developer's cost of raising capital. If the cost to raise capital is 10%, for example, the Ghana project would begin to be attractive at about $16 per tonne if there are no methodology development costs, but at $35 per tonne if a methodology needs to be developed.

Figure 4 shows several effects. First, the higher the up-front transaction costs (illustrated by the difference between paying for methodology development and using an existing methodology), the higher the CER market price needs to be before projects become attractive. Secondly, if the project could generate a larger number of CERs per year, both curves would become steeper and shift to the left, making the project viable at a lower CER price. This suggests that large-scale projects may achieve economies of scale that render transaction costs relatively insignificant,22 but small-scale projects are likely to face considerable challenges. Additionally, since methodology costs are likely to add a substantial set of up-front costs, public measures and policies to improve the availability of CDM methodologies will clearly go a long way to reducing up-front costs that make using the CDM difficult.23 In any case, there is good reason to expect that, over time and as more projects are registered, the approval process will be streamlined and a wider range of methodologies will be available, lowering initial transaction costs and allowing the CDM system to function more efficiently.

9. Preliminary market evidence on the price of carbon equivalents

If CER prices are critical to project viability, what, then, are reasonable prices to expect? Carbon and carbon equivalents have not been trading long, but some data are now available from existing exchanges. Figure 5 shows the European Climate Exchange (ECX) daily settlement prices24 for one tonne of CO2 equivalents deliverable under the European emissions trading scheme (ETS) in December 2006 (solid) and December 2008 (dotted) over the period April 2005 to August 2006. The ETS is a European Union regional GHG cap-and-trade system, designed to reduce business emissions within the EU so that EU countries achieve domestic compliance with their Kyoto Protocol targets.25

Figure 5.

Price of 1tCO2e on the European Climate Exchange, April 2005 to August 2006.

In the absence of extensive data on CER trading, the ETS offers insights on how CER prices might behave. Under ETS, businesses within specific carbon-intensive industries are assigned a quantity of European Union allowances (EUAs), valued at one tCO2e each. Companies are allowed to trade allowances as necessary to stay within emissions guidelines. The EU Linking Directive also allows ETS companies to purchase CERs and other Kyoto-approved credits and surrender them to an EU government in exchange for an equivalent number of EUAs, although few if any such exchanges have taken place.26 This substitutability does suggest that, over the long term, CER prices should approximate EUA prices, with a small discount to cover the minor administrative (transaction) costs of converting them to EUAs.

Figure 5 shows that carbon prices have ranged from highs of nearly US$40 to lows of almost US$11 per tonne of CO2 equivalent since ECX trading began in April 2005. In general, prices have tended to rise over time, but a sudden collapse in prices (70% for December 2006 delivery) took place in late April 2006 after the news that European countries had fewer-than-expected emissions for the first quarter of 2006, suggesting that total demand for carbon credits could be lower than anticipated. The event provides a concrete example of the demand collapse effects discussed earlier. Since the April 2006 price correction, tCO2e prices have nudged upwards to trade at close to $20/tCO2e at present.

Despite the price collapse, EUA prices are still likely to increase over the long term, as inexpensive domestic reductions in Annex B countries are exploited and it becomes more expensive to find additional sources of emission reductions. The emissions market may be showing this expectation by the fact that allowances for delivery in December 2006 declined by nearly 70% between 19th April and 6th May 2006, but only 45% over the same period for December 2008 delivery, and 2008 futures have been trading at a premium since.27

Daily data on CER prices are not presently available, due to low liquidity and the fact that verified CERs cannot be used for Kyoto targets until the start of the first compliance period in 2008. Nonetheless, Figure 5 does show the range of prices that “high quality” emission reductions can command in the present environment. As mentioned, since CERs can be used to acquire EUAs on a one-for-one basis, they should approximate EUA prices over time and as CER liquidity increases. In fact, the CER appears to be one of the most tradable emission credits in existence — valid for meeting Kyoto obligations, ETS, and substitutable for verified emission reduction (VER) agreements listed on the Chicago Climate Exchange (CCX).

Presently however, CERs trade at a substantial discount to EUAs, often by 50% or more. This is partly because few CERs have been issued thus far,28 impairing liquidity, and also because the CER market faces uncertainties not faced by EUAs. Limited CERs in circulation means that most CER-related contracts must be forward contracts for CERs not yet in existence, creating substantial risk of non-delivery, since project CERs may not actually be issued if the project fails to perform for any number of reasons. By contrast, EUAs already exist and have been allocated to relevant businesses on a known model, so there is little risk that EUAs will fail to materialize. There is sufficient supply and demand to create a relatively liquid spot and futures market that can be tracked daily. Finally, CERs are discounted because of regulatory and country risks that come from doing business in primarily developing countries, and from a few remaining uncertainties about the final administrative rules under the CDM, though the latter has diminished substantially in recent years. As a result, while Figure 5 shows recent EUA price trends, CERs presently trade for substantially less.

Figure 6 shows a rough approximation of the discount to CERs, comparing high, low, and average prices for CER contracts, as reported by the World Bank (2006), and high/low/average prices for EUAs, taken from ECX daily data over 2005 and the first quarter of 2006.29 Due to methodological differences, the two data sources are not perfectly comparable,30 but do give an approximation of the discount applied to CERs. Most noteworthy is the difference between the price of CERs in primary and secondary markets. Primary markets are markets in which CERs are created and initially sold; secondary markets are exchanges of already-existing CER contracts. Secondary markets have substantially less delivery risk, since they usually trade CERs that have already been issued or are otherwise guaranteed. The data show that primary market CERs trade at a discount of approximately 65% to 75% to EUAs, while secondary market CERs trade at a discount of 20% to 25% to EUAs over similar periods. The difference in discount between primary and secondary markets primarily reflects delivery risk. Whereas the discount in secondary markets may reflect insufficient demand leading to liquidity risks, and the possibility of rule changes impairing the substitutability of CERs, the substantially larger discount to primary markets reflects the possibility that CERs may simply not appear in the first place. Although the discount to primary market CERs is substantial, the data also indicate that prices in the primary market have seen the most rapid appreciation over the period covered.

Figure 6.

High, low, and average ranges for emission credits 2005 and 1st quarter 2006.
Primary and secondary CER markets compared with EUA/ECX prices.
Sources: Primary and secondary CER market prices from the World Bank (2006). EUA prices taken from European Climate Exchange and converted to dollars at US$ 1.22 per euro.

What do these prices indicate for a project like the Ghana LPG discussed earlier? Using a primary market discount of 65%, a general investor would need to see EUA prices of at least $36 per tCO2e, before the investment would start to look attractive, assuming that no new methodology would need to be approved.31 If the project developer must cover development of a new methodology, carbon prices would need to approximate $80 per tonne before a general investor would start to see positive returns.

On the other hand, if project investors have unique specialized knowledge that make them highly confident that CERs will in fact materialize and be tradable, they might use the secondary market discount of 25%, which would indicate a profitable investment at a minimum price of about $18 per EUA if no new methodology is developed, and $37 per EUA if new methodologies are required.

If these price conditions and present trends are representative, they indicate that most investors would not find the Ghana LPG project a profitable investment, since EUA prices are far away from the minimum $35 per tonne and may not reach that figure for some time. For informed specialists who have very high confidence in the project's ability to deliver registered CERs, the current price environment would favour investment, provided the project developer does not need to cover the costs of a new methodology. If a new methodology is necessary, then EUAs would need to be trading at close to their historical highs before the Ghana LPG project would start to appear attractive, even to the confident specialist. In short, assuming that the CER output is accurately estimated, the sample project in Ghana is likely be profitable enough to overcome transaction costs from the CDM approval process, but not profitable enough to overcome the full cost that includes the development of a new methodology.

10. Conclusions

This article has described transaction cost issues under the clean development mechanism and illustrated procedures for analyzing project viability in the face of these costs, using a sample project from Ghana and recent GHG price data. Transaction costs are an unavoidable part of nearly all economic activity, but they raise special concerns with the CDM because the additionality criterion places highly restrictive conditions as to what projects may receive CERs and still be economically feasible. The additionality aspect magnifies the effect of transaction costs on the viability of specific CDM projects, since projects cannot easily rely on non-CER income sources to dilute them. Ultimately, only projects with a projected net present value falling into a narrow range will be both eligible for the CDM and economically attractive to developers. The width of that range is equal to the NPV of the CER revenue stream, which in turn depends on three factors:

  • • The expected quantity of CERs;
  • • Anticipated CER market prices; and
  • • The size of transaction costs.

The higher the transaction costs, the narrower the range of viable projects, and the less successful the CDM is at achieving its objectives.

The article identifies an additional effect of CDM baseline scenarios. Such structures determine how much GHG will ultimately go into the atmosphere as a result of the project.32 If the market for CERs clears, any emissions below the baseline will be emitted elsewhere by the CER's final purchaser, except when either

  • 1) Project emissions are higher than the baseline and no CERs are issued; or
  • 2) CERs are purchased by environmental organizations for the purpose of reducing the total number of emissions available on the market.

In the first case, total emissions will be higher than the baseline scenario; in the second, total emissions will be lower. In other cases, total net emissions will exactly equal the baseline.

Some transaction costs are simply a result of the process of transferring CER titles from one owner to another and can be measured though bid-ask spreads and broker fees on GHG exchanges. As trading expands, these costs are likely to become more important. However, currently the most substantial transaction costs arise from administrative processes that CER producers must pass through in order to obtain initial title to the CERs they have created. These costs relate to project document preparation and approval, monitoring and verification, and, in some cases, design and development of a new methodology. For small-scale projects especially, these costs can be substantial and may impede project viability. Bundling projects together for joint approval is one way to achieve economies of scale and reduce the impact of transaction costs on individual projects. The CDM Executive Board is also examining methods to streamline the approval process and reduce its contribution to transaction costs through measures, such as waiving the adaptation fee for small-scale projects in developing countries.

Transaction costs will never fall to zero, however, because CERs by nature demand vigilant monitoring to safeguard the Kyoto Protocol's contribution to environmental and climate integrity. A CER cannot be pointed to or touched, as can a barrel of oil, a bar of gold, or a plot of real estate. It is difficult to “see” the absence of a tonne of carbon dioxide equivalent. As a result, the costs of enforcing CER property rights are higher than for ordinary commodities and include more costly verification methods. Without these processes, CERs would cease to be a believable commodity, defeating the purpose of the CDM both as a method of technology transfer and as a method of reducing greenhouse gas emissions at least cost.

Although transaction costs are unavoidable, there is still a need to keep them as low as possible. High transaction costs raise the price of CERs and ultimately lower the quantity of GHG removed by the CDM. This undermines two CDM objectives — the desire to reduce GHG emissions and the desire to accomplish these at lowest cost. For this reason, finding ways to keep track of and reduce transaction costs will continue to be critical. It is certainly possible that high transaction costs could render the CDM unusable. Indeed, if the price of reducing emissions is too high, there is a possibility that the Kyoto Protocol itself could expire without renewal in 2012, at the end of its first compliance period.

The good news is that these transaction costs are likely to diminish over time, and the creation of profit centres from GHG reduction may create political–economic forces that support Kyoto's continuation. With respect to transaction costs, as more projects move through the approval pipeline, document preparation processes are likely to become more streamlined. Greater experience with CDM economics and processes will lead to fewer wasted efforts and more efficient methods of preparing a project document. As more projects are approved, a larger number of acceptable methodologies will become available, reducing the number of projects that will need to consider substantial costs for creating a new methodology. Methodologies developed and approved for joint implementation projects may also become acceptable for CDM use. More experience will likely improve the confidence with which one can predict the issuance of CERs, and the delivery risk discount will diminish. Finally, there is reason to believe that, over time, the price of CO2 equivalents will rise as Annex B countries begin to exhaust the least expensive options for reducing emissions. All of these factors suggest that, while transaction costs are a substantial concern for today's projects, over time, the CDM will become more and more feasible in linking the climate change obligations of developed countries to the investment and sustainable development needs of the developing world while still addressing the challenges of climate change.


  • Bruce Chadwick is an adjunct research scholar in environmental science and policy at the School of International and Public Affairs, Columbia University, New York, NY. E-mail:

  • 1

     Annex B of the Kyoto Protocol corresponds very closely to Annex I of the UN Framework Convention on Climate Change that was signed in Rio de Janeiro in 1992. For this reason, many analyses refer to Annex I and non-Annex I countries interchangeably with Annex B and non-Annex B countries. Annex B contains countries and their specific reduction commitments and does include a few countries which were not fully independent in 1992.

  • 2

     Joint Implementation and CDM projects can also be structured between single countries and a private organization.

  • 3

     The exact amount is computed by estimating the country's emissions for the baseline year, reducing that figure by the commitment percentage listed in Annex B. This gives the allowable emissions for a single year. That number is then multiplied by five to allow for total emissions during Kyoto's first commitment period (2008−2012).

  • 4

     There is some debate as to whether the text of the Protocol permits governments to license other entities (such as, businesses, non-profit organizations, and others) to hold and transfer AAUs, but in practice, AAUs are anticipated to be exchanged between governments.

  • 5

     This assumes that CERs are eventually purchased by Annex B countries to offset emissions above their Kyoto obligations. It is theoretically possible that CERs could be purchased by an environmental (or other) group for the purpose of removing GHG emission rights from the system. Since these CERs would not be available to use as an offset under emissions trading, holdings by environmental organizations would represent GHG actually removed from the atmosphere.

  • 6

     This effect also applies to joint implementation, although the rules for baseline determination may have small differences.

  • 7

     The additionality criterion makes it more difficult for organizations to diversify their business risk by relying on revenue streams from many components of a CDM project. If it is possible to diversify their business risk, then it is likely that the project is economically feasible without CER revenue and therefore not eligible for CDM status.

  • 8

     As a practical matter, CDM CERs do not presently trade on the CCX because CER providers can obtain higher prices in Europe, where governments have binding GHG reduction commitments under Kyoto.

  • 9

     However, since the ultimate consumers of CERs are likely to be governments bound by Kyoto commitments and with public finance disclosure rules, the quantity of final CERs exchanged in over-the-counter arrangements may be estimable.

  • 10

     Many taxes – for example, tariffs, sales, and value added taxes – are transaction costs.

  • 11

     Or less costly influence, so that one party can shift the outcome to their favour with less effort than the other.

  • 12

     The main exception is where decreasing supply costs result in fewer barriers to entry. In this case, prices are dropping while existing producers lose market share. This is bad for existing producers, good for new producers, who have additional revenue, and generally good for the CDM system as a whole, since more GHGs are being accounted for by CDM activities.

  • 13

     For the entire Kyoto system, higher CER prices create an incentive for Annex B countries to achieve a larger portion of reductions through domestic emission reduction strategies.

  • 14

     If the answer to the second question is “yes, the project will make a profit without the sale of CERs,” then the project is still worth undertaking; it simply will not receive CERs. It is possible that an otherwise marginally unprofitable project will be more attractive than a slightly profitable one, since this will allow capture and sale of CERs.

  • 15

     Paraphrased from Article 12.5 of the Kyoto Protocol.

  • 16

     The 11th Conference of the Parties, held in late 2005, addressed some concerns to streamline the process, including the possibility of retroactive CER creation for slow-moving approvals.

  • 17

     Data collection and initial analysis of the Ghana LPG project under consideration by Ecosecurities, LLC. was conducted by Alex McCloskey, a graduate student for the Masters degree in Public Administration in Environmental Science, Management and Policy at Columbia University.

  • 18

     Twenty dollars per tonne is typical price for European trading system (ETS) carbon offsets as at mid-2006.

  • 19

     Some public-oriented projects could choose not to run on a profitable basis and simply use CERs to offset their losses. These may be subject to special baseline methodology considerations.

  • 20

     The CDM Executive Board has been revising its fee structure to address concerns over transaction costs and especially the effect on small-scale projects.

  • 21

     “Leakage” is when the presence of a CDM project results in additional GHG emissions in other locations, usually outside of a project's area – for example, if wood and charcoal come from local sources, the additional emissions that come from having to transport a non-local fuel source to the project area must be accounted for, or there will be GHG leakage.

  • 22

     To date, large hydrofluorocarbon (HFC) destruction projects have been one of the most cost-effective CDM strategies, given that the greenhouse potential of one tonne of HFC-23 is equivalent to nearly 12,000 tonnes CO2. Projects eliminating this greenhouse gas can generate a highly attractive number of CERs and achieve an excellent economy of scale.

  • 23

     Public subsidies or support for additional CDM methodology development is an obvious policy that will help spur CDM activity, but this is not the only option. McCloskey et al. (2005a) have suggested creating methodology licensing processes to provide incentives for private sector development of reliable methodologies for general use.

  • 24

     Prices are converted to dollars at $1.22 per euro, a typical exchange rate over the period.

  • 25

     The ETS is implemented in two phases: The first phase runs from 2006–2007 and is designed to accustom businesses to the new regulatory environment. The first phase is not entirely practice, however: fines run at 40 euro/tCO2e for each excess emission. The second phase runs from 2008 to 2012 and corresponds to the first compliance period of the Kyoto Protocol. During this phase fines for excess emissions will rise to 100 euro/tCO2e, and possibly higher.

  • 26

     The Linking Directive permits the exchange of other Kyoto-valid emissions credits for EUAs. However, there are strong incentives to bank CERs for use in the post-2008 compliance period, reducing the number of exchanges before then.

  • 27

     Part of the price difference also reflects the fact that fines for ETS shortfalls will rise from 40 euro/tCO2e in 2006–2007 to 100 euro/tCO2e after 2008. Since the cost of non-compliance is higher in 2008, the value of credits in 2008 should be higher.

  • 28

     The first CERs were issued on 20 October 2005 for two hydroelectric projects in Honduras.

  • 29

     Unfortunately, the World Bank report was published just as the EUA price collapse was occurring, so CER price data were not available for the post-correction period.

  • 30

     The World Bank reports a weighted average technique for average price, but does not describe the weighting process in detail. World Bank data were also reported in US dollars, whereas the ECX data were converted from euros using a single “typical” exchange rate of US$1.22 per euro.

  • 31

     From Figure 4, project IRR becomes positive at around $12.50 per tonne. If CERs are discounted at 65% over EUA prices, EUA prices would need to be $35.71/tonne to generate $12.50 per CER. Investors would naturally require a larger expected return than zero, which necessitates even higher carbon prices before investment will flow.

  • 32

     This is not necessarily the case for JI projects, because JI host countries have reduction commitments under the Kyoto Protocol, and these are adjusted in the AAU accounting system.


This article builds on initial research by Alexander McCloskey, Tisha Joseph, Mark Aranha, Amanda Bergqvist, Andrew Dvoracek, Takuya Kudo, Eliot Levine, Amy Lile, Heather Matsumoto, Cindy Pearl, Jessica Rogers, and Reis Lopez Rello, students for the Master of Public Administration degree at Columbia University. The research was presented during a workshop in Environmental Science, Policy and Management at the University in the spring of 2005, and to an audience at the United Nations Division of Sustainable Development. In addition, Alexander McCloskey prepared outstanding scenario analyses for the Ghana LPG case study, which was indispensable for the preparation of this article.