Document of The World Bank

Transcription

1 Public Disclosure Authorized Public Disclosure Authorized Public Disclosure Authorized Public Disclosure Authorized AFTP1 Africa Region Document of The World Bank IMPLEMENTATION COMPLETION REPORT (IDA-35630) ON A CREDIT IN THE AMOUNT OF US$6 MILLION TO THE UNION OF THE COMOROS FOR A EMERGENCY ECONOMIC RECOVERY CREDIT MAY 20, 2003 Report No: 25705

2 CURRENCY EQUIVALENTS (Exchange Rate Effective April 1, 2003) Currency Unit = Comorian Franc (KMF) KMF 1 = US$ US$ 1 = KMF F458 FISCAL YEAR July 1 June 30 ABBREVIATIONS AND ACRONYMS CURE EERC ICR GDP HIPC IMF ISS KMF QAG MOP MTR N/A OAU OP PC PCD PDO PID PIU PNAC PRSP PSR SMP TF USD/US$ Crédit d'urgence au Redressement Economique Emergency Economic Recovery Credit Implementation Completion Report Gross Domestic Product Highly Indebted Poor Country International Monteray Fund Interim Support Strategy Comoros Franc Quality Assessment Group Memorandum of the President Mid Term Review Not Applicable Organization of African Unity Operational Policy Personal Computer Project Concept Document Project Development Objective Project Information Document Project Implementation Unit Pharmacie Nationale Autonome des Comoros Poverty Reduction Strategy Paper Project Status Report Staff Monitored Program Trusd Fund United States Dollar Vice President: Country Manager/Director: Sector Manager/Director: Task Team Leader/Task Manager: Callisto Madavo Hafez Ghanem Philippe Le Houerou Colin Scott

3 COMOROS Emergency Economic Recovery Credit CONTENTS Page No. 1. Project Data 1 2. Principal Performance Ratings 1 3. Assessment of Development Objective and Design, and of Quality at Entry 2 4. Achievement of Objective and Outputs 5 5. Major Factors Affecting Implementation and Outcome 8 6. Sustainability 8 7. Bank and Borrower Performance 9 8. Lessons Learned Partner Comments Additional Information 19 Annex 1. Key Performance Indicators/Log Frame Matrix 22 Annex 2. Project Costs and Financing 23 Annex 3. Economic Costs and Benefits 24 Annex 4. Bank Inputs 25 Annex 5. Ratings for Achievement of Objectives/Outputs of Components 26 Annex 6. Ratings of Bank and Borrower Performance 27 Annex 7. List of Supporting Documents 28 Annex 8. Expenditure Table 29

4 Project ID: P Team Leader: Wolfgang Fengler Project Name: Comoros Emergency Economic Recovery Credit TL Unit: AFTP1 ICR Type: Core ICR Report Date: May 20, Project Data Name: Comoros Emergency Economic Recovery Credit L/C/TF Number: IDA Country/Department: COMOROS Region: Africa Regional Office Sector/subsector: Central government administration (100%) Theme: Conflict prevention and post-conflict reconstruction (P); Macroeconomic management (P); Decentralization (S); Other social protection and risk management (S) KEY DATES Original Revised/Actual PCD: 05/09/2001 Effective: 09/18/ /18/2001 Appraisal: 06/21/2001 MTR: Approval: 08/02/2001 Closing: 12/31/ /31/2002 Borrower/Implementing Agency: Other Partners: GOVERNMENT/PLANNING COMMISSION None STAFF Current At Appraisal Vice President: Callisto E. Madavo Callisto E. Madavo Country Director: Hafez M. H. Ghanem Hafez M. H. Ghanem Sector Manager: Philippe H. Le Houerou Cadman Atta Mills Team Leader at ICR: Wolfgang Fengler Christos Kostopoulos ICR Primary Author: Colin S. Scott; Vincent Da Cruz 2. Principal Performance Ratings (HS=Highly Satisfactory, S=Satisfactory, U=Unsatisfactory, HL=Highly Likely, L=Likely, UN=Unlikely, HUN=Highly Unlikely, HU=Highly Unsatisfactory, H=High, SU=Substantial, M=Modest, N=Negligible) Outcome: S Sustainability: UN Institutional Development Impact: M Bank Performance: S Borrower Performance: S QAG (if available) Quality at Entry: Project at Risk at Any Time: Yes ICR S

5 3. Assessment of Development Objective and Design, and of Quality at Entry 3.1 Original Objective: Country background: The Union of the Comoros comprises the three Comoros Islands that declared independence from France in 1975: Moheli, Anjouan, and Grande Comore. The central government is located on Grande Comore. For many years, the situation in Comoros has been marked by poverty, poor governance and extreme political instability, including numerous coups and coup attempts. In 1997, conditions deteriorated further: military authorities in Anjouan, the poorest of the islands, attempted to secede and both the central government and the Organization of African Unity imposed a blockade, completely isolating the island. Differences over how to proceed sparked internal disputes on Anjouan, including outbreaks of violence, but negotiations and attempts to forcibly re-integrate Anjouan failed. This failure in turn increased tensions on the island of Grande Comore, leading to the overthrow of the central government by Colonel Azali Assoumane in March Subsequently, negotiations made progress and, in February 2001, all parties signed a Framework Agreement for National Reconciliation in the town of Fomboni. This agreement outlined a path to reunification and democratization of the country, including a transition period, a referendum on a new constitution, a government of national unity and finally elections in early Hopes that real change was possible were high, including among much of the World Bank country team. In Comoros, per capita GDP has fallen throughout much of the past decade and now is below US$400. Social services delivery, never very good, deteriorated further during the secession, particularly in Anjouan. In 1998, the Bank suspended disbursements due to non-payment of arrears, but by January 2000, the new central government cleared arrears to the Bank, leading to the resumption of Bank disbursements and a review of the country s portfolio. World Bank strategy: In November 2000, the Bank adopted an Interim Support Strategy (ISS) with three scenarios: in case of renewed violence, the Bank would suspend its activities; if there was no change to the situation, the Bank would only carry out an infrastructure project, capacity building and assist the government to define a development vision; and in the high case scenario, the Bank would go ahead with an Emergency Recovery Credit, eventually leading to a PRSP and debt relief under HIPC. The high case would be triggered by (i) substantial progress in public management on all three islands, (ii) agreement on an integrated budget for the three islands, and (iii) evidence of Government transfers to, and Government expenditures in, Anjouan. This would imply tangible and decisive progress towards national reconciliation. High case scenario: A Bank mission to the Comoros in March 2001, just after the Framework Agreement was signed, concluded that there had been sufficient progress to warrant the high case scenario. At an informal donors conference in July 2001, the government presented a US$12.8 million Emergency Reconstruction and National Reconciliation Program of which the Bank was to fund about half. The Bank s country team showed strong support for the government s reconciliation strategy and rapidly prepared a US$6 million Emergency Economic Recovery Credit (EERC) under OP This credit was approved by the Board of Directors in August Six days later the local government in Anjouan was overthrown in a coup. The Bank sent a verification mission which concluded that the new authorities in Anjouan would continue to support the Framework Agreement and went ahead with the implementation of the credit. EERC objectives: The Memorandum of the President states that the credit's main objective was to finance the critical transition out of the secessionist crisis. The credit would do this by supporting a government economic recovery program which intended to alleviate distress among the poor and support the process of national reconciliation. This corresponded to the view of the government that the secession was the result of poverty on the islands concerned. Therefore, in the government s program, (i) a poverty-reduction component was to address the root causes of the conflict, (ii) a decentralization component was to pave the way for the outcome of the transition, allowing the country to function as three islands with largely autonomous governments, and (iii) a national reconciliation component was to pay for meetings, travel and seminars, thereby supporting the transition process itself. The MOP also - 2 -

6 states that credit was intended to signal that the international financial institutions were willing to help the Comoros and to leverage additional donor funding. Consistency with strategies: The Emergency Recovery Credit was a central element of the Bank's Interim Support Strategy for the Comoros, which stated that there can be no sustainable growth and poverty reduction in Comoros without a solution to the separatist crisis and a move to a broadly accepted government. Its objectives complemented the Framework Agreement and corresponded to the government s Policy for National Reconciliation. The Bank s Interim Support Strategy was therefore closely aligned with the objectives and sequencing of priorities of peace accords and rehabilitation plans agreed to by parties to the conflict (OP 2.30 Development Cooperation and Conflict ). High risk: In practice, the credit was to allow the Bank to back a largely political transition process that seemed to offer the best chance in years to pull the country out of an ever-worsening crisis. Even though this process was supported by the military rulers and opposition of both Comoros and Anjouan, the Bank was well aware that investing in it meant taking a high risk. In particular, the outcome of the transition was far from clear and the track record of the Comoros was extremely poor. The risk that the national reconciliation process could stall or reverse was stated in the MOP. The only safeguard mentioned, however, was to ensure that all stakeholders confirmed their support for the reconciliation process before the project was formulated. Choice of instrument: For this operation, the Bank chose an Emergency Credit under OP 8.50 and adapted it to the circumstances. The Comoros did not qualify for a credit under OP 8.60 and, as speed was crucial, the situation required a fast-disbursing instrument. Yet, in light of the track record of Comoros, and particularly Anjouan, the country team added controls to the standard requirements of OP Disbursements were to be made in three installments, dependent on the achievement of key milestones in the transition and satisfactory financial management. External audits were to be carried out for both imports and programs and the government s spending plan was binding, with changes in the allocation requiring no-objection from the Bank. Implementation was to follow World Bank procurement and disbursement guidelines as well as procedures to be established by the implementing agency. Drawing on the experience of Burundi, these measures were defined in a supplemental letter to the Development Credit Agreement. The Bank undertook to supervise the project closely. Indicators: Emergency Recovery Loans under OP 8.50 are designed to finance only a positive list of imports identified as necessary to a well defined recovery program. In this case, it had been agreed that the proceeds of the credit would be used for projects and activities as listed in the government s Emergency Reconstruction and National Reconciliation Program. The government s program had no logical framework matrix and was largely defined by activities and outputs rather than outcomes. Assumptions about causal relationships between national reconciliation, decentralization, poverty reduction, secession, and the transition did not have to be articulated. The absence of targets and indicators in the government s program made the evaluation of the outcomes much more difficult. However, this absence of indicators was a reflection of the limited information available at the stage of the project design and the result of the need to implement the program rapidly. Similarly, the credit documents do not state explicitly what would indicate that the transition was completed, although there was a general understanding that the objective would be achieved if the Framework Agreement was successfully implemented. With hindsight, the objectives should have been more specific. 3.2 Revised Objective: The objective was not revised

7 3.3 Original Components: The credit s MOP does not contain formal components. However, the borrower agreed with the Bank to use the Comoros EERC to finance: i) poverty reduction, ii) decentralization, and iii) national reconciliation, and the spending program was divided into these three components. For poverty reduction and decentralization, funds were allocated to specific projects which the borrower and authorities on the three islands had identified. In Anjouan, an ad hoc survey was undertaken to determine priorities. The Technical Annex and Supplemental Letter included a matrix (see Annex 2 Project cost and financing) defining how much money was to be spent for what kind of project on which island. Funds for national reconciliation were used at the central level. Emergency program components not financed by the EERC: In the government s Emergency Reconstruction and National Reconciliation Program, "national reconciliation" also included the reinsertion of young militia in Anjouan (financed by a Bank Post-Conflict Grant) and elections (financed by the European Union). A fourth component, "Economic Recovery and Financial Stabilization" covered payment of domestic arrears, wages and voluntary separation expenditures, which were to be financed by other donors. The government s program therefore offered a comprehensive package of measures to assist the transition out of the crisis. Spending plan: Beyond the allocation to components and islands, the MOP does not state what the funds are to be spent on. Details are contained in the Letter of Policy for National Reconciliation, which proposes the government s action plan to the donors. Given the sensitivity of fund allocation to the different islands, establishing how much was to be spent on each island before disbursement was an important factor for successful implementation. The spending plan was said to be binding for the borrower and any changes were subject to no-objection by the Bank. In practice, this left considerable flexibility, particularly as assumptions about how the activities would further the objectives were not made explicit. This design constituted an additional risk and therefore depended to a considerable extent on an unwritten agreement between the Bank and the borrower. Issues related to this risk were discussed within the Bank, and the prevailing view was that, on the basis of the recent track record, the borrower could implement this satisfactorily. Distress among the poor (first component, 50% of credit funds) was to be alleviated through the rehabilitation and equipment of schools and hospitals, through water and rural development projects. The institutional transition to a decentralized state (second component, 30% of credit funds) was to be assisted by constructing and rehabilitating administrative buildings and re-allocating or training civil servants. Assistance in carrying out the technical dimensions of the reconciliation process (third component, 15% of credit funds) covered unspecified activities of the commission charged with implementing the Framework Agreement and the establishing of a new fiscal structure. For this component, flexibility was an important feature as it could not be predicted with certainty what kind of actions it was going to take to attain the overall objective of assisting the transition process. The project implementation unit s running cost (another 5% of credit funds) as well as further support of the implementing partner were also officially budgeted under the heading of reconciliation. Implementing Agency: For the implementation of the activities to be financed by the credit, the government was to create a project implementation unit coordinated by the Director General for Planning. The General Directorate for Planning (later renamed Planning Commission) reports directly to the President. Two sub-committees consisting of representatives from ministries and other government agencies were responsible for mobilization and control of funds and implementation respectively. A sub-unit replicating this structure was to be set up on the island of Anjouan in order to implement projects there. This decentralization reflected the large number of activities to be carried out in Anjouan and took into account the fact that the administration of Anjouan was de facto operating independently from the central government. Both units were to be audited by internal and external auditors

8 3.4 Revised Components: The credit s MOP does not contain formal components. In the government s spending program, funds were re-allocated between the components and to other activities with the no-objection of the Bank. These re-allocations were notably to respond to overcome obstacles in the privatization of the port and to make up for delays in other donors fulfilling their commitments on financing civil servants salary arrears. 3.5 Quality at Entry: No Quality at Entry rating was carried out by the Quality Assurance Group. This ICR found the following features of the credit unsatisfactory: the spending program s emphasis on activities rather than outcomes, the absence of indicators to measure how much progress has been made and the fact that assumptions about how the activities were to serve the general objective were generally not made explicit. Different aspects of the transition - its "technical dimensions", implementation of the Framework Agreement, reunification, agreement on a joint budget and national reconciliation - were not clearly distinguished and their relationship not spelt out. These shortcomings lead to different interpretations of the overall objective and, particularly on the part of the government, to unrealistic expectations of the role the credit would play in fostering reconciliation. Satisfactory were: the overall objective and rationale as stated in the MOP, the rapid response of the credit to an urgent need (very satisfactory), the high level of government input into the planning and design of the activities to be carried out, and the appropriate flexibility of design with additional controls for a high-risk post-conflict transition. Bearing in mind the time constraints and trade-offs and that the chosen instrument had to be tailored to the specific situation of Comoros, the overall Quality at Entry is rated marginally satisfactory. 4. Achievement of Objective and Outputs 4.1 Outcome/achievement of objective: State of the transition: The uses to which the counterpart funds have been put have produced largely positive results. In terms of the transition, however, the outcome is mixed. At the closure of the credit in December 2002 and at the time of writing of the ICR, in March 2003, the transition process, as envisaged by the Framework Agreement, had not been completed. On the one hand, considerable progress had been made: the constitution had been adopted in a referendum, and presidents had been elected for the Union and for each of the three islands. Anjouan, the principal focus of the EERC, was peaceful, had overcome its isolation, and was once again participating in the federation. These outcomes, to which the credit contributed, are substantial successes. On the other hand, new lines of confrontation between the four entities in the Comoros had emerged, particularly between the newly formed government of Grande Comore and the central government located on the same island. Specific responsibilities of the four governments remained a point of contention and much of how the new federal country will work remained to be determined, let alone implemented. Whether the institutional situation of the Comoros is better today than two years ago is debatable, but it is certain that the secessionist crisis has not been overcome. Not only is the implementation of the Framework Agreement not complete, but it has also become apparent that a full transition will require more than the implementation of that agreement. Achievement of EERC objective: The credit s stated objective - financing the critical transition out of the secessionist crisis - has only been partially achieved. The project had a positive impact on poverty and it supported the transition for as long as it lasted. Nevertheless, in light of the state of political instability at the close of the credit, the ICR could not rate the outcome higher than marginally satisfactory. Even so, the political situation may still be resolved and the project may yet prove to have been an important contributing factor. 4.2 Outputs by components: Financing the transition: Through its three components, the credit successfully provided means and incentives to move forward with the transition process. It was also central to the Bank s dialogue with the - 5 -

9 Comorian governments and used in creative ways to overcome obstacles that had not been anticipated. As intended, the impact was particularly strong in Anjouan, where the poverty reduction component was able to address urgent problems that had arisen during years of isolation. The achievements financed by the credit gave considerable support to the view the island had something to gain from re-engaging with the central government and thereby helped the passing of the new constitution in a referendum. Most of the benefits of the decentralization component are not yet visible, as the major activities involve construction that is as yet unfinished. For the national reconciliation component, results are less tangible, but by helping to finance meetings and studies for the commission and committee charged with implementing the transition, the credit generally helped to move the process forward. In the absence of indicators, the contribution of the activities financed by the credit to poverty reduction, decentralization or national reconciliation cannot be quantified here. The following paragraphs highlight the main outputs. An overview of the large number of separate activities can be seen in the expenditure table in Annex 8. In general terms, although some of the activities funded seem somewhat removed from the objectives of the project, the vast majority have produced useful outputs and also had an intrinsic value through confidence building. Poverty reduction: Poverty reduction is not just subservient to the political transition but an important objective in its own right. Of the credit s funds, 47% were allocated to this component and have achieved substantial successes. The spending program contained projects in four sectors: health, education, water, and rural development, the latter including a project for the safety of fishermen. Almost all the projects have been completed as planned. Half the poverty reduction funds were allocated to the island of Anjouan, where most projects have been completed rapidly and with high visibility. More than 900 people were employed on the island in the process. A rapid survey conducted by the implementing partner in Anjouan covered 0,5% of the adult population. It claims that almost half the respondents had heard of the project, just over ten percent said the implementing partner responded to their needs and almost a quarter saw work being carried out in their village. Out of the latter, over a third said they had been consulted on the project and over 90% expressed their satisfaction with the work. Most of the poverty reduction funds were spent in the health sector (39%), with the purchase of medicine and medical gas, medical equipment, ambulances, the rehabilitation of hospitals and setting up a cholera camp during an epidemic. Much of this has been completed and is operational in the respective hospitals. The positive impact of this is evident: ambulances reduce delays in attending emergencies; new dentistry, ultra-sound and surgical equipment reduces the need to travel between islands for check-ups and treatment. In the education sector (28%), schools were built or rehabilitated and equipment purchased on the island of Anjouan, and again, the improvement of learning conditions is obvious. Much-needed teacher training, however, does not seem to have taken place. One major project (12% of total poverty reduction expenditures) to supply water to 17 villages on Grande Comore is still being completed, suffering from serious delays. On Anjouan a multitude of small infrastructure projects covering canalization and minor road works have been completed (7.5% of total poverty reduction expenditures). Also under the heading of rural development, a project on all three islands that intends to reduce the losses of fishermen at sea has been financed (12%). Unfortunately, rather than buying rescue boats, the funds were largely used to set up sophisticated communication facilities which are likely to become redundant once a mobile phone network is operational. Ratings for individual projects would seem to cover the whole range from very satisfactory to unsatisfactory and would require the definition of indicators and a beneficiary evaluation. On the basis of available information the poverty reduction component appears satisfactory overall. Decentralization: The funds of this component (18% of the credit s funds) were mostly intended for the construction and rehabilitation of public buildings, to finance redeployment of civil servants to Anjouan and training for civil servants on Grande Comore. The largest single project (42% of the decentralization expenditures) is the construction of an office complex for the administration of Moheli, which in December 2002 was yet to start. A similar amount has been invested in the building, rehabilitation and equipment of four administrative buildings on Anjouan, where work is still ongoing. Much of the rest of decentralization funds (13%) was used by the island governments to buy vehicles (eight in total) for their administrations. As much of the major work is still being completed, rating this component would be - 6 -

10 premature. Given the prevalence of cars among the outputs produced so far, the decentralization component is unsatisfactory at this stage. National reconciliation: The funds for national reconciliation (4% of the credit) supported the Committee and Commission responsible for the implementation of the Framework Agreement. These played a central role in the transition process when it was on track, and the credit helped both overcome financial obstacles and provided technical assistance in line with its objective. Because part of the budget was earmarked for their activities, Committee and Commission wanted the amount to be simply disbursed to them. However, the implementing agency kept control over the funds, initially paying them out against running costs and expenses, later only against specific outcomes, such as technical assistance on the distribution of public property and the establishment of a joint budget. Despite the progress and achievements in the area of national reconciliation, given that the transition has not been completed, this component cannot be rated higher than marginally unsatisfactory. The running costs of the Project Management Unit (4% of the total credit funds), including audits, were formally included under the heading of national reconciliation, which also covered support to the planning commission of which the PIU was part (5% of the total credit funds). This included salaries of staff contracted for the project, rehabilitation of offices on the three islands and the purchase of vehicles and equipment. Rating: satisfactory. Civil service salary arrears: Payment of civil service salary arrears was part of the government s recovery program, but belonged to a component that other donors were to finance. Following delays in the release of those funds, the Bank agreed to 10% of the total credit fund s being used for this purpose. Rating: satisfactory. Port privatization: In order to assist privatization, and with agreement of the Bank, the implementing partner set another 10% of the credit s funds aside to compensate port workers of the state maritime transport company. These funds came out of the national reconciliation and poverty reduction budgets and were a crucial factor in the successful privatization of the main port on Grande Comore. At the time of writing, the privatization of the port in Anjouan was still object of dispute so that funds for port workers there have not been disbursed there yet. Rating: satisfactory. Economic recovery: According to IMF data, the EERC was one of the main reasons why real GDP growth increased from -1.1 percent in 2000 to 1.9 percent in 2001 (the lifting of the embargo on Anjouan and favorable export prices being the other important factors) and remained at an estimated 2.5 percent in The contribution of the EERC was mainly through increased commercial activity and construction. Rating: satisfactory. Other donors: As intended, the engagement of the Bank did send a clear signal that it is willing to support the transition process in the Comoros. Other donors did contribute to the government s Emergency Reconstruction and National Reconciliation Program but the credit did not succeed in leveraging the additional donor funding it intended. Rating: marginally unsatisfactory. 4.3 Net Present Value/Economic rate of return: With the exception of minimal cost recovery at hospitals and clinics, the investments of this project do not produce revenue that could be used to quantify the economic rate of return. No data is available on the development of such revenues. As an indication of efficiency, the auditors have found prices paid for goods and services reasonable and costs were comparable to those in similar conflict affected countries. The project implementation unit s running costs represented 4% of the credit s funds and an additional 5% of the credit s funds spent on investments in the Planning Commission cross-subsidized the PIU to some extent. This is justified, given the multitude of projects and the high speed at which they were completed. The efficiency of the use of resources is rated satisfactory. 4.4 Financial rate of return: N/A - 7 -

11 4.5 Institutional development impact: The credit did not include an explicit institutional development objective. Nevertheless the experience of implementing a credit of this type which gives a large amount of autonomy to the implementing agency has had a modest capacity-building effect, particularly for island authorities which have not previously implemented projects of this type. The Planning Commission, to which the project implementation unit belongs, has become one of the few institutions of Comoros that effectively function in a deconcentrated way. Rating: satisfactory. 5. Major Factors Affecting Implementation and Outcome 5.1 Factors outside the control of government or implementing agency: No major exogenous factors affected the credit s implementation. Occasional delays in contracting and procurement, disagreements between authorities and beneficiaries, and bad weather meant that some of the deadlines, implementing partners set for their projects, were not met. 5.2 Factors generally subject to government control: The major factor affecting implementation and leading to the limited attainment of its objective in supporting the transition was ongoing political instability and disagreements between the central and island governments. Although difficulties had been anticipated, it is clear that an emergency credit alone cannot assist a transition process if the parties involved have no shared or compatible vision of where the transition is to lead to. In addition to that, several governments have pressed for re-allocation of funds to items for which the emergency recovery program was not intended. 5.3 Factors generally subject to implementing agency control: The implementing agency drafted and implemented its own procurement guidelines, but these were only formally approved by the Bank in December In a charged political atmosphere, differences between islands at times created tensions between the central and Anjouan project implementation units. Whereas it occasionally required Bank intervention to resolve the issues, such tensions did not have a negative impact on project implementation. 5.4 Costs and financing: There were no delays in disbursements. Due to exchange rate fluctuations between the US dollar and the Comorian Franc, commitments exceed available funds by 3%. 6. Sustainability 6.1 Rationale for sustainability rating: Sustainability of a project in the context of a conflict-affected country is likely to take a different form that in a conventional development project. Since the objective is supporting the transition, it is the confidence and trust built through the project with its contribution to the sustainability of peace which counts. Many of the interventions were never intended to be orthodox sustainable development. Others may prove sustainable in the conventional sense. Most of the investments under the poverty reduction and decentralization components were in buildings and equipment, which have been or will be taken over by the relevant sections of the public administration. Where ownership for these is high, it is generally very likely that their integration will not pose any particular problems. Some specific investments will probably soon be at risk: one potential difficulty is that the continued availability of trained operators for some of the medical equipment is not guaranteed. Even more importantly, declarations of intent that particular attention will be paid to maintenance will need to be translated into practice, which for some of the more sophisticated equipment will require capacity-building and technical assistance. Maintenance has been a very weak point in the Comoros in the past and much equipment has been lost due to neglect

12 The most crucial factor affecting the progress towards the objective of assisting the country out of the secessionist crisis, however, will be political instability. Whereas there has been no renewed violence and all parties continue to declare their commitment to the reconciliation process, confrontations between political leaders over the control of resources and institutions continue. The achievements of this project in terms of confidence-building are by their nature fragile and could easily be undone for political calculations. As much as one would hope that the credit will have made a lasting contribution to the country s transition towards a sustainable institutional arrangement acceptable to all parties, at this stage, given the current political situation, this seems unlikely. 6.2 Transition arrangement to regular operations: The Emergency Economic Recovery Credit was designed as a one-off operation, and no follow-up of this type is planned. Buildings and equipment are being or will be managed by the relevant island administrations and ministries. As the Emergency Credit was essentially implemented by permanent staff of the planning commission, these will be able to continue monitoring the ongoing activities even after the closure of the project. 7. Bank and Borrower Performance Bank 7.1 Lending: Despite a very uncertain environment, the Bank task team showed responsiveness to clients need in rapidly formulating a relatively simple, quick-disbursing operation aimed at supporting the transition. The team pre-appraised the credit in March/April 2001, negotiations took place in July 2001 and the first installment was disbursed in September 2001, undoubtedly providing a contribution to the ongoing stabilization efforts. The team took relevant experience for this type of credit into account and included more control mechanisms than usual into the design. While recognizing that some quality was lost due to the time pressure for the preparation of the credit, Bank performance at the lending stage is rated satisfactory. 7.2 Supervision: The Bank maintained a close dialogue with the authorities in order to deal with difficulties of the transition process that affected the implementation. In fact, throughout the credit s lifetime, different aspects of the implementation became blocked and required interventions from Bank staff. As the Bank has no office in the Comoros, four supervision missions were undertaken for this purpose within less than a year. These missions always included Anjouan and on two occasions Moheli. Bank staff invested much time and energy in the project, building a good working relationship with borrower and the implementing partners. Their responses to the changing demands of the transition showed flexibility and were, with the exception of the approval of the procedures manual, timely. PSR ratings, based on achievement of the transition s milestones and on the financial management of the project, were carried out in January 2002, May 2002 and February 2003 and were reasonable, except that the May ratings for national reconciliation and risks now seem optimistic in view of the breakdown of the transition process in July The supervision is rated highly satisfactory. 7.3 Overall Bank performance: The emergency credit was implemented under the high case scenario of the Bank s Interim Support Strategy and all disbursements were made when the underlying assumption that the reconciliation was on track was justified. In the middle of 2002 the situation deteriorated rapidly, leaving the emergency credit without a formal strategic framework. Since then, the country team has been in the process of updating its strategy to reflect the difficulty of the current situation in the Comoros. The overall performance of the Bank is rated satisfactory. Borrower 7.4 Preparation: There was close co-operation between the Bank and the Borrower in designing the project and the relevant ministries and agencies were consulted. The government showed a high commitment to the - 9 -

13 project and took the lead in allocating the funds to its poverty reduction priorities. The major shortcoming was that the government s emergency recovery program did not make explicit how the activities to be financed would contribute to the overall objective and did not include any indicators to measure its success. The Borrower s performance in the credit s preparation is rated satisfactory. 7.5 Government implementation performance: The central government appointed its Planning Commission as the implementing agency. The central government and island governments generally gave good support to the projects funded by the credit but have been unable to complete the transition and stalled the reconciliation process between them. As mentioned above, several governments pressed for re-allocation of funds to items for which the emergency recovery program was not intended, particularly vehicles. Rating: unsatisfactory. 7.6 Implementing Agency: The set-up of having the Central Planning Commission implement projects on Grande Comore, Moheli and at the central level and the Anjouan Planning Commission implementing those on Anjouan worked well for the implementation, despite some communication problems between the two commissions. Projects were implemented effectively and efficiently, and much faster than any previous Bank projects. Despite the complexity of the project, the implementing partner was able to effectively absorb disbursements at a rate of US$400,000 a month. Both commissions were capable, highly committed and worked hard for the success of the credit. No formal reporting was instituted, but the commissions produced reports for visiting supervision missions, including expenditure tables. Two internal and the first of two external audits have concluded that financial management has generally followed the agreed terms and that checks and procedures of the project management were acceptable. However, a World Bank procurement expert who was part of the December 2002 mission has expressed serious reservations on the procurement of several items which are yet to be discussed. On the assumption that these questions will be resolved, the implementing partner s performance is rated satisfactory. 7.7 Overall Borrower performance: Bearing in mind the reservation expressed above, the overall borrower performance is rated satisfactory. 8. Lessons Learned 8.1 Lessons for the ICR process: 1. The standard ICR format, its timing and assumptions may not be best suited to an emergency project of this nature. It may be too soon to judge the outcome of this EERC since a peaceful transition may still emerge from the current crisis. Sustainability in particular takes on a different meaning in this context. Given the trade-offs made in this kind of project, orthodox assessments may not be useful or fair. 2. In the changeable context, the process of generating and drafting the ICR has been as valuable as the final report, bringing increased attention on the Bank s strategy and future interventions. 8.2 Lessons for Comoros: 1. In the Comoros there is still no shared understanding of what the political transition out of the crisis and national reconciliation mean in practice. Reaching such an understanding is an important pre-requisite for moving these processes forward, which is ultimately dependent on political will. 2. The EERC has shown that service delivery and infrastructure investments can be carried out effectively even in the difficult context of the Comoros, as long as the implementation and governance of the credit meet the highest level of standards. 3. The EERC emphasized the need for improved communications between development agents on the different islands of the country. It is also desirable for them to make greater efforts to resolve problems internally before they are brought to the attention of the donor

14 8.3 Lessons for the Bank: A credit of this kind involves a number of trade-offs. Above all it promises high gain but is entered at high risk. It implies managing tensions between speed and quality both at entry and in supervision. It means mobilizing development instruments in a highly political context over which the Bank has limited control. Credits do not operate in isolation. Particularly in volatile strategic contexts such as that of the Comoros, the relationship between the Bank s operations and the Bank s country strategy needs to be continuously monitored and evaluated. This operation shows the need for an instrument allowing rapid and flexible disbursement of funds, while maintaining some control over the way the funds are spent. OP 8.50 was used to cover both, but in order to do so, Bank staff had to make a number of compromises. Formally shifting the emphasis of control from import receipts to actual expenditure of counterpart funds would reduce unnecessary bureaucracy and allow project documents to deal more immediately with actual objectives and expected outcomes. More specifically, where a credit is linked to an emergency recovery program, it is important that it is made explicit how and to what extent the activities are expected to contribute to the overall objective and that suitable indicators and procedures are defined. The development objective, in particular, needs to be precise and its implications clear to all concerned. Even where not formally required, a logical framework matrix and an operational manual would be one way of ensuring that everybody shares the same understanding of the project, its objectives and what success would mean. A complex and flexible project of this type carried out in a volatile political environment requires regular supervision and attention. It therefore comes at a heavy price in variable costs to the Bank, particularly if managed from a distance. 9. Partner Comments (a) Borrower/implementing agency: 9.1 Introduction In 1997, the secession of the island of Anjouan and the ensuing separatist crisis plunged the Comoros into political turmoil. For nearly three years, the situation in Anjouan was difficult, characterized by repeated internal clashes among opposing separatist groups, and compounded by an economic embargo imposed by the international community. This situation seriously destabilized the island's fragile economy and was partly responsible for the deterioration of its inhabitants' already difficult living conditions. The calamitous consequences of that crisis impacted on the other two islands of the Islamic Federal Republic of the Comoros, Grande Comore and Moheli. Political efforts to end the secessionist crisis led to the signature of a Framework Agreement for National Reconciliation on February 17, 2001 by the Government, the Anjouanese authorities, the opposition and civil society. That agreement provided for the formation of a New Comorian Grouping comprising all the islands of the archipelago, decentralization and a transition period to enable the country to prepare for the establishment of the new institutions. But effective resolution of the crisis required measures in support of economic recovery, rehabilitation and the reconstruction of the social sectors. The Comorian Government accordingly drew up a Reconstruction and National Reconciliation Program (RNRP) comprising four main components: poverty reduction, national reconciliation, decentralization and good governance

15 That program, designed by the Planning Commission (formerly the General Directorate for Planning), was very ambitious. It addressed almost all the urgent needs identified by the islands and by the sectors of activity to effectively end the crisis and restore social, economic and political stability to the country. It was a short-term development program, estimated at approximately US$56 million. The RNRP was used to draw up the Government's Letter of Policy for National Reconciliation submitted to the "Friends of the Comoros" conference held in Paris in July 2001 under the aegis of the World Bank. Through that conference, the Government hoped to obtain special support from partners, friends of the Comoros, for its efforts to secure national reconciliation, establish new institutions and reduce poverty during the transition period. The emergency program for which funding was sought amounted to US$12.8 million, including $800,000 for the social and vocational rehabilitation of young militiamen in Anjouan that were obtained as a post-conflict donation from the World Bank. One of the commitments made by the "Friends of the Comoros" at the Paris conference concerned an Emergency Economic Recovery Credit (EERC) Agreement in the amount of US$6 million, which was signed by the World Bank and the Government in September The EERC credit was awarded to finance the import of raw materials and capital goods. Counterpart funds generated were to be used to finance, in particular, three of the four components of the Government's emergency program, namely poverty reduction, decentralization and national reconciliation. The EERC was overseen by a funds management body, the Emergency Credit Implementation Unit, which was set up to monitor disbursements and audit activities at the national and island levels. The unit was chaired by the Planning Commissioner, who was responsible for coordinating all EERC implementation activities, reporting to the Government and to development partners, signing all correspondence and authorizing expenditure. In addition to the Planning Commission, which was responsible for the Unit's coordination, EERC-financed sectors of activity and Anjouanese stakeholders were represented in the Unit. For the purposes of fund-raising and financial oversight, the Unit was enlarged to include representatives from the financial sector. Given the particular status of Anjouan at the start of the project, one internal organization was admitted to participate specifically on that island's behalf. An EERC implementation committee was set up to facilitate project execution on the island. That committee's membership was the same as that of the EERC Implementation Unit, but it had wider powers. The Chairman of the EERC Implementation Unit was assisted in project coordination tasks by two permanent management sub-units, one national, based at Moroni, and the other accountable to an internal organization in Anjouan. The sub-units basically consisted of two senior accountants, one in Moroni and the other in Anjouan, who worked closely with two technical coordinators acting as technical assistants to the Chairman of the Unit. Each EERC management sub-unit was responsible for daily funds management, keeping the accounts used to substantiate the utilization of funds, and providing auditors with all the documents needed to draw up the necessary audit reports. In view of the EERC's emergency character, special disbursement provisions, waiving the procedures usually applicable to World Bank funds, were authorized to ensure that operations were conducted efficiently. The purpose of this EERC Final Report is to evaluate credit implementation by analyzing achievements against the established objectives. The analysis will make it possible to evaluate the performance of national and donor management teams and of oversight teams, with particular emphasis on the lessons to be learned

16 This report is in two main parts: the first focuses on EERC objectives and achievements, while the second attempts to assess and review EERC implementation. 9.2 EERC objectives and achievements Restatement of EERC objectives The main EERC objective was to support the national reconciliation process initiated in the Comoros in 2001 while laying the foundations for economic recovery and a post-transition reconstruction program. The EERC was accordingly designed to finance three priority areas: poverty reduction, decentralization and national reconciliation. In the area of poverty reduction, the objective of the emergency credit was to improve the living conditions of the poorest inhabitants, namely the rural population, by taking high-profile action to address that population group's priorities. The activities targeted the health, education, water and rural development sectors. Funds were distributed among the three Comorian islands on the basis of the criteria of population size and poverty incidence. US$3 million, i.e., 50 percent of the EERC, were earmarked for the poverty reduction component, with 50 percent allocated to Anjouan, 40 percent to Grande Comore and 10 percent to Moheli. In the area of decentralization, the specific objective pursued was balanced investment in administration. Funds were allocated for public service activities and the rehabilitation, construction and equipment of administrative buildings, depending on the institutional needs identified by each island in conformity with the principles of administrative and economic decentralization practiced by the Comorian State. US$1.8 million, i.e., 30 percent of the EERC, were earmarked for the decentralization component, with 50 percent allocated to Anjouan, 33 percent to Moheli and 17 percent to Grande Comore. In the area of national reconciliation, the objectives were to provide financial support for the activities of the two bodies involved in the process, the Monitoring Committee and the Tripartite Commission; support the units responsible for drawing up the budget; ensure efficient and transparent EERC management; and provide institutional backstopping for the Planning Commission, which coordinates the Reconstruction and National Reconciliation Program for the three islands and activities financed from external resources mobilized by the Government. US$1.2 million, i.e., 20 percent of the EERC, were earmarked for the national reconciliation component, with 42 percent allocated to transition bodies, 33 percent to management, expenditure monitoring and budget support, and 25 percent to support for planning. Evaluation of activities carried out set against initial objectives (EERC impact) To attain the three specific emergency credit objectives restated above, activities were defined and carried out on the three islands and on behalf of the Union (see the annexed "Expenditure Table"). An analysis of the results obtained in each area will be used to measure the impact of achievements in relation to each objective pursued and the links between the various objectives. EERC poverty reduction results Various EERC impact assessments among recipients of funds have shown that the poverty reduction activities financed helped considerably to improve the living conditions of the target population groups by rehabilitating and rebuilding social sectors, financing revenue-generating activities (RGAs), and creating jobs in businesses. In the health sector, activities financed on all three islands helped to enhance access to health care services, improve their quality and lower the cost of basic medicines (see the report of the country s drug procurement agency, Pharmacie Nationale Autonome des Comores, PNAC). The improvement of the