Madagascar - Antananarivo Development Project

Document of The World Bank FOR OFFICIAL USE ONLY Report No: 20108 IMPLEMENTATION COMPLETION REPORT (CREDIT 21170) ONA CREDIT IN THE AMOUNT OF US$ 30.5 MILLION TO THE REPUBLIC OF MADAGASCAR FOR THE ANTANANARIVO PLAIN DEVELOPMENT PROJECT March 14, 2000 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 30 November 1999) Currency Unit = Malagasy Franc (FMG) FMG 1= US$ .00016 US$ 1 = 6260 FMG FISCAL YEAR January 1 - December 31 ABBREVIATIONS AND ACRONYMS AFD Agence Francaise de Developpement APIPA Agence de Gestion des Ouvrages de la Plaine d'Antananarivo (Flood Control Authroity) BPPA Bureau du Projet de la Plaine d'Antananarivo FIVREN Fivondronano (Municipality) of Antananarivo ICR Implementation Completion Report JIRAMA National Water Supply and Electricity Company OED Operations Evaluation Department PPF Project Preparation Facility SAMVA Service Autonome de Maintenance de La Ville d'Antananarivo (Drainage and Solid Waste Authority) SAR Staff Appraisal Report Vice President: Callisto Madavo, AFRVP Country Director: Michael Sarris, AFC08 Sector Manager: Jeffrey S. Racki, AFTUl Task Team Leader: Arnaud Guinard, AFTUI  FOR OFFICIAL USE ONLY Republic of Madagascar Antananarivo Plain Development Project Implementation Completion Report 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 4 5. Major Factors Affecting Implementation and Outcome 6 6. Sustainability 7 7. Bank and Borrower Performance 8 8. Lessons Learned 9 9. Partner Comments 10 10. Additional Information 10 Annex 1. Key Performance Indicators/Log Frame Matrix 11 Annex 2. Project Costs and Financing 12 Annex 3. Economic Costs and Benefits 14 Annex 4. Bank Inputs 15 Annex 5. Ratings for Achievement of Objectives/Outputs of Components 16 Annex 6. Ratings of Bank and Borrower Performance 17 Annex 7. List of Supporting Documents 18 Annex 8. Borrower Comments on the ICR 19 Annex 9. Completion Report by the Borrower 20 Annex 10. Comments from AFD 31 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: P001512 Project Name: Antananarivo Plain Development Team Leader: Arnaud Marie Guinard TI, Unit: LCSFU ICR Tvpe: Core ICR Report Date: MNarch 14, 2000 1. Project Data NVame: Antananarivo Plain Development L,'C Aumber: 21170 Country/Department: MADAGASCAR Region: Africa Regional Office Sectorlsubsector: WS - Sewerage KEY DATES Original Revised/Actual PCD: 01/06/87 Effective: 01/31/91 01/31/91 Appraisal: 06/23/89 MTR: Approval: 03/29/90 Closing: 06/30/97 09/30/99 Borrower.Implementing Agency: Government of Madagascar/BPPA Other Partners: Agence Francaise de Developpment (AFD) STAFF Current At Appraisal 17ice President: Callisto Madavo Edward V.K. Jaycox Country Manager: Michael N. Sarris Sector1fanager: Jeffrey S. Racki Marc Blanc Team Leader at ICR: Arnaud Marie Guinard Gerard Tenaille ICR Primary Author: James Wright 2. Principal Performance Ratings (HS=Highly Satisfactory, S=Satisfactory, U=Unsatisfactory, HL=Highly Likely, L=Likely, UN=Unlikely, HUN=Highly Unlikely, HU=Hlighly Unsatisfactory, H=Hligh, SU=Substantial, M=Modest, N=Negligible) Outcome: S Sustainability: UN Institutional Development N Impact: Bank Performance: U Borrower Performance: S QAG (if available) ICR Quality at Entry: U Project at Risk- at Any Time: 3. Assessment of Development Objective and Design, and of Quality at Entry 3.1 Original Objective: As stated in the SAR, the main objective of the project was to improve the living conditions of the low-income population in the low areas of the city of Antananarivo and in the Antananarivo Plain. Specifically, the project would (i) create a favorable environment for private investment and economic growth in the metropolitan zone of Antananarivo; (ii) consolidate the financial and administrative  development of the municipalities which was begun under the Antananarivo Urban Development Project (1 987-MAG); (iii) establish an institutional framework to ensure the efficient operation and maintenance of works and installations financed under the project; and (iv) ensure integrated rural and urban development in the Antananarivo Plain. The Antananarivo Plain has a large concentration of the region's low income population and is the only large area available for the expansion of low income settlements and employment in the Antananarivo area. Poverty alleviation, private sector development, economic growth, municipal development and integrated rural/urban development in this area were relevant to the Bank's assistance strategy and responsive to the Borrower's concerns. However, these were very broad objectives involving different sectors, levels of govemment and institutions. This inherent complexity hampered implementation throughout the project. The objective of establishing efficient operation and maintenance of new installations was dependent on new agencies being created under the project and has proved more risky than expected (see section 3.4) 3.2 Revised Objective: The above set of objectives were broad and reflected wider goals for development in Madagascar at the time. They were not changed during project period, but, given their breadth, the main project components listed below in fact came to be used as development achievements during project implementation. 3.3 Original Components: The main components of the project were: (i) to provide 100-year flood protection through rehabilitation and construction of dikes along the Ikopa and Mamba Rivers and a flood warning system, including sites for relocating families from areas affected by these works; (ii) rehabilitation and construction of drainage canals on the Antananarivo Plain and construction of a pumping station to remove excess water from the plain into the Mamba River; (iii) irrigation improvements, including construction of an irrigation canal and rehabilitation of the drainage and irrigation network for 2,910 hectares in the Antananarivo Plain; (iv) rehabilitation and extension of the sewerage system in Antananarivo and rehabilitation of the existing solid waste composting plant; (v) improvements in two out of five drainage areas (drainage, storage basins and primary roads) in the southern part of the plain to accommodate urban development; and (vi) institutional strengthening including updating and improving the management of the cadastre, improving the land tax system, creating an institution to manage hydraulic works in the Plain, creating a user association of the irrigation system, and assisting public agencies responsible for sewerage and solid waste management. The original components were effectively designed to implement flood protection, drainage and irrigation as well as storage basins and road improvements for future urbanization in the Plain, and implement sewerage and solid waste improvements. These tasks proved to be largely within the capacity of the implementing agency. The proposed improvements in the cadastre were not directly related to the cost recovery needs of the project. Modest achievements in the land tax system could be achieved. However, the institutional arrangements to operate and maintain the main investments of the project (APIPA for flood control and drainage, SAMVA for sewerage and solid waste, and user associations for irrigation), which were of critical importance to sustain the investments, were not defined until long after appraisal and proved much more difficult to achieve than originally envisioned. -2-  3.4 Revised Cornponents: There were limited changes in some components which did not imply changes in objectives: (i) an additional 4 km. of dikes were added to the flood control component; (ii) the scope of the sewage component remained the same but after completion of a feasibility study, it was changed from new investment to rehabilitation; (iii) the cleaning out of Anosy Lake, which plays an important role in the drainage system of the central city, was added to the project in 1998; and (iv) at the request of the Municipality of Antananarivo, the component to rehabilitate the solid waste composting plant was modified in 1995 and the funds were reallocated to the reconstruction of 289 garbage containers and provision of 135 metal containers. The Credit Agreement was amended in 1996, due to a suspension in AFD financing, to finance additional drainage works (C3 canal north) and the flood warning system originally to be financed by AFD. In 1999, the flood waming system was transferred back to AFD so that the IDA credit could be closed without further delay. 3.5 Quality at Entry: The project was consistent with Govemment priorities and Bank country strategy. Improvements in urban development, private investment, agriculture and poverty alleviation in the Antananarivo metropolitan area were complementary to the structural adjustment process underway at that time. However, it was not advisable to address all of these issues in one project, especially given the disappointing results from a previous urban development project (Credit 1497-MAG) and in the absence of subsequent analysis. The project tried to address both urban and rural development issues with a large and diverse number of components involving complex engineering works, urban management issues and rural development/irrigation problems. The cadastre component, which was not directly related to the project objectives, could have been financed from another source. The technical aspects of the project were well defined through a detailed analysis financed under an engineering credit (Credit 1086-MAG) which lasted from 1981 to 1986. The engineering studies examined a number of alternatives to identify a technically and economically justifiable project for flood control and agricultural development in the Antananarivo Plain. Further studies under an ongoing Bank-assisted Antananarivo Water and Sanitation Project (Credit 1002-MAG) recommended a pumping station in the northwest corner of the protected zone to drain the flood waters into the Mamba River. In December 1986, a PPF was initiated to undertake additional economic and financial studies, including the design of urbanization schemes and local finance and institutional arrangements. In contrast to the technical aspects, the institutional arrangements were not well defined in advance and the risk to long-term financial viability was underestimated. Indeed, much of the institutional framework was left to be detailed at a future date. Experience under the Antananarivo Urban Development Project (Credit 1497-MAG) had not been positive in improving local government management and finance. Therefore, the project recommended the establishment of separate authorities (later combined into SAMVA) to manage sewerage and solid waste. For both components, operation and maintenance were required to be contracted to the private sector as a condition of disbursements; there was an additional disbursement condition that a tax on water be enacted to pay for sewerage. SAMVA's operationshave been contracted to the private sector and taxes have been established, but despite its separate status, it has not been insulated from the management and financial problems of the municipality. The project alsocalled for the establishment by December 1992 of an independent authority (APIPA) to manage flood control on the Plain. APIPA, has a broader scope of responsibilities (functional and regional), and its separate status was more appropriate. However, like SAMVA, its separate status has not been sufficient to ensure that the investments made - 3 -  under the project would be sustained. The user associations established to maintain irrigation systems remain untested. The promulgation of measures to control urbanization in the southem part of the Plain was a condition of disbursement for the flood control component, but the municipality was not sufficiently committed to this objective. Uncontrolled urbanization has continued which is especially serious in the retention pond areas necessary to create equilibrium during flooding. The project required evictions of 2,341 households, 80% of whom were squatters, about twice the number estimated during preparation. The details of the resettlement program were not folly elaborated during project preparation. Local government was not fully consulted and involved in resettlement, leaving BPPA in charge of the process. This led to long delays during implementation and poor results in resettlement. 4. Achievement of Objective and Outputs 4.1 Outcome/achievenment of objective: The living conditions of the low income population in the Antananarivo Plain have been improved. The low areas of Antananarivo are protected from 100-year floods and the infrastructure is in place to drain the Plain from rain flows. Private investment, albeit anarchic, is taking place. Less fill is necessary to safeguard plots. These improvements are reflected in increased land values. An estimated total of 24,250 man months of work were created during the last two years of the project when most works were finished. However, the institutions now in charge of operation and maintenance (APIPA for flood control; SAMVA for sewerage and solid waste) are not well anchored into government structures and their sustainability is questionable. Due to delays in the AFD-financed irrigation component, the thirteen user associations established to operate and maintain the irrigation system are not fully developed. There has been little financial and administrative development of the municipalities beyond limited improvements in the potential municipal real estate tax. The relevant central and local govemment departments have not proven their ability to assure integrated rural and urban development of the Plain and designated retention pond areas are not being adequately protected from development. Nevertheless, given the successful completion of the project works under difficult circumstances, the overall outcome is rated marginally satisfactory. 4.2 Outputs by components: Physical Objectives Most of the physical objectives have been met. This successful completion of such a complex set of works is an important achievement. Dikes have been rehabilitated along the Mamba and Ikopa Rivers. Drainage canals in the Antananarivo Plain have been constructed and rehabilitated and a major pumping station has been put in place to extract water from the protected area. Sewerage rehabilitation has improved sanitation in a previously urbanized 67 hectare area, although sewage is discharged without treatment into the Ikopa River. New solid waste receptacles have been rehabilitated and solid waste management is now managed under private contract. Lake Anosy has been cleaned out. A new irrigation canal has reduced the pollution of rice paddies and the irrigation network in the north of the Plain will soon be improved with AFD financing. A flood warning system will also soon be put in place with AFD financing. Involuntarv Resettlement. While the project has improved the living conditions of many poor households in the Antananarivo Plain, many households subject to resettlement are in worse conditions than before. The original plan to give each family a resettlement plot in addition to compensation was changed to one of reduced compensation, temporary shelter and an option to buy a plot in three designated areas far from most employment opportunities. In 1998, only three of the 2,341 households displaced were living in the -4 -  three designated resettlement sites. Most households used their compensation to settle nearer their original homes and many suffered a deterioration in tenure status and housing quality. More than 60% of former owners and tenants in a survey on resettled households were not satisfied with resettlement. The resettlement process did not provide assistance with the move, support at the resettlement site or assistance with improving living standards as would now be required (appraisal was one year before OD 4.30 on involuntary resettlement was issued). Financial Objectives. The institutions established for operation and maintenance of works financed under the project are not yet financially viable. APIPA is authorized to recover: (a) a fee for new landfills, but for 1998 it recovered only FMG 2.3 billion of the 7 billion expected based on landfill authorizations to date (part is overdue payments; some expected landfills have not yet been made); and (b) a 5% surtax, based on updated property values (see section 4.5) on properties benefiting from flood control, to be collected by the city and transferred to APIPA. This annual surtax is not yet being collected, but the first tax roles should be issued the end of 1999. With the present rate of filling and with a tax surcharge rate of 5%, it would take five to seven years for APIPA to be self sufficient. With a lower surtax rate of 3%, which was suggested bythe local authorities, APIPA would not cover its costs until 2010. These projections assume that some depreciation and maintenance costs will be subsidized from presently unidentified Government sources. The long-term operation and maintenance prospects for the complex flood waming system are particularly poor. SAMVA is supported by a special solid waste management tax on built-up property which is a percentage of the property tax, but the tax rate depends each year on the municipal council. Solid waste removal fees were insufficient to cover SAMVA's cost in 1998. For 1999, FMG 4 billion have been collected and transferred to SAMVA, compared to the FMG 7 billion necessary to cover full costs, possibly leading to curtailment of service. The management of the sewerage system is supposed to be covered by a surcharge on water fees collected by JIRAMA, currently 10% of the amount due for water consumption. Present transfers from JIRAMA enable SAMVA to maintain only 15 km. of the 47 km. network. Even if the tax were raised to 15% in 2000, it would not be able to carry out the necessary maintenance. The problems are likely be more severe than reflected here. Much of APIPA's infrastructure maintenance is still covered under works and equipment contracts. Both agencies infrastructure is new. T'he cost of deferred maintenance will not be apparent for several years. 4.3 Net Present Value,/Economic rate of return: The economic rate of return of the project calculated at appraisal was 18%. At completion, the irrigation component was incomplete, and the benefits of the sewage, solid waste management and urban finance and management components could not be quantified. A rate of return of 17% was calculated for the flood protection and drainage components which account for 49% of project cost, confirming the substantial economic return of protecting the 635 hectare area of the southern plain (see Annex 3). 4.4 Financial rate of return: Not applicable. 4.5 Institutional development impact: The objective of consolidating the financial and administrative capacity of the municipalities in the area covered by the project was achieved in part through the improvement of the tax base (property ownership study, updating property values and taxes, and assessing municipal resource and expenditure). This will have a city-wide impact on tax receipts, including revenues for SAMVA and APIPA which are based in part on property values. However, one recommendation of the study was to increase the tax rate on higher - 5 -  valued properties. This has proven difficult to implement. The cadastre has been updated based on aerial photography and digitized maps, but this work is having a lower impact. Only 10% ofthe property records have been entered into the new system, and FMG 2.5 billion would still be required to complete this work. The separate para-public entities created to operate and maintain the investments made under the project were slow to develop and their future is uncertain. Laws establishing SAMfVA and APIPA were not enacted until 1995 and the taxes and fees necessary for operation and maintenance were not established until 1996. Disbursements on works contracts could not begin until management contracts were signed by SAMVA with private operators and the taxes and fees were fully agreed to by the government. These actions did not occur until 1998. SAMVA lacks the technical capacity to carry out its functions and both agencies lack financial capacity. The creation of SAMVA as a separate entity to manage solid waste and sewerage was not sufficient to ensure continuity of service. A strong financial base would be essential for this, but that would have required more effort to build the financial base of the municipality. 5. Major Factors Affecting Implementation and Outcome 5.1 Factors outside the control of government or implementing agency: Not applicable. 5.2 Factors generally subject to government control: * AFD funding was suspended from 1994 to 1997 due to arrears of government with AFD. This stopped AFD disbursements and work on the main pumping station, and IDA-financed drainage works could not proceed. Eventually a special arrangement was made with the French Government to finance the pumping station. The drainage works which the French Ministry of Cooperation could not finance were finished with a 2 1/2 year delay with IDA financing. * The process of land expropriation was stopped several times during elections and due to delays in funding from the government. The entire eviction/reallocation process lasted six years from January 1993 and December 1998. Relocation of households affected by flood control work were not managed by FIRVEN (the local government) in accordance with its convention with BPPA. This led to long delays in the construction of dikes and canals. * There were some early delays in counterpart funding from Government. The problem was more severe at the end of the project because about 70% of the project had to be implemented after 1997 when work on the pumping station could resume. During this period, large amounts of counterpart funds were required and budget provisions were made. However, the funds were not disbursed to the project in a timely manner. There were also some delays in the signing of contracts. 5.3 Faclors generally subject lo implementing agency control: * The accounting system of BPPA involved complexities of multiple donors and components. The accounting system put in place in 1993 was insufficient and led to delays in producing financial reports acceptable to the donors. An improved version was put in place in 1997 and has proven to be satisfactory. 5.4 Costs andfinancing: The original cost of the project was US$ 68.6 million, and the final cost is US$ 63.3 million. The original - 6 -  credit amount was US$ 30.5 million (SDR 23 million). When AFD financing was resumed in 1997, the closing date of the IDA credit was extended from June 30, 1997 to June 31, 1999. A further three-month extension was made to September 30, 1999 to complete the execution of the remaining works. Because of the suspension, IDA agreed to finance a much larger percentage of drainage canals and it reduced its participation in the pumping station and in irrigation. At the time of suspension, IDA also agreed to finance the flood warning system, but this was transferred back to AFD in 1999 so that the IDA credit could close without fiurther delay. At that time, SDR 5 million were canceled from the project leaving the final amount of IDA financing at SDR 18 million (US$ 23.4 million equivalent). 6. Sustainability 6.1 Rationale for sustainability rating: BPPA provided strong leadership during implementation, but its future role is still unclear following the closing of AFD financing expected in December 2000. A study is currently being undertaken to examine future options for BPPA. The ability of APIPA and SAMVA to operate and maintain the works is not certain. These are new agencies with serious financial problems (see section 4.2). The municipality must approve the fees they collect and subsidize costs they cannot cover, but local govermnent finances are also constrained. The inability of central and local govemment to control development is also a serious issue (see section 3.2). User associations were set up to manage irrigation but are untested. Given the delay in implementing the irrigation component, they would need to be revitalized when the irrigation systems are operational. 6.2 Transition arrangement to regular operations: SAMVA and APIPA are supervising their respective components jointly with BPPA during the final stages of the project. There has been a provisional handover of all components except some of the retention basins and the C3 drainage canal. SAMVA and APIPA have sufficient fumding to pay their staff but funding of contracts for private operating contractors is less secure. Some payments to contractors by SAMVA are already in arrears. Relevant performance indicators: SAMVA a) Amount of sewage fee collected (% of JIRAMA water fee); % of costs covered b) Amount of solid waste fee collected c) Continuing operation of sewage pumping stations d) Amount of solid waste collected; coverage of city e) Contractors for solid waste collection and sewerage O&M being paid regularly f) Contractors meeting technical requirements of their contracts APIPA a) Amount of landfill fees and flood protection surtax collected; % of costs covered b) No breaches/erosion on dikes c) Dikes protected from squatters and retention basins protected from illegal landfills d) Private O&M contractors being paid regularly e) Contractors meeting technical requirements of their contracts f) Continuing operational condition of pumping stations IRRIGATION a) User associations managing and maintaining systems effectively - 7 -  Recommended follow-up by the Bank Sector work has been initiated by the Bank to take a program approach to the urban problems in -Madagascar, especially those of the Capital region. This approach should lead to more focused and better phased project interventions with donors expected to finance discrete parts of a complementary program. An OED impact evaluation is recommended after two to three years of implementation experience and after the irrigation components are complete, especially to assess the institutional development and sustainability aspects of the project. 7. Bank and Borrower Performance Bank 7.1 Lending: The project was very complex involving too many components, both the urban and rural sectors, and difficult engineering and urban development issues (see section 3.4). For the engineering aspects, there was a thorough weighing of technical alternatives, and strong technical implementation arrangements put in place. However, the institutional arrangements were to be defined during implementation. The resources required to finance operation and management were not assessed or agreed upon, and the lessons from the First Urban Project, especially the need for local government development, were not taken into account. 7.2 Supervision: There were 19 supervision missions during the ten-year implementation period. These were characterized by a good skill mix, almost always with an engineer or urban planner and a financial analyst. There were only four team leaders during the long supervision/ICR period. Nine missions included a municipal financial specialist to assess institutional/financial arrangements and four missions included sociologists to assess the resettlement component. The Bank's performance on supervision is rated marginally satisfactory given the strong technical support from Bank missions enabling the implementation of a complex set of engineering works under difficult circumstances. The redirection ofsewerage works from new investment to rehabilitation also demonstrated the Bank's flexibility in a changing environment. However, when faced with a critical 2 1/2 year suspension by AFD, the Bank did not take the opportunity to reflect fully on the project and consider possible restructuring. There was no mid-term or other systematic review of project objectives which continued to be judged satisfactory despite the suspension and the lack of progress on institutional issues. 7.3 Overall Bank performance: The technical design and supervision of the project was strong. However, the lack of initial sector analysis and stock taking of previous project experience, the complexity of the project with very broad objectives and the inadequate preparation of the institutional aspects were serious problems which could have been mitigated or avoided. The Bank's response to the 2 1/2 year suspension by AFD was insufficient. Therefore, overall Bank performance is rated unsatisfactory. Borrower 7.4 Preparation: The technical aspects of the project were well prepared by the Government. The Government established the BPPA during project preparation to manage implementation. Under the engineering loan and the subsequent PPF, options were evaluated and project proposals were prepared with consultants with - 8-  minimum delays. 7.5 Governnment implementation performance: Support from Government was irregular. There were many changes in govemment which adversely affected implementation. The inter-ministerial steering committee in charge of coordinating the project and addressing key institutional and financial issues often failed to meet. This led todelays in the promulgation of laws and creation of APIPA and SAMVA. There were also some delays in the provision of counterpart funds and the signing of contracts, and the project was delayed for 2 1/2 years when the Govemment was in arrears to AFD. 7.6 Implementing Agency: The agency generally managed well. It overcome early accounting problems which were due to the complexity of the project (multiple donors, many components and institutions). The 1993 audit report of BPPA indicated irregularities in accounting and inadequate control, and a committee was established to examine this. As a result, errors in three reimbursement claims were rectified in 1994 and improvements recommended by the auditors were introduced. Progress reports were submitted in a timely manner. The BPPA managed well the large procurement and supervision burden at the end of the project. 7.7 Overall Borrower performance: On balance, the Borrower's performance is judged generally satisfactory, especially given the complex set of works which were well executed under difficult circumstances, with renewed efforts during the last two years of the project after AFD resumed financing. 8. Lessons Learned v The full range of complex problems in large metropolitan areas need to be carefully assessed and are unlikely to be effectively addressed with a single all-encompassing project. It would have been preferable to undertake a review of issues and opportunities in the region in order to define a set of priorities. Focused individual and more manageable projects with clear objectives would then be identified to meet the priorities. Thus, when faced with a range of issues such as flood control, agricultural development and urban development, a set of more focused, complementary projects could achieve better results within a broader development program for the region. * When possible, cofinancing should be targeted to discrete projects of a larger program or to discrete, physically unlinked components of a single project to avoid interruption when problems occur with one cofinancier. * Flood control projects of this magnitude require support and funding from national budgets as well as from local sources to make sure the system will operate over the long term. Sustained commitment is needed from high levels of government. Occasional coordinating committee meetings are not enough. * When projects are substantially delayed, the Bank should review the project objectives to assess whether restructuring is advisable. * The definition of the institutional and resource mobilization arrangements to ensure sustainability after implementation should not be postponed until long after appraisal. Alternatives should be carefully assessed, and an explicit plan should be agreed upon at appraisal. -9-  * Most types of municipal functions cannot be successfully quarantined from municipal govemments. The development of sound municipal govemment, particularly its revenue base, is critical for sustainability. * Resettlement components should be planned with clear objectives and in more detail during project preparation to avoid delays. The capacity and will of local government to implement agreed resettlement plans should be carefully assessed. This is especially important if living standards and income eaming capacity are to be preserved. 9. Partner Comments (a) Borrowerlinmplementing agency: Comments were received from BPPA on the draft ICR These comments were limited to a few technical details in the text which were taken into consideration in the final report. The Borrower has indicated in letter dated February 22, 2000 that it has no further comments on the report (see Annex 8). (b,) Cofinanciers: (c) Other partners (NGOs/private sector): See Annex 10 for comments received from AFD. 10. Additional Information - 10-  Annex 1. Key Performance Indicators/Log Frame Matrix Outcome I Impact Indicators: A. FLOOD PROTECTION 1) High waters do not flood the plain 1) flood control system in place 2) High waters alert system installed 2) Not yet; to be financed by AFD B). DRAINAGE 1) Antananarivo canal rehabilitated 1) canal rehabilitated 2) Pumping station in operation 2) pumping station in operation 3) Drainage canal C3 built 3) C3 built and operating C) URBAN INFRASTRUCTURE 1) Priority sewers rehabiltated 1) Sewers and pumping stations rehabilitated 2) Priority South Plain infrastructure in place 2) Infrastructure in place D) AGRICULTURE DEVELOPMENT 1) Irrigation Canal GR built 1) Not yet completed. Being finance by AFD 2) 3,000 ha. of irrigation rehabilitated. 2) Not yet completed. Being financed by AFD E) INSTITUTIONAL DEVELOPMENT 1) 90,000 plots of cadastre updated 1)12% updated 2) Property valuation survey carded out 2) survey carried out 3) Flood protection agency established 3) established; not financially viable 4) Sanitation agency established 4) established; not yet technically or financially viable. Output Indicators: (Output indicators were not stipulated when the project was retrofitted with performance indicators). End of project - 11 -  Annex 2. Project Costs and Financing Project Cost by Component (in US$ million equivalent) Relocation 1.20 0.55 1.8 Dikes 4.00 5.00 5.8 Flood warning system 0.90 4.04 1.3 Drainage canals 3.60 5.27 5.3 Pumping station 2.60 3.34 3.8 Irrigation canal and system rehabilitatiton 10.40 18.08 15.1 Sewerage and solid waste 8.90 8.79 12.9 Southern Plain drainage 3.20 3.88 4.7 Land acquisition 3.50 2.75 5.1 Cadastre and real estate tax 4.60 2.73 6.7 Equipment 2.50 0.72 .3.6 Studies, supervision, project office, PPF and TA 7.10 8.13 10.3 Total Baseline Cost 52.50 63.28 Physical Contingencies 5.60 8.2 Price Contingencies 10.50 15.3 Total Project Costs 68.60 63.28 Total Financing Required 68.60 63.28 Project Costs by Procurement Arrangements (Appraisal Estimate) (US$ million equivalent) .. j iggll B ............................................... ::B'. @:B Bhet0 .... .B0F. 1. Works 34.70 8.00 0.00 0.00 42.70 (19.40) (1.50) (0.00) (0.00) (20.90) 2. Goods 5.60 4.40 0.00 0.00 10.00 (I.00) (2.00) (0.00) (0.00) (3.00) 3. Services 0.00 0.00 8.10 0.00 8.10 (TA,Training and (0.00) (0.00) (3.20) (0.00) (3.20) Professional Services) 4. Miscellaneous 0.00 0.00 0.00 0.00 0.00 (0.00) (0.00) (0.00) (0.00) (0.00) 5. Administration costs 0.00 0.00 0.00 6.40 6.40 and Land acquistition (0.00) (0.00) (0.00) (2.20) (2.20) 6. Refinancing of PPF 0.00 0.00 0.00 1.40 1.40 and AFD advance (0.00) (0.00) (0.00) (1.20) (1.20) Total 40.30 12.40 8.10 7.80 68.60 (20.40) (3.50) (3.20) (3.40) (30.50) - 12 -  Project Costs by Procurement Arrangements (Actual/Latest Estimate) (US$ million equivalent) Procremet 1tetd 1. Works 33.29 3.88 0.36 0.00 37.53 (13.58) (2.42) (0.03) (0.00) (16.03) 2. Goods 3.28 0.94 0.07 0.00 4.29 (0.37) (0.94) (0.07) (0.00) (1.38) 3. Services 7.00 1.69 2.78 0.00 11.47 (TA,Training and (4.16) (0.42) (0.74) (0.00) (5.32) Professional Services) 4. Miscellaneous 0.00 0.00 0.00 4.50 4.50 (0.00) (0.00) (0.00) (0.00) (0.00) 5. Administration costs 0.02 0.27 0.04 5.16 5.49 and Land acquistition (0.02) (0.07) (0.00) (0.62) (0.71) 6. Refinancing of PPF 0.00 0.00 0.00 0.00 0.00 and AFD advance (0.00) (0.00) (0.00) (0.00) (0.00) Total 43.59 6.78 3.25 9.66 63.28 (18.13) (3.85) (0.84) (0.62) (23.44) Figures in parenthesis are the amounts to be financed by the Bank Loan. All costs include contingencies 2' Iicludes 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 govermilent units. Project Financing by Component (in US$ million equivalent) Percentage of Appraisal Component Ap raisal Estimate Actual/Latest Estimate Bank Govt. CoF. Bank Govt. CoF. Bank Govt. CoF. Land Acquisition & 1.28 3.64 2.93 0.17 0.0 80.5 0.0 relocation Dikes 4.15 0.78 3.87 0.87 0.36 93.3 111.5 0.0 Flood Warning 0.39 0.76 0.07 1.02 2.97 0.0 261.5 390.8 Drainage canals 1.11 0.91 2.98 2.29 2.59 0.15 206.3 284.6 5.0 Pumping station 0.74 2.48 0.00 2.02 1.32 0.0 273.0 53.2 Irrigation 2.37 11.33 0.39 0.39 5.44 0.0 16.5 48.0 Sewerage & solid waste 10.55 2.21 0.32 7.52 1.58 0.00 71.3 71.5 0.0 So.Plain drainage 4.03 0.75 3.33 0.69 0.00 82.6 92.0 0.0 Cadastre and R.E. Tax 4.22 1.58 2.15 0.65 50.9 41.1 0.0 Equipment 1.59 1.12 0.56 0.31 0.18 0.20 19.5 16.1 35.7 Studies, supervision, 2.34 0.93 3.89 3.62 1.03 2.94 154.7 110.8 75.6 project office, PPF and TA - 13-  Annex 3: Economic Costs and Benefits The economic rate of return (ERR) of the project calculated at appraisal was 18%, with a lower estimate of 15% using somewhat more moderate estimates of land price increases. The appraisal used a complex methodology for estimating an ERR for most of the project. At completion, the irrigation component was incomplete and its benefits could not yet be quantified. The economic benefits of the sewage, solid waste management and urban finance and management components do not lend themselves to accurate quantification. An ex-post economic rate of return was calculated for the flood protection and drainage components which accounts for 49% of project cost. The calculation of benefits was based on increases in the US$ equivalent value of land prices in the 635 hectares of the southem plain affected by improvements in flood protection and drainage. The net amount of land usable for urbanization increased from 97 hectares to 490 hectares, and land dedicated to agriculture has decreased from 538 hectares to 97 hectares. The difference is accounted for by the 47 hectares now dedicated to infrastructure created under the project. Much of the increase in the value of urbanizable land was due to the reduced need for costly filling in the protected area. Although irrigation improvements have not yet been realized, agricultural land values have increased due to improved drainage which controls water flows and enables a wider variety of crops. The combined value of urbanizable and agricultural land in the area is estimated to have increased during the implementation from 1992 to 1999 by about eleven-fold. In such a rapidly urbanizing area, all of the increase in land value cannot be attributed to the flood protection and drainage works alone. Several assumptions were tested when assessing benefits attributable to the project against the costs (see below). The yearly capital costs in US$ net of taxes were used for the flood control and drainage works together with prorated amounts for miscellaneous and overhead costs. Together these costs amounted to 49% of the project costs. The projected O&M costs of APIPA were used as operating costs for a twenty-year period extending after implementation. With the assumption that 50% of the total increase in land values is attributable to flood protection and drainage, these investments yield a 17% ERR If 40% of land value increases are attributable to the project investments, the ERR would be only 11%. Conversely, if 60% of land value increases are attributed to the project, the ERR would be 23%. Other projects in Madagascar during the same period, such as the Social Fund II and the Antananarivo Urban Works projects, have had as a primary objective the creation of employment. While this was not a primary objective of this project, most of the works such as dredging canals and preparing retention ponds were undertaken using very labor intensive techniques. It is estimated that a total of 24,250 man-months of work were created during the last two years of project implementation when the bulk of the works were carried out. - 14 -  Annex 4. Bank Inputs (a) Missions: Stage of Project Ccle No. of Personsand Specialty Performmance (e.g.-2 Economnists 1 EMS, et) .-Implementaton :Deepment Mboiflh/Yeax C:ount............. - Specilty --.prgess Objeve v Identification/Preparation I Eng 2/86 12/86 1 Eng 2/88 2 Eng,FMS 10/88 4 Eng, UrbPlan, Econ FMS 2/89 6 2Eng. FMS,Urb Plan,Econ,Ag Appraisal/Negotiation 7/89 6 2Eng,Econ,2FMS,Ag 12/89 3 2Eng,FMS Supervision 2/91 3 2Eng,FMS 2/92 3 2Eng,FMS 7/92 2 Eng,FMS 2/93 2 Eng,FMS 3/93 2 2 Soc 7/93 3 Eng,FMS,Soc. 12/93 4 Eng,FMS,Soc,Research 6/94 3 Eng,FMS,Res S S 11/94 4 Eng,2FMS S U 5/95 3 Eng,2FMS,Research U U 10/95 4 Eng,2FMS,Research U S 6/96 4 2Eng,2FMS U S 10/96 4 Eng,Econ,FMS,Research U S 3/97 5 Eng,2FMS,Econ,Research U S ICR 12/97 3 Eng,2FMS U S 4/97 3 Planner,Eng,Soc U S 10/97 3 Planner,Eng,FMS S S 3/99 3 Planner,FMS,Research S U 10/99 (ICR) 3 Planner,Econ,FMS S U (b) Staff: Stage of Project Cycle Actual/Latest Estimate ...____ __..___ _..__..No. Staff weeks US$ t,00.0) Identification/Preparation Appraisal/Negotiation 57.20 131.2 Supervision 213.40 757.7 ICR . Total 270.60 888.9 - 15 -  Annex 5. Ratings for Achievement of Objectives/Outputs of Components (H=High, SU=Substantial, M=Modest, N=Negligible, NA=Not Applicable) Rating O Macro policies O H OSUOM O N * NA Z Sector Policies O H O SU * M O N O NA Z Physical O H * SU O M O N O NA Z Financial O H OSUOM * N O NA O Institutional Development 0 H O SU O M 0 N 0 NA Z Environmental O H OSUOM O N O NA Social O Poverty Reduction OH OSU*M ON O NA X Gender O H OSUOM O N * NA Z Other (Please specify) OH OSUOM O N O NA Resettlement Z Private sector development 0 H 0 SU 0 M 0 N 0 NA O Public sector management 0 H O SU *M 0 N 0 NA O Other (Please specify) - 16-  Annex 6. Ratings of Bank and Borrower Performance (HS=Highly Satisfactory, S=Satisfactory, U=Unsatisfactory, HIU=Highly Unsatisfactory) 6.1 Bank performance Rating