IDA14 IDA's Performance-Based Allocation System Update on Outstanding Issues International Development Association February 2004 - i - Acronyms and Abbreviations AfDB African Development Bank AsDB Asian Development Bank ARPP Annual Review of Project Performance CAS Country Assistance Strategy CPIA Country Policy and Institutional Assessment GNI Gross National Income IBRD International Bank for Reconstruction and Development ICP IDA Country Performance Rating IDA International Development Association IPRSP Interim Poverty Reduction Strategy Paper LICUS Low-Income Countries Under Stress LSR Lending Strategy Review OED Operations Evaluation Department PBA Performance-Based Allocation System PCPA Per Capita Per Annum PR IDA Country Performance Rating PRSP Poverty Reduction Strategy Paper SDR Special Drawing Rights - ii - Table of Contents I. Introduction..........................................................................................................................1 II. Towards Disclosure of IDA Country Performance Ratings ................................................1 III. Calibrating Governance in the PBA System........................................................................2 IV. Weighting Poverty in the PBA System................................................................................6 V. Low-Income Countries Under Stress: Addressing LICUS Needs Within the PBA Framework...........................................................................................................................8 VI. Summary and the Ongoing Evolution of the PBA System ...............................................13 Table 1: MDB Cluster Weights ......................................................................................................3 Table 2: Calibrating Governance in the PBA System ....................................................................4 Annex I: Summary of the CPIA Criteria ...................................................................................14 Annex II: IDA: Alternative Cluster Weights Criteria ................................................................15 Annex III: Low-Income Countries Under Stress (LICUS) Typologies .......................................16 I. INTRODUCTION 1. At the time of the IDA13 Mid-Term Review IDA Deputies expressed support for IDA's Performance-Based Allocation (PBA) System and the work program set out to refine it.1 This paper examines before the start of the IDA14 deliberations four issues that are part of that work program and on which IDA Deputies requested further analysis or information. The four issues discussed in this paper concern: (i) the next steps towards disclosure of the IDA Country Performance Ratings; (ii) a discussion of calibrating governance in the PBA System; (iii) a more detailed rationale for the weighting of poverty in the PBA System, and (iv) a description of how IDA can address special needs of LICUS countries within the PBA framework. The last three issues are of particular import for countries at the lower end of the performance and income spectra. They all involve a search for a way to address the particular needs of these countries while maintaining the integrity of the PBA System. II. TOWARDS DISCLOSURE OF IDA COUNTRY PERFORMANCE RATINGS 2. Over the past two months Management has consulted with the Board on how to implement the IDA13 undertaking of moving towards disclosure of IDA Country Performance Ratings (ICP). In response to suggestions by several Board members it has assembled a Panel of Experts, including participants from developing and developed countries and drawing on both academics and public officials to review the coverage of the present rating system, methodology, database and cross-country comparability, and to propose options to strengthen the CPIA. The action plan laid out in the revised Board paper2 recommended that the Executive Directors approve by late January 2004, starting with the results of the 2003 CPIA exercise, the public disclosure of each country's ratings for the overall CPIA, its four dimensions, the portfolio (ARPP), the ICP, and the governance score in the form of half-point ranges. 3. During an informal Board discussion of the disclosure action plan in January, it was decided to defer the disclosure in the form of half-point ranges, and that Management would get back to the Board in March with its proposed plan of action, in light of the findings and recommendations of the Panel of Experts. Upon its request, the Terms of Reference of the Panel of Experts have been shared with the Board. The panel meeting has been scheduled for February 17-18 in Washington, D.C. 1 For a summary of the PBA system, see Additions to IDA Resources: Thirteenth Replenishment, Annex 1, September 17, 2002, IDA/SecM2002-0488. For a review of the PBA System's recent evolution and results, see Allocating IDA Funds Based on Performance - Fourth Annual Report on IDA's Country Assessment and Allocation Process, March 26, 2003, IDA/SecM2003-0130. For the PBA paper discussed at the time of the IDA13 Mid-Term Review, see IDA's Performance-Based Allocation System: Current and Emerging Issues, October 27, 2003, IDA/R2003-0203. 2 Disclosing IDA Country Performance Ratings, December 24, 2003, IDA/R2003-0187/2 (previous version, dated October 18, 2003, IDA/R2003-0187). - 2 - III. CALIBRATING GOVERNANCE IN THE PBA SYSTEM 3.1 Governance and Operational Continuity 4. As reported at the time of the IDA13 Mid-Term Review, over the past three years the governance factor has proven to be an effective means to place governance at the center of the allocation of IDA funds and in the country dialogue. At the same time, relatively modest changes in the ratings of governance criteria were found to result in rather severe allocation changes. This raised the question whether the effective governance weight of 67% in the IDA Country Performance Rating, versus 33% for economic management, structural policy, social policy, and portfolio performance combined, might require a recalibration. Accordingly Management has been looking for a calibration of the weight of governance that would strike the right balance between the CPIA clusters, while both keeping governance as a central policy focus and avoiding sharp allocation shifts as a result of modest governance rating changes. 5. While considering the need for a recalibration of governance's role in the PBA System, a number of Deputies have also pointed at the need to make that role more transparent than is currently the case: governance scores are counted twice, first in the CPIA and second through the governance factor which draws on selected CPIA and Annual Review of Project Performance (ARPP) criteria.3 A simplification of the allocation formula could contribute significantly to increased public understanding of the IDA rating system at a time when it is increasingly shared with development partners. 3.2 Calibrating and Simplifying the Role of Governance 6. This section discusses two options of how to calibrate and simplify the role of governance in the PBA System. One option was mentioned at the time of the Mid-Term Review, involving the elimination of the exponent (of 1.5) currently applied to the governance factor (Option I).4 Such linear governance factor would modestly mitigate the current impact of governance, reducing its effective weight in the system from 67% to 60%, and increasing the average per capita per annum allocations to countries in the bottom quintile by about one third from the currently very low level. This approach would not, however, simplify the treatment of governance in the system. 7. An alternative simplified approach has been used by the Asian Development Bank (AsDB) and African Development Bank (AfDB). Instead of applying a governance factor, they give extra weight to the governance cluster in the country policy and institutional assessment framework. In the case of IDA, this approach would involve increasing the weight of the governance cluster within the CPIA. Table 1 compares weights that IDA currently assigns to 3 The governance factor is based on seven criteria including six governance-related CPIA criteria (4 and 16-20, see Annex I), plus the procurement practices criterion included in the ARPP rating (the governance discount was based on the same seven criteria). The factor is calculated by dividing the average rating on these seven criteria by 3.5 (the mid-point of 1-6) and applying an exponent of 1.5 to this ratio: governance factor = (average governance rating / 3.5)1. .5 4 See Table 2. - 3 - each of the policy and institutional assessment clusters as well as to the country's portfolio rating with those applied by the Asian and African Development Banks. Since in IDA's current approach six out of twenty equally weighted criteria (i.e. 30%) in the CPIA concern governance, and the CPIA in turn is weighted 80% (vs 20% for the ARPP rating), the governance weight is 24% as shown in the table. However, because IDA applies a governance factor, the overall effect is to give governance an effective weight of 67%.5 This compares to 30% in the AsDB, and 21% in the AfDB.6 Table 1: MDB Cluster Weights IDA AsDB AfDB Macro / Structural Criteria 36% 25% 28% Social Criteria 20% 30% 21% Total Non-Governance Criteria 56% 55% 49% Governance Criteria 24% a/ 30% 21% ARPP (Portfolio Rating) 20% 15% 30% Total 100% 100% 100% a/ Including the Governance Factor, the effective weight is 67% 8. In Option II the governance factor is eliminated and instead each of the five governance criteria is weighed three times as much as each of the 15 other CPIA criteria (10% versus 3.3%). This would imply giving equal weight to the governance cluster (five criteria weighting 10% each) as to the other three clusters combined (macro, structural, and social: fifteen criteria weighting 3.33% each). Fifty percent weight to governance within the CPIA would result in a 40% weight in the overall IDA rating (given the CPIA's 80% weight vs the ARPP's 20% weight). This would be a lower figure than under the current approach (67%), but still a significant weight, well above those of the AsDB (30%) and AfDB (21%). Table 2 presents the two alternative options, together with the current approach and the scenario in which no extra emphasis is given to governance. 5 The effective weight is based on: (i) a calculation of the impact of a one point change in the rating of a governance criterion on the IDA country rating as compared to the impact of such rating change of a non- governance criterion; and (ii) a calculation of governance's effective weight taking into account the number of governance criteria in the CPIA and ARPP. 6 Within the AfDB's CPIA the governance cluster is assigned a 30% weight, but since that CPIA is weighted 70% vs 30% for the portfolio rating, the AfDB's effective governance weight is 21%. - 4 - Table 2: Calibrating Governance in the PBA System Current Option II. Approach: Option I. High Weight Governance Governance for CPIA criteria Factor Factor Governance in weighted Expon. Formula Linear Formula CPIA equally Governance Factor? Yes (power 1.5) Yes (power 1.0) No No Governance Criteria 7 7 5 5 CPIA #16-20 CPIA #16-20 CPIA #16-20 CPIA # 16-20 CPIA #4 CPIA #4 ARPP - Proc. ARPP - Proc. Weight of each CPIA Governance Criterion 5% 5% 10% 5% Weight of each CPIA Non-Governance Criterion 5% 5% 3.3% 5% Governance / Non- Governance IDA Rating Impact Ratio 1/ 6.2 4.5 3.0 1.0 Allocation reduction resulting from a one point drop in a Governance Criterion 13.8% 10.0% 4.5% 2.2% Effective Weight of Governance in IDA Country Rating 67% 60% 40% 20% 1/ Ratio of the IDA country rating changes resulting from: (i) one point reduction in the rating of a Governance Criterion (e.g. 3.5 to 2.5) and (ii) a similar change in the rating of a non-governance criterion. - 5 - 9. Option II would be simpler, since it would eliminate the governance factor and focus exclusively on the CPIA's governance cluster (criteria number16-20),7 while giving significantly more weight to each of the five criteria in the governance cluster than to the non-governance criteria in the CPIA. Currently, in addition to the five criteria in the CPIA's governance cluster (number16-20), the governance factor draws on criterion number 4 (covering the sustainability of the development program), and the procurement criterion of the portfolio (ARPP) rating. Since the score of criterion number 4 turns out to be highly correlated (0.9) to the average score of criteria number16-20, the approach can be simplified by the special focus on criteria number 16-20. The procurement rating's correlation with the governance cluster, on the other hand, is less close (0.5). Still, dropping special consideration of this governance criterion would eliminate the double counting of governance criteria and increase the PBA System's simplicity.8 10. Under Option II: (i) a one point drop in one governance criterion would typically reduce the IDA rating by some 2.2%, resulting in a reduction in the IDA allocation of about 4.5% (since performance is squared in the IDA allocation formula);9 and (ii) a given rating change in a governance criterion would affect the IDA rating three times as much as a similar rating change in a non-governance criterion. This compares to over six times as much under the current governance factor approach. This would be a significant shift. Compared to the 2003 process, which applied the current governance factor for the FY05-07 allocations, Option II would result in a doubling of the average per capita per annum allocation of countries in the bottom quintile. Thus Option II's effort to moderating the year-to-year volatility of allocations would come at the price of weakening IDA's governance-allocation link. Switching to this approach - similar to those of the AsDB and AfDB - would thus significantly change the current stress on governance by flattening the performance-based allocation curve. 11. Fine-tuning the role of governance in the PBA System is a challenge since it requires a balancing of three objectives: (i) to maintain the dominant performance- allocation link, while keeping the focus on governance as a key performance dimension; (ii) to avoid excessive year- to-year volatility in the country allocations as a result of modest changes in governance ratings; and (iii) to increase the transparency of governance's role in the system. The two identified options each have their appeal and drawbacks. Option I leaves the strong governance ­ allocation link largely intact, and would somewhat reduce year-to-year allocation volatility. But it would not increase the transparency of the governance's role in the system. Option II, on the other hand, weakens the governance ­ allocation link substantially, but would result in a commensurate moderation of allocation volatility and would be more transparent with respect to the role of governance in the PBA System. Management looks forward to the Deputies' views on these difficult trade-offs. It also looks forward to the findings and recommendations of the CPIA Panel of Experts, which are expected to be finalized in March, 2004, and may have implications for the CPIA's governance criteria. 7 See Annex I. 8 The procurement dimension would continue, of course, to be an element of the (ARPP) portfolio rating. 9 For the resulting governance and non-governance cluster weights of each of these options see Annex II. - 6 - IV. WEIGHTING POVERTY IN THE PBA SYSTEM 12. It is generally agreed that aid should be primarily targeted to the poorest countries if it is to have maximum positive effect helping developing countries reach their economic and social objectives. To this end, IDA limits eligibility for its funding normally to countries with a per capita income below its operational cut-off level, currently $865. There is also broad consensus that among low-income countries, large-scale financial aid has more of an impact in an environment of sound institutions and policies. Accordingly, IDA has made its Country Performance Rating the dominant factor in its allocation among its eligible countries. At the same time, IDA's allocation formula provides a modest extra bias in favor of the poorest countries, by using an exponential of ­1/8 for GNI/capita in its allocation formula. 13. In preparation for IDA14, Management has given further thought to the relative weight of how poverty is treated in the allocation of IDA resources. A first issue is what measure of poverty to use. Specifically, is per capita income an appropriate proxy for targeting poverty? Statistical studies suggest a high correlation between headcount poverty and per capita income for most countries. Large discrepancies between the headcount poverty measure10 and the per capita income proxy are found only in a few cases. These may reflect countries' different income distributions, in which case the poverty headcount gives a better indication of the level of poverty. However, the discrepancies can also arise from errors in the household surveys, which in general are less reliable than national accounts. For example, a survey in Tanzania in the mid- 1990s generated a poverty estimate well below the level in some middle-income countries, which seems implausible in light of other information on Tanzania, including its per capita GNI. Furthermore, up-to-date direct poverty measures, such as headcount poverty or poverty gap, are difficult to obtain because they are based on household surveys which are conducted periodically ­ in some countries at ten-year intervals. Variations in the determination of the poverty line and in the methodology for poverty assessment also make the comparison of poverty levels among countries unreliable. Thus, Management continues to use per capita income as a proxy for poverty because: (i) it is more readily available for most countries on an annual basis; (ii) it is less subject to serious errors; and (iii) it is simple and transparent. 14. A second question raised by some of the IDA donors concerns the weight of favor of poverty in the allocation formula. It is worth noting that IDA uses a two-step per capita income two-step approach: countries above a certain level of income are not eligible at all, while for eligible countries the allocation formula has a coefficient of ­1/8 on the log of per capita income.11 The question then is whether a further bias in favor of the lowest income IDA eligible countries would increase the effectiveness of IDA resources in combating poverty. In preparation for the IDA Replenishment and as part of the global monitoring requested by the Development Committee, the Bank's research group has examined the poverty focus of different types of ODA. This analysis of poverty selectivity based on data across all developing countries and controlling for population and policy shows that IDA's PBA System is the most poverty 10For example: number of people living on less than $1 per day. 11IDA's Performance-Based Allocation System: Current and Emerging Issues, October 2003, page 3. - 7 - focused among development agencies.12 In light of this, the question raised by some IDA Deputies, whether the weight of poverty in IDA's PBA System should be increased further to favor the lowest income countries, needs to be approached with caution. 15. At this time, IDA is sharply focused on the poorest countries, and among these the system favors those that are better governed. Increasing the poverty weight in the allocation among the poor countries would de facto reduce the weight put on the quality of policies and institutions. Management is of the view that this would lead to less effectiveness in fighting poverty. While it is natural to focus on how the allocation formula distributes aid across poor countries, it should be kept in mind that it also affects how poor countries are treated over time. The negative coefficient on per capita GNI is essentially a tax on growth. Starting from the lowest level of per capita GNI among IDA countries, as a country attains some success in development (and its per capita income rises), the evidence suggests that the need for public investment in infrastructure and social services increases in absolute terms. Management has looked at how other aid allocation systems treat poverty. One research paper13 has estimated the allocation of aid that would have the maximum effect on poverty, under a certain set of assumptions. This "poverty efficient" allocation actually rises with per capita GNI up to a level of about $800, because of the fore-mentioned increasing need for public investment and increasing ability to absorb aid. However, to implement this "poverty efficient" allocation would require a complex formula. Still, it makes the useful point that the ability to absorb aid productively increases as GNI rises from an extremely low level. Beyond a certain level of per capita GNI, countries can turn to private markets and their own savings. The IDA approach roughly fits this pattern, with little distinction based on per capita GNI among the poorest countries, and then graduation to IBRD beyond a similar income level (currently $865). 16. The Bank's global monitoring analysis also found that most donor agencies have adopted this pattern of allocation. The allocation of bilateral aid as a whole actually has a positive relationship with per capita GNI among IDA-eligible countries. With its negative relationship, IDA is far more poverty focused than bilateral aid and more focused than most individual aid agencies, bilateral or multilateral. However, there are a couple of individual donors that have higher poverty selectivity among the IDA-eligible countries, but the differences are not statistically significant. All of these donors face the issue that if aid has a sharply negative relationship with per capita GDP among these low-income countries, this in effect is a heavy tax on growth. It would require poor countries that make a good start on growth to be weaned off of concessional assistance very rapidly, just when capacity to effectively use the resources may be increasing. 17. In conclusion, IDA is highly focused on poverty because only low-income countries are eligible. Given the widespread poverty in these countries and the limited IDA resources, the PBA System is used to steer those funds to where they are most likely to be effectively used to reduce poverty. Management remains of the view that a further increasing of the poverty weight would come at the cost of reducing effectiveness of the use of its scarce resources. It would in 12David Dollar and Victor Levin. The Changing Quality of and Quantity of Aid, 1984-2002. Paper under preparation for the spring 2004 meeting of the Development Committee. The CPIA is used to control policy. 13Paul Collier and David Dollar (2002). Aid Allocation and Poverty Reduction, European Economic Review. - 8 - effect be a tax on growth in these countries and would have the effect of forcing them too quickly out of IDA. V. LOW-INCOME COUNTRIES UNDER STRESS: ADDRESSING LICUS NEEDS WITHIN THE PBA FRAMEWORK 5.1 Overview 18. While taking steps noted above to calibrate the governance in the PBA System and make its role more transparent, Management has also been concerned about how the system can best support low-income countries with weak institutions and difficult governance conditions. This group of countries has been labeled Low-Income Countries Under Stress (LICUS). LICUS rate poorly on the CPIA and governance rating scales. While in line with the views expressed by the Board, there is no official list of LICUS, for internal monitoring purposes, the Bank classified 26 countries as LICUS in FY04, based on a rating of 3.0 or less on the CPIA and governance rating scales.14 Annex III summarizes the typology of LICUS in FY04. About one-third of LICUS are in non-accrual status and, hence, have no access to IDA financing. Other LICUS are meeting their financial obligations to IDA and are receiving IDA lending. 19. During the IDA13 Mid-Term Review discussions, Deputies requested further work on a framework for IDA's assistance to LICUS. The main issue is how to respond to specific LICUS challenges in a manner that responds to their varied circumstances while preserving the PBA System. This section discusses the scope of current IDA assistance to LICUS, including recent measures taken to strengthen IDA's capacity to provide effective support to these countries. 20. IDA has established a central unit to support assistance to LICUS at the country-level. The evolving framework of assistance is anchored in the PBA System and is responsive to varying country circumstances. In the case of LICUS in non-accrual status with difficult governance conditions, assistance is focused on: (i) working with partners to find a resolution to governance issues; (ii) building the statistical and knowledge-base to design appropriate interventions; (iii) supporting simple and feasible reforms; and (iv) exploring innovative approaches to aid the delivery of selected services for vulnerable groups, especially in the areas of HIV/AIDS and epidemic diseases. In the case of LICUS initiating steps towards normal relations with IDA and the international community, the form of assistance emphasizes pre- arrears clearance support in preparing transitional activities, including capacity building, and engaging country counterparts to take advantage of reform opportunities. In the more stable, active IDA countries with weak policies and governance, the LICUS initiative focuses on increasing the selectivity of reforms, grounded in sound socio-political analysis, and on expanding capacity-building efforts. The LICUS Implementation Overview report discussed by the Board on January 15, 2003, provides a discussion of progress under the LICUS initiative.15 14The cut-off point is based on the FY02 CPIA Ratings. 15Low-Income Countries Under Stress: Implementation Overview, December 19, 2003, IDA/SecM2003-0647. - 9 - 21. In view of their weak policy and institutional environment, LICUS receive lower per capita IDA allocations per annum than other IDA countries. Over recent years, LICUS' share in total IDA resources decreased. This is in line with the approach to assisting LICUS, emphasizing knowledge and capacity-building, rather than immediate lending. There are significant differences within the LICUS group however, and IDA has gradually expanded its menu of options in response. For instance, in IDA12 Deputies authorized special pre-arrears clearance allocations to aid eligible countries coming out of active conflict and in the process of normalizing IDA relations.16 In IDA13, Deputies further authorized use of IDA grants to support the recovery and stabilization of eligible post-conflict countries, most of which are LICUS. The rest of this section discusses, in turn, measures concerning: (i) all LICUS; (ii) post-conflict LICUS; and (iii) LICUS in non-accrual status. 5.2 General LICUS Approach 22. Recent analytical work and experience of LICUS assistance efforts suggest a need to strengthen IDA's assistance, including enhancing the applicability of the PBA System to countries at the bottom of the performance spectrum, focused capacity-building, appropriate phasing of post-conflict allocations to match absorptive capacity and supporting early reform for LICUS in non-accrual status. Management is adopting measures to respond to these concerns. · Adequate and Reliable Data. LICUS generally lack adequate and good quality data for policy diagnosis and the application of the PBA System. The problem is more pronounced in conflict afflicted countries facing loss of skilled personnel and the degradation of institutions and facilities for data collection and processing. Restoration of statistical and managerial capacity and knowledge-building are top priorities for LICUS and is reflected, for example, in work being done in Angola, Burundi, Central Africa Republic, Somalia, and Sudan. In the Africa Region, which covers many of the post-conflict LICUS, a plan to fill gaps in data and analytical work is being implemented. The numbers and quality of core analytical work have increased and household surveys needed to track poverty and social development trends are being compiled. · Adjusting the Performance Rating Scale. The current CPIA rating scale is broadly viewed as an appropriate tool for assessing country performance. However, there is a concern that it does not capture adequately changes in performance at the bottom end of the scale for LICUS making efforts to improve their performance. This is because the current rating system, starting from "2" instead of "1," limits differentiation in performance (currently there is no very weak "1" rating: a country receives a "1" in case it rates "2" for at least three consecutive years). At the same time, relatively small improvements in governance and other indicators at the bottom could have marked implications in terms of improving the overall ratings and associated IDA allocations. To help capture more accurately differences in performance, Management recommends amending the CPIA scale. The measure will involve extending the rating scale by adding 16Access to this assistance is a last resort and requires: (i) the country to be making progress towards peace; (ii) concerted international effort to support a program of social economic recovery; and (iii) creditors to agree to arrears accumulation. - 10 - below the current weak "2" rating a very weak "1" rating and defining for each of the criteria the corresponding performance-level and scoring guidelines.17 While this change has only modest implications for allocations, it does allow for finer differentiation of performance, especially in looking at relative status of LICUS countries. · Institutional Development Fund. All LICUS are characterized by weak institutions and inadequate capacity, and stronger capacity-building is, therefore, a critical element of IDA's programming. Management has recently approved the inclusion of LICUS as a focus area under the Institutional Development Fund (IDF). The measure will enable priority capacity-building in social service delivery institutions for LICUS. 5.3 Conflict and Post-Conflict LICUS18 23. The Bank provides assistance for conflict-afflicted countries as well as countries emerging from conflict. The Post-Conflict Fund (PCF), for example, works with development partners to support a range of activities, including knowledge-building and delivery of selected services.19 In several countries, PCF-supported conflict analysis is helping to deepen understanding of conflict and the design of appropriate interventions. The Bank's country operational budget and the World Bank Institute support LICUS capacity-building efforts. IDA12 initiated a special pre-arrears clearance grant for eligible countries emerging from conflict to enable quick and flexible support in the immediate post-conflict phase. IDA13 provided for expanded grant support for eligible post-conflict countries. Operational Policy (OP) 2.30 provided for exceptional financial assistance, subject to the Bank's financial rules, to meet the transitional financial needs in a timely manner for post-conflict countries normalizing relations with IDA. Management is implementing the following adjustments in the current support system to strengthen support for post-conflict countries. · Post-Conflict Progress Indicators. LICUS emerging from conflict are rated on the basis of special post-conflict progress indicators (PCPI). In this connection, the PCPI replaces the normal CPIA within the PBA System. Management is in the process of completing the full set of characterizations of each of the rating levels for the PCPI criteria, with a view to limiting further the scope for subjectivity in the rating process. · Stretching Post-Conflict Allocations over Time. Recent analytical work raised concerns about the matching of implementation capacity relative to resource flows in post-conflict countries. While resource inflows tend to be larger in the early years, countries tend to 17Ratings at the top of the CPIA scale will also be redefined by adding a score of 6 to permit better differentiation between strong and excellent performance levels. This extended 1-6 scale would apply to all IDA countries. Under the current system countries score "6" in case they rated "5" for at least three consecutive years. 18"Post-conflict" has various definitions, depending on the intensity of the conflict afflicting the countries. The designation in this paper is within the narrow context of countries that are eligible for special IDA allocations. 19The Post-Conflict Trust Fund (PCF) supports activities which are not normally funded under country operational budgets. For example, PCF assistance enabled the Bank to work with Sudan's Central Bureau of Statistics and UNICEF to undertake a Multiple Cluster Indicator Survey to update human development indicators. In Somalia, the PCF supported non-state providers to deliver health services targeted at women and children. - 11 - achieve improved absorptive capacity in the subsequent years.20 Sustaining peace- building efforts and improving aid productivity in these countries require adjusting the pattern of aid flows to take account of capacity improvements. During the IDA13 Mid- Term Review, Deputies supported measures to stretch out the phasing of the special post- conflict allocations while keeping the same total allocation. Eligible countries may now receive the special post-conflict allocation for four years instead of three, and there will be three years of transition to the normal PBA allocation instead of two. This approach gives country teams greater flexibility in phasing commitments appropriately to support country programs. 5.4 Other Special LICUS Situations 24. Beyond the already difficult LICUS cases described above, there are some cases where because of their non-accrual status IDA support is not appropriate. Countries in this situation include Central Africa Republic, Haiti, Myanmar, Togo and Zimbabwe. In addition, some countries coming out of conflict may not be eligible for post-conflict pre-arrears clearance support. Management is taking the following steps to fill this gap in the current assistance framework. · LICUS Trust Fund. A proposal to establish a LICUS Trust Fund of $25 million, financed through transfers from IBRD surpluses was discussed by the Bank Board on January 15, 2004. The proposal was endorsed and is now awaiting final approval from the Board of Governors. The LICUS Trust Fund will focus mainly on providing assistance, prior to arrears clearance, to LICUS in non-accrual status mired in difficult institutional and governance conditions.21 In exceptional cases, where the use of IDA resources would not be appropriate, the Trust Fund could support a Bank contribution to a multi-donor effort in other LICUS. The Fund will seek to leverage efforts of other development partners and will focus on: (a) knowledge and capacity building; (b) aiding country counterparts in LICUS to implement early reform opportunities; and (c) on assisting with the delivery of selected services to stem deterioration in social development indicators. The fund will be administered by the existing Post-Conflict Unit and will have a life span of three years. · Regional Projects. During FY04, Management initiated the implementation of a pilot program of regional projects. As set out in the paper describing the pilot program,22 the initiative involves exceptional inclusion of LICUS in non-accrual status in case exclusion of neighbors in non-accrual status would undermine coherent implementation and results. This problem applies especially to HIV/AIDS projects along regional transport corridors with high risks of cross-border infections. The initiative involves a pilot envelope of up 20Paul Collier: Aid Policy and Growth in Post-Conflict Countries, September, 2002. 21In exceptional cases, the LICUS Trust Fund will finance Bank contribution to a multi-donor effort for LICUS which may not be in accrual status. 22Pilot Program for Regional Projects, September 30, 2003, IDA/SecM2003-0532. Information Note provided to the World Bank Executive Board. - 12 - to $300 million per annum during FY04-FY05 to support such regional projects.23 Non- state suppliers would implement the part of the project in the LICUS in non-accrual status. Eligible projects would involve three or more countries; they would be based on a regional strategy with clear evidence of regional ownership and spill-over effects across national boundaries. A pilot is under way for the Pan Caribbean Community HIV/AIDS Project and the Lagos-Abidjan Corridor HIV/AIDS Project. · Partnerships. Donor coordination is particularly crucial in non-accrual LICUS, where the capacity does not exist to manage fragmented donor programs. LICUS re- engagement often requires multi-donor effort, for example, the joint donor missions now underway in Sudan and CAR. IDA is also working to build partnerships to assist LICUS in such areas as joint analytical work and preparation of projects that could be fully or partially funded by other donors. For example, IDA encourages LICUS countries to develop project proposals to be funded by the Global Fund for Aids, Malaria and Tuberculosis; and in Sudan, the Bank prepared an irrigation reform report which was then supported by IFAD on a pilot basis. 25. The experience with these measures that address LICUS circumstances while remaining consistent with the PBA System, will continue to be monitored with a view to learning what works on the ground and to make adjustments where warranted. 23An additional $150 million is expected to be funded from participating countries' allocations. - 13 - VI. SUMMARY AND THE ONGOING EVOLUTION OF THE PBA SYSTEM 26. The PBA issues revisited here - poverty weighting, the calibration of governance, and the LICUS framework - all interact at the weak performing country-level. Management believes that the PBA's modest bias in favor of the poorest IDA countries should not be increased, since it may reduce the effectiveness of the use of IDA resources. Instead, the LICUS framework provides an approach, guided by a typology of weak performing countries, which tailors the mostly non-lending support to the country's specific condition, be it the country's coming out of civil conflict, or its having slipped into arrears on servicing its IDA debt. As for the possible recalibration of the PBA's stress on governance, Management looks forward to the Deputies views on the observed trade-off between maintaining the governance-allocation link and efforts to: (i) reduce excessive year-to-year allocation volatility; and (ii) increase the transparency of the role of governance ratings. It also awaits the findings of the CPIA Panel of Experts, to see to what extent they might have implications for IDA's approach to stressing governance in the PBA System. At this time Management recommends: · To leave the poverty weighting in the PBA allocation formula unchanged (paras 12-17); and · To revise the CPIA 1-6 assessment scale by adding at either end of the scale a very weak and a very strong performance-level for each criterion (para 22). 27. IDA's PBA System has evolved over the last two decades and will continue to do so during IDA14 and beyond. A number of factors can be identified at this time that will provide a special impetus to this evolutionary process, including: (i) in March 2004 the CPIA Panel of Experts is expected to report on their findings and recommendations on how the assessment system could be strengthened; (ii) the planned increased disclosure of IDA Country Performance Ratings should strengthen the outside peer review of the assessment process on which the PBA System is based. - 14 - Annex I ­ Summary of the CPIA Criteria Box 1 - CPIA Criteria A. Economic Management 1. Management of Inflation and Macroeconomic Imbalances 2. Fiscal Policy 3. Management of External Debt 4. Management and Sustainability of the Development Program B. Structural Policies 5. Trade Policy and Foreign Exchange Regime 6 Financial Stability and Depth 7 Banking Sector Efficiency and Resource Mobilization 8. Competitive Environment for the Private Sector 9. Factor and Product Markets 10. Policies and Institutions for Environmental Sustainability C. Policies for Social Inclusion/Equity 11. Gender 12. Equity of Public Resource Use 13. Building Human Resources 14. Social Protection and Labor 15. Monitoring and Analysis of Poverty Outcomes and Impacts D. Public Sector Management and Institutions 16. Property Rights and Rule-Based Governance 17. Quality of Budgetary and Financial Management 18. Efficiency of Revenue Mobilization 19. Quality of Public Administration 20. Transparency, Accountability and Corruption in the Public Sector - 15 - Annex II ­ IDA: Alternative Cluster Weights IDA Cluster Weights ­ Alternative Ways of Stressing Governance GF - Current Option I Option II Macro / Structural Criteria 36% 36% 27% Social Criteria 20% 20% 13% Non-Governance Criteria 56% 56% 40% Governance Criteria 24% a/ 24% b/ 40% Portfolio 20% 20% 20% Total 100% 100% 100% a/ Including the Governance Factor the effective weight is 67% b/ Including the Governance Factor the effective weight is 60% - 16 - Annex III ­ LICUS Typology C&PC QI QII QII Afghanistan Angola Burundi Congo Dem. Rep Congo Republic Sierra Leone Liberia Timor Leste Somalia Guinea Bissau Sudan NON-ACCR ACCR QIII QIV CAR QComoros IV Kosovo Haiti Laos PDR Myanmar Nigeria Togo Papua New Guinea Zimbabwe Sao Tome & Principe Solomon Islands Tajikistan Uzbekistan NPC ACCR Accrual status NON-ACCR Non-Accrual status C&PC Conflict &Post-Conflict NPC Non-Post-Conflict