Document of The World Bank FOR OFFICIAL USE ONLY Report No: 23373 IMPLEMENTATION COMPLETION REPORT (IDA-23790) ON A CREDIT IN THE AMOUNT OF XDR17.60 MILLION (US$24 MILLION EQUIVALENT) TO THE REPUBLIC OF MALAWI FOR A LOCAL GOVERNMENT DEVELOPMENT PROJECT DECEMBER 19, 2001 This document has a restricted distribution and may be used by recipients only in the performance of their official duties. Its contents may not otherwise be disclosed without World Bank authorization. CURRENCY EQUIVALENTS (Exchange Rate Effective 11/26/2001) Currency Unit = Malawi Kwacha (MWK) MWK I = US$ 0.02 US$ I = MWK 63.25 FISCAL YEAR July I June 30 ABBREVIATIONS AND ACRONYMS BCC Blantyre City Council BMLP Building Materials Loan Program CDP Council Developmeit Plans CSP Country Strategy Paper CTA Chief Technical Advisor DEVPOL - Statement pf Development Policies DFLA - Development Fund for Local Authorities DLV - Department of Lands and Valuation DTCP - Department of Town and Country Planning ECMAC - Entrepreneurship and Capital Market Adjustrnent Program ESAF - Enhanced Structural Adjustment Facility GDP - Gross Domestic Product GOM - Government of Malawi -1RID - Human Resource and Institutional Development Project ICB - Intemational Competitive Bidding IDA - Intemational Development Association LAs - Local Authorities LASC - Local Authorities Service Commission LCC - Lilongwe City Council LGDP - Local Government Development Project LGSC - Local Government Service Commission LCS - Local Competitive Bidding MHC - Malawi Housing Corporation MIM - Malawi Institute of Management MLG - Ministry of Local Government MOF - Ministry of Finance MOW - Ministry of Works NDF - Nordic Development Fund ODA - Overseas Development Administration OPC - Office of the President and Cabinet PPF - Project Preparation Facility PSIP - Public Sector Investment Program SCDP - Secondary Centers Development Program SRP - Special Rating Procedure STC - Staff Training College (Mpemba) TA - Technical Assistance THAs - Traditional Housing Areas USAID - United States Agency for International Developmnent WTP - Willingness to Pay ZMC - Zomba Municipal Council Vice President: Callisto E. Madavo Country Manager/Director: Darius Mans Sector Manager/Director: Jeffrey S. Racki Task Team Leader/Task Manager: Alan G. Carroll FOR OFFICIAL USE ONLY MALAWI LOCAL GOVT. CONTENTS Page No. 1. Project Data 1 2. Principal Performance Ratings 1 3. Assessment of Development Objective and Design, and of Quality at Entry 1 4. Achievement of Objective and Outputs 7 5. Major Factors Affecting Implementation and Outcome 19 6. Sustainability 23 7. Bank and Borrower Performance 27 8. Lessons Learned 30 9. Partner Comments 31 10. Additional Information 32 Annex 1. Key Performance Indicators/Log Frame Matrix 33 Annex 2. Project Costs and Financing 39 Annex 3. Economic Costs and Benefits 41 Annex 4. Bank Inputs 42 Annex 5. Ratings for Achievement of Objectives/Outputs of Components 46 Annex 6. Ratings of Bank and Borrower Performance 47 Annex 7. List of Supporting Documents 48 This document has a restricted distribution and may be used by recipients only in the performance of their official duties. Its contents may not otherwise be disclosed without World Bank authorization. Project ID: P001636 Project Name: LOCAL GOVT. Team Leader: Alan G. Carroll TL Unit: AFTUI ICR Type: Core ICR Report Date: December 19, 2001 1. Project Data Name: LOCAL GOVT. L/C/TFNumber: IDA-23790 Country/Department: MALAWI Region: Africa Regional Office Sector/subsector: UM -Urban Management KEY DATES Original Revised/Actual PCD: 08/15/1989 Effective: 10/01/1992 12/22/1992 Appraisal: 05/14/1991 MTR: 05/15/1996 Approval: 06/04/1992 Closing: 12/31/1999 06/30/2001 Borrower/lmplementing Agency: GOVERNMENT OF MALAWI/MIN. LOC. GOVT. Other Partners: STAFF Current At Appraisal Vice President: Callisto E. Madavo Edward Jaycox Country Manager: Darius Mans Stephen Denning Sector Manager: Jeffrey Racki Isaac Sam Team Leader at ICR: Alan G. Carroll Anders G. Zeijlon ICR Primary.4uthor: Colleen Butcher 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: U Sustainability: UN Institutional Development Impact: M Bank Performance: U Borrower Performance: U QAG (if available) ICR Quality at Entry: U Project at Risk at Any Time: Yes Yes, rated at risk in F1'1998 as achiev-emenit of DO was rated unsatisfactory (PSR Seq. Nos. 17 and 18). 3. Assessment of Development Objective and Design, and of Quality at Entry 3.1 Original Objective: The objectives of the project as stated in the Staff Appraisal Report (SAR) were stated as follows: "The main objective is to put in place a combination of related actions that will collectively bring about a stronger, more effective system of local government in Malawi. These are building blocks for enhancing the financial, managerial, and technical capacity of the local government system to (a) provide and maintain essential municipal services and infrastructure at a level consistent with affordability and with the size and growth rate of the population; and (b) finance those endeavors mainly with locally generatedfunds, while improving thefinancial management of such funds. Particular emphasis is on strengthening local authorities' ability to provide serviced land, sanitation, road maintenance and certain commercial services, i.e., markets, for the growing urban population. " The SAR recognized that the building of "an effective operating local government system is a long-term (10 - 15 year) process" (SAR, p.16). The project identification process (which brought together the MLG, local authorities, and representatives from other relevant GOM ministries) resulted in a consensus that the project should contain three major components: an overhaul of the local government system; institutional strengthening of the MLG and the largest local authorities; and provision of inputs to ensure the operations and maintenance of existing infrastructure investments. The project objectives respond to these priorities in a clear way. The Bank's assistance strategy in Malawi in the early 1990s was to pursue with the government "the process of further liberalization and opening of the economy in order to boost economic growth and alleviate poverty" (SAR, p. 1). The objective of creating strong, effective local governments would contribute to the essential conditions for making towns and cities more productive, for extending services to the lower income groups, and providing tangible public health benefits through improved environmental sanitation. The project objectives also complemented the activities of a number of donors which were "supporting Government efforts by providing technical assistance, training and infrastructure in pilot towns" (SAR, p. 1 7). By today's standards of practice in the Bank, the project's objectives were not formulated clearly enough and were too ambitious. In particular, the objectives should have reflected a more realistic vision of what one project could accomplish, given the inherently risky nature of governance reforms, the severely limited implementation capacity of both the central and local governments, and the long time horizon needed for local governments to achieve the capacity to provide and maintain essential services. 3.2 Revised Objective: The objective was not revised during implementation 3.3 Original Components: The following are the components description summaries as stated in the SAR: (a) Policy Reform Component (US$2.787 m.): "Modifications to the legal and institutional framework for the local Government system will be carried out through a series of policy studies, reform advocacy workshops, and the implementation of associated action plans. This will be organized in four areas (i) legislation review, (ii) duties and responsibilities of central and local - 2 - government agencies, (iii) revenue mobilization and financial management, and (iv) human resources." (b) Institutional Strengthening (US$9.672 ni.): "(i) modifications in MLG's organization and procedures supported by the provision of vehicles, equipment and other incremental operating costs related to the project for MLG and selected councils; (ii) training for 'MLG and councils staff at local training institutions; (iii) specialized assistance to help reorganize the MLG and implement a phased decentralization of certain MLG functions; (iv) establishment of improved intemal operating procedures; and (v) building technical capacity within MLG divisions." (c) Provision of Infrastructure and Services (US$17.168 m): "(i) plant, vehicles, equipment and services required to rehabilitate and improve maintenance of existing infrastructure and other assets in nine local councils, and (ii) new infrastructure investments in Lilongwe, Blantyre and Zomba. Funds for infrastructure investments, vehicles and equipment will be on-lent to the participating local authorities and the repayments will be re-lent to other local authorities able to carry out essential infrastructure and services through a Development Fund for Local Authorities (DFLA)." The design of the three components was clearly linked to the objectives described in 3.1 above. The credit provided financial support for local governments to respond to immediate maintenance needs (through the procurement of municipal vehicles and equipment) as well as to prepare infrastructure investment plans (Council Development Plans), all within the larger context of organizational restructuring, training of key staff, establishing improved financial management systems within the local governments, and carrying out legislative and policy reviews and amendments. The SAR elaborated a timetable for adoption of the recommendations of the studies and workshops during implementation (SAR, Annex 2). The SAR (p.16) specifically noted that the project design took into account the following lessons derived from the Infrastructure Project (7551 - MAI) and the First Urban Project (5200 - MAI): * The need for simplified implementation structures. * Enhanced training, technical assistance, and studies are needed to address weakening implementation capacity in Malawi, especially for cost recovery, institutional development, and maintenance. A major omission from the project design was an explicit monitoring mechanism for assessing the impact of the policy and institutional strengthening components and the perfornance of the local authorities. I The participating local authorities were the City Assemblies of Blantyre, Lilongwe and Mzuzu; the Municipality of Zomba; the Town Assemblies of Salima and Luchenza; and the District Assemblies of M'Mbelwa, Ntcheu and Nsanje. -3- In light of the assessment of objectives in 3.1 above, the project could have been designed more modestly, to initially address the gaps in the policy framework, especially the lack of an intergovernmental system of fiscal and human resource policies. This could have been done, for example, by focusing the project as a pilot and/or a technical assistance operation precedent to further stages of support. 3.4 Revised Components: A number of amendments were made to the Credit Agreement at various times to try to improve the implementation of the project. The amendments did not, overall, address the fundamental problems that emerged during implementation (see Section 5 below). Following discussions between GOM and the Bank, Section 3.01(c) of the Credit Agreement was amended on March 11, 1994 to incorporate an element of grant funding in the on-lending arrangements to the local authorities made through the DFLA. Increased grant proportions were approved to individual local authorities based on the proportion of non-revenue-generating activities to revenue-generating ones. Approval of the grants were to take into account: (i) the LAs' financial position and projections, in particular their level of existing indebtedness and ability to service new debt; and (ii) the ratio of the activities to be carried out with high social content and poverty alleviation potential compared to those which more readily could earn a commercial return, i.e. to provide support to sub-projects that might have a high social need, but difficult to justify on a commercial basis. The amendment allowed for 90% of the amount passed on to the smaller Town and District Assemblies to be grants and 35% to 50% in the case of the larger Urban Assemblies. On July 8, 1994, Schedule 2, Part C(2) of the Credit was further amended to incorporate new sub-projects for Blantyre City Assembly. On June 26, 1995, Schedule 5, para. I(c) was amended to increase the authorized allocation in the Special Account from $750,000 to $1,250,000. The on-lending arrangements (under Category 1, Part C) were once again amended on August 2, 1996, increasing the grant proportion of allocated funds for the four larger participating local authorities (Lilongwe, Blantyre, Mzuzu and Zomba) to 80 percent. The amendment was introduced so as to improve the cash flow situations of the local authorities (from the savings in interest that would be realized), whilst at the same time continuing to seek an element of cost recovery from the local authorities and provide an incentive to them to prioritize investrnents in sub-projects which could show a positive rate of return. On October 22, 1996, following agreements reached with the Government during the Mid-Term Review, Schedule I of the Credit Agreement was amended to reallocate funds between categories so as to accommodate new works (principally in Blantyre) and increase the allocation for studies and consultants' services, to provide greater support to the policy reform component (which had been making slow progress). The amount for Incremental and Operating Expenditures (the PCU operating costs) was also increased through raising the IDA financing share from 35 percent to 70 percent. -4 - In March 2000 the cost of the Infrastructure and Services component was reduced to $13.3 m., and the expected project costs reduced accordingly from the previous $36.3 m. to $28.0 m. The original credit amount of SDR17.6m remained unchanged. (PSR Seq. No. 22). The reduced component amount was based on a revised disbursement forecast for GOM to complete the construction of ongoing and agreed upon new sub-projects and the agreed technical assistance to support the government's decentralization policy. In July 1999, the DFLA Board resolved to capitalize the accrued and still unpaid interest on local authorities' outstanding loans, add this to the outstanding principal, and refinance the total outstanding debt at a new fixed interest rate of 15 percent (well below the Reserve Bank rate at the time of 46 percent). The World Bank judged that it was not worthwhile to amend the Credit Agreement (Section 3.01 (c)) once again to reflect the new terms and conditions adopted by the DFLA Board, but on April 19, 2000, the credit was amended simply to make future allocations to the Assemblies under Part C as 100% grants (up to $1.0m), effectively bypassing the DFLA altogether. At the same time, the credit closing date was extended from December 31, 1999 to June 30, 2001 so that outstanding sub-projects could be completed (most works sub-projects had contract completion dates of six months). Due to the slow implementation and the cancellation of all but one of the remaining site and service sub-projects under the project (because of poor cost recovery efforts on completed schemes), an amount of SDR 500,000 was also cancelled from the credit. 3.5 Quality at Entry: The project's Quality at Entry is assessed on the basis of (a) the consistency of objectives with the CAS and government priorities, (b) the Bank's safeguard policies, (c) the quality of the design, and (d) the reasonableness of assumptions regarding relevant external risks. The results are mixed, but on balance the quality at entry is rated Unsatisfactory. The project's objectives were consistent with the CAS and government priorities. The SAR (pp. 4-5) set the context within which the project objectives were prepared and stressed that the reforms envisaged for the local government sector were part of a lengthy and complex process. The need for strengthened local governments was clearly articulated with regard to reducing central fiscal demands as well as building long-term sustainability in the provision and maintenance of infrastructure through accountable and efficient local governments. Furthermore, the objectives fitted in well with the government's development policy which foresaw the local government system benefiting from a careful decentralization of both responsibilities and authority. The project supported the Bank's strategy for Malawi in (i) addressing human resource development and institutional capacity building through a number of categories under the LGDP Credit; (ii) developing strong coordination with other donors, with the LGDP explicitly recognizing the role of bilateral grant support in complementary policy reforms and provision of local government technical advisors; (iii) acting as a key catalyst to mobilize technical and financial assistance to support improved public sector management performance at the sub-national level; and (iv) targeting some of the project's infrastructure investments to lower income groups (sites and services, environmental sanitation). - 5 - The environmental impact of the project was assessed during appraisal and rated as "C", i.e. that the project would lead to significant improvement and tangible benefits to the environment in the Assembly areas through the provision of potable water, additional latrines, storm water drainage, improved solid waste management, improvements to and around existing markets, and road maintenance. There was no involuntary resettlement envisaged under the project (although subsequently during implementation, 13 households and 16 crop plots had to be compensated and relocated within the Chilimoni site and service area). No significant environmental or social issues arose during project implementation. The SAR identified the following risks: (i) the central government's willingness to implement the action plans to be produced under the policy component in a timely manner; (ii) the ability of the institutionally weak Ministry of Local Government and local authorities to implement such actions; and (iii) the pace at which the entrenched control behavior of the central government could be modified to create an environment of self reliance and independent decision-making in the local authorities. The SAR considered that these could be mitigated by orienting a significant part of the Project towards technical assistance and training and by carrying out annual reviews in which progress in the project's policy and institution strengthening component would be a prerequisite for continued implementation of the physical investment components. The actual implementation of the project revealed four notable project design flaws against which the risk mitigation measures outlined in the SAR proved to be inadequate. (a) The implementation capacities of the central and local governments were seriously overestimated. (b) The project design failed to include a system to monitor project impact and local government performance. As a result, neither the government nor the Bank were provided with information to make timely corrections to the project design or implementation arrangements. (c) The robustness of the DFLA mechanism in the face of potential macro-economic risks and the weak revenue base of the borrowing local authorities was not rigorously appraised. Flaws in the structural design of DFLA led to a lack of clarity and commitment on the part of GOM regarding the operating principles that should have been enforced to ensure that DFLA would be managed in a technically sound and financially self-sustainable manner. (d) The project design overestimated the government's commitment to local government development. It also overestimated the ability of technical assistance, training, and periodic reviews to influence and strengthen the policy regime and to improve the financial capacities of the local authorities. In the end, the interventions built into the project were not sufficient to overcome the larger political, macro-economic, and institutional forces that held back local government reform. It is notable that in November 1994, only two years after credit effectiveness, the elected local govermnent councils were dissolved. It took until November 2000 for local elections to be held. As a result, the local authorities remained relatively unempowered and unaccountable to their constituents during most of the project's lifespan. This contributed to - 6 - various implementation problems mentioned in Section 5 below. 4. Achievement of Objective and Outputs 4.1 Outcome/achievement of objective: If the project achieved anything, it did -- albeit belatedly -- help to lay a "foundation of related actions" which should lead to the development, over the long term, of an effective local government system in Malawi. The project influenced the adoption of reforms from 1998 onwards in two ways: * The project's Policy Reform Component produced a series of influential studies which fostered policy dialogue within the government and with the donor community on decentralization. * The project provided a platform for the World Bank to join with other donors, notably the UNDP, UNCDF, DANIDA, and DFID, to actively promote the local government reform agenda over a crucial period of years. This was instrumental in the realization of the actions listed below. The significant steps taken during the last years of the project which have built a foundation for the further development of the local government system are: * The adoption of a Decentralization Policy by Cabinet in 1998. * The passage of an amended Local Government Act (covering the responsibilities of both Urban and District Assemblies) in 1998. * The democratic election of local Assemblies in all 39 of the country's local government jurisdictions in November 2000 (the first local government elections to have been held since 1994). * The establishment in 2001 of a Local Government Finance Committee (LGFC) with responsibility for overseeing, approving and reporting to Parliament on the annual financial budgets of the local Assemblies and of recommending to Cabinet a formula for the annual transfer of intergovernmental fiscal transfers (established under Sec. 149 of the constitution). * The approval by Parliament of the transfer of significant resources (five percent of national revenues) from central to local governnents commencing in FY2001/02; * The adoption by all local Assemblies of a standardized Financial Management and Accounting Handbook. * The completion in 2000 of a Functional Review of the DLG and the initiatiation of a thorough restructuring of the Department to achieve its core objectives, including the recruitment of new, more highly skilled staff. * The appointment in 2001 of a highly capable and experienced Permanent Secretary for the Department of Local Government. -7- The project's secondary objective was to "enhance the financial, managerial and technical capacity of the local government system to: (a) provide and maintain essential municipal services at a level consistent with affordability and the size and growth rate of the population; and (b) finance those endeavors mainly with locally generated funds, while improving the financial management of such funds." As the subsequent sections of this ICR indicate, this secondary objective was not achieved. It was envisioned that policy reforms could be put in place within the first two to three years of the project, and that this framework could then facilitate real results at the local level within the project's life span. This vision proved to be too ambitious; actual experience showed that the time frame for improving local government capacity is far longer than expected. From its early years the operating environment for the project was extremely constrained, as explained in Section 5 below. By today's operational standards, the kinds of objectives sought by the project would be approached in a long-term context. The Bank's current practices would call for such a project to be a pilot or the first stage of a long-term program of support, perhaps through an APL. The objectives of the first stage operation would be stated much more modestly, with "triggers" - in terms of specific policy reforms or other actions - set as conditions for moving from one phase to another. Also, current good practice would dictate that the capital investment component be designed in a programmatic way, based on eligibility and performance criteria to provide incentives for capacity building of the local councils. It might be argued that the project was a necessary - though not sufficient - factor in the realization of the local government reforms to date and thus achieved its main objective of setting in place a foundation of policies and actions to facilitate local government development over a long-term period after the project's conclusion. However, the reforms so far, while significant, do not consititute a sufficiently robust and sustainable foundation to justify this conclusion. Political and economic uncertainties may yet undermine or stave off the crucial follow-up actions needed to implement and uphold the reform agenda. This, together with the unsatisfactory results of the institutional strengthening component and the DFLA, lead to the project's overall outcome being rated Unsatisfactory. * The project's Policy Reform Component produced a series of influential studies which fostered policy dialogue within the government and with the donor community on decentralization. * The project provided a platform for the World Bank to join with other donors, notably the UNDP, UNCDF, DANIDA, and DFID, to actively promote the local government reform agenda over a crucial period of years. The significant steps taken during the last years of the project which have culminated in a sustainable foundation for the local government system are: - 8- The project's secondary objective was to "enhance the financial, managerial and technical capacity of the local government system to: (a) provide and maintain essential municipal services at a level consistent with affordability and the size and growth rate of the population; and (b) finance those endeavors mainly with locally generated funds, while improving the financial management of such funds." As the subsequent sections of this ICR indicate, this secondary objective was not achieved. It was envisioned that policy reforms could be put in place within the first two to three years of the project, and that this framework could then facilitate real results at the local level within the project's life span. This vision proved to be too ambitious; actual experience showed that the time horizon for improving local government capacity is far longer than expected. From its early years the operating environment for the project was extremely constrained, as explained in Section 5 below. By today's operational standards, the kinds of objectives sought by the project would be approached in a long-term context. The Bank's current practices would call for such a project to be a pilot or the first stage of a long-term program of support, perhaps through an APL. The objectives of the first stage operation would be stated much more modestly, with "triggers" - in terms of specific policy reforms or other actions - set as conditions for moving from one phase to another. Also, current good practice would dictate that the capital investment component be designed in a programmatic way, based on eligibility and performance criteria to provide incentives for capacity building of the local councils. It might be argued that the project was a necessary -- though not sufficient -- factor in the realization of the local government reforms to date and thus achieved its main objective of putting in place a foundation of policies and actions to facilitate local government development over a long-term period after the project's conclusion. However, the reforms so far, while significant, do not constitute a sufficiently robust and sustainable foundation to justify this conclusion. Political and economic uncertainties may yet undermine or stave off the crucial follow-up actions needed to implement and uphold the reform agenda'. This, together with the unsatisfactory results of the institutional strengthening component and the DFLA, lead to the project's overall outcome being rated Unsatisfactory. Including (i) the establishment of a predictable, transparent, and adequately funded system for inter-governmental transfers, (ii) the election of Assembly Chief Executives, (iii) the formal transfer of basic competencies from selected sector ministries to Assemblies, and (iv) the implementation of effective capacity building for local officials and staff. -9- 4.2 Outputs by components: (a) Policy Reform Component ($3.193 m. SAR, $7.034 m. DCA, $5.772 m. actual) Recognizing that GOM's ability to improve how local governments operate was constrained by an inadequate policy framework, this component set out to assist GOM in formulating and analyzing alternatives, removing bottlenecks, facilitating discussion, and implementing appropriate actions. This would be achieved by carrying out 14 policy studies' (focusing on a legislative review, the division of responsibilities and functions between the two tiers of government, financial resources management, and human resources management) and holding associated high-level workshops to review and adopt the recommendations of the studies (with associated outputs such as redrafted legislation, procedural handbooks and similar). A team of MLG staff and four technical advisors would monitor progress in adoption of policy reforms. The Mid-Term Review (May 1996) reported that: * Two out of the 14 studies had been completed and recommendations were in use (the Local Authority Budget Systems Manual and the Local Authority Accounting Systems Manual). A further 10 of the 14 studies had been completed (with the 1th recently contracted) but the associated recommendations had not been adopted. Reasons for the lack of follow up on the studies and recommendations cited in the supervision missions and MTR mission included. * Only three of the four technical/resident advisors had been recruited (the financial controller, systems maintenance engineer and housing advisor) and the unit in the MLGRD assigned for monitoring implementation of the policy reforms clearly lacked both the technical capacity and political support to take forward the recommendations in a proactive way; * Some of the reports (for example, the redrafted Local Government Act) had been overtaken by events on the ground and would need further revision; * Many of the policy reform recommendations were put on hold due to the absence of an overarching Government policy on decentralization (what model to adopt, timetable for implementation, etc.). A general policy on decentralization was adopted in 1996 and a Cabinet Committee on Decentralization established, but it was only in 1998 that the new Local Governnent Act was passed and in 1999 that a detailed Decentralization Implementation Plan adopted by GOM. Following the Bank's July 1999 mission, which performed the first serious review of the DFLA since the mid-term review in 1996, it was agreed at the Bank's urging that a detailed assessment study of the DFLA should be carried out. This was eventually carried out by a Bank mission in May 2001, shortly before project closing, as the government had been unable to contract a suitable consultant to complete the study before the project Including (i) the establishment of a predictable, transparent, and adequately funded system for inter-govetorment transfers, (ii) the election of Assembly Chief Executives, (iii) the formal transfer of basic competencies from selected sector ministries to Assemblies, and (iv) the implementation of effective capacity building for local officials and staff. - 10 - closed. The study confirmed the extent of financial difficulties facing the DFLA, analyzed the roots of problems, and provided GOM with alternative options for closing or managing down the Fund. Overall, the studies were completed in a timely manner and were of a high standard. Until December 1995, the project supervision missions comrnented that the Government remained committed to implementing policy reforms under the project. However, the crucial adoption of the recommendations of the studies and the political commitment to implement them were lacking throughout most of the life of the project. From December 1995 the policy reforn component was reported as making slow progress and was eventually downgraded to Unsatisfactory by mid- 1997. Political commitment to decentralization and the recognition of the need for strong and accountable local governments was given a new impetus in the final years of the project. In mid-1999, the component was upgraded to being Satisfactory as a result of the approval of a new Local Government Act and the Cabinet approval of a new Decentralization Policy. Local Government elections were held in November 2000, a Local Government Finance Committee was established, and in FY200l/02 the MOF approved the transfer of 5 percent of national revenues to local governments for recurrent and development expenditures. While the recent advances in local government policy reforms in Malawi resulted from a number of influences, it is clear that the LGDP deserves credit for facilitating the reforms through the studies it sponsored. These opened up crucial policy discussions over a lengthy period of time (nine years) on key reform areas and ultimately contributed to the original objective of assisting GOM to put in place the key foundations of a stronger, more effective local government system. For this reason, the component is judged satisfactory. (b) Institutional Strengthening Component ($11.666 m. SAR, $2.543 m. DCA, $2.197 m. actual) At the time of appraisal, GOM recognized the need to strengthen the capacity of local governments as a key element in the development process and as one of the important ingredients of macroeconomic adjustment and stabilization in Malawi. This component was therefore seen as essential. It had three principal elements: reinforcing the MLG through changes in the staffing and organization (at both the center and in the regional offices); institutional strengthening of the four main urban councils and a package of management and skills enhancement assistance to other local authorities; and developing a national training program for local authority elected officials and staff. The SAR detailed that this would be achieved by: * Preparing a five-year action plan for MLG reorganization involving a new mission statement, supportive equipment and operating expenditures and technical assistance. * In the case of the four main urban councils, providing incremental staff and running costs for administering the Traditional Housing Areas (THAs); refurbishing vehicle fleets; studies for THA upgrading; technical assistance for development of maintenance systems; project development; property and housing management; and finance and accounting strengthening (making use of manuals and systems developed under the policy reform component); * In-service training, entry level training, and academic training for central and local government staff; All of these were specified in a detailed work program (SAR, Annex 3). The LGDP delivered a significant amount of training in the first six years of the project, during which time 638 individuals had undergone either short-term or long-term training in a wide range of topics including upgrading of low income settlements, financial management, procurement, and planning and urban management. There was a hiatus in training activities in 1999, and then in 2000/01 it picked up again, in particular with the induction training of all newly elected local government councilors after the November 2000 elections. Long term technical advisors provided useful inputs, in particular to the four cities in the administration of housing areas (THAs) and in systems maintenance engineering. The technical assistance component for a housing estate development adviser was implemented between August 1994 and December 1999. The adviser worked with Blantyre, Lilongwe and Mzuzu council assembly in the management and administration of the site and services/upgrading schemes and carried out on the job-training for cost recovery calculations based on the projects implemented. Both Blantyre and Lilongwe city assemblies have gained practical experience and knowledge on development of project proposals, procurement, document preparation and procedures, contract management and cost recovery and user charges calculations. Mzuzu city assembly to a limited extent also acquired some experience and knowledge on development of project proposals and project management. The systems maintenance engineering was successfully completed with training assistance provided to the staff of Blantyre, Lilongwe, Zomba and Mzuzu assemblies focusing on roads, buildings, vehicles and equipment. The assemblies were also assisted to develop a maintenance policy. Furthermore, the project financed the procurement of a range of municipal equipment (88% disbursement percentage under this category in the Credit): 76 vehicles (station wagons, pickup trucks, trailers, 7 ton truck), plant (dump truck, water tanker, 3 skip handlers, 90 skips, pit emptier) and equipment (computers, photocopiers) and motorcycles for distribution to the Assemblies, Ministry of Local Government and the Mpemba Staff Development Institute. Also procured were the spares for most of the vehicles. In addition, 20 vehicles, plant and equipment were procured under works and services contracts. The functional review of the MLG was slow to be undertaken and was finally only completed in 1999 with grant funding from another donor. Full implementation of the Review, including the re-staffmg of the Department of Local Government, was still being effected at the time of the ICR and so the full impact of the Review is yet to be evaluated. 12 - Supervision missions consistently rated this component as performing satisfactorily up until December l999', noting that technical expertise (in planning, procurement, engineering design and construction supervision) at the local level was improving and that local governments were gaining valuable lessons from the implementation of the infrastructure component. However, despite the significant financial inputs provided under the project ($2.2m), capacity within both tiers of Government remained weak at the close of the project. At the central government level, the DLG continued to rely heavily on outside units such as the Decentralization Secretariat to monitor the activities of the local authorities and to coordinate policy reform activities, and on the PCU for financial control of the project funds (a recently completed external audit of the Project accounts cited extensive qualifications by the auditors on the financial records kept by the DLG itself). Similarly, the capacity within the Local Assemblies, with the possible exceptions of Blantyre CA and Lilongwe CA, remained weak across the board (poor revenue collection and financial management, late or non-submission of annual audited accounts, consistent defaulting on repayments to the DFLA, costly overruns on works contracts on the sub-projects in most of the centers, poor cost recovery procedures on site and service schemes, to name some key examples). While most of the tangible outputs of this component were delivered (advisers, training, studies, equipment), these have not been transformed, in the main, into tangible results in terms of institutional strengthening. For this reason the component is rated as Unsatisfactory. (c) Infrastructure and Equipment Component ($21.394 m. SAR, $13.399 m. DCA, $11.08 m. actual) Implementation of the infrastructure and services was achieved slowly but in general satisfactorily under the project. The financing mechanism (DFLA) set up by the project was a failure. By the close of the project, 80 percent of the funds in the works categories under the credit were disbursed. Two works contracts were not completed largely because of poor performance by the contractors: Blantyre Ndirande market extension (90% completed) and Lilongwe Works Yard Improvement Phase 2 (60% completed). One other works contract, the Ntcheu Kampepuza market, was not fully completed (98 percent) due to the inability of the District Assembly to meet its counterpart payment toward the contractor s certificates. At the end of March 2001 the Government, in consultation with the Bank, cancelled one large works project before it was awarded - the upgrading of Phwetekere sites and service scheme Phase B - because a number of households had settled informally on the site, requiring a resettlement plan and because there was insufficient time to complete the scheme before the project closing date. 4 With one exception, in September 1998, but in April 1999, the rating improved once again to satisfactory) - 13- The following details the infrastructure provided under the project in each of the participating Assemblies: Lilongwe City Assembly: The first phase of Phwetekere sites and service/upgrading scheme was completed, with 521 plots were demarcated and allocated to members of the public. The paving of the Malangalanga Road and the improvement of the link between the MI and Malangalanga Road and market was implemented and has generated considerable additional traffic and economic activity in and around the market. The improvement of the Phase I of the Work Depot was completed and the depot has since been put to use. Construction of Phase 2 of the Works Depot was only 60 percent completed at credit closing. The Assembly is in the process of reviewing the scope of work with a view to reducing it to the level they can fund. The equipment delivered for Solid Waste Management assisted in the cleanliness of the city. Between 50-60% of the solid waste generated within Lilongwe city is moved and disposed at the dump site. Evacuation of septic tank effluent was improved through the provision of septic tank emptiers. Though landfill equipment was procured, it was not fully utilized as the landfill component of the project did not go beyond the detailed design. The vehicles delivered assisted the city assembly to step up its rate collection and movement of officials. Blantyre City Assembly: Except for the Ndirande Market Extension, which was 90 percent completed at project closing, all physical works including procurement of vehicles and equipment were satisfactorily completed. The Ndirande market comprises various types of sheds with a total of about 920 stalls, an ablution block, drainage, the boundary wall and paving of the market area. The uncompleted works are the ablution block and the paving of the market area. The City Assembly reported that it plans to use its own funds to complete these elements. The Nkolokoti and Chilomoni THA upgrading; and Ndirande informal settlement upgrading involved upgrading of a range of essential infrastructure and services in these poor neighbourhoods. About 3.5km of roads and drains were provided and provision was made for sanitation improvement. To reduce the incidence of night-time accidents along Zalewa Road (which has a high level of pedestrian traffic) 8km street lighting was constructed. The road is now well lit at night, enhancing the security within the community and also assisting business activities in the area. The Nkolokoti Parkway Upgrading involved the upgrading of 3.1 km of road. The completion of the road has assisted traffic circulation and has removed congestion which previously characterized this section of the road. Limbe Market Improvement and Extension involved the construction of a new food hall, refurbishment of the existing market, provision of a forecourt (for delivery of produce) provision of two public toilets consisting of 12 rooms, refurbishment of the refuse bunker, construction of boundary wall, Market Master's office and electrical works. The market has since become operational. Although the occupancy rate of the market is high, revenue collection is low. The completion of the Limbe market did not affect street vendoring, as many small vendors consider the market unaffordable and therefore still prefer to sell their wares on the streets. The subsidiary market for Chilomoni involved the construction of 12 open stalls, paved market - 14 - area, refurbishment of existing shed and construction of new shed for sale of vegetables. Though the market was completed, the rate of occupancy was low (approximately 30%.) This was attributed to smaller informal markets springing up in the vicinity of Chilomoni. Limbe Bus Depot: The Limbe bus depot involved the construction of access road, provision of a service block, bus island and canopy, gate houses and erection of boundary wall. Despite the successful completion of the bus depot, it was not in use at project closing because Blantyre City Assembly had not yet worked out an acceptable mode oi operating it. Discussions with bus operators and other stakeholders are continuing. Chilomoni South Site and Service Scheme: The scheme's main objective of providing affordable decent land to low- to middle-income people so that they can build and own permanent houses, thereby improving their standard of living was achieved. A total of 1.6km of surfaced road; 5.6km of gravel road, 9.8km of storm water drains, road furniture, landscaping electricity and water supply were provided. Also 542 plots comprising of 483 residential plots, 50 commercial plots and 9 institutional plots were demarcated. By August 2001, 501 plots have been allocated and some of them fully developed. Ndirande Market Extension: The main objective of embarking on the extension of Ndirande Market is to reduce congestion at the existing market and create room for vendors who hitherto trade on the street. The project was only 90% completed at project closing. Mzuzu City Assembly: The Luwinga site and services and upgrading project which envisaged the provision of serviced plots for low income housing was satisfactorily completed. 8.4 km. of road of varying width, 0.8km of lined drain and 1.1 km. of unlined drain, sanitation facilities were provided. About 662 plots were demarcated. However, no plots had been allocated as of project closing. Advertisement for the sale of plots has been made. The Assembly expects that the Luwinga site and service will enhance the revenue accruing to it through the sale of plots and payment of ground rents. Zomba Municipal Assembly: The bus depot which involved the construction of access road, service block, bus Island canopy and toilets was successfully completed and has since been put to use. Work on the upgrading of Chinanwali and Ngongonwa Markets were substantially completed, however, at loan closing there were minor outstanding works. Despite these minor outstanding works, the two markets are being used. District Assemblies: (i) The refurbishment of M'mbelwa Resthouse at Mzimba Boma was only 80% complete at loan closing. Due to the non-completion of the resthouse, the District Assembly is unable to put it to its intended use. (ii) The improvements to Mphate market in Ntcheu District Assembly entailed building of - 15- stalls, coldroom, electrification, toilets and perimeter fence. The improvement works were substantially completed with only the electrical works outstanding. The improvement work on the Kampepuza market which involved erection of sheds, Market Master' s office, toilets, slaughter house and perimeter fence was substantially completed (98%). The markets are however, in use. (iii) The additional work to Nsanje market involved provision of lock-up shops, selling sheds, toilets, canteen and construction of perimeter fence. The work was successfully completed. Role of the Development Fund for Local Authorities (DFLA) The Credit Agreement provided that GOM should on-lend funds to local authorities to finance infrastructure and services. The Project SAR explained that "Funds for infrastructure, vehicles and equipment component of the project will be on-lent to local councils through a Development Fund for Local Authorities (DFLA). DFLA will be created through a government trust fund and will receive the repayments of all on-lending" (SAR, p.25). The underlying idea was that repayments accruing to the Trust would form a revolving fund pool which could provide the source of funds for lending to local authorities on an ongoing basis into the future. The DFLA's primary capital assets were derived from the loans provided by GOM to the nine participating Assemblies with funds sourced from IDA. The initial amount totaled MK13m (US$3.3m); by June 2001, MK137.7m (US$4.9m) had accrued to DFLA on this basis. The Fund Administrators referred to loans under the LGDP as "indirect loans" to the Assemblies; loans filmded from other income of the Fund (principally Treasury bills) were called "direct loans". The life of the Fund can be divided into a number of different phases. From 1993 - 1996, it was managed in-house by the DLGA and loan disbursement only occurred in "indirect" form. In 1996, the Fund administration was outsourced, "indirect" loan disbursement continued and "direct" loans were allocated to a number of Assemblies. Original loan terms were charged during this period. In 1999, with the Reserve Bank of Malawi rates at a high 47 percent, the Board decided to capitalize all accrued and unpaid interest amounts, amortizing these over the remaining lives of the loans, and reduce interest rates to a fixed 15 percent. By mid-2000, the DFLA did not have the fumds to provide any loans to Assemblies and all new lending activity had ceased. In the words of the Fund Administrators, it had become "dormant and illiquid". The basic factor behind this situation was the chronic failure of local authorities to service their debt obligations, almost from the inception of the project. Repayments were extremely and consistently poor - averaging well below 25 percent of amounts due annually. In consequence, there was a growing cumulative arrear amount. In addition, the Fund did not write off bad debts and due to the combined effects of high inflation and local currency depreciation, the real value of the Fund's original capital base was considerably eroded. At historic prices, the original value of the indirect loans amounted to US$4.9 m. By the end of the Project, the book value of the portfolio had deteriorated to about $2.1 m., the cost of this depreciation being borne by the central fiscus. The DFLA therefore is unsustainable, and there is little prospect of it being able to resume lending - 16 - to Assemblies. According to a detailed report produced by a Bank mission in June, 2001, the DFLA has become an increasing burden on the central fiscus which has ultimately had to cover the cost of the Fund's asset depreciation. DFLA also has suffered from political influence over its management, in part because its Board has been composed exclusively of government officials with little or no experience in business or financial management. The DFLA apparently never incorporated any systems to assess the creditworthiness of the local authorities to which it "lent". In practice, the costs of new sub-projects financed under the LGDP would automatically appear as "loans" to the local authorities with no ex ante analysis, discussion, or agreements. Other loans outside the framework of the LGDP were made in similar ways. Although the SAR stated that "provision of funds for infrastructure will be linked to will be linked to the financial performance of the local authorities", there is no evidence that any such mechanism was designed or used. The Bank study recommended termination or withdrawal from the DFLA. There are two ways in which this could be done: (i) attempt to attract a private sector buyer for the loan portfolio which would then manage the portfolio as it saw fit; (ii) freeze the DFLA operations and, with the assets parked either in the existing vehicle (the Trust) or transferred to a government Ministry, manage the existing portfolio down to zero. In either case, government will need to develop an approach to deal with the distribution of the accumulated debt burden between itself and the borrower local authorities. It is recommended that the responsibility for dealing with the detailed policy and implementation of this option be delegated to the Local Government Finance Committee. Overall, in that the works were largely completed as planned (albeit generally behind schedule), the provision of infrastructure and services was achieved satisfactorily. However, the DFLA did not succeed as a self-sustaining financing intermediary for capital investments by the Assemblies, and it must be evaluated as unsatisfactory. 4.3 Net Present Value/Economic rate of return: The SAR presented economic rates of return (ERRs) for 11 of the physical investments including sites and services schemes, road improvements, and waste management in Lilongwe and Blantyre; markets in Blantyre and Zomba; and a bus depot in Zomba. The weighted average ERR for these subprojects, which represented 52 percent of the total project cost, was 16 percent. The economic benefits from the Policy Reform and Institutional Strengthening components were not quantified, but qualitative descriptions of the benefits were outlined. No data were available to permit ex post re-estimations of the economic benefits of the actual capital investments. The collection of actual quantitative benefits data for the various sub-projects (e.g., actual payments for plots in sites and services schemes, traffic counts on roads, actual costs per m3 of solid waste collected, actual incremental revenues from markets and bus parks) would have been an enormous task beyond the capabilities of the implementing agencies involved. Given the severe limitations on implementation capacity that persisted toward the end of the project in the Department of Local Government and the Project Coordination Unit, it was judged neither feasible nor worthwhile to organize a consulting contract to carry out such an assessment. - 17 - 4.4 Financial rate of return: Not applicable, as no financial rates of return were presented in the SAR. 4.5 Institutional development impact: IDI at central government level Through its Policy Reform Component, the LGDP set in place an effective platform for dialogue within central government, between the two tiers of government, and with donors over policies and institutional development of local governments and decentralization more broadly. Fourteen key studies were carried out with project funding and although not all the recommendations were finally adopted by government, they nonetheless provided continual "entry points" for keeping discussions alive. The momentum of the LGDP was complemented by parallel support to GOM by other donors, in particular UNDP and UNCDF through the 5th Country Assistance Program and the Local Govemance and Development Management Program. These sustained activities contributed significantly to the formulation and adoption of the legal, policy, and institutional framework described in Section 4.1 above. IDI at local government level The IDI at the local government level had inadequate results. One positive aspect is that, during the project, there was an improvement in financial accounting by the local Assemblies. At the inception of the project, most local assemblies were four years in arrears in maintenance of their books and having them audited. Financial information was not easily obtainable. The requirement that the participating assemblies submit their entity audited accounts nine months after year end did bring about some much needed discipline in the Assemblies to keep their books up-to-date, and audit reports of the participating assemblies have in most cases been submitted in time. For the most part, however, the impact of the project in terms of strengthening the capacity of the Assemblies to procure and manage contracts, to raise revenues, and to maintain basic infrastructure and services has been disappointing (see Section 4.2). On balance the institutional development impact of the project is rated as Modest. 5. Major Factors Affecting Implementation and Outcome 5.1 Factors outside the control of government or implementing agency: (a) Several ongoing donor-ftmded programs contributed directly to project preparation: * USAID funded a needs assessment preliminary strategy in 1989 and a resident MLG training advisor to assess the local government sector administrative capacity and supervise the design of training programs. * ODA funded a consultant reviewing local government legislation as well as technical assistance to three urban councils and assured its commitment to fund an urban development - 18- advisor to MLG for three years as well as to continue financing the legislative review. * KfW funded investments and technical assistance through the Department of Housing and Physical Planning Secondary Centers Development Project. * UNDP provided technical assistance to MLG for preparation of a project proposal under its Local Governance and Development Management Program, subsequently refinanced through a PPF. At appraisal, UNDP committed to further supporting the project and agreed to consider funding other technical assistance positions of the institutional strengthening components. At the time of appraisal, significant co-financing was envisaged under the project to complement IDA support to institutional strengthening activities (the restructuring of the DLG, the strengthening of the Assemblies, and higher level training). Grant co-financing for these activities was to be provided as follows: * Municipal Development Program ($0.3 million) for studies and t-aining for Assemblies. * ODA (DFID) (S0.8 million) for technical assistance. * UNDP ($1.7 million) for four of the long-term advisor positions and two UN volunteers at each of the three local government regional offices. * USAID ($0.7 million) for technical assistance. By the supervision mission of July 1993, however, it became clear that none of these funds would be forthcoming. The available documentation does not explain why the co-financing fell through. As a result, at least two of the four long-term technical advisors did not materialize, and this impacted negatively on the adoption of recommendations from the various policy studies and the capacity of MLG to monitor the performance of the Assemblies. Subsequently, UNDP did provide separate financing for several studies and technical assistance activities under its Local Governance and Development Management Program. (b) The poor performance of the contractors and supervising consultants on some of the works contracts led to their non-completion. For example, the contractor's cash flow problems (resulting in difficulties in timely procurement of materials and equipment for the contract) affected the completion of the Ndirande Market. A similar situation occurred with the Mzimba rest house. While the bulk of the civil works contracts were finalized before the end of the project, many of them overran their original schedules for completion, with their defects liability periods running beyond the project closing date. (c) Some turnover of officials and staff at the central and local levels due to HIV/AIDS had a negative effect on capacity-building. 5.2 Factors generally subject to government control. (a) Substantial devaluations of the local currency throughout the project period resulted in a - 19 - reduced requirement for the Credit's US dollars. Partially in response to this, SDR 500,000 was cancelled from the Credit in April 2000. The deteriorating value of the local currency also caused numerous contract management problems, as local contractors submitted claims for price escalations and requests to "peg" contract prices to the US dollar, contributing to slow completion rates on the works projects and substantial overruns on the scheduled completion dates. (b) High inflation and the accompanying high interest rate regime (around 46 percent per year in mid-1999) impacted strongly on the project. The Assemblies found it extremely difficult to budget accurately for sub-projects, and this resulted in slow implementation progress on many contracts as cost overruns had to be verified. High inflation had a devastating impact on the DFLA onlending scheme. In particular, the initial decision to charge a floating interest rate linked to the Central Bank rate subjected local authorities to strong upwards interest rate movements (increasing from 15 percent at the start of the project to 48 percent in 1998, to a high 60% in 2000). Without commensurate elasticities in their revenue base, this created major debt service difficulties for the Assemblies. The capping of DFLA interest rates at 15% in 1999 helped to diminish this problem but only by expanding the interest rate subsidy received by the Local Assemblies and exacerbating the sustainability problems confronting the DFLA. This, together with other factors, finally resulted in the Fund becoming insolvent. (c) In the assessment of Project Risks, the SAR noted that "there is a risk that the pace at which the entrenched control behavior of the central government can be modified to create an environment of self-reliance and independent decision-making in the Councils will be slow." The central government's commitment to creating an enabling environment for the development of efficient and effective local governments was noted by supervision missions as an ongoing risk throughout the life of the project. In fact, in 1994 the Government dissolved all existing local Councils (on the grounds that it wished to make way for fresh, multi-party elections at the local level). It took until November 2000 for the government to carry out new local elections. During the six-year interim, the Councils were run by appointed officials. This undermined the downward accountability of local governments, an essential requirement for the improvement of local government performance. Between 1994 and 1998 the government also held back on the legal, institutional, and financial reform agenda for local government, leaving the assemblies relatively "disempowered" in terms of financial and human resources. For example, during this period the government reneged on its commitments to provide transfers to the local authorities from the central fiscus and to pay subventions in lieu of property taxes on government properties. The lack of an enabling environment contributed to the increasingly poor financial performance of the Councils, which in turn led to significant negative impacts on the project. Firstly, it gave rise to very low loan servicing levels by the Local Assemblies to the DFLA, ultimately forcing the DFLA to the point of insolvency. Secondly, at various times during the course of the project, the Local Assemblies were unable to make good on their 15 percent contributions to the sub-projects being constructed. Faced with late or non-payment of the local counterpart funds, contractors' work programs slipped and a few sub-projects were not completed. Finally, it was possibly also a contributing factor to the lack of 'ownership' by local governments over their particular sub-projects, as evidenced for example, by their endorsing contractors' claims for escalations in price and for contract variations without thorough prior review. - 20 - (d) At appraisal, the Ministry of Local Government was identified to be the focal point for the project, responsible for monitoring the overall progress of the project and coordinating the respective assemblies and other agencies. These roles were not effectively performed by the Ministry due to lack of adequately skilled personnel and frequent changes of officials. (e) At the start of the project, as provided for in the SAR, a project coordinator was appointed in MLG "to ensure satisfactory and timely implementation of the project" as well as to present annual project progress reports to IDA. This proved to be ineffective, and in June 1994, the GOM and the Bank agreed to the establishment of a dedicated Project Coordination Unit (PCU) staffed by a full-time project manager, a full-time financial controller, and a part-time project engineer. While a useful improvement in principle, the PCU did not live up to expectations in practice. The assignment of roles and accountability between the Ministry and the PCU and between the PCU and the project Steering Committee (inter-Ministerial) were not well defined, which delayed key decisions and actions. Appointments to the PCU came to be seen as "political" rather than on the grounds of technical competence. Financial accounting, requiring coordination between the Ministry and the PCU, was poorly managed until the last two or three years of the project. The PCU never adopted a pro-active, problem-solving management approach. This combination of factors resulted in inadequate support to the Local Assemblies (in particular the smaller Assemblies). This manifested itself in poor contract management of works projects and missed opportunities for providing feedback to Assemblies to improve their performance. (f) Institutional development impacts were limited by frequent changes of officials and staff at central and local levels. (g) The above-mentioned factors combined to take away from the local authorities whatever sense of ownership they may have built up during project preparation. The project gradually came to be seen as belonging mainly to the central government or to the Bank, and as such did not receive optimum levels of staff commitment and involvement in implementation (this was especially true of the smaller Assemblies). This was exacerbated by the absence of technically qualified engineering and project management staff in either the Department of Local Government or the PCU, which meant that opportunities for iterative "learning-by-doing" were lost. 5.3 Factors generally subject to implementing agency control: (a) The PCU, as mentioned above, operated in a generally passive, reactive way. This led to inadequate supervision of implementation at the local authority level and delays in coming to grips with problems between World Bank missions. (b) As implementing agencies for the investments sub-projects, the local authorities were poorly equipped with either procurement or contract administration skills and experience (except for Blantyre and Lilongwe City Assemblies). It took some time for the Assemblies to understand and accept Bank procurement requirements, and there were instances of Assemblies giving verbal advice or approvals to contractors which subsequently had to be changed to conform with Bank guidelines. - 21 - (c) Cost recovery on the site and service schemes was weak. Actual construction costs were higher than planned. The plots appear to have been allocated mainly to middle income rather than lower income households. 5.4 Costs andfinancing: The total project cost at appraisal was projected at US$36.3 million, of which US$24.0 million would be financed by IDA, US$9.1 million by GOM and the Local Assemblies, and US$3.2 million would be grant co-financing. The IDA Credit, approved on June 4, 1992, provided the original principal of SDR 17.6 million (approximately US$24.47 m.), close to the appraisal figure. However, as noted in Section 5.1, the grant co-financing did not materialize. In February 2000, the project cost was reviewed, resulting in a new estimate of $28 m. IDA would contribute $22.6 m., and the remaining $5.4 m. would be met by GOM and the Local Assemblies. In light of the revised estimates for IDA financing, SDR 500,000 ($672,265) was cancelled from the Credit on March 17, 2000. The reduction in the project cost ($7 m.) was made from the Infrastructure and Services component after the feasibility of starting new infrastructure sub-projects was assessed by the Bank. Four main criteria were used in the assessment: (a) the track record in cost recovery and maintenance of previous investments; (b) the likely abilities of the Local Assemblies to adequately finance maintenance in the future; (c) Local Assemblies' capacities to manage implementation of the new sub-projects; and (d) the likelihood of completion within the extension period. A further contributing factor to the revised cost estimate (over and above the reduced number of infrastructure sub-projects) was the substantial devaluation of the local currency over the life of the project, resulting in a reduced requirement for the credit's US dollars. (At the time of project appraisal, $1 = MK2.96, but this had fallen to $1 = MK70 by the close of the project). By the end of the project (data as of November 15, 2001), total IDA disbursements amounted to $19.89 m., lower than the estimate of February 2000. The difference of $2.7 m. between the estimate made in 2000 and the final expended amount was primarily due to (a) slow completion of ongoing projects (please refer to Section 4.2.c for contracts not completed at project closing); (b) the completion of two consultancy studies by the government using grant funds from UNDP rather than the IDA credit (the study on the design of the intergovernmental fiscal transfer system and the study into enhancing local government revenues, to an amount of $335,000); and (c) the cancellation in April 2001 of the Phwetekere sites and services scheme Phase B (an amount of $635,000) when it became clear that the project could not be completed before the project closing date (in addition to the late start up of contract award, the risks were exacerbated by the need to resettle and compensate 40 homesteads which had moved into the site after it had been designed). 6. Sustainability 6.1 Rationale for sustainability rating: - 22 - For this project, evaluation of sustainability is based on three elements. First is the issue of sustainable financing for Assembly capital investments. Second is the ability of the Assemblies to sustain these investments through proper operations and maintenance procedures. Third is the sustainability of the "enabling framework" of laws, policies, and institutions for local government development and eventual service provision by the Assemblies. Each of these considerations is discussed below. (a) Sustainable financing of Assembly investments As explained in Section 4.2.c. above, the DFLA as a financing mechanism for Local Assembly capital investments is unsustainable, and there is little prospect of it being able to resume lending to Assemblies. The DFLA has become an increasing burden on the central fiscus which has ultimately had to cover the cost of the Fund's asset depreciation. The Ministry of Finance is currently reviewing options to either sell off the loan portfolio of the Fund or to manage it down in a debt sharing relationship with the Assemblies. The recognition of this situation has in turn given rise to a discussion within Malawi as to the extent to which credit is an appropriate means of financing investments for local authorities. The government is shifting its focus to unconditional block grants to Assemblies (through the intergovernmental fiscal transfer system) complemented by local government own-source revenue enhancement measures. (b) Sustainabilitv of the investments by the Assemblies Regarding the extent to which the Assemblies are able to operate and maintain the investments made under the project, the experiences to date have been mixed. All Assemblies visited by the ICR mission were operating and maintaining equipment which had been procured under the project (primarily municipal service vehicles and office equipment) and in one case (Luchenza Town Council), leasing out a vehicle on a commercial basis to derive revenues for its maintenance (although the Assembly had not done a cost-benefit analysis to determine if the additional operating costs would in fact be recouped). The markets completed under the project and now in use (for example, Ngongomwa and Chinamwali in Zomba, and Limbe in Blantyre) had high occupancy rates, and the Assemblies were deriving significant monthly incomes from them. An analysis of income and expenditures6 by Blantyre City Assembly on the Limbe market for July 2000 through April 2001 showed that revenues were 2.6 times expenditures (and in ten out of twelve markets in the city, the Assembly made a profit over the period). However, the roads which had been upgraded under the project (for example, the Ndirande Loop Road (Blantyre) and the local access roads in Phwetekere sites and services scheme (Lilongwe) showed physical signs of low levels of maintenance. Storm water drains were blocked with silt and garbage resulting in drainage down and erosion to the road reserve itself. Culverts providing off-street access were blocked and in some cases Market fees and rent 6 Water, electricity, Telephone, Salaries and wages, Bus fare and Maintenence - 23 - collapsed. Similarly, a detailed assessment of cost recovery procedures in the site and service schemes carried out by the Bank mission in July 1999 indicated that recovery rates were low and that there was a reluctance by both local and central government to attempt to recover full costs on the schemes. Overall it must be recognized that the own source revenue potential of Malawi's local authorities is structurally limited. Malawi has one of the lowest per capita incomes in the world, and only in the four main urban centers is there any possibility of levying or collecting any significant amounts from property taxes (rates). In the District Assemblies, over half of all revenues derive from market fees. In addition to these structural constraints, actual collection rates are low and erratic. For example, over the six-month period July - December 2000, actual revenue inflows in Blantyre stood at about 75 percent of the budgeted total; for Lilongwe the figure was only 58 percent. The Constitution of Malawi explicitly recognizes the need for the devolution of responsibilities to local governments to be matched by appropriate resources. This is to be achieved not only by improved revenue collection efficiencies, but also by an appropriate system of fiscal transfers from the center. In this context, the government recently commenced a fundamental restructuring of the intergovernmental fiscal transfer system. From FY2001/02, five percent of national revenues are being transferred, based on an allocative formula, to all Assemblies as unconditional block grants. Initially at least 60% of the grants are intended for recurrent expenditures. The central government has also recently agreed to begin paying rates on its properties to Assemblies more consistently and at higher levels (50%) than it has done in the past which will in turn free up resources for O&M requirements of the Assemblies. Much more still needs to be done, both with respect to policy (agreements on revenue sharing principles, long-t erm intergovernmental fiscal transfers) and to strengthening the capacity of Assemblies improvements to revenue collection systems and budgeting and prioritizing development and maintenance needs. (c) Facilitative operating environment for local authorities As outlined in Section 4.2 (a) above, the government has now put in place a series of legal, policy, and institutional reforms, for which the project can claim partial credit, constituting a good foundation for sustaining progress, over the long term, on strengthening local governments to provide basic services. Conclusion: The financing mechanism for local capital investments failed; the government did not put in place the intergovernmental fiscal transfer system; and own-source revenue collection by the local authorities did not improve. On the other hand, some key foundations of the policy and institutional framework were finally laid, and the government is now seeking donor support for a - 24 - comprehensive Decentralization Implementation Plan. However, the resiliency of these gains to future political and macro-economic risks remains uncertain. On balance, overall sustainability of the project is evaluated as Unlikely 6.2 Transition arrangement to regular operations: None of the Assemblies prepared a plan for the operation and maintenance of the investments delivered through the project. However, during the project completion mission, the Assemblies expressed their readiness to undertake the maintenance of the facilities in their respective jurisdictions. Roads: All the roads upgraded under the project are concentrated in Lilongwe and Blantyre cities. Both LCA and BCA agreed to include in the budget beginning from 2001/2002, funds for operation and maintenance. Markets: The project supported the upgrading of three markets in Blantyre, two in Zomba, one in Nsanje and two in Ntcheu. All of the markets are functional. The two markets in Zomba and the two in Ntcheu are in use despite the outstanding works which the Assemblies plan to complete with their own funds. The increased revenues accruing from the operation of the markets would suggest that the Assemblies should be able to include in their budgets funds for the maintenance of these markets. Site and Services: The site and services subprojects in Blantyre, Lilongwe and Mzuzu were successfully completed. In the case of the Chilomoni site and services, BCA plan to maintain the facilities through user charges and collection of ground rents. LCA has had difficulty with cost recovery on the Phwetekere site and services scheme due to lack of proper records of payment, a faulty data base, and high staff turnover. Vehicles and Equipment: Site visits during the completion mission indicate that the vehicles and equipment acquired through the project are being maintained. However, there is need for the Assemblies to develop plans for the regular maintenance of these items. With regard to the Infrastructure and Services component in general, as noted in Section 6.1 (ii) above, the ability of the Assemblies to adequately operate and maintain the works and equipment over the medium term is not a given, even in those instances where the facilities are generating incomes for the Assemblies (e.g. markets). The Government has recently initiated a pilot project in seven Assemblies to carry out in-depth analysis of local tax collections and the administrative procedures used. However, the effectiveness of these measures are yet to be seen and will need to be constantly monitored in the future. A future impact evaluation would be prudent during the second year following full completion of all sub-projects and after the expiry of the defects liability periods At that time, Assembly budget allocations for O&M could be better monitored. Additional capital financing is required for the Assemblies to undertake new capital investments to keep pace with population growth. Moderate amounts of funding are available through the District Development Fund (UNCDF supported) and the Malawi Social Action Fund (IDA supported), but in neither case are sufficient levels of finance available for the types and size of investments required in the larger towns and cities. Given that the new fiscal transfers are - 25 - intended for recurrent costs, no financing arrangement exists to support new local government investments. With regard to the Policy Reform and Institutional Strengthening components, additional financing would be required to maintain the momentum gained under the project and to complement declining resources from the UNDP Local Governance and Management Development Program. GOM has recognized the need for this and has incorporated ongoing training, policy studies and institutional development activities into its adopted Decentralization Implementation Plan. A Donors' Roundtable, facilitated by the DLG's Decentralization Secretariat, was held in August 2001 to solicit donor support to the activities budgeted for in the Plan, and it does appear likely that some grant finance will be available over the next five years to support some of these activities. Joint Donor/GOM quarterly meetings are being held, and annual appraisal joint missions are to be fielded to monitor progress on decentralization implementation on the basis of an agreed log-frame. 7. Bank and Borrower Performance Bank 7.1 Lending: The Bank fielded a sufficient number of reasonably staffed missions and carried out adequate consultations during project identification, preparation and appraisal, but a number of critical risks were misjudged in the project design. These issues are covered in Section 3.5, Quality at Entry. In summary, the Bank misjudged the risks regarding: * The government's commitment to policy reforms for decentralization. * The sensitivity of the local authorities' financial situation to downturns in macro-economic factors. * The influence of technical assistance and training on local institutional and financial capacity (see SAR para. 3.3). * The implementation capacity of MLG. * The viability of the DLFA in relation macro-economic factors and the creditworthiness of the LAs (see paras. 2.12 and 2.13 of the SAR--there is no evidence of any risk analysis or creditworthiness analysis). Neither the "yellow cover review" memo (2/28/92) nor the "pre appraisal review" memo (5/10/91) come to grips with these issues. The only mention of these risks is one reference in the yellow cover memo to the weak finances of the councils. The memo states that this risk will be mitigated by (i) the fact that most of the investments to be financed will be "income generating" and (ii) the provision of technical assistance to strengthen the LAs. A serious omission in the initial project design was the failure to establish indicators to monitor - 26 - project progress and impact with respect to any of the components. No monitoring framework was put in place for assessing the impact of the government's policies or the project's interventions on the local authorities' financial or technical performance. A more realistic appraisal of the risks might have produced a more modest project, perhaps a pilot or a technical assistance operation to test the government's commitment to local government development and lay a foundation of policy reforms - especially the intergovernmental fiscal and human resource framework - before moving to major capacity building and investment components. 7.2 Supervision: Again, the Bank fielded an adequate number of reasonably staffed supervision missions during the life of the project. However, the record shows that the Bank was insufficiently pro-active in addressing the problems that arose. The Bank never effectively addressed the problem of performance monitoring. The Bank failed to grasp the fundamental structural deficiencies of the DFLA which became evident early in the project. The Bank intervened on the margins several times to increase the shares of grant financing and reduce the shares of loan financing, but it never came to grips with the larger forces that were rendering the DFLA unviable. The mid-term review in 1996 was a major missed opportunity for re-thinking the whole concept. By the time a new Bank team took over in 1999 and realized the magnitude of the problem, it was too late. The Bank is to be commended at least for carrying out a detailed study and a workshop in June 2001 to analyze the problems and propose a strategy to resolve them. In the face of the stagnation of policy reform from 1994 onwards, the Bank could have intervened forcefully to restructure the project, perhaps by scaling back the objectives, canceling components or sub-components, re-thinking the financing mechanism as mentioned above, and re-designing the implementation arrangements to provide more direct support to the local authorities. No such actions were taken. The Bank did not intervene forcefully to correct the obvious inadequacies of the Project Coordination Unit. The PCU could have been overhauled at mid-term. There also were delays at various times on the Bank's side with regard to reviewing documents and giving necessary clearances (no objections). At certain points this was due to changes of task managers responsible for the project. 7.3 Overall Bank performance. The Bank's performance is rated Unsatisfactory overall Borrower 7.4 Preparation: The Ministry of local Government showed strong commitment to enhancing the effectiveness of the local government system in Malawi from the earliest phase of projecl. preparation. This - 27 - emerged in particular during the workshop on "Strengthening Local Government in Malawi" held in Mangochi in 1990. The government's commitment to a participatory approach was shown by its decision to base the initial project proposal on the Council Development Plans developed by the Local Authorities. The borrower's approach was also grounded in realism, as shown in 1991, when the Department of Economic Planning and Development and the Ministry of Finance requested (i) that the LGDP be considered as the initial phase of a long-term process aimed at strengthening the local authorities' fiscal and administrative capacity and (ii) that it be followed by a second investment operation about three years from project effectiveness. Realism was also shown in the borrower's recommendation to drop the "Economic Development" component and to focus on activities more directly related to supporting the local authorities selected for infrastructure investment. The government also supported limiting the scope of the infrastructure investments and reducing the overall size of the project to about $50 million. Unfamiliarity with World Bank procedures and inadequate communication among government agencies and between the participating Councils and MLG resulted in some delays (e.g., 11 months between appraisal and negotiations and a three month delay in effectiveness). In addition, project management, located in MLG, had competing demands on its time and was not able to proactively follow up on day to day management issues. In fact, it emerged that structure of the IvMLG was inadequate to support implementation of the project. The restructuring proposed during the appraisal mission of May 1991 included devolution of management functions from MLG to regional offices as well as to local authorities. The position of Commissioner of Local Government was introduced for the purpose of providing policy oversight and managing policy formulation. A Local Government Management and Finance Unit was also set up in MLG to handle multiple management and administrative functions. In the event, these changes proved to be inadequate to assure a sufficient level of performance by MLG once implementation started. The above problems were exacerbated by the low-key role played by the inter-ministerial project steering committee. The project did not gain "ownership" from key ministries such as Finance and Works 7.5 Government implementation performance: After the second supervision mission, less than a year into project implementation, it became clear that new arrangements needed to be put in place to strengthen project management and coordination. The Institutional Strengthening component was delayed by a lack of capacity in MLG's training unit. The procurement of most studies under the Policy Reform component did not proceed on time. During the supervision mission of November 1993, it was recommended that a Project Coordination Unit be formally established within the MLG structure (accountable to the Principal Secretary) but with the specific responsibilities of implementing the Project and including a Project Manager, a Project Accountant and a Project Engineer. However, the PCU was not formally established until the end 1995, as preparation of the organizational structure and the operating budget for the PCU were delayed. Major reasons were the restructuring of MLG, which had in the meantime assumed responsibility over rural development, and the dissolution of - 28 - the elected councils in 1994. As of 1996, the establishment of the PCUJ, coupled with the incorporation of an element of grant funding for infrastructure and equipment investments, led to an increased pace of disbursement and progress being made on the project activities. A number of significant problems with the borrower's performance are presented in Section 5.2, namely: * A loss of macro-economic stability, manifested in high inflation and steep currency devaluation over the life of the project. * Dissolution of the elected councils in 1994, without new elections until late 2000 which rendered the councils unaccountable to their constituents. * A hiatus in policy reforms from 1994 to 1998 which left the councils unempowered. * Frequent changes of officials and staff. Financial management for the project within the Ministry was weak A recently completed. independent audit of the project for 1993/94 through 1999/2000 reported that there were many intemal control and accounting problems that were common to each year, including: * No financial statements produced in any year to give a total comprehensive view of the income and expenditure of the project. * Withdrawals into the local Malawi Kwacha account were not based on actual expenditures. * A significant amount of expenditures did not have proper supporting documents. * No fixed asset register available for any of the seven years. * Bank reconciliations were not properly prepared and were not produced on a timely basis. 7.6 Implementing Agency: Please refer to Section 5.3 regarding shortcomings in the PCU and the local authorities. 7.7 Overall Borrowerperformance: The Borrower's perfornance is rated Unsatisfactory overall. 8. Lessons Learned (a) The shortcomings of the government in policy formulation and project implementation are not unique to this project'. One of the clear lessons to emerge from the LGDP experience is the need to establish a strong project unit. Such a unit needs to be well managed and have strong financial and technical (engineering, procurement, enviromnental, community liaison, etc.) skills 7 As of November 2001, 3 of 12 active projects in the Bank's portfolio had long standing implementation problems, and difficulties with policy dialogue affect several sectors. - 29 - recruited into it on a contract (suitably remunerated) basis. This in turn requires that clear lines of responsibility and accountability between the unit and the ministry be established during project design. (b) Given the risks inherent in govemance reform and capacity building, project objectives need to be appropriately modest. Risk assessment and mitigation in project design must be much more rigorous, including the scaling down of projects if necessary. (c) Bank supervision missions need to be (i) critically honest in assessing progress in project implementation and achievement of the Development Objective and (ii) suitably interventionist when components are clearly under-performing. (d) Building on the points made above, projects need to be designed in a performance driven way (achievement of certain triggers before a next set of resources are provided or activities undertaken - the Bank's APL is an instrument well-suited to this approach) and this in turn requires the early design, adoption and implementation of a rigorous monitoring system (based on the use of parsimonious but quantifiable indicators). (e) Local government projects can usefully link capacity building to capital investment (and rehabilitation and maintenance) programs. The challenge is to ensure that positive incentives are provided to the local authorities and that these are transparently linked to increased levels of performance management (the antithesis of the perverse incentives provided under the LGDP, where, for example, the better performing Assemblies were allocated smaller grant percentages through the DFLA than weaker Assemblies). (f) Today the Bank is much more cautious about supporting special loan mechanisms for local governments than it was when the LGDP was formulated. Whenever sub-national lending is contemplated, a much more rigorous analysis should be carried out on both the supply and demand sides. The burden is on the side ofjustifying (i) whether local governments are creditworthy at all and (ii) if so, why a special financial intermediary is needed. (g) The LGDP has shown that, in the case of Malawi, the Assemblies will require grant or matching grant funds for the foreseeable future. The allocation of these grants could be linked to a mix of criteria based on Assembly performance, poverty indicators and investment needs (based on, for example, regional needs). (h) By the 1990s, the Bank had phased out its support for sites and services schemes, based on world-wide experience similar to that which occurred in the LGDP. Direct poverty targeting in urban/local government projects is generally focused on community upgrading on low-cost water and environmental sanitation. 9. Partner Comments (a) Borrower/implementing agency: As of the date of filing, the Borrower's Comments and Contribution had not been received despite various reminders - 30 - (b) Cofinanciers. Not applicable (c) Other partners (NGOs/private sector): Not applicable 10. Additional Information As of the date of filing, the Borrower had not subrnitted its separate evaluation despite various reminders. - 31 - Annex 1. Key Performance Indicators/Log Frame Matrix Note: No performance indicators were identified in the SAR as the Log Frame was not in place at the time Projected Outcomes Outputs POLICY REFORM COMPONENT Local Authority budget system study Completed (1994) Accounting and internal control systems in place - Accounting system manual for Local Authorities Budget and accounting procedures manual completed (KPMG Peat Marwick, 1992) Grants to Council Study completed (Price Waterhouse Panafrica) Rating and Valuation Review Completed Local Authority Financial Administration Analysis Financial management of LAs study completed (Deloitte and Touche, 1992) Financial Management Workshop completed (Deloitte and Touche, 1995) Policy related training See table 1.1 below Local Government Legislation Review: a) Finalize review of LA legislation and propose Local Government legislation review completed modifications to describe expanded role and (Solace International) responsibilities for councils - policy workshop leading to action plan b) draft legislation, as required, based on the above Adoption of a Decentralization Policy (1998) Passage of amended Local Government Act (1998) c) draft handbook clarifying the relationship of LAs Not completed to central government, including guidelines for council operation Policy workshop on role of the District Policy workshop on role of district commissioner Administration vis the local Authorities versus the local authorities completed Study identifying the Financial, Human and Completed (1992) Management Resources requirements and procedures for administering the THAs Study for integrating LG revolving funds Not completed Workshop defining role of MLG and LAs in Not completed developing social services Study and policy workshop to assess constraints on Completed (M.G. Tsoka, 1994) economic activities, deregulation, and to identify the economic development potential of LAs Development Fund for Local Authorities Completed (2000) -32 - INSTITUTIONAL STRENGTHENING COMPONENT Preparation and implementation of a business plan Five-year plan for MLG completed. based on the MLG's new mandate in supporting the development of councils. MLG senior management and unit chiefs must adopt new operating procedures based on implementing the business plan Development and implementation of a more Achieved decentralized operating procedure and decision making process Reorganization and staffing of the MLG Recruitment of key personnel of Chief local government officer, principal Finance Officer, Principal law Officer and Principal planning Officer completed Specialized advisory assistance to the MLG Urban Management Adviser Not hired due to failure of co-financing plan Chief Technical Adviser Not hired due to failure of co-financing plan Training Consultancy Not completed due to failure of co-financing plan Financial Analyst and Accountant Hired Estate Development and Management Adviser Hired Municipal Engineering Adviser Not hired due to failure of co-financing plan Short Term Technical Assistance Partially completed - six economists/financial analysts hired Purchase of vehicles, equipment and short life assets Completed - equipment and support for incremental operation costs provided Commercial related buildings Completed Municipal infrastructure Completed MLG courses and workshops Local government staff trained, including in-service training, entry level training, and academic training See Table 1.2 below for details INFRASTRUCTURE AND EQUIPMENT COMPONENT Lilongwe City Assembly Provision of New Shelter - 3,200 serviced plots Partially completed - 521 plots demarcated and developed allocated to public Traditional Housing Upgrade Not completed Waste Management Completed - Equipment for solid waste management delivered and put to use - 33 - Works Yard Upgrading Improvement of market completed Work Depot put to use Geographical Information System Completed Malangalanga Altemative Route Paving of Malalanga Road completed Improvement of link ml -Malalanga Road completed Blantyre City Assembly ___ Site and Service Scheme Chilomoni South Site and Service Scheme Completed - 1.6km of surfaced road; 5.6km of gravel road, 9.8km of storm water drains, road fumiture, landscaping electricity and water supply were provided. Completed - affordable land plots for low-middle income households provided - 483 residential plots, 50 commercial plots, 9 institutional plots demarcated THA Pilot Upgrade completed - 3.5km of roads and drains provided and provision made for sanitation improvement Produce Market Developments Limbe Market - New food hall, fore court, 12-room public toilets, refuse bunker refurbished, boundary wall constructed, market master's office, electrical works and refurbishment Chilomoni Market - 12 open stalls constructed, market area paved, existing sted refurbished and new shed constructed Ndirande Market extension not completed Solid Waste Management Not completed Nkolokoti Parkway Link Upgrading Completed - 3.1 km road upgraded Zalewa Road Streetlighting Completed - 8 km street lighting constructed Computer Requirements - Computer Department Not completed Zomba City Assemblv Upgrade of Ngongomwa and Chinamwali Secondary Completed Markets Bus Depot Improvements Completed - access road constructed, service block provided, bus island canopy provided, toilets completed Vehicle Inventory and Acquisitions Not completed Works Completed not Planned in SAR - 34 - Mzuzu City Assembly Luwinga site - 662 serviced plots for low-income households completed 8.4 km road comnpleted 0.8 km lined drain completed l . 1 km unlined drain completed Blantyre City Assembly Limbe Bus Depot - access road constructed, service block provided, bus island and canopy provided, gate houses and erection of boundary wall completed District Assemblies M'mbelwa Resthouse at Mzimba Boma 80% completed Mphate market in Ntcheu District - stalls built, coldroom completed, electrical works not completed Kampepuza Market in Ntcheu District - sheds built, market master's office completed slaughter house and perimeter fence completed Nsanje Market - lock-up shops, selling sheds, toilets, canteen, perimeter fence completed - 35 - Table 1.1 Name of Course Date No. Of Participants Basic Local Government Induction Course July 1994 25 Workshop on Management of Local Authorities in the Nov. 1994 2 Era of and Reforms in Africa Study Tour on Squatter Upgrade and Low Income Aug. 1996 8 Housing Development Transfomnation of City and Town Councils in Africa Aug 1997 2 Local Government Reform and Decentralization Aug. 1997 I management Integrated Development Strategy Study Tour August 1998 8 Africities 2000 Summit May 2000 8 Pan African Workshop on Community Participation June 2000 1 and HIV/AIDS Regional Roundtable on Upgrading Low Income Oct. 2000 3 Settlements Workshop for Chief Executives Nov. 2000 African Local Govemment Action Forum Distance Nov, Dec. 2000 46 Leaming Workshops March, April, May, June 2001 Orientation program for Assembly Members April/May 2001 Approx. 740 Table 1.2 Name of Course Date No. Of Trainees Effective Manager Program 1993 10 Advanced Leadership and Management Course 1993 I Modem Techniques of Managing Committee 1993 1 Works/Meetings Financial Management for Donor Funded Projects 1993 10 Building Foreman Course 1994 7 Construction management Program 1994 5 Certificate of Accounting 1994 25 Effective Management by objectives 1994 26 Management of Training 1995 1 Accounting Technician Course 1995 Training Priorities in Local Authority Technical 1995 44 Departments Works Procurement Management Program 1995 5 Works Procurement Management 1995 57 Training of Trainers Workshop 1995 24 Councilors' Handbook 1995 Local Authorities' Administration and Management 1995 Handbook _ Accounting Skills Application Workshop 1995 57 Management Skills Application Workshop 1995 63 Local Government Administration and Finance Course 1995 25 Computer Training Course 1995 12 Course on Local Administration and Finance 1996 25 IDA Procurement Seminar 1996 3 - 36 - Masters in development Planning and Project Planning 1996 1 Training Delivery Skills 1996 1 Masters in Development Planning and Project Planning 1997 1 Masters in Planning 1997 2 Masters in Urban Management/ Town Planning 1997 2 Masters in Local Government and Development 1997 1 Project Management 1997 2 Orientation to Training 1997 1997 20 Local Government Administration and Finance 1997 23 Local Authority Accounting 1997 36 Systems Maintenance Management 1997 35 Financial Management of Donor Funded Projects 1997 2 Goods and Equipment Procurement Course 1997 3 Diploma in development Studies 1997 1 Micro-Computer Based Data Management 1997 3 LCA - Training on Budget and Budgetary Controls 1997 approx. 12 Human resources Management Course 1998 2 Local Government Management Course 1998 Environmental Health Course 1998 8 Accounting, Budgeting and Budgetary Control 1998 Diploma Program on Environmental Engineering 1998 1 Diploma Course in Accounting 1998 16 Financial Management in Public Sector Course 1998 3 Systems Engineering Seminar 1998 15 Workshop on Procurement 1999 8 Masters' Program on Environmental Engineering 2000 1 Procurement Capacity Building Training Workshop 2000 4 Training of Trainers 2000 27 - 37 - Annex 2. Project Costs and Financing Project Cost by Component (in US$ million equivalent) _ Appraisal Actual/Latest Percentage of Estimate Estimate Appraisal Project Cost By Component US$ million US$ million Policy Reform 2.80 1.48 53 Institutional Strengthening 9.70 4.97 51 Provision of Infrastructure 17.00 10.63 63 Total Baseline Cost 29.50 17.08 Physical Contingencies 2.90 Price Contingencies 3.90 Total Project Costs 36.30 17.08 Total Financing Required 36.30 17.08 'roject Costs by Procureme nt Arrangements (Appraisal Estimate) (US$ million equivalent) Procurement Method Expenditure Category ICB NCB Other2 N.B.F. Total Cost 1. Works 15.00 3.80 0.00 0.00 18.80 (15.00) (0.05) (0.00) (0.00) (15.05) 2. Goods 3.70 0.00 0.00 0.00 3.70 (3.70) (0.00) (0.00) (0.00) (3.70) 3. Services 0.00 0.00 4.20 3.50 7.70 (0.00) (0.00) (4.10) (0.00) (4.10) 4. Miscellaneous ( 0.00 0.00 6.10 0.00 6.10 Operating Expenses and (0.00) (0.00) (0.07) (0.00) (0.07) Other) 5. Miscellaneous 0.00 0.00 0.00 0.00 0.00 (0.00) (0.00) (0.00) (0.00) (0.00) 6. Miscellaneous 0.00 0.00 0.00 0.00 0.00 (0.00) (0.00) (0.00) (0.00) (0.00) Total 18.70 3.80 10.30 3.50 36.30 (18.70) (0.05) (4.17) (0.00) (22.92) Project Costs by Procureme nt Arrangements (Actual/Latest Estimate) (US$ million equival nt) Procurement Method Expenditure Category ICB NCB Other2 N.B.F. Total Cost 1. Works 3.37 8.68 0.20 0.00 12.25 (0.00) (0.00) (0.00) (0.00) (0.00) 2. Goods 3.32 0.00 0.08 0.00 3.40 (0.00) (0.00) (0.00) (0.00) (0.00) 3. Services 0.00 0.00 5.13 0.00 5.13 (0.00) (0.00) (0.00) (0.00) (0.00) - 38 - 4. Miscellaneous( 0.00 0.00 0.00 Operating Expenses and (0.00) (0.00) (0.00) (0.00) (0.00) Other) 5. Miscellaneous 0.00 0.00 0.00 0.00 0.00 (0.00) (0.00) (0.00) (0.00) (0.00) 6. Miscellaneous 0.00 0.00 0.00 0.00 0.00 (0.00) (0.00) (0.00) (0.00) (0.00) Total 6.69 8.68 5.41 0.00 20.78 (0.00) (0.00) (0.00) (0.00) (0.00) The figures for the Waste Management component of this project are included under the Services Category. The costs for Waste Management were $1,080,067. ` Figures in parenthesis are the amounts to be financed by the Bank Loan. All costs include contingencies. 2' Includes civil works and goods to be procured through national shopping, consulting services, services of contracted staff of the project management office, training, technical assistance services, and incremental operating costs related to (i) managing the project, and (ii) re-lending project funds to local government units. Project Financing by Component (in US$ million equivalent) Percentage of Appraisal Component Appraisal Estimate Actual/Latest Estimate l Bank Govt. CoF. Bank Govt. CoF. Bank Govt. ICOF. Policy Reform 0.95 2.30 0.0 0.0 Institutional Strengthening 5.08 5.66 0.65 0.0 0.0 0.0 Infrastructure and 18.00 3.40 0.0 0.0 Equipment -39 - Annex 3. Economic Costs and Benefits The SAR presented economic rates of return (ERRs) for 11 of the physical investments including sites and services schemes, road improvements, and waste management in Lilongwe and Blantyre; markets in Blantyre and Zomba; and a bus depot in Zomba. The weighted average ERR for these subprojects, which represented 52 percent of the total project cost, was 16 percent. The economic benefits from the Policy Reform and Institutional Strengthening components were not quantified, but qualitative descriptions of the benefits were outlined. No data were available to permit ex post re-estimations of the economic benefits of the actual capital investments. The collection of actual quantitative benefits data for the various sub-projects (e.g., actual payments for plots in sites and services schemes, traffic counts on roads, actual costs per m3 of solid waste collected, actual incremental revenues from markets and bus parks) would have been an enormous task beyond the capabilities of the implementing agencies involved. Given the severe limitations on implementation capacity that persisted toward the end of the project in the Department of Local Government and the Project Coordination Unit, it was judged neither feasible nor worthwhile to organize a consulting contract to carry out such an assessment. - 40 - Annex 4. Bank Inputs (a) Missions: Stage of Project Cycle No. of Persons and Specialty Performance Rating (e.g. 2 Economists, I FMS, etc.) Implementation Development Month/Year Count Specialty Progress Objective Identification/Preparation FY85-FY9 1 Appraisal/Negotiation May 15 - 30, 3 Mr. A. Zeijon, (Economist, 1991 Mission Leader Mr. G. Gattoni, Urban Planner Mr. C. Banes, Municipal Engineer (Consultant) December 9 - 18, 6 Mr. A. Zeijlon -Mission Leader 1991 Mr. G. Gattoni, Principal Urban Planner Ms. G. Kajubi, Urban Planner Mr. E. Clarke, Municipal Engineer (Consultant) Mr. K.E. Lindberg, Financial Analyst (Consultant) Mr. H. Hartikainen, GIS Expert (Consultant) Supervision September I - 10, 4 Mr. A. Zeijlon, Mission 1992 - Project Leader Launch Ms. G. Kajubi Urban Workshop) Planner Mr. C. Banes, Municipal Engineer (Consultant) Mr. P. Berrningham, Financial Analyst (Consultant) May 10 - 21, 1993 6 Ms. Naa Dei Nikoi, Mission Leader Ms. Gibwa Kajubi, Urban Planner Mr. C. Banes, Municipal Engineer (Consultant) Mr. P. Bermingham, Financial Analyst (Consultant) Mr. David Kithakye, Municipal Development Program Mr. George Matovu, Municipal Development Program September 6 - 17, 3 Ms. Naa Dei Nikoi, Mission 1993 Leader Ms. Gibwa Kajubi, Urban Planner Mr. P. Bermingham, Financial Analyst _____ _ - 41 - April 18 - 28, 1994 4 Ms. Naa Dei Nikoi ,Mission S S Leader Mr. D. Gallagher, Financial Analyst Mr. C. Banes, Municipal Engineer Mr. G. Matovu, Municipal Development Program September 26 - 4 Ms. Naa Dei Nikoi, Mission S S October 6, 1994 Leader Mr. D. Gallagher Financial Analyst Mr. C. Banes, Municipal Engineer Mr. G. Matovu of Municipal Development Program May I - 12, 1995 4 Ms. Naa Dei Nikoi, Operations S S Officer Mr. D. Gallagher, Financial Analyst Mr. C. Banes, Municipal Engineer Mr. G. Matovu, Municipal Development Program October 16 - 27, 6 Ms. Naa Dei Nikoi S S 1995 Ms. Judy Makanda, Financial Analyst Mr. Chris Banes Ms. Damiam Gallagher Ms. Mila Zlatic Mr. George Matovu, Municipal Development Program May 6 - 22, 1996 7 Ms. Naa Dei Nikoi, Mission S S Leader, Ms. Judy Makanda, Financial Analyst Ms. Marianne Fay, Economist Mr. Chris Banes, Municipal Engineer, (Consultant) Mr. Damian Gallagher Ms. Mila Zlatic Mr. George Matovu,, Municipal Development Program November 19 - 22, 1 Ms. Judy Makanda, Financial S S 1996 Analyst January 19 - 29, 4 Ms. Naa Dei Nikoi, Operations S S 1997) Officer Ms. Judy Makanda, Financial Analyst Mr. Chris Banes, Municipal Engineer (Consultant) Ms. Marianne Fay, Economist -42 - November 19 - 28, 4 M;s Judy Makanda, Mission S U 1997 Leader, Financial Analyst Ms. Asamanetch Fantaye, Operations Analyst Mr. Chris Banes, Municipal Engineer (Consultant) Mr. George Matovu, Municipal Development Program April 26 - May 2, 1 Ms. Asamanetch Fantaye, S U 1998 Operations Analyst July 6 - 24, 1998 4 Ms. Judy Makanda, Mission U U Leader, Financial Analyst Mr. Alan Carroll, Senior Urban Specialist Mr. John Kandulu, Infrastructure Specialist Mr. Chris Banes, Municipal Engineer (Consultant) January 18 - 3 Ms. Judy Makanda, Mission S S February 1, 1999 Leader, Financial Analyst Mr. Chris Banes, Municipal Engineer (Consultant) Mr. Francis M'buka July 7 - 16, 1999 4 Mr. Alan Carroll, Task Team S S Leader Mr. Lance Morrell, Senior Financial Analyst Mr. Jonadab Metibaiye, Civil Engineer Mr. Donald Mphande, Financial Management Specialist February 8 - 23, 5 Mr. Lance Morrell, Senior S S 2000 Financial Analyst Ms. Colleen Butcher, Local Government Specialist Mr. Jonadab Metibaiye, Civil Engineer Mr. Donald Mphande, Financial Management Specialist Mr. Francis M'buka, Responsible for Decentralization September 4 - 16, 5 Ms. Colleen Butcher, Senior S S 2000 Urban Specialist Mr. Jonadab Metibaiye, Civil Engineer Mr. Alan Carroll, Task Team Leader Mr. Francis M'buka, Responsible for C'entralization Mr. Donald Mphande, Financial Management Specialist -43 - March 26 - April 6, 6 Mr. Alan Carroll, Mission S S 2001 Leader Ms. Colleen Butcher (Senior Urban Specialist) Mr. Jonadab Metibaiye, Civil Engineer Ms. Christine Kimes, Deputy Resident Representative/Senior Operations Officer Mr. Francis M'buka, Senior Program Officer Mr. Donald Mphande, Financial Specialist ICR August 24 - 3 Colleen Butcher, Senior S S September 7, Urban Specialist and task 2001 manager Mr. Jonadab Metibaiye, Engineering Specialist Donald Mphande, Financial Specialist Note: IP and DO ratings are indicated for the period after this rating system was introduced (b) Staff Stage of Project Cycle Actual/Latest Estimate No. Staff weeks US$ ('000) Identification/Preparation 72.1 148.8 Appraisal/Negotiation 52.4 100.3 Supervision 208.5 450.7 ICR 5.6 8.7 Total 338.6 708.5 * * Dollar equivalent of staff weeks during each phase of the project cycle, excluding travel, fiduciary, trust fundsand including consultants funded under BB. -44 - Annex 5. Ratings for Achievement of Objectives/Outputs of Components (H-High, SU=Substantial, M=NModest, N=Negligible, NA=Not Applicable) Rating b Macro policies O H OSUOM O N * NA l Sector Policies O H O SU * M O N O NA O Physical O H *SUOM O N O NA O Financial O H OSUOM * N O NA b Institutional Development 0 H O SU * M 0 N 0 NA El Environmental O H OSUOM O N * NA Social E Poverty Reduction O H OSUOM O N * NA LE Gender OH OSUOM ON *NA El Other (Please specify) O H OSUOM O N * NA El Private sector development 0 H 0 SU 0 M 0 N * NA LI Public sector management 0 H 0 SU * M 0 N 0 NA Li Other (Please specify) O H OSUOM O N * NA -45 - Annex 6. Ratings of Bank and Borrower Performance (HS=Highly Satisfactory, S=Satisfactory, U=Unsatisfactory, HU=Highly Unsatisfactory) 6.1 Bank performance Rating E Lending OHS OS *U OHU O Supervision OHS OS *U OHU El Overall O HS O S * U O HU 6.2 Borrowerperformance Rating E Preparation O HS * S O U O HU O Government implementation performance 0 HS 0 S 0 U 0 HU O Implementation agency performance 0 HS 0 S 0 U 0 HU O Overall O HS O S * U O HU - 46 - Annex 7. List of Supporting Documents 1. Malawi Local Government Development Project, Credit No. C2379 - MAI, Development Credit Agreement 2. Staff Appraisal Report, Malawi Local Government Development Project, May 12, 1992 3. An Assessment of the Development Fund for Local Authorities and Proposals for Options Regarding its Future", World Bank, June 2001. 4. Audit Report of LGDP Project Accounts, FY1993/94 to Fyl999/2000, Deloitte and Touche, Malawi, October 2001. 5. "Decentralization Process in Malawi: A Technical Cooperation Programme 2001 - 2004", Government of Malawi, July 2001. 6. "Report on Donor Roundtable Conference on Decentralization", Government of Malawi, August 2001. 7. "Malawi Intergovernmental Fiscal Transfers Study", Georgia State University, April 2001. 8. "Improving Revenue Mobilization in Malawi: Study on Property Rates and Business Licensing", Kennedy School of Government, Harvard University, April 2001. 9. "Appraisal of Malawi's Decentralization Implementation Plan", Joint Donor/GOM Appraisal Team, August 2000 and October 2001. - 47 - MAP SECTION IBRD 23560 32 T. Dar Es 34 36 C&½ 0w -A 1D5 %i:') 10 t \s * oN.^a* \ vMALAWI This snap has boon -O MALAWI - byctr deTh Worltd Baheonff3niehr3 t 9 ~s0 LOCAL GOVERNMENT internaupse.ofnThe WorldoBnk 0'ln 5sn.s DEVELOPMENT PROJECT and the bounddries shown ) -, ;porhis of TdWordSo1nkGfo ), RL bl; on)ugen nte o @ XKt,s ° stoftot an"y torrituY or ony o Jr endoarsment or acceptncr -oZ sucb booundones. , > s1h NOdTr2rTVN ) O PROJECr SITES / Eks4 ... ll ' \_ i PRINCIPSL ROADS JF0 T\C.VNS 0 W <>> OSL lu;,or. * DISTRiCT CAPITALS Mzirr-N. N i )1 / NSONL CAPITAL 12 L^X.Ic3. Chinumuo DISTRICT BOUNDARIES 12- 12- ",CO1 ZAMBIA - - Lkwa I REGION BOUNDARIES i_} > t INTiERN;TIONAL BOUNDARIES I , . | ILGr.5 Fb :.~ 40 6 0 oO 100 . n~ O sa..c t r'Ž *%o , knlIa h M*I.t, .!. .;. 40 60 t a '- SNtn FA irri NAIC)ZNAMBI(B LIE To Luska EGION - D;cc .14 0.kurX j \\ / o \ 14 Es N-1b < - < -- ~~~~~* MarU.gxt. .11 ski3>, MOZAMBIQL'E E OL.u,s.. R WNA *KWA S Z issuE / XToTele A -'I T- < ,.b> C i +hrs |16' Chikn 16 To Beira N'- _ / rTH) J <34 36 April 1992