Issues and options in implementing the Montreal protocol in developing countries

THE WORLD BANK SECTOR POLICY AND RESEARCH STAFF Environment Department Issues and Options in Implementing the Montreal Protocol in Developing Countries Mohan Munasinghe Kenneth King September 1991 Environment Working Paper No. 49 This paper has been prepared for internal use. The views and interpretations herein are those of the author(s) and should not be attributed to the World Bank, to its affiliated organizations or to any individual acting on their behalf. ISSUES AND OPTIONS IN IMPLEMENTING THE MOATREAL PROTOCOL IN DEVELOPING COUNTRIES L INTRODUCTION In recent decades, there has been increasing concern worldwide regarding the depletion of the high altitude ozone layer that protects the biosphere from harmful solar radiation. Almost all the past chlorofluorocarbon (CFC) accumulation (which is the primary cause of the ozone layer depletion) has been the result of economic activity in the industrialized world. Meanwhile, the ozone layer appears to be thinnest in higher latitudes, and persons living in those regions are likely to be more vulnerable. There is also an unknown but potentially serious global effect due to the generalized impact of increased ultraviolet rays on plant arid animal life. It is not surprising that the initial recognition of the problem and impetus for action to limit the emission of ozone depleting substances (ODS) has arisen in the developed world. At the same time, the developing countries (LDCs) share the deep worldwide concerns about environmental degradation, and some are taking steps already to address the ozone issue. However, they also face other urgent issues like poverty, hunger and disease, as well as rapid population growth and high expectations. The paucity of resources available to address all these problems constrains the ability of LDCs to undertake costly measures to protect the global commons. In brief, the crucial dilemma for LDCs is how to reconcile development goals and the elimination of poverty -- which will require increased use of environmental resources -- with responsible stewardship of the environment, and without overburdening already weak economies. The recent report of the Bruntland Commission (WCED 1987), which has been widely circulated and accepted, has presented arguments along the theme of sustainable development, which consists of the interaction of two components: needs, especially those of the poor segments of the world's population and limitations, which are imposed by the ability of the environment to meet those needs. The development of the presently industrialized countries took place in a setting which emphasized needs and de-emphasized limitations. The development of these societies have effectively exhausted a disproportionately large share of global resources - broadly defined to include both the resources that are consumed in productive activity (such as oil, gas and minerals), as well as environmental assets that absorb the waste products of economic activity and those that provide irreplaceable life support functions (like the ozone layer). Clearly, any reasonable growth scenario for developing nations that followed the same material-intensive path as the industrialized world, would result in unacceptably high levels of future ODS accumulation as well as more general depletion of natural resources. The division of responsibility in this global effort emerges from the above arguments. The unbalanced use of common resources in the past should be one important basis on which the developed and developing countries can work together to share and preserve what remains. The developed countries can afford to substitute envirotimental protection for further growth of material output. On the other hand, the developing countries can be expected to participate in the global effort only to the extent that tmis participation is fully consistent with and complementary to their immediate economic and social development objectives. Since most LDCs can hardly afford to finance even their present industrial development, to address global environmental concerns they will need financial (and technical) assistance on concessionary terms that is additional to existing conventional aid. 1.1 The Montreal Protocol The international community, reacting through the United Nations, has recognized the danger presented by ozone destruction and decided to limit the damage ind the risk by stopping the emissions of ODS (CFCs, halons, and substances having similar effects). This resolve took the form of the Mot.treal Protocol, signed in September 1987. The Protocol is an agreement among nations which establishes schedules for gradual reductions in the consumption of ozone depleting substances based on use within each participating country during 1986. The Protocol also places restrictions on trade in ODSs with non-parties to the Protocol. 1.2 The Multilateral Fund The Parties to the Protocol recognized the special situation of developing countries and undertook to facilitate their access to environ-entally safe alternative substances and technology and to assist them in making expeditious use of such alternatives. The Parties also undertook to facilitate bilaterally or multilaterally the provision of subsidies to developing countries for this purpose. They realized, in essence, that the reluctance of developing countries to ratify the Protocol was due to lack of financial resources necessary to meet the obligations without impairing their development efforts. 2 Consequently a special fund (the Multilateral Fund) was set up to provide concessional finance and outright grants additional to those available from existing aid programs. The Fund is jointly administered by the World Bank, the United Nations Development Programme, and the United Nations Environment Programme, with the Bank handling the project financing arrangements. Grant financing is to be made available to eligible countries (i.e., developing countries who are Parties to the Protocol and who consume less than 0.3 kg of ODS per head per year) for eligible expenditures (the incremental costs of various phase-out projects). The items on which expenditures are reimbursable were determined by the Parties in June 1990, in what are known as "the London Agreements." This list includes things such as incremental costs of production (e.g., to reconfigure a plant to produce substitutes), equipment manufacture (e.g., to retool refrigerator manufacturing to use alternatives to ODS), recycling, technical assistance, and training. 1.3 National Ozone Policies The phase-out in each country takes place in a policy context comprising overall economic policy, industrial strategy, and specific measures adopted to comply with the Protocol ("National Ozone Policy"). As finance from the Multilateral Fund is, in a sense, compensation by the developed countries for developing country action toward mitigating a global problem, the finance is not conditional upon any policy reform or change in industrial strategy. There is, however, a requirement that signatory countries will produce a national implementation plan for their phase-out activities which would show that the projects proposed for financing will meet the obligations under the Protocol and do so at least cost. Specific institutional steps and regulations or other instruments of National Ozone Policy will also be needed. In this paper we present some of the policies and issues involved in the implementation of the Montreal Protocol; first the issues confronting the Fund, which provides the backdrop for National Ozone Policies, and second the issues to be addressed by those policies. There are three broad requirements one might reasonably expect that both the Fund and National Ozone Policy should meet and which we touch on in the following discussions. They are: - Economic efficiency. Can the goals be met in a cheaper way? - Equity. Who bears the burden of adjustment and will they be compensated fairly? 3 - Effectiveness. Will the specific mechanisms proposed be effective in achieving the goals? As is often the case, these broad requirements partially conflict and, as we shall see below, this is also the case in implementing the Protocol. 4 ISSUES AND OPTIONS IN IMPLEMENTING THE MONTREAL PROTOCOL IN DEVELOPING COUNTRIES I. INTRODUCTION In recent decades, there has been increasing concern worldwide regarding the depletion of the high altitude ozone layer that protects the biosphere from harmful solar radiation. Almost all the past chlorofluorocarbon (CFC) accumulation (which is the primary cause of the ozone layer depletion) has been the result of economic activity in the industrialized world. Meanwhile, the ozone layer appears to be thinnest in higher latitudes, and persons living in those regions are likely to be more vulnerable. There is also an tinknown but potentially serious global effect due to the generalized impact of increased ultraviolet rays on plant and animal life. It is not surprising that the initial recognition of the problem and impetus for action to limit the emission of ozone depleting substances (ODS) has arisen in the developed world. At the same time, the developing countries (LDCs) share the deep worldwide concerns about environmental degradation, and some are taking steps already to address the ozone issue. However, they also face other urgent issues like poverty, hunger and disease, as well as rapid population growth and high expectations. The paucity of resources available to address all these problems constrains the ability of LDCs to undertake costly measures to protect the global commons. In brief, the crucial dilemma for LDCs is how to reconcile development goals and the elimination of poverty -- which will require increased use of environmental resources -- with responsible stewardship of the environment, and without overburdening already weak economies. The recent report of the Bruntland Commission (WCED 1987), which has been widely circulated and accepted, has presented arguments along the theme of sustainable development, which consists of the interaction of two components: needs, especially those of the poor segments of the world's population and limitations, which are imposed by the ability of the environment to meet those needs. The development of the presently industrialized countries took place in a setting which emphasized needs and de-emphasized limitations. The development of these societies have effectively exhausted a disproportionately large share of global resources -- broadly defined to include both the resources that are consumed in productive activity (such as oil, gas and minerals), as well as environmental assets that absorb the waste products of economic activity and those that provide irreplaceable life support functions (like the ozone layer). Clearly, any reasonable growth scenario for developing nations that followed the same material-intensive path as the industrialized world, would result in unacceptably high levels of future ODS accumulation as well as more general depletion of natural resources. The division of responsibility in this global effort emerges from the above arguments. The unbalanced use of common resources in the past should be one important basis on which the developed and developing countries can work together to share and preserve what remains. The developed countries can afford to substitute environmental protection for further growth of material output. On the other hand, the developing countries can be expected to participate in the global effort only to the extent that this participation is fully consistent with and complementary to their immediate economic and social development objectives. Since most LDCs can hardly afford to finance even their present industrial development, to address global environmental concerns they will need financial (and technical) assistance on concessionary terms that is additional to existing conventional aid. 1.1 The Montreal Protocol The international community, reacting through the United Nations, has recognized the danger presented by ozone destruction and decided to limit the damage and the risk by stopping the emissions of ODS (CFCs, halons, and substances having similar effects). This resolve took the form of the Montreal Protocol, signed in September 1987. The Protocol is an agreement among nations which establishes schedules for gradual reductions in the consumption of ozone depleting substances based on use within each participating country during 1986. The Protocol also places restrictions on trade in ODSs with non-parties to the Protocol. 1.2 The Multilateral Fund The Parties to the Protocol recognize.' the special situation of developing countries and undertook to facilitate their access to environmentally safe alternative substances and technology and to assist them in making expeditious use of such alternatives. The Parties also undertook to facilitate bilaterally or multilaterally the provision of subsidies to developing countries for this purpose. They realized, in essence, that the reluctance of developing countries to ratify the Protocol was due to lack of financial resources necessary to meet the obligations without impairing their development efforts. 2 Consequently a special fund (the Multilateral Fund) was set up to provide concessional finance and outright grants additional to those available from existing aid programs. The Fund is jointly administered by the World Bank, the United Nations Development Programme, and the United Nations Environment Programme, with the Bank handling the project financing arrangements. Grant financing is to be made available to eligible countries (i.e., developing countries who are Parties to the Protocol and who consume less than 0.3 kg of ODS per head per year) for eligible expenditures (the incremental costs of various phase-out projects). The items on which expenditures are reimbursable were determined by the Parties in June 1990, in what are known as "the London Agreements." This list includes things such as incremental costs of production (e.g., to reconfigure a plant to produce substitutes), equipment manufacture (e.g., to retool refrigerator manufacturing to use alternatives to ODS), recycling, technical assistance, and training. 1.3 National Ozone Policies The phase-out in each country takes place in a poll hy context comprising overall economic policy, industrial strategy, and specific measures adopted to comply with the Protocol ("National Ozone Policy"). As finance from the Multilateral Fund is, in a sense, compensation by the developed countries for developing country action toward mitigating a global problem, the finance is not conditional upon any policy reform or change in industrial strategy. There is, however, a requirement that signatory countries will produce a national implementation plan for their phase-out activities which would show that the projects proposed for financing will meet the obligations under the Protocol and do so at least cost. Specific institutional steps and regulations or other instruments of National Ozone Policy will also be needed. In this paper we present some of the policies and issues involved in the implementation of the Montreal Protocol; first the issues confronting the Fund, which provides the backdrop for National Ozone Policies, and second the issues to be addressed by those policies. There are three broad requirements one might reasonably expect that both the Fund and National Ozone Policy should meet and which we touch on in the following discussions. They are: - Economic efficiency. Can the goals be met in a cheaper way? - Equity. Who bears the burden of adjustment and will they be compensated fairly? 3 - Effediveness. Will the specific mechanisms proposed be effective in achieving the goals? As is often the case, these broad requirements partially conflict and, as we shall see below, this is also the case in implementing the Protocol. 4 II. ISSUES IN THE IMPLEMENTATION OF THE MULTILATERAL FUND 2.1 Global Efficdency of the Montreal Proog At the outset, it is interesting to consider how the efficiency of the Protocol as it now stands compares with alternative approaches. First, as shown in Figure 2-1, if we assume that global benefits (B) of ODS reduction exceed the costs (C) of implementation, then the Protocol-bounded path yields ODS Use .Without* Case .............- .. (absence of Protocol) Protocol Bounds D. Time Figure 2-1 Benefits and Costs of the Phase-Out significant net benefits over the unbounded base case. Since the Protocol specifies a minimum compliance scenario, the basic objective for LDCs is to identify the path such as D which will satisfy the Protocol requirements at minimum incremental cost to the country. Suppose we write the global net benefit of ODS reduction as: NB = B - C. If the objective of the world community is to maximize NB, and if the benefits of a unit ODS reduction greatly 5 exceed the costs (at least, for initial decreases), then the optimal path may be E which lies inside D. In fact, depending on how high incremental benefits are relative to costs, there will be a family of curves (such as E and F) all of which favor faster ODS phase-out than the Protocol. This still leaves us with the difficult practical problem of valuing the economic benefits of ODS use reduction. Finally, if we recognize that there are problems (administrative, institutional, human resource etc.) that are very likely to place implementation constraints on the rate of ODS phase-out in LDCs, then curve G may be the appropriate limiting scenario. In this case, the valuation of benefits is not required in practice. Two issues arise from the foregoing discussion. The first concerns the extent to which incentives should be provided for countries to do better than the Protocol, perhaps to the limit of their implementation capability. The second concerns the exploitation, through market mechanisms and economic incentives, of the significant gains in speeding up ODS reduction in certain countries relative to others. First, let us consider the issue of incentives for accelerated phaseout. In Figure 2-2, ICD is the curve of incremental costs for phasing out ODS in a typical developing country, and IC is the corresponding curve in a relatively higher cost industrialized country. IB is the overall benefit to the global community, and this is assumed to greatly exceed incremental costs. MN represents the typical compensation received by the developing country for eliminating one unit of ODS. The net global benefit of this activity (i.e., global benefit minus the cost of providing compensation) is NS. This large surplus accrues to the global community generally; only a small fraction represents benefits for the host nation. Since IB is unknwn (but large), suppose we use MR (Le., KL) as a proxy for benefits based on the known, avoided incremental cost in the industrialized country. Even with this conservative assumption, there is still a net global benefit NR associated with the ODS eliminating activity in the developing country. The conclusion from the foregoing is that, if indeed incremental costs such as MN are small compared to net potential benefits, then the compensation provided to the developing country is the 6 Developing Country Industrialized Country Increm. Increm. IC Cost - Cost ICD ODS Substitution K ODS Substitution Figure 2-2 Incremental Costs and Benefits of ODS Substitution minimal amount. Therefore, some part of the significant net global benefit or surplus (e.g., NS, or more conservatively NR) could be used as an added incentive to accelerate ODS phaseout. 2.2 Factors Affecting Incremental Costs A number of general factors will affect the incremental costs of the transition, apart from the specific factors applying to individual ODSs. These are the role of market forces, the timing of intervention in developing countries, expenditure on informational and administrative matters, and the overall strategy adopted. Market forces. The signifirant reductions in CFC use which have occurred to date in most countries have been market driven. In some cases the use of ODS alternatives has actually been cost effective and operating costs have been reduced (e.g., in replacing CFCs in aerosols). Export and local markets have been protected where industries have stopped ODS use before being required to 7 do so by bans. Where satisfactory alternative technology exists, consumer preferences have been satisfied by non-ODS devices and preferences. External markets also exert a powerful influence over the user sectors. For example, Brazil exports compressors to the US and elsewhere, and Tunisia expects to export more than half of its refrigerator production to Europe. The ODS ban and new standards in these large markets (comprised of Protocol signatory countries) will by itself force compliance in the exporting countries. Exports of foam products and aerosols will be similarly affected. Imports of ODS and equipment will bear the international price of these substitutes. In other cases, where the objectives of the Protocol cannot be met by the free market, policy instruments that use market mechanisms are important because they can ensure that adjustment takes place in a cost-effective manner. Timing. The developing countries have been given a ten-year grace period under the Montreal Protocol. That is, while developed countries are assigned a 1986 baseline consumption (from which the percentage reductions are computed) developing countries can continue expanding consumption until 1996 after which the reductions take effect. There is some evidence that this has introduced a note of complacency in developing countries who do not feel the urgency underlying the Protocol, and also led to certain other concerns discussed below. The first concern is that because LDCs will be using a 1996 baseline for ODS use, the potential exists for unintentional or deliberate misforecasting of the 1996 ODS consumption leveL The second concern is that the Protocol does not provide incentives for aiming for a lower 1996 ODS usage, rather than ramping up consumption (especially when the levels of today are low). Illustrative growth rates for near term consumption of ODSs are given in Table 2-1. While the actions underway or planned in several countries will actually reduce ODS consumption, the larger nations will expand consumption. The costs of delayed action will be high in countries where ODS use is expanding dramatically (e.g., China, India, Brazil) because the capital stock that ultimately requires modification or replacement would likewise be growing rapidly. 8 There are two types of delay, each of Table 2-1 which imposes costs. First, delaying the target Table 2- date allows the problem to grow much worse Indicatve GrwM Rates V Ia ODS Consnpdon before corrective action is applied. Second, Current Usage h/ Average Annual Growth delays in taking action to meet a given target Counr rooo_t Rate Expected can lead to inefficient adjustment because the China gy 46 12% action taken is then too abrupt. India £y 11 15% Brazil c/ 11 8% The first type of delay is reflected in the Meico r/ 7.9 2% ten-year grace period granted to developing Yugoslavia !y 7.5 -10% countries for the implementation of the phase- Turkey d/ 4.6 -4% out targets. While this grace period is based on Egypt A/ 2.7 -6% considerations of equity, to soften the blow in Tunisia 41 0.73 6% countries least able to afford the phase-out, it 1/ Source: World Bank estimates based on various studies. may have the perverse effect of increasing their k/ Consumption in 1990 in terms of CFC-1 I equivalent i.e., Ozone Depleting Potential (01P). ultimate costs of adjustment. Delays in uf Based on projected change in ODP to 1996. pursuing ODS reduction in countries with p/ Based on projected change in ODP to 1993. potentially high growth in ODS use, will increase the difficulty of ODS elimination. Data presented at a World Bank Seminar in October, 1990 suggest that if implementation were delayed, overall incremental costs could in some countries increase by about 60 percent, and consumer costs would increase six-fold in the course of about a decade. This results from enormous growth in the domestic market for refrigerators and other CFC-using appliances which may more than offset declining ODS-substitute manufacturing costs resulting from rapid mastery of their production in developed market economies. Not only will the incremental costs be increased, but the relative burden will be shifted more to consumers as refrigerators are purchased. In China and India, it has been a conscious policy to increase the proportion of the population having access to refrigeration. Information, administration, and infrastrueture. Costs will be incurred on essential background information, administration, and infrastructure such as: - studying the ODS industry within a country; 9 - establishing the framework for administering that country's obligations under the Protocol; - developing a database on ODS production, trade, and consumption; - conducting workshops on the new technologies; - training a core group of technical experts in the application of non-ODS technologies; - demonstrating new technologies; - administering a tradeable ODS permit scheme within a country (if this is forms part of government policy to phase out ODSs); and - designing infrastructural support, e.g., recovery and recycling facilities, supply of alternative propellants like deodorized LPG; - ODS recycling and ODS destruction. 2.3 Country Incremental Costs The first major distinction to be faced in deciding what "incremental costs" are, is the one between economic costs (costs to the country as a whole) and financial costs (those incurred by individual agents, such as firms and consumers). Financial costs include various transfer payments (such as interest payments, taxes net of rebates, and duties) which are not economic costs as they do not involve the use of resources or added value. Economic costs, on the other hand, include costs borne by others even if not paid for by the firm whose actions give rise to these costs. Additional unemployment, the external costs caused by pollution and additional infrastructural costs are examples of economic costs which are frequently not "internalized," i.e., do not enter the financial calculations of firms. The overall cost to the country of compliance would be the sum of all the economic costs of all the actions undertaken. The sum of all the financial costs is not a good indicator of country costs because the financial transfer payments are not costs to the country as a whole (they are transactions among agents like banks, government, and other firms within the country) and because some significant costs are not included in financial costs. In fact, there may be large differences between the overall national economic costs and the sum of all the private financial costs (see also Box 2-2). 10 The transition from a development path without constraints on ODS use to one constrained to meet the Protocol targets will impose resource costs on the economy. In principle, by implementing an appropriate "National Ozone Policy" and by judicious selection of projects, one would like to minimize these economic costs. In any case, it is necessary to understand the nature of these overall country incremental costs in order to sharpen the discussion of the issues involved in practical estimation of incremental costs that are eligible for financing under the Multilateral Fund. The principle is to minimize the incremental economic costs subject to the constraints defined by the obligation under the Protocol. In essence the economic problem is the same as developing a least cost investment program for (say) the power sector. Electric sector plans begin with demand forecasts and a range of technical options that meet the demand. Total capital and operating costs, appropriately discounted, are minimized using an algorithm that determines the selection and timing of feasible investments subject to a variety of imposed -onstraintu. The reasonableness of initial assumptions concerning demand (a function of price) can I -.ested by extracting estimates of supply cost from the calculation. If necessary, these assumptions can be revised and the whole calculation can be performed iteratively until a desired degree of consistency and convergence is reached. In the case of ODS phase-out, there are four special technical features to take into account which make the problem more difficult than least-cost programming in other sectors. The first special feature is the nature of the "without" and "with" situations that are being contrasted in order to arrive at an incremental economic cost. These scenarios are not always clearly spelled out, which can lead to confusion. Since there is no obviously right choice, selection will have to be made according to the purpose in hand. In the "without" situation, a choice has to be made whether losses connected to trade with developed countries who have ratified the Protocol are included (say, for compensation purposes). So, for example, if we wish to compensate a developing country because it will lose export markets or incur incremental costs in making its export goods suitable for countries which have signed the Protocol, then we need to define our "without" scenario as the one representing what would have been the case if no importing countries had imposed any of the Protocol's restrictions on trade or their internal markets. If, on the other hand, we regard trade losses as non-compensable (since there is no precedent for compensation for trade losses), our "without" scenario must assume that developed countries have ratified but that the LDC in question has not done so yet. 11 In the "with" situation, a major question is whether the cost structure (and hence the price signals and resulting demand patterns) have been in any way affected by assumed levels of compensation by way of government transfer payments or grant finance from the Multilateral Fund. That is, if we assume the adjustments are made without compensation by the Fund or the government, incremental costs will be passed on by the producer or manufacturer down to his customers who, in response to higher prices, will consume less of the substitute. If, on the other hand, we assume that compensation for incremental costs flows through to the firm, his actual financial costs will be offset. To the extent the market is a competitive one, his prices will be less than they would have been and projected consumption of demand for ODS services need not be adjusted from what it would have been had no phase-out taken place. Depending on which "without" and which "with" scenario are selected, different estimates of incremental cost can be derived. The second special feature is that the demand for the services of existing ODSs must be disaggregated by ODS. There are six groups of controlled substances defined by the Protocol, each with its own chemical properties, type of use and sector of application. In principle, as outlined in Section 2-1, one should compare total cost and benefit streams obtained in an integrated framework, (i) using the benefits specific to the reduction of each particular ODS, say by weighting emissions in terms of their Ozone-Depleting Potentials,' and (ii) discounting future benefits of reduced emissions, which would value more highly the earlier reductions and encourage earlier compliance. In practice, the Protocol stipulates phase-out schedules for the substances and hence the benefits are fixed, only the costs need to be minimized. Furthermore, the phase-out schedules are applicable on a group basis: this means the problem is the sum of six separate reduction strategies with no tradeoffs allowed between substances in different groups.2 Ideally one should trade off reduction rates among all the ODSs until the marginal benefit of a unit reduction in ODS is the same for all. Also, because the Protocol stipulates reduction targets (e.g., a 50 percent reduction on baseline consumption for Group II ODSs by year 2006) the benefit stream in each scenario is implicitly the same with no acknowledgement of any additional discounted benefits resulting from earlier than normal compliance. I Because of some scientific uncertainties and data gaps on proposed weights (such as Ozone Depleting Potential and Chlorine Loading Potential) this may not be practical in all cases. ' Within groups, tradeoffs are implicitly allowed as the consumption subject to control under the Protocol is calculated as the ODP-weighted sum of consumptions in each group. 12 The third special aspect concerns data adequacy. In the case of (say) power and water sector development plans there are reasonably accurate demand forecasts and good technoeconomic data for costing alternative supply options at different scales of operation. This is not the case for ODS substitution where (i) demand patterns and price elasticities have not been extensively researched; (ii) rapid technical change is expected to have a large but unknown impact on supply costs within the planning period; and (iii) the cost reductions due to learning and the scaling up of new techniques can only be guessed at. While several alternative technologies are commercially available and have good cost data (e.g., aqueous cleaning as a substitute for CFC-113 cleaning and CFC-12 recycling system for mobile air conditioners), the cost of emerging technologies (e.g., HCFC- 141b for solvent cleaning and HFC-134a for refrigeration) are less certain. The fourth special feature is that the costs of ODS phase out will be highly distributed; they are not borne by one or two utilities or corporations but by a variety of producers, users and consumers. There may also be costs imposed on the government in the form of incremental administrative costs and additional infrastructure, and on the economy as a whole to the extent there is (say) consequential unemployment, although these may not be large (or easily quantified). Finally, since we are interested in incremental costs, the calculations of cost have to be performed twice; once for the least-cost, Protocol constrained (or "with") solution and once for the unconstrained (or "without") case. Incremental cost is the difference between two scenarios. The analysis requires projections of demand for ODS services, the constraints placed on individual ODS use (i.e., phase out profiles), a technically feasible program af appropriate substitutions and alternative technologies to meet the demand for ODS services within the constraints set for the ODSs themselves, cost data on the current technologies and the alternatives, and the choice of the least-cost approach. (See Box 2-1 for a summary of these steps.) 2.4 Effect of Country Policy on Country Incremental Costs Compliance costs are the country incremental costs calculated as the difference in economic costs between one "with" and one "without" scenario such as those listed in Table 2-2. Compliance costs should be minimized (as indicated above) by appropriate selection and timing of investments and operations to phase out ODS use. However, compliance takes place against a background of a 13 country's policies that are specific to the phase out (i.e., the National Ozone Policies discussed in 2.1 .Calculating Country Compliamce Costs Section III), as well as those that are general in application (but assumed to be fixed). These Step 1: Project demand for ODSs Step 2: Ascertain the Protocol's constraints on policies will not only influence the overall level of consumption compliance cost, but also distribute the burden of Step 3: Technically assess the most appropriate adjustment differently among various sectors of the substitutes and alternative technologies for replacing ODSs. economy. Let us consider this aspect in more Step 4: Propose alternative scenarios using these detail, substitutes/alternatives to satisfy the demand for the services of ODSs, but within the constraints for ODS consumption. Economic and industrial p icy. A Step 5: Calculate the incremental cost of each scenario relative to the without Protocol case. country's general policies will strongly influence the Step 6: Select the scenario with the minimum costs of compliance. An extreme exampic would be incremental cost; this is the compliance cost. a country that had encouraged small-scale, inefficient domestic production of CFC behind high trade barriers; in this case compliance with the Protocol would actually remove a distortion and incur a negative incremental cost. More generally, key factors will include: * The country's development strategy (e.g., degree of outward orientation); * Industry strategy (including trade protection and restrictions on ownership and competition); O Regulatory and incentive structure (including legislation and the tax structure); and o Institutional capacity to implement regulations. National Ozone Policy. In addition to the compliance costs (i.e., the minimum cost of the phase-out, given the general policies of the country) there will be inevitable additional costs introduced by inefficiencies in the adjustment process. One goal of the government adopting a national strategy for protection of the ozone layer ("National Ozone Policy") would be to minimize 14 these inefficiencies as well.? The distinction between compliance costs and inefficiencies in the adjustment process is useful to make, because these costs are related to different policies (the pre- existing industrial policy versus the adopted National Ozone Policy), and can be controlled in different ways. It may also be easy to lose sight of adjustment inefficiencies in cases, such as the example given in the last paragraph, where compliance costs are negative. 2.5 Incremental Costs to Firms and Individuals Economic agents (firms of producers and users, and individual consumers) will f4ce incremental costs as the country moves, under the influence of National Ozone Policy, to satisfy the Protocol targets. In many ways the issues and analyses of these private incremental costs parallel those of the compliance costs of countries. Thus, economic agents operating in the environment established by general government policy (as modified by National Ozone Policy) face incremental financial costs of adjustment, just as countries operating within the international political and trading environment (as modified by the Protocol) face incremental economic costs of adjustment. Rational economic agents will minimize their own financial costs, although often this will be within a highly distorted policy environment. Cost minimization may also entail strategic behavior, especially where Ozone Policy provides for compensation and other transfers. Actual incremental costs need to be evaluated in a dynamic framework -- a program with a time dimension. Consumers' incremental costs may be those incurred over the lifetime of new appliances, e.g., "premature" replacement of a CFC-based refrigerator, increased energy requirements for operation, and higher recharge expenses, followed by the earlier subsequent replacement of the refrigerator. User firms may experience one-time retooling costs followed by continued higher operating costs. Producers too would need to alter the selection and sequence of investment projects and operations. 'A good National Ozone Policy is still required even when compliance costs are negative. For example, scrapping an industry that is CFC-dependent but also uneconomic, might result in a certain net gain if handled well. If handled badly, the resulting adjustment inefficiencies might reduce this net economic gain. 15 As in the case of country compliance costs, the calculation of incremental costs has to recognize certain technical features and imposed characteristics. First, there are six groups of controlled substances. while there is some flexibility of substitution within a group (e.g., reducing relatively more halon-1211 than halon-1301 while still maintaining the overall target reduction for Group II as a whole) substitution between groups is not credited for compliance purposes, even between groups having the same target reduction profile. So economic agents must respond to at least six sets of regulations. Second, data are scarce. As the elasticities of demand for ODSs and related equipment and products are barely known, price effects on demand will be very poorly understood. This will be important for a producer who shifts to a high-cost substitute. Added to the lack of data on elasticities is the uncertainty of supply costs of substitutes from sources outside the country, or from domestic producers affiliated with multinational producers. In fact, after the Montreal Protocol, major ODS producers and users in industrialized countries invested substantial financial resources on development of substitute processes. The two large U.S. producers, Du Pont and Allied Signals have announced plans to substitute for ozone depleting CFC products by the year 2000. The substitute products are likely to have more specialized applications than existing products. For example, industry sources, expect about eight chemicals to replace the many applications made at present using CFC-11 and CFC-12. Therefore, there is considerable uncertainty about whether the substitutes will be able to benefit from the same economies of scale as the existing ODS. The decline in prices of ODS substitute manufacturing and utilization equipment would also depend on economies of scale in production in industrialized countries, which in turn would reduce the incremental costs for developing country producers. CFC prices in the past have declined by constant percentages each time production doubled. As the substitute products are being developed by the same multinational firms (Du Pont, Atochem, Allied Signals, ICI) which are price setters for specialty chemicals world-wide, one could expect past trends to continue. Third, there is the issue of passing on the costs between the affected groups. Under normal competitive conditions, incremental producer costs would be partially passed on in the form of higher prices to downstream user groups and ultimately consumers, depending on supply and demand. In this case there are additional complications. One is the extent to which incremental costs are 16 compensated by the government (or the Fund) and thus not passed on, and the other is the extent to which other aspects of National Ozone Policy are redistributive. Financial and Economic Incremental CosL Example: Modified plant to produce substitutes Catemories of Cost Private inrmental Financial/ Economic/ Public Financial Cost Component Rs million Rs million Rs million Capital N 862 862 . Taxe and dutes h/ 141 - -141 Additional fuel / 71 187 116 Retrainlag g 43 43 - Electridc ty/ 20 38 18 ladU tty .subsidy it -110 - 110 .Rad widening and new local government waste dsposal method - 4 4 Pol610on damage ~- 10 10 Total incremental 1027 1144 117g/ a. At intenital capital expenditure made in a single year. b. Customs duties on Imported components and other taxes paid are financial tiansfers but do not represent economic value-added. c. ThIs is an oil exporting county where internal (financial) petroleum product prices are below world (economic) prices, as part of national industrial strategy. In this example, future financial and economic costs are discounted at the same rate d. These are one-time worker retraining costs. c. The financial cost of electricity to the firm is below the (economic) long run marginal costs of production to the country. f. The government stimulates investment in manufacturing with a tax rebate offset against income from any source. g. This "balancing" cost is borne by the government and public at large. 2.6 Effect of Country Policy on Incremental Costs of Firms and Individuals Economic and industrial policy. General economic policy will definitely impact on the incremental private financial costs. (See Box 2-2 for an illustration of the impact of an investment tax credit.) This shows not only that different policies will entail different incremental costs for the economic agents involved in the phase-out, but also that incremental financial and incremental economic costs can differ substantially too as a result of transfer payments inherent in the tax system 17 (and as a result of other costs borne by the government and the public at large shown in the last "balancing" column). National Ozone Policy. National Ozone Policy seeks to implement the Protocol and, in the process, may redistribute the incremental costs between groups. This will have to be considered in the calculation of the incremental costs actually experienced. For example, ozone tax credits or subsidies will lessen the burden for some, perhaps even affecting the incentive structure and the demand. One important feature of ozone policy will be in providing a "level playing field," so that those firms undertaking incremental investments for compliance purposes will not be disadvantaged in a competitive market for their goods and services vis a vis more tardy firms. 18 M. ISSUES AND OPTIONS FOR NATIONAL OZONE POLICIES IN DEVELOPING COUNTRIES The standard types of policy instruments for the control of pollution are regulation, pollution taxes, marketable permits and deposit/refund schemes. These should be viewed in the light of special feature: of the pollution problem of ozone-depleting substances. 3.1 Special F?atures Affecting Choice of Policy Instruments First, there is the incentive framework provided by the Fund. Specifically, the Multilateral Fund under the Protocol has been created to reimburse the incremental costs of specified projects within the developing signatory countries. Thus the only incentive for governments in these countries to take difficult domestic policy decisions to minimize these incremental costs is moral suasion and international pressure. This is likely to be rather difficult politically because the ozone problem is seen as largely the result of developed country action. Transfers under the Fund are also seen as compensation for the costs borne by developing countries in mitigating a problem which will have its greatest impact in higher latitudes (i.e., fall more heavily on developed countries). Furthermore, the incentives within countries may tend to maximize the transfer of resources by ensuring that no eligible project or incremental cost goes unnoticed. One interpretation of the incremental costs that are eligible for reimbursement by the Fund is that they are only the private financial costs borne by firms. If so, there is still some incentive for the government to ensure that those adjustment costs not included in the financial costs are minimized. For example, it is important for a country to get the long-term industrial structure right to avoid misallocation of investment and other resources. This means among other things that consumer investment in CFC-techrclogies should be limited to prevent long-term welfare losses due to lack of serviceability. Second, there is a complex distributional issue. Supposing it were decided to speed up the phase-out in order to maximize benefits, as suggested in Section 2-1. In that case, restrictions on CFC production and use would be sudden and severe, and might impose heavy losses on entrepreneurs, e.g., owners of production facilities in the developing countries. Thus if ODS bans are implemented immediately, governments will be tempted to provide adequate compensation to firms penalized by these requirements, rather than to adopt a "polluter pays principle" whereby the 19 losses are simply absorbed where they fall. However, reimbursement of a producer's incremental costs is a subsidy for the new technology (e.g., for the manufacture HCFCs) and will be inefficient since it also subsidizes commercial risk and implicitly penalizes other forms of substitution. However, for domestic distributional reasons and the way the Fund is set up, this efficiency loss may be unavoidable if a quick response is made. In fact though, LDCs are not under this time pressure and have been granted a ten-year grace period. However, taking full advantage of this grace period may not be in a country's best interests for the reasons mentioned in Section 2-2. Also, depending on how competitive the various ODS industries are within the country, the benefits of any grant financing for incremental costs may or may not be passed on to final consumers. Manufacturers may in fact have a windfall gain. Third, it is generally acknowledged that the benefits of CFC phase out (avoided damage due to ozone depletion) far exceed the relatively modest cost of the phase out. See for example, UNEP (1989). This greatly favors the rise of quantitative instruments (bans, percentage reductions etc.) over price instruments. The latter are less certain in their effects (entailing a risk of significant loss of benefits) even if less efficient (although efficiency losses are likely to be small in comparison). Fourth, the urgency of the problem makes it unlikely that a new philosophy of environmental pollution control can be established for ozone-depleting substances. Historically, pollution has almost always been the subject of "command-and-control" approaches, and this is still generally the case in the U.S. today. Sophisticated approaches such as marketable permits should be used wherever such systems are used for domestic pollution control, but there will be few such instances in developing countries. Fr more effective approaches, given the short time span available, will be mandated technology standards, bans, reductions, and direct government investments. Fifth, there will be minimal concern over the dynamic incentives of the policy instruments. Developing countries are followers in the field of ODSs and ODS substitutes and the most effective way for them to acquire new technology will be through technology transfer from multinational corporations in the developed signatory countries. 20 Sixth, the problem is a transitional problem, not one of constructing an efficient long-r-in equilibrium. Ozone-depleting substances will have to be phased out completely (i.e., the long-t.it "optimum discharge rate" is zero or very close to zero). Seventh, the industry structure is stch that there are very few producers and importers (sometimes just a couple), several user firms (e.g., refrigerator manufactures) and vast numbers of consumers. Thus requirements of centralized information and ease of enforceability clearly suggest that direct controls be applied on upstream producers and users, while downstream users should be left to respond to market forces. Eighth, because of obvious informational and enforcement problems. control must be exercised on the production and disposal of the ODSs, not their emissions as such. It is far easier to monitor and control production, import, recycling and disposal of CFCs and halons, rather than their direct release to the atmosphere which occurs after long lag times from millions of point sources. 3.2 National Strategies There are three broad country types insofar as national ozone policies are concerned, distinguished by import dependency. * Those that import everything, i.e., ODSs and ODS using equipment. Most of the smaller developing countries are in this category. * Those that manufacture some equipment themselves, such as refrigerators, but import ODSs. Several medium-sized developing countries are in this category such as Egypt, Yugoslavia, and Turkey. * Those that manufacture at least some ODS (and generally also the equipment). Some larger developing countries are in this category, notably China, India, Brazil and Mexico. 21 Several countries are also linked to external marketi through exports as well as through imports. Many of these linkages are with signatory countries and, to protect their overseas markets, the ODS industries in these countries are likely to retool and adjust anyway. For example, Brazil exports refrigerant compressors to the United States and Tunisia hopes to expand into the European market for domestic refrigerators. Adjustments they make for export markets will be available for servicing internal markets as well. In this case, the major national policy issue in adjusting to the Protocol are to structure incentives for limiting purely domestic use. The general policy imperative, would be to minimize the domestic (economic) costs of adjustment net of any external transfers of technology (from multinational corporations) and resources (through the Fund) likely to be available. As a minimum, and whatever category the country falls in, National Ozone Policy will entail: (i) giving domestic firms and consumers early warning of the future measures and cutbacks required. In this way, adjustment will be less disruptive and costly; (ii) administering the compensation payments available under the Fund, i.e., identifying eligible projects and assisting in quantification of incremental costs; and (iii) providing information, and possibly technical assistance and retraining, to facilitate adoption of technologies already cost-effective (such as use of alternative aerosol propellants), or adjustment to emerging changes in export markets. It is unlikely that there will be significant employment or other macroeconomic effects of the phase out In most cases, the ODS industries constitute a very small part of total industrial activity. The general approach suggested in the discussion on policy instruments, is one centered on quantities of ODS in circulation, not directly on emissions. This is also in line with the Protocol itself which seeks limitations on total ODS consumption, defined as net imports plus production less destruction. Direct regulation on the amounts available upstream provide the certainty needed to 22 meet phase out targets, but allowing consequential price rises to flow onto downstream users provides some of the efficiency of charge-based policy instruments. Other general interventions would be those that facilitate the adjustment, including market responses where appropriate, such as provision of information or technical advice and retraining of servicemen to limit consumers welfare losses associated with existing ODS appliances. Some special attention may be needed if there are general economic distortions that interfere with economic adjustment and if there are any "loopholes" due, for example, to imports of equipment containing ODS, stockpiling, or recycling. Controls affecting consumption. These will affect all three categories of country: the ones fully import-dependent through to ones with their own production capacity. In this case the primary policy instrument will be a regulation on the ODS supply quantities. While this will provide sure control over the phase out, there are some problems. First, there may be more than one supplier (importers and producers) and the supply quotas will have to be allocated among them. Auctioning marketable permits may suffer from buyer collusion as it is unlikely that there are many importers. This may not matter very much, the main advantage of these permits is certainty of impact. Allocation could also be done pragmatically either by voluntary industry agreements (as in Mexico) or by grandfathering, i.e., by equal percentage reduction for existing importers. Second, price increases resulting from the quantity controls on ODS are likely to give suppliers a windfall profit until the phase-out is complete. On equity grounds, (and perhaps to raise revenue) the government could introduce a windfall profit tax, or impose a compensating excise duty on the ODS. Note that this fiscal measure is not a primary instrument, it merely corrects for some of the bluntness of the regulatory instrument which is primary. Third, although the Protocol stipulates a few target dates for specified reductions, it is otherwise silent on the required reductions in any given year. National Ozone Policy can fill in the targets for the intervening years as well. In fact this would help to minimize overall adjustment costs, since doing nothing until the target date would make the transition abrupt and costly, with steep consequential price changes flowing on to downstream users. 23 Fourth, the broad approach advocated above assumes that the market downstream will allocate restricted supplies of ODS among users and induce switches in a cost-effective manner: rising ODS prices will cause substitution by those users who can switch most cheaply. While this is a reasonable presumption, it does not necessarily assure a cost-effective outcome for the country. Major distortions in the ODS market may affect the outcome and negate its economic cost- effectiveness. For example, taxes and/or subsidies may affect the relative prices of substitutes for various users. If there are large distortions like this some corrective may be needed, otherwise they should be ignored since the time and cost of information gathering may make it not worthwhile pursuing. Fifth, the control of supplies alone does not ensure that ODS use is reduced. The quota could be partly circumvented by stockpiling ODS in the years before the quotas take effect (not only creating an alternative source of supply but also building up a bigger baseline of imports) or by importing goods with ODS incorporated in them (such as aerosol cans and refrigerators). If these appear to be anything other than minor or temporary flaws, then these sources of ODS should be included in the total sources of supply and subject to the same supply reductions. For a given consumption target, there are supplementary means by which a government can reduce welfare losses. One is to promote conservation of ODS. For example, large quantities of CFCs are vented during the servicing of domestic refrigerators in developing countries; appropriate training and technical assistance to servicemen can reduce such losses dramatically and help sustain the economic life of existing ODS-dependent appliances. A second approach is for recovery of ODS from old appliances for recycling, as recycled supplies are not part of consumption controlled by the Protocol. Even if it is uneconomic to recycle recovered ODS, they should be destroyed (say, by incineration) because destruction increases the allowable level of imports (the Protocol target is net of destroyed ODS). A third method is to encourage retrofits on large equipment to make it suitable for using alternatives. The costs of training technical assistance, recovery, recycling, destruction and retrofits are "incremental" and eligible for grant financing under the Montreal Protocol. Ideally recovery and recycling schemes should be instituted to the extent that they are cost- effective. The benefits are either that a greater amount of ODS can remain within the system for any given target on net consumption, thereby reducing adjustment costs, or that ODS consumption 24 Sixth, the problem is a transitional problem, not one of constructing an efficient long-run equilibrium. Ozone-depleting substances will have to be phased out completely (i.e., the long-run "optimum discharge rate" is zero or very close to zero). Seventh, the industry structure is such that there are very few producers and importers (sometimes just a couple), several user firms (e.g., refrigerator manufactures) and vast numbers of consumers. Thus requirements of centralized information and ease of enforceability clearly suggest that direct controls be applied on upstream producers and users, while downstream users should be left to respond to market forces. Eighth, because of obvious informational and enforcement problems control must be exercised on the production and disposal of the ODSs, not their emissions as such. It is far easier to monitor and control production, import, recycling and disposal of CFCs and halons, rather than their direct release to the atmosphere which occurs after long lag times from millions of point sources. 3.2 National Strategies There are three broad country types insofar as national ozone policies are concerned, distinguished by import dependency. * Those that import everything, i.e., ODSs and ODS using equipment. Most of the smaller developing countries are in this category. * Those that manufacture some equipment themselves, such as refrigerators, but import ODSs. Several medium-sized developing countries are in this category such as Egypt, Yugoslavia, and Turkey. * Those that manufacture at least some ODS (and generally also the equipment). Some larger developing countries are in this category, notably China, India, Brazil and Mexico. 21 Several countries are also linked to external markets through exports as well as through imports. Many of these linkages are with signatory countries and, to protect their overseas markets, the ODS industries in these countries are likely to retool and adjust anyway. For example, Brazil exports refrigerant compressors to the United States and Tunisia hopes to expand into the European market for domestic refrigerators. Adjustments they make for export markets will be available for servicing internal markets as well. In this case, the major national policy issue in adjusting to the Protocol are to structure incentives for limiting purely domestic use. The general policy imperative, would be to minimize the domestic (economic) costs of adjustment net of any external transfers of technology (from multinational corporations) and resources (through the Fund) likely to be available. As a minimum, and whatever category the country falls in, National Ozone Policy will entail: (i) giving domestic firms and consumers early warning of the future measures and cutbacks required. In this way, adjustment will be less disruptive and costly; (ii) administering the compensation payments available under the Fund, i.e., identifying eligible projects and assisting in quantification of incremental costs; and (iii) providing information, and possibly technical assistance and retraining, to facilitate adoption of technologies already cost-effective (such as use of alternative aerosol propellants), or adjustment to emerging changes in export markets. It is unlikely that there will be significant employment or other macroeconomic effects of the phase out. In most cases, the ODS industries constitute a very small part of total industrial activity. The general approach suggested in the discussion on policy instruments, is one centered on quantities of ODS in circulation, not directly on emissions. This is also in line with the Protocol itself which seeks limitations on total ODS consumption, defined as net imports plus production less destruction. Direct regulation on the amounts available upstream provide the certainty needed to 22 meet phase out targets, but allowing consequential price rises to flow onto downstream users provides some of the efficiency of charge-based policy instruments. Other general interventions would be those that facilitate the adjustment, including market responses where appropriate, such as provision of information or technical advice and retraining of servicemen to limit consumers welfare losses associated with existing ODS appliances. Some special attention may be needed if there are general economic distortions that interfere with economic adjustment and if there are any "loopholes" due, for example, to imports of equipment containing ODS, stockpiling, or recycling. Controls affecting consumption. These will affect all three categories of country: the ones fully import-dependent through to ones with their own production capacity. In this case the primary policy instrument will be a regulation on the ODS supply quantities. While this will provide sure control over the phase out, there are some problems. First, there may be more than one supplier (importers and producers) and the supply quotas will have to be allocated among them. Auctioning marketable permits may suffer from buyer collusion as it is unlikely that there are many importers. This may not matter very much, the main advantage of these permits is certainty of impact. Allocation could also be done pragmatically either by voluntary industry agreements (as in Mexico) or by grandfathering, i.e., by equal percentage reduction for existing importers. Second, price increases resulting from the quantity controls on ODS are likely to give suppliers a windfall profit until the phase-out is complete. On equity grounds, (and perhaps to raise revenue) the government could introduce a windfall profit tax, or impose a compensating excise duty on the ODS. Note that this fiscal measure is not a pnmary instrument, it merely corrects for some of the bluntness of the regulatory instrument which is primary. Third, although the Protocol stipulates a few target dates for specified reductions, it is otherwise silent on the required reductions in any given year. National Ozone Policy can fill in the targets for the intervening years as well. In fact this would help to minimize overall adjustment costs, since doing nothing until the target date would make the transition abrupt and costly, with steep consequential price changes flowing on to downstream users. 23 Fourth, the broad approach advocated above assumes that the market downstream will allocate restricted supplies of ODS among users and induce switches in a cost-effective manner: rising ODS prices will cause substituticn by those users who can switch most cheaply. While this is a reasonable presumption, it does not necessarily assure a cost-elective outcome for the country. Major distortions in the ODS market may affect the outcome and negate its economic cost- effectiveness. For example, taxes and/or subsidies may affect the relative prices of substitutes for various users. If there are large distortions like this some corrective may be needed, otherwise they should be ignored since the time and cost of information gathering may make it not worthwhile pursuing. Fifth, the control of supplies alone does not ensure that ODS use is reduced. The quota could be partly circumvented by stockpiling ODS in the years before the quotas take effect (not only creating an alternative source of supply but also building up a bigger baseline of imports) or by importing goods with ODS incorporated in them (such as aerosol cans and refrigerators). If these appear to be anything other than minor or temporary flaws, then these sources of ODS should be included in the total sources of supply and subject to the same supply reductions. For a given consumption target, there are supplementary means by which a government can reduce welfare losses. One is to promote conservation of ODS. For example, large quantities of CFCs are vented during the servicing of domestic refrigerators in developing countries; appropriate training and technical assistance to servicemen can reduce such losses dramatically and help sustain the economic life of existing ODS-dependent appliances. A second approach is for recovery of ODS from old appliances for recycling, as recycled supplies are not part of consumption controlled by the Protocol. Even if it is uneconomic to recycle recovered ODS, they should be destroyed (say, by incineration) because destruction increases the allowable level of imports (the Protocol target is net of destroyed ODS). A third method is to encourage retrofits on large equipment to make it suitable for using alternatives. The costs of training technical assistance, recovery, recycling, destruction and retrofits are "incremental" and eligible for grant financing under the Montreal Protocol. Ideally recovery and recycling schemes should be instituted to the extent that they are cost- effective. The benefits are either that a greater amount of ODS can remain within the system for any given target on net consumption, thereby reducing adjustment costs, or that ODS consumption 24