THE WORLD BANK
SECTOR POLICY AND RESEARCH STAFF
Environment Department
Policies and Measures
to Implement the Montreal Protocol
Herman G. van der Tak
September 1991
Environment Working Paper No. 48
This paper has been prepared for internal use. The views and interpretations herein are those of
the author(s) and should not be attributed to the World Bank, to its affiliated organizations or to
any individual acting on their behalf.


THE WORLD BANK
SECTOR POLICY AND RESEARCH STAFF
Environment Department
Policies and Measures
to Implement the Montreal Protocol
Herman G. van der Tak
September, 1991
Environment Working Paper No. 48
This paper has been prepared for internal use. The views and interpretations herein are those
of the author(s) and should not be attributed to the World Bank, to its affiliated oranizations
or to any individual acting on their behalf


ACKNOWLEDGEMENT
The paper has been prepared by Herman G. van der Tak, a Consultant to the Environment
Department. The author is indebted to Ken Newcombe, Mohan Munasinghe, Ken King, and Sherif
Arif for helpful comments. The paper reports on findings that are particularly significant in the
context of the overall work program of the Environmental Policy and Research Division.
Departmental Working Papers are not formal publications of the World Bank. They present
rough and tentative results of country and sector analysis or tesearch, and are circulated to encourage
discussion and comment. Citation and use of such a paper should, therefore, take account of its
provisional character. The findings, interpretations and conclusions expressed in this paper are
entirely those of the author and should not be attributed in any manier to the World Bank, to its
affiliated organizations, or to members of its Board of Executive Directors or the countries they
represent.
Because of the informal nature of this paper, and in order to present the results of research
with the least possible delay, the typescript has not been prepared in accordance with the procedures
appropriate to formal printed texts, and the World Bank accepts no responsibility for errors.
i


ABSTRACT
The establishment of the Interim Multilateral Fund of the Montreal Protocol reflects the
judgment that substantial transfers of resources to developing countries (to compensate them at least
in part for the cost of the phase out of ozone depleting substances) would strengthen adherence to
the Protocol, ease adjustment to the phase out, and share the burden of the phase out more
equitably.
The author takes as given the judgment that the benefits of the phase out outweigh the costs,
the requirements of the Protocol, and the rationale for the Fund. He considers two sets of issues
which emerge at this point: first, what policies are the most suitable ones for countries to follow to
meet their obligations under the Protocol, and second, how such a Fund can be operated most
effectively to support efficient and equitable implementation of the Protocol. The author examines
these issues from a broad economic perspective, and is not bound by the actual mechanisms which
are currently being used by the Fund.
With respect to national policies, the author concludes that countries should preferably
control the supply of ozone-depleting substances rather than the consumption, leaving the market
free to allocate the increasingly scarce supplies among the various users. He makes a number of
specific recommendations for the controls that can and should be used. With respect to fund
strategy, he concludes that the resources available should be allocated in accordance to the relative
cost of adjustment, but with a heavy bias in favor of the poorest countries. Finance should be
conditional upon the adoption of appropriate country policies and should be made on a program
basis, rather than project by project.
ii


POLICIES AND MEASURES TO IMPLEMENT THE MONTREAL PROTOCOL
TABLE OF CONTENTS
ACKNOWLEDGENT ........................................ i
ABSTRACT    .................................................  ii
I Nation altzone1Policies  .............................................
Cost-effectiveprograms  .........................................
Policyr kstrume   .....................................  .......  3
Basicp olicypapp ach  ...............................................  3
Regulation of supplies with market
alocadon amonguers .........................................  3
Allocation among suppliers.......................................  4
W indfallpr>fits  ...............................................  4
Tunephasingpriortotargetyears  .................................  4
M arketd torons  .............................................  5
Stoc kping  ...................................................  5
Indirect timpor  ...............................................  7
Alternatie polpappr7aches  .........................................  7
Poluti on tax..................................................  7
Mar ketbleseprmit  .........................................  8
Subsidies  ....................................................  9
Reyl ingadspidspo   ...............................................11
Recycling  ....................................................11
D isposa  .....................................................12
onclusns  ....................................................... 14
II International Policies: The Interim Multilateral Fund of the Montreal Protocol ... 15
Internationaloabatementpolies ...................................15
Inte rnationalasppportu  ......................................16
Country alloca ons.............................................17
Promotion of cost-effective abatement ..............................18
Categories of financing  .......................................... 19
Pro jectsstandar  .............................................. 21
Condit ona y.................................................21
Financig er2s ...............................................22
Institutionalss  .............................................23
Conclusions  ......................  ............................... 25
llLinkstoExistingFrameworkfor Fund Operations ........................25
iii


POLICIES AND MEASURES TO IMPLEMENT THE MONTREAL PROTOCOL
1.    The Montreal Protocol, as subsequently strengthened in London, requires its signatories to
reduce sharply or phase out altogether the production and consumption of a wide range of
chlorofluorocarbons (CFCs) and other ozone-depleting substances (ODSs) in accordance with a
specified timetable. This note does not discuss the merits of the Protocol. The signatories
presumably considered that the benefits of a rapid phase-out of ODS would outweigh its costs. While
the magnitude of the benefits (globally and for individual countries) is uncertain, this judgment is
supported by the apparently low costs involved. For developing countries these costs are to be
reduced further by contributions from an international Interim Multilateral Fund of the Montreal
Protocol. Again, this note does not examine the rationale for such a Fund. Its establishment reflects
a judgment that substantial transfers to the poorer countries to compensate them to some extent at
least for the cost of the mandated phase-out would increase adherence to the Protocol, ease
adjustment to the phase-out, and share its burden more equitably.
2.    Given the requirements of the Protocol and the existence of a compensatory fund, this note
addresses two sets of issues. It considers, first, what policies are most suitable for countries to follow
in order to meet their obligations under the Protocol. And second, how such a Fund can be operated
most effectively to support an efficient and equitable implementation of the Montreal Protocol.
L     National Ozone Policies
3.    Cost-effective programs. It is desirable that national programs to phase out ODS be cost-
effective; that is, that countries try to minimize the economic costs of any particular reduction in ODS
emissions. This applies to the ODS phase-out envisaged under the Montreal Protocol, as well as any
other more rapid phase-out that countries may choose to adopt. (Equivalently, one might try to
maximize the reduction in ODS emissions for any given level of adjustment costs). Comparison of
alternative phase-out programs is not straightforward. Various ODS differ in their ozone depleting
potential (ODP). Equivalent reductions in ODS emissions should therefore be measured in terms
of ODP, not simply in terms of tons of ODS. Earlier reductions in emissions of ODS are more
valuable than later ones, quite apart from the usual discounting problems. Furthermore, ODS differ
not only in their ODP, but also in their global warming potential. Since it is difficult to make
plausible allowances for some of these factors, comparison of alternative programs is likely to be
rough and ready at best.


4.     However, the principle is clear: each country has a variety of possible ODS phase-out
scenarios for meeting its requirements under the Montreal Protocol. These scenarios have different
economic costs, including not only the cost of conversion to ozone-friendly processes by
manufacturing firms currently using ODS, but also the cost of technical assistance, research and
development, recycling, disposal, training, and preparation of phase-out programs and policies. All
costs are assessed in economic terms, corresponding to the common practices in cost-benefit analysis,
and the different cost streams are compared in terms of present value. Costs of a program may be
higher because the costs of its substitutes and other phase-out measures are inherently greater, or
because it accelerates phase-out activities. Cost differences should be adjusted, qualitatively if not
quantitatively, for any differences in benefits associated with alternative programs. These may include
differences in the timing of the ODS phase-out, as well as in the quality of substitute products.
Subject to that important qualification, the least-cost alternative should be selected as the most cost-
effective phase-out program (see also below, on country allocations and incremental economic costs.)
5.     Few country studies have been made thus far to design such cost-effective programs for the
phase-out of ODS, or at least to determine their broad outline and major features. But many
building blocks for such studies exist, indicating the available substitutes and their costs. Some of the
alternatives, such as the use of LPG propellants in aerosols, appear to be more efficient and reduce
costs, even without taking into account the benefits of reducing ODS emissions. Others, such as the
use of new thermodynamic cycles for refrigeration that do not depend on CFCs, are expected to
increase costs. Some are commercially available now, while others are expected to become available
in the next few years, or in the more distant fiuture.
6.     What does this mean for the design of a cost-minimizing national ODS reduction program?
One would expect it to start with those investments that are profitable to the country, i.e., produce
net cost savings, in any case; and to postpone ODS reducing investments that involve a substantial
increase in net costs to the country as long as possible. Somewhere in between are investments that
are cost-saving, but not sufficiently so to outweigh the additional cost of accelerated replacement at
this time; they will become economical after a few years. This broad picture is somewhat muddled
by the uncertainty about future alternatives and their costs. This suggests that it is a good idea to
keep your options open, that is, to start with short-lived ODS reducing investments, and to delay
longer investments until later when the choices have become clearer. To the extent that there is
2


some correlation between the cheap/easy and short-lived substitutes on the one hand, and between
the expensive and long-lived investments on the other, the uncertainty factor supports the earlier
conclusion.
7.     Policy instruments. What policies are most effective in implementing such programs for ODS
reduction? Preferably such policies should be cost-effective, easy to design and administer, and to
monitor compliance, be equitable in their impact, adapt themselves easily to uhanging circumstances,
and promote technological progress and further cost savings. The standard policy instruments for the
control of pollution are regulation ("command-and-control"), pollution taxes, marketable permits, and
deposit/refund schemes. These approaches are not mutually exclusive, but are typically used in
various combinations, depending on circumstances and country preferences. The remainder of this
section considers what policy packages appear most promising for the control of ODS pollution.
Basic policy approach
8.     Regulation of supplies with market allocation among users. A useful starting point is to note
that there are usually few ODS suppliers (manufacturers or importers of CFC's, halons, etc.),
especially in the developing countries, which in many cases rely (nearly) exclusively on imports. But
there are many users, who use ODS in the manufacture of refrigerants, foam products and aerosol
products, as cleaning solvents, and in fire extinguishers. It is much more convenient therefore to
regulate the suppliers than the users of ODS.
9.     This suggests the following approach: limit total ODS supplies (production plus imports) to
the desired amounts, such as the amounts permitted under the Montreal Protocol for the various
target years. The supply constraint will tend to raise the prices of ODS. The price increases will
induce substitution: those users who can substitute at low costs will substitute first when the price
ruses; those with higher substitution costs will continue to use ODS until the price increase exceeds
the costs of substitution. Thus the rationing by price of increasingly scarce supplies ensures a cost-
effective adaptation by the users. Note that this is achieved by direc% controls over the ODS
suppliers, not the ODS users. It would be difficult to duplicate the cost-effective result by direct
regulation of the users: it would require detailed knowledge of the substitution costs of each of many
users. That is clearly not a feasible solution. Note also that this approach relies neither completely
3


on command-and-control, nor on market mechanisms. Instead it uses direct regulation of ODS
suppliers, in combination with a market-based allocation of limited supplies among ODS users.
10. Allocation among suppliers. While this approach appears a promising starting point, in terms
of cost-effectiveness, easy administration, and sure control over the ODS phase-out, there are some
problems that need to be addressed. First, ODS suppliers may be few in a country, but there is often
more than one. That raises the question how the reduction in supplies should be allocated among
them. If there is more than one supplier of any individual substance (otherwise there is obviously
no problem), the country should try to reduce the supplies from the highest-cost supplier, and so on.
But how? It is difficult to determine the relative costs of various suppliers, especially with
overcapacity and foreign ownership being common. With few firms, competition is likely to be
limited. Competitive bids for the permitted supplies are likely to suffer from collusion. Perhaps the
best one can do ib *egin with reducing imports, and subsequently reduce local suppliers
proportionately. Or alternatively, try to reach a negotiated solution.
11.   Windfall profits. Second, the price increases resulting from the controls over the supplies of
ODS are likely to give the suppliers (producers and/or importers) a windfall profit, at least until the
phase-out is completed. That does not appear desirable. Why should ODS suppliers profit from the
environmental controls? The profit might be creamed off by a windfall profit tax. Alternatively, the
government might impose a compensating tax on ODS production and imports. Note that in this case
the production/import tax on ODS is supplementary to, and not a substitute for, direct regulation of
ODS supplies. (The possibility of an ODS pollution tax as a substitute for direct regulation is
considered further below.)
12. Time phasing prior to target years. Third, the Montreal Protocol specifies the required
reductions in the production and use of ODS, and their subsequent total phase-out, for particular
target years. That determines the necessary controls over ODS supplies in those target years. But
what about the supply controls in the intervening years leading up to those target years? The
Protocol is silent on this point: it does not impose any requirements for the interim periods. What
then is a sensible approach? One possibility is to do nothing until the target year. This postpones
the adjustment costs as long as possible. It also keeps the options open for choosing the most cost-
effective substitutes in a rapidly evolving field. But the sudden large price increases in the target
4


years, when the controls on ODS are sharply tightened, and the resulting uncertainty about which
user will be induced or forced to 3witch to substitutes are likely to add to the adjustment costs.
Perhaps more important, postponement of action until the target date also minimizes the abatement
of ODS emissions. The intention of the gradual phase-out under the Protocol is to mitigate the
adjustment costs, and not to reduce ODS abatement.
13.   Some abatement is possible now at little if .ny costs. That should be encouraged. This
suggests that supply controls should be gradually tightened, so that prices of ODS would gradually
increase, during the period leading up to the target years. That would promote the easier
substitutions; keep the options open for those uses where switches are more difficult and costly at
present; and increase ODS abatement. Such a gradual approach appears preferable to a "wait-until-
the-target year" approach. It would probably reduce overall adjustment costs as well as ODS
emissions. Beyond some point, however, further acceleration of the phase-out would involve a trade-
off between the costs and benefits of ODS abatement.
14.   Market distortions. Fourth, the broad approach advocated above assumes that the market
will allocate restricted supplies of ODS among users and induce switches to substitutes in a cost-
effective manner: rising ODS prices will cause substitution by those users who can switch most
cheaply. While this is a reasonable presumption, it does not necessarily assure a cost-effective
outcome to the country. Major distortions in the ODS user market may affect the outcome and
negate its economic cost-effectiveness. For example, taxes and/or subsidies may affect the relative
cost of using ODS and substitutes for various users. As a result switches may be induced by a user
whose economic costs of substitution are higher than those of another user. (Financial costs continue
to be minimized, but not the economic costs to the country.) Should this approach therefore be
abandoned? That would be an overreaction. The alternative of allocation by fiat is most likely to
be worse: to be c -st-effective, it would need a great deal of cost information on the use of ODS and
its substitutes by many users. It would appear preferable to be alert to market distortions and take
corrective action as necessary in case of major distortions in ODS user markets. (Minor distortions
can probably be safely ignored.)
15.   Stockpilinu. Fifth, the control of new supplies of ODS (from production and/or imports) does
not ensure that the use of ODS is reduced accordingly. There are several possible loopholes or
5


leakages. They include (i) stockpiling of ODS by users; (ii) "indirect" imports of ODS incorporated
in intermediate products, such as refrigerants, foam products and solvents, or in final products such
as refrigerators, furniture and aerosol products; and (iii) recycled supplies. The first two are
considered here. Recycling issues are discussed later.
16.   Some ODS users may want to stockpile ODS as insurance against (expected) high substitution
costs. Since the total available (new) supplies are controlled, this will drive up the price of ODS and
result in greater abatement by other users. Total use of ODS during net stockpiling periods will be
less than it otherwise would have been, and greater later when the stockpiles are reduced. Is
stockpiling of ODS to be avoided? Its negative effects, if any, appear to be minor. Stockpiling does
not affect total use of ODS; it merely shifts ODS use forward to the future when the stock is being
used. The immediate reduction in ODS use when the stockpiling takes place is presumably welcome.
The subsequent increase in ODS use when the stock is reduced could be compensated (roughly) by
tightening the controls over (new) supplies accordingly. Similarly, if stockpiling takes place in
anticipation of ODS supply controls, as is likely, this can be offset by an adjustment in the subsequent
controls over new supplies, so as to keep the total volume of prior stocks and new supplies to the
desired level.
17.   Such an offset is obviously not feasible if at that time all (new) supplies have been phased out.
That is, when supplies of ODS are controlled, but use of ODS pse is not, stockpiling undermines
perfect compliance with a requirement for a complete phase-out of both production and use of ODS
by a certain target year. However, this appears to be a minor, and temporary, flaw of approaches that
control supplies rather than uses of ODS. If it is considered important that all use of an ODS is
strictly phased out by the same year as (new) supplies, one could supplement supply controls with a
flat prohibition of all uses by that time. Users would be put on notice that all stockpiles of the ODS
concerned would have to be used prior to that date. If not, they would have to be disposed of
through proper channels (see below). That would still avoid the administrative problems, rigidities,
and higher abatement costs that would be inherent in general reliance on control of ODS uses, rather
than control of supplies as advocated in this paper, during the whole phase-out period. Enforcement
of compliance by a few stockpilers with a strict ban on ODS use after the phase-out date, backed up
with heavy penalties, should be easier than the design and monitoring of a wide range of controls on
a large number of users during a long adjustment period.
6


1&    IJiret ionggrts. There is another "loophole," as noted above. Control of the production and
direct imports of ODS does not preclude "indirect" imports of ODS incorporated in intermediate
products such as refrigerants and foam products, or in final products such as refrigerators or vehicles.
It is not clear whether such indirect ODS imports are in fact significant. Intermediate products are
likely to be more important in this regard than final products. And sources of imports of ODS
containing products will presumably dwindle if ODS controls imposed by the Montreal Protocol begin
to bite overseas. However, if this loophole is considered serious, it can be closed in several ways.
The controlled supply of ODS itself might be (roughly) adjusted downward to allow for indirect ODS
imports. Or imports of some major offending (i.e., ODS containing) products might be frozen,
curtailed and ultimately banned altogether. While the possibility of "indirect" ODS imports adds
somewhat to the complexity of an effective ODS abatement policy, the few supplementary measures
that they may require do not greatly affect the advantages of the basic approach supported here,
which primarily relies on the control of the supply of the various ODS themselves.
Alternative policy approaches
19.   Pollution tax. The policy outlined above is, of course, not the only possible policy approach
to the abatement of ODS pollution, as required under the Montreal Protocol. Earlier we noted that
desirability of a tax on the production and imports of ODS to eliminate excess profits for suppliers
when ODS supplies are controlled, and their prices rise, to protect the ozone layer. If such a tax
were set just right, it would limit supplies to the desired level. It would also allocate the permitted
supplies to the lowest-cost suppliers: for higher-cost suppliers it would no longer be profitable to
supply, given the tax (subject to the usual qualifications wi!h respect to market distortions). This is,
of course, the pollution tax solution. It obviates the need for direct control of suppL. -% is well as
users. Given the tax, the market tends to come up with the cost-effective solution for tb . allocation
of restricted supplies among both the suppliers and the users. (The basic approach discussed earlier
is a hybrid in this regard, with direct control of the suppliers, but a market solution for the allocation
among the users.)
20.   The trouble with the pollution tax approach is that it is impossible, in practice, to determine
the right level of the tax except through trial and error. Even a rough guess is difficult, since both
the ODS cost price and the demand price, at the supply level to be arrived at for ODS abatement
reasons, are unknown beforehand. When a tax is picked, one cannot be certain that the desired
7


abatement level will be reached: it may be higher or lower. On the other hand, one is certain of the
maximum (unit) abatement cost, since each firm will abate up to the point where the unit cost of
abatement equals the tax. Conversely, with direct control over supply of ODS, as in the basic
approach discussed above, one is certain of the overall ODS abatement level, but not of the unit cost
of abatement, which is set by the market (scarcity) price of ODS.
21.   In the ODS case, where abatement is considered important (that is, benefits are judged to
be high) and the costs relatively small, it appears better to get a firm grip on the level of abatement
than the cost That suggests that in the ODS case the direct control of the abatement level (that is
the supply of ODS), as in the basic approach discussed earlier, is preferable to the pollution tax
approach. Furthermore, trial and error changes in the tax may make it difficult for firms to plan for
cost-effective abatement measures. However the cost of this uncertainty is probably minor in the
ODS case, assuming that taxes are raised gradually during the interim period leading up to the target
years (see earlier discussion of what to do during those interim years). In that case any lagging
behind or running ahead of the desired interim abatement levels can be dealt with by speeding up
or slowing down the tax increases in the subsequent years. Up and down adjustment levels in the tax
level can be avoided.'
22.    Marketable use permits. A third approach to the ODS abatement required under the
Montreal Protocol is through marketable use permits. Total authorized uses of the various ODS are
controlled in this case. Ignoring recycling complications for the time being (they will be considered
later), this would be equivalent to the control of supplies in the first option discussed above. Given
any initial allocation of the permits among the users, trading in the permits will result in a cost-
effective abatement program: each user will sell permits as long as the price of the permit is higher
than his abatement costs; and buy permits when the abatement costs exceed the permit price. This
results in a cost-effective solution (with the usual qualifications), which corresponds exactly to the
market allocation of the limited supply of ODS in the basic (hybrid) approach discussed above.
' This problem of the interim years, as well as those of market distortions, stockpiling, and
"indirect" ODS imports discussed earlier, are common to both the "basic" and pollution tax
approaches, and the marketable permits approach to be considered below.
8


23.    However, the market in permits may not work as smoothly as the market in the ODS
themselves. And the marketable permit scheme is certainly more complicated, since it superimposes
a market for permits on the market for the ODS themselves. Probably most important, enforcement
problems are much more formidable with controls over many users than over only a few suppliers.
While equivalent in principle, from an administrative point of view control over supplies together with
a market-based allocation of the limited supplies among the users, as in the basic approach outlined
above, appears to have substantial advantages over the marketable permit scheme for the abatement
of ODS.
24.    As noted above, control over ODS supplies together with a free market for ODS users results
in windfall profits for the suppliers. With a marketable use permit system, if the initial allocation of
the permits is by flat and free of charge, equivalent windfalls accrue to the initial recipients of the
permits (whether or not they turn out to be net buyers or sellers of the permits). Such initial
allocation is likely to be (roughly) based on previous ODS usage, i.e., previous ODS pollution. That
would seem an unfortunate reward of the polluters. (It might also lead to an undesirable surge in
ODS usage in order to establish a profitable base). Any windfall profits of suppliers can be
eliminated, as discussed earlier, by imposing a proper pollution tax on them. With marketable use
permits, the equivalent result can be achieved by auctioning of the use permits.' The price of the
permits corresponds to the tax. If a marketable permit scheme is adopted, initial allocation by auction
is far preferable to allocation by fiat, free of charge. Any desirable subsidies can be considered
separately, on their merits.
25.    Subsidies. If any of the above policies are followed, subsidies are unlikely to be necessary to
provide incentives to ODS users to switch to substitutes. The higher cost of ODS use resulting from
price increases when ODS supplies are restricted, ODS pollution taxes and/or the price paid for use
permits, may be expected to give ample incentives for switching away from ODS. As noted earlier,
exceptions may arise when major market distortions intervene. Even in that case, however, it is
usually preferable to get rid of the distortions rather than compensate for them with special subsidies.
Bids on supply permits were rejected above because of limited competition and the risk of
collusion among very few suppliers. That is unlikely to be a problem with use permits.
9


26.    Nor are there any obvious equity reasons for subsidies: the polluter pays is a good general
principle. Furthermore, the cost of switching is often relatively small, and most of the burden of
adjustment is likely to be passed on to final consumers in the form of higher prices. At any rate,
many of the losers will not be in the lower-income groups. It would appear advisable to limit
subsidies/transfers to demonstrated hardship cases, if any.3
27.    All the policies considered above rely on increasing the cost of ODS use to provide the
necessary incentive for switching to substitutes. In that case subsidies play at best a secondary,
supporting role. However, it is also possible to provide the switching incentives by subsidizing the
substitutes, rather than "taxing" the ODSs. A subsidy on ODS use foregone would have similar
effects on substitution as a corresponding "tax" on ODS. But the combined ODS/substitutes industry
would tend to be larger, in the absence of rising ODS/substitute prices and hence of decreasing
demand. To that extent, the subsidy is not efficient. As with the pollution tax, the "right" level of
the subsidy required can only be determined by trial and error. The equity effects of such subsidies
are dubious: the polluters clearly do not pay: the burden falls on the general taxpayer. Perhaps most
troubling, implementation of the subsidy requires detailed information for each firm on its initial use
of ODS and subsequent reductions. Subsidies as the primary instrument for driving the phase-out
do not appear to be attractive or feasible."
28.    On balance, it may be better for the government, to use available funds (for example, from
ODS pollution or windfall taxes, auctioning of marketable use permits, and the Interim Multilateral
Fund of the Montreal Protocol) to support overhead facilities, such as research and development of
substitutes, and information and training, and to subsidize recycling and disposal, as discussed below,
and pilot and demonstration projects.
IConversely, as noted earlier, there are good equity reasons to impose offsetting taxes of
suppliers of ODS, the major (temporary) gainers from the phase-out.
A general flat percentage subsidy on the cost of all substitutes is another possibility, with similar
drawbacks.
10


Recycling and disposal
29.   Recycling. This brings us to the role of ODS recycling. It is clearly desirable from an ODS
abatement point of view. As a minimum, it will postpone ODS emissions. To the extent that
recycled ODS/substitute for the use of new supplies, they reduce the total volume of ODS emissions
over time. In addition, the cost of recycled ODS may be less than new supplies. Even if not,
recycling is likely often to be cheaper than alternative ODS abatement measures.
30.   If ODS abatement is to be maximized, recycled supplies should be included in the overall
supply limits established under the basic policy approach outlined earlier.' Recycling then reduces
new supplies of ODS, and thus ultimate emissions, without imposing any additional burden on users.
Recycling competes in this case with new supplies. When the cost of recycled ODS supplies is less
than the production/import costs of new supplies, recycled supplies will substitute for new ones. The
excess profits tax should extend to all supplies, including the recycled supplies, but its love. sht uld
remain geared to the profits on new supplies so as not to eliminate the incentive to recyc;e.
However, if ODS recycling costs are higher than for new supplies, the net tax on recycled supplies
only would have to be adjusted downward to provide an adequate incentive for recycling. This can
probably be done most easily (and transparently) by maintaining the general tax on all ODS supplies
and providing a separate subsidy for recycling.
31.   Marketable use permits, and pollution taxes, raise no new issues in this regard. The results
are the same as above, when use of recycled ODS supplies is taken into account in setting the total
of the permits; and similarly, when the pollution tax is adjusted accordingly. Again, a (gross) subsidy
on recycling is necessary if the cost of recycling exceeds that of new ODS supplies.
32.   However, it may reasonably be argued that the ODS phase-out schedule, and the
corresponding supply limits, should be fixed for new ODS supplies only, in the light of the additional
emissioti that are considered tolerable. Since recycled ODS supplies do not add to the ultimate ODS
emissions, they are then considered additional to the ODS supply limits. They permit &- increase in
ODS use beyond that provided by new supplies. In fact, if recycling were free of leakage, ODS use
' This applies to recycled ODS from supplies both before and after imposition of the supply
controls. Both will help reduce total supplies of ODS over time.
11


could continue for ever even after new ODS supplies had been phased out completely. On this
interpretation, which is apparently the intention of the Protocol, recycling only postpones ODS
emissions; it does not serve to reduce the total emissions over time.
33.    Nevertheless, the benefits from recycling are likely to remain substantialh some ODS emissions
are postponed, and there are likely to be substantial cost savings. Recycling in this case competes
with further abatement (rather than new supplies), and recycling costs are likely to be less than the
incremental costs of further ODS abatement. (All cheap abatement will already have been done, as
indicated by increasing ODS prices/pollution taxes when supplies are tightened.) Recycling subsidies
are then unnecessary. If recycling costs were to exceed incremental abatement costs, recycling would
require subsidies. Whether they are desirable in this case is another matter. Subsidies to postpone
ODS emissions (with recycling outside the supply limits) are less persuasive than subsidies to reduce
total emissions (with recycling within the supply limits). Basically, recycling within the limit results
in lower emissions and higher abatement costs than recycling outside the limits.'
34.    Whether or not recycled supplies are included in the supply limits is clearly irrelevant for
recycled supplies becoming available after all (new) supplies have been phased out. This is similar
to the stockpiling complication discussed earlier. Again, if this is considered a serious loophole, it
can be dealt with by a flat prohibition of all ODS use at that time.
35.    Disposal. Finally, there is the problem of proper disposal of ODS containing products, such
as refrigerators, foam products, and solvents. The ozone literature tends to assume that all ODS,
once produced, will ultimately end up in the atmosphere. This is apparently not quite tru -: proper
disposal can make a difference. What can be done to encourage proper disposal? The government
can offer to pay for proper disposal. The price to be paid should cover the actual disposal cost, the
collection and transport cost to the disposal site, and an inducement to the owner of the ODS
containing products to dispose of the discards through the proper channel, minus the costs, of any,
to the owner of improper disposal. The role of the government would depend on the response from
' Note also that with recycling "outside the supply limit" ODS prices will be determined by the
recycling costs, and not by the scarcity imposed by the supply limits (and corresponding abatement
costs). Recycling outside the supply limit may be expected to lower the ODS price, and hence the
excess profits of suppliers within the limit.
1e


the private sector. As a minimum it will have to set standards for acceptable disposal. But it may
also have to help finance establishment of disposal plants, regulate monopolistic behavior (probably
more likely at the processing end than the collection end of the disposal business), or take on (part
of) the operation in the public sector.
36.   The price the government is willing to pay for the proper disposal of ODS containing goods
is presumably worth the expected benefits from ODS abatement. The government can try to recoup
all or part of the costs by imposing a tax on the final user/consumer when purchasing items that later
give rise to the disposal problems. The tax might equal the subsequent refund (incentive) to be paid
to the user to induce compliance with proper disposal procedures. Or more drastically, the tax might
equal the full cost to the government of proper disposal. However, one may reasonably object that
the latter is overkill: the user/consumer will already have paid a (quasi) pollution tax on ODS under
the policies recommended earlier (with the exception of products already in use or in the pipeline).
At any rate, the desirability of the refundlincentive features of the disposal scheme does not depend
on the "deposit"/financing arrangements.
37.   Disposal problems take on a special edge when the target date for the complete phase-out
of ODS use is reached. Apart from other "leakages" noted above, such as through stockpiling and
recycling of ODS, there will be a stock of ODS containing products which continue to emit ODS
beyond the phase-out date. If it is considered important to eliminate such emissions as much as
possible, it is essential to have an effective dispw-al system, with strong inducements/penalties to
ensure compliance with the phase-out timetable set for this remaining stock.
38.   Disposal may be more than a temporary problem. If permitted use/consumption of ODS is
defined net of proper disposal/destruction (as in the Protocol), there may be a continuing flow of new
supplies of ODS after the phase-out date. This is similar to the quasi-permanent overhang of a stock
of ODS if recycled supplies are left outside the ODS supply limits. While in principle these
approaches do not add to ODS emissions, they appear at odds with the basic notion that ODS use
is dangerous and should be phased out as soon as reasonably possible.
13


Conclusions
39.    The major elements of a national ozone policy to meet the obligations under the Montreal
Protocol, as discussed above, may be summarized as follows: The Country should preferably control
ODS supplies (production and imports), rather than ODS uses. It should allow the market to
allocate scarce ODS supplies among the many users. This would tend to ensure cost-effective
abatement, provided that market distortions are kept within reasonable bounds. Excess profits of
suppliers, because of the rise in ODS prices when supplies are restricted, should be eliminated by a
(quasi) pollution tax on ODS production and imports. The scheme should be phased in gradually so
as to avoid unnecessary shocks and uncertainty, and to enhance/accelerate ODS abatement. Some
supplementary measures may be necessary to deal with the possible effects of stockpiling of ODS,
and of "indirect" imports of ODS incorporated in various products.
40.    Recycling of ODS should be encouraged, preferably by including recycled supplies within the
overall ODS supply limits. When recycled supplies substitute for new production and/or imports of
ODS, they will reduce its aggregate supply and use, and thus reduce ODS emissions. Such recycling
should be subsidized, if necessary, to make it competitive with new supplies of ODS. When recycling
substitutes for ODS abatement, it reduces the adjustment costs, and postpones ODS emissions.
Subsidies (to make recycling competitive with abatement costs) are probably not necessary in this case
and questionable if they are. Proper (ozone-friendly) disposal of goods incorporating ODS should
be encouraged through refund schemes, standard setting, and public payment of the cost of disposal.
41.    This set of policies primarily relies on controls for a few suppliers (producers and importers),
but leaves maximum freedom to the many and varied users to adapt to these supply restrictions as
best they can. It appears effective in curtailing/phasing out the supply and use of ODS; relatively
easy to design, implement and enforce; cost-effective in its abatement, and reasonably fair in its
impact. Policies based on pollution taxes or marketable use permits would, in principle, have the
same results. But in practice they would be more difficult to design and enforze, or more uncertain
in this control over total ODS abatement over time. The administrative advantage appears to lie
clearly with the proposed hybrid scheme based on control over ODS supplies, rather then uses,
market allocation of scarce supplies among the users, and (quasi) pollution charges.
14


42.    How important is adoption of the policies advocated here to reducing the economic costs to
a country of meeting its obligations under the Montreal Protocol? The differences in cost of
alternative ODS phase-out scenarios, if available, may give some indication of what can be at stake.
The policies supported here are conducive to minimizing the cost of any chosen time path for the
phase-out (A slower phase-out may have even lower costs.) Other market-based policies, such as
pollution taxes or marketable permits, are likely to have similar cost results, with the qualifications
noted earlier. Command-and control policies, on the other hand, would probably result in
substantially higher costs, since it is difficult to estimate and implement the minimum-cost solution
among many producers/users. They also appear an anomaly in the increasing number of countries
whose economic policy regime is generally market-based. Conversely, the policies advocated here are
hardly suitable in countries that broadly adhere to central planning.
II.    International Policies: The Interim Multilateral Fund of the Montreal Protocol
43.   International abatement policies. The Montreal Protocol mandates the same ODS phase-out
targets for all participating countries, with the major exception of a ten-year grace period for
qualifying developing countries (as defined under the Protocol, i.e., countries using no more than 0.3
kg of ODS per head). It does not attempt to influence the phase-out pattern among (or within)
countries, except to the minor extent of a few interim targets. Just as at the national level adjustment
to the phase-out is easier/less costly for some users of ODS than for others, so the cost of adjustment
will be greater for some countries than for others. From a global point of view it would be more
efficient if, for a given level of ODS abatement, the adjustments were concentrated in those countries
where this can be done at least cost. (Any inefficiencies on this score are, of course, resolved when
the phase-out is complete, as envisaged under the Protocol for most ODS.) But the Protocol does
not provide a mechanism to encourage this. The grace period for developing countries was inspired
by equity rather than efficiency considerations.
44.   In principle, one might try to promote a more globally efficient ODS abatement pattern by
(i) specifying interim targets suitably differentiated by country, so as to accelerate adjustment in
countries where costs of substitution are low, and slow it down in other countries; (ii) controlling
global supplies of ODS, and letting the market allocate the restricted supplies among user countries
in accordance with their relative ease of switching to alternatives; (iii) levying international ODS
taxes, which would encourage phase-out in countries where the costs of doing so are less than the
15


tax, and discourage abatement in countries where paying the tax is cheaper than phasing out ODS;
and (iv) by issuing internationally marketable permits for the use of ODS, with similar effects as ODS
taxes. The advantages and disadvantages of these approaches in the international arena are similar
to those discussed above at the national level. But practical problems of design and implementation
are even greater at the international level, and the difficulties of reaching political agreement appear
formidable.
45.   International support funds. At any rate, the Protocol does not address these issues of how
the phase-out of ODS could be achieved most efficiently. It contents itself with the establishment
of an international fund (the Interim Multilateral Fund of the Montreal Protocol) to ease the burden
of adjustment in the poorer countries. The remainder of this section considers how such a fund can
best be used to facilitate an efficient and equitable phase-out of ODS in the recipient countries.
46.    What then should be the operating rules for an international fund such as the Interim
Multilateral Fund of the Montreal Protocol? The resources of such funds are always likely to fall
far short of being able to compensate qualifying countries for all potential costs involved in the ODS
phase-out. hat underscores the importance of using them so as to maximize the Fund's impact on
the efficient and equitable reduction and ultimate elimination of ODS pollution in the recipient
countries. Questions to be considered in this context include:
(i) principles of country allocation
(ii) promotion of cost-effective country policies and programs
(iii) eligible categories of financing
(iv) appraisal criteria
(v) funding conditions
(vi) financing terms
16


(vii) institutional arrangements.
These issues will be briefly reviewed below.
47.   Country allocations. What is a suitable basis for allocating the available international support
funds among countries? Any such allocation would, of course, be national and subject to revision if,
for whatever reason, countries did not take up the funds set aside for them. With this proviso, there
are several possibilities. A good starting point would appear to be allocation in (rough) proportion
to the burden on countries of compliance with the Protocol.
48.   In principle, the incremental economic cost to the country is the difference in cost with and
without the ODS phase-out program required under the Protocol (adjusted for any differences in
benefits). The relevant "with" program is the minimum cost program, was discussed earlier. The
higher cost of other possible phase-out programs is not attributable to the need to comply with the
Protocol, but rather to inappropriate country policies. The "without" program would simulate the
corresponding development in the absence of the Protocol. In practice, the with/without comparison
would focus on the micro-costs of ODS producers/users/consumers, together with related program
costs of recycling, technical assistance, etc., as discussed previously.!
49.   The problems and pitfalls of assessing incremental costs have been discussed at some length
elsewhere,' and will not be reviewed here. In the absence of reasonably reliable and comparable
estimates of this kind, the relative use of ODS in the base period (say, the late eighties) might serve
' To the extent that an ODS phase-out program also reduces global warming effects, this adds
to the global benefits of the ODS phase-out, but does not raise the incremental cost of the phase-out
program for the country concerned. This may increase the willingness of donors to contribute to the
Fund, but does not appear to be a reason to change the country allocation. However, if costs were
to be minimized for meeting both ODS and greenhouse standards, the cost of the ODS component
is likely to be higher than for an ODS program by itself that disregards the greenhouse effects. And
similarly for the greenhouse component. Because of the joint costs involved, the overall costs of an
ODS OpM greenhouse program would be lower than for two separate programs. This might be
reflected in the country allocations of a joint ozone/greenhouse fund. But in the absence of
agreement on a greenhouse phase-out, there seems to be little basis for taking these factors into
account in the operations of the Ozone Fund.
' See Incremental Costs: Options and Priorities, Discussion Paper, The World Bank, March 1991.
17


as a very rough proxy. This could be somewhat refined by making allowances for differences in ODS
use patterns among countries: some uses, where the costs of substitution are expected to be high,
could be given greater weight than others.
50.   A cost-based allocation is unlikely to be acceptable unless it is strongly biased in favor of the
poorer countries. A large first step in this direction is to limit funding to the developing countries
(as stipulated by the Interim Multilateral Fund of the Montreal Protocol). But one might go further,
in the interest of equitable burden sharing, and fund the adjustment costs of the poorest countries
in full, while funding only a fraction of the costs of better-off developing countries.
51.   Alternatively, funds might simply be allocated on a first-come, first-served basis, as an
incentive for countries to make an early start with ODS abatement. This would also have the great
practical advantage that the Fund gets used quickly, and does not have to wait too long for customers.
But the results are unlikely to be equitable. It might be combined, therefore, with country allocations
based on cost and poverty by making such allocations for a three-or five year period. The quick
starters would use up their "quota" already in the first year; the laggards only later (or not at all).
52.   Access to the Fund is presumably restricted to signatories of the Protocol, in part as a (small)
inducement for countries to sign. But there is some argument for making the Fund available also for
studies by a country of the cost of compliance with the requirements of the Protocol, as a preliminary
to joining. This would serve the general purpose of such Funds to promote ODS abatement.
53.   More generally, country allocations along the lines set out above would serve the purposes
of the Fund to facilitate and ease the burden of compliance with ODS phase-out requirements under
the ProtocoL However, they entail the risk that the bulk, or all, of the available funding will go to
only a few, large countries. This does not seem desirable. It may be necessary therefore, to put a
cap of say, 20 percent, on the share of the Fund that may be granted to any one country.
54.   Promotion of cost-effective abatement. The previous section has discussed the major
elements of cost-effective country programs and policies for ODS abatement, so as to meet the
requirements of the Montreal ProtocoL What can an international support fund do to promote such
programs and policies? Basically, it can do three things: (i) the Fund can make its financing
18


conditional on agreement with the country on sensible ODS abatement programs and policies; (ii)
it can help finance preparation of such programs and policies, and (iii) it can finance (part of) the
phase-out program itself.
55.   As with other development lending, the contribution of the Fund will be multiplied manifold
to the extent that it reaches out beyond the ODS abatement projects it finances directly to the
broader ODS phase-out program and policies of the country. As much as possible, therefore, it
should use its leverage to help ensure that the country's overall phase-out strategy and its
implementation are sound. This suggests that financial support by the Fund should be conditional
on agreement with the government on a sensible ODS abatement policy framework
56.   However, at this time, few countries have prepared such ODS phase-out plans, and it will be
some time before they become available. The start of operations of the Fund would probably be
delayed, possibly by many months or even years if its initial financing were to be conditional on
agreement on an acceptable ODS phase-out program and policies. At the same time, it will often
be possible to identify ODS abatement projects that are clearly sensible and of high priority in any
cost-effective ODS phase-out program to be formulated subsequently. This suggests the following
approach: the Fund should be willing to finance ODS abatement activities that appear of high priority
and are otherwise well-prepared, even in the absence of a proper ODS phase-out plan. But it should
do so conditional on the preparation and review of such a plan (that is, an ODS phase-out program
and the policies to implement it, as discussed above) within a reasonable period of, say, six months.
Any subsequent financing by the Fund would be contingent on agreement on an ODS phase-out
program and policies. The initial financing would support preparation of such plans.
57.    Categories of financing. This leads to the broader question of what the Fund should be
willing to finance. What are eligible categories for tinancing, and what are the priorities? A good
general principle in this regard is for the Fund to try to maximize its impact on the preparation and
implementation of sound ODS phase-out policies and programs. The foregoing argument suggested
making financing by the Fund conditional upon a satisfactory ODS abatement plan. The Fuzid might
well go further, and preferably help finance the phase-out program as a whole, rather than specific
abatement projects. This would substantially ease the administrative problems, and costs, for the
Fund or its Agents of having to evaluate and supervise mostly very small abatement projects, as well
19


as focus attention even more sharply on the need for satisfactory ODS abatement policies and
programs.' This would appear to be a sensible and attractive approach.
58.    However, as noted earlier, in the initial stages of the Fund few countries have ODS phase-out
plans that offer a satisfactory basis for such broad financial support. And some donors may prefer
more narrowly-focussed project lending for other reasons. So one may expect some financing of sub-
projects/components of the program by the Fund, at least initially. In that case it would appear
desirable to give priority in funding to those activities that have a broader impact beyond the
immediate project being financed. This suggests a rough order of priority for financing of sub-
projects in the program as follows:
(i) preparation of sound country ODS phase-out programs and policies, as discussed
above;
(ii) technical assisiance, including spreading relevant information, transferring
technology, training of staff involved in the design and implementation of country
ODS abatement policies and programs;
(iii) infrastructure necessary to support sound ODS abatement programs, for example,
recycling facilities, disposal plants, research and development facilities;
(iv) pilot or demonstration projects that, if successful, may be expected to generate a
lot of other projects substituting for ODS; and
(v) regular projects of which the primary purpose is to reduce the use of ODS, that is
for substitution and modifiration of existing production capacity (rather than
expansion of capacity, which should not be eligible).
'Funds could be disbursed against eligible expenditures under the program, or possibly as general
budget support.
20


59.   The Fund would finance the last category only after the needs of categories (i) to (iv) have
been considered. Regular substitution projects may become important in the later stages. Initially
at least, the bulk of Fund financing would be expected to support the first four categories. This
would have the added advantage of largely by-passing the difficulties of estimating the incremental
costs of ozone-friendly industrial investments."o
60.   Project standards. The appraisal standards for projects and programs to be financed by the
Fund would properly be based on those of international financing agencies with regard to their
technical, financial, economic, environmental, institutional and other aspects. Most likely the Fund
would use a particular agency, such as for example the World Bank, as its Executing Agent. In that
case the Fund's appraisal standards would follow those of the Agency concerned. Eligibility to
participate in procurement financed by the Fund would presumably be limited to members of the
Fund. Simplified procedures normally used for small projects would be expected to be applicable to
most projects financed by the Fund. This would often involve a suitable local intermediary which
would appraise sub-projects, and act as a conduit for funding, in accordance with standards agreed
with the Executing Agency. The key standard would be that investments financed by the Fund would
need to be cost-effective in abating ODS emissions.
61.   Conditionality. The appropriate conditionality to be attached to financing by the Fund has
been discussed already above. Basically, it would be two-fold: (i) agreement on the preparation and
implementation of a sound ODS phase-out program and policies, either, where appropriate, as a
prior condition for initial financing by the Fund, or as a condition for subsequent financing; and (ii)
agreement on suitable project/program standards. Any follow-up financing would be contingent on
satisfactory progress with implementation of the agreed abatement policies and programs, and/or the
0 See Incremental Costs: Options and Priorities, op,cit.
21


projects/program being financed.u Conditions are to be strictly limited to the ODS phase-out
program and policies, and not to be extended to broader issues of economic and industrial policy.
62.    Financing terms. At what terms should the Fund provide its financing? This involves two
different questions: (i) the financing terms to the country, and (ii) the on-lending terms, if applicable.
With respect to terms to countries, straight grants appear to be the most appropriate. This is in line
with the intended purpose of the Fund to provide some (small) compensation for the incremental
cost to the country of the ODS phase-out, as required under the Montreal Protocol; and it greatly
simplifies the administration of the Fund. Objections to grant financing for higher-income countries
can be met by adjusting the country allocations accordingly (see above). The question of on-lending
does not arise with the financing of the preparation of country programs and policies, technical
assistance in support of these programs, and probably at least some of the necessary infrastructure.
Such activities are presumably carried out by government departments. What fraction of the cost
should the Fund cover in such cases? It is probably desirable for the country to pay part of the cost,
in the interest of "ownership" of the project. There may be some advantage in spreading out the
financing over more activities in order to increase the leverage of the Fund. But too small a
contribution ra*ses the administrative costs to the Fund. On balance, somewhere between 50 and 80
percent may be a reasonable share of financing by the Fund of such projects.
63.    Recycling projects and pilot and demonstration projects are more likely to entail on-lending,
to tLe extent that they are carried out by private and public enterprises. The benefits of such
projects are expected to accrue not just to the enterprises undertaking them, but to spill over more
widely to other firms. Furthermore, they may not be undertaken without some sort of subsidy to
make them profitable. However, as argued before there should not be a presumption that recycling
necessarily needs to be subsidized. The same applies to pilot and demonstration projects of
alternatives to using ODS. If subsidies are in fact required, they might be provided through
concessionary on-lending terms. It may well be preferable, however, to maintain standard market
n It has been suggested that if a country is fully/or largely compensated anyway, it may have no
incentive to minimize the phase-out costs. But there is little risk of this happening, since the
resources of the Fund are limited. There is no open-ended commitment. Countries may be tempted
to overstate their phase-out costs, as a basis for a larger country allocation. But only the minimum-
cost phase-out program would be considered in the allocation, and only cost-effective programs and
projects would be eligible for compensation by the Fund.
22


terms for all on-lending, and provide separate subsidies, if and as required. That would result in
greater transparency, easier administration, and greater flexibility to adjust the necessary subsidy to
the particular circumstances.
64.   The same line of argument suggests that there is normally little reason to provide on-lending
for regular ODS abatement projects at concessionary terms. If such projects are undertaken to meet
the phase-out requirements under the Protocol, incentive reasons for subsidies are weak, since
appropriate policies would be expected to give all the necessary incentives, through rising ODS prices,
for the desired pace of the phase-out. Equity reasons for subsidies, to compensate for losses, also
appear implausible, as noted earlier. Market on-lending terms, without subsidies, appear to be the
most appropriate for such projects.
65.    In some cases, the phase-out induced by the adopted ODS abatement policies may be
considered slower than desirable. One might then wish to use concessionary on-lending terms to
stimulate a faster phase-out. However, it would be preferable in that case to adjust the abatement
policies accordingly, by tightening the ODS supply limits, raising ODS prices, and hence promoting
substitution. If this is not possible, for whatever reason, it would still seem advisable, as noted
above, to provide direct subsidies, together with market on-lending terms.
66.   The general conclusions with regard to the financing terms of the Fund are therefore that
grants are appropriate for the terms to the country, but that on-lending, if relevant, should normally
be at market terms. The Fund should not provide subsidies, through its on-lending terms, which run
counter to the national ozone policy. Any desirable subsidies, and they often are not necessary, are
better provided directly, rather than through concessionary on-lending terms.
67.   Institutional issues. The operations of the Fund raise also some institutional issues for (i) the
country, (ii) the Fund itself, and (iii) its Executing Agent(s). The country will, of course, need to
decide which government department has primary responsibility for dealing with the Fund and
reaching agreement on Fund-supported ODS abatement activities in the country. It will also need
n An accelerated ODS phase-out tends to increase its costs to the country and hence, in principle,
the country allocation by the Fund (as argued above). This might be explicitly recognized in the
allocation policy.
23


to assign responsibility for designing and implementing suitable ODS phase-out poiicies and programs,
so as to meet the country's obligations under the Montreal Protocol, and for monitoring their
progress. Presumably the same agency would be given responsibility for both these functions. A
plausible candidate would be the Ministry of Industry, since the ODS phase-out demands adjustments
mostly of industry. Some countries may prefer to assign the primary responsibility to their Ministry
for the Environment.
68.   The Fund will have to decide on its general policies as reflected in operating rules such as
discussed above. It will also have to work out the relationship with its Executing Agent(s), if any.
In particular, how much authority is being delegated to the Executing Agent. For example, does the
Fund determine broad policy principles, but leave elaboration of more specific operating rules to the
Executive Agent? Or does the Fund rather keep close control over individual financing decisions?
It is unlikely that cooperation between the Fund and an Executive Agent, such as the World Bank,
is workable without a reasonable delegation of authority and discretionary decision-making by the
Executive Agent, within broad policy guidelines set by the Fund. Executive Agency relations are
further complicated if more than one such Agency is involved, as is the case with the Interim
Multilateral Fund of the Montreal Protocol. This requires a clear agreement with the Fund and
among Agencies concerned regarding their respective roles. It is unlikely to work well unless one
agency is assigned the leading role.
69.    Finally, for an Executive Agency the institutional implications of the Interim Multilateral Fund
of the Montreal Protocol are straightforward. It needs a clear understanding of the operating rules,
as worked out in cooperation with the Executive Committee of the Montreal Protocol. And it needs
to assign responsibility within its own organization for formulating the operating rules and policies,
and for carrying out the actual operations to be financed by the Fund. In the case of the World
Bank, for example, drafting of operating rules might be assigned to the Office of the SVP Operations,
in consultation with the Environmental Department; and actual Fund operations would be the
responsibility of Regional staff, and probably be assigned more specifically to the Industrial Divisions.
I Other important institutional issues, such as the Fund's internal structure, the respective roles
of its the Executive Committee and Secretariat of the Interim Multilateral Fund of the Montreal
Protocol, and how they are being financed, are ignored here.
24


As in any new area of lending, staff training and consultant services are likely to be necessary to
ensure adequate expertise.
Conclusions
70.   This second part of the paper has briefly reviewed how an international fund can maximize
its impact on the efficient and equitable phase-out of ODS pollution. It has concluded that available
support funds should be roughly allocated among (developing) countries in accordance with their
relative cost of adjustment to the phase-out, but with a heavy bias in favor of the poorest countries.
Financing should be conditional on preparation and implementation of sensible, cost-effective ODS
abatement policies and programs (with a flexible interpretation permitting a rapid start-up of the
Fund). The Fund should preferably finance ODS abatement programs, rather than projects. Any
project financing should give priority to preparation of ODS abatement programs and policies,
technical assistance, recycling and disposal facilities, and pilot and demonstration projects, in this
order. Financing of regular ODS substitution projects should be a last resort. Appraisal standards
would be essentially the same as those of the Fund's Executive Agent. Fund financing should be
provided to countries as grants. On-lending terms, where relevant, should preferably be at market
rates. Subsidies, if and when necessary at all, are best provided separately. The main institutional
requirements are clear assignment of responsibilities, common understanding of the Fund's operating
rules, and appropriate delegation of authority to the Executing Agent.
III.  Links to Existing Framework for Fund Operations
71.   The operational policies currently being developed by the Executive Committee of the Interim
Multilateral Fund of the Montreal Protocol have a somewhat different emphasis from those favored
above. But they are in many ways compatible."' They envisage preparation of country programs by
recipient countries outlining, inter ulia, their phase-out programs and policies for implementation of
the Protocol. "A country programme approved by the Executive Committee shall serve as a basis for
project preparation and further cooperation between the Party and the implementing agencies.
However, individual projects prepared prior to the completion of a country programme shall also
" See "Implementation Guidelines and Criteria for Project Selection", Annex III, UNEP/OzL
Pro/ExCOM/3/18, Executive Committee (1991).
25


qualify for funding if consistent with project eligibility criteria .........approved by the Executive
Committee" (ibidem, para. 5 of Section II). This is in line with the suggestions in this paper.
72.    The section on (interim) project eligibility criteria specifies that assistance from the Fund shall
be available to qualifying parties only "......... to finance the incremental cost incurred in meeting
(their) requirements (under the Protocol." While this appears to refer to incremental costs to the
country, the remainder of the section deals only with project criteria, and with incremental costs in
that context. There is no discussion of country allocation criteria.
73.    The draft guidelines include among the categories eligible for assistance by the Fund ".........
inter alia training, technical assistance, pre-investment studies, country programs,' or capital
investments to modify or establish a manufacturing facility." They state explicitly that "Projects other
than capital investment projects shall also qualify for assistance under the Fund (e.g., technical
assistance and clearing house facilities." This is wholly in the spirit of the present paper. But the
further discussion of eligibility for financing in the guidelines is nearly exclusively concerned with
capital investment projects, and the definition of their incremental costs, which runs contrary to the
financing priorities envisaged in this paper. The guidelines do note the need for setting priorities "in
the event of a shortage of available funds," which would appear to be a near-certainty. However they
limit themselves to the following ordering:
"1.    Projects with potential for the most cost-effective and efficient reduction of emission
of controlled substances.
"2. Projects involving broad geographic balance.
"3.   Projects with ease of replication and technology transfer to other (qualifying)
parties........" (Section IV of draft guidelines.)
nThe intention here is probably the preparation of country programs, but it appears to leave
open the possibility of financing the programs themselves, as proposed in this paper.
26


74.    This gives little guidance on the envisaged priorities in financing the various elements of ODS
phase-out programs, such as preparation of such programs, technical assistance, training, pre-
investment studies, recycling facilities, disposal plants, research and development, pilot projects, and
regular substitution projects, all of which appear eligible for assistance by the Fund. Basically, it
appears to leave priorities to the discretion of the implementing agencies.
75.    However, the guidelines recognize that the eligibility criteria will be clarified in the light of
experience, are "a flexible document, and may be revised after one year...... And they specifically
mention the need for further refinement in defining incremental costs eligible for financing.
76.    The guidelines provide for grant financing by the Fund, except "where the investment project
will be profitable in a short period of time (i.e., 1-2 years)......", when concessional terms by the Fund,
or (commercial) loans by existing financial institutions may be appropriate. No distinction is made
between country terms and on-lending terms.
77.    This paper argued earlier in favor 3f grants to countries, and commercial terms for on-lending,
where relevant, with subsidies (or concessional terms) as necessary and desirable. It does not believe
that there should be a presumption in favor of subsidies, especially if countries follow sensible phase-
out policies. Nevertheless, in practice there may well be a substantial convergence of the two
approaches, depending on the projects being financed and the interpretation of the profitability
criteria in the guidelines.
78 From this brief review of the link between the initial operating policies of the Fund as
outlined in the draft guidelines, and the operating rules proposed in this paper, it would appear that
there are no unreconcilable differences. While the initial emphasis in the guidelines is on capital
investment projects to help manufacturing enterprises to switch to ozone-friendly production facilities,
there is nothing in them to prevent moving towards a broader program and policy approach as
advocated in the present paper. This could be clarified further in the revision and refinement of the
guidelines that are envisaged in the current draft.
79.    Perhaps the major differences between the guidelines and the approach of this paper relate
to the treatment of incremental costs. This paper is mainly concerned with the incremental
27


(economic) cost to the country, as a basis for country allocations. The guidelines (and its Appendix)
focus on incremental (financial) cost to enterprises converting to more ozone-friendly production
methods, as a basis for project financing. As elaborated in the World Bank paper on incremental
costs referred to earlier (para. 49), the relevant incremental costs are essentially the cost of
converting or replacing existing capacity, net of its amortization to date and its remaining salvage
value, pLus (minu) any operating cost increases (savings).'
80.   If the costs of substitution capacity are roughly the same as for the existing capacity, the
incremental costs (and the profitability of substitution) depend largely on the degree of amortization.
If existing capacity is already mostly amortized, incremental costs would be small and substitution
would tend to be profitable. Any financing would then appropriately be at commercial terms, as
suggested by the guidelines. Conversely, with relatively new existing equipment, early substitution
would entail relatively high incremental costs, and low profitability. Grant financing may appear
appropriate in such cases, in accordance with the guidelines, as an incentive to accelerate ozone-
friendly substitution, and implementation of the Montreal Protocol. Howev--r, this approach ignores
that the profitability of substitution depends also on the national phase-out policies being followed.
By suitable measures to raise the prices of ODS, the government can stimulate cost-effective
substitution and accelerate the phase-out, and eliminate the need for subsidies (see the discussion of
nationa! ozone policies above).
81.    The World Bank paper on incremental costs suggests that in such cases as recycling plants,
patents, retraining, technical assistance, and one might presumably add preparation of country
programs, ODS disposal plants and research and development," all reasonable costs would be deemed
incremental." This appears eminently sensible. Indeed, by focusing its financial assistance on such
items, as suggested in this paper, the Fund would minimize the vexing problems inherent in defining
and estimating the incremental costs of the ODS phase-out for individual manufacturing enterprises.
It would greatly reduce a major obstacle to its smooth and efficient operation.
* Pragmatically limited to a period of 5 years.
28