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: 26 November 2020

The International Finance Facility Proposal

The International Finance Facility Proposal

(p.110) 6 The International Finance Facility Proposal
New Sources of Development Finance

George Mavrotas

Oxford University Press

Abstract and Keywords

Proposals for new institutional arrangements for Official Development Assistance (ODA) that exploit techniques for securitisation in the capital market can function as an innovative source for generating the funds necessary for the achievement of Millennium Development Goals (MDGs). This chapter takes as an example of this type of proposal the International Finance Facility (IFF) proposal published in the United Kingdom in January 2003 jointly by HM Treasury and the UK Department for International Development (DFID). It discusses the IFF proposal in detail, describing its main technical details and financial structure, and elaborating on crucially important administration and implementation issues. The proposal is evaluated in terms of both its potential advantages and its shortcomings, as well as suggesting ways to strengthen the proposal further. The key challenges for IFF are summarised in the concluding section.

Keywords:   administration, development funding, evaluation, financial structure, implementation, institutional arrangements, International Finance Facility, Official Development Assistance, United Kingdom


Proposals for new institutional arrangements for official development assistance (ODA) that exploit techniques for securitization in the capital market can function as an innovative source for generating funds necessary for the achievement of Millennium Development Goals (MDGs). The present chapter takes the proposal for an International Finance Facility (IFF), published in the United Kingdom in January 2003 jointly by HM Treasury and the UK Department for International Development (DFID) (HM Treasury–DFID 2003a,b) as a concrete illustration of this type of proposals.1

The IFF proposal could be viewed as part of the ‘Global New Deal: A Modern Marshall Plan for the Developing World’, which was put forward by the UK Treasury in 2002 (HM Treasury 2002) to tackle poverty in the developing world.2

The generic features of the HM Treasury–DFID proposal on IFF are as follows:

  1. 1. A substantial increase in ODA of US$50 billion a year today to US$100 billion per year.

  2. 2. Making a pre-commitment, so that the promises can be ‘banked’.

  3. (p.111) 3. Annual commitments would start from the roughly US$15–16 billion of aggregate Monterrey and post-Monterrey additional sums pledged and would rise by 4 per cent (in real terms) per year.

  4. 4. The increase is of limited duration, timed to achieve the internationally agreed MDGs by 2015.

The above generic features of the IFF scheme seem to suggest that the proposal can be analysed either as a net addition to existing development aid, or as a comparison with ‘straight’ ODA of the same net present value (NPV), or as a comparison with ODA with the same time path.3 In the first case, we have all the benefits of the increased flows, together with some potential problems related to absorptive-capacity constraints of aid recipients. In the second case, we have the benefit of being able to bring forward disbursements. Finally, in the third case, we have the certainty of the flows with the difference being much smaller.

It is notable that the internationally agreed MDGs in Monterrey pose an important challenge to the international community, namely to increase aid flows substantially to meet the MDGs or to think of alternative ways of development financing.4 The joint HM Treasury–DFID proposal on IFF focuses exactly on this important issue.5

In view of the significance of the above interesting and promising IFF proposal, the present chapter will seek inter alia to contribute to the current debate on exploring innovative ways to raise funds to meet the MDGs by:

  1. 1. Discussing in detail the IFF proposal, its main technical details and financial structure (Section 6.2).

  2. 2. Elaborating administration and implementation issues of crucial importance in the current proposal (Section 6.3).

  3. 3. Evaluating the proposal in terms of both its potential advantages (substantial increase in aid flows, predictable and stable nature of flows over the next ten years, tried and tested principle for raising finance in international capital markets, among others) and its shortcomings (heavy reliance on political coordination among donor countries, possible absorptive-capacity constraints in aid-recipient countries related to the substantial increase in aid and continuous commitment on behalf of the donor community towards the implementation of the IFF during the thirty years of its life among others), as well as suggesting ways to strengthen further the proposal (Section 6.4).

  4. 4. Summing up the key challenges for IFF (concluding Section 6.5).


According to the HM Treasury–DFID proposal on IFF, the Facility is specifically designed to achieve both the additional finance and the value for money necessary to reach the MDGs.6 The whole idea is based on securitization structures used extensively in the capital markets, that is, leverage in additional finance by borrowing through bonds issued in the international capital markets against long-term commitments for aid by donor countries. Along these lines, the Facility would essentially frontload long-term aid flows so that the MDGs are reached by 2015. The Facility will be structured so that the bonds it issues can achieve the highest possible ratings. This will result in a cost of leverage, which would be comparable to that achieved by existing multilateral organizations.

At the same time, it is argued that the Facility is designed so that aid effectiveness can be improved substantially by focusing aid disbursement on pro-poor priorities of IFF aid-recipient countries and also by improving the predictability and stability of longer-term aid. This, as has been recently argued, would strengthen recipients' efforts to adopt policies that foster sustainable growth, create the right environment for trade and investment, and ensure that the new resources are used efficiently for poverty reduction.7

The Facility would involve two main parties, the donor countries and the recipient countries. The donor countries will be the shareholders of the IFF and at the same time party to the IFF's Articles of Association; they will also make commitments to provide annual payments to the Facility. Finally, they need to agree to a set of overarching principles in relation to the effectiveness of their aid management and aid policies. On the other hand, the recipient countries, according to the IFF proposal, will be the world's poorest countries. It is notable that the relevant HM Treasury–DFID documents on IFF are not clear at all at this stage regarding which countries will be in the final list of the world's poorest countries eligible for IFF funding; obviously the final decision/list of countries will be reached by discussion between all countries, both donor and recipient, in future meetings in which the final proposal on IFF will be discussed in detail. On this front, the IFF proposal states that countries which received financing in the last IDA replenishment could be potential IFF-recipient countries (although there is not any clear commitment in the proposal regarding this).8

(p.113) It is notable that the IFF will not disburse funds directly to recipient countries. On the basis of the agreed IFF overarching principles (see relevant discussion below) it will instead provide funds for disbursement by existing bilateral and multilateral aid-delivery channels, which may include the World Bank and the Global Health Fund as well as specific agencies of the governments of donor countries (e.g. DFID in the United Kingdom). These agencies, both bilateral and multilateral, would be acting as agents on behalf of the IFF. They would also manage the disbursements in line with the allocation of funds agreed by the Facility.

The life of the Facility will be rather limited, in the sense that the IFF will terminate upon repayment in full of all bonds issued and other liabilities incurred by the IFF. A rough idea on how the Facility would work can be provided by the flow diagram (Fig. 6.1).

The International Finance Facility Proposal

Figure 6.1. Overview of the IFF

Source HM Treasury–DFID (2003a).

The IFF proposal states that the Facility will be created by an international treaty among participating donor countries (the shareholders of the IFF). Issues related to the principal objectives of the IFF, its constitution and governance structure would not be covered by the Articles of Association, which would be negotiated at an IFF founding conference. The overarching principles (OPs) would also be defined by the Articles of Association. The OPs, to which all donors would sign up, would have to be met for the disbursement of all funds raised through the IFF. The plan is for OPs to be agreed by donor countries at the founding conference and they could include conditions such as that funds raised by the Facility must be disbursed to recipients on the basis of sound aid-effectiveness principles as well as the domestic policy environment and need. Although the proposal does not provide (at least at this stage) further details regarding the precise nature of the OPs, it clearly states that the funds should be (i) used for poverty reduction; (ii) not tied to contracts using specific national suppliers; (iii) provided in multi-year programmes of at least three years' duration; (iv) disbursed (p.114) mainly in grant form, including debt relief, with some highly concessional loans where appropriate; and (v) targeted at low-income countries.

The assets of the Facility will consist primarily of donor commitments to provide streams of annual payments to the IFF. Thus, the Facility will be relying on donor commitments to meet its obligations under the bonds. The proposal also states that the IFF's ability to service the debt under the bonds will not depend on any repayment by the recipient countries of the aid disbursed to them.9

Donors will contractually commit to provide streams of payments annually to the Facility. The plan is for each stream to comprise fifteen annual payments (under discussion at the moment). To decide on these commitments, donor countries should meet on a regular basis. One option is for shareholders to participate in pledging rounds on a tri-annual basis. Another option is that donor countries should participate in rolling annual pledging rounds. It is notable that participation in one pledging round would not commit a donor country to participate in following pledging rounds. This is a potential weakness of the proposal in the sense that the continuous commitment of the donor countries to the Facility, absolutely vital for its sustainability and success, is not guaranteed (see Section 6.4 for further discussion).

Furthermore, donors will be contractually bound to make annual payments subject to certain fundamental conditions being met by the recipient countries (called ‘high-level financing conditions’; not precisely defined at the moment). These conditions would be few according to the proposal, clearly defined in advance and capable of independent determination; a possible financing condition is that a recipient is not in protracted arrears to the IMF (International Monetary Fund); defined as continuous arrears for more than six months. Note, however, that an IFF donor will not be legally bound to make its annual payments to the Facility in regard to notional amounts for a particular recipient if that recipient fails to meet the high-level financing conditions.

Also, donors would be separately, not jointly, liable for making their payments to the Facility and they would not have any responsibility for making good on payments for which another donor country had defaulted. The IFF, on the other hand, will have the right to suspend disbursements to programmes if any donor countries are in arrears on any payments due to the Facility.

Regarding disbursements by the IFF, these will be linked to a stream of annual payments. In the case where a recipient country programme disburses over five years, it will be financed via payments from five streams. Furthermore, it would be possible for donors to change the length of a specific stream subject to the minimum length necessary to support a bond issuance. Note also that it is not necessary that every payment profile should be considered as attached to a particular stream. The IFF income will be the result of annual payments made each year by donor countries to the Facility of an amount equal to their share of each outstanding committed stream.


Figure 6.2 shows how donor commitments to the Facility can be made through a number of pledging rounds.

The International Finance Facility Proposal

Figure 6.2. Stylized representation of donor commitments

Source HM Treasury–DFID (2003a).

The above-stylized representation feeds into a broader illustrative model, which assumes the following parameters:

  1. (i) fifteen funding rounds, one every year;

  2. (ii) a defined life-span of thirty years for the IFF;

  3. (iii) disbursements of funds from the Facility increase from US$10 billion in the first year to US$50 billion in 5 years, to remain constant for five years before declining to zero over the final 5 years;

  4. (iv) the average cost of funds for the IFF is 5 per cent;

  5. (v) no more than 85 per cent of the net present value is raised as debt (the leverage limit; see below).

In line with the above assumptions, Figs 6.3 and 6.4 show the funds which could be raised based on annual donor contributions starting at US$3.65 billion, rising to US$44 billion from 2017 to 2020 and thereafter falling (Fig. 6.3) and the resulting profile of IFF bonds outstanding (Fig. 6.4).

The International Finance Facility Proposal

Figure 6.3. IFF inflows (income) and outflows (disbursements)

Source HM Treasury–DFID (2003a).

The International Finance Facility Proposal

Figure 6.4. IFF debt profile during the life of the Facility

Source HM Treasury–DFID (2003a).

The underlying assumption of the illustrative example of IFF income and disbursement in Fig. 6.3 is that donors will commit themselves to provide constant nominal streams of annual payments to the Facility. However, it may be possible that donor commitments will be phased in a different way. It may be preferable, for instance, for donors to provide a more even spread of aggregate payments across the lifetime of the Facility, rising in line with donor income. This will result in the phased streams illustrated in Fig. 6.5.

The International Finance Facility Proposal

Figure 6.5. Illustrative example of IFF income and disbursement patterns—phased streams

Source HM Treasury–DFID (2003a).

It is notable that in line with Fig. 6.5, disbursements as projected fall away sharply after 2015. Most conventional wisdom would suppose, however, that if we were planning a pattern of aid flows, we would not necessarily plan it like this. Inter alia, it would (p.116) rather tend to disrupt economic stability. Added to this is the situation projected in which (according to Fig. 6.5), from 2020 on to 2032, the extra donor contributions (above the 2002 base) gradually rise until they are running at more than US$40 billion a year (in 2002 prices) without a cent of this extra going to the recipient countries; it is all debt-servicing and replenishment of the Facility's reserves.10

Turning to leverage issues of crucial importance in the proposed Facility, in order to achieve and preserve the highest possible ratings of all bonds issued by the IFF, the Facility will limit the degree to which the donor commitments may be levered. More (p.117) precisely, at each disbursement the Facility will allocate a fixed proportion of the donor commitment made by a donor country in the pledging round to that disbursement. This allocation needs to take into account the prevailing cost of long-term debt for the IFF in the donor country's currency and the leverage limit.11

Under the assumption that donor commitments are binding and perceived as credible by financial markets, the IFF leverage would depend on a careful assessment of the likelihood that the high-level financing conditions (see above) are met by the recipients to whom finance raised through the Facility is disbursed. In case the high-level financing condition is that recipients are not in protracted arrears to the IMF (see above) then a careful assessment of the historical data, according to internal work by HM Treasury, could provide a sound basis for preliminary work to evaluate the possible leverage limit.

Internal work by HM Treasury includes some examples to illustrate the above. For instance, the historical record of all IDA-eligible countries' experience in going into protracted arrears with the IMF has been considered and has been used as a basis for projections. Under the assumption that commitments by donors are evenly distributed across the 75 countries receiving funds in the IDA-13 replenishment, it was calculated that it is 99 per cent likely that at least 87 per cent of the NPV of donor commitments would actually be paid to the Facility.12

Relaxing the unrealistic assumption that donor countries are evenly distributed across the IDA countries and in line with the disbursement profile of IDA-13, the (p.118) leverage limit would now need to be lower in order to achieve and maintain AAA equivalent ratings. The estimates in this case seem to suggest that it is 99 per cent likely that at least 80 per cent of the net present value of donor commitments would actually be paid to the Facility, compared to 87 per cent in the previous case.13

It is also notable that although the leverage applied to donor commitments will be restricted by the leverage limit, the full value of donor commitments will be disbursed over time in the sense that a lower leverage limit would affect the volume of funds that could be disbursed in the early years of the Facility but not the cost of these funds. More precisely, as the Facility receives annual payments from donors, the value of donor commitments will rise relative to the funds disbursed. This will have two possible effects: it will either reduce the level of commitments donors would need to make for new programmes or release additional funds that may be disbursed under existing (or proposed) programmes.

A relevant issue to the above discussion of the leverage limit is the rating of bonds. The working assumption in the IFF proposal is that the Facility will operate under a leverage limit that will enable bonds issued by the Facility to secure AAA/Aaa/AAA ratings by Fitch/IBCA, Moody's, and Standard & Poor's, respectively.

The HM Treasury–DFID argue that under the assumption that the Facility operates under a prudent leverage limit (in the range 80–87 per cent) and with AAA credit ratings, it is expected that the IFF's debt will be perceived by the market as a risk comparable to the existing multilateral debt.

A rather dubious feature of the above interesting proposal and, at the same time, an issue that possibly deserves further discussion is related to the ‘high-level financing condition’, that is, the licence for a donor to stop servicing portion of the debt for which it is responsible if the relevant recipient has, for example, defaulted on its obligations to the IMF. It might well be argued that the servicing obligations should be treated instead in the normal way (especially since the failings of the recipient country in no way affect the capacity of the donor to service the loan) and since the cessation of servicing by the donor constitutes in itself no punishment or deterrent for the erring recipient. Building on this it might be suggested, for instance, to drop the high-level financing condition and raise the leverage limit since no one would suffer from this or would be subjected to moral hazard as a result.14


The present section briefly discusses a possible governance structure of the IFF by drawing on an illustrative model used in the internal work by the HM Treasury (p.119) (work still in progress at the time of writing). Obviously, the final governance structure has to be agreed at the establishment of the IFF at the founding conference. One possible governance structure of the IFF might comprise two main committees:

  1. 1. A shareholders' committee comprising all of the shareholders of the IFF, each appointed by a donor country. The committee would be responsible for ensuring that the Facility is managed according to the Articles of Association (see Section 6.2). The committee should meet annually to review the activities of the IFF.

  2. 2. A management committee, which would be appointed by the shareholders, comprising the executive directors (each one to be appointed by the shareholders) in the board of executive directors. The board of executive directors would be responsible for overseeing the overall management of the IFF. It would be also responsible for approving programmes to which the Facility can disburse and ensuring a prudent borrowing strategy such that the IFF retains the highest possible credit ratings (e.g., AAA type ratings). The committee would be chaired by a chief executive who could be appointed for a fixed term by the shareholders by means of a transparent and open process (further details on this not available at the time of writing).

A small team of professional staff might also be appointed by the chief executive to carry out the treasury function of the Facility and to oversee the allocation of funds.

Regarding the voting rights of the members of the management committee, it could be suggested that these rights should reflect the level of participation of the shareholders in the IFF. This would be subject to a maximum limit in the sense that no shareholder can hold more than 49 per cent of the voting rights. In relation to the above, it might be prudent for certain matters to be subject to approval by super-majority and/or veto by one or more shareholders (again, discussion on these issues is still ongoing and a final decision will be reached at the founding conference).

The main duties of the management committee would be to maintain the financial integrity of the Facility, which may involve inter alia:

  1. (i) deciding on the level of leverage to be achieved by the IFF;

  2. (ii) establishing the bond issuance programme of the IFF;

  3. (iii) deciding how much cash reserve the IFF will hold and how this will be managed;

  4. (iv) deciding the appropriate strategy for the allocation of funds across recipient countries to achieve an acceptable spread of risk;

  5. (v) deciding how much finance will be disbursed each year.

Finally, the location of the main offices of the Facility needs to be decided at the IFF founding conference.


6.4.1 Potential advantages of the IFF

Revenue-raising potential and accelerating progress regarding the MDGs

Undoubtedly, a key advantage of the proposed Facility is its revenue-raising potential. Indeed, the revenue-raising potential of the IFF is quite substantial. The Facility could double existing ODA from US$50–100 billion per year during the crucial years of 2010-15, thus allowing the MDGs to be met.15 This is a great advantage of the proposed Facility combined with the plan to disburse grants rather than loans to the recipient countries participating in the IFF, although some concessional loans will also be disbursed where appropriate.16 In view of the above, the scheme can deploy a critical mass of aid flows as investment over the next few years (in line with the worked examples of Section 6.2 when it will have the most impact on achieving the MDGs.

Predictable and stable aid

Another major advantage of the proposed Facility is related to the nature of the aid flows, which will be disbursed to the IFF-recipient countries. The proposal clearly states that the aid flows will be predictable and stable, thus minimizing the negative effects associated with unpredictable and volatile aid. Indeed, unpredictable aid imposes a serious constraint in recipient countries with regard to future public expenditure planning and causes problems related to the achievement of sound macroeconomic management.17

Distinguishing predictable from unpredictable aid flows is also relevant to the issue of aid heterogeneity, which has been neglected (until very recently) by the vast literature analysing the impact of aid in aid-recipient countries. More precisely, until very recently existing evidence on the macroeconomic impact of aid lacked a systematic treatment of the aid disaggregation issue and of the way different types of aid affect key macroeconomic variables in aid-recipient countries. One of the main features of the vast quantitative literature of the effectiveness of development aid in recipient countries has been the employment of a single figure for aid. However, this is likely to be misleading for reaching conclusions on aid effectiveness, since we can distinguish at least four different categories of aid: project aid with a rather lengthy gestation period, programme aid that disburses rapidly as free foreign exchange, technical assistance, and food aid and other commodity aid which adds directly to consumption (Cassen 1994; Mavrotas 2002a, b). To the above four types of foreign aid, emergency or relief (p.121) aid could be added as a separate category, given its increasing importance in recent years (Addison 2000).18

Recent work in this promising research area seems to suggest that aid disaggregation does matter for our overall understanding of the macroeconomic impact of aid in aid-recipients (Mavrotas 2002a, 2003a, b; Mavrotas and Ouattara 2003a, b). The relevant policy lessons for both the donor and aid-recipient communities are quite significant: understanding how aid works, and in particular, how different types of aid work is of paramount importance for designing and implementing policies aiming at improving further aid effectiveness. However, there is a clear need for further work in this promising area at the individual country, regional and global level so that important conclusions and robust policy guidelines can be derived.

Comparative advantage compared to other proposed funding additional to existing ODA: IFF vis-à-vis MCA

An advantage of IFF as compared to other possible ways of increasing aid flows so that MDGs are achieved, and in particular the much discussed recently Millennium Challenge Account (MCA) of the US administration, is that the IFF does not deal with recipient countries on a bilateral basis only, as the MCA does.19 By doing so, the MCA could possibly undermine significant progress made recently in improving donor coordination (see, for instance, the recent Aid Harmonization Initiative on this front). As has been recently argued, ‘backsliding in this area could condemn poor countries to the unhappy position of having to court myriad donors and wade through competing and conflicting regulations’ (Sperling and Hart 2003: 10).

A relevant question here might be, what would happen if the rest of the donor countries, apart from the United States, set up a fund (let us say the IFF) with the United States going its own way (with the MCA, for instance)—which, indeed, is the most likely outcome in view of the recent US (but also German) opposition to the IFF proposal at the IMF–WB Spring Meetings in Washington DC in April 2003 (see further details later). Crowding-out issues are of relevance in this case (i.e. introducing a new revenue source may displace other sources for development funding) under the assumption that the IFF proposal would be finally accepted by some donor countries. A proper discussion of the above issue should adopt a common framework to evaluate (p.122) different forms of development financing including the proposed Facility (Atkinson, Chapter 1, this volume).20

Indeed, this is also relevant to the IFF as compared to the IDA. IDA disbursed US$5.6 billion (US$5 in 2001 [World Bank 2003]) in 2002. Recently there has been a lot of discussion regarding the restructuring of IDA with a part of its lending being in the form of grants rather than loans. The IDA-13 replenishment was approved in mid-2002 and it was agreed that between 18 and 21 per cent of its overall resources would be provided in the form of grants. Bearing also in mind that IDA focuses on low-income countries, the question is, what would happen if the IFF was finally set up to disburse aid (mostly in the form of grants as discussed in Section 6.2 with a focus on low-income countries also). Most likely, the countries receiving aid flows from the IFF would be the same countries receiving grants (but also loans) from IDA. This is an important issue that needs to be addressed by the IFF proposal in view of the above recent developments in the IDA.21

IFF is based on a tried and tested principle for raising finance in international capital markets

It could be well argued that the securitization principle for raising finance through the IFF has been tried and tested for raising finance in international capital markets, thus adding significantly to the robustness of the overall proposal. Needless to say, the overall success of the proposed Facility will also be determined by other factors, such as the number of donors participating in the Facility (as well as their importance as perceived in international capital markets) and their continuous commitment to the Facility during its thirty-year life (see below for a discussion).

6.4.2 Potential Disadvantages of the Proposed Facility and Issues for Further Discussion

Possible undesirable effects of increasing aid flows through the IFF

Doubling aid through IFF may cause some potential undesirable effects in aid-recipient countries such as:

  1. 1. Absorptive-capacity problems (i.e. how recipient countries can cope with high levels of aid) and diminishing marginal rates of return to increased aid;22

  2. (p.123) 2. Dutch-disease type of effects (i.e. the impact of aid on relative prices in aid-recipient countries);

  3. 3. The impact of aid on the public sector in recipient countries (i.e. does aid result in reduced taxation effort in recipient countries? How aid affects the fiscal sector of the aid recipient).23

Absorptive-capacity issues

It has been recently estimated that if aid donors were to meet the ODA target of 0.7 per cent of donor country GNP, aid flows would increase to about US$175 billion, that is more than three times the current levels. This, obviously, would help a lot with the achievement of the MDGs, but at the same time it would pose a number of challenges for aid-recipient countries at both the micro and the macro level (Heller and Gupta 2002). Although not implying an increase of aid flows equivalent to 0.7 per cent of donor country GNP, the proposed IFF is associated with a sharp increase in aid in the next decade up to 2015, which may pose crucial challenges for potential recipients of IFF flows. An important relevant issue is the possible absorptive-capacity constraint in a number of aid-recipient countries.24 More precisely, doubling aid through the Facility would face diminishing marginal rates of return and this could become a serious problem at very high aid-to-GDP ratios. However, doubling aid in real terms by 2005 would bring only fourteen countries with a combined population of 109 million above 20 per cent aid/GNI. At the same time, it would be fair to say that there is no evidence that countries which receive high levels of aid have performed poorly. Some recent success stories in Africa clearly show how large amounts of aid can yield substantial rates of return: Uganda received more than 20 per cent of aid as share of GDP in the early 1990s but managed to register high growth rates above 7 per cent, reducing at the same time poverty (mainly through the Poverty Eradication Action Fund) by 20 per cent. Mozambique, with a 50 per cent aid-to-GDP ratio in the 1990s, achieved high growth rates, reaching 12 per cent in 1998. Recipients with high aid levels (above 20 per cent of income), most of them in Africa, increased on average their per capita GDP by 1.3 per cent per year over the period 1995–2000 (World Bank 2003).25 Needless to say, these trends cannot suggest a robust conclusion regarding the impact of aid on growth (and the overall effectiveness of development aid) in countries receiving huge amounts of aid, since the aid-growth nexus can be affected by a number of factors which are not captured by simple correlation statistics.

(p.124) On the issue of diminishing marginal rates of return related to increased aid, HM Treasury has recently argued that a minimum criterion for public investment, including what comes through aid, is that the test discount rate must not be less than the rate used by the donor country for domestic spending (currently 3.5 per cent in the United Kingdom) or the social time preference rate of the recipient country, estimated at 5.8 per cent for low-income countries. IFF-related assistance at average rates of return of 25 per cent (HM Treasury estimate) would yield returns of more than three times the borrowing rate of 6 per cent, and far above test discount, and social time preference rates.

Dutch disease effects

Regarding the Dutch disease type of effects, it may be well argued that large and sustained aid flows (i.e. IFF-type flows) may cause some appreciation of the nominal and real exchange rates in recipient countries. Indeed, since most aid is provided to governments whose expenditure is mainly on non-tradables such as public services, there is clearly a likelihood of short-run Dutch disease effects. In this case, aid may have two effects: a distortionary effect on price incentives and a direct positive income effect. Obviously, the final outcome will be determined by how the economy responds to the distortion. However, if substantial aid flows are supported by appropriate economic policies in recipient countries, the net gains to higher sustained levels of aid will remain strongly positive and donors should therefore continue to make these resources available to recipient countries pursuing sound macroeconomic policies (DFID 2002).26

The impact of aid on the public sector

An issue relevant to potential absorptive-capacity problems in countries receiving large amounts of aid through the IFF concerns the impact of aid on the public sector in the aid-recipient economy. One of the key criticisms of the aid–growth literature is that it fails to recognize explicitly that aid is given primarily to governments in aid-recipient countries, and hence any impact of aid on the macroeconomy will depend on government behaviour, in particular how fiscal decisions on taxation and expenditure are effected by the presence of aid. This is exactly what motivates the so-called ‘fiscal response’ literature, that is, modelling how the impact of aid is mediated by public sector behaviour.27 The analysis of fiscal response is also important because it helps to open one of the many ‘black boxes’ of the aid–growth nexus (McGillivray and Morrissey 2000; Mavrotas 2002a; Mavrotas and Ouattara 2003a). Long ago it was argued that aid, inter alia, may have a negative effect on recipient economies (p.125) since recipient-country governments often use aid money to increase government consumption rather than direct aid flows towards developmental government investment (Griffin 1970). These potential negative effects of foreign aid could be viewed further within the context of the fungibility literature; the impact of aid on fiscal variables in the recipient economy and the related issue of aid fungibility have been the subject of a booming empirical literature in recent years. The ‘fiscal response’ literature, however, is not conclusive about the overall impact of aid on the fiscal sector of recipient countries.28

Issues related to donor coordination and commitment: the key-challenge

An issue very relevant to the overall success of proposals like the IFF is donor coordination: improving donor coordination (as well as maintaining it throughout the life of the Facility) is absolutely vital for the successful completion of the IFF. This is one of the disadvantages of the proposal, namely its heavy reliance on political coordination among the donor countries that will finally participate in the Facility. At the same time, the underlying assumption in the IFF proposal is that there will be continuous commitment on behalf of the donor community towards the implementation of the IFF during the thirty years of its life. This, to our view, is rather optimistic since, in a dynamic and uncertain world, this type of commitment is rather difficult to guarantee.

The conditions to be attached to the outflows from the Facility would be politically difficult to be agreed by all donors participating in the IFF

The conditions related to the IFF (see the discussion on OPs in Section 6.2) will be rather difficult to be agreed among donors and multilateral agencies participating in the Facility. Since the delivery channels of the IFF will be various (including multilateral agencies and bilateral channels), there is no guarantee that the conditions to be met by the recipients will necessarily be the same in all cases. Indeed, there are plenty of reasons to argue the opposite, in view of the lack of consensus among donors on issues related to the role of conditionality in development aid in the past. This may also undermine one of the main advantages of the current IFF proposal, namely that it will channel predictable and stable flows to low-income countries, thus allowing them to adopt pro-poor growth strategies. If we assume that there will be no widespread agreement regarding the OPs among donors participating in the IFF but that instead different conditions will be adopted by different donors, this may possibly make the disbursements less predictable and stable thus affecting the achievement of MDGs. However, possibly the right route for the IFF to take at this stage is to link the whole effort with recent initiatives on the aid harmonization front.29

(p.126) Would the IFF be large enough (in terms of donor participation) to be able to deliver what is being proposed?

Obviously, the issue (and the main challenge for the proposed Facility) is really whether enough donor governments will commit enough money to get the Facility working to the right scale (Financial Times 2003a). This clearly emphasizes that the biggest task in the implementation of the IFF proposal is political. The proposal does not discuss details regarding possible donor participation in the Facility in terms of the ‘right’ number and/or size needed to raise US$50 billion a year until 2015. However, this is a crucial factor, which will eventually determine the future of this promising proposal. If the HM Treasury–DFID manage to convince a substantial number of donors to go ahead with the proposed Facility, even without the United States as an IFF donor country (in view of the recent developments related to the MCA, see above), the Facility may be in a position to deliver what it is promising at the moment. In the opposite case, the prospects for the Facility would be rather bleak.

How would the IFF be treated in the national accounts of donor governments?: Implications for debt/GDP ratios

Another potential disadvantage of the proposed Facility is related to the way it may be treated in the national accounts of the donor countries participating in the Facility. It may well be argued here that the future commitments of donor countries to the Facility will be registered as government liabilities in the national accounts of the donor countries resulting in higher debt-to-GDP ratios in these countries. This would make the Facility rather unattractive to those donor countries wishing to reduce their debt-to-GDP ratios.

Regarding the above issue, the HM Treasury–DFID argues in the revised proposal (work still in progress) that the IFF is expected to be classified as an international organization (under Eurostat guidelines) and so within in-the-rest-of-the-world sector in national accounts and therefore outside the general government sector in individual countries' national accounts. Along these lines, the proposal goes further by arguing that the Facility will be ‘owned’ by the several donor governments but will be responsible for its own debt servicing, that is, bondholders will have no call on donor governments participating in the Facility. HM Treasury–DFID also argue that donor countries will need to consider whether, under their own accounting conventions, the funds raised through the securitization process would be classified as IFF borrowing rather than as borrowing by the individual donor governments.30

Under the assumption that the IFF will be classified in this manner, there is no potential disadvantage arising from the Facility regarding its treatment in the national accounts of the participating donor governments. However, as the proponents of the Facility also admit, this will crucially depend on two factors, namely (i) the nature of the donor commitments and (ii) the IFF's decisionmaking process for allocating (p.127) funds for disbursement (to be agreed at the proposed founding conference). Needless to say, even on the assumption that there is some ineluctable logic determining how the presentation for national-accounts purposes should be done given the nature of the financial obligations (which might make the first condition important), it is very hard to see how the manner in which the allocation decisions are made with the borrowed funds (the second condition) has any bearing on whether those funds should be counted as donor-government borrowing.31

The IFF may imply a significant change in organizational terms since it would involve a new international treaty and possibly substantial organizational costs

It could well be argued that a new scheme, along the lines of the proposed IFF, implies a significant change in organizational terms since it would involve a new international treaty as well as potentially substantial organizational costs as compared to the existing institutions. One might even go further to argue that it is rather unclear what advantage the new Facility would provide as compared to existing institutional structures. Along these lines, one might possibly wonder whether increasing the callable capital of the existing MDBs could achieve similar results.

However, distinguishing between grants and non-concessional loans is crucial in dealing with the above issue. In the case of non-concessional loans (which is not the focus of the proposed Facility) the most efficient means would probably be an increase in the callable capital of the existing MDBs. This would not be the case if we considered grants (and concessional loans) as in the case of IFF, since the IFF would be a more efficient way to finance grants and concessional loans than the callable-capital structure of the MDBs. For example, IDA cannot borrow in the international capital markets and the International Bank for Reconstruction and Development (IBRD) may not give grants. In the latter case, internal work by the HM Treasury seems to suggest that, if the World Bank itself were to finance loans on IDA-type concessional terms through market borrowing, and wanted to maintain its status as a stand-alone financial institution (without any donor subsidy), it would require an increase in capital at least equal to the grant element of the loans and probably significantly more, reflecting credit risk; this would result in a leverage which would be very low. Thus, it would be fair to conclude that increasing the callable capital of the existing MDBs would be unable to accelerate grant finance the way the IFF does.

Could the additionality of the money committed through the IFF be guaranteed?

Finally, an important issue is related to the additionality element of the proposed Facility regarding existing ODA flows from donor countries. It would be difficult (if not impossible) to guarantee that the money committed through the IFF would be additional to ODA flows. Regarding this, it could possibly be argued that, even if contributions are not entirely additional, the IFF process will bind them in and give (p.128) them leverage. This is clearly an issue that needs further discussion and elaboration in forthcoming G7 meetings.32


In a globalized world, we need solutions of a global scale to deal with important global issues (such as the MDGs). In view of this, a proposal to set up an IFF should, in principle, be welcomed by the international community. In the present chapter, we have discussed in detail the IFF, the joint HM Treasury–DFID proposal to increase development aid substantially so that the MDGs are achieved by 2015. The main conclusion of the chapter is that the proposed IFF is a promising and forward-looking proposal, in view of the substantial increase in fresh, predictable, and stable aid it is envisaging as well as its robust financial structure.

However, there are a number of concerns about some potential shortcomings of the proposal, namely its underlying assumptions of continuous commitment on behalf of the donor community towards the implementation of the IFF during the thirty years of its life, and most importantly its heavy reliance on political coordination among donor countries participating in the proposed Facility. Achieving its huge political task as well as relaxing the crucial constraints on the successful implementation of this innovative proposal would be the main future challenge of the IFF.33

Furthermore, the discussion of potential disadvantages of the proposed Facility, particularly those related to donor coordination and continuous commitment of the donor community during the life of the Facility to contribute to the IFF in line with the Articles of Association, clearly suggests the need for future research to discuss in-depth political economy aspects of sources of development finance, including the IFF, since ‘the political economy of different proposals is an important part of the story, requiring a nuanced political analysis of the likely coalitions of support and opposition, as well as a careful specification of the exact nature of the proposals’ (Atkinson 2003: 22).

To sum up, our own view is that this is a promising, forward-looking, and creative proposal which needs to be carefully considered at international fora in the very near future so that it is turned from a proposal into a practical solution to the problem of development financing for the achievement of the MDGs. Obviously the political constraints regarding the possible implementation of the proposal in the near future are extremely important and should not be overlooked. Many donor countries and international agencies may be sceptical about a proposal which may reduce part of their autonomy and independence regarding aid allocation. Indeed this sort of scepticism is reflected in some of the comments already made by some donors at the IMF–WB Spring Meetings in April 2003. The United States and Germany, for instance, fear that (p.129) the proposal may be impractical since it would bind the hands of future governments and mortgage future aid budgets to repay the borrowings (Financial Times 2003b). Others, however, including the IMF's former managing director Horst Kohler and France, welcomed the proposal thus crediting it with more time for further discussions and revision at this stage.34

At the Pocantico Conference on Feasible Additional Sources of Finance for Development (May 2003) an alternative option was discussed in case binding explicit commitments from all potential major donors proved impossible to achieve. It has been argued that it might be fruitful to ask the potential donors to issue the relevant volumes of bonds on an individual basis, with the proceeds still to be used under whatever ground-rules the donors corporately could agree to follow; each donor country's own credit would then be at issue in maintaining the annual payments. Along these lines, whatever bonds each had issued, these it would certainly service. However, in view of the Facility's multilateral approach to aid allocation, coordination issues could be more difficult to deal with here as compared to the IFF option.

Accelerating aid flows through the IFF needs to take into account absorptive-capacity constraints in aid-dependent economies to ensure that marginal returns do not fall below the minimum thresholds we derived. In view of this, borrowing against future aid, as proposed by the IFF, can be worthwhile provided returns exceed the borrowing cost (Foster 2003).35

The final point to make is that arguing as well as campaigning for the IFF (as is the case with other sources of development finance), even if the proposed Facility finally does not succeed in being adopted in the near future, can be extremely valuable in maintaining a sense of urgency over the need for additional finance for the achievement of MDGs and in challenging the major economic powers to find an efficient way of providing it. Furthermore, coalitions among donors trying to promote new major (p.130) initiatives on the development finance front (such as the IFF) could be possibly more efficient if they included politicians representing both North and South; this would give to the proposed initiatives more international support and credibility as well as higher chances of being finally adopted.


Bibliography references:

Atkinson, A. (2003). ‘Innovative Sources for Development Finance: Global Public Economics’. Paper presented at the World Bank ABCDE Conference, 15–16 May, Paris.

Addison, T. (2000). ‘Aid and Conflict’, in F. Tarp (ed.), Foreign Aid and Development: Lessons Learnt and Directions for the Future. London: Routledge Studies in Development Economics 17.

Brainard, L., C. Graham, N. Purvis, S. Radelet, and G. Smith (2003). The Other War: Global Poverty and the Millennium Challenge Account. Washington, DC: Centre for Global Development and Brookings Institution.

Cassen, R. (1994). Does Aid Work?, 2nd edition. Oxford: Oxford University Press.

Catholic Relief Services (CRS) (2002). ‘Improving Development Effectiveness: Recommendations for the Millennium Challenge Account’. Report, June.

DFID (2002). ‘The Macroeconomic Effects of Aid’. A policy paper by the Department for International Development. London, December.

Financial Times (2003a). ‘G-7 to Hear Brown's Anti-Poverty Plan’. 21 February.

—— (2003b). ‘Brown Plan to Reform Global Financial System Meets Opposition’. 14 April.

Foster, M. (2003). ‘Criteria for Assessing the Case for Overseas Aid: A Note’. Development Policy Review, 21(3): 293–300.

Griffin, K. (1970). ‘Foreign, Domestic savings and Economic Development’. Oxford Bulletin of Economics and Statistics, 32(2): 99–112.

Heller, P. and S. Gupta (2002). ‘More Aid—Making It Work for the Poor’. World Economics, 3(4): 131–46.

HM Treasury (2002). Tackling Poverty: A Global New Deal—A Modern Marshall Plan for the Developing World. London, February.

HM Treasury–DFID (2003a). International Finance Facility. London, January.

—— (2003b). International Finance Facility: A Technical Note. London, February.

Kapur, D. (2002). ‘Do As I Say Not As I Do: A Critique of G-7 Proposals on Reforming the MDBs’. Working Paper No 16. Washington, DC: Centre for Global Development.

Levy, V. (1987). ‘Anticipated Development Assistance, Temporary Relief Aid, and Consumption Behaviour of Low-Income Countries’. Economic Journal, 97(June): 446–58.

Mavrotas, G. (2002a). ‘Foreign Aid and Fiscal Response: Does Aid Disaggregation Matter?’. Weltwirtschaftliches Archiv, 138(3): 534–59.

—— (2002b). ‘Aid and Growth in India: Some Evidence from Disaggregated Aid Data’. South Asia Economic Journal, 3(1): 19–49.

—— (2003a). ‘Which Types of Aid Have the Most Impact?’. WIDER Discussion Paper No. 2003/85. Helsinki: UNU-WIDER.

—— (2003b). ‘Assessing Aid Effectiveness in Uganda: An Aid Disaggregation Approach’. Report to the UK Department for International Development, January.

—— and B. Ouattara (2003a). ‘The Composition of Aid and the Fiscal Sector in an Aid-Recipient Economy: A Model’. WIDER Discussion Paper No. 2003/11. Helsinki: UNU-WIDER.

(p.131) —— —— (2003b). ‘Aid Disaggregation, Endogenous Aid and the Public Sector in Aid-Recipient Economies: Evidence from Côte d'Ivoire’. WIDER Discussion Paper No. 2003/15. Helsinki: UNU-WIDER.

McGillivray, M. (2000). ‘Aid and Public Sector Fiscal Behaviour in Developing Countries’. Review of Development Economics, 2(2): 156–63.

—— and O. Morrissey (2000). ‘Aid Fungibility in Assessing Aid: Red Herring or True Concern?’. Journal of International Development, 12: 413–28.

Radelet, S. (2003). Challenging Foreign Aid: A Policymaker's Guide to the Millennium Challenge Account. Washington, DC: Center for Global Development.

Sperling, G. and T. Hart (2003). ‘A Better Way to Fight Global Poverty: Broadening the Millennium Challenge Account’. Foreign Affairs, 82 (March/April): 9–14.

White, H. (1992). ‘The Macroeconomic Impact of Development Aid: A Critical Survey’. Journal of Development Studies, 28: 163–240.

World Bank (2003). Global Development Finance 2003. Washington, DC: World Bank.


The chapter is part of the UNU–WIDER and UN–DESA project on Innovative Sources for Development Finance directed by Tony Atkinson, Nuffield College, University of Oxford. The views expressed in the chapter are those of the author and do not necessarily represent UN–DESA and UNU–WIDER. I am most grateful to Tony Atkinson, Ernest Aryeetey, Nicholas Vaughan, and two anonymous referees for comments and suggestions which substantially improved the Chapter. I am also indebted to Tony Addison, Yilmaz Akyuz, Tony Burdon, Andrew Kenningham, Ian Kinniburgh, Tony Shorrocks, and Rachel Turner for helpful background discussions. I would like also to thank the participants of the project meeting on Innovative Sources for Development Finance (WIDER, Helsinki, May 2003) as well as the participants of the Sharing Global Prosperity WIDER Conference (Helsinki, 6–7 September 2003) for their constructive comments and suggestions. None of the above should, however, be held responsible for the views expressed in the chapter.

(1) The scheme has the full support, at the time of writing, of the United Kingdom and France while other donors are still considering their position (Report of Pocantico Conference, New York, 29–31 May 2003).

(3) Note that ODA flows are defined from the perspective of the donor countries.

(4) See Atkinson (2003) for a detailed discussion on this.

(5) It is notable that the IFF is an ongoing and evolving process at the time of writing thus the present study cannot be considered as a final assessment of the proposal since its technical details, implementation procedure, and overall administration structure can change dramatically in the very near future. Indeed, HM Treasury and DFID are currently trying to iron out the technical details of the current proposal in the aftermath of discussions at the G7 Meeting in Paris in February 2003 and the IMF–World Bank Spring Meetings in April 2003.

(6) The present section (as well as Section 6.3) draws heavily on recent documents published by the HM Treasury and DFID on IFF (see www.hmtreasury.gov.uk/documents/international_issues/int_gnd_intfinance.cfm) as well a technical note from the HM Treasury–DFID not in the public domain at the time of writing. It is notable that there is no other bibliography on IFF (external to the HM Treasury–DFID) at present apart from the relevant documents published by HM Treasury–DFID as well as press releases following the Chancellor's presentation of the Facility in recent G7 Meetings in Paris (February 2003) and Washington DC (IMF–World Bank Spring Meetings, April 2003).

(7) Speech given by the Chancellor of the Exchequer, Gordon Brown, at the Financing Sustainable Development, Poverty Reduction, and the Private Sector Conference, London, 22 January 2003.

(8) In view of the fact that the IFF is designed to help achieve the MDGs, and since its niche is in grants, this would suggest the poorest countries.

(9) It is notable that the bonds issued by the IFF will be backed by the aggregate of all donor commitments; this will essentially provide bondholders with a direct claim against the IFF, if it fails to meet its obligations under the bonds.

(10) We would like to thank an anonymous referee for raising this issue.

(11) By the leverage limit it is assumed that only a certain proportion of the donors' commitment would be allocated to loan servicing; the rest would be a reserve kept in the Facility, principally to cover donor default, either under the ‘high-level funding condition’ or otherwise.

(12) This means that if the IFF were to borrow 87% of the net present value of donor commitments over the next fifteen years, there would only be a 1% chance of the IFF not having sufficient funds to repay its borrowings.

(13) Clearly, the larger the portfolio of commitments and the more even the distribution, then the greater the leverage, all other things being equal. However, it is rather unrealistic to assume that all countries have the same probability of entering protracted arrears with the IMF. Countries receiving a high proportion of funds within the IDA-13 replenishment would tend to have a lower probability of entering protracted arrears with the IMF. Hence, while the more unequal distribution within IDA-13 would tend to lower the leverage, this would be offset, to some extent, by the better creditworthiness of countries receiving an above-average proportion of funds.

(14) We thank an anonymous referee for alerting us on this issue.

(15) This is equivalent to an average aid ratio of 0.47% of GNP, a clear departure from current levels.

(16) This is an important comparative advantage of IFF in accelerating grant finance in view of the need to make substantial progress regarding the MDGs.

(17) Empirical findings of an important early study on the negative effects of unanticipated aid (Levy 1987), clearly indicate different tendencies of anticipated and unanticipated aid in thirty-nine countries over the period 1970-80: unanticipated aid was fully consumed but more than 40% of predictable aid was invested thus contributing significantly to the growth process in recipient countries.

(18) There are three relevant points here: (i) different types of aid operate in different ways (and with different lag-structure) in the recipient country thus resulting in different macro effects; (ii) because of different conditions regarding each in different countries (e.g. the state of aid coordination may vary among aid recipients), there is also an extra reason to expect different effects of aid in each country—the ceteris paribus assumptions of the econometrics of aid may be disturbed by such considerations; (iii) and perhaps most important, aid disaggregation might not matter so much if the proportions of different types of aid were constant; if they were, there could be sense in a coefficient relating aid (measured by one number) to some aspect of effectiveness; but if the proportions are changing, as they are, and changing in different degrees for different countries, this will definitely disturb the empirical results (Mavrotas 2002a, b).

(19) The US administration has recently proposed the creation of a federal corporation, separate from the Agency for International Development, to administer its proposed fund by targeting a small group of ‘high-performing’ countries on the basis of some eligibility criteria. For a detailed discussion of MCA see Radelet (2003), Brainard et al. (2003), and Catholic Relief Services Report on MCA (2002) among others.

(20) Obviously, in considering the likely extent of crowding out, the specification of the counter-factual is also crucial (Atkinson 2003).

(21) The reader can refer to a recent paper by Kapur (2002) on MDBs and IDA for a detailed discussion of IDA restructuring and the ‘aidization’ of the World Bank.

(22) It is notable that even if aid increases substantially by US$50 billion a year up to 2015, as the IFF proposal predicts, this does not necessarily guarantee that the MDGs will be reached. This is also relevant to absorptive-capacity issues; needless to say, this might be also the case with other types of development financing under discussion in the international community at the time of writing. There is some degree of optimism in the IFF proposal regarding this issue. It is assumed that by doubling aid flows through the Facility the targets will be met. However, we should be a little sceptical about this, in view of the lessons emerged from the voluminous literature on aid effectiveness, and particularly recent developments on this front.

(23) The above issues may be also relevant to all forms of increased transfer; for example, the absorptive-capacity issue, the Dutch disease problem, and issues related to fiscal response could all arise if the transfer was funded by a Tobin tax.

(24) Absorptive-capacity constraints and recipient need may be more unpredictable and uncontrollable than the steep bell-curve of rising and declining aid flows during the life of the Facility that the IFF seems to suggest. Heller and Gupta (2002) have recently argued that donors should put funds in trust today in order to finance increased aid in future when absorptive capacity allows.

(25) This is lower than China and India but greater than the average for all low-income countries (World Bank 2003).

(26) Potential Dutch disease effects of increased aid flows are also related to the classical ‘resource transfer problem’ that has been discussed extensively in the trade literature. In the case of IFF, since the pattern of transfers is advanced, the terms of trade effect may become more conspicuous relative to a smoother time path; this may not necessarily outweigh the advantages of earlier disbursement, however, it needs to be put into balance (see Atkinson, Chapter 2 in this volume).

(27) The term is attributed to White (1992).

(28) See McGillivray and Morrissey (2000), McGillivray (2000), Mavrotas (2002a), and Mavrotas and Ouattara (2003b) for a detailed discussion of the relevant literature.

(29) See the High-Level Aid Harmonization Meeting in Rome, February 2003.

(30) In the former case, it is argued that the Facility's disbursement of those funds would be recorded as IFF expenditure rather than expenditure by the donor governments and government expenditure would be recorded when donor governments make their actual payments to the Facility.

(31) We are grateful to an anonymous referee on this point.

(32) Furthermore, it is not very clear from the above proposal what happens after 2015 when a part of the aid budget is diverted into funding the bonds that have been issued to support increased transfers prior to 2015; in particular, it is not certain whether this intertemporal change in resource transfer will increase welfare.

(33) As has been correctly argued recently, ‘unless there is the commitment to implement the IFF, the Facility will remain just a proposal’ (Financial Times, 21 February 2003a).

(34) The recent response of the IMF to the IFF proposal was overall positive, as reflected in the comments made by its managing director at the recent Spring IMF–World Bank Meetings in Washington, DC: ‘I think this is an intelligent idea. It raises a number of questions. I know that there is some skepticism, but I do think we should remain open to this suggestion, because we need more financing for development. We know that public budgets are tight and, therefore, we should also look to some creative ideas, and this is one.’ Press conference at the IMF following the Spring Meetings, Washington, DC, April 2003 (www.imf.org/external/np/tr/2003/tr030410.htm).

(35) Another fair point (raised by an anonymous referee) is that behind the IFF proposal seems to lie a supposition about a rather strange pattern of discount rates (and productivity rates on investment) over time within the recipient countries. The pattern of disbursements (as illustrated in Figs 6.3 and 6.5) seems to make sense only on the assumption that the marginal productivity-of-investment function moves sharply downward from 2015 to 2020. Although it makes sense (in view of the need to meet the internationally agreed MDGs by 2015) to assume an upward sloping marginal productivity-of-investment function prior to 2015, it is not necessarily reasonable to assume a downward sloping curve after 2015. The present chapter examines whether returns on investment in the recipients are likely to be high enough to justify borrowing at AAA rates to finance them and concludes that they very probably are. However, one of the distinctive features of the IFF is that it not only increases disbursements up to 2015 but also (in comparison with a straight disbursement of the projected donor contributions as they occur) reduces them thereafter. Do we know of any good reason why that is likely to be efficient? Or is it just an inefficiency that is worth accepting because the whole package will generate more predictable donor contributions, one of the main advantages of the above proposal?