E N-V I R O N M E N T r 63 ~~~D E P A R T M E N T PA P-ERS PAPER NO. 053 TOWARD ENVIRONMENTALLY AND SOCIALLY SUSTAINABLE DEVELOPMENT Environmental EConomics Series Market Based Instruments For The Implementation Of The Montreal Protocol In Developing Countries Nikhil Desai Subodh Mathur November 1996 Environmentally Sustainable Development The World Bank ESD i t Global Environment Division Market Based Instruments For The Implementation Of The Montreal Protocol In Developing Countries November 1996 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 Division, Room S-2145. This paper was prepared under the direction of Andres Liebenthal (The World Bank). Various colleagues reviewed parts or the whole of earlier drafts since September 1995; we are in particular indebted to Jessica Poppele, Ajay Mathur, and Joachim von Amsberg. The final draft was reviewed by Roy Pepper, David Hanrahan, Andres Liebenthal, and Stephen DeCanio; their questions and recommendations are deeply appreciated. Of course, the authors alone are responsible for the views presented and any remaining errors. Contents Section Page Glossary Summary 1. Introduction 1 2. National Obligations vs. Firm-level Controls 5 3. Scope for Market Based Instruments 11 4. Bid Auctions for ODS Phaseout 19 5. Other Market Based Instruments 31 6. Conclusions and Recommendations 41 Bibliography 43 Annex: Supplemental Tables 47 Environmental Economics Series Market Based Instruments for the Implementation of Montreal Protocol in the Developing Countries Environment Department Papers Glossary ABBREVIATIONS CFC (class) Chloro-fluorocarbons HCFC (class) Hydrochloro-fluorocarbons MFIMP Mulfilateral Fund for the Implementation of Montreal Protocol (the "Fund") MFEC Executive Committee for the Fund ODP Ozone Depletion Potential ODS Ozone Depleting Substance UN United Nations UNDP United Nations Development Programme UNEP United Nations Environment Programme UNIDO United Nations Industrial Developent Organization PRINCIPAL ODS AND THEIR ODP (MONTREAL PROTOCOL VALUES) Trade name Formula Relative ODP (to CFC-11) Carbon tetrachloride (CTC) CCl4 1.1 CFC-11 CFC1. 1.0 CFC-12 CF2C4 1.0 Halon-1211 CF.BrCl 3.0 Halon-1301 CF3Br 10.0 Methyl chloroform (MCF) CH3CC13 0.1 NarATioNs: This paper's references to ODS are strictly to the five CFCs and three halons in the Protocol Annex A. References to $ are for US dollars. Environmental Economics Series Envirorunent Department Papers Summary 1. Established in 1990, the Multilateral according to the UN assessment rules. Fund ("the Fund") for the Implementation of the Operationsal authority over the use of the Montreal Protocol on Substances that Deplete Fund's resources is vested in its Executive the Ozone Layer ("the Protocol") is provided Committee (MIFEC). The Fund disburses these with the mandate "to meet, on a grant or resources via its Implementing Agencies (the concessional basis, the agreed incremental World Bank, UNDP, UNIDO, and UNEP). Its costs" of developing country Parties (those financial assistance to date primarily has operating under Article 5.1 of the Protocol) taken the form of grant awards to individual "to enable their compliance with the control firms engaged in the use of ODS undertaking measures" of the Protocol. These control phaseout via conversion to permissible ODS measures mandate reduction and eventual substitutes. These grants are approved by the elimination of the "consumption" (defined as MFEC on a project-by-project basis according aggregate production plus net imports plus to administratively established assessment stock increase) of virgin (newly produced) criteria for certain categories of "eligible ozone depleting substances (ODS) according incremental costs". Willing candidate firms to schedules established by the Protocol. have to be identified and canvassed by the Implementing Agencies, who submit the 2. The Protocol's control measures are not grant requests on behalf of the national yet binding on the developing country Parties. governments. Considerable administrative While many developing countries have ex- resources at the multilateral level are ex- pressed an intent to achieve an earlier compli- pended to identify and develop the projects, ance than the Protocol commits them to, no calculate the "eligible" need, prepare grant control measures currently apply at the level of proposals, negotiate the amounts, and imple- individual firms in most developing countries. ment the approval, disbursement, and verifi- The Fund has been empowered by the Parties, cation mechanisms. however, to use its assistance "as an incentive for early adoption of ozone-protecting technolo- 4. The Fund's current approach of need gies". The Fund's assistance to date and for assessment implicitly presumes that the firm's some more time may therefore be character- optimal phaseout (in terms of timing and ized as voluntary transactions for an acceler- technological option) and associated capital ated phaseout, or over-compliance, with the and operating costs can be administratively Protocol. determined, and that the firm's conversion project can be isolated from its overall opera- 3. The developed country Parties (i.e., tions and objectives. It ignores certain firm- those not operating under Article 5.1 of the specific parameters (e.g., remaining economic Protocol) are obliged to meet the agreed life of existing equipment, conversion project incremental costs of developing countries' economic life, discount rate, future expectations, compliance with the Protocol. In practice, the and collateral economic benefits of conversion) Fund receives periodic voluntary grant which determine the firm's incremental cost of contributions from the developed countries conversion and its choice of when to convert, Environmental Economics Series Market Based Instruments for ODS Phaseout in the Developing Countries because it is difficult or impossible to know for a national system of controls, in particular these variables. There is a risk that adminis- establishing enforceable firm-level limits on trative assessment distorts the firms' behavior ODS use, necessary to ensure that the Fund's from least-cost choice of technology and assistance, and the country's phaseout strategy, timing and that the Fund's grants meet more is effective. The emphasis is on a simple than the incremental costs a firm may be exposition of options at the Fund's disposal, wiDing to accept for conversion. subject to individual developing countries' consent (i.e., with no need for multilateral 5. The project-by-project approach, and agreement). Design and implementation details the associated administrative determination of require further analysis for specific situations. qualified grant level, may have been necessary in the early years of the Fund's operations 8. Toward the first objective, the paper when there was little experience with ODS- proposes that the Fund's assistance to firms substitute technologies, few candidates willing using ODS be seen as a "market" for ODS to convert, and no legal limits on ODS use. phaseout - the firms providing a verifiable, Over time, these conditions have changed and quantifiable service in return for a unit price ($/ will continue to change, creating greater kg ODP) - and advances a conceptual pro- incentives and lesser risks to conversion. posal to use sequential competitive bid auctions Already, the demand for Fund grants have at the national level. Such a scheme could exceeded available resources, as more projects apply to all firms across all sectors and project are developed. At the same time, projects types (including shutdown), or could be limited approved to date have shown a wide variation to certain segments at a time. Each participat- (across countries and end-use sectors, project ing firm would submit a technical proposal and size, and over time) in project-level unit grant a sealed cost proposal. Evaluation of the levels ($/kg ODP, termed "cost-effectiveness" technical proposal would be limited to the by the Fund) assessed and approved under extent of verifying the amounts of ODS use the current rules. firm is entitled to, and of the verifiability of the service offered (via conversion or closing the 6. Beginning in 1995, the Fund has used operations). Subsequently, technically qualified some sector-level "cost-effectiveness threshold firms' cost proposals would be ranked accord- values" (CETV, in $/kg ODP terms), for its ing to their "offer price" (in $/kg ODP) and grants, and has also permitted "partial fund- their offers accepted until the pre-set available ing" of assessed eligible expenditures if the funds are exhausted. beneficiary firm elects to meet the remainder so as to maintain the grant within the threshold 9. A competitive bid auction is fair in that values. While this approach has implicitly a technically qualified firm's offer price may be introduced partial elements of competition rejected only if other firms' cheaper offers among firms, within particular use sectors, the exhaust the available funds; the winning firms' projects are approved essentially on a "first- incremental costs of compliance are met pre- come, first-served" basis. Firms have an cisely as they perceive them to be. At the same incentive to choose technologies and timing just time, competitive pressures would drive the so as to meet the CETV, not less; sectoral CEIVs participating firms to bid close to their true vary considerably; and, self-selection is at best costs (net of expected benefits) in order to limited. An explicit scheme to require firms to maximize the probability of winning and self-select, and to ensure that phaseout pro- minimize that of losing out to a competitor. ceeds along the efficient path of achieving the cheaper reductions before the more expensive 10. A competitive bid auction is also ones is desirable. efficient in the sense that individual firms have an incentive to choose least-cost options and 7. This paper (i) explores alternatives to that, among firms, cheaper phaseout projects the current approach that offer efficiency gains are implemented before more expensive ones. - i.e., increasing the ODS use phaseout for a Precise savings as compared to the current given sum of money - to the Fund and an administrative assessment approach would individual developing country; and (b) argues depend on the design and implementation of a ii Environment Department Papers Summary competitive bid auction scheme in country- specific situations. This paper does not offer a (d) A competitive bid auction particular model of conducting a bid auction; provides such a mechanism. In a competitive specific timing and rules of such auctions auction, a firm's "offer price", or "willingness depend on country-specific factors (e.g, market to accept", reveals its own estimate of the size and composition). Examples of altemative "incremental cost" incorporating its own rules that can be used in auctions include: expectations of the costs and benefits of conver- guaranteed floor price, maximum acceptable sion. It may well offer a higher price than what price, "set-asides" (i.e., separate auctions with it is willing to accept, but in so doing it risks different quota) or "standard price" (i.e., fixed losing out to a competitor. If it estimates that its unit price) options for small firms, sector- or offer price would be lower some years hence, it substance-specific auction without a follow-on could either wait until a later auction round, or auction, permissibility of re-bidding, and the enter a bid now and take a chance that its like. current offer is indeed cheaper as compared to its competitors. Thus, competitive bidding 11. The essential argument rests on the would come close to revealing the firms' following propositions: minimum "willingness to accept' and optimize the timing of conversion across firms. (a) Accepting that the Fund's assistance is to be applied at the firm level, (e) An accepted bid, in turn, rather than Party (country) level, and to firms constitutes a mutual agreement between the engaged in ODS use rather than "consumption" firm and the Fund that the total amount (production and imports) only, the incremental awarded is adequate compensation for the costs that ought to be financed by the Fund are service to be provided; the "agreed incremental the minimum net costs; there is no assurance that cost" criterion is therefore satisfied, without the project-by-project "need assessment" having to externally assess the firm's costs and approach under current rules does that. benefits. (To reiterate for emphasis, it is impos- sible to administratively determine the mini- (b) It is impossible to administra- mum net costs of conversion). The potential for tively determine the minimum net costs of one firm gaining competitive advantage over conversion. Conversion project costs and others is minimized, and the country is able to benefits vary across firms and over time, and fulfill its phaseout obligations in the least-cost are influenced by variables - such as discount manner. rate, expectations of market conditions (includ- ing for ODS, their substitutes, and final prod- 12. Rejected bidders can compete in a ucts firms sell), and market strategy - which subsequent round of phaseout so long as the are specific to individual firms, and best known Fund continues to assist the country in meeting to the firms themselves. its compliance obligations. Successive auctions need not result in higher offer prices, however; (c) Viewed from the overall to the extent that over time, the cost of prema- perspective of protecting the atmosphere, a kg. ture replacement of existing capital equipment of phaseout offered by one firm is no different declnes, the price differential between ODS and from that offered by another; given a budget its substitute declines, technological growth constraint, the Fund should choose cheaper and adaptation reduce net conversion costs, phaseout before the more expensive one. The and the firms perceive greater advantage in challenge is to devise a mechanism whereby all conversion due to market shifts, their offer offerors of phaseout compete at the same time, prices can be expected to decline correspond- instead of administratively selecting the offerors ingly. and then paying them the amounts deemed to be "eligible" even as other offerors may have a 13. A national-level bid auction would be lower unit price. It is not necessary to separate based on competitive procurement practices in "incremental capital costs" and "incremental use in a given country, and would require only operating costs", or distinguish by use sectors, a slightly different administrative structure in as under current rules. ODS phaseout as compared to the current one; Environmental Economics Series i Market Based Instruments for ODS Phaseout in the Developing Countries for instance, expanding technical assistance and limiting windfall gains or losses. While activities at national and sectoral levels (so that they involve flows of funds away from firms all firms can better evaluate their options), and rather than to them (which is the Fund's providing at least the first-level authority to objective), the resultant revenues may be used accept phaseout offers at the national level to supplement the Fund's resources. These (instead of MFEC approval for every firm's instruments are not within the Fund's author- conversion). Bid auctions as proposed here ity to use, and may only affect firm-level should be applied only at the national level, not financial calculus without necessarily affect- at the international level (whether between ing national economic costs of compliance; countries or between firms from different however, their use may supplement the countries). developing country government resources to implement the ODS phaseout, and provide 14. Whether or not the Fund continues flexibility and predictability to firms. Also, with the project-by-project need assessment compliance with the Protocol means honoring approach or permits the use of national aggregate quantity limits according to a competitive bidding for grants, it is impera- particular schedule; therefore, pricing policies tive that national legal and regulatory struc- alone risk being insufficient or overshooting tures be established to translate Party-level the target. compliance commitments into firm-level permits or "entitlements", with legal limits on 16. Specific recommendations are that (i) market entry (by new firms) or re-entry or the Fund accept bid auctions as a means of expansion (by existing firms). Without such grant transfer within those developing coun- enabling legislation, there is a risk that the net tries which wish to use them, and encourage ODS phaseout may be less than what the their efforts via pilot schemes; (ii) the Fund Fund's assistance provides for (if the aggre- provide incentives to national governments, gate use rises even if some individual firms perhaps in the context of a broader policy convert). Corresponding to such controls, dialogue, to use other market-based instruments administrative structures to maintain regis- (pricing policies and/or use of tradeable tries of eligible firms and enforce the phaseout permits for ODS consumption or use) which and bans are also needed. can offer greater overall effectiveness of the Fund's grants; (iii) developing countries 15. Bid auctions are a particular form of establish firm-level limits (via permits) on ODS market-based instrument (MBI) to achieve a use, production and trade, both in order to regulatory goal. Other market-based instru- monitor compliance with the Protocol and to ments such as pricing policies for ODS (e.g., protect the phaseout achieved; and, (iv) the taxes) and substitutes (e.g., subsidies) and Fund accord greater priority to conversion of inter-firm tradeability of ODS production, ODS production (to substitute production) and import or use permits can be very powerful in recycling and reclamation activities than it has providing the appropriate signals to firms so far. iv Environment Department Papers 1 Introduction Under the 1987 Montreal Protocol (the with the control measures of the Protocol". "Protocol") on Substances that Deplete the (UNEP 1995b, p. 1) The Fund receives assessed Ozone Layer and its amendments subse- grant contributions from the developed quently adopted in London (in 1990) and country Parties (i.e., those not operating under Copenhagen (1992), the signatory national Article 5.1 of the Protocol). Operational Parties have committed themselves to a authority over the use of the Fund's resources gradual phaseout of the domestic production is vested in its Executive Committee (MFEC). as well as consumption (defined as domestic The Fund disburses these resources via its production minus destruction plus net im- Implementing Agencies (the World Bank, ports) of certain virgin (i.e., newly produced) UNDP, UNIDO, and UNEP). The agencies are ozone depleting substances (ODS)1. As of required to "apply only those considerations January 1996, the developed countries2 ceased relevant to effective and economically efficient virgin "consumption" of certain ODS whose programmes and projects consistent with pre-Protocol "consumption" level was over criteria adopted by the Parties." (ibid., p. 13, 1,000,000 tons per year; the use of these ODS emphasis added). may continue for some time, depending on the inventories of previously produced (including While no control measures currently recycled) ODS. The developing countries are apply to developing countries, the Fund has granted a grace period for compliance with the been empowered by the Parties to use its Protocol, whose "consumption" control assistance "as an incentive for early adoption of measures will become binding upon them ozone-protecting technologies" or, in other beginning 1999 (with a freeze, followed by words, for an accelerated phaseout or over- reductions beginning 2005). Many developing compliance with the Protocol. The Fund's countries have expressed a desire to accelerate assistance to date primarily has taken the form their compliance schedule, indicating that an of grants to individual firms engaged in the use accelerated phaseout may be beneficial to of ODS undertaking phaseout via conversion to them. permissible ODS substitutes. These grants are approved by the MFEC on a project-by-project Overall, the task is to phase out annual basis according to administratively established virgin ODS "consumption" in the developing assessment criteria for certain categories of countries of about 150,000 to 200,000 tons per "eligible incremental costs" for each firm and year (or more, up to limits established by the conversion project. As of end-1995, the Fund average between 1995 and 1997), and a smaller has approved over $400 million of grants, quantity of virgin ODS production. In 1990, the expected to result in elimination of over 60,000 Parties established a Multilateral Fund for the tons per year of ODS use. By the end of 1995, Implementation of Montreal Protocol ("MFIMP" about 7,500 tons of ODS use (but a negligible or the "Fund"), to fulfill the Protocol mandate to amount of ODS production) had been phased "meet, on a grant or concessional basis as out with Fund's assistance. appropriate, and according to the criteria to be decided upon by the Parties, the agreed incre- For the first four years of the Fund's mental costs" of these countries' "compliance operations, available funds exceeded the Environmental Economics Series Market Based Instruments for the Implementation of Montreal Protocol in the Developing Countries demand for grants; beginning 1995, the to facilitate the ODS use phaseout: (a) allocation situation reversed. This suggests that it would of Fund grants via competitive auctions (similar be useful to devise economically efficient to other bulk procurement schemes); (b) making methods that would bring about the greatest use permits or entitlements tradeable, whether ODS phaseout with the available funds they are issued by an administrative mecha- (including via reduction in the administrative nism or sold to firms via competitive auction; costs of grant delivery). Also, in the absence of and (c) providing appropriate price signals to developing countries' adoption of domestic firms, particularly via taxes on sales or use of regulatory measures to control ODS use, there ODS and/or subsidies on sales or use of is a risk that the Fund's assistance may be permissible substitutes. The primary attention frustrated by subsequent growth. Finally, in this paper is on the first type of MBI, ad- because the Protocol only requires controls on vancing a conceptual proposal to use national- domestic "consumption" (production and net level sequential bid auctions for firms who imports), phasing out ODS use without corre- offer "ODS phaseout" in return for Fund sponding reductions in production and net grants. imports may provide perverse price signals to firms and permit an increase in ODS stockpile The essential argument for a competi- which may negate the gains of an accelerated tive bid auction, in contrast to the project-by- phaseout3. project approach, is as follows: the intrinsic or "true" incremental costs, net of expected cost Certain shortcomings of the Fund's savings and collateral benefits, are best known current project-by-project approach with to the firm itself, and cannot be determined administratively (i.e., external to the firm) externally by using general rules. "ODS assessed "eligible incremental costs" are as phaseout" (measured in ODP terms) is a service follows: (i) firms' phaseout decisions may be of uniform quality, i.e., one kg. of ODP phaseout distorted away from a least-cost path; (ii) firms from one source is indistinguishable from that may be overcompensated for replacement of of another. Accepting firms' offers of ODS existing equipment with limited remaining phaseout in the increasing order of $/kg ODP economic life; (iii) cost savings and collateral ensures that the cheaper conversions are benefits to the firms as a result of conversion achieved before the more expensive ones. Also, cannot be captured; (iv) the incentives to a technically qualified firm's offer may be minimize the grant request are at best limited; refused only because another firm has beaten its and, (v) in-depth reviews and considerable offer price; hence, the auction approach is likely scrutiny at multiple levels increase the admin- to be seen, at least from the firms' point of view, istrative costs of grant delivery. as fairer and more transparent than the admin- istrative project-by-project approach. This paper has two broad objectives: (a) to explore alternatives to the current approach The key point is that, under a competi- to the phaseout of ODS use that may offer tive auction or procurement scheme, firms are efficiency gains - i.e., increasing the ODP compelled to reveal their intrinsic incremental phasedout for a given sum of money - to the costs in the form of "offer price" or "willingness Fund; and, (b) to argue for a national system of to accept" (WTA). An accepted bid constitutes controls, in particular establishing enforceable the agreement between the firm and the Fund firm-level limits on ODS use, necessary to (or the national-level auctioneer as the Fund's ensure effective use of the Fund's assistance representative) on the "incremental costs" of (so that use reductions also result in consump- phaseout at a given point in time, and would tion reductions). The specific focus is on thus qualify as an "agreed incremental cost". If laying out a conceptual framework for the the maximum acceptable bid (in $/kg ODP) is Fund's use of market-based instruments set low enough to achieve greater phaseout for a (MBIs) that provide incentives to firms for self- given sum of money, the competitive auction selection, continued innovation, and choosing would clearly guarantee savings as compared the least-cost path to phaseout to the current approach. Precise levels of savings would depend on sector, country, and Three types of MBIs may be considered design of the auction/procurement scheme. 2 Environrment Department Papers Introduction This paper does not provide a model for any ered beyond the scope of this paper, primarily particular country or scheme, nor does it because they are subject to the jurisdiction of propose that bid auctions be used at the inter- national governments, not the Fund. Some country level. Rather, it recommends that the limited observations are offered, however, Fund encourage efforts to use national-level outlining conceptual and design issues as well bid auctions and other MBIs for ODS phaseout as political and administrative requirements of and provide incentives to countries which the use of these other MBIs. Used in combina- choose to engage in compliance strategies tion with firm-level controls on ODS production which offer greater overall effectiveness of the and imports, they may well supplement the Fund's grants. developing country government resources for ODS phaseout; provide flexibility and predict- Environmental mandates which ability to firms; and, affect the firms' "willing- impose aggregate time-bound quantity commit- ness to accept" (expressed in their auction ments, as does the Protocol, require that corre- offers), and thereby "agreed incremental costs" sponding obligations be imposed on individual (i.e., accepted offers). Indeed, the Fund's use of firms which cause the environmental damage. bid auctions can be seen as a limited case of Thus, creation of legal and regulatory frame- tradeability wherein the Fund "buys out" the works establishing firm-level permits or "en- entitlements and "retires" them. A full- titlements" for ODS use with legal limits on fledged tradeable permits regime would also market entry (by new firms) or re-entry or reveal the competitive valuation of ODS use expansion (by existing firms), is critical to an entitlement, a very useful piece of information effective use phaseout, whether or not there are for the government as well as the Fund. incremental costs associated with it and whether or not these costs are financed by the The rest of the paper is organized in Fund and how. Such framework would at the five sections: Section 2 provides a summary of very least ensure that the Fund-assisted phase- the developing country Parties' compliance out is effective, and would also enable the use of obligations and the Fund's mandate, conclud- other MBIs such as pricing policies or inter-firm ing with a brief discussion of the need for trade of permits or " entitlements". At the national firm-level controls; Section 3 discusses minimum, a registry of user firms with a pre- the Fund's current grant rules and their short- determined quantity of ODS phaseout it may comings in the context of compliance costs; offer is required. (As the controls on ODS Section 4 outlines the concepts behind a production and net imports become binding, competitive bid auction for ODS phaseout; and developing countries will also need to use Section 5 discusses the potential for other MBIs. similar controls on the supply side). Section 6 concludes with a summary and recommendations. A detailed discussion of pricing policies or a regime of inter-firm trade of entitlements ("tradeable permits") is consid- 1/ References to "ODS" in this paper are to the five CFCs and three halons in the Protocol Annex A, though the main argument is unaffected by including any other controlled ODS. All ODS quantities in this report are in terms of Ozone Depleting Potential (ODP) equivalent. ODP of a substance is a weighted value of its potential harm to stratospheric ozone, relative to CFC-11. (That is, ODP of CFC-11 is 1.0). 2/ Parties operating under Article 5(1) of the Protocol are referred to as "developing countries", and "non- Article 5(1) parties" as "developed countries". ?/ That is, if some firms convert but other firms expand the use or build up inventories of ODS, especially if ODS price falls as a result of the former group's conversion. Environmental Economics Series 3 Market Based Instruments for the Implementation of Montreal Protocol in the Developing Countries 4 Environment Department Papers National Obligations vs. 2 Firm-Level Controls Compliance Obligations tion" well in advance of the schedule imposed by the Protocol. Implementation of the control The Protocol and its Amendments measures via reduction targets upon individu- impose controls on aggregate national produc- als or firms is subject to national jurisdiction; in tion as well as consumption (defined as pro- most of the developing countries, no firm-level duction minus destruction plus net imports) of reduction targets or schedules apply as yet. a large number of chemicals subject to different schedules. From the developing countries' The Fund's Mandate and Operations perspective, the most significant controlled substances at this stage are the five CFCs and The Fund has been empowered by the three halons covered by Annex A to the Protocol Parties to use its assistance 'as an incentive for and referred to as ODS in this paper. Table 1 early adoption of ozone-protecting technolo- shows the control schedules binding upon the gies". The primary vehicle the MFEC has opted developing countries for these substances. to use is in the form of grants to individual firms engaged in the use of ODS undertaking Once the Protocol's control measures phaseout via conversion to permissible ODS become binding, imports by developing coun- substitutes. These grants are approved by the tries are permitted only for "basic domestic MFEC on a project-by-project basis according to needs". Net imports of used (reclaimed) ODS administratively established assessment criteria are to be taken into account for determining for certain categories of "eligible incremental the base year quantities, but not afterwards. costs" for each firm and conversion project. Many developing countries have expressed a Implementing agencies submit grant proposals willingness to reduce their ODS "consump- to the MFEC Secretariat on behalf of the firm TABLE 1: DEVELOPING COUNTRY OBLIGATIONS FOR ODS CONTROLS Production and net imports of virgin controlled substancesl/: Annex A: Group I Primary CFCs (CFC 11-12,113-5) Baseline: average of 1995-97 Freeze 7/1/1999; 50% cut by 1/1/2005; 85% cut by 1/1/2007; 100% cut by 2010. Annex A, Group II: Halons (1211,1301,2402) Baseline: average of 1995-97 Freeze 1/1/2002; 50% cut by 1/1/2005; 100% cut by 1/1/2010. 1/ Imports permitted only for basic domestic need; essential uses exempted. "Destruction" credits applicable for additional production or net imports.Sources: UNEP (1996). Also see UNEP (1993), Ahmed (1995), Parson and Greene (1995), UNEP (1995c), and Bojkov (1995), including for background to the evolution of the Protocol. It is unclear if more stringent control measures bind the few developing country signatories to the Copenhagen Amendments. Environmental Economics Series 5 Market Based Instruments for the Implementation of Montreal Protocol in the Developing Countries and the developing country government; the provided by the Protocol. Secretariat reviews the proposals for consis- tency with procedures and guidelines, and Recipients of assistance: While the recommends them to MFEC for approval. Protocol contemplates assistance to "Parties", Considerable administrative resources at the the Fund's resources are provided directly to multilateral level are expended to individually individual firms, who in most cases as yet face identify and develop each project, have it no control measures and, in the absence of reviewed by technical experts, select a national regulations, may have none until the "baseline", calculate "eligible" need, prepare ultimate phaseout date, so long as their own grant proposal, negotiate MFEC Secretariat ODS requirements are met (This is because recommendation, and oversee the disburse- even as interim controls on national ment, implementation and verification mecha- "consumption"are applied, some firms would nisms. be able to continue unrestricted use so long as their own requirements are met ven with lower The principal guiding document aggregate supply). characterizing the Parties' consensus on the types of expenditures that may be covered is the Form of assistance: The Parties have "Indicative List of Categories of Incremental accepted "eligible incremental costs" as an Costs" (the "List")' The cost categories operational guide, but have leftitto the MFEC broadly covered by the List include: costs of to determine precise rules of eligibility and have new capital equipment and of premature also left "agreed incremental costs" undefined. modification or replacement of existing capital The MFEC, therefore, is permitted to choose any equipment, and incremental operating (i.e., vehicle for grant delivery so long as an agree- non-capital or recurring) costs including those ment is reached between the Fund and the of royalties, retraining, and R&D. For each recipient; it is not necessary that the current end-use sector (e.g. solvents, aerosols, different approach of administrative assessment of types of refrigeration and air-conditioning or eligible incremental costs, and associated foam manufacturing), the MFEC has precluded arbitrary rules, be adhered to. certain technical options deemed to be inap- propriate in expert judgment and has set A departure from the strict terms of the operational criteria for the calculation of grant Protocol is understandable: the decisions to eligibility. convert to ODS-substitute production or use are made and implemented by firms, not govern- Four features of the Fund's operational ments; also, without user firms' conversion to strategy are worth noting: ODS substitutes, the market for substitutes would not exist, and producer firms would not Timing of assistance: While a strict be willing to convert. Similarly, when there was interpretation of the Fund's mandate would little experience with substitute technologies imply that the "agreed incremental costs" to be and no legal limits on ODS use, user firms had met are those of compliance with the Protocol's to be solicited, provided extensive technical control measures, the Parties have agreed that assistance, and project implementation closely the Fund's resources may be utilized for ODS monitored (for information exchange and phaseout before such control measures become learnig, in addition to verification of conver- binding upon the developing countries. In sion). The "early preventive action" on the other words, the Fund's grants may meet Fund's part has also served to limit the expan- incremental costs of accelerated phaseout, not sion of ODS consumption and associated just those of compliance. equipment stock in the grace period allowed to the developing countries. Scope of assistance: While the Protocol's control measures refer solely to However, the Fund's operations expose production and net imports of virgin ODS, the its program to three particular risks: Fund's resources are permitted to be utilized for reducing ODS use even as ODS production and (a) In the absence of national-level net imports may continue to rise or stabilize as 6 Environmnent Department Papers National Obligations vs. Firm-level Controls legally enforceable limits on market entry by words, there is a considerable risk of under- new firms and expansion of ODS use by shooting or overshooting the target. Other existing firms, the net ODS phaseout may be less MBIs like bid auctions or tradeable permits than what the Fund's assistance provides for.2 require government action to create firm-level controls. Such controls are in any event (b) In the absence of correspond- necessary for the implementation of a ing limits on the production and net imports of country's obligations under the Protocol, ODS, an exclusive focus on reducing the ODS irrespective of who finances the incremental use may lead to an imbalance in the market for costs of compliance or how. Before discussing ODS and their substitutes3. While the full the scope for MBIs or advancing a particular effects of ODS phaseout in the developed proposal for grant allocation, it is therefore countries on the prices of ODS and substitutes necessary to describe the need for national in the developing countries are yet to be seen, a firm-level controls. situation of excess capacity for ODS supplies and/or capacity constraints for the supply of Need for Firm-level Controls substitutes may lead to perverse price signals to ODS-using firms. Even as ODS use is phased To effectuate compliance with the out, if production and imports continue up to ProtocoL national governments need exercise the limit permitted by the Protocol (to add to their legal authority upon firms' doing busi- ODS stocks for later use), the environmental ness in their boundaries to restrict, and as gains of "accelerated phaseout" may be ne- necessary prohibit, them from taking actions gated. which violate the country's legal commitment. In other words, the national aggregate obliga- (c) Since administrative proce- tions need to be translated into firm-level dures for project identification and develop- obligations with the corresponding legislative ment, and for associated need assessment, and/or regulatory authority for credible preclude consideration of the benefits of enforcement. Moreover, such controls provide conversion that accrue to firms (because these additional motivation (i.e., in addition to other cannot be externally determined or verified), forces such as market changes) for firms to or a determination of the optimal timing of shift from ODS production or use, and may conversion, there is a risk that the Fund's generate a more predictable environment in assistance is not as efficient as it could be - which they may undertake such actions While i.e., by maximizing the amount of phaseout for recognizing the conceptual or legal and a given sum of money. administrative complexity of the design, monitoring, verification, and reporting activi- The third risk is the main theme of this ties that may be involved, we discuss below, in paper, with an aim to propose a particular rather simple terms, the elements of firm-level market-based instrument (MBI) - competitive control strategies. bid auctions - as an altemative to the current administrative approach to grant allocation. Compliance with the Protocol requires, But before discussing MBIs, it is necessary to first, monitoring the flows of ODS into and out recognize that government action creating a of the economy via domestic production and commodity or a property right and/or setting intemational trade, and second, imposing firm- the rules of market transactions are a pre- level quantity controls on these activities. This requisite to the effective use of MBIs. can be accomplished by developing registries and record-keeping systems, on the one hand, Certain MBIs "e.g., pricing policies" and controlling the additions and deletions to may require no additional government action the registry, in other words "entry" and "exit". because the commodities as such are already The latter may be voluntary or as a result of defined; only tax levels and collection mecha- some other scheme such as for compensation nisms need to be specified. However, pricing (whether or not financed by the Fund) or a instruments alone are insufficient to achieve mandate by the government. Additionally, if quantity obligations, especially time-bound the country wishes to ensure that any conver- obligations such as set by the Protocol; in other sion by use sector firms - whether or not Environmental Economics Series 7 Market Based Instruments for the Implementation of Montreal Protocol in the Developing Countries financed by the Fund - is "protected" (in that provided over a long period (e.g., through 2010, other firms do not expand their use), registry, subject to interim "decay" in accordance with record-keeping and quantity controls on use the Protocol) and would apply only to the are essential. specific ODS as engaged in historically. On the other hand, since it is only the "net imports" The key to translating aggregate limits quantity that is to be controlled, import permits into firm-level controls is of course allocation of in the amount of baseline "net imports" is better permits (which may also be called licenses or auctioned off, and may be granted for shorter allowances or entitlements); as the aggregate periods (e.g., a year or two). Import permits may limits change, according to the Protocol sched- be substance-specific; if not, they must at least ule or a more accelerated schedule if the country apply to the grouping of ODS corresponding to wishes to impose one, the permits have to be the Protocol (i.e., any of the principal CFCs or either re-allocated or a certain "decay rate" any of the halons). Both production and import (reduction in the physical value of the permit) permits would become effective 1999. has to be built in. To the extent that historical patterns of activity are to be any guide to the All import permits and at least a initial allocation or the re-allocation processes, portion of the production permits should be the registries in the previous paragraph can permitted to be exchanged, during the period of serve as the basis. validity, among firms (or within a firm, if it engages in more than one activity). This is The first step is a statutory or regula- because no permits are as such made available tory requirement that only the firms holding for exports; a ton of exports permits a produc- permits for a particular activity may engage in tion or import of an extra ton, and an an extra that activity, and only up to the level permit- ton may be imported only against a ton of ted; all other activity is legally prohibited. In production reduction or export. Because the next step, several alternatives are available production permits are longer-term, their for initial allocation of permits to firms: exchanges for import or export must be of "grandfathering", i.e., on the basis of historical limited duration (a year or two), after which patterns; competitive bidding (or auction); they revert to the production account. Addi- "first-come, first-served" basis (if it appears tional "tradeability" --of production permits that the demand for permits is likely to be less among the producing firms, between produc- than the available aggregate); or, voluntary tion and import permits (to firms already agreements with industry associations. The authorized to do either or to new firms, if at all permit system can be made functional once the interested) -- and "offsets" - against reclaimed aggregate limit and firm-level historical record ODS or destruction of ODS - may be permitted, for the baseline are known, or in advance, with but is not necessary, and may give rise to certain lee-way retained by the government to enforcement problems and high transaction make changes at a later time according to pre- costs as discussed in Section 5. Unless inter- set rules-. A simple scheme on both the firm exchange of production permits is allowed, "consumption" and the "use" sides, combining a firm may permanently "return" its production grandfathering and auction approaches, is permit only to the government, for re-allocation described below6; some of its implications, and to another firm, or "retire" it by offering it to the additional policy instruments that may be Fund in return for Fund grants for conversion. based upon it, are discussed later in Section 5. These concepts are further discussed in Section 5; for now, it may be obvious that the situation "Consumption" Permits: On the is much simpler for developing countries who "consumption" side, different firms could be have no domestic ODS production as of the engaged in just one or more of the three activi- baseline period and prohibit new production. ties (production, imports, exports). Political expediency as well as recognition of the indivis- Beyond the initial allocation, the ible nature of investments in the production grandfathered production permits would be activities would appear to dictate that initial subject to a "decay rate" corresponding to the allocation of production permits be based on control schedules of the Protocol. For example, the historical record, i.e., grandfathered; be a permit to produce one ton of a particular CFC 8 Envirornent Department Papers National Obligations vs. Firm-level Controls between 1999 and 2004 becomes a permit to stockpiling that is permissible under the produce a half a ton of that CFC between 2005 Protocol (and has indeed occurred in the and 2006, and so on until 2010, when all developed countries) is unavoidable, unless production is banned (except for the amounts the country periodically re-determines its permitted due to verified destruction). Permits aggregate supply quotas in accordance with for net imports offer considerable leeway, since the reduction in use, or undertakes other their allocation can be restricted to shorter measures (such as pricing policies to discour- periods. That is, permits for net imports may age stockpiling, or prohibitions on stockpiling). be auctioned every year or every two years, This matter is further discussed in Section 5. and the aggregate available quota may be re- determined as frequently. For a variety of reasons, some govern- ments may be wiling to undertake measures Use permits: Controls on user firms are that are not mandated by the Protocol7. Some not necessary for strict compliance with the alternatives for a more aggressive approach are: Protocol. Rather, they are necessitated by the Fund's choice of providing grants for use sector (i) Accelerating the schedule conversion. These controls, in turn, satisfy two and/or reduction requirements: for example, needs: (a) fix the amount of ODS that a conver- permits may be made effective earlier than 1999; sion project is expected to phaseout, and (b) 50% reduction in an individual firm's produc- ensure that any phaseout, whether or not tion permit or in the country's aggregate import financed by the Fund, is not negated by subse- limit to be auctioned may be advanced to 2002, quent growth in use. A registry of user firms say; or reductions may be phased in more combined with prohibition on market entry or gradualy by placing interim limits. growth would essentially parallel the scheme (ii) Partial permitting: for example, described above, with some exceptions. First, a government may "withhold" a fraction of the the initial aggregate level of "use" may not aggregate import limit, and "release" it at a later match aggregate "consumption" due to stock date to moderate ODS price increases. "With- changes at the supplier or user level, and due to holding" production or use permits may require lack of data on certain classes of users (e.g., an advance ban on some firms, unless they especially those engaged in the service sector). were to cease operations and exit the market Therefore, the use permits may not be able to any way. cover aDl uses and aDl users precisely. Second, (iii) Substance or sectoral bans: all use permits would probably have to be government may impose differential reduction granted initially at a historical base year level or "decay" rates in the production, imports or (i.e., via "grandfathering" a particular historical use permits, if a specific substance or class of level). Third, use permits need not have any firms is seen to have especially low cost of reduction or "decay" from the initial level. conversion or to benefit from it. Fourth, because of monitoring and enforce- (iv) "Downstream" regulations: If ment difficulties, if inter-firm trade in user servicing requirements constitute a high (e.g., > permits is allowed, its amount and frequency 20%) portion of the ODS use, a government may should probably be restricted; i.e., a firm can wish to impose additional regulations on the only offer its entire permit to another user service sector, including for recycling, reclama- firm, and may not buy it back or buy from tion, and operator certification. another firm. As discussed in Section 5, a firm-level From the perspective of the Fund, any control scheme is also necessary for the use of phaseout financed on either the production or other market-based instruments. But first we the use side entails permanent removal of the turn to the discussion of the Fund's current beneficiary firm from the registry. However, process for grant allocation to finance use sector even such a permit system may not adequately conversion. protect against risk (b) described above. ODS Environmental Economics Series 9 Market Based Instruments for the Implementation of Montreal Protocol in the Developing Countries 1/ UNEP (1995b), Annex IV.1. 2/ Despite the MFECs decision in July, 1995 to exclude ODS use capacity installed after that date from its assistance, the potential exists that a firm may profitably begin or expand ODS use and continue doing so as long as virgin or reclaimed ODS are available. 3/ The seventh MOP in December, 1995 made a decision that "from December 7,1995, no Party should install or cormission any new capacity for the production of CFCs, halons, CTC and MCF" (Decision VII/9(7)). Recogniz- ing the problem of oversupply and dumping, the seventh MOP also decided that all Parties engaged in international trade of ODS "should monitor and regulate this trade by means of import and export licenses" (Decision VII/9(3)). It appears that neither of the decisions is binding. 4/ The term "firms" is meant to include all economic agents. S For strict compliance with the Protocol, there is a problem in that neither the aggregate limit nor the firm- level patterns would be known until some time in 1998. On the other hand, the apparatus for registration and record-keeping would have to be prepared before then; it appears that most developing countries have already accomplished this task as a part of their "country programmes". §/ Also see Munasinghe and King (1991), UNEP (1995a) and SEI (1995). _Z/ Countries which are willing and able to carry out and implement more stringent control measures (as encouraged under Article 2, ¶ 11 of the Protocol) could be given preferential treatment by the Fund. 10 Environment Department Papers 3 Scope for Market Based Instruments This Section briefly reviews the The MFEC approves an estimate of Fund's current grant allocation rules and these two components when a project is process, then evaluates them in the context of presented to it; the disbursement takes place different types of incremental costs. It then upon the actual, demonstrated value of the advances the argument that the firm-level gross capital expenditures and the incremen- incremental costs depend on economic tal operating cost for the pre-defined dura- parameters which (i) are best known to the tion2. While any "budgetary savings" due to firm itself, and (ii) vary across firms. Charac- lower actual costs than estimated are re- terizing the Fund's program as a market in turned to the Fund, and any "cost overruns" which the Fund buys ODS phaseout offered due to higher actual costs have to be ap- by different firms provides the conceptual proved by the MFEC, essentially the Fund context for discussing market-based instru- absorbs the entire risk of errors in cost estima- ments discussed in the following two sections. tion. (Such an approach is worthwhile when project implementation is in doubt and costs Current approach for grant allocation are expected to decline over the implementa- tion period). Finally, the Implementing Under current rules (applying to user Agencies are reimbursed for their administra- firms' conversion), the "eligible incremental tive costs, and sometimes financial intermedi- costs" are assessed as the sum of: aries in the developing countries are provided a fee. (i) Eligible Capital Costs These are defined to be all the gross capital expendi- Projects approved to date have shown a tures of the new equipment that needs to be wide variation (across countries and end-use installed (subject to guidelines and technical sectors, project size, and over time) in unit grant experts' approval), in general without regard levels ($/kg ODP) assessed, as shown in Annex to the remaining economic lifetime of the Table A.2 for projects implemented by the existing equipment to be replaced or to the World Bank. This suggests that there are no firm's discount rate. "model" projects as such (to establish a fixed average grant, for example), and that some firms (ii) Eligible Operating Costs indeed find cheaper conversion options than These are defined as the excess of the operat- some others. If so, it is in principle possible to ing costs associated with the use of the ODS choose cheaper projects before more expensive substitute over those with the use of ODS, for ones if firms are provided the opportunity to a limited period of timel. Except when self-select, instead of the current "first come, demonstrated and quantified, the benefits first served" approach. that may accrue to the firm as a result of conversion are ignored, as are changes in The MFEC has defined "cost-effective- operating costs (e.g., due to changes in the ness" as the unit grant level approved: total price differentials between the ODS and its approved grants divided by the annual quan- substitutes) beyond the initial qualifying tity of ODS to be phased out. Until 1995, the period. Environmental Economics Series 1l Market Based Instruments for the Implementation of Montreal Protocol in the Developing Countries Fund's available grant resources exceeded the or reveal their willingness to accept a lower demand for grants from candidate projects in unit grant which would give them a priority the implementing agencies' pipeline, and over (by being less than the unit grant re- grants were approved without specific quests of) other firms.3 Combined with the consideration for "cost-effectiveness". administrative identification and selection of Beginning in 1995, when the potential de- participant firms, and preparation of project mand for grants as expressed in the project scope, there is at best limited room for self- submissions to MFEC exceeded the Fund's selection by firms, or for firms' own assess- available budgetary resources, the MFEC set ment and selection of project scope and specific "cost-effective threshold values" timing. There is a need for a mechanism (CETVs) as cut-off points to be used to select providing for self-selection and competition from among project submissions for certain among firms. Before proposing one such sectors. (See Annex Table A.1, along with a mechanism, however, it is useful to address comparison to MFEC's prior estimates of certain conceptual issues relating to incre- average grants for the purposes of budgetary mental costs and the implications of the programming). In most cases, the 1995 Fund's current rules. CETVs were selected in such a way as to match a uniform portion of the projects in the Costs of compliance vs. accelerated phaseout Implementing Agencies' pipeline against the funds deemed to be available to the MFEC for As noted earlier, the Fund's mandate is approval during 1995; conceivably, the to meet the "agreed incremental costs" of CETVs could have been different if the Fund's "compliance" with the Protocol, but it has also financial situation or the composition of the been empowered to provide its grants for an projects in the pipeline were different. accelerated phaseout via broad guidelines for "eligible incremental costs". The Protocol One result of the MFEC's announce- imposes no controls on ODS use, and an ment of CETVs in March 1995 was that a individual user firm may in principle continue number of project submissions were revised so ODS use so long as its existing capital equip- as to lower the grant requests to match the unit ment permits it to4. The incremental costs of grant levels to CETVs. It appears that in some over-compliance can be higher than those of cases, the beneficiary enterprises searched for strict compliance, but different firms would still lower-cost alternatives while in some other have different incremental costs. The project- cases they were willing to accept a lower level by-project approach does not allow for covering of grant (corresponding to CETVs) than that lower-cost phaseout before higher-cost ones at a corresponding to the "eligible costs" criterion given point in time, and the administrative (meeting the rest of the costs by themselves). At grant assessment rules may risk meeting more its July, 1995 meeting, the MFEC decided to than the incremental costs and providing permit such "partial funding" of assessed incentives to individual firms to choose higher- eligible expenditures. In response, some firms cost options and timing. This is because, even claimed they were willing to accept only a if an accelerated phaseout gives rise to a portion of their eligible capital costs, while higher cost, that cost should still be mini- some claimed they were willing to accept only a mized-. portion of their eligible operating costs, confirm- ing that the total amount of grant is the deter- Following King and Munasinghe mining factor, not its administrative character- (1991), we can define minimum cost of compli- ization. ance ("MCC") as the difference in the economic costs incurred by the country in two scenarios: While the use of CETVs and "partial (i) without the Protocol, and (ii) a compliance funding" have implicitly introduced some scenario with the Protocol in which the country elements of competition among firmswithin has devised and implemented the least-cost certain sectors, the projects are approved path to meeting the requirements imposed6. To essentially on "first-come, first-served" basis, the extent that a country or a firm does not and there is no particular incentive for firms to adopt the least-cost path, its actual cost of limit their unit grant requests below the CETVs, compliance ("ACC") may be greater than its 12 Environment Department Papers Scope for Market Based Instruments MCC7. While recognizing that country direction of the MCC path. In other words, incremental costs also include government even if there is a wedge between financial and costs of compliance and changes in public economic costs, minimizing the former would welfare, we focus on the firm-level incremen- also minimize the latter, absent serious tal costs, since that is the primary target for market distortions. the Fund. There are practical problems with Several features of these concepts need estimating the country-level MCC, and also to be emphasized: with applying the concept of country incre- mental costs, embodied in the Protocol, to the (i) Both the MCC and ACC level of individual firms9. At a minimum, concepts are based on economic costs, i.e., cost of estimating the country-level MCC requires the resources used in the process, and not on knowledge of firm-level costs and benefits at financial costs, which may be affected by different points in time and, to the extent that market distortions. One implication is that a one firm's actions influence another firm's country's MCC or ACC is not reduced by any decisions, the knowledge of all firms' likely grants provided by the Fund, since such grants behavior. arefinancial flows that do not affect the cost of the underlying resources. On the other hand, even as the Fund's assistance has been made available in advance (ii) Both are net of the benefits of of the Protocol's compliance schedules, admin- compliance actions. That is, if conversion to the istrative assessments of "eligible costs" have an use of substitutes lowers the cost and/or inherent limitation in that only gross expendi- improves the quality of the final product or tures can be observed externally. To derive the service, those benefits are, at least in principle, net costs from the gross expenditures, and the to be subtracted from the cost of compliance. incremental costs from the total costs, requires This may be impossible in practice, in part information that is best known to the firm, may because the benefits may not be measured in not be in its interest to disclose, and may not be advance, and in part because they may well subject to verification or control. Even from the pass on to consumers. firm's viewpoint, some of this information is quantifiable - for example, its discount rate (iii) Both refer to the costs which and its expectations of costs of all its inputs and are incremental to a baseline scenario of no firm- prices for its products - whereas some is not - level restrictions on continued ODS use. For for example, product development and market- example, to the extent that conversion to ing strategy'". However, any firm uses both substitute use involves premature replacement types of information in making any decision to of existing capital equipment, the incremental invest, including that to convert to ODS substi- cost is not the full amount of gross capital tute. As argued next, the Fund's current rules of expenditures, but only that part which is due to grant calculation provide no incentives to firms premature investment (including the difference to minimize their own ACC, which is best in the cost of the two types of equipment)8. known to the firm themselves; may indeed Similarly, if the post-conversion operating costs distort the firms' behavior from least-cost choice are higher than those under current practice, of technology and timing; and may provide the difference over the entire duration of the more than the net incremental costs of conver- new project is incremental. These points are sion. further elaborated upon below. Limitations of current Fund rules (iv) Individual firms operating in a competitive market seek to minimize their net Ideally, the Fund would structure its financial costs of conversion. In the absence of grants so that: (a) an individual firm deems the serious economic distortions, which would Fund's grants to be adequate, based on its own bring about significant differences between assessment of overall costs and benefits of economic and financial costs, this would also conversion; (b) the individual firms are moved tend to move the individual firms in the in the direction of the MCC least-cost path; and, Envirornental Economics Series 13 Market Based Instruments for the Implementation of Montreal Protocol in the Developing Countries (c) at any given time, the Fund maximizes the the Fund, the user firm may be seen to bear the ODS phaseout with the available resources. entire burden of incremental operating costs for The Fund's current approach (including use of the remainder period. Obviously a firm would CETVs) certainly meets the first objective, but rationally convert only if (i) it sees this burden falls short of meeting the other two. as minimal or even negative, i.e., if it expected the incremental operating cost to vanish and/or First, calculation rules for "eligible the conversion to provide it with collateral incremental costs" are only weakly related to benefits (such as higher overall productivity, the firm's net incremental costs of conversion, improved productquality, higher market share); and provide incentives to increase the ACC as or, (ii) it believes its under-compensation for the well as permit higher grant levels than ACC. incremental operating cost to be matched by an This can be shown as follows. Conceptually, over-compensation for the capital costs. the firm's incremental costs of conversion are the sum of: (a) the costs of premature replace- Current rules do not permit a determi- ment of its equipment, and (b) the net present nation of what the firm's "true" incremental value of the incremental operating costs, net of capital costs or incremental operating costs (net benefits, over the economic life of new invest- of collateral benefits) are, and, more impor- ment. Considering that tantly, what choice of timing and technology would minimize these costs. This is because (i) the investment for conversion the necessary variables - which the firm itself typically involves substitution of existing capital would use in making such determinations - stock rather than addition to it; and, are ignored and in any event may not be accurately available for an external assessor. (ii) the remaining economic life of its For example, the cost of premature replacement existing equipment is finite, and some firms is a function of the firm's discount rate and the may well have to invest in replacement capacity remaiing economic life of existing equipment, before they have a legal obligation for compli- the latter in turn being an elastic measure ance with the Protocol11, indeed whether or not influenced by changes in market conditions, the Protocol existed; and, further assuming that, technologies, and government policies. Simi- larly, the benefits of conversion that may reduce (iii) at the time of such replacement, a or outweigh the firm's incremental operating firm will choose the non-ODS using equip- costs are in turn dependent on the firm's Ment2, expectations, and procurement and marketing strategies. Even if some of these variables could it follows that the policy of funding gross be obtained - for example, the discount rate capital expenditures creates differential "rents" and price forecasts - it is impossible to disen- - the difference between opportunity cost to tangle the firm's conversion project from its the firm and the grant amount it is awarded - overall operations and objectives'3. to the firms depending on the distribution of remaining economic life of existing capital Second, the project-by-project approach stock. Under current rules, the Fund gener- does not permit the Fund to discriminate in ally covers all the gross capital expenditures, favor of firms with lower phaseout costs. Not and not just the costs of premature replace- only are incremental costs of conversion ment which constitute the incremental costs of dependent on firm-specific variables, but there conversion. This willingness to over-compen- need be no uniformity in such variables across sate provides incentive to the firm to both firms, even within a sector or size grouping. accelerate its replacement decision and to Given the lengthy process of project identifica- prefer capital-intensive options. tion to approval under the current approach, it is practically impossible to permit firms to self- On the other hand, the Fund restricts select or for the Fund to reject project proposals. duration of financing incremental operating The use of CETVs and "partial funding" partly costs. Inasmuch as the economic life of new addresses this problem. However, even the use investment in ODS-substitute technologies is of CETVs does not provide any explicit incen- likely to be longer than the period covered by tives for the firms to minimize the grant re- 14 Environment Department Papers Scope for Market Based Instruments quested from the Fund, nor does it permit the In the early years of the Fund's Fund to explicitly recognize that the firm may operations, ODS substitute technologies were well benefit from conversion and that such entirely new and risky, sometimes even in the benefits are not a part of "incremental costs"4'. developed countries. This may still be true At best, this is a "satisficing" approach in the with some technologies and some developing sense that projects that meet the threshold value countries, but clearly not so for some others, are said to meet, pnimafacie, the cost-effective- judging at least from the Implementing Agen- ness criterion. In any case, there is no assur- cies' project pipeline. An alternative mecha- ance that the project cost data upon which nism to the Fund's current "micro-managing" CETVs are based reflect incremental cost approach then becomes necessary; this alterna- minimization calculus at the firm-level. tive should (i) permit firms to self-select, i.e., let the firms determine their incremental costs of Finally, the current approach requires compliance they seek the Fund's financing for, an in-depth external assessment of "eligible and (ii) introduce competition among firms in incremental costs" but does not require that the grant allocation process. A competitive each firm's phaseout decision is optimized with bidding mechanism would permit these respect to time, or that grant decisions are objectives and would naturally guide firms to optimized across users within an end-use choose phaseout strategies consistent with the sector and across end-use sectors15. As the least-cost path using their own values for firm-specific incremental costs are best known economic parameters and market expectations, to the users themselves, efforts to collect and because by doing a lower-cost route, a firm analyze information by the implementing improves its own competitive position. This agency involves a duplication of effort and would ensure that even with an accelerated additional administrative costs. phaseout at the country level, individual firms' - and therefore, the country's - actual cost of Despite these shortcomings, adminis- compliance is as close to the minimum as trative need assessment and the associated possible, and that the Fund would be able to project-by-project approach, is not only appro- obtain cheaper phaseout before more expensive priate but sometimes the only practicable one in one. Additional efficiency gains accrue to the certain situations. These are where firms are Fund because such a mechanism would also being asked to adopt technologies which are in obviate the duplication of effort inherent in the experimental or demonstration stage with little project-by-project approach with multiple levels experience of performance, their costs and of decision-making, avoid the need for external benefits to firms are difficult to predict reliably. hypotheses about firm-level behavior17, and Even if firms seem to demand more than their reduce the associated administrative costs. As incremental costs of adopting new technologies, will become clear in the subsequent discussion it may be argued that the apparent over- of designing and implementing various MBIs, compensation is actually a proper compensa- the project-by-project approach provides critical tion for the risk premium the firms place on initial information on which such MBIs can be adopting new technologies. Over time, some based. technologies are rejected and others mature, i.e. more operating experience with them is gained, Characterizing the Market for ODS Phaseout their costs and benefits become reliable. More importantly, market expansion and learning In a situation where firms self-select effects lead to cost reductions for new technolo- and compete as described above, the firms offer gies. Correspondingly, firms' "search costs" to "produce" a verifiable service called "ODS and "transaction costs" are reduced, and they phaseout", and the Fund offers to "buy" this increasingly find it in their own interest to service. The flow of funds is from the Fund (via adopt new technologies and develop further the government) to the firm, and the amount of changes"6. There may still be incremental costs the Fund's grants must be acceptable to the associated with the use of new technologies as firms that convert. Two alternatives in these compared to the older ones, but firms then self- transactions are: select according to their own cost calculus. Environmental Economics Series 15 Market Based Instruments for the Implementation of Montreal Protocol in the Developing Countries (i) the Fund offers to buy "ODS phase- recognizes the impracticality of the regulator out" at a certain price ($/kg ODP) and the being able to (i) allocate a multi-product firm has the option of accepting or rejecting firm's costs among product lines, (ii) disen- the offer. If the grant is lower than what the tangle operating expenditures from invest- firm perceives to be its incremental cost (net ments, and (iii) disentangle the cost associ- of collateral benefits), it would be irrational ated with producing the product whose price for it to convert is being regulated from that associated with producing a non-verifiable benefit."8 (ii) the firm offers to sell "ODS phase- out" at a certain price and the Fund has the It is the overall amount of grant (in $/kg option of accepting or rejecting the offer. If ODP) that expresses the firm's "offer price" or another firm offers a lower price, it would be "willingness to accept" (WTA). Different inefficient for the Fund to accept the first firm's firms would compete on; whether or how its offer. components are characterized is immaterial and unnecessary. This was implicitly recog- The Fund is the sole buyer from many nized in the MFEC decision in 1995 to permit offeror firms of ODS phaseout. In a competitive "partial funding" of projects whose assessed auction, described in detail in the next section, eligible grants came in above the CETVs. If the offerors decide when and at what price they the assessed eligible grant were itself in excess wish to offer their phaseout; the Fund accepts of the firm's unrevealed "true" incremental firms' offers up until the point that the available costs, the so-called "partial funding" by the funds are exhausted. Fund may well cover all of the true costs (or more). The firms' "supply price for ODS phaseout" - i.e., the minimum unit grant (in $/ In the conventional language of kg ODP) that each present user deems adequate economics textbooks, each firm's WTA is a - cannot be precisely assessed administra- point (because the decision to convert is of a tively. This is because the firms "produce" not "yes/no" type), and all such points together merely ODS phaseout but other products in reflect the market supply curve for ODS phase- conjunction (such as refrigerators, foams, out at a given time. Granting a fixed price aerosols), so that the activity of "producing" provides an excess to all the suppliers of phaseout also influences their costs and phaseout with WTA below the fixed price; a benefits overall. In any event, the Fund has no competitive bid auction - similar to bulk way of knowing or separating or controlling for procurement - is the mechanism to encourage all the costs and benefits. In generaL where a the suppliers to bid close to their WTA, and is regulator seeks to administratively determine explored in the next section. the price of a product, economic theory 1/ Ranging from six months for halon projects to four years for aerosol and solvent projects; see footnote below for domestic refrigeration sector. Source: UNEP/OzL.Pro/ExCom/19/64. V/ With the exception of domestic refrigeration sector projects where firms are given the choice among: 10% of the eligible capital costs up-front; 6 months of eligible incremental operating costs at ex ante prices, up front; or, 12 months of eligible incremental operating costs at actual prevailing prices). l/ There is also the risk that the CETVs set for 1995 may be perceived as "model" values to apply over a longer period, essentially as the Fund's "offer price" against which the firms should prepare their phaseout plans. 4/ That is, if other firms' phaseout actions sufficed to meet the interim aggregate reduction target. If suffi- cient quantities of reclaimed quantities were available, an individual firm could go on using the ODS indefinitely. It may be argued that the government need resort to regulating the end-user firms only if it cannot meet its Protocol obligations by just regulating the firms engaged in production and trade. 16 Environment Department Papers Scope for Market Based Instruments 5/ As argued later in the paper, it is impossible to externally determine a firm's "true" incremental cost; accordingly, no a priori determination can be made about whether an accelerated phaseout necessarily increases all firm's incremental costs. f/ They use the term "compliance cost" for this concept (least-cost phaseout against the background of existing economic and industrial policies) and point out the difficulties in distinguishing it from what they call "minimum incremental cost" (least-cost phaseout in the absence of market distortions and failures). We have called it the minimum compliance cost in order to emphasize the least-cost aspect, and also to clearly differentiate it from the actual cost For further details, also see Munasinghe and King (1991), and King and Munasinghe (1995). Z/ King and Munasinghe (1991, 1995) use the term adjustment cost for this concept. E/ Because we are dealing with economic measures, merely the fact that a plant may be fully depreciated in the financial accounting sense does not mean that its remaining economic life is zero. 2/ See UNEP/OzL.Pro/ExCom/7/21 for an elaboration. 1O/ To put this differently, the model of incremental cost calculation that can, in principle, be applied to a closely regulated industry (e.g., public utilities) is not useful in industries without such close regulation of price and/or product quality. W If ODS-using firms - whether in final use, such as solvents, or in intermediate use, to produce consumer equipment charged with ODS - generally have capital stock with an economic life of less than 10 years, it is plausible to argue that the 10-year grace period allowed to the developing countries sought to limit their incre- mental cost Ignoring search, royalty, and training costs, the incremental capital cost for some user firms may well be zero, especially if no new investments in ODS-using technologies were made since the country ratified the Protocol. 12/ This assumption is made only for the convenience of abstracting away from the question of differential costs of user firms' capital equipment for alternative technologies with and without the Protocol. Similarly, we abstract away from the possibility that the existence of the Protocol itself may affect the remaining economic life of a firm's existing equipment - e.g., consumer awareness about the ozone depletion issue and the Protocol may have led to shifts in consumer preferences. 3/ ODS phaseout is very different from the traditional utility regulation where a regulator could obtain information on all the firm-specific cost variables and is in a position to regulate the firm's profit rate as well as make adjustments in allowed costs at a later time. 1/ The report under Article 5.8 of the Protocol (ICF et al., 1994) did, however, point to the potential to reduce Fund grant requirements if firm benefits could be captured. Model calculations of aggregate (i.e., for all firms) incremental costs are subject to the same limits as the project-by-project need assessment approach due to their inability to capture all significant parameters that affect firm-level choice and the variations in such parameters across firns. This is a familiar problem with technology adaptation modeling, and shows up, for example, in costing energy efficiency gains. l With the exception of the MFEC's acceptance of a "sectoral approach" for China. 1/ It is often argued that the actual costs of compliance with environmental mandates turn out to be less than the ex ante estimates of such costs. In the context of MP compliance, see Ahmed (1995) and Cook (19%). There may well be inherent upward biases in any ex ante external assessments of compliance costs where technologies for compliance change rapidly. AZ/ For example, about what the firm would have done in the absence of the Protocol, which can easily deteriorate to exercises in persuasion. i8/ See Laffont and Tirole (1993). Environmental Economics Series 17 Market Based Instruments for the Implementation of Montreal Protocol in the Developing Countries 18 Environment Department Papers 4 Bid Auctions for ODS Phaseout In this Section, we first introduce the to reduce their load according to a ranking concept of accepted bids in a competitive based on, say, $ per megawatt of load reduc- auction as "agreed incremental cost" corre- tion offered. By accepting the lowest bids, the sponding to the Protocol requirement, fol- utility is able to meet its load-management lowed by an illustrative example for the target in the least-cost mannere. The situa- domestic refrigeration sector and comments tion in ODS phaseout is similar, with the Fund regarding implementation of such an auction. corresponding to the power utility, and the ODS using firms corresponding to the power Equivalence of accepted bids and agreed utility's customers. incremental costs The basic idea behind bid auctions is Because "ODS phaseout" (measured simple: different firms have different willing- in ODP terms) is a service of uniform quality ness to accept (WTA, in terms of kg of ODS - one kg. of ODP phaseout from one source is phased out per $ of Fund grant) depending on indistinguishable from that of another - it is a variety of factors2, many of which are only ideally suited to competitive bid auctions or or best known to the firms themselves, and bulk procurement. Here the buyer's (i.e., the that a given firm would also have different Fund's or its agent's) budget is assumed to be WTA at different points in time. An adminis- fixed at the national level for a particular trative calculation of firm-level incremental period, say a year or two. Competing firms costs is necessarily arbitrary and fails to indicate their quantity of ODS phaseout as capture the difference between the firm's well as the unit grant level (in $/kg ODP) that willingness to accept and the administratively they are willng to accept from the Fund. Bids calculated grant. In a bid auction, a firm's are ranked according to their "offer price" "offer price", or "willingness to accept", and accepted until the pre-set available funds reveals its own estimate of the "incremental are exhausted. Auctions of varying complex- cost' incorporating its own expectations of ity can be devised, and may be repeated at the costs and benefits of conversion. An specific time intervals. accepted bid, in turn, constitutes a mutual agreement between the firm and the Fund The use of competitive bidding in that the total amount awarded is adequate grant allocations is similar to the traditional compensation for the service to be provided; system of Government procurement, with the the "agreed incremental cost' criterion is product being procured - ODS phaseout - therefore satisfied, without having to exter- classified as a verifiable service. Competitive nally assess the firm's costs and benefits3. procurement is also used by private firms for a wide variety of goods and services. The In preparing its bid, each firm would approach has a number of precedents in the consider its WTA in comparison with its energy sector. For example, power utilities estimate of the competing firms' WTA4. If it interested in measures to manage their loads indicates a higher offer price than what it is may invite bids from their customers to willing to accept, it risks losing out to a reduce their power loads; it pays its customers competitor. If it estimates that its offer price Environmental Economics Series 19 Market Based Instruments for the Implementation of Montreal Protocol in the Developing Countries would and take a chance that its current offer Illustration: Domestic Refrigeration is indeed cheaper as compared to its competi- The bid auction proposal relies tors. Thus, competitive bidding would come essentially on differential "wplingness to close to revealng the firms' minimum WTA and otimiz thetimin of cnverion arossaccept" across firms (and for a given firm, and optimize the .iming of conversion across over time). Before discussing the proposal, we firms as seen from their own perspective, first discuss the hypothetical revealed "will- Further, a bid auction approach ingness to accept" constructed from selected preserves all of the firm's incentives to domestic refrigeration manufacturing conver- minimize its incremental financial cost. sion projects recently approved by the MFEC. ,hen a firm's bid is accepted by the Fund, the These projects range from 22 to 165 ODP tons assistance foigfper year of ODS use, and while all used HFC- assistandrce gafls from the Fund to ter 134a as the substitute for CFC-12, the substi- is a direct grant- that is not tied to any tt o F-1(sfaigaet a particular aspect of the firm's operations. tute for CFC-11 (as foaming agent) was Consequently, it does not introduce any distortionary incentives such as overcompen- Table 2 shows calculations of incre- sation for capital costs or under-compensa- mental costs as a function of changes in four tion for incremental operating costs; there is variables - each of which are firm-specific no tendency for the Fund's support to in and best (or only) known to the firm itself - crease the firmi's and the country's ACC. for a hypothetical project. This hypothetical project was constructed from the average of AXpart from leading to economic the "representative" sample in Annex Table efficiency, the bid auction approach has several practical advantages. It could be A.3, a list of all domestic refrigeration invest- implemented within the existing administra- ment projects approved by the MFEC at its tive framework of conventional procurement, 13th, 15th, and 17th meetings7. The average and its administrative costs would be lower unit capital grant (gross capital expenditure) than under the current approach, since many is $14.3/kg ODP and average unit first-year of the expensive functions, such as project incremental operating cost (IOC) is $9.9/kg . . ~~ODP. Under the MFEC's current rules, the preparation and need assessment, would no longer be necessary. Most importantly, it is firm would obtain a grant of $19.2/kg ODP. fair in that a technically qualified firm's offer The Table compares this grant with price may be rejected only if other firms' incremental conversion cost computed from cheaper offers exhaust the available funds; the firm's viewpoint the sum of incremental the winning firms' incremental costs of capital cost (computed as the carrying cost for compliance are met precisely as they perceive premature replacement) and the NPV of them to be. incremental operating costs, under alterna- tive assumptions. For simplicity, it is as- Rejcte bidr,a cmeei sumed that the replacement cost of existing subsequent round of phaseout so long as the e regacmsntheosn of its Fund continues to assist the country in equipment when it reaches the end of its meeting its compliance obligations. Succes- economically useful life is equal to the gross sive auctions need not result in higher offer capital expenditure for the non-ODS using prices. To the extent that over time, the cost of equipment; the argument remains the same premature replacement of existing capital even if this assumption is relaxed. equipment declines, the price differential The current rules ignore the expected between ODS and its substitute declines, remaining economic lIfe of existing equip- technological growth and adaptation reduce ment, discount rate, collateral benefits, net conversion costs, and the firms perceive expected project life, and expected change in greater advantage in conversion due to the IOC. Even when implementing agencies market shifts, their offer prices can be ex- such as the World Bank adopt a particular pected to decline correspondingly. discount rate or project lifetime for their 20 Environment Department Papers Bid Auctions for ODS Phaseout calculations, such assumptions are necessarily in IOC 30% p.a., assuming that the firm does arbitrary and may bear no relation to the not derive any savings or collateral benefits firm's own cost of capital or business develop- from conversion over the project life. If it also ment plans. On the other hand, the neglect of expects to earn such savings or benefits, then such variables essentially implies a presump- it ought to be willing to accept less than tion that the remaining economic life of $19.2/kg ODP. The point is that firm-level existing equipment is infinite, the discount economic parameters which are ignored rate is zero, there are no expected collateral under the administrative rules strongly benefits of conversion and expected change in influence the choice to convert, and that these IOC is also zero. In other words, only this parameters vary across firms. A corollary is combination of these (ignored and unobserv- that government actions which influence the able) parameters would equate the Fund's firm's valuation of incremental costs would TABLE 2: FUND-APPROVED DOMESTIC REFRIGERATION PROJECTS - "REPRESENTATIVE" AVERAG APPROVED GRANT VS. INCREMENTAL CONVERSION COSTS Hypothetical values based on average of "representative" projects $/kg ODP abated (a) Gross capital expenditure $14.3 (b) Incremental operating cost (IOC), Year 1 9.9 (c) Total grant approved (a + b/2) 19.2 Calculated conversion cost = [incremental capital cost ICC) = net present value (NPV) of the carrying cost of premature replacement] + [NPV of IOC over project life], at Remaining Discount Project Annual ICC NPV Total economic rate life change in of life IOC IOC Infinite 0% p.a. 10 years 0% p.a. $14. 3 $99.0 $113.3 10 years 10% 10 -10% 8. 8 42.8 51.6 5 20 5 -20% 10.8 26.3 37.1 0 25 4 -30% 2. 9 16.2 19.1 Revealed "willingness to accept' = "agreed incremental cost" as % of calculated conversion cost (range) 17% - 100% grant with the firm's own valuation of also affect its "agreed incremental costs" - incremental cost (i.e., its willingness to for example, the economic life of its existing accept)." and replacement equipment via accelerated depreciation allowance or investment tax When would the firm be willing to credits; discount rate via concessional financ- accept $19.2/kg ODP? This question can be ing; and, the IOC via an excise tax on ODS examined by varying the assumed parameters and/or price subsidy for substitutes. of expected remaining economic life, discount rate, project life, and expected change in IOC. Table 2 also shows that if this hypo- A number of combinations of these param- thetical firm's "willingness to accept" were to eters may show the calculated total conver- be measured as the ratio of accepted grant sion project cost to be close to $19.2/kg ODP. ($19.2/kg ODP) to total project cost, it can One such combination is when the remaining range from as little as 17% to over 100%. The economic life is zero, the discount rate 25% "true" costs and collateral benefits as per- p.a., project life 4 years, and expected change ceived by the firm can never be known Environmental Economics Series 21 Market Based Instruments for the Implementation of Montreal Protocol in the Developing Countries precisely, nor is it at all necessary to know the cost of non-ODS-using capital equip- them. However, once the firm has agreed to ment'0; (v) incremental operating costs are accept $19.2/kg ODP, it represents at least positive for the foreseeable future11; and, (vi) 100% of its "agreed incremental cost". If its collateral benefits to the firm and local own valuation of the incremental costs results environmental extemalities are negligible. in a lower "willingness to accept", or "offer The main argument - that the "willingness price", say $10/kg ODP - based on its to accept' is determined by the firm-specific perceptions of expected savings or collateral parameters known to the firm itself, and that benefits, say - then the $19.2/kg ODP grant it would differ across firms - does not provides it an excess over what is needed to depend on any of these assumptions; the buy phaseout from that firm. design of a specific bid auction, however, would need to examine each of the assump- The illustration in Table 2 also shows tions. that by covering the entire capital expendi- ture, rather than the incremental capital cost, Table 3 shows an illustrative calcula- the Fund's current rules provide an incentive tion similar to that in Table 2. Unit gross to firms to phaseout their ODS use prema- capital expenditure is taken to be $9/kg ODP turely. If the price difference between ODS because safety and training costs are ab- and substitutes is expected to decline over stracted from (and because a typical firm in time, there are additional costs of premature Brazil is larger than those represented in replacement. The costs of premature replace- Table 2). First year IOC is taken to be $10/kg ment can be quite high, unless offset by ODP, in part based on the chemical price data incremental operating savings. If there are no for Brazil (see Annex Table A.5)"2. Both these such savings, the chemical (ODS vs. substi- variables - gross capital expenditure and first tute) price differentials dominate the conver- year IOC - and only those variables are sion costs, suggesting that the user firms' assumed to be observable. For simplicity, all "willingness to accept" may be significantly cost estimates are presented in unit ($/kg lowered if these price differentials can be ODP) terms. narrowed (e.g., by pricing interventions, as discussed in the next section). Under the current approach, the firm would qualify for a grant of $14/kg ODP, In order to examine the implications which is roughly equal to the CETV set for of firm-specific differences in costs and 1995/96 for this sector. Whether the firm's implied "willingness to accept" at a country- "willingness to accept" would be higher or wide level, a simulation exercise was under- lower than this amount depends on other taken for the "investment" subsectors (i.e. economic variables which are either known those projects where investments for conver- only to the firm or cannot be verified indepen- sion are to be made)- domestic refrigeration, dently. Some of these variables - strategic commercial refrigeration, and rigid polyure- market interests, or risk aversion, for example thane foams - for Brazil. A number of - are difficult to quantify and model; follow- simplifying assumptions were made for the ing the example in Table 2, let us assess the exercise; these include: (i) each firm operates influence of remaining economic life, discount in one end-use sector and its quantity of rate, and expectations of decline in annual phaseout offer is fixed and known in advance; IOC. (ii) each firm knows its technological alterna- tives and associated costs (i.e, each firm finds Remaining economic life depends in a particular substitute to be the cheapest); (iii) part on firm's discount rate, market strategy, the firm will convert to the non-ODS option and expectations; the same applies to "project at the earlier of when the economic life of life", i.e., the firm's expectation of how long existing equipment ends8, which is known in the conversion investment in non-ODS advance, or at a fixed date in future, dictated technology is going to last before it has to be by the national policy on Protocol implemen- replaced. To keep the illustration manage- tation9; (iv) the replacement cost of existing able, Table 3 shows calculations based on the ODS-using capital equipment is the same as assumption that the firm determines the 22 Environment Department Papers Bid Auctions for ODS Phaseout TABLE 3: ILLUSTRATIVE EXAMPLE OF SAVINGS UNDER COMPETITIVE BIDS (DOMESTIC REFRIGERATION) Assumptions Schedule - Mandated conversion at the end of 7 years - Remaining economic life of existing equipment 4 years Cost $/kg ODP abated (a) Gross capital expenditure $ 9 (b) Incremental operation cost (IOC), Year 1 10 (c) Of which, price differential between ODS and substitute 8 (d) Total grant under current rules (a + b/2) $ 14 Calculated conversion cost = [incremental capital cost (ICC) = net present value (NPV) of the carrying cost of premature replacement] + [NPV of IOC over project life], at Phase out in year 5: Minimum compliance cost (MCC) Discount Project Annual MCC, in $ of rate life change in year of phaseout MCC, in today's $ IOC 10% p.a. 10 years -10% p.a. $35.0 $21.7 20% 10 -20% 16.4 6.6 25% 10 -30% 10.6 3.5 Phaseout today: Cost of accelerated phaseout (IOC calculated only for 4 years) Discount Project Annual Incentive required for accelerated rate life change in phaseout (today's $) IOC 10% p.a. 10 years -10% p.a. $31.3 20% 10 -20% $25.9 25% 10 -30% 23.2 remaining economic life of existing equip- discussing the revealed willingness to accept ment to be four years and the project life to be for a hypothetical average of recent domestic 10 years, keeping in mind that neither are refrigeration projects - it would perceive its observable and could change. If the firm's MCC to be $10.6/kg ODP (four years from discount rate is 10% p.a. and it expects the now) and $3.5/kg ODP (in today's dollars). annual IOC to decline at the rate of 10% p.a., In the latter case, if the firm were given the it would perceive its MCC - the net present choice of accepting the grant today and value of the incremental capital cost plus the converting four years hence, it should be incremental operating cost, calculated for the satisfied with $3.5/kg ODP. Of course, it year in which the phaseout would occur - to would accept any higher amount if offered; if be $35/kg ODP four years from now, or it had to compete, however, and risk losing $21.7/kg ODP in today's dollars. With a the grant to its competitor, it would wish to discount rate of 25% p.a. and expectation of offer its phaseout for an amount close to $3.5/ IOC declining at 30% p.a. - a pattem consis- kgODP. tent with the example given in Table 2 Environmental Economics Series 23 Market Based Instruments for the Implementation of Montreal Protocol in the Developing Countries Let us suppose the firm weighs the latter firm's remaining economic life were choice of converting today versus four years nearly zero'5, its bid would be lower. The hence when it has to. Such premature ODS firm modeled in Table 3 may choose between phaseout would increase the firm's economic not bidding in an auction today and take its capital and operating costs of adjustment. chance next time, or participate if it believes First, the capital costs would increase because its bid to be lower than what it perceives its equipment is replaced even though it has a competitor's bid is likely to be. If it fears a significant remaining economic life. The loss of competitive advantage, it might even economic costs of early replacement are given lower its own bid below its own estimate of by the carrying costs of the new equipment, incremental cost because in doing so, it only calculated for the number of years of prema- hopes to gain a benefit (competitive advan- ture replacement. Second, the operating costs tage). also increase because the firm begins to incur the differential operating cost before it has to, Savings via competitive bidding and because the differential operating cost is lower in the later years as a result of (as- Any estimate of the potential savings sumed) expected decline in price differen- of the Fund's resources under the phaseout tials"3. bid approach requires a specification of the firms' bids as well as grant levels under Given the assumption that the firm current rules. Clearly, this is a difficult task, would switch to the non-ODS technology since the behavior of each firm would depend when it replaces its existing equipment (in upon its own circumstances, as well as its year 5), from the firm's perspective today, if it perceptions about the behavior of other firms. were to convert today in return for a grant, it An auctioneer can never precisely know what would compare the cost of phaseout today to bids would be offered; rather, s/he can only that in year 5, and not to that of continued use guess the offers based on market research and of ODS technology permissible in the absence then devise the rules of auction accordingly16. of the Protocol. This is calculated in the lower portion of Table 3 as the carrying cost of To demonstrate savings, one would premature replacement and the IOC over four have to select arbitrary assumptions and years. behavioral rules for each firm. For instance, one could specify that the firm would choose It is clear that any incentives for to accept 50% of its MCC calculated at 10% premature replacement have the potential for discount rate and expected change in IOC of adding significantly to the costs of adjustment 10% p.a., and thus derive savings of 22% as so long as the firm expects the IOC to remain compared to $14/kg ODP. Or one could positive and any collateral benefits of adjust- specify that the firm would only accept 100% ment to be negligible. On the other hand, as of its MCC at 25% discount rate and expected the discussion for Table 2 observed, the change in IOC of 30% pa., and show savings "revealed willingness to accept" in projects of 75%. (In each case, the firm's MCC is thus far imply that at least some of the firms expressed in today's dollars for a conversion that have accepted the grants thus far have a in year 5). But the firm's discount rate and combination of higher discount rates, shorter expectations, as also its perception of remain- project evaluation horizons, and an expecta- ing economic life, its market strategy and risk tion of faster decline in lOCs'4. Parameters aversion, or its collateral benefits of conver- for the particular hypothetical firm modelled sion, may not be known or determined in Table 3 are not known in advance (nor can administratively'7. An actual bid would only they be). In a bid auction, however, many show the price and quantities demanded, not other firms with varying circumstances what percentage it may be of the firm's would participate. Some other firm with (unknown and unknowable) MCC. To different parameters would have different maximize the chances of winning, the firm costs - not externally knowable - and can also be expected to seek to lower its costs different "willingness to accept", which its as much as possible - e.g., identify an ODS phaseout bid would reveal; for instance, if the substitute at a lower price'8 - and to choose 24 Enviromnent Department Papers Bid Auctions for ODS Phaseout its technology option in such a way as to framework in which to design and conduct an maximize its savings and collateral benefits. auction. Firms enter an auction with their assessment of relative likelihood of success, Competitive pressures are likely to but are likely to avoid them altogether if they ensure that the firm would bid as close to its cannot sufficiently rely on there being some "willingness to accept" as possible or risk successes. In order to avoid the so-called losing the grant to other firms"9. Once the "disbursement problem" that arises when the firm makes an offer, it represents its "willing- budget is fixed upwards as well as down- ness to accept" and thus at least 100% of its wards (i.e., a fixed sum has to be disbursed in "agreed incremental cost". a particular period), the country should not be required to spend the entire sum in a given auction, but rather permitted to carry over Implementation of bid auctions amounts not committed to subsequent roundsO. Also, when bid auctions are applied Bid auctions of varying complexity in multiple countries, a country should not be can be designed to meet the needs of specific penalized in a subsequent round of pre- situations. The simplest from the Fund's allocations simply because it was able to perspective would be periodic (say, annual or obtain significantly cheaper reductions in a biennial) sealed-bid auctions, where each previous round. In other words, because one bidder offers different quantities and prices, auction's results are not necessarily good similar to many other conventional systems of indicators for the future, subsequent aggre- bulk procurement. Bidders would self-select gate pre-allocation at the country level should the year in which they bid. Each bid would not penalize good performance in an earlier follow a standard contract proposal format, round. which would consist of a technical proposal to confirm its phaseout approach and a In the absence of prior information separate cost proposal constituting its "offer indicating that a particular class of partici- price" or grant request without revealing its pating firms may share certain characteristics "true" (and unknowable) costs'. which justify breaking the auction across different groups, the simplest option is to Table 4 provides a general outline of have a single auction across all firms, with a the main issues and options that arise in the maximum acceptable bid price announced in design of bid auctions. For simplicity, it is advance and the right to reject any or all bids. assumed that a government agency is the auctioneer and sets the rules; this function Even if the auctioneer is permitted to may well be performed by an independent carry over amounts from one period to next, contractor. The relative importance of issues, he may be concerned that not enough firms and the relative merits of options, are specific would participate. The issue is not what their to particular situations; the listing here is offer prices would be, just that they would not meant to be illustrative, though it does reveal the offer prices. Some firms may well address a larger set of alternatives than the appear to be "not convinced"; perhaps their illustrative example above. existing equipment is younger, perhaps they expect CFCs to get cheaper as demand for For instance, a variety of altematives them falls, perhaps they do not derive any may be considered by the MFEC to determine collateral benefits of conversion. An auction- the country-level aggregate sum; the simplest eer may well make participation mandatory, would be setting a minimum target for the especially for a certain class of firms (by ODS, aggregate phaseout offers - in tons per million use sector, or size group). Additional pressure dollars, say (or, conversely, a maximum may be exercised by adopting a "use it or lose aggregate average price). Pre-allocation of it" rule - for example, if the firm does not the aggregate sum may be necessary to help submit a winning bid (i.e., in competition with establish confidence among bidders and other firms) for three successive rounds, and provide the national ozone units a stable continues to be using ODS five years ahead, Environmental Economics Series 25 Market Based Instruments for the Implementation of Montreal Protocol in the Developing Countries its ODS entitlement (the quantity it is permit- round of auction may "fail" in that there ted to offer for phaseout) would be reduced would be some money left to be carried over by, say, 50%. Arguably this stretches the to the next round, whose rules may have to be meaning of "agreed incremental costs"; the set differently. A similar problem may rise at idea is simply that an auctioneer, once the level of a class of firms. Let us suppose, determined to obtain a certain amount of for example, that a certain class of firms may phaseout, can use a variety of tools at his have inherently much lower WTA than disposal. Another, less drastic, practical another such class, and individual firms in solution to get around the problem of "non- both classes are aware of this. Firms in the participation" is to set a minimum guaran- former group may be able to bid uniformly teed price (perhaps even different price bands higher prices up to what each of them per- for different end-use sectors)22. ceives to be the likely lowest bid from the latter group. If the auctioneer also knows of If a likely reason for the firms' reluc- such a possibility, he may choose to set tance to participate in bid auctions is their "divided auctions" with different maximum perceived difficulties in obtaining the techni- acceptable bid price for different classes of cal and economic information required to bidders. Such "divided auctions" should be prepare the bids, the Fund should expand the run concurrently, without setting a sector- technical assistance activities at the national specific budget limit. level. As compared to the project-by-project approach, a bid auction approach shifts the In short, even as competitive bidding burden of evaluating technical and economic practices are very much rooted in the national options from the Implementing Agencies and legal and organizational (business and their consultants to the firms. On the other government procurement) cultures, auctions hand, the project-by-project approach limits can be custom-tailored to meet the efficiency the transmission of information to one firm challenges in specific situations2. As com- (or a group of firms) at a time. If the national- pared to the project-by-project approach, one level technical assistance activities are can gather a large number of competitive bids expanded, not only do all firms share equally in one sweep; indeed, if the Fund can commit in the information and expert advice avail- itself to multi-year allocations in advance, a able (which may encourage them to partici- national auctioneer may well permit the firms pate in the auction), but the unit costs (per a wider latitude in choice of phaseout timing ODP phased out) of providing such services and have multi-period bids in a single auc- will also be lower. Another advantage (as tion. That is, firms offering phaseout four comapred to firm-level technical assistance) years hence now at a $/kg price to be paid would be that the government would obtain then may compete with firms offering phase- greater information and understanding of the out a year hence at a $/kg price to be paid ODS markets, enabling it to create more now. (To the extent that premature replace- effective policy frameworks and design more ment costs and the price differentials between efficient auctions (e.g., by creating due ODS and substitutes decline over time, diligence criteria to minimize the risks of non- phaseout offers can be expected to be lower in performance). the future.) Similarly, although heterogeneity Unlike other routine bulk procure- among ODS uses and users may well ensure ment processes, it is impossible in the case of competition, the auctioneer may be con- ODS phaseout to observe another actual cerned, based on market intelligence, about procurement and get a sense of prevailing the possibility of collusion among bidders. He market price or different firms' bids. In large could set a "maximum acceptable countries, they would at least reduce the price"which somewhat protects against administrative costs of project-by-project collusion. If some of the bidders find the need assessment and may reduce the monitor- maximum acceptable price too low, the first ing and verification costs as well. On the 26 Enviromnent Department Papers Bid Auctions for ODS Phaseout TABLE 4: LIKELY ISSUES AND OPTIONS IN DESIGN OF BID AUCTiONS 1. Determination of country-level total grant (assuming a given aggregate sum at the Fund level): a. Pro rata according to the countries' per capita ODP baseline consumption. b. Pro rata according to the countries' per capita ODP "consumption" phaseout commitment in excess of Protocol requirement. c. International comparison: The Fund sets an aggregate maximum average price based on prior experience and market analysis, including sectoral composition of ODS use and indications of sectoral variations in cost components. d. Competitive inter-Party (not cross-national inter-firm) bidding: The Fund permits Party bids, combined with pre-set price ranges and/or ODP quantities. 2. Terms of aggregate (country-level) grant: a. Disbursement (i) Up-front; (ii) Upon bid selection and award; (iii) linked to disbursement to firms; (iv) combination. b. Duration: (i) fixed period (e.g., annual); (ii) roll-over permitted. c. Grants for implementation and enforcement costs: lump-sum or prorated according to phaseout success or other criteria (e.g., regulatory measures, stricter limits than those imposed by the Protocol). d. Implementing Agencies' role: nationul "portfolios" (rather than project portfolios or sectoral portfolios; one manager responsible for an entire country); review and approval of bidding rules and bid selection; implementation monitoring and reporting. 3. Determination of firm-level ODS use (phaseout offer) entitlement: a. Period: Target year only, or interim annual, or cumulative over a period. b. Quantity: (i) based on historical amount ("grandfathering") and reduced according to the Protocol's schedule for the entire country; (ii) based on initial auction of "use permits". 4. Auction rules: a. Participation: voluntary or mandatory ("use it or lose it"); general or specific to substance (e.g. halons), use-sector (e.g. domestic refrigeration), or size groups; user firms only or third party (agent/broker) bids. b. Aggregate quantity targets: (i) none; (ii) minimum total from acceptable offers; (iii) maximum total from acceptable offers. c. Performance schedule: (i) at a fixed time (e.g. 2 years from date of bid award); (ii) at any time up to a pre-set schedule (e.g. any time by 2004). d. Acceptable bid price ranges: (i) none; (ii) maximum acceptable price; (iii) minimum guaranteed bid price. e. Selection criteria: (i) award bid price, subject to applicable maximum; (ii) award a market clearing price. f. Bid rejection criteria: (i) unacceptable price; (ii) unacceptable verification procedure. g. Rebidding within the same auction: (i) not permitted; (ii) permitted. h. Government participation: May choose to "withhold" a certain quantity of entitlements as permitted under the Protocol, or create and hold entitlements to "reclaimed" ODS, in order to make up for any shortfalls in an auction. 5. Terms of grants to firms: a. Disbursement: (i) up-front cash; (ii) up-front promissory notes or vouchers, encashed upon verification; (iii) combination. b. Verification: (i) by governrment agencies according to detailed pre-set rules; (ii) left to firms and industry groups to develop standards, and demonstrate acceptable compliance to broad, general rules; (iii) combination. Notes: 1. More options are suggested here than discussed in the example in the main text; some of these options may impose different firm-level control requirements than those imposed at the Party level by the Protocol. Environmental Economics Series 27 Market Based Instruments for the Implementation of Montreal Protocol in the Developing Countries would be much higher than they need be. other hand, if a large number of user firms are encouraged to compete, the rate of conver- Finally, projects involving conversion of sions may be expedited, with ratcheting ODS production facilities, or reclamation activi- effects on market composition (ODS vs. their ties, are unlikely to be suitable candidates for substitutes, and products containing them)2'. national-level bid auctions together with the user The only way to know the prevailing market firms. This is because in both these types of price or the scale effects of conversion is via projects, a firm's expectations are crucially depen- conducting an actual, pilot auction. dent on the aggregate decisions of all the user firms who are participating in the auction. If the It bears emphasizing that bid auctions user firms reveal especially low WTA and the as proposed here be applied only at the auctioneer is able to buy a great quantity of national level, not at the inter-country level phaseout with the available budget, the producing (whether between countries or between firms companies have a high economic incentive to shift from different countries, as is implicitly production but the reclamation activities would permitted in the Fund's use of CETVs). Inter- collapse. The converse also applies. As discussed country bid auctions are likely to be impracti- briefly in the next section, the issue of production cal, inequitable, and may even be inefficient. sector conversions and reclamation activities This is because governments have no way of needs to be addressed in a comprehensive manner estimating the optimal conversion costs of since the markets for controlled virgin ODS, individual firms, and firms in different recovered ODS, and their substitutes are essen- countries operate face different markets and tially worldwide in nature and price of any one policy environments. The potential for firms' should affect the other two. forecasting and estimation errors is far greater (than when they are competing for grants against other firms in the same coun- try), with the likely result that the bid prices 1/ A number of US electric utilities have sought additional capacity via competitive bidding including from demand reduction. For a broad overview of US bidding programs, rules and results, see Swezey (1993) and Kliman (1994). A number of developing countries have also used competitive bidding programs for privatizing existing electric utility assets and/or acquiring new generating assets. 2/ For example, the end-use markets (refrigeration, foams, aerosols), remaining economic life of existing equip- ment, discount rate (in turn based on cost of equity and borrowing), marketing strategy, market power vis-a-vis its suppliers and competitors, and benefits of conversion. 3/ Valuation of existing assets affected by deregulation of the electricity industry in the US - i.e., of the so-called "stranded assets" - has certain similarities to the problem here of determining the firm-level economic costs of regulatory change. See Baxter (1995) for a discussion of altemative approaches, including "market-based", to this problem. At least one US electric utility has proposed competitive bidding as a means to determine such values. 4/ A firm knows only its own costs but not that of its competitor, and the auctioneer knows neither except as actually revealed in their offer price. Sequential auctions have the advantage that the auctioneer gains knowledge from the first auction that he can use in the next round, as can all the firms whose bids are rejected. _/ In terms of economic theory, this is a "lump-sum" subsidy that has no distortionary effects because it does not affect the prices and costs faced by the firm at the margin. 6/ Note that "excess demand for grants", which in turn has required changing the rules of the game, is a practi- 28 Environment Department Papers Bid Auctions for ODS Phaseout cal problem for the Fund and the implementing agencies only under the project-by-project approach. Strictly speaking, an auction would also have a residual, unmet "excess demand"; the distinction is that the uncertainty and the costs of preparing the bid are borne by the enterprise, who is therefore likely to adopt a strategy which would maximize its probability of winning (i.e., all other things being equal, come as close to its own perceived value of the phaseout grant). Z/ In order to derive a "representative" sample, a significant number of projects had to be eliminated for certain reasons: the projects included compressor manufacturing conversion, or had partial foreign ownership (which limits Fund's grants), or did not contain data on prices for ODS and their substitutes, or showed unit grant or cost levels which appeared to be considerably outside the range represented by others. 8/ A firm may well choose to replace existing equipment with the same technology instead of converting if it deems the potential cost of premature replacementin response to the Protocol requirements outweighs the avoided incremental operating cost in the interim. 9/ For example, assuming that the govermnent imposes the same phaseout schedule on user firms as that on production and imports. This assumption is needed to avoid a separate analysis of the options to stockpile virgin ODS or create/purchase reclaimed ODS. 10/ While licensing, testing and training costs for a new technology may be significant, they are assumed to be at least offset by productivity gains or other collateral benefits. 11/ At least to the extent that the net present value, if negative, of incremental operating costs does not more than offset the initial capital cost, rendering the conversion project as an economic one which would be undertaken by the firm without the Fund's support. Projects with such "net savings" may not need the Fund's support but, under the Fund's current procedures, are eligible for grants. 12/ International price data and projections are not available except for the expected prices for selected firms (from the Fund's project documents), which show tremendous variations. Although Brazil produces ODS substi- tutes already, the reported HFC-134a price in Annex Table A.4.3 is significantly higher than the generic values used by OORG (1994) or ICF et al. (1994). If HFC-134a prices can be lowered to $4-6/kg, as observed for some other countries, the project cost estimates shown here would be substantially lower. 13/ Note that the proposition "the later the cheaper" is valid only when the present value of net incremental costs (capital plus operating), as calculated from the firm's perspective, is positive. 14/ For ease of modeling, this expectation has been defined in terms of the change in the price differential between ODS and its substitute. It can also serve as a proxy for expected collateral benefits or remaining economic life. D In which case its choice would be between bidding now and converting to non-ODS technology with the amount it can win and replacing existing equipment with the same (ODS-using) technology and bidding later. The Fund's current rules do not allow it the latter option, but it may still find it in its economic interest to continue using ODS until some later date (up until it is legally banned from doing so) even if it does not obtain any conver- sion grant. 16/ For example, maximum acceptable bid and/or minimum guaranteed bids (for small users, say), by sector or across sectors. It is beyond the scope of this paper to discuss alternative auction types and corresponding bidding strategies. For an introductory review, see Milgrom (1989). 17/ And, to the extent that the Fund may have some prior infornation about some of these variables, under current rules it would have to engage in detailed project review and negotiation; note, however, that at the very least only the total sum need be negotiated. 18/ Note that the reported price differential at this time between CFC-12 and IFC-134a in Brazil is much higher than observed elsewhere or as perceived to be "normal" by the OORG. 19/ Although the illustration provided above is for the domestic refrigeration sector, the auction is assumed to be Environmental Economics Series 29 Market Based Instruments for the Implementation of Montreal Protocol in the Developing Countries available to firms in all sectors. The auctioneer may choose to announce in advance a "maximum acceptable price" for each sector, but competition across sectors must be permitted. 20/ This assumes a sealed-bid discriminatory auction, where each bidder gets just the price in his/her bid. Other types of auctions and rules - open auctions, or permitting re-bids - can be envisioned as well, but require more research for specific situations. In principle at least, because the Protocol allows a significant interval until devel- oping country firms are obligated to phaseout, an auction may issue "entitlement certificates" valid for a certain number of years, and these can themselves be traded among firms in tum. Similarly, an open auction permitting rebids may use "kg ODP per $' rule and allow firms to bid partial quantities at different prices. There are many theoretical possibilities; practicality beyond the simple concept presented here depends on specific situations. 21/ For example, because not enough "acceptable" offers are received which would permit the country to meet the terms of its aggregate grant from the Fund. More specific rules would need to be developed; e.g., a "use it or lose it" (partially or fully) rule for firms in a specific sector may reduce the likelihood of such an outcome. 22/ The "sectoral approach" now under consideration for China addresses precisely this "non-participation" issue. On the other hand, administrative assessments of sectoral incremental costs may also be prone to error; some firms from a sector believed to have higher offer prices may in fact underbid some firms from the sector believed to have a lower offer price, if cross-sectoral competitive bidding were allowed. 23/ Some parallels are found in the electricity industry. In the early days of promoting non-utility generation in Califomia, utilities were required to contract with cogenerators and independent power producers at the projected avoided cost. For projects below a certain size, utilities were required to use the "standard offer" - akin to "fixed price" - contract. For larger projects, contracts were individually negotiated, but with the avoided cost calcula- tions as the benchmark. Later, as the avoided cost declined, as did the prices offered. State agencies and utilities still had an interest in buying renewable sources of energy, so they devised a "renewables only" competitive bidding. In the UK, the govemment did not wish all renewable technologies to compete against each other, so they employed technology-specific indicative "price bands", leaving actual negotiatirig details to the project de- veloper and the distribution company. 24/ As the use of ODS substitutes and products containing them expands, domestic market forces may become the "drivers" behind expanding their market share even more. 30 Environment Department Papers 5 Other Market-Based Instruments Conceptually, competitive bid auc- tax credits, or concessional financing, tar- tions represent a market-based instrument geted at conversion projects; or, differential that requires neither the price nor the quan- tax treatment of goods containing ODS vs. tity of ODS phaseout to be pre-determined; it those containing the substitutes. Some is a means of allocating the grant budget'. general features of these MBIs are: Two other broad categories of MBIs are also possible: 1. Flow of funds: Because permit auctions and ODS taxes involve flow of funds Tradeable permits for ODS "consumption" or away from the firms to the government, their use: Once the national government has political acceptability may be low. On the elected to implement the Protocol at the level other hand, governments may elect to either of domestic firms by granting them permits to recycle the revenues in the form of tax credits produce, import, or use ODS as described in to firms which undertake conversion or Section 2, it may further allow them to trade subsidies on ODS substitutes. Another these permits among themselves upon altemative is to dedicate the revenues to a mutually acceptable terms. That is, the trust fund (perhaps with control shared government determines the aggregate allow- among industry representatives and the able quantity, and the market determines the government) to finance industry-wide prices as well as who gets to produce, import activities such as information-sharing, or use how much of the ODS. Additionally, technology and market intelligence, training, the government's initial allocation of permits and research and development. (The Fund may itself be market-based, e.g., when they itself may choose to participate in such are auctioned off rather than conferred to national trust funds as a means of meeting firms on the basis of historical record. incremental costs which are not specific to individual conversion projects.)2 Pricing (tax/subsidy) policies: Via pricing policies, the government influences the prices 2 Financial vs. Economic Calculus: of ODS and/or their substitutes, and the Generally speaking, both permit auctions and market determines the firm-level prices and taxes or subsidies affect firm-level financial quantities of chemicals as well as the aggre- calculus without affecting national economic gate quantities of chemicals. Main examples costs of compliance, except to the extent that are a tax on ODS and/or a subsidy on substi- they reduce uncertainty, motivate search for tutes; more generally, a differential taxation lower-cost solutions, and increase the flow of scheme according to the ODP of individual information (if the bids in permit auctions are substances. The rates of tax and/or subsidy made public). On the other hand, they both may also be varied over time. do influence the "agreed incremental costs" in the market for ODS phaseout, i.e., the Other market-based policy instru- transactions between the Fund and the firms ments are also available to governments: for which are willing to convert. That is, a firm example, use of fiscal policies such as acceler- which has to purchase a permit from the ated depreciation allowance or investment government (in the initial auction) or from Environmental Economics Series 31 Market Based Instruments for the Implementation of Montreal Protocol in the Developing Countries another firm would weigh its decision against powerful complements to the Fund's grant the choice to convert, offering its phaseout allocation process and may well lower the service to the Fund. Similarly, a tax on ODS governments' costs of monitoring and en- and/or a subsidy on the substitute affects the forcement. Some comments on the issues firm's valuation of the incremental operating involved in their design and use are offered cost associated with conversion. Two impli- below for further analysis and research. cations follow: (i) these other MBIs can be powerful complements to the Fund's grant Tradeable permits assistance to firms; and, (ii) to the extent that government's use of domestic transfer A tradeable permit regime is suitable mechanisms reduces the Fund's grants to for situations where only the aggregate firms but not necessarily the national eco- environmental damage from multiple sources nomic costs, the Fund should provide prefer- is sought to be contained, and the cost of ential treatment to governments willing to so containment varies among individual collaborate (e.g., via direct grants to govern- sources. Compliance with the Protocol is ments or domestic revenue-recycling mecha- certainly an aggregate obligation, and from nisms such as suggested above). the phaseout projects undertaken thus far it appears that the cost of conversion to ODS 3. Timing of policies: Because devel- substitutes does vary significantly across oping countries are not bound by any aggre- firms. A tradeable permits regime would gate limits on ODS consumption until 1999, appear to be applicable to such a situation. and because the effect of developed countries' production phaseout (with a limited exemp- In fact, Section 3's characterization of tion for production to export to developing transactions between the Fund and the user countries) on ODS markets and use patterns firms as a "market in ODS phaseout" is only a are yet to be determined, some latitude is special sub-market in a tradeable permits available for the design and implementation regime. It is easy to see how this is so. The of a tradeable permits regime (in production, firms may be seen as holding a certain prop- imports, or use) or a pricing regime. Proper erty right, a permit - to use ODS, that is - design of pricing policies or of "rules of the and they may choose to enjoy that right game" in a tradeable permit regime would themselves, or sell it to another firm, or "sell" require harmonization between different it to the Fund which "extinguishes" this types of permits (production, imports, and property right. The sole distinction is that if use) and perhaps international policy harmo- another firm bought the permit, it would nization as well (so as to have an inter- likely "exercise" the permit - i.e., use ODS national regime in production permits, for during the period permitted - while the example, or so as to avoid undesirable im- Fund buys it off to permanently "retire" it. pacts on the domestic firms' competitiveness The situation is similar in the case of permits in world trade). In the case of ODS tax for ODS production. A particular advantage policies, for example, developing country is that the observed prices in inter-firm trade governments may well wait to use them if provide information to the Fund about how and when the risk of "under-phaseout" different firms value ODS use, and conversely becomes imminent - at any time between the observed prices in the Fund's competitive 1999 and 2004, say - and then apply taxes to auction of grants provide similar information the level of severity required. to firms. If a firm can offer its permit to another firm at a higher price than the value This paper deals primarily with it has to exercise that permit for its own improving the efficiency of the Fund's grant benefit, it is rational for it to switch to ODS allocation process (with the flows of funds to substitute; in turn, if it can obtain a higher firms) and actions the Fund is authorized to price for the permit from another firm than take, while these other MBIs are subject to from the Fund, it is rational for it to offer it to national governments' jurisdiction. In due the former. In either transaction, the offeror course, developing countries may well wish to has competition; indeed, under a tradeable take advantage of these MBIs which can be permit regime, the Fund also has competition 32 Environment Departnent Papers Other Market-Based Instruments from other buyers. In practice, of course, all ODS tradeable permits regime may be quite these transactions are not likely to be struck at high, as compared to, say, the celebrated a single point in time with identical informa- sulfur dioxide (SO2) emission allowance tion available to all participants; therefore, scheme in the US (under the so-called "acid there will be divergences between the prices rain program"). Indeed, it appears that even in inter-firm transactions and those in the though the US and the European Union Fund's bid auction. permitted trading of ODS production and "consumption" allowances among firms, the Permit regimes of varying complexity actual volume of inter-company, inter- can be created. For example, firms may be substance and inter-national trades thus far is allowed to trade permits for one ODS for reportedly extremely small. The likely another within a Protocol grouping (e.g., CFC- reasons, and the implied differences from the 11 for CFC-12) for their own use or for US acid rain program, may be summarized as trading to another firm. Firms may also trade below. permits for any future period. Transactions between firms may be permitted to be bilat- (i) In many countries, the use of eral and private, with only the record of certain ODS is scattered among a large transaction to be submitted to the govern- number of small firms. In such situations, the ment, or may be made in a central exchange, government's costs of tracking ODS permit with all transactions made public. exchanges would be very high, as would the firms' transaction costs. The risks of "non- National tradeable permit regimes performance" - i.e., firms exceeding their may also be linked so as to create interna- permitted use or continuing the use after the tional tradeable permit regimes. Indeed, the permits have been transferred to another firm Protocol explicitly permits inter-Party trans- - may also be very high. This is different from fer of production and/or net imports3. The the SO2 emissions situation, where a small US and the European Union have adopted the number of large emitters have to be moni- permissibility of such transfers in their tored, and the costs of monitoring systems are respective legislation or regulations4. Beyond minuscule as compared to the overall budgets 1995, the allowable production in developed of emitters. ODS markets also differ in that countries for export to the developing coun- ODS can be stocked and that reclaimed ODS tries to meet the latter's "basic domestic may be used in unrestricted quantities; this needs"' is also transferrable. Inasmuch as the suggests that firms may more wish to take markets in ODS and their substitutes are advantage of flexibility afforded internally indeed global, and that country-level compe- rather than trade with other firms. tition in converting ODS production may be (ii) Electric utilities have a high limited (see below), an international tradeable degree of flexibility to vary their SO2 emis- permits regime for ODS production may also sions; firms producing or using ODS may facilitate an international auction for produc- have extremely limited flexibility, if at all. tion sector conversion. Their choice in most cases is likely to be of a "yes"/"no" type - i.e., continue or convert On the other hand, even though (via any among multiple options). If they tradeable permits regime offer many opportu- expect some fluctuations in operations, they nities for policy innovation and can theoreti- are more likely to increase or deplete their cally offer significant cost reductions in ODS stock rather than purchase above-quota country compliance with the Protocol, the permits from other firms. On the other hand, actual savings achieved depend on the firms that for their own economic reasons volume and frequency of trading6, which to a (other than the market value of permits, that certain extent depend on the design of the is) produce or use less ODS than they are scheme, which can sometimes be a politically entitled to (or close operations) would earn a thorny issue. Several features of ODS markets windfall profit. (Note that national aggregate and the Protocol indicate that the volume of limits on production and imports apply every trading may really be quite small, and that single year, and inter-period transfers are not the costs of implementing and enforcing an permitted under the Protocol). Environmental Economics Series 33 Market Based Instruments for the Implementation of Montreal Protocol in the Developing Countries (iii) Some observers of the US acid or developing countries, in turn partly rain program suggest that firms are usually reducing the uncertainty about availability more inclined to enter the permit trade and costs of ODS over a longer term (say, five market if they have obtained permits by years or more). If the experience is hearten- paying for them (e.g., via an auction or on a ing, and if several large countries join in sliding-scale fee basis) rather than just developing national and international trade- receiving them. On the other hand, confer- able permits schemes, the volume of trade ring permits on the basis of historical produc- may grow and indeed deliver the savings tion or use is likely to be far more politically such schemes promise in principle9. acceptable than requiring fees or auctioning the permits. Because ODS production and Taxes and subsidies most of the uses are tied to long-lived capital stock, these permits would not be traded Taxes on ODS and subsidies on often, if at all. Even if import permits are substitutes - or, in broad terms, pricing auctioned, they are likely to be of short policies which may also include price stabili- duration (e.g., a year or two), and the permit zation schemes (floors for ODS and ceilings holders may have no desire to trade in the for substitutes) - may be somewhat easier to interim. use than a tradeable permit scheme. The key (iv) The Protocol's control sched- issues in using these instruments are adminis- ules apply to a large number of ODS, different trative efficacy and accuracy of the tax/ groupings of which have different reduction subsidy levels selected. Because compliance and phaseout schedules7. Unless inter- with the Protocol means honoring aggregate substance and inter-temporal trade were quantity linits according to a particular permitted, firms may just not have enough schedule, pricing policies alone risk being information - or, collecting such information insufficient or overshooting the target. may be very costly - to make reasonable valuations of the permits. Any new tax is likely to meet with resistance from those who have to pay it With the use of innovative designs of (though a subsidy would be welcome by those permits and "rules of the game", these and who benefit from it). Also, imposition of a other difficulties are not insurmountable. For new tax may well require legislative authority example, tradeability may be limited to a and collection and enforcement activities certain group of ODS and activities or firms. would require involvement of a number of Some other ODS or uses may be subjected to a government agencies. More importantly, it is "sectoral ban", if the costs of a forced switch difficult to determine, a priori, the appropriate are judged to be minor. Or the government tax/subsidy rate without a reasonable esti- may "withhold" certain quantity from the mate of price elasticities of demand, espe- initial allocation, if by auction, and use its cially over time (and, inasmuch as the Proto- own holdings to influence market behavior col constraints both certain ODS as well as and prices at a later point in time. their substitutes, albeit with different control schedules, the cross-price elasticities of In the near term, however, developing demand)"0. Prohibitive tax levels of the type countries may wish to try using annual or used in the US are simply not necessary until biennial auctions for tradeable permits for the national phaseout limits are approached" ODS imports8, and use the experience gained in developing countries. to create a broader, longer-term tradeable permit regime for ODS production and/or A tax/subsidy scheme would also use. As tradeable import permits regimes are require harmonization with other domestic established in a greater number of large policies and, if the country's exports of goods enough countries, firms may well find it in containing ODS or substitutes are significant, their advantage to enter in long-term con- with the tax/subsidy policies in major trade tracts with exporting firms in the developed partners as well. For example, if the country 34 Environment Department Papers Other Market-Based Instruments has implemented a scheme to auction ODS with the "permit and prohibition" regime import permits, the bids in such an auction described in Section 2. At the same time, their would provide some indicators of the scarcity governments may wish to (a) initiate the value of ODS to the user firms. For "imports legislative action required for imposition and only" countries, it may be argued that no collection of taxes (and the ability to adjust additional taxation is justified. For countries tax rates as deemed necessary) at a later date, with domestic production, import permit and (b) use the registry and permit systems auctions may lead to certain windfall gains discussed in Section 2 to study potential for domestic producers. It may be difficult to collection and enforcement costs and rev- estimate the extent of such gains in advance, enues under altemative tax regimes. The because as the ODS phaseout of recent years latter would also help determine the merits of begins to show its impacts, ODS markets may government- or Fund-financed centralized well swing one year to the next. Frequent ODS reclamation and banking schemes, and changes in income tax rates on domestic associated regulations to control the ODS producers or excise taxes on domestic produc- requirements to service the existing stock of tion are neither easy nor desirable. In some cooling equipment. Again, as in the case of situations, the govenmment may have to use instituting a tradeable permits regime, lump-sum taxes or fees, or non-price mea- national governments may need to be pro- sures to induce conversion by domestic vided direct incentives (grants) from the Fund producers (e.g., voluntary agreements). to undertake the steps which lower the firm- level financial costs of use phaseout (i.e., via Assuming that the administrative lowering the WTA in a bid auction) but not difficulties of policy harmonization and necessarily the country economic costs. coordination among government agencies are overcome, the determination of appropriate Note: Special Role of the CFC Production tax/subsidy rates would remain more a and Reclamation Activities matter of trial and error, which in turn may lead to some problems. For example, if tax It was pointed out at the end of revenues are supposed to be dedicated to a Section 4 that projects involving conversion specific phaseout assistance scheme, the risk of ODS production facilities, or reclamation exists that the revenues from users that activities, may not be suitable candidates for continue use (and thus pay taxes) fall short of bid auctions for user firms. There are other the expenditures for those who convert. One reasons to suggest that the nature of competi- answer to such a contingency would be to tion in the ODS production activities may be accumulate tax revenues in a "reserve pool" fundamentally different from that in the ODS or a fund to expedite phaseout at a later date use sectors. if necessary. First, unlike the use sectors, produc- Despite such doubts and difficulties, tion of specific ODS in a country is likely to be the powerful role of pricing policies in effect- concentrated in a very small number of ing compliance cannot be denied; it is pre- firms'2. Both ODS and their substitutes are cisely because of such power that their use generally bulk chemicals. Their production has to be cautioned against, otherwise the appears to be characterized by sizeable country may end up incurring excessive cost economies of scale and significant learning for compliance with the Protocol. For ex- effects, and sometimes patent monopolies. If ample, a prohibitive tax of the type used in the existing ODS producers are operating at the US may force a pre-mature retirement of sub-optimal size, and further if financing end-use equipment (such as domestic or their conversion is likely to lead to sub- mobile air-conditioners) without compensa- optimal size in substitute production, that is tion. At this point and for a few years still, clearly not in the country's economic inter- the main task for the developing countries is ests. to curtail the growth of use and consumption of virgin ODS, which can be effectively met Second, in the absence of import Environmental Economics Series 35 Market Based Instruments for the Implementation of Montreal Protocol in the Developing Countries restrictions, the markets in ODS and their current consumption rate in the developing substitutes are indeed global in nature, and countries; similarly, the cumulative perrnis- the Protocol's control measures clearly sible production far exceeds the likely cumu- provide for substitution of domestic produc- lative consumption. tion for imports (and vice versa) during the phasedown period. If importing ODS - from One may expect an increase in the other developing countries or from the production rate for 1995-97 period, if only developed countries, under the exemption because of the motivation provided by the granted for meeting the "basic domestic Protocol to establish a higher baseline; even needs" of developing countries - is cheaper then, aggregate capacity constraints are not than maintaining domestic production, the likely to occur14. In the presence of such latter should retrench. Similarly, if importing excess production capacity, reclamation may the substitutes is cheaper than converting the well be commercially unattractive. However, domestic production capacity, the country the question to ask is not whether it is cheaper should import rather than convert. to produce a ton of ODS than to reclaim it from existing stocks, but whether it is cheaper The calculus is further complicated by to convert to substitutes than to reclaim. (a) the permissibility of ODS stock increase Essentially, ODS phaseout (for non-halons) is during the grace period, (b) the availability of likely to become more of a "stock manage- destruction credits to increase permissible ment" problem as the "emnissive uses" (such production13, and (c) the exclusion of re- as in aerosols, solvents and foam-making) are claimed ODS from the control limits. This is curtailed1,. In developed countries such as because, even as ODS use is phased out, the the US, heavy taxation of sales of virgin ODS aggregate demand may not fall pan passu (whether from current production, as was the because of possible stockpiling: (a) by firms case until the beginning from this year, or who wish to service, i.e., re-charge, the stock from stockpile) combined with tax exemption of existing cooling equipment, and (b) by user for the reclaimed ODS, seems to have helped firms who do not yet convert, even if it is limit the stockpiling of virgin ODS and assumed that they do not increase their provided incentives for commercial ODS annual ODS use rate (if only because such an destruction and recycling activity, though increase is barred by the conditions of their perhaps not at as high a level as might have permit). Of course, stockpiling only shifts the been anticipated. time pattern of demand; ultimately the demand would have to fall. Similarly, the While it is not self-evident that aggregate supplies may also not fall, if the similar measures are yet useful for the devel- producers (or importers) choose to maintain oping countries, reclamation may be economi- their production (or imports) up to the cally as well as environmentally more attrac- permissible quota and add to their own stocks tive: the former since it yields a useful prod- for possible sales in later periods. uct, and extends the remaining economic life of existing ODS-using equipment (either in The essential issue is relative costs of the firms which use ODS to manufacture supplies for virgin and reclaimed ODS from other products or in the firms and individual different sources, domestic as well as imports, who use ODS to service existing stock of ODS- especially for CFCs. These price differentials containing goods), and the latter since it at are crucial to the economics of conversion (of least postpones emissions and may avoid use or production), of reclamation, and to the them altogether if ODS can be continuously design of pricing policies which seek to recycled. The Ozone Operations Resource influence the price differentials. Unfortu- Group (OORG) has thus recommended it as nately, data on historic or projected trends in "a first priority" (OORG, 1994, p. 27). availability and price of ODS and substitutes are extremely sparse. As shown in Annex Yet, only a few recovery and recycling Table A.6, the production rate of virgin CFCs projects have been undertaken thus far in the permitted by the Protocol far exceeds the developing countries, most of them at pilot or 36 Environment Department Papers Other Market-Based Instruments demonstration scale. One reason appears to activities for the CFCs - the situation seems be the greater difficulties of preparing and to be different for the halons - it is notewor- implementing a project with numerous, thy that the price differentials between CFCs small-scale firms in the so-called "service and their substitutes reported in recent sector". More importantly, there may be no project submissions (reported as a composite commercial incentive until the Protocol limits in Annex Table A. 3) appear to be much on virgin ODS become binding and/or their greater than those in the developed countries prices increase. Even then, optimal invest- (ICF et al. estimates as shown in Annex Table ments in recycling would remain subject to A.5). As the pace of conversion accelerates in uncertainties about the size - an unexpect- the developing countries, the prices of ODS edly rapid retirement of ODS-using freezers substitutes may be expected to fall, and the and chillers, for example, may lead to under- Fund's assistance to production sector con- utilization of recovery, recycling and recla- versions may help in this regard. Previous mation investments - and distribution of the discussion on the nature of competition on the end-use stock"6 and prices for virgin and supply side, and on the importance of recla- reclaimed ODS. The problem of establishing a mation activities, suggests that all sources of one-to-one relation between the Fund's supply be viewed in an integrated framework, accelerated phaseout program and a which is currently restricted by lack of country's compliance obligations may be relevant price data and cost indicators. more serious for the production sector con- version (than for the use sector): if the Fund So far, the Fund has not provided finances production conversion but the grants for production sector conversions country's imports make up for it up to the (except for some shutdown) and has financed amount permitted by the Protocol, not only is only demonstration projects on recycling and accelerated phaseout negated, but the CFC reclamation. Because assistance in produc- prices may not rise to the level that may tion sector conversions and recycling and justify reclamation 7 reclamation activities may both lower the economic costs of compliance and reduce the While the excess capacity in produc- uncertainties associated with optimal timing tion may have kept prices low and discour- of phaseout, greater priority ought to be given aged reclamation and banking (stockpiling) to such projects. 1/ Although the Protocol permits concessional lending approach for the financial mechanism, it has not been used thus far. Especially for those end-use sectors where technology assessments and market research suggest that ODS substitution project may be economical from the firm's perspective - i.e., the so-called "net savings" projects where the operating savings more than outweigh initial gross capital expenditures - concessional lending, loan guarantees or equipment leasing may be more appropriate than grants. In principle, market-based instruments can also be devised for concessional lending, though they are not addressed in this paper. 2/ Vincent (1996) discusses many such funds and associated issues. 3/ Protocol Article 2, paragraph 5 for production for London Amendment signatories and for consumption (i.e., production and net imports) for Copenhagen Amendment signatories. Governments, for example in the US and the European Union, have in turn allocated their allowable national consumption to domestic firms and permitted inter-firm trades. The Protocol mentions "transfer" rather than "trade" (and so do the US and European Union legislation as well), probably because the Parties as such may not claim an ownership interest in ODS. Governments thus recognize the inter-firm "trade" by the means of an inter-Party "transfer". 4/ For the US, see 40 CFR (Code of Federal Regulations) § 82.12 and a detailed discussion in the Federal Register, 60:90, 10 May 1995, p. 24970-25009. Also, see Lee (1996). With the introduction of "destruction credits" for CFCs Environmental Economics Series 37 Market Based Instruments for the Implementation of Montreal Protocol in the Developing Countries whose ODP value can be applied to expand a firm's allowance for HCFCs, the US has in effect expanded the scope of inter-substance exchange. For the European Union, see European Communities (1994). 5/ Protocol Article 2. The Protocol language is unclear whether the "basic domestic needs" criterion applies to one developing country at a time, or to all developing countries as a group. Inasmuch as inter-Party transfers of consumption (which indude imports) are permitted for the developing countries, the latter interpretation would appear to hold valid at least for the Copenhagen Amendment signatories among them. 6/ See Hahn and May (1994) and US GAO (1994) for analyses of factors limiting SO2 trading; the latter's critique of SO2 allowance auction by the US Environmental Protection Agency is also illuminating for comparable situations. Also see Hays (1995) for a critique of the US SO2 allowance trading program. Z/ In addition, the Parties have yet to determine developing countries' control schedules for CFC substitutes which also have ozone depleting potential and whose use will grow over time. Unless the limits were set in aggre- gate ODP-equivalent terms, national quantitative restrictions for each substance individually may create unnecessary disruption in industrial activity and distort firms' current choice of substitute technology. 8/ For imports-only countries, such schemes may be relatively easy to implement, as has been the case in Singapore and New Zealand, non-Article 5 countries. See Chapter 4 in UNEP (1995a) and SEI (1995). For a review of auction theory and application to replacing import quotas, see Takacs (1994). Countries which both produce and import may find it significantly harder. (Mexico also planned to have tradeable production and import permits, but implementation appears to have stalled.) 9/ Kerr (1995) argues for a transition from an international funding mechanism such as the MFIMP toward an inter-national tradeable permit regime, if the transaction costs are not too high. 10/ See Barthold (1994) and UNEP (1995a), Chapter 4. 11/ See Annex Table A.4.2 for US list price and tax rate information, US Internal Revenue Code 26 § 4681-2 for enabling legislation, and 26 CFR Part 52 for implementing regulations. Fay (1995) reports an estimate that the US has collected more than $6 billion in ODS taxes, and proposes that a part of the revenues be used for financing ODS phaseout in the developing countries, e.g., via the Fund. Also see Hoerner (1996.) 12/ Also, unlike the user firms, ODS phaseout (i.e., substitute production) is in the main line of business for the producing firms. 13/ The Protocol assumes that all production of virgin ODS ever will eventually result in atmospheric emissions. On the other hand, if any quantity of virgin ODS or ODS in the user stock (e.g., in the refrigeration and cooling equipment such as refrigerators, air-conditioners, and chillers) is physically or chemically destroyed (i.e., converted to substances which cannot escape in the atmosphere), a corresponding credit (in ODP terms) is granted to the country so that an ODP-equivalent virgin ODS is permitted to be produced, in exemption of the overall prohibition. The US has incorporated this provision in its national regulations. 14/ Under certain scenarios, one may well expect a sharp increase in ODS trade among the developing coun- tries, if imports from the developed countries become more expensive due to the plant closures in the latter. L5/ See Chapter 7 in UNEP (1995a) for an elaboration. Indeed, certain level of demands in the developing or developed countries can be met indefinitely by reclaimed ODS. 16/ Incentive size and structure, and the choice between retrofit and replacement of cooling equipment, depend on the end uses. The distribution of existing ODS stock in the cooling equipment of developing countries may be markedly different from that in the developed countries - for instance, a greater share in domestic refrigeration and air-conditioning equipment, which tends to have smaller leakage rates than in, say, commercial freezers and chillers or vehicular cooling systems. 17/ This problem was pointed out by the TEAP 1991 assessment: "If producers increase their demand for recycled materials in response to rising prices for newly produced ODS materials.. then markets for recycled materi- 38 Environment Department Papers Other Market-Based Instruments als will be stimulated.. However, the recycling market can be weakened by uncertainty about future costs and availability of substitutes and by inadequate policy pressures on the prices of ODS chemicals." (UNEP, 1991, summary para S.21). Envirortnental Economics Series 39 Market Based Instruments for the Implementation of Montreal Protocol in the Developing Countries 40 Environment Department Papers 6 Conclusions and Recommendations The years 1995-97 mark a significant than the current approach as well as other stage in the evolving history of the Protocol, as types of market-based instruments. We recom- the developed countries' ban on the production mend that the MFEC consider bid auctions as and imports of Protocol Annex A ODS has an acceptable means of grant transfers for those become effective and as the developing coun- countries which wish to use them. A pilot tries' baseline levels of production and imports scheme may be a simple way to begin, wherein come to be established. How these develop- the MFEC commits a fixed maximum sum of ments affect the global distribution of ODS and money over a particular period for a fixed substitute supplies and prices remains to be minimum quantity of phaseout to be obtained seen, but it is clear that the focus of ODS via a national auction program for a certain phaseout will shift to the developing countries. group of end-use sectors. The developing country, in turn, forbids any firm-level project In the context of the Fund's grant- proposal in those sectors to be directly submit- financing of conversion of ODS use to substi- ted to the MFEC (as is the current procedure) tutes, we have argued that its current proce- during that period. If the concept is agreed to dures have a potential to distort firms' choice of by a developing country and the MFEC, addi- optimal timing and technology, that adminis- tional country-level market research would be trative "need assessment" is inherently con- necessary in order to prepare a detailed auction strained in its ability to determine firm-level program proposal'. incremental costs, and that the project-by- project approach does not permit the Fund to In addition, we have pointed to the make an efficient use of its resources and need for developing countries to establish firm- involves high administrative costs. While these level limits (via permits) on ODS use, produc- procedures were necessary in the initial phase tion and trade, both in order to monitor compli- of the Fund's operations and have yielded ance with the Protocol and to protect the invaluable information and understanding, a phaseout achieved from being negated by transition to alternative, market-based ap- subsequent growth. The development and proaches is warranted, at least in the case of implementation of these firm-level controls large developing countries. would also help analyze the relative usefulness of other market-based instruments - a trade- We have provided a conceptual able permits regime and/or pricing policies - argument for an alternative scheme that permits that were briefly commented on in this paper. If self-selection and competition among firms, and carefully designed and implemented, these argued that competitive bid auctions for the other MBIs hold considerable potential for Fund's grants provide incentives to the firms to achieving an orderly compliance with the minimize their incremental costs and permits Protocol, at lower economic costs; promoting the Fund to acquire cheaper phaseout first. It optimal levels of ODS reclamation; and, supple- appears to us that such auctions are also menting the Fund's resources. We recommend administratively simpler and less costly, and that the Fund provide incentives to govern- may also have greater political acceptability, ments to engage in these efforts, perhaps as a Environmental Economics Series 41 Market Based Instruments for the Inplementation of Montreal Protocol in the Developing Countries part of a broader policy dialogue between the (iii) the potential for recovery, recycling, and Fund and the developing country governments reclamation activities worldwide to avoid via the Implementing Agencies. longer-term supply constraints of ODS (for servicing enduse equipment) and to reduce Further investigations are warranted their atmospheric emissions; and (iv) the on the following topics: (i) global availability potential for accelerated conversions in the and price prospects for virgin Protocol Annex A "production sector". ODS and their substitutes; (ii) design and implementation alternatives for national and inter-national tradeable permits and pricing regimes for virgin ODS and their substitutes; 1/ Since the initial draft of this paper, the MFEC has approved an auction plan for Chile (19th Meeting, May 1996) 42 Environment Department Papers Bibliography Ahmed, Kulsum (1995). Technological Development and Pollution Abatement: A study of how enterprises are finding alternatives to chlorofluorocarbons. Technical Paper No. 271, Washington, DC: The World Bank. Barthold, Thomas A. (1994), "Issues in the Design of Environmental Excise Taxes," Journal of Economic Perspectives, 8:1, Winter 1994, p. 133-151. Baxter, Les (1995). Different Approaches to Estimating Transition Costs in the Electric Utility Industry. ORNL/ CON-423. Oak Ridge, TN: Oak Ridge National Laboratory. Bojkov, Rumen D. (1995). The Changing Ozone Layer. A joint publication of the World Meteorogical Organi- zation and UNEP. Cook, Elizabeth (1996). "Marking a Milestone in Ozone Protection: Learning from the CFC Phase-out", WRI Issues and Ideas, January. Washington, DC: World Resources Institute. COWkonsult in association with Goss Gilroy Inc. (1995) Study on the Financial Mechanism of the Montreal Protocol, draft final report, March 1995. Nairobi: Ozone Secretariat, UNEP. European Communities (1994). "Council Regulation (EC) No. 3093/94 of 15 December 1994 on substances that deplete the ozone layer," Official Journal of the European Communities, L333, vol.37,22 December 1994 (English edition). Fay, Kevin (1995). Testimony before (US) House Committee on Science, Subcommittee on Energy and Envi- ronment, September 20,1995. Arlington, VA: Alliance for Responsible Atmospheric Policy. Fernando, Chitru S., Paul R. Kleindorfer, and Mohan Munasinghe (1993). "Economic Design for Implement- ing the Montreal Protocol: Country Plans and Global Efficiency", (mimeo). Pollution and Environmental Economics Division, Environment Department Washington, DC: The World Bank. Greenpeace International (1995). The Ozone Layer Destroyers: Whose Chlorine and Bromine Is It? Washington, DC. Hahn, Robert W. and Carol A. May (1994). "The Behavior of Allowance Markets: Theory and Evidence", mimeo. Cambridge, MA: Center for Science and International Affairs, John F. Kennedy School of Govern- ment, Harvard University. Hays, Sam (1995). "Emissions Trading Mythology," The Environmental Forum, January/February, pp. 15-20. ICF et al. (ICF Incorporated, with Center for Environmental Technologies, Tata Energy Research Institute, and Center for Global Change, 1994). Report under Paragraph 8 of Article 5 of the Montreal Protocol. Submitted to Sub-Committee for the Report, Multilateral Fund for the Implementation of the Montreal Protocol, Montreal, Canada. Hoemer, Andrew (1996). "Taxing Pollution," in Elizabeth Cook, ed. (1996), Ozone Protection in the United States: Elements of Success. Washington, DC: World Resources Institute. Environmental Economics Series 43 Market Based Instruments for the Implementation of the Montreal Protocol in the Developing Countries Kerr, Suzi (1995). Markets versus International Funds for Implementing International Environmental Agreements: Ozone Depletion and the Montreal Protocol. Working Paper 95-12. Cambridge, MA: The Center for Intemational Affairs, Harvard University. King, Kenneth and Mohan Munasinghe (1991). Incremental Costs of Phasing Out Ozone Depleting Sub- stances, Environment Working Paper No. 47. Washington, DC: The World Bank. King, Kenneth and Mohan Munasinghe, (1995), "Country Incremental Costs of Phasing Out Ozone Deplet- ing Substances, in King and Munasinghe, eds (1995). Ozone Layer Protection: Country Economic Costs, Wash- ington, DC: Global Environmental Facility and The World Bank. Kliman, Mel (1994). "Competitive bidding for independent power: Developments in the USA," Energy Policy, January, pp. 41-56. Laffont, Jean-Jacques and Jean Tirole (1993). A Theory of Incentives in Procurement and Regulation. Cambridge, MA: MIT Press. Lee, David (1996). "Trading Pollution," in Cook, ed., op. cit. Milgrom, Paul (1989). "Auctions and Bidding: A Primer," Journal of Economic Perspectives, 3:3, Summer 1989, p. 3-22. Munasinghe, Mohan and Kenneth King (1991). Issues and Options in Implementing the Montreal Protocol in Developing Countries. Environment Working Paper No. 49. Washington, DC: The Wo*ld Bank. OORG (Ozone Operations Resource Group), (1994). Domestic Refrigeration RefrigerantAlternatives. OORG Refrigeration Working Group Report No.7. Montreal Protocol, Global Environment Coordination Division, Enviroinment Department. Washington, DC: The World Bank. Parson, Edward A. and Owen Greene (1995). "The Complex Chemistry of the International Ozone Agree- ments," Environment, 37:2, March 1995, p.1642. SEI (Stockholm Environment Institute, with UNEP E's OzonAction Programme), (1995). Monitoring Imports of Ozone Depleting Substances: A Study of ODS Monitoring Systems and their Relationship to Enacted Control Systems in ODSONET/SEAP Countries. Supported by MFIMP and Sida. Stockholm. Swezey, Blair (1993). The Impact of Competitive Bidding on the Market Prospects for Renewable Electric Technolo- gies. NREL/TP-462-5479. Golden, CO: National Renewable Energy Laboratory. Takacs, Wendy (1994). "Import License Auctions and Trade Liberalization: Theory and Experience," (mimeo), July 1994 draft. International Trade Division, International Economics Department. Washington, DC: The World Bank. UNEP (United Nations Environment Programme), (1991).1991 Assessment: Report of the Technology and Economic Assessment Panel. UNEP Ozone Secretariat, Nairobi. UNEP (1993). Handbookfor the Montreal Protocol on Substances that Deplete the Ozone Layer. Third edition. UNEP Ozone Secretariat, Nairobi. UNEP (1995a). 1994 Report of the Economic Options Committee (1995 Assessment). UNEP Ozone Secretariat, Nairobi. UNEP (1995b). Multilateral Fundfor the Implementation of the Montreal Protocol: Policies, Procedures, Guidelines and Criteria. (Draft, as at July 1995). UNEP (1995c). Vienna plus ten: The Vienna Convention: 10 years of achievement. UNEP OzonAction Special Supplement No.3. Paris. UNEP (1996, draft). June 1996 Report of the TEAP Task Force on the Multilateral Fund Replenishmentfor 1997-99. The Hague, The Netherlands. 44 Environment Department Papers US GAO (General Accounting Office, 1994). Allowance Trading Offers an Opportunity to Reduce Emissions at Less Cost. GAO/RCED-95-30. Washington, DC. van der Tak, Herman (1991). Policies and Measures to Implement the Montreal Protocol. Environment Working Paper No. 48. Washington, DC: The World Bank. Vincent, Jeffrey (1996). "Environmental Funds in Developing and Transition Economies," in Third Expert Group Meeting on Financial Issues ofAgenda 21, 6-8 February 1996, Manila, the Philippines. New York: United Nations. World Bank (1995). The World Bank Portfolio of Montreal Protocol Investment Projects, March 1995. Global Environment Coordination Division, Environment Department. Washington, DC: The World Bank. Environmental Economics Series 45 Market Based Instruments for the Implementation of the Montreal Protocol ih the Developing Countries 46 Enviromnent Department Papers Annex: Supplemental Tables Annex Supplemental Tables A.1 Fund Budgetary Programming and CETVs A.2 Fund Grants for Approved World Bank Investment Sub-projects A.3 Approved Costs of Domestic Refrigeration Projects, 13th + 15th + 17th MFEC Meetings A.4 US List Prices and Taxes for CFC-11 and CFC-12 A.5 Comparison of Domestic Refrigeration Materials Prices A.6 Prospects for CFC Availability Environmental Economics Series 47 Market Based Instruments for the Implementation of Montreal Protocol in the Developing Countries TABLE A.1: FuND BUDGETARY PROGRAMMING AND CETVs MFEC (8/93) BUDGETARY MFEC (3/95) REQUIREMENTS & TARGETS COST-EFFECTIVENESS THRESHOLD VALUES (CETVs) Secton Avg. $/k Target 1995 Max. $/kg Available ODP (tons) 1995 1995 ODP (tons) Production RAC (Refrigeration/air- conditioning) 2,047 3,705 Domestic $34.80 $13.76 3,470 Commercial $6.50 $15.21 235 Chillers $18.89 NA MAC (Mobile a/c) $11.76 NA Halons Substitu- $4.35 2,478 $1.48 1,200 tion _ $13.91 NA Banking - Aerosols $1.67 3,835 $4.40 3,913 Foams 5,268 4,992 Unclassified -- $9.53 2,590 Flexible pu $7.03 $6.23 424 Rigid $5.69 $7.83 1,372 Integral skin NA $16.86 21 Poly-styrene/urethane $8.00 $8.22 585 Solvents $12.00 1,242 NA 153 CFC-113 $19.73 153 TCA $38.50 0 Notes: 1/ Figures in US$ of different years. 2. For comparison of average grants for all projects (including those by other implementing agencies) and across sectors and country groupings, see COWIconsult (1995). Environment Department Papers Annex: Supplemental Tables TABLE A.2: FUND GRANTS FOR APPROVED WORLD BANK INVESTMENT SUB-PROJECTS SECTOR No. OF TOTAL TOTAL ODP FUND GRANT (US$/KG) PROJECTS GRANTS PHASEOUT2/ WT. AVERAGE RANGE (US$M)V (TONS) Production RAC (Refrigeration/ airl conditioning): (total) 54 $70.7 7,105 $10.0 Not meaningful Domestic 26 $32.2 3,338 $9.6 $0.8-$74.4 Commercial 11 $18.4 1,545 $11.9 $6.9-$33.3 Chillers 1 $1.1 14 $78.6 N.A. MAC (mobile a/c) 8 $13.3 1,183 $11.2 $1.7-47.5 Other 8 $5.7 1,025 $5.6 $3.8-$16.3 Halon 5 $3.7 1,655 $2.3 $1.8-$3.2 Aerosol 10 $9.2 17,711 $0.5 $0.3-$5.0 Foams: 57 $26.1 6,361 $4.1 Not meaningful (total) 24 $6.8 1,467 $4.6 $0.6-$17.2 Flexible 16 $10.2 2,172 $4.7 $1.3-$385.0 Rigid 17 $9.2 2,722 $3.4 $0.1-19.6 Other l Solvent 15 $8.7 764 $11.4 $1.9-$142.0 Recovery/recycling - - - - Total 143 $125.2 34,396 $3.64 Not meaningful (incl. 3 unclassified) Total, excluding China 109 $84.2 12,546 $6.71 Source: The World Bank Portfolio of Montreal Protocol Investment Projects, March 1995. Notes: 1/ In mnixed-year, as-comnmitted US$. 2/ As approved by the MFEC. Environmental Economics Series 49 Market Based Instruments for the Implementation of Montreal Protocol in the Developing Countries TABLE A.3: APPROVED GRANTS FOR DOMESTIC REFRIGERATION PROJECTS13TH+15TH+17TH MFEC MEETINGS Capital Incremental Composite "Unit Country Company ODP grant operating price Averag Abate- abated cost, Year 1 differential grant ment (t) Cost1' (assump tions _____ _____ __ _ _____ ____ _ ___ ____ belo w) ._________ . _____ All in US$ per kg.ODP abated "Representative" projects: _ i Indonesia Topjaya 91 $14.1 $2.8 $12.4 $14.2 $5.1 Colombia Corelsa 22 $19.7 $21.0 $10.8 $30.2 $24.2 Colombia ICASA 72 $9.8 $10.7 $10.8 $15.2 $12.3 Colombia HACEB 69 $4.2 $17.5 $10.8 $12.9 $18.2 Colombia POLARIX 63 $8.0 $20.1 $10.8 $18.1 $21.4 Peru Selva 35 $7.3 $9.1 $8.9 $11.8 $10.3 Peru Coldex 67 $7.1 $8.1 $8.9 $11.1 $9.2 India BPL 165 $17.5 $5.8 $6.5 $20.9 $8.7 Jordan HAMCO 47 $15.3 $2.3 $4.9 $16.5 $4.8 Syria Penguin 82 $17.1 $7.6 $4.8 $20.9 $10.4 Syria El-Hafez 107 $22.5 $8.9 $1.2 $26.9 $12.5 Cameroon FAEM 62 $28.8 $5.0 $1.2 $31.3 $9.7 Sum/simple average 882 $14.3 $9.9 $7.7 $19.2 $12.2 standard deviation $7.0 $6.1 $3.7 $6.7 $5.8 Not included: Combined with compressor: Mexico Mabe (all 3) 427.0 $6.1 $7.4 $6.8 $9.8 $8.4 Mexico Vitro (all 3) 456.0 $3.5 $7.9 $5.4 $7.5 $8.5 With foreign ownership Indonesia NABEL 110 $13.0 $4.0 $0.0 $14.2 $6.1 Turkey PEKEL 160 $3.8 $1.3 $0.0 $4.6 $1.9 Indonesia Sharp Yasonta 173 $9.7 $4.7 $0.0 $12.0 $6.3 Malaysia Oyl 22.0 $26.6 $3.7 NA $28.4 $8.0 No material price data Egypt Delta 97 $18.3 $17.3 $26.9 $20.2 Egypt Electrostar 46 $29.1 $5.2 $31.7 $9.9 Egypt Kiriazi 137 $8.7 $6.6 $12.0 $8.0 Environment Department Papers Annex: Supplemental Tables TABLE A.3: APPROVED GRANTS FOR DOMESTIC REFRIGERATION PROJECTS13TH+15TH+17TH MFEC MEETINGS Capital Incremental Composite "Unit Country Company ODP grant operating price Averag Abate- abated cost, Year I differential grant ment (t) Cost" tions __________ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ ____ ___ ______ _________ __________ below ) Not included (continued) "Outliers" Argentina NEBA 29 $24.9 $12.7 $14.3 $25.6 $16.7 Argentina Briket 30 $24.9 $14.2 $14.3 $25.7 $18.3 Argentina Piragua 61 $20.2 $11.5 $14.3 $20.8 $14.8 Argentina McLean 74 $15.1 $34.9 $13.0 $33.0 $37.3 Argentina Helametal 62 $32.9 $27.7 $13.0 $47.7 $33.1 Argentina Fribe La Rioja 37 $31.9 $32.8 $11.8 $49.3 $38.0 Argentina Fribe La Rioja 26 $40.9 $30.2 $11.8 $57.2 $36.9 Argentina Autosal 22 $42.9 $4.2 $8.9 $44.2 $11.2 Argentina Aurora 53 $36.0 $3.0 $8.9 $37.0 $8.9 Peru Andina 5 $39.6 $9.2 $8.9 $44.2 $15.6 Jordan NRC 14 $53.0 $7.8 $4.9 $56.9 $16.5 Jordan MEEI 17 $48.3 $9.7 $4.9 $53.2 $17.6 Syria Barada 109 $6.9 $4.4 $2.9 $9.1 $5.5 Peru Lenche 6 $29.5 $11.7 $8.9 $35.3 $16.5 Peru Alfa 7 $32.6 $9.0 $8.9 $37.2 $14.4 Source: Authors' calculations based on subproject documents. Notes: 1. Each project is to use HFC-134a as the CFC-12 substitute for refrigerant; substitutes for foaming agent CFC-11 vary (HCFC-141b or hydrocarbons). 2. "Composite price difference" refers to controlled ODS and the substitute for refrigerant as well as foaming agent use. 3. Average grant and Unit Abatement Cost (UAC) calculated before implementing agency costs. 4. UAC calculated for 10 years project life and 10% discount rate, assuming no change in IOC over time. Environmental Economics Series 51 Market Based Instruments for the Implementation of Montreal Protocol in the Developing Countries 199 M.29 , 2.75 0 0ii 1990 2.1.7 $2.89 $30301 I~~~~~~~~~~~~~~~~~ 1989 N.A. 2.A1 0 0 1990 2.41 2.89 $3.01 $3.01 1991 2.35 2.62 3.01 0.00 1992 2.90 4.16 3.67 0.66 1993 3.75 5.63 7.37 3.70 1994 N.A. N.A. 9.57 2.20 1995 N.A. N.A. 11.77 2.20 1996 12.76 N.A. Sources: List prices obtained from Professor Stephen DeCanio (personal communication, July 1995), in turn based on monthly data from Chemical Marketing Reporter. Retail prices are reported to be considerably higher than wholesale list prices. Tax rates from Chapter 7 of UNEP (1995a) and US Internal Revenue Code, 20 USC 4681. TABLE A.5: COMPARISON OF DOMESTIC REFRIGERATION MATERIALS PRICES ($/KG) Source/date/place CFC-12 HFC-134a CFC-11 HCFC-141b ICF, et al. (1994) - Generic $2.50 $7.35 $2.15 $4.63 World Bank internal data, $4.32 $13.17 $3.53 $4.78 August 1995 - Brazil 52 Envirorunent Departmnent Papers Annex: Supplemental Tables Potential production (thousand tons) Annual Cumulative (1995-2010) Developing countries - 1993 rate 78.3 896.7 Developing countries - maximrum capacity 149.4 1,755.3| Developed countries - allowed production 151.8 2,277.2 Substance End-use Stock in use Productionl CFC-11 Chillers 61 25 CFC-12 MAC (Mobile a/c) 250 Domestic refrigeration 150 Commercial refrigeration 31 total 431 155 Sources: For (A), UNEP/OzL.Pro/ExCom/15/43 (Limited distribution) Draft. Capacity maybe used for CFC-11 or CFC-12 (and other CFCs). Also see Greenpeace (1995). For (B), Chapter 7 in UNEP (1995a). Developing countries' production rate may be constrained by the baseline consumption levels of 1995 to 1997 except to the extent of exports to other developing countries. Environmental Economics Series 53 Market Based Instruments for the Implementation of Montreal Protocol in the Developing Countries 54 Environment Department Papers Environment.Department The World Bank 1818 H Street, NW. Washington, D.C. 20433 202 473 36441 202 477 0565 FX :Printed on 100% post-consumer recycled paper