IDA ELIGIBILITY, TERMS AND GRADUATION POLICIES International Development Association January 2001 Acronyms and Abbreviations AfDF African Development Fund AsDF Asian Development Fund CAE Country Assistance Evaluation CAS Country Assistance Strategy CPIA Country Policy and Institutional Assessment DOD Debt Outstanding and Disbursed ESW Economic and Sector Work FDI Foreign Direct Investment FSU Former Soviet Union GNI Gross National Income HII Higher Income IDA HIPC Heavily Indebted Poor Country IBRD International Bank for Reconstruction and Development IDA International Development Association IDB-FSO Inter-American Development Bank-Fund for Special Operations IFC International Finance Corporation MIGA Multilateral Investment Guarantee Agency NGO Non-Governmental Organization OECS Organization of Eastern Caribbean States PPP Purchasing Power Parity PRSP Poverty Reduction Strategy Paper UNDP United Nations Development Program Table of Contents 1. IDA Eligibility ...............................................................................................................1 Per Capita Income.........................................................................................................2 Creditworthiness............................................................................................................3 Performance...................................................................................................................4 2. IDA Terms .....................................................................................................................5 Less Concessional IDA Terms .......................................................................................5 IDA Grants.....................................................................................................................8 Financial Impact of Possible Changes in IDA Terms .................................................10 3. Graduation from IDA.................................................................................................12 Graduation Policy and Its Implementation..................................................................12 The Small Island Economies........................................................................................13 India.............................................................................................................................14 Indonesia......................................................................................................................15 Lowering the Per Capita Income Threshold for Graduation ......................................15 4. Issues for Discussion...................................................................................................16 Annex I, Note on a Possible Composite Social Index....................................................19 Annex II, Annex Table 1 .................................................................................................22 Annex II, Annex Table 2 .................................................................................................23 Annex II, Annex Table 3 .................................................................................................24 IDA ELIGIBILITY, TERMS AND GRADUATION POLICIES 1. Under the IDA12 Agreement, a commitment was made to review IDA's graduation policies by end 2000. Developments since then -- including the enhancement of HIPC, changes in the IDA status of a number of important countries, and renewed innovation and debate about international support for poor countries' poverty reduction efforts -- make it sensible to cast the review in somewhat broader terms to include related considerations of IDA eligibility and pricing as well. Each of these questions -- eligibility, pricing and graduation policy -- give rise to issues with considerable overlap, but it is nevertheless worth treating them separately. 2. This paper is intended to stimulate a discussion of these issues, and of possible proposals for change, as IDA members move towards the 13th replenishment. It canvasses some options for changes in eligibility and pricing in particular, with the objective of increasing IDA's effectiveness and outreach, without jeopardizing its sustainability. But at this point much further debate and reflection are still needed to decide the specifics of how the IDA framework should be adapted to a changing development landscape. This paper is a first step in this process. 3. In approaching these matters, it is as well to be clear about the underlying philosophy which informs the discussion below. The basic objective of IDA is to help poor countries to reduce poverty by growing faster, more equitably and on a sustainable basis. In doing so, IDA provides highly concessional support designed to enable poor countries to move up an escalator of increasingly robust engagement in the world economy, and towards correspondingly normal relationships with international financial markets. IDA is, therefore, a transitional instrument of concessional support, from which most countries should over time graduate; it is also predominantly a lending institution, based on a strong framework embodying performance-based allocation of funds, and the maintenance of high standards of credit discipline and fiduciary responsibility. These principles have served IDA and its borrowers well, and remain highly relevant for the future. The challenge is to ensure that IDA's policies -- including those fundamental rules governing access to its resources which are discussed in this paper -- apply these principles creatively to deal with the growing diversity of IDA borrowers' circumstances, and with the increasing volatility of the global environment in which all countries must now operate. 1. IDA Eligibility 4. The fundamental tests of a country's eligibility for access to IDA resources are its poverty and its lack of creditworthiness for market-based borrowing, from both commercial sources and from IBRD1. For the majority of IDA countries, these two 1See para. 41 of Additions to IDA Resources: Twelfth Replenishment - A Partnership for Poverty Reduction, IDA/R98-195, December 23, 1998. - 2 - criteria coincide. In recent years, moreover, economic and political crises have led to income declines and deteriorating creditworthiness in a number of blend countries, and even to renewed access to IDA, sometimes on an exceptional basis, by countries which had previously graduated. Eligibility has become a more hazardous and complicated two-way street, the net result of which is an increase in the number of countries needing access to IDA, and within that an increase in the number of effectively "IDA-only" borrowers. It should, also, be noted that IDA eligibility and the "IDA-only" category have implications beyond IDA itself, as they are the starting points for consideration for concessional funds and debt restructuring terms by other multilateral and bilateral institutions, as well as HIPC eligibility. IDA eligibility criteria need to be reviewed against this background. Per Capita Income 5. The ceiling for IDA eligibility (currently called the historical cutoff), initially set at $250 per capita in 1964, has been revised to account for inflation, reaching $1,445 in 2000. In the early eighties, the availability of IDA resources was not adequate to fund programs for all countries below this eligibility ceiling. As a result, IDA ceased to lend to countries at the upper end of the notionally eligible per capita income scale. This created a second and lower "operational cutoff" which was formally recognized by IDA donors in IDA8. The operational per capita income cutoff has been reaffirmed by the donors in each subsequent replenishment, and now stands at $8852. The historical cutoff is used only for determining preference in civil works procurement. 6. The IDA universe has been limited to low-income countries (see Annex II, Annex Table 1). There are currently 78 IDA-eligible countries, of which 10 are small island states, which are in a somewhat special category. Of the remaining 68 IDA-eligible countries, 38 have per capita incomes of $400 or less, i.e., less than half the operational cutoff. Today, the IDA recipients with the highest per capita national income -- other than the small islands -- are Bolivia and FYR Macedonia (due soon to graduate from IDA) both of which are a little above the operational ceiling, and Albania and Sri Lanka, with per capita incomes of $870 and $820, respectively, a little below it. 7. Per capita income by the World Bank Atlas methodology is IDA's basic indicator or proxy for poverty. While purchasing power parity (PPP) income estimates are, in theory, better suited for cross-country comparisons, the lack of timely and reliable price data reduces considerably the usefulness of PPP income and poverty estimates for operational purposes. 2 The IDA per capita income operational cutoff is updated annually with the international inflation (SDR deflator). Usually, the SDR deflator in US dollars has increased, and the income thresholds have moved up. But in 1998 and 1999, the thresholds moved down because the US dollar appreciated against the SDR. Consequently, the IDA income cutoff has evolved as follows during the last five years: Year GNP Measured 1995 1996 1997 1998 1999 Fiscal Year FY97 FY98 FY99 FY00 FY01 GNI (GNP) per Capita ($) 905 925 925 895 885 - 3 - 8. Adequacy of Per Capita Income as a Poverty Indicator. Per capita income is used as a proxy for individuals' potential access to income to enhance their well being, but it does not indicate how equitably income is shared within the country, the extent of existing poverty, and where the country stands with respect to international development goals. This raises an important issue: is the use of per capita income appropriate, as the sole indicator of relative poverty? One possibility to address this concern would be to design an appropriate and comparable social index to be used in conjunction with the per capita income criterion in order to identify the most vulnerable countries socially, and adapt IDA's response accordingly. The development of a social index to guide IDA's policy decisions would be consistent with IDA's efforts to better impact and monitor social indicators through Poverty Reduction Strategy Papers (PRSP), and the Bank's overall mission of poverty alleviation and helping countries reach international development goals. Such an index could be developed jointly with other agencies involved in this area. In this paper, a purely illustrative social index has been constructed to explore the relative mapping of client countries with respect to their per capita incomes and social indicators (for further details and issues in constructing such an index, see the Note on a Possible Composite Social Index in Annex I). 9. To the extent that a suitable composite social index can be developed, it could, in principle, be a potentially useful sensitivity measure to complement per capita income estimates, and possibly be an additional criterion of IDA eligibility.3 A social index could help explore the possibility of introducing differential pricing (including grants) of IDA funds. For example, the composite social index may provide the mapping needed to identify countries for which limited use of IDA grants may be most appropriate. The social index could contribute to more informed decisions on whether or not IDA eligibility for sectors is warranted, and under which conditions. (These issues are discussed in greater detail in paras. 22 and 23). Such an index could also help monitor IDA's effectiveness in helping reach international development goals. Creditworthiness 10. In general, creditworthiness has been defined as "the ability to service new external debt at market interest rates over the long term". Creditworthiness considerations have always guided IDA lending policies, since the Articles of Agreement limit IDA from providing assistance if financing is "available from private sources on terms which are reasonable for the recipient or could be provided by a loan of the type made by Bank". As a result, some countries with per capita incomes below IDA operational cutoff do not receive IDA credits if they have been able to obtain substantial loans on conventional terms, including from IBRD. Conversely, a lack of creditworthiness implies a need for concessional resources as part of a sustainable long-term financing package for a country's development program. 3 It should be noted, however, that no country should be denied IDA eligibility for having social indicators that are relatively high for their income level, for both incentive and efficiency reasons. For example, many IDA countries in Former Soviet Union (FSU) have social indicators well above the norm for their income levels, as a result of past heavy investments in social sectors. However, those social programs have proved to be financially not sustainable in a market economy, and, in fact, these countries' social indicators have started to deteriorate. It is more efficient to prevent social conditions in these countries from worsening further, with IDA assistance, rather than trying to improve them afterward. - 4 - 11. Within the group of countries with per capita incomes below/above the operational cutoff, the criterion of creditworthiness, i.e., access to IBRD funds and international financial markets, is used to determine which countries are eligible for IDA. Some countries, with incomes below the operational income cutoff but with access to commercial credits and IBRD funds, do not receive IDA funds. Conversely, there are some countries -- the so called "gap countries" -- with incomes above the operational cutoff, but which are not creditworthy for IBRD lending, and might therefore find themselves without access either to IBRD or IDA resources for their development. Also eligible, as an exception to the per capita income operational cutoff, are a number of small island countries which face special size-related problems and which would otherwise have little or no access to Bank Group assistance because of their limited creditworthiness for IBRD lending4. 12. Blend countries constitute a special category of countries with access to IDA, but also able to borrow from IBRD. These are countries that are IDA-eligible on the income criterion, but which also have been creditworthy for limited IBRD lending. However, the number of blend countries deemed creditworthy for IBRD lending has been shrinking, with some major countries including Pakistan and Nigeria having practically no capacity to take on further IBRD borrowing. 13. Moreover, the "blend" countries of the former Soviet Union (FSU) are not likely to gain creditworthiness, any time soon, as it becomes clear that the shocks of the transition were underestimated. Initially, it was thought that these countries would be IDA recipients only for a limited period of time. However, a reevaluation of the economic and financial prospects of some FSU countries, i.e., Armenia, Georgia, Kyrgyz Republic and Moldova has recently resulted in a change in their status from "blend" to "IDA-only". Moreover, some relatively rich but poorly managed countries of the FSU -- Ukraine, Uzbekistan and Turkmenistan -- currently have per capita incomes hovering around the operational cutoff and marginal creditworthiness, posing special challenges for IDA eligibility. Performance 14. Until IDA11, IDA eligibility criteria explicitly included a performance criterion: "the capacity to use IDA resources effectively". The IDA12 Agreement emphasized this criterion within the context of a much strengthened performance framework that governs IDA's allocations to countries. IDA's management has consistently taken into account the performance of a country when deciding on its IDA eligibility, in order to avoid an outcome in which concessional IDA funds seemingly become a "reward" for economic mismanagement. As more and more countries become eligible for IDA (or lose blend creditworthiness for IBRD) because of pronounced poor governance, the performance criterion to determine IDA eligibility at the outset, becomes particularly important. 4 In the past, small island countries were typically considered as "gap" countries. - 5 - 15. Thus, if IDA resources were to be made available to a previously IBRD-only country, and/or a blend country is to be reverted to IDA-only status on the ground that its inadequate economic management and poor governance had resulted in a decline in its creditworthiness, IDA eligibility will continue to be complemented by a performance criterion requiring that appropriate standards of performance be restored for some period before the IDA eligibility of such countries can be activated. These "appropriate standards of performance" -- of which macroeconomic stability is an important one -- would need to be decided according to individual country circumstances. However, there may be country cases where some flexibility would be needed: in some poor performing countries, IDA resources may be needed for formulation and implementation of reforms. Then, a case for IDA eligibility could be made based on government commitment and prior actions. In others, it may be possible to design a small program capable of achieving its development objectives with community participation, intensive supervision, etc. In such situations, an IDA-eligibility case could be made explaining the safeguards for the efficient use of IDA funds, and how the program objectives could be met despite a poor policy environment. 2. IDA Terms 16. Historically, IDA has been a single-price entity, though with minor modifications of maturities for blend countries5. Yet, IDA countries do not constitute a homogeneous group. There are fundamental differences between Azerbaijan and Tajikistan, between Nigeria and Burkina Faso, between India and Nepal, and between Indonesia and Laos, in terms of resource endowment, growth potential, and debt service capacity in the medium to longer term. Given the increasing diversity among IDA countries, it is timely to consider the introduction of greater differentiation in the pricing of IDA funds (including less concessional terms and grants), in order to have the flexibility to adequately meet the needs of individual countries, taking into account their social conditions and economic prospects. The section below presents possible cases for differentiated pricing of IDA funds. Less Concessional IDA Terms 17. Blend Countries. During the IDA11 period, IDA lending to blend countries6 accounted for 36 percent of total IDA lending. After China and Egypt stopped borrowing from IDA in FY00, IDA lending to blend countries declined to 27 percent of total IDA lending commitments in FY00. 5 Until 1987, IDA lending terms were the same for all borrowers: credits had 50 years maturity and 10 years grace. Since then, IDA credits to blend countries have had 35 years maturity and 10 years grace, while those of IDA-only countries have had 40 years maturity and 10 years grace. These rates have created a marginal difference in concessionality (3 percent) between IDA-only and blend countries. 6 Currently, the blend countries are: Azerbaijan, Bosnia and Herzegovina, fYR Macedonia, Indonesia, India, Nigeria, Pakistan, Zimbabwe, and four small islands (Dominica, Grenada, St. Lucia and St. Vincent). - 6 - 18. IDA's share is relatively high in total lending to blend countries. As Table 1 below shows, IDA commitments accounted for over 40 percent of total World Bank commitments to blend countries over the FY98-00 period. If Indonesia -- which was an IBRD-only country until FY99 -- is excluded, this ratio rises to 57 percent. If India as well as Indonesia are taken from the sample, IDA's share in total commitments to blend countries is even higher at almost 70 percent. Table 1: IBRD and IDA Commitments to Blend Countries FY98-00 (US$ million) IBRD (Actual) IDA (Actual) Country FY98-00 FY98-00 IDA/Total (%) Azerbaijan 0 212 100 Bosnia-Herzegovina 0 378 100 Dominica 3 0 - Grenada 1 0 - India 2,402 2,595 51 Indonesia 3,321 256 7 Macedonia, FYR 100 100 50 Nigeria 0 80 100 Pakistan 600 648 52 St. Lucia 5 2 28 St. Vincent 0 0 - Zimbabwe 1 136 100 OECS countries 0 8 100 Total 6,433 4,415 41 Memorandum item: Total without Indonesia 3,112 4,159 57 Total without Indonesia and India 710 1,564 69 19. During IDA12 negotiations, the feasibility and desirability of hardening the terms of IDA lending for blend countries was discussed. It was noted that "an equivalent financial effect for the IDA recipients could be achieved by hardening the blend -- that is, increasing the share of IBRD lending in total IDA and IBRD lending to a country as its economic position strengthens". However, in the short to medium term, there are very few countries whose economic position is likely to strengthen considerably. There is, thus, very little scope for hardening the terms of loans offered to "blend" countries by changing the ratios of IDA and IBRD funds, with the exception of India (see para. 47). In fact, if anything, IDA is likely to become more prominent in these countries, because of creditworthiness issues (Pakistan, Nigeria, Azerbaijan) and IBRD exposure (Indonesia, Pakistan). According to preliminary projections for FY01-03, IBRD commitments in these countries would be about half of their level in FY98-FY00, which, in turn, would imply that IDA may need to fill the gap as long as the countries' per capita incomes remain below the operational cutoff. Also, in the future, some other countries, traditionally in the IBRD category could qualify for IDA eligibility. 20. Over the last few years, the marginal or deteriorating creditworthiness of most of the blend countries has resulted in their being in practice only "notional" blends, with no actual access to IBRD resources, a fact that underscores the need to introduce greater - 7 - differentiation in IDA terms: with intermediate lending terms between those of the IBRD and IDA, the same level of concessionality as blending can be achieved, consistent with the economic prospects and debt servicing capacity of the countries concerned. 21. "Gap" Countries. There are also other cases where less concessional IDA terms could be justified. These include countries whose graduation from IDA is imminent because their per capita income has consistently exceeded the IDA operational per capita income cutoff, but have very limited creditworthiness (the case of "gap" countries). As a result: (a) IBRD cannot lend; or (b) IBRD lending is severely limited; and/or (c) the terms of IBRD loans are not consistent with the current debt servicing capacity of the country. Providing such countries with hardened IDA terms would facilitate a more efficient and smoother transition from IDA-only to IBRD-only borrowing. 22. Eligibility by Sector. The issue of IDA eligibility by sector, i.e., lending to social sectors in less developed regions of countries that are not IDA eligible, has been raised during previous IDA replenishment discussions. It had been concluded that broadening IDA's eligibility concept in such a way would not only over-stretch limited IDA resources, but it would also make IDA-eligibility criteria exceedingly subject to judgment and discretion. There are, however, a number of countries not eligible for IDA, with very low overall social index rankings, which might be considered for limited access to IDA. (These countries are situated in the upper left quadrant of chart in Annex I.) While IBRD has the same mission of poverty alleviation as IDA, most countries are unwilling to borrow from IBRD for social sectors, because borrowing at non-concessional terms for "soft" sectors is seen as potentially leading to a fiscal problem: the debt needs to be repaid in commercial terms but the benefits (although very high from a national social perspective) occur only over time (in the form of higher fiscal revenues). As a result, some critical activities toward reaching international development goals may not be undertaken at all. Also, the Bank may lose its ability to engage such countries in a dialogue on important poverty policy issues. 23. The use of IDA funds with harder terms for social sectors in countries that are not IDA eligible, could, thus, be considered. IDA lending on harder terms to social sectors in "high-income and low-social index" countries would be consistent with IDA's overall objective of poverty alleviation. This would, however, inevitably, spread IDA's resources to slightly less poor countries. The issue, in practice, may involve a trade-off between poor countries and poor people. In any case, this type of IDA eligibility would need to be clearly defined and strictly limited. It could, for example, be restricted to countries with incomes below the historical income cutoff and with a low overall social index ranking. Moreover, access to IDA by such countries would need to be time bound (say 3-6 years) and tied to specific monitorable programs to improve their worst social indicators. 24. In short, if IDA were to consider moving to a more differentiated price structure, harder IDA terms could be considered for: (i) "blend" countries, as a way of hardening the mix when sufficient capacity for IBRD lending is not available (this would not apply to India where hardening lending terms is possible through IDA/IBRD blending); (ii) countries with incomes above the operational cutoff but have not yet established their creditworthiness for other sources of fund; and (iii) for specific social programs, in - 8 - countries not eligible for IDA (because of relatively high per capita incomes), but with low ranking on social indicators and with per capita incomes below the historical cutoff. IDA Grants 25. Starting with IDA11, IDA was authorized to provide direct grant funding in exceptional cases, as approved by its Executive Directors. Currently, IDA can provide grants: (a) in the context of joint Bank/Fund HIPC debt reduction initiative; and (b) to post-conflict countries prior to arrears clearance, as a last resort. There have not, as yet, been any post-conflict grants, because so far possible candidates for such grants have either remained in conflict, or had sufficient funds available to finance necessary early reconstruction activities. During IDA12, IDA grant funding for HIPC operations is expected to be used primarily for interim relief for Honduras, Cameroon and Cote d'Ivoire to reimburse a portion (i.e., 50 percent) of their IBRD debt service payments during that period. Total HIPC grant funding during IDA12 is likely to amount to about $165 million. 26. Long-Term Debt Sustainability. The issue of greater use of IDA grants is now also being raised in the context of debt sustainability of HIPC countries. It needs to be stressed, however, that it is quite implausible to assert that additional lending on standard IDA terms -- with over 60 percent grant element -- could of itself cause a recurrence of unsustainable levels of debt. It is also evident that IDA's long grace period and extended maturities mean that debt service on new post-HIPC IDA credits cannot be a problematic component of future debt obligations for a very long time -- in fact not at all for a decade. Rather, debt sustainability is likely to become an issue for different reasons: for example, if the level of old non-cancelled debt remains too high because it is based on overly optimistic projections of economic and export growths; and/or because external financing other than from IDA is not forthcoming in sufficient quantities and on adequately concessional terms. 27. Despite these important caveats, there is a case to be made for some expansion of grants to the countries most vulnerable to renewed debt problems -- countries with low income and social index rankings, dependent on few major export crops, thus highly vulnerable to terms of trade shocks -- especially since IDA is likely to continue as the key element of their external development financing over the long term. When such recourse to IDA grants is considered for a specific country because of the country's inability to sustain debt, the CAS would need to make a strong and clear case on the basis of realistic assessment of the country's growth and balance of payments prospects, an analysis of debt sustainability, and the burden sharing among various multilateral and bilateral aid institutions. Clearly, the criteria for such extensions of grant financing would need to be carefully defined, to guide specific country decisions, without providing a sense of "automatic" entitlement. In that regard, an analysis of the long-term debt sustainability of HIPC countries is underway, and should be available for discussion early in 2001. 28. Poor Performers. It is generally recognized that IDA should not ignore poor performing countries, in particular the poor people living in these countries. In such cases, IDA's main emphasis would be to encourage better government performance, and develop stronger institutions within the country. But this is a long term undertaking. In - 9 - the short term, if the weak state is the main constraint to development, instruments/delivery mechanisms may need to be found to reach poor beneficiaries directly. In such environments -- where IDA lending to governments cannot be effectively used because of very weak public institutions -- IDA grants in support of small-scale community based activities could be more appropriate, based on decentralized community development approaches being pursued in Africa and elsewhere. The implementation of small, grant-funded projects could be contracted out to other international agencies and/or NGOs experienced in community based small programs, as IDA may have little comparative advantage in implementing such projects. Such grants would allow a quicker IDA response, and would help bridge the gap between relief and development in post-conflict countries. 29. Global and Regional Public Goods. Another area where the use of IDA grants could be considered is global public goods and regional initiatives. A recent paper7 defines public goods, and provides an overview of main areas for IDA/IBRD concentration. Among them, communicable disease control is deemed to be a critical global public good, for which urgent action is required. IDA's role in combating communicable diseases was discussed in IDA12 mid-term review meeting in June, 2000. IDA's strong emphasis on HIV/AIDS and other communicable disease -- with at least $1billion to be allocated over the next three years -- was welcomed. In this area, it was acknowledged that there may be a good case for carefully limited, strategic use of grants for regional and/or global public goods, when other sources of grant funding are not available. 30. The Bank is already closely involved in international efforts to counter communicable diseases through participation in and financial support for key partnership and programs. The Bank has also stepped up its lending for HIV/AIDS. Since 1999, there have been five new projects solely directed to HIV/AIDS eradication for about $506 million. There is a growing sense, however, that these efforts are not nearly enough. The HIV/AIDS epidemic is not only a serious public health concern, but also a major impediment to economic growth and poverty alleviation, particularly in African countries with high prevalence of HIV/AIDS. Therefore, careful consideration needs to be given to the modality/terms with which $1billion would be allocated among programs within countries and regions. Making IDA assistance to HIV/AIDS programs available in the form of grants, would provide a powerful incentive for countries to prepare strong programs in this area, help put AIDS issues prominently on the governments' policy agenda, and eliminate the countries' reluctance to borrow -- however soft IDA's terms may be -- for projects/programs whose benefits would be felt in the long term. 7Poverty Reduction and Global Public Goods - Issues for the World Bank in Supporting Global Collective Action", DC/2000-12, September 6, 2000. - 10 - Financial Impact of Possible Changes in IDA Terms 31. Changing IDA's terms will clearly have ramifications for IDA's finances over the long term, and for lending capacity. IDA lending terms may be hardened by introducing interest rates, shortening the maturity, decreasing the grace period, or through a combination of all these. However, shortening the maturity and the grace periods of IDA credits are not to be recommended: IDA credits with less concessional terms would be used mainly by countries that are experiencing medium term -- hopefully transitional -- balance of payments difficulties, but which have, nonetheless, promising longer term prospects. The intention would be to reduce the debt service obligations of these countries over this transitional period, during which they may be experiencing, in addition to marginal creditworthiness, some liquidity problems. Therefore, hardening IDA terms through decreasing the grace and maturity periods would not to suit the needs of these countries. 32. Tables 2-A and 2-B below present some illustrative information on the impact of a combination of hardened IDA terms plus expanded use of grants8. Table 2-B shows, for example, the medium to long-term impact of having 20 percent of IDA's lending on harder terms, and 20 percent of IDA funds allocated in the form of grants. As can be seen from this table, the gains for IDA's finances resulting from harder -- yet still highly concessional with a 40 percent grant element -- lending terms outweigh the losses incurred by having a modest share of IDA allocated as grants, both in the short and long term. 33. While the proportions used in Table 2-B are presented for illustrative purposes, they indicate that it is possible to structure IDA's lending terms -- to ensure a more equitable treatment across countries in the pursuit of poverty alleviation objective, and to be more responsive to clients' needs -- in such a way that changes in terms would have only a small impact on IDA's finances overall. Clearly decisions would need to be made on the level of harder IDA terms, and the amount of resources that would be provided in that manner, as well as through grants. This would involve striking an appropriate balance between operational objectives, country needs, and financial realities. 8For comparison, the lending terms of other multilateral financial institutions are presented in Annex II, Annex Table 3. - 11 - Table 2-A: Current and Illustrative IDA Pricing and Lending Terms Grant Element a/ Interest Rate Service Charge Maturity Grace Period Current and Illustrative Lending Terms (%) (%) (%) (Years) (Years) IDA Only Terms 64 0 0.75 40 10 Blend Terms 62 0 0.75 35 10 Grants 100 0 0.5 b/ N/A N/A Hardened Terms: 10/40 Amortization with a 3% Interest 40 3 0 40 10 Table 2-B: Financial Impact of Illustrative Changes in IDA's Lending Portfolios c/ Percentage of Each Instrument in IDA's Scenarios of Future Aggregate Lending Portfolios Short-Medium Term Financial Impact Medium-Long Term Financial Impact Lending Portfolios Current Terms Hardened Years: 1-10 Total Gains (Losses) Years: 11-20 Total Gains (Losses) (IDA Only & Blend) Grants Terms ($billion) ($billion) 1. 100% Current Terms 100% 0 0 2. 20% Grants 80% 20% (0.4) (4.3) 80% Current Terms 3. 20% Grants 20% Hardened Terms 60% 20% 20% 0.6 1.2 60% Current Terms Notes: a/ Commercial Interest Reference Rate of 6.3% is used as the discount rate for grant element calculations. b/ Service charge for grants will be 0.5% of the total commitment amount and will be charged for six years. c/ This simulation analysis on financial impact is done under the following assumptions used in the current base-case cashflow scenario: A) lending remains stable at IDA12 framework level ($SDR15.2b) in real terms (assuming a 2.3% inflation rate); B) donor contribution remains stable at IDA12 level (SDR8.6b) in nominal terms; C) 5% investment income from liquid assets; D) 5% non-accrual; E) administrative expenses remain stable in real terms; F) the share of programmatic lending will increase from 25% for IDA12 to 35% for IDA14; G) future IBRD net income transfer $300m a year; H) HIPC debt relief costs are fully reimbursed by donors; etc. - 12 - 3. Graduation from IDA Graduation Policy and Its Implementation 34. The process of graduation from IDA is normally triggered when a country exceeds the operational per capita income guideline. Because income levels fluctuate from year to year, countries whose per capita income has risen above the threshold by a marginal amount, would normally begin a graduation process that, in most cases, lasts for several years. This graduated approach avoids situations in which income fluctuations could allow only intermittent access to IDA funding, which would complicate IDA allocations as well as rational planning by the country. It also recognizes that the national income data are imperfect, and that a sudden termination of IDA would not be desirable from a developmental point of view. 35. Some countries, however, have graduated from IDA on an accelerated basis. This may occur where improved information becomes available showing that a country's income is substantially higher than previously estimated. In such cases, where the new income data indicates that country income is substantially above the IDA operational cut- off, the country may be graduated at the end of the fiscal year. A country where this has recently occurred is Egypt, which was graduated from IDA at the end of FY99. In other cases, a substantial amount of petroleum or other natural resources may be developed which greatly increases a country's per capita income, especially if the country is small in population. In such cases, the country would also be graduated on an accelerated basis. Equatorial Guinea that graduated during FY99 is a case in point. 36. Graduation can also occur when a country achieves creditworthiness for adequate amounts of IBRD and other commercial sources of funds, even though its per capita income remains below the operational cutoff. Such countries would typically already be blend countries, with access to both IDA and IBRD lending. Countries move to blend status with per capita incomes below the operational cutoff, when they have improved creditworthiness. In such cases, IDA lending to the country is curtailed and replaced to some extent with IBRD lending. With an even greater improvement in creditworthiness, such blend countries can graduate from IDA, even though their per capita incomes remain below the operational cutoff. These countries would normally have strong export earnings and large reserves, and a demonstrated track record of successfully borrowing on the international commercial markets to meet much of their capital needs. China is the most recent example of a country which ceased to borrow from IDA at the end of FY99 with a per capita income of $860, less than the operational cutoff of $925 at the time, but with substantial access to capital on commercial terms. 37. When IDA determines that a country should graduate, a graduation program is formulated usually in the CAS, comprising: · A phase-out of IDA lending; · A phase-in of IBRD lending; · Special ESW and technical assistance to help address transition issues, such as improving access to commercial sources of lending; - 13 - · Increased role of MIGA, IFC to improve private sector capital inflows; and · Application of triggers to accelerate the repayment of IDA credits9. 38. Each of these measures is tailored to the particular circumstances of the country, except for the acceleration of repayments of IDA credits, which is governed by a standard formulation. Accelerated repayments are triggered when a country is creditworthy for IBRD borrowing and its per capita GNP reaches or exceeds the operational cut-off for IDA eligibility for three consecutive years10. 39. It is clear from the interaction of income and creditworthiness considerations, and also from the volatility and fragility of economic and political progress in many IDA countries, that graduation cannot be solely driven by a mechanistic formula. A total of 31 countries have graduated from IDA during FY61-99 (Annex II, Annex Table 2 provides a list of countries that graduated since 1961). Because of adverse developments subsequent to their graduation, 8 of these graduates have been granted renewed access to IDA, leaving the total number of net graduates during these years at 23. This experience illustrates the importance of a graduation policy which is unwavering in its basic intent, but applied with flexibility and due regard to the diversity of country circumstances, as is clear from the discussion below of country cases and approaches. 40. Two countries that are, now, on the path to graduation are Bolivia and fYR Macedonia. FYR Macedonia's CAS Progress Report of May 2000 confirms that the country will graduate from IDA at the end of FY01. As for Bolivia, it has exceeded the operational per capita cutoff for a number of years. The 1998 CAS11 recommended that Bolivia cease to be IDA eligible after FY02. However, the CAS also stated that many factors are likely to play a role in Bolivia's creditworthiness, including "smooth democratic transition without significant social unrest". The recent social crisis which paralyzed the country and resulted in significant economic losses has demonstrated how fragile and vulnerable Bolivia's achievements are. A creditworthiness study for IBRD lending is underway. Even if Bolivia is found creditworthy for IBRD lending, under current circumstances a more gradual graduation path may be warranted; alternatively, Bolivia could be a candidate for harder IDA lending terms in its transition period, if such a policy change to IDA lending were to be adopted. The Small Island Economies 41. The small island economy exception permits provision of IDA resources to small island economies, with per capita income above the operational cutoff for IDA eligibility, 9 See Modification to Accelerated Repayment Terms of IDA Credits, IDA/R96-145, 1996. 10 IDA borrowers most likely to be affected by accelerated repayment provisions are those having the highest per capita incomes, strong growth prospects, and full creditworthiness. However, recently, not many IDA countries have had these characteristics, limiting the impact of these provisions. 11 Country Assistance Strategy of the World Bank Group for the Republic of Bolivia, May 21,1998, Report No.17890-BO. - 14 - but with marginal creditworthiness12. This exception reflects a recognition that these economies typically have high transportation costs, fewer opportunities to pursue economies of scale and severe human capital constraints because of their small populations. Total lending to the small island countries during FY98-00 was 0.6 percent of total IDA lending. Given their small claim on IDA resources and limited creditworthiness, it seems prudent to continue the small island exception. The Bank is currently implementing a program to better assist these countries and address their needs. 42. There are two large IDA countries, India and Indonesia, for which the issue of graduation and IDA eligibility has arisen on several occasions. The section below provides an overview of Indonesia's and India's circumstances which would warrant their continued access to IDA resources. India 43. India is home to 300 million poor, more than in all of sub-Saharan Africa. India has 36 percent of the world's poor, and 32 percent of world's children not enrolled in primary education. It accounts for: (i) 36 percent of the world's gender gap in primary and secondary education; (ii) 23 percent of all under-5 child deaths; (iii) 20 percent of all maternal deaths; (iv) 29 percent of the world's unmet need for reproductive health services; and (v) 30 percent of world's death from poor access to water and sanitation. 44. The great bulk of IDA support goes for the social sectors and rural poverty reduction projects. IDA funds are used, for example, to support the Bank's largest elementary education and nutrition programs, both targeted at girls. IDA also supports a large number of health programs, in order to combat AIDS, malaria, and leprosy. Without IDA, the Bank would not be able to lend to social sectors, because of the Government's reluctance to borrow non concessional funds for "soft" sectors. Moreover, although at the macro level, IDA funding means little to India, IDA's impact has greatly increased by adopting a state focus. 45. India makes effective use of IDA resources. India consistently figures in the top or upper quintile of Country Performance Ratings which include ratings on Country Policy and Institutional Assessment (CPIA) and ratings on implementation of IDA projects (portfolio performance review). 46. OED is now finalizing its Country Assistance Evaluation (CAE) for Bank (IBRD/IDA) assistance to India over the nineties. The draft CAE concludes that Bank assistance over this period was "moderately successful". It also notes, however, that "the relevance of the assistance strategy, however, has improved substantially over the past two years through a more sharpened focus on poverty reduction, a more selective approach to state assistance and greater attention to governance and institutions. Although it is too early to gauge its efficacy ..., recent initiatives hold strong promise of substantial improvement in performance". To reduce IDA assistance at this particular 12 There are a number of such economies that have received IDA funds: Cape Verde, Dominica, Grenada, Kiribati, Maldives, Samoa, St. Lucia, St. Vincent and the Grenadines, Tonga, Vanuatu. - 15 - juncture would marginalize the Bank and threaten the progress made in the last few years in promoting reforms. 47. India's access to international capital markets is far below its potential. India's foreign direct investment (FDI), for example, is at less than $4 billion, a fraction of what China has achieved at almost $40 billion. To be able to attract much more FDI, India needs fundamental reforms in infrastructure, governance, and fiscal areas, especially at the state level. Combined IDA and IBRD assistance is critical in that regard. Given the fact that India has the largest concentration of the poor in the world, a reduction in IDA resources would not be desirable. The country's increasing creditworthiness could be reflected in a hardening of the blend through increased IBRD lending relative to IDA, rather than a decline in its IDA allocation. Indonesia 48. Following the Asian crisis and the sharp drop in its per capita income, Indonesia became once more IDA eligible as a blend country during FY99, along with an understanding that only small volumes of IDA would be made available for a limited time, while the country's economy recovered. Despite many positive developments, including signs that recovery is taking hold, political stability has been fragile, financial markets have remained skeptical, and severe government indebtedness has limited public spending critical for Indonesia's poor. 49. In June 2000, IDA Deputies agreed with the Bank staff assessment that the process of reform and improving governance in Indonesia should be strongly supported, including through enhanced IDA allocations. The CAS covering FY01-03, includes an increased IDA allocation in a high case scenario13. Access to higher IDA allocations would depend on sustained performance in structural and governance reforms and development and implementation of a poverty strategy: if these benchmarks can be met, IDA resources will clearly play a critical role in sustained recovery. In this connection, it should be noted that even in a favorable scenario -- in which growth would climb to six percent in 2003 -- Indonesia's per capita income would stay below the IDA threshold through 2004, and possibly beyond, while its creditworthiness is likely to remain highly constrained throughout this period. Lowering the Per Capita Income Threshold for Graduation 50. Aside from the above specific country considerations, the most obvious potential change in IDA graduation policy which could affect a wider category of countries would be a reduction in the per capita income cutoff. A reduction, for example, of about 20 percent in graduation income threshold to $700 would affect a diverse group of IDA-eligible countries, the majority of which are good performers14: Albania, Bolivia, Bosnia and Herzegovina15, Cote d'Ivoire, Djibouti, Guyana, Honduras, Macedonia FYR, 13The FY01-03 Country Assistance Strategy for Indonesia is due for Board discussion in early 2001. 14Within this group, 99 percent of IDA resources are allocated to 9 countries in the top quintile, 4 countries in the upper quintile, and 2 countries in the middle quintile of the 2000 IDA Country Performance Ratings (combined CPIA and portfolio ratings). - 16 - Sri Lanka, as well as 10 small island countries (Cape Verde, Dominica, Grenada, Kiribati, Maldives, Samoa, Solomon Islands, St. Lucia, St. Vincent and Vanuatu).16 51. These relatively higher income IDA (HII) countries, however, accounted for only 12.8 percent of total IDA lending during FY98-00. Also, there are important differences between the HII countries and the IBRD countries, in terms of development indicators as well as access to foreign capital. Social indicators in the HII countries are below those in the middle-income IBRD countries (except in the case of small islands and former socialist countries). For example, adult illiteracy is about twice as high in the HII countries as in the IBRD countries. Female literacy and female primary school enrollment are also significantly lower in the HII countries. Moreover, most of the HII countries would not be eligible for IBRD lending for quite some time, because of creditworthiness issues. Thus, graduating them prematurely would increase the number of "gap" countries, and leave them with costlier resources to satisfy their development needs. 52. Therefore, there does not seem to be any compelling reason to change IDA's current graduation policy. Individual countries should continue to be monitored and evaluated carefully as they begin to approach graduation guidelines. IDA should continue its flexible approach to graduation to ensure that countries are not prematurely pushed to borrow on IBRD and on commercial terms. For that purpose, IDA resources with hardened, yet still highly concessional terms could be provided: (a) to countries -- that have met graduation's income and creditworthiness criteria -- for a limited period of time (say 3-5 years); and (b) to countries -- that have met only the income criterion of graduation -- until they also satisfy the requirements of full creditworthiness (as discussed in para. 21). 4. Issues for Discussion 53. The increasing diversity of country conditions and global issues have called for a re-evaluation of IDA's eligibility criteria, terms and graduation policies in order to best fulfill IDA's objective of poverty alleviation and adequately address the needs of individual countries and emerging priorities, without compromising the close link between performance, IDA eligibility and allocation. To that end, IDA could consider the following in its operational policies. 15Bosnia and Herzegovina's per capita income is controversial because of unreliable population statistics. 16 All these small island countries, except Solomon Islands, are now IDA eligible under the small island exception. - 17 - Adapting IDA's Instruments to the Diversity of its Clients · IDA could introduce differential pricing for its clients: (a) regular "IDA-only" terms (64 percent grant element); (b) hardened IDA terms (about 40 percent grant element); and (c) grants. · This differentiated pricing could be based, in part, on the existing categories of IDA countries: (a) regular IDA terms for IDA-only countries. (b) hardened terms for: (i) "notional blends" where there is almost no scope for hardening the lending terms through appropriate IBRD/IDA blending; (ii) "gap countries" with per capita incomes above the operational cutoff but with marginal or no creditworthiness in the medium term; and (iii) graduating countries for a limited period of time (say 3-5 years). (c) part of the regular IDA allocation could be provided as grants to countries most susceptible to renewed debt problems: low-income countries with poor social indicators, dependent on few major export crops, thus highly vulnerable to terms of trade shocks. Increasing IDA's Effectiveness and Outreach · IDA grants could be provided: (a) for global and regional public goods such as eradication of communicable diseases, in particular HIV/AIDS which is not only a serious public health problem, but also a major impediment to growth and poverty alleviation in Africa; and (b) to international agencies and/or NGOs experienced in community-based programs to reach the poor directly in countries where the public institutions are extremely weak. · IDA could explore the possibility of developing a suitable and comparable composite social index incorporating international development goals and other relevant poverty indicators, as a joint effort with other agencies involved in this area. If such an index could be successfully developed, it could complement IDA's per capita income indicator, and help monitor IDA's effectiveness in poverty alleviation and in reaching international development goals. It could also help identify: (a) socially most vulnerable IDA countries for which grants may be needed; and - 18 - (b) non-IDA eligible countries for which IDA eligibility by sector may be warranted (see below). · IDA could lend (with hardened terms) for social programs in countries with per capita incomes above the operational cutoff, but with low rankings on social indicators, subject to overall good performance. Lending can be time-bound and tied to specific programs aimed at improving the countries' worst social indicators. Keeping Current Graduation Policies · Current per capita income graduation threshold need not be lowered, as this would have only a marginal impact on the demand for IDA funds, would penalize mostly good performers, and increase the number of "gap" countries, leaving them with higher cost funds to meet their development needs. · Small island exception should be maintained because their economies continue to face serious development challenges. · It is not advisable to reduce IDA resources to India in order to help eradicate global poverty, and not to jeopardize the progress the Bank has made in recent years in improving its impact and relevance in this country. India's increasing creditworthiness could be reflected in a hardening of the blend through increased IBRD lending relative to IDA, rather than a decline in its IDA allocation. · Indonesia may have access to increased IDA resources, based on annual performance criteria/triggers to be set once the country meets the requirements of the high case scenario of the CAS. Graduation would happen as economic growth gains momentum, investors' confidence is restored, and the per capita income threshold is surpassed. - 19 - ANNEX I Page 1 of 3 NOTE ON A POSSIBLE COMPOSITE SOCIAL INDEX 1. In this paper, a composite social index has been developed for illustrative purposes only to complement per capita income estimates. The indicators included in the index are all taken from the set of 22 indicators identified as part of "the international development goals", shown below. The International Development Goals · Reduce the proportion of people living in extreme poverty by half between 1990 and 2015; · Enroll all children in primary school by 2015; · Make progress toward gender equality and the empowerment of women by eliminating gender disparities in prima in primary and secondary education by 2005; · Reduce infant and child mortality rates by two thirds between 1990 and 2015; · Reduce maternal mortality ratios by three quarters between 1990 and 2015 · By 2015, provide access to reproductive health services to all who need them; and · Implement national strategies for sustainable development by 2005 so as to reverse the loss of environmental resources by 2015. 2. The indicators chosen to develop the social index used in this paper, are: primary school net enrollment; ratio of girls to boys in primary school gross enrollments; infant mortality rate; under-age-5 mortality rate; illiteracy rate of adults; contraceptive prevalence rate; births attended by health personnel; and under-age-5 malnutrition17. It should be noted, however, that this set of indicators is far from being comprehensive. Some of the indicators necessary to measure progress toward the goals are not available: for example, there is considerable debate over how to measure maternal mortality rates, and there is no direct information on access to reproductive health services. 17 Each indicator was standardized by taking the natural logarithm, subtracting its mean, and dividing by its standard error. The overall social index is the sum of the 8 standardized indicators, re-scaled to a range of 0 to 100. - 20 - ANNEX I Page 2 of 3 3. If the use of a composite social index were to be adopted as playing a complementary role in IDA eligibility, practical problems in constructing such an index would need to be addressed. There should be a comprehensive analysis of how the relevant dimensions of poverty and social indicators would be included, how the statistical and data problems would be addressed, and how aggregation problems (relative weights of various indicators) would be solved. Also, the links between any such social index and the PRSP should be spelled out, as PRSPs should identify specific monitoring indicators, consistent with international development goals, but rooted in country specific situations. 4. The chart below that includes both IDA and IBRD countries, shows the distribution of countries, according to their per capita income and social index rankings. As can be seen from this chart, high per capita income is generally closely linked with a high social index18. Countries that are in the lower-left (low income and low social index) and upper-right (high income and high social index) of the chart are unambiguously below or above the IDA operational cutoff, respectively. The countries that raise particular questions concerning the use and role of IDA resources are located in the upper-left and the lower-right quadrants of the chart. The countries in the lower-right are those that are IDA eligible based on per capita income, but have relatively high social index values. Those countries are: Kyrgyz Republic, Moldova, Azerbaijan, Tajikistan, Vietnam, Mongolia and Nicaragua. These are formerly socialist economies that invested heavily in social sectors. A key challenge for these countries and IDA is to be able to maintain these good social indicators. 5. The countries in the upper-left of the chart (Namibia, Botswana, Swaziland, Gabon, Guatemala, Egypt, Morocco, and Papua New Guinea) are not IDA eligible because their per capita incomes are above the operational cutoff, but they rank, nonetheless, relatively low on social indicators. Two of these countries are from North Africa, four are from Sub-Saharan Africa, of which three have very high HIV infections rates. Five of these countries are mineral rich economies that do not seem to have been very effective in using their public resources to improve their social indicators. This raises the issue of whether IDA's involvement could have made a difference in these countries, given the fact most countries are unwilling to borrow on IBRD terms for social sectors. 18 The mapping of countries is also quite similar, when UNDP's Human Development Index and GNP per capita rankings are plotted. - 21 - ANNEX I Page 3 of 3 Social Index GNPPC Social Index GNPPC Income per capita versus social index ranking Country Ranking Ranking Country Ranking Ranking Argentina 1 2 Solomon Islands 57 62 Chile 2 6 Algeria 58 40 Croatia 3 8 Samoa 59 52 Brazil 4 7 Maldives 60 51 Costa Rica 5 23 Bolivia 61 55 Rank on social index (1994-1998) Bahrain 6 3 Cape Verde 62 49 Colombia 7 26 Honduras 63 64 101 81 61 41 21 1 Russian Federation 8 29 Zimbabwe 64 68 14 10 1 1 Venezuela, RB 9 18 Namibia 65 33 6 Korea, Rep. 10 1 Indonesia 66 67 High income and low High income and 2625 432 Kazakhstan 11 44 Botswana 67 22 High income and low social index social index high social index 23 12 Hungary 12 10 Egypt, Arab Rep. 68 46 70 31 15 Kyrgyz Republic 13 79 Swaziland 69 42 37 22 16 Dominican Republic 14 35 Gabon 70 11 36 9 Poland 15 12 Lesotho 71 70 34 48 30 21 Malaysia 16 16 Sudan 72 96 67 47 5 Jamaica 17 37 Ghana 73 78 19 38 7 Moldova 18 80 Kenya 74 87 27 8 Lithuania 19 25 Guatemala 75 39 45 35 28 Bulgaria 20 48 Congo, Rep. 76 66 65 51 14 Romania 21 43 Morocco 77 47 49 39 17 Mauritius 22 14 India 78 75 75 58 Trinidad and Tobago 23 9 Nigeria 79 94 69 46 41 21 Albania 24 59 Papua New Guinea 80 58 11 68 42 Uruguay 25 4 Comoros 81 85 77 20 Czech Republic 26 5 Cameroon 82 69 62 33 Latvia 27 28 Cambodia 83 98 60595655 Thailand 28 31 Gambia, The 84 89 61 44 32 Azerbaijan 29 73 Haiti 85 76 80 24 South Africa 30 20 Zambia 86 90 41 57 52 61 1998 Mexico 31 13 Togo 87 91 99 63 40 Uzbekistan 32 57 Uganda 88 93 76 6664 capita Jordan 33 50 Mauritania 89 77 82 71 Estonia 34 19 Tanzania 90 103 9896 per 29 Tunisia 35 32 Madagascar 91 99 93 85 78 Lebanon 36 17 Malawi 92 104 89 GNP 73 13 Slovak Republic 37 15 Pakistan 93 74 on 102 53 18 81 Peru 38 27 Congo, Dem. Rep. 94 111 54 43 Iran, Islamic Rep. 39 38 Bangladesh 95 88 81 74 50 Rank China 40 63 Senegal 96 72 95 Sri Lanka 41 60 Central African Republic 97 95 100 878684 Macedonia, FYR 42 45 Guinea 98 71 88 97 79 Tajikistan 43 83 Cote d'Ivoire 99 65 103 72 83 Georgia 44 56 Lao PDR 100 92 111109 110 91 101 Belarus 45 30 Burundi 101 110 Ecuador 46 41 Benin 102 82 9290 108105 104 Low income and Low income and Belize 47 24 Yemen, Rep. 103 97 112 low social index high social index 106 Turkey 48 21 Nepal 104 105 101 107 94 Paraguay 49 36 Eritrea 105 107 Vietnam 50 86 Guinea-Bissau 106 109 El Salvador 51 34 Ethiopia 107 112 IDA Countries Line of Equality Non-IDA Countries Guyana 52 61 Mozambique 108 106 Mongolia 53 81 Burkina Faso 109 101 Nicaragua 54 84 Mali 110 100 Philippines 55 53 Chad 111 102 Syrian Arab Republic 56 54 Niger 112 108 *Numbers in the graph designatethe Social Index Ranking of countries in this sample - 22 - ANNEX II Page 1 of 3 Annex Table 1: IDA Borrowers and Their Per Capita Income, 1999 (US$, Atlas Methodology) 100-400 400-500 500-600 600-700 700 and over 700 Angola Armenia Azerbaijan * Congo, Rep. Albania Bangladesh Haiti Bhutan Georgia Bolivia Benin India * Cameroon Cape Verde @ Burkina Faso Pakistan * Guinea Cote d'Ivoire Burundi Nicaragua Indonesia * Djibouti Cambodia Lesotho Dominica @* Central African Rep. Senegal Grenada @* Chad Zimbabwe * Guyana Comoros Honduras Eritrea Kiribati @ Ethiopia Maldives @ Gambia, The Macedonia, FYR * Ghana Samoa @ Guinea-Bissau Solomon Islands Kenya Sri Lanka Kyrgyz Republic St. Lucia @* Laos St. Vincent & Gre @* Madagascar Tonga @ Malawi Vanuatu @ Mali Mauritania Moldova Mongolia Mozambique Nepal Niger Nigeria * Rwanda Sao Tome & Principe Sierra Leone Sudan Tajikistan Tanzania Togo Uganda Vietnam Yemen, Republic of Zambia @ Small Island Exception. * Blend Countries. NA: Afghanistan, Congo DR, Bosnia and Herzegovina, Liberia, Myanmar, Somalia. - 23 - ANNEX II Page 2 of 3 Annex Table 2: Countries Ceasing to Borrow from IDA Between FY61-99 Country Last IDA Credit Re-access to IDA Botswana FY74 Cameroon FY81 FY94 Chile FY61 China FY99 Colombia FY62 Congo (Rep. of) FY82 FY94 Costa Rica FY62 Cote d'Ivoire FY73 FY92 Dominican Republic FY73 Ecuador FY74 Eq. Guinea FY99 a/ Egypt FY81; FY99 (FY91) ) b/ El Salvador FY77 Honduras FY80 FY91 Indonesia FY80 FY99 Jordan FY78 Korea FY73 Mauritius FY75 Morocco FY75 Nicaragua FY81 FY91 Nigeria FY65 FY89 Papua New Guinea FY83 Paraguay FY77 Philippines FY79; FY93 (FY91) c/ St. Kitts FY94 Swaziland FY75 Syria FY74 Thailand FY79 Tunisia FY77 Turkey FY73 Zimbabwe FY83 FY92 a/ Remained IDA eligible until FY99. b/ Graduated again in FY99. c/ Graduated in FY93. - 24 - ANNEX II Page 3 of 3 Annex Table 3: Lending Terms of IDA and Other Multilateral Development Banks IDA AsDF AfDF IDB-FSO Maturity - years IDA only borrowers 40 Blend borrowers 35 All - project loans 32 All - sector loans 24 All borrowers 50 40 Grace period - years 10 8 10 10 Charges Service charge on DOD balance (%) 0.75 0.75 Interest charge on DOD balance During grace period (%) 1.0 1.0 During amortization period (%) 1.5 2.0 Commitment fee on undisbursed balance (%) 0 - 0.50 0.50 0.50 Front-end fee for supervision and inspection (%) 1.00 Grant element a/ (%) 61-64 41-47 64 47 a/ Assumptions on grant element calculation: 1 Discount rate used is the Commercial Interest Reference Rate of 6.3%. 2 Disbursement profile (% of face value) used: Yr1-9.6; Yr2-15.6; Yr3-18.6; Yr4-15.6; Yr5-14.6; Yr6-12.6; Yr7-7.6; Yr8-5.5.