Jump to ContentJump to Main Navigation
New Sources of Development Finance$

A. B. Atkinson

Print publication date: 2004

Print ISBN-13: 9780199278558

Published to Oxford Scholarship Online: January 2005

DOI: 10.1093/0199278555.001.0001

Show Summary Details
Page of

PRINTED FROM OXFORD SCHOLARSHIP ONLINE (oxford.universitypressscholarship.com). (c) Copyright Oxford University Press, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use. date: 27 November 2020

A Development-Focused Allocation of the Special Drawing Rights

A Development-Focused Allocation of the Special Drawing Rights

(p.90) 5 A Development-Focused Allocation of the Special Drawing Rights
New Sources of Development Finance

Ernest Aryeetey

Oxford University Press

Abstract and Keywords

The original purpose of the Special Drawing Rights (SDRs) of the International Monetary Fund (IMF) was to increase international liquidity, but here the emphasis is on the potential role of SDR creation in providing funds for development finance. The chapter describes the historical development of ‘regular’ SDRs, including their purpose and the problems associated with their creation and allocation. It then introduces the recent proposals for development‐oriented SDRs, and analyses these proposals with a focus on the arguments for and against them. Next, the mechanisms proposed for the creation of development SDRs are discussed, including the institutional and structural changes in international finance that may support their creation. The concluding section of the chapter assesses the effects of development SDRs on the resource situation in poor (developing) countries and how they will be perceived.

Keywords:   creation of development Special Drawing Rights, developing countries, development finance, development funding, development Special Drawing Rights, institutional changes, international finance, International Monetary Fund, Special Drawing Rights


The campaign for the issue of development-focused Special Drawing Rights (SDRs) by the International Monetary Fund (IMF) has been on the development agenda for many years. This was certainly reflected by the Brandt Commission's report (1970). In more recent times the campaign has been propelled, first, by the Asian financial crises as the liquidity problems of developing nations took a new dimension with the collapse of private capital markets, and then second, with the search for financing options towards the achievement of the Millennium Development Goals (MDGs). Since the Asian crises broke, there has been a shrinking of capital flows to small nations, making it more difficult for them to hold adequate reserves, even if they chose to borrow for that. It is reckoned that when borrowing for reserves becomes difficult and a need is felt to increase reserves, as after the Asian crisis, many poor nations are obliged to make consumption and investment decisions that harm future growth and development. This is one of the reasons for the growing calls for multilateral finance institutions to be innovative in increasing the financing options available to these countries, including an expanded and enhanced role for SDRs.

The SDRs are international finance instruments created for the purpose of providing an increase in the world stock of monetary reserves from time to time without making countries run surpluses or deficits. Indeed, the idea behind the SDRs is that large imbalances force countries to incur costs in earning or borrowing reserves, and this should be contained with the IMF allocation of SDRs. It was born out of what has been referred to as Triffin's dilemma (Triffin 1960): ‘whereby additions to official dollar holdings were seen as undermining the stability of the system, given the tendency on the part of some central banks to convert their dollar reserves into gold, thereby drawing down the limited US gold stock’ (Clark and Polak 2002: 3).

Extending the uses to which SDRs can be put results from the growing demands on the international financial system to respond to the development finance needs of poor nations. Apart from the need to provide emergency funds in times of crises and the whole area of crisis prevention, increasingly the facilitation of development in poor countries and assistance for making the best policy decisions are considered (p.91) crucial. It is the growing complexity of the requirements of the international financial system that leads to the frequent suggestion that the IMF should increasingly play the role of a lender of last resort, including the issuing of SDRs from time to time. Some of the strongest advocates for this idea have been associated with the Fund. Shortly before retirement as IMF managing director, Michel Camdessus proposed that in times of a systemic credit crunch, ‘the IMF be authorized to inject additional liquidity—and to withdraw it when the need has passed—in a manner analogous to that of a national central bank, through the creation and selective allocation of SDRs’ (Camdessus 2000: 3).

But this chapter on SDRs separates, despite the obvious links, the issues of international liquidity reform and development finance. The focus here is on development finance. The recent calls to redistribute SDRs for development purposes were given a boost by the Zedillo Report (UN 2001) and by the appeal from George Soros (2002), and the contributions from Stiglitz (2003). Proponents for the creation of development SDRs argue that such additional funds for development can assist in meeting the MDGs. Such SDR creation would be the means for reducing the costs to developing countries of the increased holding of reserves, which has become quite common after the 1997–8 East Asian crises. In the report of the Zedillo Panel, it is argued that a resumption of the issue of SDRs would reduce the demand for US-dollar holdings, a development that might discourage the indefinite increase of US short-term debt. Soros (2002) suggests that there should be a resumption of SDRs, by which developed countries would re-allocate their share first to the provision of global public goods, and secondly to supplement aid flows to individual countries. If the approved allocation of 1997 were made active and the re-allocation took place, he expects that US$18 billion could be made available for global public goods or additional aid.

It is important that the arguments for the creation of SDRs for development are explored in some detail, particularly from a developing-country perspective. While showing the likely benefits of such SDRs on the development process in poor economies, the arguments that are often raised against such an endeavour must be explored, both with regard to the global economy and to the poor nations. The entire exploration should investigate the feasibility (both technical and political) of the undertaking, and the likely consequences on the overall flow of resources for development.

This chapter discusses in Section 5.2 the historical development of SDRs, including their purpose and how they are created. This is basically about regular SDRs and the problems associated with their creation and allocation. Section 5.3 introduces the recent proposals for development-oriented SDRs, and Section 5.4 analyses these proposals with a focus on the arguments for and against them. In Section 5.5, the chapter discusses the mechanisms proposed for the creation of development SDRs, including the institutional and structural changes in international finance that may support their creation. Section 5.6 concludes with an assessment of the effects of development SDRs on the resource situation in poor countries and how they will be perceived.


The SDRs are a special reserve asset allocated to IMF member countries participating in the SDR department proportionate to their IMF quotas. SDRs are not a claim on the IMF but are potentially a claim on the freely usable currencies of IMF members, in that holders of SDRs can exchange their SDRs for these currencies (IMF 2002a). Its value as a reserve asset comes from the commitments of members to hold and accept it. The IMF members have, indeed, undertaken to honour all obligations related to the SDR system. The Fund ensures that the SDRs' claim on freely usable currencies is honoured by first ‘designating IMF members with a strong external position to purchase SDRs from members with weak external positions, and through the arrangement of voluntary exchanges between participating members in a managed market’ (IMF 2002a). The least developed countries (LDCs) may use the SDRs (i) to repay the IMF; (ii) to repay Paris Club debt; (iii) to help countries in foreign exchange crisis get hard currency from IMF by exchanging at the Fund their SDRs for US dollars with the IMF matching them up with a source of dollars, and (iv) to release hard currency reserves for use in such transactions.

Indeed, in the basic form, SDRs are promissory notes issued by the IMF to member states on the basis of a quota that is related to their relative strength in the world economy. Members that receive these notes may either hold them or exchange a part of them over time for hard currency, through the Fund itself and through central banks. In a regular sense, SDRs may be perceived as liquid assets that are created by the IMF in the same manner that national monetary authorities issue their currencies as liabilities against themselves and can affect the supply of money. In this limited context, the Fund may behave like a central bank so long as it commands sufficient credibility. On the other hand, however, because of the peculiar treatment that SDRs are given, such as the fact that they are allocated by quotas and only central banks can hold them. No other asset has to be exchanged for them, and they cannot be treated as money. If countries received their allocations of SDRs and held them, there is no interest cost to them.1 Countries actually receive interest income from the Fund for their SDR holdings, but they only pay interest at the same rate on their total cumulative allocations. When countries seek to exchange SDRs for hard currencies, the new holders then earn the accompanying interest from the Fund (Fig. 5.1). In effect, it is only when a country's holdings of SDRs are less than its cumulative allocation that it becomes a net payer of interest. So, obviously when poor countries exchange SDRs for hard currency, they have to pay the interest on these.

A Development-Focused Allocation of the Special Drawing Rights

Figure 5.1. SDR flows


In addition to the general allocations of SDRs for the purpose of supplementing existing official reserve assets of member countries, there is currently a proposal to create a special one-time allocation of SDRs that would enable all IMF members to receive such SDRs on an equitable basis, making amends for the fact that a fifth of the membership have never received an SDR allocation. The proposal followed the fourth amendment of the Articles of Agreement by the IMF board of governors in 1997, and will be effective only when the United States gives its backing to what 73.34 per cent of the total voting power already supports.

Even though the original intention behind the creation of SDRs was to increase what was thought of as a total international liquidity, Clark and Polak (2002) argue that that rationale is no longer relevant. They maintain that if international liquidity were simply an aggregation of the foreign currency and gold holdings of all the economies of the world, then there was adequate liquidity and this had been growing as fast as the world's economies, if not faster. Hence, the concept of aggregate international liquidity was no longer relevant for creating SDRs. They also point out that the conditions in the global economy that necessitated the creation of SDRs changed shortly after these reserves were created. This included the adoption of flexible exchange rate regimes, which may have reduced the size of the reserves needed, compared to what was required to maintain a fixed exchange rate regime. But they argue that this situation notwithstanding, there were still good enough reasons to create SDRs in view of efficiency gains resulting from the low-cost access to reserves and the reduction in systemic risk. Developing countries, which would otherwise incur costs in adding to their international reserves, would thereby be enabled to do so costlessly. Their subsequent increased readiness to hold reserves would reduce the risk of financial crisis to them—and to the international system.

5.2.1 Trends in SDR Allocation and the Global Reserve System

There have so far been only two rounds of creation of SDRs, each spread over three years. The first allocation was in 1970, for a total amount of SDR9.3 billion. The most recent allocation, made on 1 January 1981, brought the cumulative total of SDR (p.94) allocations to SDR21.4 billion. The IMF executive board discussed the possibility of an SDR allocation during the fourth, fifth, sixth, and seventh basic periods, that is up to 31 December 2001, but there has not been enough support for an allocation.

A major reason for no SDRs being allocated since 1981 is that industrial countries no longer see clear benefits from receiving such allocations, particularly in the presence of thriving capital markets with full capital mobility. Low-income developing countries, with generally far less stable international earnings, could make very good use of SDRs, but their small quotas mean that they get very small shares of any allocation in relation to those of the big industrial countries that do not want or need them. But any allocation that is not in accordance with IMF quotas would require amendment of the Articles of Agreement. Thus, because of the need to amend the Articles, the 1997 agreement on a special one-time allocation of SDRs, which would equalize the ratio of cumulative allocations to current quotas for all member countries, is unratified six years later.

The situation has extensively reduced the role of SDRs as a reserve asset. By April 2002, SDRs accounted for less than 1.25 per cent of IMF members' non-gold reserves, even though the holding of reserves was growing worldwide (see Table 5.1). Indeed, the developing countries added little to reserves between 1980 and 1995, but then (p.95) added SDR 37 billion in the following five years.2 When the Zedillo Panel called for the resumption of the issue of SDRs, the panel members argued that, ‘the cessation of allocations has severely prejudiced the interests of developing countries’ (UN 2001). They suggested that developing countries were not in a position to borrow additional reserves in the market on terms similar to SDRs. But many developing countries were trying to build up their reserves in order to reduce their vulnerability to crises. With the emerging markets included, they were estimated to hold reserves of over US$850 billion, which was almost US$300 billion more than before the Asian crisis. These additional reserves had been borrowed largely on terms clearly more difficult than the SDR issues. At the time, emerging markets paid an average premium of about 8 per cent more than the US Treasury bond rates. The panel noted that, ‘the result is a large flow of what is sometimes called ‘reverse aid’, which in the aggregate is not far short of the flow of conventional aid from the DAC countries’ (UN 2001).

Table 5.1 Worldwide non-gold reserves, 1970–2000a (in billions of SDRs)








Advanced economies, of which:b
















Hong Kong SAR




















Emerging markets, of which:b




































Developing countriesc of which:b
































United Arab Emirates















(a) The increase in worldwide reserves between 1970 and 1995 is slightly overestimated because data for a few economies become available only in the latter part of the period.

(b) Economies with the largest increase in reserves (in billions of SDRs) between 1995 and 2000.

(c) Excluding economies that are included as emerging markets.

Source: Clark and Polak (2002).

Stiglitz estimates that there are currently US$2.4 trillion held in reserves around the world, held in a variety of forms including US Treasury Bills for most developing countries:

While the United States may benefit from the resulting increased demand for US Treasury Bills, the cost to the developing countries is high. Today they receive a return of 1.25 per cent—a negative real return rate—even though investments yield high returns in their own countries. This is the price developing states have to pay to insure against unpredictable market events. (Stiglitz 2003: 56)

There are a limited number of private financial instruments that are denominated in SDRs. Because of their limited use, the main function of the SDRs' is to serve as the unit of account of the IMF and some other international organizations (IMF 2002a). It is thus used almost exclusively in transactions between the IMF and its members.

5.2.2 The IMF, Developing Countries, and SDR Creation

The creation of SDRs has always been seen in developing countries as an inadequate but necessary tool for countering the usual problem of low reserves. It has been perceived as inadequate in relation to the financing needs arising from increasing volatility in exchange rate instability and foreign exchange earnings. While inadequate, its creation has sometimes been seen as an attempt by developed or industrial economies to cover their reluctance to deal with the real financing issues confronting developing countries through IMF that has for long been seen as a rich men's club. Aboyade (1983) notes strongly that,

The Fund has always related its (foreign) currency sales and stand-by facilities for member countries to their respective quotas. This means the richer countries, which have (and have always had) the highest quotas, can also borrow the most … It is generally unsympathetic to exchange rate policies, which offer any strong prospect of affecting the existing pattern and structure of international economic power, with the excuse that they may hurt international (p.96) trade. Most of its innovations over the last decade (for example, the creation of SDRs and the establishment of a Substitution Account) are more to help the currencies of the rich nations and preserve international stability, than as direct answers to the clamour of the poor countries. (Aboyade 1983: 30)

Wade (2002) comments on similar sentiments expressed by Triffin (1968) who had suggested that since the allocation went strictly in proportion to the countries' quotas, it was ‘as indefensible economically as it (was) morally’. At the time, the two biggest economies received one-third of the total. In Triffin's (1968) view, ‘the SDR designers had created an asset that made the rich even better off’.

The role that the Fund's board plays is crucial in understanding some of the sentiments often expressed about the issue of SDRs. The board of governors of the IMF (in which all member governments are represented) has to approve each issue of SDRs by an 85 per cent majority.3 The voting system is weighted so that a small number of industrial countries can veto any new creation. This allows the United States alone to do the same. It is the lack of interest of the industrial economies, for reasons already provided, that has ensured that there was no new approval between 1978 and 1997. Opposition from Germany, Japan, the United Kingdom, and the United States effectively made new allocations impossible in the period. The status quo is that any changes in the quota arrangements would require an amendment to the IMF's Articles of Agreement.

It is interesting that in 1997 when an attempt was made to issue new allocations in order to make the cumulative proportions of SDRs received up to that time equal to the various member-countries' current quotas, this received a negative response only outside of the Fund itself. Indeed, the necessary changes to bring about the Fourth Amendment to the Articles of Agreement, which would allow an allocation other than in proportion to current quotas, had been made, but the amendment failed year after year to reach the required level of ratification by the 110 members who constitute 85 per cent of the voting power. By the end of April 2001, only 107 members had ratified it. The main obstacle has been that the US Congress remained opposed to it, despite the support of the Clinton administration.

The US Congress and other industrial country governments believe that there are a number of good reasons why they should worry about changes in the power structure that pertains at the IMF, particularly if that change means an increasing use of SDRs for reserve holdings in many countries. Increasing the stock of SDRs would mean a likely reduction in the holding for reserves of US bonds, ‘Certainly expansion of the SDR stock touches America closely because of the prominence of US dollar holdings among existing reserves. Cutting the world's dependence on dollar reserves would reduce Americans’ access to a deepening well of cheap credit' (Clunies-Ross 2002: 30).

But economists at the Fund have made some of the strongest arguments for the resumption of regular SDRs. Clark and Polak (2002) provide an interesting argument (p.97) for allocating new SDRs. After indicating that the original justifications for creating SDRs were no longer relevant as the concept of aggregate international liquidity lost its meaning in the post-Bretton Woods era, they argue strongly that the individual developing countries' need for reserves was still significant, and this was important for improving the operation of the international monetary system. Considering that most countries still needed to increase their reserve holdings in view of expanding current and capital account fluctuations, if countries were to attempt doing this by generating a balance-of-payments surplus this would be costly in terms of foregone consumption and investment. On the other hand, attempting to borrow such reserves on the international capital markets is very costly for poor countries, while some countries have no access to such markets. They also indicate that while borrowed reserves may substitute, to some extent, for owned reserves, ‘volatile capital flows demonstrate that undue reliance on international capital markets for (the) purpose (of holding reserves) can be risky’ (Clark and Polak 2002: 11). Their main argument (following an earlier point of Mussa 1996) is that SDRs offer a costless reserve asset which, if properly managed as required by the Articles of Agreement, leads to enormous efficiency gains for the world economy compared to the cost of foregoing consumption and investment, and the cost of borrowing from the capital markets. The risk involved in increasing SDR allocations to developing countries can also be shown to not worsen. They observe that:

a number of considerations suggest that the provision of reserves in the form of SDRs would in fact reduce credit risk. Allocations of SDRs make more external resources available to a country, enabling it to weather potential balance-of-payments crises without undue reliance on import compression or the imposition of trade and other restrictions. (Clark and Polak 2002: 19)

Clark and Polak (2002) additionally argue that SDR allocations would contribute to reducing systemic risk. This is because they are a permanent addition to the world's stock of reserves since the Fund is unlikely to cancel any stock of SDR holdings. They contrast this with reserves acquired from borrowing on the capital market, which may be ‘withdrawn under inauspicious circumstances’. The example from the crises of the 1990s that made it difficult for countries to refinance their debt is given to support this point.

In the Zedillo Report, the panel members argue that now will be a good time for resuming allocations, ‘in that the original concern was not just with the cost to a typical country of having to earn or borrow a secular increase in its reserve holding, but also with the impact on the financial fragility of the country issuing reserves’ (UN 2001).

They argue that the financial fragility of the countries that issued the reserves was not much of a concern before, but it is now. This they attribute to the unprecedented size of the US current account deficit. This is partly a consequence of the desire all over to build up dollar reserves. These have become too large for comfort and a source of discomfort in the financial markets. ‘Substantial SDR allocations might help to shrink the US deficit while allowing other countries to continue to build up the reserves they feel they need to guard against financial crises’ (UN 2001).


As already mentioned, there is a long history of proposals to resume allocations of SDRs for specific purposes, often development, or to redistribute holdings of SDRs. Indeed, it would appear that the campaign for SDRs to support development is being pushed along two complementary lines. The first is for the resumption of the regular SDRs to deal with what has become the frequent shortage of liquidity in developing countries (Stiglitz 2003), while the second looks at the occasional injection that mainly targets the developing countries (Soros 2002). The second is intended to be an allocation beyond the scope of the regular SDR. It might have elements of regular SDRs (allocation by quota) with a modification coming by way of donations to other users. Further extensions of this might take the form of alterations to the agreed quota as was, indeed, negotiated in 1997. It might, therefore, favour some countries and institutions, namely those that may be assessed to be in need of development assistance well beyond what traditional assistance packages may afford.

The IMF, as an institution, however, has tended over the years to view proposals to finance specific development initiatives with SDRs more cautiously, suggesting that the use of SDRs for such purposes would generally require a change in the Articles of Agreement, except if industrialized countries voluntarily transferred their SDRs to other countries. The Fund has always observed that there was nothing preventing countries from engaging in such voluntary transfers. On the proposals to supplement fund resources, the Fund has responded that:

To the extent that these proposals involve balance-of-payments financing with conditionality they can be viewed as essentially substituting for an increase in IMF quotas or IMF borrowing. The key difference among them is the degree of IMF involvement in intermediating redistributed SDRs, and the implications of this for conditionality and the assumption of credit risk. (IMF 2002b)

The IMF places SDR proposals into two broad categories as follows:

  • Proposals to supplement Fund resources. These proposals seek to direct SDRs allocated to industrial countries to countries with more severe international liquidity needs.

  • Proposals to finance development. These are generally of two types, those that may involve voluntary donations to a prescribed holder or to another country and not requiring a change in the Articles of Agreement; and those that call for a redistribution of quotas.

These proposals may be classified further as those requiring mild reforms in the international financial system or no change to the status quo, and those requiring more radical or substantial reforms that would imply changes to the governance arrangements of the IMF. While the proposals from the Zedillo Panel seem to call for both broad categories of change in the way SDRs were allocated, it may be noted that most of the calls over the years have been related to the first category of change. A slight departure from the category of mild reforms is the Soros proposal, the best-known scheme. A central part (p.99) of the proposal from his book George Soros on Globalization (2002), is that there should be periodic creations of SDRs and that the rich countries should agree to make their allocations available for global public goods and for aid to development in individual countries. He would like to see the process started with the activation of the proposed allocation of 1997. The Soros proposal was intended to achieve additional aid resources in a manner that was more or less automatic. The pooling of funds was also expected to engender proper coordination. In this proposal, SDR donations from the industrial economies would first be paid into an escrow account, and there would be no recorded budgetary cost to them until the SDRs were withdrawn from the account in order to pay for the approved development projects.

Stiglitz (2003) has suggested clearly there is a need for more than just a one-time issuance of SDRs. He calls for a complete overhaul of the global reserve system, which he blames for being at the centre of the failures of the global financial system. His argument is also based on the fact that developing countries' reserves are growing much faster than they can afford to, as they set aside huge reserves against a variety of contingencies, including decreases in foreign investor confidence and declines in export demand. With imports growing at 10 per cent per annum, about US$160 billion has to be set aside each year for reserves, and a part of this could easily be made available for health and education. In advocating more frequent issues of SDRs to finance development, Stiglitz writes that:

Keynes, during the founding of the IMF, envisaged the issuance of ‘global greenbacks’, more familiarly known as Special Drawing Rights (SDR) … Global greenbacks could be used to finance global public goods, such as improving the environment, preventing the spread of diseases like AIDS, increasing literacy in the developing world, and providing humanitarian and broader development assistance. (Stiglitz 2003: 57)

Some countries may receive more than they put into reserves, which they can exchange for conventional currencies, while countries receiving less than they put into reserves may supplement these reserves, freeing up money that would otherwise have been set aside.

Richard Cooper (2002) has proposed an amendment of the Articles of Agreement to allow the IMF to create SDRs on a large enough and temporary basis to counter financial crises and to forestall creditor panic (Fig. 5.2). Interestingly, while the Monterrey Conference on Financing Development discussed the various proposals for using SDRs among other schemes, it did not provide support for any particular scheme.

A Development-Focused Allocation of the Special Drawing Rights

Figure 5.2. Proposed SDR flows


The discussion here focuses on the perceived role of SDRs within the context of the debate for and against the issuance of such development focused SDRs. Noting the general concern about the effect of development SDRs on international stabilization, Clunies-Ross (2002) poses the question of whether the use of specific-purpose SDRs for allocating resources to global public goods is consistent with their use for stabilization purposes as originally intended, and if so what should be the nature of the institutional (p.100) changes required to bring that about. His response to these questions is that, with minor modifications, it is possible to issue development SDRs for supplementing aid and providing global public goods in a manner that does not compromise international stabilization.

The idea that development SDRs will not harm global stabilization is even more strongly put across by both Clark and Polak (2002) and also by Stiglitz (2003). The latter has written that: ‘this scheme would not be inflationary: rather it would offset the inherent downward bias of the current regime. Relative to global income—some US$40 trillion—the magnitudes of monetary emissions would be minuscule’ (Stiglitz 2003: 5).

There appear to be quite strong arguments for the issuance of SDRs, both the regular and the specific. But the arguments against cannot be dismissed.

5.4.1 The Argument for a Development SDR

Stiglitz sums up his discussion of the need for development SDRs as follows: ‘In effect, these reserves are a commitment of the world to help each other in times of difficulty … This policy would end the logic of instability that is built into the current system, for it would allow some deficits without inevitable crisis’ (2003: 5).

As noted earlier, the idea behind the calls for these new issues of SDRs is basically to provide developing countries an opportunity to devote resources that would otherwise have been devoted to enlarging reserves to providing services that facilitate development.

But the calls for development SDRs go beyond the need for less expensive reserves, as we saw with Stiglitz (2003). The growing calls for the creation of development (p.101) SDRs are also closely associated with the need for a faster development of global public goods a la Soros (2002) and Stiglitz (2003). The argument is that effective delivery of global public goods enhances the achievement of the development goals of poor nations. Some of the more comprehensive argumentation for the creation of development SDRs has been put together by Clunies-Ross (2002). In his preamble to the arguments for considering various ways for mobilizing resources for social and economic development, including development SDRs, he indicates that the methods and resources must be (i) technically and administratively accessible; (ii) unlikely to impose any unduly high excess burden of costs through misallocation; (iii) equitably distributed; (iv) not politically out of the question for ever, and (v) so far not fully exploited. Clunies-Ross (2002) sees the attraction of SDRs as a source of globally available funds, first, in the fact that it is created by an international institution, hence belonging to the world as a whole, and making it useful for the maintenance of the global public goods for global stability and full employment of resources. Second, he anticipates that in certain circumstances the rise in the resulting world income and output may either equal or exceed the value of the funds assigned for the purpose. Third, since SDRs come to national monetary authorities without payment, and their uses are somewhat restricted, those countries with more than adequate reserves could give them up for global causes while incurring minimal overt sacrifices (Clunies-Ross 2002: 26).

Wade (2002) has had an interesting look at the Soros proposal for new SDRs, and notes that, the ‘proposal focuses public attention on apparently arcane monetary issues, which have a huge impact on the performance of the world economy yet receive rather little public attention’. He suggests that the injection of a modest amount of US$27 billion of SDR equivalent under the proposal for the first issue makes it doable, and this could produce better performance from the world economy as a whole. Wade (2002) sees the link made by Soros between monetary/payment issues and the supply of global public goods as credible ‘and the mechanism of choosing which goods will be supplied, by whom, and financed by whom, is an interesting one when put alongside the present arrangements’. In Soros' proposal there is a group of eminent persons or an independent jury, working with a trust fund, and chosen for the purpose of deciding which global public goods may be funded. This is not left for the World Bank, currently the most significant supplier of global public goods, to determine.

Wade (2002) notes that even though the proposal is quite modest in scale, it could easily be enlarged to make it a significant contributor ‘to solving the chronic tendency in the world economy at large towards excess capacity reflecting insufficient demand’. If the possible uses of SDRs were broadened, Wade expects the allocations of these development SDRs, which favour poor nations, to raise their consumption significantly. He expects such a growing consumption to lead largely to an increasing demand for goods from industrial economies. In effect, industrial economies do not necessarily lose by consenting to the creation of the development SDRs.

Having observed that international agreement on assistance to poor nations often takes too long to achieve, Clunies-Ross (2002) notes that the Soros proposal for regular SDRs to be issued to member countries, with the richer ones among them donating (p.102) theirs to aid and global public goods, has the advantage of ensuring that allocations to such aid and global public goods do not have to wait for universal agreement.

5.4.2 The Argument ‘Against’ Development SDRs

The arguments against the issuance of development SDRs are basically an extension of the arguments made against regular SDRs. These have been summarized by Lissaker's in the Pocantico Report (2003) as (i) legal, referring to the requirement that there must be long-term global need to supplement international liquidity; (ii) moral, in that non-conditional financing encourages bad policies; (iii) efficiency, implying that the SDRs are unlikely to go to the developing countries; and (iv) historical, suggesting that the environment of floating exchange rate regimes and the existence of multiple media for reserve holdings made SDRs unnecessary. But these are the points that Clark and Polak (2002) strongly argue against. In effect, the main arguments against the issue of development SDRs on account of international stabilization derive from the anxiety that not enough measures are being put in place to restrain the IMF from ‘flooding’ the financial markets with excess new liquidity and cash-strapped poor nations will go on a spending spree.

In addition to the above arguments, most of the points of concern that have often been raised about the use of SDRs for development are centred on a couple of issues. First, it is not obvious what advantages the use of SDRs brings to poor developing nations that are not derived from traditional aid packages. And second, the impact on the stability of the world economy is not clear, and finally, the institutional adjustments that are required for creating the special SDRs that affect country quotas are probably not politically appealing to industrial economies and hence not likely to happen.

Wade (2002) points out the fact that Soros is not clear on what the poorer countries would do with their SDR allocations. He notes that beyond the effect to be generated by the global public goods, it is not obvious what direct benefits countries would receive. ‘Presumably, the direct benefits are those they could have had all along from conventional SDRs, and the Soros proposal does not contain anything new in this respect.’

But Clunies-Ross discusses even more extensively the difficulty in identifying the potential benefits from development SDRs. He sees a possible problem with generating enough desire on the part of potential SDR holders to want them, particularly when the recipient of the SDRs is not a central bank but a trust set up to administer global public goods after these allocations have been transferred to the trust by a recipient central bank. The problem arises from the obligation of interest payments on total cumulative allocation. The interest would, in that case, have to be transferred to whoever held the donated SDRs or spent their proceeds. Thus, the ‘development SDRs’ would come to the governments or other agents representing their beneficiaries not in the form of grants, but rather in the form of low-interest term loans. He explores a number of possible uses of the development SDRs as follows:

If the global fund were to provide guarantees, for example, of markets for new drugs or vaccines, these assets could be held costlessly against the call to make good the guarantees without having (p.103) in that contingency to raise large amounts in the financial markets. Or, if one of the functions of the fund were to reduce the debt burden of highly indebted poor countries, buying out their debt by exchanging it for the proceeds (and the obligations) of SDRs could be valuable to these debtor countries even though it left them with interest obligations at international short-term rates. The servicing obligations would be much more favourable than those attached to many of the loans that might be available to developing and transitional economies (though less so than those on IDA loans). (Clunies-Ross 2002: 29)

He admits, however, that this ‘quirk of SDR arrangements does take some of the attractiveness from the Soros and similar schemes’ (Clunies-Ross 2002: 28).

It has been suggested that new proposals for the creation of development SDRs ignore the question of how to entice the US government into accepting the proposals (Wade 2002). In view of this, it is obvious that if the new SDRs are seen as additional grants to developing nations the US government will not be enthused about it. Aside from its reluctance to see a reduction in the demand for dollar reserves, the US government is also not very interested in seeing the IMF become a central bank to the world.

Indeed, Wade (2002) also suggests that the Soros proposal does not say enough on the governance issues related to the creation of SDRs. The lack of clarity leaves the question of how the global public goods would be prioritized unresolved. In the Soros proposal, infectious diseases, judicial reform, education, and bridging the digital divide are prioritized for attention. But this is observed to be different from the priority list of the G7 countries. For the World Bank, infectious diseases, environmental improvement, trade promotion, and greater financial stability should be at the top of the list for global public goods. Under the New Partnership for African Development (NEPAD), the priority areas are given as (i) peace, security, and governance; (ii) investing in Africa's people; (iii) diversification of Africa's production and exports; (iv) investing in ICT (information and communication technology) and other basic infrastructure, and (v) developing financing mechanisms. It is obvious that several different lists are possible and accommodating all of them may lead to a longer list that may not easily lend itself to feasible agreed actions. This point is buttressed by the fact that the Stiglitz list is only a subset of all of these.


A number of the proposals for the issue of development SDRs do not provide details of how this can be done. A good example of this is the Stiglitz (2003) paper. Details of what mechanisms may be applied in creating new SDRs for development can at best be pieced together from several sources, and this is done in this section.

5.5.1 The IMF Articles of Agreement and the Creation of a Development SDR

A complication would arise in the creation of development SDRs if the proposed scheme were to make an allocation in other than exact proportion to IMF quotas. This (p.104) would require an amendment of the IMF Articles of Agreement, and the difficulty of meeting this requirement would, therefore, impede the use of SDRs in ad hoc schemes intended to benefit particular groups of countries. If there were no changes in the allocation quotas, in order for high-income recipients of SDRs to give them up simply as grants to others would require no alteration to the Articles. It has been pointed out by Boughton (2001) that the Fund's executive board has already altered the working practice to allow recipients of SDRs to use them as grants. The Fund is, indeed, in a position to name a global fund that receives the grants from SDR recipients as a ‘prescribed holder’, as has been done for some international institutions. This will allow the global fund to hold SDRs and also to trade them with central banks.

As indicated earlier, however, someone must bear the net interest cost of the SDRs once they have been spent on global public goods or spent by recipient developing countries. If it is not to be the donor countries themselves, then it must presumably be the bodies (recipient governments or other agencies) that do the spending. From their point of view, spending of the proceeds would in that case be like spending financed by a low-interest loan. On the other hand if the donors were to bear the full interest burden, the transfers of SDRs would entail the same costs to the donors themselves as equal amounts of official development assistance (ODA) given in the form of grants, and there would be no advantage in undertaking the process through SDRs.

Clunies-Ross (2002) discusses at length various scenarios for dealing with the interest payment issue, if industrial countries were to decide to follow the Soros proposal for donating any new issues of SDRs immediately to a global fund. The likely outcome would be that the wealthier nations would essentially be providing only a termless loan at a standard short-term interest rate to the global fund or poorer recipient countries, and this may be utilized in any of the manners discussed earlier. It is, indeed, the lack of clarity of whether or not donor countries should continue with the interest payment that is the main issue. Making donor countries pay the interest effectively turns the SDRs into additional financial assistance to the recipient.

5.5.2 Conditions for Creating Development SDRs and Links to Debt Relief

It is important to point out that the Soros proposal for donations was intended to overcome the difficulty that an alteration to the SDRs allocation by quotas was considered unlikely to happen soon. While the quotas remained, voluntary donations were thus a way of ensuring less antagonism from some industrial economies. Clunies-Ross (2002) suggests, however, that eventually there might be possibilities for other allocations, as was indeed the case with the (unratified) 1997 Fourth Amendment of the Articles of Agreement, and that considerable discretion might be given (under guidelines) to the management of the Fund.

If there is to be a special allocation outside the quotas, who should be eligible for it? In the Soros proposal, the SDRs going to the global fund are all from donations, so there is not a major problem. But it should be possible to link the gains from the (p.105) new development SDRs to current initiatives on debt relief and the attainment of the MDGs. If the concern with the new SDRs is the likely effect of rapid expansion in international liquidity, one could limit the consideration for special allocation to the heavily indebted poor country (HIPCs) only, instead of the low-income and middle-income countries, as proposed by Clunies-Ross. It is observed from the enhanced HIPC Initiative that countries will continue to borrow even as they receive relief in order to settle other obligations in the pursuit of poverty reduction goals. It is important that payments on these do not slow down growth. It is obvious that debt relief must be recognized by creditor countries as additional to new and increased ODA with a focus on enhancing and sustaining both growth and poverty reduction explicitly. New SDRs that provide HIPC countries with an opportunity to engage in further debt exchanges certainly enhance the benefits of their involvement in the HIPC Initiative.

There is tension between quick debt relief and comprehensive country-owned poverty reduction strategies. The solution to this problem is to make countries focus on their medium- and long-term development frameworks, showing the anticipated growth paths and how these promote poverty reduction. In this situation, new SDRs that create new comfort levels in reserve holdings allow countries to pursue development programmes of a longer orientation than the shorter-term programmes of the last two decades under the direction of the Bretton Woods institutions. Essentially, the development SDRs will provide them with termless credit facilities that have somewhat reduced short-term interest rates. In effect, specific or development SDRs may be essential so long as there are HIPC countries that need greater liquidity than their trade volumes would permit and want greater flexibility in the international financial markets.

Linked to the conditions for creating development SDRs is the issue of their timing. The report of the Zedillo Panel (UN 2001) indicates that there should be an immediate resumption of the issue of regular SDRs. Throughout that report, there is emphasis on the need to augment financial support to developing countries in a manner that deals with their cyclical problems related to the nature of their engagement with world trade. When compounded with the cyclical nature of international capital flows, especially private, the need to develop mechanisms that are counter-cyclical cannot be overemphasized.

5.5.3 Proposed Institutional Re-organization for Creating a Development SDR

The forms that development SDRs may take are often linked to the question of whether the IMF should behave increasingly like a central bank. Mohammed has noted that: ‘developing countries would continue to press for an exploration of the merits of establishing an effective international lender of last resort, that is, one able to create international liquidity freely and to deploy it rapidly to deal with widespread financial crises’ (2000: 201).

(p.106) If SDRs are to be used actively to provide global public goods or aid to developing countries, it is argued that the IMF needs to have the authority and structure to take decisions and act upon them (though probably under imposed guidelines) as independent central banks do. While the Soros proposal deals with the problems of providing support for development largely through informal agreements to transfer reserves from one party to another, there appears to be a major demand from developing countries for the formalization of such arrangements (Wade 2002). Clunies-Ross (2002) suggests that it might be possible to begin with the informal approach to providing additional support for development with SDRs, but eventually to move to a more formalized way of using SDR allocations so as to bring about both more targeted development finance and counter-cyclical activity for international stabilization. This would be effected by giving greater discretionary authority to the Fund's structures, working through the managing director's recommendations, in concert with the executive board, to the board of governors. The managing director might, for example, be empowered to seek authorization for allocation in a particular year to be made to a certain group of countries only.

As the process of selective allocations of SDRs is formalized, it is inevitable that the IMF would have to act a little more like a central bank—and that requires an amendment of the Articles of Agreement: ‘to provide more powers to its existing governing institutions, and possibly also delegation by them, or further amendment, to give more actual decision-making power to say a technical committee [operating of course under guidelines that the governing institutions had laid down]’ (Clunies-Ross 2002).

Putting together ideas from Soros' proposal and sentiments from the Zedillo Panel, Clunies-Ross (2002) outlines five stages that could be followed in promoting the use of SDRs for stabilization and for the provision of global public goods. At the first stage, the Fourth Amendment to the Articles of Agreement is to be ratified by the US Congress in order to allow the 1997 allocation to proceed. This is followed at the second stage by the high-income or industrial economies agreeing to donate their new holdings to an appropriately constituted international body for the delivery of global public goods. That body is to be accepted by all and made a prescribed holder by the IMF. At the third stage, the IMF board of governors agree informally over a period to accept recommendations from the managing director and the executive board to issue and cancel SDRs according to an agreed formula that reflects a relationship between the growth rate of total international reserves and the trend rate of growth of real international transactions. At the fourth stage, the Articles of Agreement are altered to allow the distribution to be done possibly according to criteria other than the prescribed quota, while such alterations are approved each time by the board of governors, following proposals from the managing director and the executive board. Finally, the Articles of Agreement could be further altered to create a technical body, which will work on the appropriate quantities of SDRs to be allocated and how these can be timed. It is this body, guided by the executive board and the managing director, which will generate the information necessary for decisionmaking by the board of governors.


There are two issues that we address in this concluding assessment, namely whether development SDRs are likely to bring significant additional benefit to developing countries, and the manner in which new SDR issues can be managed.

5.6.1 Development SDRs versus Other Forms of Assistance

In the work by Clark and Polak (2002) on regular SDRs, they draw attention to what they refer to as ‘the allocation of SDRs versus the provision of conditional Fund credit’. The question suggests that the perceived direct benefits of an expanded SDR allocation to developing countries, even if it is not directed at development, need to be more clearly identified. They are obviously concerned about the question of whether regular conditional credit from the Fund does not deal with the problems that new and regular SDRs are supposed to tackle, namely creating the means for increasing reserve holdings. Clark and Polak (2002) cite the IMF's 1965 annual report, which observed that, ‘ideally countries’ need for additional liquidity could be met by adequate increases in conditional liquidity. In practice, however, countries do not appear to treat conditional and unconditional liquidity as interchangeable'. Clark and Polak (2002) fear that inducing countries to meet their liquidity constraints with increased conditional lending from the Fund would force them to adopt balance of payments programmes that may have largely negative consequences for growth and development. But this does not deal with the substitution of conditional credit by new grants. Ultimately, should new grants not be preferred to new SDRs? Maybe that is what is intended by proposing the new development SDRs.4

As indicated earlier, Wade (2002) questions in his critique of the Soros proposal how developing countries would gain directly from a large increase in SDRs. He then proposes his own list of what could be done with SDRs, but these are a basic extension of what countries would generally do with owned reserves as opposed to borrowed reserves.

Indeed, there are a number of people who have suggested that if the idea behind the campaign for new development SDRs is simply a matter of getting more development assistance for developing countries, then one should tackle the issue directly. In response, it should be pointed out that it is more difficult to argue for increased ODA which involves a significant direct cost to one party, namely the development assistance donor, than for the SDRs, which involve no direct cost to those developed countries.

In our view, given the existence of traditional ODA, the question of whether or not new SDRs should be created should ultimately be tied to the issue of whether such SDRs are a substitute for ODA or complement it. The difference between SDRs and other flows to developing countries lies in the fact that the use of SDRs is determined (p.108) largely by the recipient, and not by the donor. Recipient countries, therefore, are free to use these essentially termless facilities over a period, despite the possible low short-term interest costs if the recipients have to transact business with the development SDR provider. A developing country that receives an SDR allocation targeted at such countries basically has the option of using the opportunities created by the facility to increase spending on items not typically funded by donor ODA. This includes the massively expanded support to the private sector. This characteristic ensures that additional SDRs without a decline in ODA will give developing countries greater flexibility in managing their economies than they possibly will have with ODA only. Using the development SDRs in a complementary manner is what may be useful.

If donor countries accept to continue paying the interest discussed earlier, this, in effect, would be the main form of support for development that they are providing under this arrangement. But such interest payment cannot be expected to continue indefinitely. The time limits will be determined largely by the usual requirements for assisting developing countries to overcome structural deficiencies.

5.6.2 Structuring the Issuance and Management of New SDR

There is the obvious question of how much development SDRs can be issued at a time. Wade's estimate of an equivalent of US$27 billion of SDR as the amount required to satisfy the needs of the Soros' proposal is comparable to the Clark and Polak (2002) suggestion. They propose an annual issue of regular SDRs with an upper limit of 10 per cent of combined quotas, which would yield about SDR20 billion. Assuming that the annual issues suggested by Clark and Polak were utilized mainly for development finance, with industrial economies donating their quota share for the purpose, an amount of US$25–30 billion would be available as an additional resource for development finance. And this is about one-half of the initial additional resources needed for the achievement of the MDGs.

One area in which the creation of SDRs for global public goods is seen to be likely to generate good outcomes for developing countries is governance. Wade (2002) thinks that Soros' ideas of establishing a trust fund with an independent jury, the shopping for recipient programmes, and the addition of SDRs to the reserves of the poorer member countries are important and should be treated seriously. The element of having an independent third party organize the distribution of global public goods reduces the influence of the World Bank in this area and supposedly brings greater transparency and accountability into what poor nations may or may not have. Shopping for recipient programmes ensures that developing countries are not saddled with what some international technocrat believes is good for them.

But even more important is the question of which of the three possibilities should be adopted in resuming SDR issues: (i) should the IMF allocate the SDRs differentially? (ii) should the IMF allocation follow quotas, and should there be an international agreement of donors to transfer according to some formula? and (iii) should the IMF allocation follow quotas and individual members decide how to transfer their allocation? This takes us back to the issue of whether or not the transfer arrangements should be (p.109) more formalized. There seems to be the growing consensus that the third option is the way to start, with a gradual shift to more structured transfers later. This is generally seen by the supporters of the resumption of SDRs as the best chance of gaining wider support.


Bibliography references:

Aboyade, O. (1983). Integrated Economics, A Study of Developing Economies. London: Addison-Wesley Publishers.

Boughton, J. M. (2001). Silent Revolution: The International Monetary Fund 1979–1989. Washington, DC: IMF.

Camdessus, M. (2000). ‘An Agenda for the IMF at the Start of the 21st Century’. Remarks at the Council on Foreign Relations, February, New York.

Clark, P. B. and J. J. Polak (2002). ‘International Liquidity and the Role of the SDR in the International Monetary System’. IMF Working Paper WP/02/217. Washington, DC: IMF.

Clunies-Ross, A. (2002). ‘Resources for Social Development’. Paper prepared for the World Commission on the Social Dimensions of Globalization.

Cooper, R. (2002). ‘Chapter 11 for Countries?’ Foreign Affairs, 81(4): 90–103.

International Monetary Fund (IMF) (2002a). ‘Special Drawing Rights (SDRs): A Factsheet’. IMF Website. Available at: www.imf.org/external/np/exr/facts/sdr.htm. Accessed 20 August.

—— (2002b). ‘Allocations of Special Drawing Rights (SDRs): A Factsheet’. IMF Website. Available at: www.imf.org/external/np/exr/facts/sdrall.htm. Accessed 22 August.

Mohammed, A. A. (2000). ‘The Future Role of the IMF: A Developing Country Point of View’, in J. J. Tuenissen (ed.), Reforming the International Financial System. The Hague: FONDAD.

Mussa, M. (1996). ‘Is There a Case for Allocation Under the Present Articles?’, in M. Mussa, J. M. Boughton, and P. Isard (eds), The Future of the SDR in the Light of Changes in the International Financial System. Washington, DC: IMF.

Pocantico Report (2003). Report of the Conference on Additional Sources of Finance for Development, 29–31 May, Pocantico, Tarrytown, NY. New York: Friedrich Ebert Stiftung.

Soros, G. (2002). George Soros on Globalization. Oxford: Public Affairs.

Stiglitz, J. (2003). ‘How to Reform the Global Financial System’. Harvard Relations Council International Review, 25(1): 54–9.

Triffin, R. (1960). Gold and the Dollar Crisis. New Haven: Yale University Press.

—— (1968). Our International Monetary System: Yesterday, Today, and Tomorrow. New York: Random House.

United Nations (UN) (2001). ‘Technical Report of the High-level Panel on Financing for Development’. Available at: www.tradeobservatory.org/library/uploadedfiles/Technical_Report_of_the_High-Level_Panel_on_Fi.htm. Accessed 20 August.

Wade, R. H. (2002). ‘On Soros: Are Special Drawing Rights the Deus ex Machina of the World Economy?’. Challenge, September–October. Available at: www.findarticles.com/cf_dls/m1093/5_45/91659838/p1/article.jhtml. Accessed 18 June.


(1) The SDR interest rate provides the basis for calculating the interest charges on regular (i.e. to say, non-concessional) IMF financing and the interest paid to members who are creditors to the IMF. The SDR interest rate is determined weekly and is based on a weighted average of representative interest rates on short-term debt in the money markets of the SDR basket countries. The yields on three-month treasury bills serve as the representative interest rates for the United States and the United Kingdom. In keeping with the changes introduced to the SDR basket on 1 January 2001, the three-month Euribor (Euro Interbank Offered Rate) became the representative rate for the euro area, replacing the national financial instruments of France and Germany. The representative interest rate for the Japanese yen was changed from the three-month rate on certificates of deposit to the yield on Japanese government thirteen-week financing bills.

(2) This does not include the emerging markets.

(3) The decision to make a general allocation has to be based on the finding that there is a long-term global need to supplement existing reserve assets. The decision of the board of governors is on the basis of a proposal by the managing director with the concurrence of the executive board, with an 85% majority of the total voting power. Decisions on general allocations are made in the context of five-year basic periods.

(4) Wade (2002) suggests that, ‘it is basically a way to arrange for the rich countries to cough up more grants to poor countries. All the fancy talk is really just an elaboration of this very familiar idea’.