1 ~~~~~E N V I R 0- N M E N T D: E P A R T M E N T -X *^ P A P E R - S . . PAPER NO. 005 TOWARD ENVIRONMENTALLY AND SOCIALLY SLUSTAINBt LE DEVELOPMENT CLIMATE CHANGE SERIES Joint Im.plem.entation. of Climate Change Measures Robert J. Anderson, Jr. March 1995 Environmentally Sustainable Development The World Bank ESD Global Environment Coordination Division Joint Implementation of Climate Change Measures Robert J. Anderson, Jr. March 1995 Papers in this series are not formal publications of the World Bank. They are circulated to encourage thought and discussion. The use and citation of this paper should take this into account. The views expressed are those of the authors and should not be attributed to the World Bank. Copies are available from the World Bank's Environment Department. Global Environment Coordination Division, Room S- 2135. Contents 1. Introduction 2 2. A Brief Description of the Case Study Projects 3 Mexico ILUMEX 3 Poland Coal-to-Gas Conversion 4 3. Basic Joint Implementation Project Design and Measurement Issues 5 Determination of the Net Abatement Effect 5 Determination of Prices of Joint Implementation Projects 8 Performance Issues 15 Procedural/Documentation Issues 16 4. Concluding Observations 19 Annex A 23 Annex B 25 Annex C 31 Climate Change Series 1 Introduction This paper discusses some of the issues that in any other form of implementation as well. may arise in joint implementation of climate The paper considers, but does not propose or change measures under the United Nations endorse specific solutions. Framework Convention for Climate Change (FCCC). The basic idea behind joint The plan of the paper is as follows. Section II implementation is simple: one or more parties briefly describes the two "case study" projects in one or more countries would contract with that provide a real context for subsequent one or more parties in another country to discussion of the issues. One of these projects, implement a project (or, perhaps, modify a ILUMEX, will demonstrate the use of high- policy) for the purpose of reducing emissions efficiency lighting technology in two major of greenhouse gases (GHGs) in that country. Mexican cities. The other, Poland Coal-to-Gas The contracting parties would be able to seek Conversion, will demonstrate the credit for these emission reductions against establishment of institutional and financial their individual obligations for greenhouse gas mechanisms to encourage the conversion of emission reductions under the FCCC. The coal fired boilers to gas in selected situations motivation or incentive for such contracting in which this would not otherwise be would be that the parties to the contract could economical. lower their total costs of meeting these obligations by working together to take Section III examines four classes of issues that advantage of relatively low-cost abatement arise in the joint implementation context: (i) opportunities. determination of the net abatement effect of a policy or project intervention; (ii) Most of the issues surrounding joint determination of prices of or compensation implementation have been aired extensively in due for joint implementation projects; (iii) recently prepared papers or papers now under performance issues such as verification of preparation as background for the FCCC abatement effects and handling of risk; and Conference of Parties' upcoming deliberations (iv) documentation and procedures. In on, inter alia, criteria for joint examining these issues, the paper keeps in implementation.' These papers examine the mind five criteria:2 issues from a policy perspective. The present paper examines some of these issues from a * efficiency: whatever degree of abatement is "bottoms up", project-based perspective, based obtained, good approaches to joint on experience gained under the pilot phase of implementation should promote the GEF. For this purpose, it draws heavily on attainment of that degree of abatement at experience with the preparation of two the lowest possible global expenditure of projects. Its objective is solely to illustrate, and resources; thereby hopefully clarify, some project level * equity: the distribution of gains from joint considerations that may have implications for implementation should be fair; the choices between alternative policies * effective: joint implementation governing joint implementation. Since these arrangements should be consonant with are project level issues, they also clearly arise the attainment of agreed FCCC goals; Climate Change Series I Joint Implementation * robust: joint implementation arrangements Section IV offers some concluding should be sufficiently flexible to adjust as observations. It also raises some additional the provisions of the climate convention issues that do not arise in the projects evolve; examined in this paper but are nonetheless * practical: joint implementation should be important joint implementation issues. simple and transparent to arrange, administer, and monitor. 1 See, for example: H. Merkus, The Framework Convention on Climate Change: Some Thoughts on Toint Implementation Ministry of Housing, Physical Planning and Environment, November 1992; and Tom Jones, Operational Criteria for Toint Implementation, Organization for Economic Cooperation and Development, OCDE/GD(93)88. 2 The term "criteria" as used here is not to be confused with the term "criteria" as it is used in the FCCC. The FCCC contemplates operational criteria for joint implementation to be developed by the COP. As used here, the term means the general properties that a good system for joint implementation would possess. 2 Environment Department Papers 2 A Brief Description of the Case Study Proj ects Two projects, one in Mexico (ILUMEX) and existing network of local administrative one in Poland (Coal-to-Gas Conversion) offices. These are familiar to customers since provide a concrete context for examination of the vast majority pay their electric bills in many of the issues that would arise in joint person at these locations. implementation projects. Both countries rank among the 15 largest GHG emitters in the CFL lamp sales stations will be established in world according to current estimates, due each of the utility's local offices in the two primarily to energy-system related emnissions.' project cities. Lamps will be offered at The energy systems in both are heavily subsidized prices (efectively sharing some dependent on fossil fuels. Over 98% of portion of the utility's projected savings with Mexico's commercial energy prodution is customers), promoted by an extensive accounted for by fossil fuels, with oil advertizing and public education campaign. accounting for over 80% of the total.2 In Customers will have the option to pay cash for Poland, fossil fuels account for over 99% of the lamps, or pay over time, with their total commercial energy production, with coal electricity bill, at a rate that is established to acounting for over 96% of the total.3 create a net positive cash flow to the customer. Estimated 1989 CO2 emissions from industrial If these two options fail to generate enough sources in Mexico and Poland (by far the demand for lamps, the utility stands ready to predominant sources in both countries) were increase the subsidy to lamp purchases, respectively about 320 million and 441 million and/or to pursue a neighborhood sales plan in tons.4 which mobile sales vans would be sent to neighborhoods and a neighborhood sales force Mexico ILUMEX dispatched door-to-door. ILUMEX will procure and place roughly 1.7 A primary objective of the project, apart from million compact fluorescent lamps over a two- demonstration of CFL technology, is to year period in households in the cities of establish the capacity within the national Guadalajara and Monterrey, two of Mexico's utility to deliver demand side management major urban areas. The procurement process resources over the long term. The project is will be based on a set of performance structured to ensure that the original specifications developed specifically for the investment in the project will be recovered. Mexican market that reflect the technical This will enable the utility to extend demand operating characteristics of the Mexican power side management programs to other regions system and the climates of the two and other technologies. demonstration cities. The procurement process will also involve testing of lamps The total cost of the project is currently against test specifications. The lamps will be estimated at US$ 23.25 million, to be financed distributed initially through the utility's by a grant of US $ 10.0 million from the GET, Climate Change Series 3 Joint Implementation of Climate Change Measures $10 million provided by the Government of measures; (ii) connection to the fuel Mexico, $ 3.0 million provided as joint network; and (iii) intergrated monitoring cofinancing by the Kingdom of Norway, and $ systems; and 0.25 million of parallel cofinancing provided . a Technical Assistance Component by USAID. (US$1.3 million), covering project administration, management, consultancy services, training, and monitoring system. The Poland Coal-to-Gas Conversion (CTG) At the time of appraisal of the CTG project by Project is essentially an umbrella project. It the GEF, two pilot investment subprojects had has many features in common with been identified and associated engineering development finance institution projects prefeasibility studies conducted. In one which establish lines of credit. GEF funds subproject, a local district heating company would be administered by the Polish Bank for boiler in the old city center of Krakow will be Environmental Protection (Bank Ochrony replaced by a gas-fired condensing boiler. In Srodowiska, SA). The project has two the other, a new gas-fired boilerhouse will components: replace two coal fired boilerhouses, and will consist of a combination of one baseload an investment component (US$43.5 million), cogeneration unit, steam boilers and several covering the conversion to gas-firing of a peaking hot water boiler units. The power number of coal-fired boiler houses. The generated will be sold to the grid. conversion technologies selected would include a possible mix of small-packaged The total cost of the overall project is gas-fired cogeneration schemes and gas- estimated at US$ 44.8 million, financed by a fired high efficiency condensing boilers. GEF grant of US$ 25.0, US$ 18.8 million The final mix of conversion technologies provided by Poland, and approxirnately US$ will depend on Polish priorities. This mix 1.0 million of grant cofinancing provided by will affect the number of individual the Kingdom of Norway. The two pilot subprojects because of the large investment sub-projects described in the differences in costs of the two preceding paragraph are estimated to cost US$ technologies. Each individual project 415 thousand and US $ 5,296 thousand design will consider a number of respectively, of which the GEF-financed elements: (i) supplementary end-user amounts would be US$ 133 thousand and US$ energy efficiency and conservation 4,159 thousand respectively. 1 Ibid. 2 World Resources, 1992-93, World Resources Institute (New York: Oxford University Press) 1992. 3 Ibid. - 4 Ibid. 4 Environrment Department Papers 3 Basic Joint Implementation Project Design and Measurement Issues There are four classes of issues that arise in want to know about the emissions generated designing joint implementation projects: (i) by subsidized power prices as well as the determination of the net abatement effect; (ii) emissions reductions associated with the use establishing the price or compensation due in of more efficient lighting. We would also payment for a joint implementation project; want to know whether the production and (iii) performance issues such as verification distribution of CFLs generated, directly or and the treatment of risk; and (iv) procedural indirectly, significant additional GHG issues, including the nature of contracts and emissions. records. These are discussed in the following paragraphs, drawing on experience gained The conceptual principles of proper through the ILUMEX and CTG projects. determination of net abatement effect are thus absolutely clear: the net abatement effect of an Determination of the Net Abatement intervention (policy or project) is the Effect difference between the time path of global emissions with the project and the time path of The science of global warming implies that global emissions without the project. global accumulations of GHGs are responsible From the perspective of the FCCC, the for anthropogenic climate change. This has rom t of thesumablythe clear implications for the way in which the relevant unit of analysis presumably is abatement effects of GHG interventions national emissions since that is the should be measured. Strictly speaking, what reporting/planning entity defined by the we would like to know is what effect each and Convention. It has been suggested that net every intervention (policy and/or project), abatement effects should be measured against individually and collectively, will have on targets or baselines established in national future global emissions and accumnulations of studies/plans. This would require that plans GHGs. If, for example, country x proposed to contain emissions estimates/projections at the convert boilers from oil to gas based on project intervention level of detail. To imports of gas from country y, we would want determine the net abatement effect of a to know not only about the reduced emissions proposed joint implementation project under due to fuel substitution in country x, but also this approach, the "without project" plans of about possible increased emissions due to all project participants' countries would have increased fuel production and transport t to be examined and the effect of the project on country y. To consider another example, "without project" emission trajectories would country y. ~~~~~~~~have to be estimated. suppose that country z supported the placement of high efficiency light bulbs in its As a practical matter, it is currently almost own country and used the resource savings impossible to quantify baselines or emission generated by lower lighting demand to reductions at very specific geographic, subsidize the purchase of power, we would sectoral, technological or temporal levels. In Climate Change Series 5 Joint Implementation of Climate Change Measures any event, data do not exist for Poland and cities as well, based on information and Mexico which would allow us to define a experience generated by the project. These meaningful detailed national baseline "information effects" can be substantial. In projection. what has been termed the "free driver" effect, Mills4 reports that experience in several Another suggestion is that a somewhat less countries in Europe shows that utility-run CFL quantitative "rule of reason" standard be distribution programs appear to stimulate applied.1 This approach would give credit for retail purchases of CFLs through ordinary emnissions reductions achieved by projects commercial channels. According to Mills, the provided these projects were also believed to Swedish lamp manufacturers' trade stand a "reasonable chance" of reducing global organization estimates that 75,000 lamps emnissions. The reasonable chance standard distributed through a rebate check program would be defined in terms of certain criteria. also resulted in the purchase of 41,000 For example, it could be stipulated that additional lamps under ordinary commercial projects implemented in an otherwise terms. In the Netherlands, one utility-run favorable environment for GHG emissions program that resulted in direct sales of 25,000 abatement (e.g., no energy subsidies, lamps reportedly spurred an additional 50,000 demonstrated national commitment to indirect sales; another program generated solution of local environmental problems, 60,000 direct sales and 40,000 indirect sales. In emnissions inventory and national plan that Switzerland, program placement of 7,000 meets FCCC requirements) presumptively lamps reportedly resulted in sales through stand a reasonable chance of reducing global normal channels of an additional 70,000 units. emissions. The amount of the net abatement due to the project, under this approach, would Second, the revolving fund and the creation of be based on calculations against a "reasonable" a DSM entity in the utility company will project-specific baseline. While materially directly and indirectly spur further diffusion simpler than estimating abatement effects of efficient lighting technology, as well as against projected national emissions baselines, other energy efficient technologies. The the task is still daunting.2 emnissions reductions associated with the continuing process of innovation and diffusion The difficulties are well-illustrated by the could also properly be attributed to the ILUMEX and CTG projects. The ILUMEX ILUMEX project. project demonstrates a technology which, provided that power tariffs are adjusted to The ILUMEX project should, in principle, be and maintained at reasonable levels, could be credited with all of these effects. Failure to do expected to penetrate the Mexican market so would reduce the attractiveness of whether or not a demonstration project is investing in such projects and, other things mounted.3 The project also provides for the being equal, would tend to depress investment creation of a revolving fund and institutional in them below optimum levels. In practice, infrastructure that will be used to expand the however, it is scientifically impossible to reach (both geographic and technological) of estimate all of these effects with any degree of Mexico's energy efficiency initiatives. confidence. While there is a vast theoretical and empirical literature on the determinants of Viewed in these terms, the ILUMEX project the rates of innovation and diffusion of new will affect Mexico's future emissions path in a technologies, the current state of the art does number of ways. First, it will, other things not permit reliable calculations/predictions of being equal, speed up the diffusion of efficient the effects of a demonstration project like that lighting technology. Lamps will be replaced being mounted. Nor, unless very carefully not only based on lamp procured from utility- controlled experiments are designed, does it established sales outlets, but also as other permit accurate ex post determination of these households buy lamps through other channels. effects. The procedure followed in the Lamps will be replaced not only in the two cities participating in the project, but in other 6 Environment Department Papers Project Design & Measurement Issues Item Value Source Incandescent Bulb (watts) 67.2 Market survey CFL replacement lamp (watts) 16.8 Calculated based on 4/1 Factor Energy Savings/Lamp (watts) 50.4 Calculated Number of Lamps Replaced 1,500,000 Project assumption Daily Usage of Lamps (hours) 4.0 Assumption/market survey Energy loss factor (%) 18.0 Technical data Energy generation savings (GWH/year) 134.6 Calculated C02 emission factor (kg/kwh) 0.75 Assumption Methane emission factor (kg/kwh) 1.61 x 103 Assumption C02 emission reductions (kg/year) 101.0 x 106 Calculated Methane emission reductions (kg/year) 216.7 x 103 Calculated Methane global warming potential 20.0 Net abatement effect (kg C02/yr) 105.3 x 106 Calculated Figure 1: Mexico ILUMEX Net Abatement Effect illustrative calculations of the net abatement would tend to reduce the marginal emission effect of the ILUMEX project, therefore, takes a reduction associated with CFLs. much narrower view. The calculation is summarized in Figure 1. As can be seen, the On balance, however, the omitted "free driver" calculation basically involves four groups of and other factors (e.g., DSM capacity-building factors: (i) factors relating to average wattage are likely to far outweigh factors that would reduced per replacement lamp; (ii) number of tend to increase emissions per lamp lamps replaced; (iii) average usage of lamps; replaced).5 The net effect estimnated above is and (iv) factors relating to GHG emissions per thus likely to be a substantial underestimate of unit of lamp usage. Based on these the contribution of the project to GHG measurements/assumptions, it was estimated abatement. that annual emissions of C02 would be reduced by 101 thousand tons, and that The effect of the CTG project on likely future annual emissions of methane would be erissions trajectories is qualitatively different reduced by 217 tons. No estimate is made for from the effect of the ILUMEX project in tat any indirect emissions or abatement of GHGs. the boiler conversions/technology mnodifications to be financed under this project While this procedure underestimates the net would not occur under present economic abatement effect of the project on account of circumstances. In principle, thus, there are no the "free driver", organizational learning, and induced or accelerated boiler conversions that revolving fund aspects of the project, it is should be attributed to this project.6 There likely to yield an overestimate of the are, however, some organizational learming abatement achieved by the replacement of the and information externalities associated with lamps per se for two reasons. First, CFLs the project that should lower the cost of at lower the cost of lighting services. As a result, least certain types of future GHG abatement households may (i) choose to burn lamps projects and thus contribute to additional longer each day, (ii) substitute lamps with GHG abatement. greater light output for lamps with lower light The CTG calculation, lke the ILUMEX output, and (iii) increase the number of calculation, lke enginEX sockets in the household. Each of these calculation, is fundamentally an engineering behavioral adjustmnents would tend to reduce calculation based on three factors: (i) heat the actual impact of the bulbs placed by the output; (ii) heat input; and (iii) emissions per project estimated as described above. Second, unit of heat input. Data were obtained from the per kilowatt hour saved emnission factor engineering prefeasibility studies which, utilized is probably also too high for the latter rather than investigating straight substitution years of the project. Mexico is likely, for other of gas for coal, also examined some other reasons, to be shifting its capacity mnix and aspects of the energy system to optimize the altering its dispatching in a manner that balance between heat supply and efficiency of Climate Change Series 7 Joint Implementation of Climate Change Measures Value Source Facility I Energy demand (MWh/yr) 2300 Prefeasibility study Coal demand (MWh/yr) 2949 Prefeasibility study Coal C02 emission factor (tons/MWh) 0.33 Assumption C02 emissions (tons/year) 977 Calculated Gas demand (MWh/yr) 2421 Prefeasability study Gas C02 emission factor (tons/MWh) 0.19 Assurnption C02 emissions (tons/year) 483 Calculated C02 emissions reduction (tons/year) 494 Calculated Facility II Energy demand (MWh/yr) 12770 Prefeasibility study Coal demand (MWh/year) 16372 Prefeasibility study Coal C02 emission factor (tons/MWh) 0.33 Assumption C02 emissions (tons/year) 8100 Calculated Energy demand (MWh/yr) 12770 Prefeasibility study Energy demand for cogeneration unit (MWh/yr) 5500 Prefeasibility study Gas demand (MWh/yr) 23138 Prefeasibility study Gas C02 emissions factor (tons/MWh) 0.20 Assumption C02 emissions (tons/year) 4615 Calculated C02 emissions difference (tons/year) 807 Calculated C02 reduction from power production (tons/yr) 5358 Calculated C02 emissions reduction (tons/yr) 6165 Calculated Figure 2: Poland CTG Net Abatement Effect heat transfer/use. The calculation is Facility II, the reductions would be an summarized in Figure 2. As can be seen, it is estimated 2,681 tpy (i.e., from 8,104 tpy to estimated that annual emission reductions of 5,422 tpy). If, instead, it were assumed that, 494 and 6,165 tons of C02 would be achieved absent the project, the facility operators would respectively by the two suprojects each year. continue to operate their existing boilers, the emnissions reductions attributable to the project Despite the engineering simplifications, the would be much larger. net abatement effects of the subprojects are probably well-approximated by these Determination of Prices of joint calculations. However, although conceptually Implementation Projects clearer than the calculation for the ILUMEX project discussed above, the CTG calculation A variety of approaches have been put also raises serious issues. The most difficult of forward for determining the prices of or these is identifying what the two boiler facility compensation due for the emission reduction operators would do absent the project. The produced by a joint implementation project. above calculations assume that they would One approach would be to leave this replace old boilers after 30 years in service determination to project participants through with new coal-fired boilers, and re-engineer negotiations. The distribution of gains the installation to take advantage of privately obtained thusly would be presumptively fair, profitable energy efficiency improvements. although the negotiation process to arrive at it These assumptions are based on calculations may be very costly. Still another approach which suggest that re-engineering and would be to establish a market. This could be replacement would be optimal from the done by setting up a market maker prepared facility operators' standpoints. Replacing to buy or sell emission reductions at stated boilers with new, more efficient downsized (to prices, via an auction of projects, or take advantage of energy efficiency independently by any entity offering to (say) improvements) would result in substantial buy emission reduction projects. The reductions of C02 emissions. At Facility I, C02 distribution of gains from market or quasi- emissions would be reduced by an estimated market arrangements would depend on 716 tpy (i.e., from 1,693 tpy to 977 tpy); at elasticities of supply and demand in the 8 Environment Departrnent Papers Project Design & Measurement Issues market, as does the distribution of gains from may also be important. Where nations are trade of conventional goods and services. involved, international relations considerations may arise. Or, an investor These points are well-illustrated in terms of might perceive some longer-term market GEF experience with the ILUMEX and CTG advantage unrelated to GHG abatement projects. It was impractical to hold mock considerations to a joint implementation negotiations concerning pricing and shares of arrangement and thus be willing to pay more the net abatement achieved by the projects, or than would be indicated by cost of other GHG to simulate the workings of a market. It is, abatement options. An implementor might however, possible to say something on first perceive some macroeconomic advantage principles about the some of the determinants (e.g., stimulation of employment and domestic of the range of probable outcomes of such output) associated with large-scale negotiations were they possible to hold. From implementation of GHG abatement projects the investor's standpoint, it would make little and thus be Uilling to accept less than sense to pay more for an emission reduction marginal cost than would be obtained through implementation of an equivalent domestic These considerations suggest that the GHG abatement project or through equivalent negotiated price of an emission reduction alternative joint implementation opportunities. produced by any individual joint These alternatives define an upper bound, implementation project could end up at almost 'reservation" price that the investor would be any level. Calculations presented in the willing to pay per unit of emission reduction. following paragraphs based on specific From the implementing entity's perspective, it reference alternatives for investors and would make little sense to accept as implementors provide at best a very rough compensation less than the marginal cost,7 guide to the possible range of outcomes. It is considering the best alternative project, of important to keep this caveat in mind. producing GHG emission reductions via the project. This alternative defines a lower Investors' Reservation Prices. In the present bound reservation price that the implementing example, on the investors' side, Norway's entity would be willing to accept. marginal cost of abating GHG emissions through domestically-implemented measures Price expectations will also play a role in is believed to be relatively high. It has been determining reservation prices of estimated that stabilizing emissions at their implementors and investors. If prices of 1989 levels in the year 2000 would entail an equivalent emission reductions projects marginal cost in the year 2000 of about $180 generally are expected to increase, per ton of C02. Discounted back to 1993 at a 7 implementors may be unwilling to accept a percent social discount rate,8 this implies an price based on today's marginal costs. marginal cost in 1993 of about US$ 112 per Similarly, investors may be willing to pay ton. On purely GHG abatement grounds, more than what they believe to be the current there would be no reason for Norway to pay 'going" price. any more than this for a C02 emission The "equivalent" qualifier in the two preceding reductions. In fact, its willingness to pay paragraphs is important because GHG would probably be somewhat lower than this abatement projects typically involve a number due to alternative possibilities for joint of attributes which affect their potential value implementation with other OECD and non- to implementors and investors. These include OECD countries. There are numerous such important project factors as the untapped opportunities for abatement at less institutional capacity of the implementor, that US$ 100/ton of C02 equivalent. Still reliability of the technology, complexity of the another, somewhat lower benchmark of project, arrangements for dealing with willingness to pay is provided by the C02 tax contingencies, and reputations of the which Norway presently levies on fossil fuel participants. Non-project-specific attributes Climate Change Series 9 Joint Implementation of Climate Change Measures Item Value Source Real Social Discount Rate (%) 10 Assumpion Real Utility Discount Rate (%) 12 Utility cost of capital Real Private Discount Rate (%/6) 24 Credit card real interest rates CFL life (hours) 9000 Assumpiion Incandescent life (hours) 750 Assumption Daily usage (hours) 4 Market survey Incandescent wattage (watts) 67.2 Market survey CFL wattage (watts) 16.8 Calculated at 4/1 CFL Cost (US $/lamp) $10.00 Assumed Incandescent cost (US $/bulb) $0.75 Assumed ILUMEX program costs (US $/lamp) $1.69 Estimated Incremental annual cost/lamp E 10% -$2.88 Calculated Incremental annual cost/lamp @ 12% -$3.50 Calculated Incremental annual cost/lamp @ 24% -$3.61 Calculated Incremental annual cost/ton C02 @ 10% -$0.041 Calculated Incremental annual cost/ton C02 @ 12% -$0.050 Calculated Incremental annual cost/ton CO2 @24% -$0.051 Calculated Figure 3: Mexico ILUMEX Conventional Estimate of Marginal Cost combustion. This tax is roughly equivalent to the conventional calculation are shown in US $ 45 per ton of CO2. Figure 3. Note that marginal costs are negative by this calculation, ranging between - Implementors' Reservation Prices. On the $0.04 and -$0.05 per ton of C02 equivalent implementors' side, as noted above, one reduced depending upon the discount rate, benchmark reservation price is provided by even in the absence of any attempt to adjust the marginal cost of GHG abatement via the for locally accruing benefits such as reductions project. It is worthwhile spending some time in concentrations of suspended particulates discussing this benchmark since it also figures and sulfur dioxide. prominently in discussions of other methods for price determination (e.g., price set If these calculations are even approximately administratively by some international body) accurate, CFLs should immediately replace and in discussions of the operations of incandescent bulbs in a wide variety of multilateral funding mechanisms such as the applications on a pure cost-effectiveness basis, GEF and the financial mechanism of the without regard for global and local external FCCC. benefits. In fact, the process of diffusion of CFL technology will take time for a number of The principles of determination of marginal reasons. Foremost, consumers need to become cost, like the principles of determination of net informed about the cost-savings they can abatement effects, are relatively clear. The achieve with CFLs and the quality and difference between the least cost of an reliability of the product. Someone has to activity/project/policy with and without a produce and disseminate this information. GHG emnissions constraint imposed is the This requires resources. For many products, marginal cost of GHG emissions control.9 In particularly products where brand names are calculating costs with and without the important, manufacturers will provide this emissions constraint, any other direct or information through advertising. For "spillover" benefits associated with the products like light bulbs, where brand name is emissions constraint that would accrue to the not very important, however, manufacturers project host country should be deducted. have little incentive to advertise. By the same token, utilities, which derive their income The application of these prles actice from sales of power on a cost-plus basis, also raises some interesting problems. have little or not incentive to help consumers Conventionally calculated, the marginal costs acquire information about CFLs. of the ILUMEX project are negative. That is why the project is designated, for GEF There frequently are additional factors in purposes, as a Type I Project. The details of developing countries that may slow the 10 Environrnent Department Papers Project Design & Measurement Issues diffusion of CFL technology. CFLs may not applications by overcoming some of these perform as expected under developing barriers. On this interpretation of the project, country power supply conditions. Many marginal costs would be positive. The basic consumers may find the initial purchase price idea behind this interpretation is illustrated in prohibitive. In many countries, highly Figure 4. This figure shows hypothetical subsidized power tariffs blunt incentives to stylized "with" and "without" demonstration economize on power usage. In many of these project emission paths associated with same countries, the power sector is under compact fluorescent lamp penetration of the public ownership and management, and is market. "Without" project diffusion is slower heavily dependent on the national budget. (and perhaps the ultimate penetration of the This also tends to blunt incentives to market is less deep) due to the effects of economize. information externalities, possible capital market constraints, and perhaps other factors The process of diffusion of the technology is as well. The vertical distance between the essentially a process of overcoming these paths is the abatement effect achieved by the informational and institutional barriers. The project at anv point in time, and is thus the ILUMEX project can be expected to accelerate marginal contribution of the project to GHG the diffusion of CFL technology in residential abatement.10 Without Project With Project Time Figure 4: Effect of Project on Emissions Caused by Residential Lighting Climate Change Series 11 Joint Implementation of Climate Change Measures On this interpretation, the gross (of any likely to rise sharply after the first spillover local benefits) marginal costs of the demonstration project. This will tend to make project would be positive, and would be equal investment in successive projects -- roughly - to the total cost of the project, or progressively less appealing on a marginal US $ 23.25 million in the ILUMEX case. Net cost basis. marginal costs would be these costs less any benefits received by Mexico. The majority of Rate Benefits these benefits assume one of two forms. First 10 Percent USS 5.7 million are the benefits associated with the breaking of 2 Percent USS 6.5 million informational and institutional barriers to Percent USS 9.4 milhon efficient lighting technology. These are Figure 5: Benefits to Mexico of Accelerating measured by the value of the advance (i.e., Diffusion By One Year bringing closer to the present) of the time In sum, the above considerations suggest a stream of resource savings that is caused by negative marginal cost of the ILUMEX project the project. Second are the benefits associated on the conventional marginal cost with local envirounental improvements that interpretation, and a positive marginal cost of come about as a consequence of the project approximately the total cost of the project ( US (e.g. lower concentrations of suspended $ 23.25 million), less some undetermined and particulate matter). perhaps undeterminable deduction for As a practical matter, the methodology for benefits accruing to Mexico in virtue of the estimating these latter benefits is controversial, advance in time of resource savings, other involving calculations of the value of reduced local environmental extestalities, morbidity and mortality, increased informa,onal exteoorei,es that accrue te productivity, and enhanced environmental Mexico, and, of course, the future climate amenities. The data required for these change damages prevented or postponed. As cakulations is sorely lacking in developing a practical matter, given the current state of countries. With regard to the benefits the art, it is probably infeasible to do anything associated with bringing resource savings other than treat the total less an approximate closer to the present, one rnight for purposes value for the advance in time of resource of making a rough calculation assume savings as a rough, albeit high, eshmate of conservatively (and in a manner that is marginal cost. On this reasoning, an logically consistent with procedure used appropriate reservation price or compensation above to calculate the net abatement effect of for the emission reduction on a present value the project) that the effect of the project is to marginal cost basis would be roughly on the advance the realization of savings associated order of US $ 13 million to US $ 18 million, or with lamps placed by the project by one year. on an annualized basis roughly US$ 30 to US$ Under this assumption, the present value of 33 per ton of C02-equivalent reduced. benefits associated with advancing the time The Govenment of Mexico recognizes that the stream of benefits range between US$ 5.7 project wllH confer benefits on Mexico and, million and US$ 9.4 million, depending upon therefore, has agreed to finance $10.0 million the discount rate, as shown in Figure 5. of project costs. If this were taken as a rough It should be noted that this project rationale estimate of domestic benefits, marginal cost of does not provide a justification for the project would be $13.25 million. This latter proliferation of projects that have negative figure is the figure being used by the GEF as marginal costs as conventionally measured. its estimate of project marginal cost. Any reasonable methodology would conclude In the case of the CTG project, there clearly are that the marginal effect on the diffusion path positive marginal costs. As noted above, of a second demonstration project is likely to operators of small boiler facilities would not be much smaller than the effects of a first convert to gas under the present incentive project, or perhaps even negligible. That is, structure. The CTG project estimates marginal the marginal cost of accelerating diffusion are 12 Environment Department Papers Project Design & Measurement Issues Item Value Facility I Average Annual Cash Flows (US $ '000) New Coal-Fired Boiler 46 New Gas-Fired Boiler 34 Incremental Operating Costs (Savings) -12 Investment Costs Present Value @ 25% (US$ '000) New Coal-Fired Boiler 235 New Gas-Fired Boiler 415 Incremental Investment Costs (Savings) 180 Service Life (years) 17 Internal Rate of Return (percent) 1.4% Capital Subsidy Required to Achieve 25% IRR (US$ '000) 133 C02 Emnissions Reduction (TPY) 494 Incremental Cost ($/Ton) 33 Facility II Average Annual Cash Flows (US$ '000) New Coal-Fired Boiler 297 New Gas-Fired Boiler/Cogen (net) 296 Incremental Operating Cost (Savings) -1 Investment Costs Present Value @ 25% (US$ '000) New Coal-Fired Boiler 1134 New Gas-Fired Boiler/Cogen 5236 Incremental Investment Cost (Savings) 4162 Service Life (years) 17 Internal Rate of Return (percent) 0 Capital Subsidy Required to Achieve 25% IRR (US$ '000) 4159 C02 Emission Reduction (TPY) 6165 Incremental Cost ($/Ton) 84 Figure 6: Poland Gas-to-Coal Subproject Marginal Costs cost as the capital subsidy required to bring operator. In macroecononically stable the rate of return on boiler conversion up to envirornments with well-functioning capital the facility operator's required rate of return. markets, this is fairly straightforward to The calculations for the two subprojects that estimate. In economies undergoing upheaval, have been identified are shown in Figure 6. where posted interest rates bear little Marginal cost on an annualized per ton basis relationship to the true cost of capital for Facilities I and II respectively are about resources and where all kinds of quantitative US$ 33 and US$ 84. restrictions on lending/borrowing may apply, inevitable elements of judgment and This calculation raises two serious issues. ambiguity enter the calculation. First, as noted above (see paragraph 29 ), there is some ambiguity about the appropriate In principle, two further adjustments should "without project" baseline. If continued use of be made to the costs calculated in Figure 6. old boilers were the baseline, incremental C02 First, some portion of the other costs of the reductions would be somewhat greater. project that are not directly attributable to Marginal investment costs and marginal these two subprojects are nonetheless in some operating cost savings would also be larger. sense caused by them. These costs should be The net result, however, would be an marginal added to the capital subsidies calculated in the cost per ton of C02-equivalent reduced that is figure. Second, as was the case in ILUMEX, much lower. these subprojects will produce local benefits that are properly deductible. Another serious difficulty in these calculations is estimating required intemal rates of return | Pollutant Units Level on some objective basis. An estimate of 25 SO2 $/ton 73 percent is used in the calculations shown in Flyash $/ton 36 the figure. Lower rates would result in lower NO, $/ton 73 estimates of marginal cost. The conceptually igure 7: Emission Fees Levied in Poland correct basis is the cost of capital to the facility Climate Change Series 13 Joint Implementation of Climate Change Measures Poland levies a fee on emissions of selected air monitored. These include annual electricity pollutants emanating from combustion savings attributable to CFLs installed under sources. These fees are as shown in Figure 7. the project during the first year of operation, The calculation of marginal costs summarized net energy savings and critical period load in Figure 6 include (in operating cost savings) reductions directly attributable to the project, reductions in emnissions fees associated with changes in utilization of fixtures with CFLs these pollutants. To the extent that these fees installed under the project, effect of the project represent a reasonable estimate of the on sales of CFLs by retail outlets outside the incremental benefits Poland derives from project, effectiveness of the publicity and sales control of local air pollution, the calculation promotion strategies in creating awareness thus already includes a rough allowance for and stimulating sales of CFLs, and reductions this type of domestic benefit."1 There are of emissions of pollution from power plants probably also other local benefits (e.g., attributable to the project. In the CTG project, organizational learning), however, for which each individual subproject financed by the no allowance has been made. facility will include pre and post conversion calculations and measurement of emissions of To sum up, the above considerations greenhouse gases and other air pollutants. concerning (i) Norway's domestic costs of implementing GHG abatement measures and There are several different kinds of risks that (ii) marginal costs to Mexico and Poland of are common to project investments, whatever implementing the GHG abatement measures their purpose. First, there is the risk that one suggest a very wide range of possible joint or more of the participants in the project will implementation project price outcomes in the fail to discharge their obligations. The best two cases considered here, ranging from about way to ensure that this does not happen is to US$ 30/ton C02-equivalent to as much as US$ structure the agreement between the parties in 100/ton C02-equivalent or more. It must be such a fashion that, at each point during the remembered, moreover, that the prices implementation of the project, all parties have actually established through negotiations an incentive to perform as agreed. One way of could lie outside of this range. This could accomplishing this insofar as the implementor happen due to other factors, some of which of the project is concerned is to require, in are discussed above, that could be important general, that the implementor retain an to either the implementor or the investor. ownership stake in the project during its initial implementation. If the project is successful, Performance Issues this creates an asset that subsequently could be sold or retained, as the implementor The calculations of net abatement effects prefers. summarized above unavoidably are based on a number of assumptions. It is thus quite Many projects will depend on development of likely that, even within a relatively simple new institutions, new markets, and/or framework for accounting for net abatement, adaptation of relatively new technologies to estimates made prior to project new environments. Both the ILUMEX and the implementation will turn out to be wrong. CTG projects have these features. In the case Some mechanism for verification is thus of ILUMEX, there are risks that CFE will not clearly required. Moreover, as in any be able to place all of the bulbs, or only be able contractual relationship, a proper incentive to do so at substantially higher cost than framework requires that there be an agreed envisioned in the project design. There are upon means to deternine whether all parties further risks that CFLs will fail under Mexican have discharged their obligations properly. power system conditions. In the case of CTG, risks include delays in obtaining local Both projects considered here have a financing for the project and difficulties in verification procedure built in. In the identifying suitable subprojects for support. ILUMEX project, a number of different While the project designs address these risks dimensions of project performance are to be adequately, there can be no absolute assurance 14 Environment Department Papers Project Design & Measurement Issues that these other problems will not arise. The global level. Provided the legal and result could be materially smaller net institutional arrangements are structured abatement effects and/or significantly higher satisfactorily so as not to provide an project costs. systematic incentive to project failure, the global portfolio of all projects would be As with pricing issues, arrangements for expected to perform on average about as handling project risks would probably best be expected. This is because projects that fail to left to the contracting parties. One can live up to expectations would tend to be envision a wide variety of possible balanced by project that exceed expectations.12 arrangements, ranging from requiring only 'best efforts" from an implementor (but Procedural/Documentation Issues otherwise holding the implementor harmless for partial or complete failure of the project) to The approach taken in preparation of requiring performance-to-specification and a illustrative legal instruments for joint schedule of penalties for failure to perform. implementation is modeled on an investment As noted above, the way in which risks are project syndication. This approach is taken for handled will also have a bearing on the two reasons. First, it builds on approaches pricing of the emnission reductions produced that have been taken to cofinancing of GHG by a joint implementation project. abatement projects during the pilot phase of the GEF, and arrangements negotiated by Whatever arrangements are made between power producers to procure carbon contracting parties with respect to risk, the sequestration services in other countries. investors will necessarily bear the ultimate Second, it recognizes that the FCCC is not at risk associated with failure to meet obligations present sufficiently structured to support the under the FCCC or other cognizant regulatory development of a full-fledged market based authorities. There are a number of strategies approach to global GHG abatement. The available for coping with this risk. Countries absence of global emissions limits and might, for example, plan to abate (or contract timetables and standard measurement for the abatement of) emnissions by more than protocols precludes the definition of a the required amount, using the cushion as standard commodity or instrument that could insurance against the failure of some projects. be traded. Trade provides another route for dealing with this risk. An investor with a serendipitous or There would be three basic documents under intentional surplus could sell or rent to a party the syndication approach. The first would be needing additional credit to meet its a Project Placement Memorandum. A draft obligations. It is probable that forward and example of this document for the ILUMEX options transactions would emerge as well. project is contained in Annex B of this paper. This document would offer a specified Still another way to reduce risk would be to investment opportunity to selected potential invest in a portfolio of projects. This could be investors on stated terms and conditions. In done individually or through a project fund particular, it would contain (i) a detailed like the GEF. In the present case, the CTG technical description of the project, including a project, which will fund a number of quantitative description of the estimated net subprojects, diversifies against risk associated GHG abatement to be achieved by the with individual facility subprojects, although project;13 (ii) the price or cost at which shares it does not address the risks associated with in the project are being offered; (iii) the the possible failure of an implementor or a general terms and conditions under which the rather narrow class of technologies. offering is made; and (iv) representations and The risks described above are risks that disclaimers. individual implementors and investors face in The Project Placement Memorandum would individual projects. This does not, it should be be the vehicle for soliciting investment, stressed, necessarily translate into risks at the establishing the outlines of the agreement Climate Change Series 15 Joint Implementation of Climate Change Measures between the cooperating parties, and for The third document would be an Annual soliciting a determination under the FCCC or Report, prepared by the implementing entity under various national laws and regulation of for the project. This document would inform the eligibility of the project for abatement investors concerning the status of the project, credit. The degree of technical detail would be as well as provide information required by sufficient to permit relevant national regulatory bodies to make a determination regulatory bodies and/or the relevant body concerning the abatement of GHGs actually under the FCCC to determine, subject to obtained by the project. The information verification, the presumptive net abatement presented would include independent effect (i.e., the net reduction in global GHG verification (perhaps like an auditor's emissions) of the project. statement in a financial report) of this net abatement. Given the uncertainty about both the degree of abatement that will actually be achieved by "Unofficial" (i.e., prior to operation of the the project and the amount of abatement that FCCC) joint implementation could begin will be allowed by the responsible regulatory immediately based on these procedures. authorities, the investment opportunity would Investors and implementors alike would, to be be stated, in both the Project Placement sure, be speculating. In some cases, investors Memorandum and the Legal Agreement in might be able to use investments in projects in terms of shares of an unknown total, to be another country to meet domestic obligations. determined by the responsible regulatory For example, if a company resident in a authority. country that taxes CO2 emissions were to invest in one of the two projects considered The second document would be a Legal here, it might seek some credit against its Agreement between the various participants obligations for C02 taxes on fossil fuels. To in the investment project. A draft example of this end, it could present the Project Placement this document for the ILUMEX project is Memorandum to the tax authorities for a contained in Annex C of this paper. This preliminary determination concerning document would spell out the rights and whether offsets against domestic tax liabilities responsibilities of each project participant, the would be granted. A final determination remedies available for nonperformance or would be based on an executed Legal other forms of project failure, as well as the Agreement establishing the company's shares of net abatement due to each. This ownership of a share of the net abatement document would be the basis for submission achieved, and Annual Reports that establish to regulatory authorities for determination of the amount of abatement actually achieved. credits, subject to verification, against the project participants' individual abatement As the apparatus for implementing the FCCC responsibilities. comes into being, essentially this same approach could be followed. The Project The Project Placement Memorandum and Placement Memorandum could be submitted Legal Agreement would both provide for the to the implementing arm of the FCCC for subsequent transfer of investors' interests to preliminary determination that the proposed other parties. This would facilitate the project meets its requirements for joint development of a secondary market in project implementation. The Legal Agreement would participations, as well as introducing fix shares in a project under implementation, possibilities for development of forward and and the Annual Report would provide the options markets. Transfers would, of course, information needed for a determination of be subject to regulation by national authorities how much abatement would be credited to and by the FCCC.14 individual project participants as an offset to their individual obligations under the FCCC. 16 Environment Department Papers 1 Jones, op. cit. 2 It should be pointed out that the problerns that arise are not problems that have to do with joint implementation per se. Rather, they are basic practical difficulties in measuring the full effects of policies and or projects. 3 See paragraphs 38 and 39 for an explanation as to why CFL technology would be expected to penetrate the Mexican market eventually. 4 Mills, Evan. "Using Financial Incentives to promote Compact Fluorescent Lamps in Europe: Cost Effectiveness and Consumer Response in 10 Countries", in Right Light, Bright Light, ed. Evan Mills, 1992. 5 Elasticities of demand for lighting services are probably low and the extent to which the capacity mix and dispatching will shift over the life of the lamps placed by the project is also relatively small. 6 Poland is expected to tighten substantially its controls and regulations on local air pollution. Under these circumstances, it is probable that boiler conversions that are not privately profitable now would become so. To this extent, the net effect of the CTG project, like the ILUMEX project, is to move forward in time the time at which GHG emnission reductions are realized. 7 The term "incremental cost" is frequently used in discussions of the cost of GHG abatement. It is also used in the FCCC. To avoid confusion with this latter usage, which will be defined by the Conference of Parties, the term "marginal cost"-meaning the additional economic cost caused by GHG reduction -- is used here. A precise definition is given in paragraph 37 of the paper. 8 This is the social discount rate currently applied by the Kingdom of Norway to evaluate investment projects. 9 See King, Kenneth and Mohan Munasinghe, Cost-Effective Means to Limit the Emissions of Greenhouse Gases In Developing Countries, World Bank Environment Department Divisional Working Paper 1992-30, pp. 7- 11, for a more complete discussion. 10 The economic rationale underlying this interpretation of the marginal cost of projects that advance the time stream of global benefits associated with GHG abatement is further explained in Annex A. 11 Rough estimates of the benefits of reducing local air pollution suggest that the marginal benefits are in fact somewhat larger than is implied by the tax rates in Figure 7. 12 There may, of course, be some sources of systematic risk that are not easily diversified. If GEF projects are like normal development projects, for example, the portfolio of projects is likely to perform less well under difficult world economic conditions than when the world economy is basically healthy. Note in this case, however, that the lower levels of economic activity associated with weak global economic performance would tend to reduce emission in weak performance years, compensating in some degree, for project portfolio performance. There may, of course, be other forms of systematic risk that do not possess this serendipitous characteristic. 13 The description of the project would be prepared in accordance with standards established under the FCCC in sufficient detail to permit the Conference of Parties or its designee to determine whether the project concept is consistent with the joint implementation provisions of the Convention, and what the total net abatement credit, subject to verification by the FCCC or other appropriate regulatory body, would be. 14 At a minimum, both national authorities and the FCCC would have to know the nationality of ownership of offsets to determine whether nations are in compliance with their responsibilities under the FCCC. Participation Series 17 4 Concluding Observations Consideration of the issues raised above As project experience accumulates, both under suggests a number of conclusions with respect joint implementation and under the financing to the questions to be confronted as mechanism, it should be possible to strengthen procedures for joint implementation under the considerably the procedures for estimating net FCCC are fleshed out. Perhaps foremost, it is abatement effects of projects ex ante and clear that the fundamental practical problem verifying their effectiveness ex post. The GEF to be confronted by the Conference of Parties pilot phase portfolio provides a particularly in making operational the joint fruitful testbed for the development of new implementation provisions of the FCCC will methods. Implementation of GEF projects be ensuring -- witiin a reasonable margin of under a successor phase (assuming that doubt - that global GHG emissions are negotiations for a restructuring and reduced relative to what they would have replenishment of the GEF are concluded been absent joint implementation. This successfully), as well as the first few years of problem is particularly difficult in the case implementation under the financing where joint implementation teams up Annex I mechanism and joint implementation with non-Annex I partners since the latter provisions of the FCCC could also play an have, as yet, assumed no monitorable important role. It will be important in setting obligations with respect to emissions up the policies and procedures that govern reductions.' It should be pointed out, these various forms of implementation to however, that this difficulty is not unique to build-in studies, based on careful project-by- joint implemenentation. In the absence of project monitoring and evaluation, and studies cornmittments to ernissions limnits, any like PRINCE2 which also examine and abatement effect (be it achieved through joint evaluate experience in preparing and revising implmentation or some other form) will country studies and plans, designed expressly necessarily have to be measured against some for the purpose of strengthening purely hypothetical baseline. methodological approaches. It will also be important to provide for revisiting, on a The technical/analytical considerations regular (say, every 3 to 5 years) basis, the illustrated by the ILUMEX and CTG cases regulatory/procedural framework based on suggest that, at least initiatially, the approach the accumulating lessons of experience taken to estimating and verifying net learned from this growing portfolio. abatement probably will have to be relatively simple. It does not seem practical at this Pending the development of better juncture to prescribe approaches that would methodology, one way of maximizing the require detailed data and very sophisticated likelihood that joint implementation would calculations or projections. At present, there result in global net abatement is to impose appear to be few if any viable altematives to some general restrictions on participation in proceding on a project-by-project basis, joint implementation arrangements. For allowing considerable latitude and scope for example, it could be stipulated that both judgement, as illustrated by the analyses of the investor and implementor nations would have case studies presented here. to have "climate change friendly" policy Climate Change Series 19 Joint Implementation of Climate Change Measures environments, as determined under Another difficult question, related to the procedures to be defined by the Conference of determination of net abatement effects, may Parties. Examples of criteria that could be also confront the Conference of Parties: established to define "climate change friendly" whether to place restrictions on joint would be (i) compliance with all key implementation of "no regrets" interventions procedural obligations (e.g., current inventory like ILUMEX. Policy reforms, although not and national plan meeting FCCC-established examined in this paper, also represent another methodological standards of anthropogenic potential type of "no regrets" intervention that emissions of GHGs, and current in meeting could be proposed for joint implementation. obligations to contribute to the financing For example, it might be proposed that one mechanism) under the FCCC and (ii) no country provide support to another country to subsidies, implicit or exlicit, to net emissions (say) raise its electricity tariffs to the level of of GHGs (e.g., subsidies to energy long run marginal cost. consumption, subsidies/taxes that favor GHG-intensive energy forms such as coal There are two potential problems with "no relative to oil or gas, subsidies/taxes that regrets" options when the proposed encourage deforestation). implementor country has not assumed specific obligations with respect to limitation of its Other suggestions have also been put forward overall GHG emissions. First, true "no regrets" with the "net abatement effectiveness" options would be undertaken anyway. There objective, in whole or in part, in mind. Many is thus, it could be argued, no net abatement of these would limit joint implementation effect associated with their implementation. according to countries' classifications under Second, if "no regrets" projects and/or policies the FCCC. One such proposal is that joint were eligible for joint implementation, implementation be limited to countries that potential implementors not otherwise bound have agreed to specific emissions limitations by specific emissions limitation obligations under the FCCC (i.e., the Annex I countries). might perceive incentives to create or Another proposal is that Annex I countries' exaggerate existing barriers to the adoption of reliance on joint implementation with non- globally beneficial technologies and policies Annex I countries be limited to some with the hope of attracting additional percentage of the projected emissions international resource transfers. In the worst reductions they need to acheive to meet their of circumstances, situations could be FCCC obligations. Still another proposal is envisioned in which the seeking of such that joint implementation with non-Annex I transfers could actually result in an increase in countries be deferred until Annex I countries global emissions and accumulations from have met their GHG emission stabilization those which might otherwise obtain. objectives. While these restrictions would tend to increase the likelihood that Annex I There is a broad range of positions that could countries' emissions will meet targets, their be taken with regard to the eligibility of "no overall effect on global emissions is uncertain. regrets" interventions for joint Joint implementation, by lowering costs and implementation. One possibility would be to providing potential gains to both investor and simply rule out project and/or policy implementor alike, may be a very effective initiatives that are "no regrets" under the instrument for encouraging efforts by Annex I conventional calculus of costs and benefits for and non-Annex I countries alike in furtherance joint implementation. This would effectively of the purposes of the FCCC. In addition, deal with the perverse incentives problem, but restrictions on joint implementation with non- at the possible cost of sacrificing some globally Annex I countries may substantially reduce beneficial abatement opportunities. The the cost-effectiveness of whatever degree of ILUMEX project considered here, for example, global GHG abatment is attained. would be ineligible for joint implementation under such an approach. 20 Environment Department Papers Concluding Observations Another possibility, already outlined above, does not necessarily mean that the would be to take the position that a suitable intervention is not "no-regrets". The economic policy environment is a precondition for, but benefits that accrue domestically may or may not eligible for, support under joint not exceed the costs of the intervention, implementation. This posture would not rule including the costs associated with out support for projects like the ILUMEX overcoming these barriers. project provided that a legitimate case can be made that the project will contribute to an Second, as illustrated by the ILUMEX project, acceleration of net abatement of GHGs. It even in the case of interventions that appear to would, however, rule out direct support for be "no regrets" when all domestic costs and policy reforms such as increasing energy benefits are counted conventionally, there is tariffs.3 still the issue of the effect of the intervention on the timing and the ultimate extent of GHG Still another alternative would be to place no a emissions reductions. Like the ILUMEX case, priori restrictions on "no regrets" policy many interventions conceptually have the reforms and projects as vehicles for joint effect of accelerating GHG emissions implementation. As in the case of "no regrets" abatement. If a global benefit is associated projects, however, a legitimate case would with more rapid/earlier abatement per se, have to be made, on a case-by-case basis, that then there is a good rationale for not ruling there are in fact marginal marginal global out ipso facto joint implementation of such benefits that would be accrued. This would interventions. preserve the maximum range of choice for joint implementation (and hence should Another clear implication of the case study contribute to cost effective implementation of experience examined here is that there is a the FCCC), but would entail entail great deal to be said for implementation considerable exercise of judgement. In the approaches that do not require official or case of policy interventions, estimation and explicit determination of marginal cost.4 The verificationi of net abatement would involve "operational" methodology, like that for policy conterfactuals that are even more determination of net abatement effects, daunting than the project counterfactuals requires a great deal of judgement, and official illustrated here, and could involve the determiinations of marginal cost plays no Conference of Parties in monitoring and essential role in joint imphmentation. While supervision of economic policy reform - a role prospective parties to a joint implmentation for which it may not be well suited. agreement may want to make marginal cost calculations as a point of reference for the There are two additional fundamental points bargaining positions they take, this is not to be kept in mind when weighing the costs necessary. and benefits of alternative postures with regard to restricting joint implementation of Finally, it should be noted that there are a "no regrets" interventions. First, many (if not number of other important systemic issues most) interventions that typically are believed with regard to joint imnplementation that are to be "no regrets" in fact have real costs not discussed in this paper because they do attached to them. These may be costs not arise in the context of the two projects associated with information externalities (e.g., examined here. Perhaps the most important of consumer information about new products, these concern the structuring of the organizational knowledge about how to relationship between inplementation under supply a product or service), the joint implementation provisions of the transitional/adaptational costs (including convention and implementation through the difficult, perhaps impossible to estimate financing mechanism to be established under political costs) associated with changing the the FCCC. The challenge to the Conference of way things are done, and/or capital Parties will be to give structure to these two constraints. The presence of these other costs instrumnents so as to maximize their complementarity, and thereby promote more Climate Change Series 21 Joint Implementation of Climate Change Measures efficient, effective, equitable GHG abatement than would be attainable via either of the implementation approaches in isolation. 1 Cases in which Annex I countries are involved in joint implementation do not raise this difficult since it can presumably be determined whether their actual emissions les net holdings of credits for joint implmentation iwth other Annex I countries are less than the emissions limits to which they have agreed. 2 Program for Measuring Incremental Costs for the Environment. Administrator's Office. Global Environment Facility. 3 Joint implementation of such projects might indirectly support such reforms, however. Energy efficiency projects, for example, help to blunt the impact on consumers' energy expenditures of tariff increases. 4 Merkus, op cit, proposes, for example that a company that wihes to invest in emissionsreductions via joint implementation with a company in another country should be required to show that the investment is cost- effective. 22 Environment Department Papers Annex A Additional Explanation of the purposes. All that is important is that there Calculation of the Marginal Cost of are positive global benefits associated with the the ILUMEX Project project. This annex briefly explains the basic economic Assuming an interest rate of r,1 the present logic behind the calculation of the marginal value at the present time of domestic and cost of the ILUMEX project, and conceptually global benefits if the project is implemented in similar projects. For this purpose, it is year T are respectively: convenient to develop a simple, stylized De-r economic analysis of the optimal time of implementation of a project that has both domestic and global benefits. This analysis is and outlined in the following paragraphs. Ge-rT To simplify, let us assume that the present Let us further assume that the present value at values in the year implemented of both domestic and global benefits are independent the present time of the costs of implementing of the year in which the project is imnplemented, and are equal to D and G relationship C(T), and that C(T) declines as T respectively. In the case of the ILUMEX increases; that is, the present value of project, for example, D would represent the implementation costs go down as the date at present value in the year incandescent lamps which the project is to be implemented is were replaced by compact fluorescent lamps extended further into the future. There are werthe preplaced byecoompac falueorhesent y lseveral reasons that this might occur. Pushing of the present economic value of the energy the project further into the future presumably savings associated with use of these energy efficient lamps. If it were possible to estimate means that at least some cost outlays can be the present value of other domestically- postponed. Discounting reduces the present 'accruing benefits such as the benefits value of these costs. More importantly, associated with lower emiissions of local air allowing more time prior to implementation pollutants such as particulate matter and may result in lower costs ipso facto. In the sulfur dioxide, these also would be included i case of ILUMEX, for example, the sooner the the calculation of D. time that CFLs are to be placed, the bigger and more expensive the public information, In principle, global benefits associated with advertising, and marketing outlays that would reductions of concentrations of GHGs in the be required. The relative price of project atmosphere would be the major component of technologies may decline with time, as the calculation of G. Any portion of this appears to be happening with the CFL benefit accruing in the technology that wil be employed in the domestic/implementing country would be ILUMEX. And finally, "haste makes waste" is included in the calculation of D rather than G. a basic if not terribly scientific principle of project implementation. In practice, of course, it is extremely difficult to calculate G. This is not essential for our 1 Annex A 1 6 1 4 t12 0 -2 C4~ C4 4 C4 1 ) ell -W~ - UV.f q.m qj*U Years Figure A.1: Effect of Accounting for Global Benefits on Optimal Implementation Date From the domestic (i.e., implementing) value of benefit curve upward and shifts the country's perspective, the optimal course of point at which it reaches a peak leftward.2 action is to select a project implementation This is shown by the higher of the two curves date that maximizes the difference between shown in the figure, which implies an the present value of domestic benefits of the optimum implementation date of some 4.4 project and the present value of costs. This is years hence. illustrated in Figure A.1, where the lower of the two curves shown represent the net In other words, form a global perspective, it is present value of the project, as a function of optimal to implement the project earlier than it the implementation date, to the implementor. is from a domestic perspective. The additional The domestically optimal implementation date costs associated with eariier implementation, is some 5.3 years hence. less the additional domestic benefits realized, are the marginal cost of the project. Factoring global benefits into the decision changes the calculus. It shifts the net present 1 Domestic and global interest rates need not be the same. For simplicity, however, we assume here that they are. 2 A necessary condition for maximization of the net present value of benefits is that the marginal contribution ot the present value of benefits of advancing the implementation date in time equal the marginal contribution of the present value of costs of doing so. Mathematically, this requires that -rBe-rT-C'(T)=O and r2Be-rT-C"(T)