Mali - Urban Development Project

Document of The World Bank FOR OFFICIAL USE ONLY Report No. 7880 PROJECT PERFORMANCE AUDIT REPORT MALI FIRST URBAN DEVELOPMENT PROJECT (CREDIT 943-MLI) JUNE 23, 1989 Operations Evaluation Department S3 -be' by re 1 in the performance of VOR OFFICIAL USE ONLY THE WORLD BANK Washington. D.C. 20433 U.S.A. Ofce of Diuctor-Cenral Operatinm vatM June 23, 1989 MEMORANDUM TO THE EXECUTIVE DIRECTORS AND THE PRESIDENT SUBJECT: Project Performance Audit Report on Mali - First Urban Development Proiect (Credit 943-MLI) Attached, for information, is a copy of a report entitled *Project Performance Audit Report on Mali - First Urban Development Project (Credit 943-MLI)m prepared by the Operations Evaluation Department. Attachment 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. i ALV. PROJECT PERFORMANCZ AUDIT REPORT MALI FIRST URBAN DEVELOPMENT PROJECT (CREDIT 943-MLI) TABLE OF CONTENTS Paie No. Preface............................................. I Basic Data Sheet ............................................ ii Evaluation Summary................................................. v PROJECT PERFORMANCE AUDIT MEMORANDUM I. SECTOR BACKGROUND .............................. II. THE PROJECT ................................... ........... 3 A. Project Preparation .................................... 3 B. Project Objectives .................................... 5 C. Project Description .................................... 7 III. PROJECT IMIPLEMENTATION AND ACHIEVEMENTS ................... 8 A. Time Delays and Overruns ..................... 8 B. Revisions in Project Scope ......................... 9 C. Project Achievements and Continuity..................... 10 1. Civil Works ..................... .. ........ 10 2. Institutional Develoament .......................... 14 3. Resource Mobilization ............... ............. 15 4. Technical Assistance ...........*........................1 IV. POINTS OF INTEREST .. .. ............................. 19 A. Physical Impact ...........................................19 B. Institutional Impact ................ ........... 20 C. Policy Impact............................................ 21 1. Urban Development........................... 21 2. Revenue obilization ......................... 22 3. Real Estate and Property Codes.........................22 D. Poverty Impact ........ ......................... 22 E. Other Points of Interest ......................... 24 V. ROLE OF THE BANK ................................ . 26 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. TABLE OF CONTENTS (continued) Paxe No. VI. LESSONS TO BE LEARNED ....... . ......... 27 A. Implications for Project Design ........ . 27 B. Implications for Physical Works .. ................. 28 C. Implications for Revenue Generating Programs 28 D. Implications for Urban Development .................... 28 E. Tmplications for Municipal Strengthening ............... 29 F. Implications for Urban/National Economic Development Strategies ............ ...... 29 VII. CONCLUSIONS .................. . .................. 30 ANNEXES 1 - Comparison of Initial and Final Expenditures.......... 31 2 - Photographs............. ..... ... ........ ... 32 PROJECT COMPLETION REPORT I. Introduction ............. .................. ...... 36 II. Project Identification, Preparation and Appraisal... 37 III. Project Implementation....... ... ............ .......... 39 IV. Project Implementation Performance............... ...... 41 V. Institutional, Financial and Economic Performance.......... 49 VI. The Role of the Bank............................... 52 VII. Conclusions ................ ...... .................... 52 ANNEXES MAPS PROJECT PERFORMANCE AUDIT REPORT MALI FIRST URBAN DEVELOPMENT PROJECT (CREDIT 943-MLI) PREFACE 1. This is a Project Performance Audit Report (PPAR) on the Mali First Urban Development Program. Approved on June 26, 1979, the project was supported by Credit 943-MLI for US$15.3 million. 2. The PPAR consists of: (i) an Evaluation Summary and a Project Performance Audit Memorandum (PPAM), written by the Operations Evaluation Department; and (ii) a Project Completion Report (PCR), drafted by the staff of the Infrastructure Operations Division of the Sahalien Department (AF5IN). 3. The PCR provides a short account of project experience, including the achievements and failures, wittout elaborating on the reasons for these achievements and failures. To givc a wider perspective, OED has attempted to provide additional information about the history of the project, its eventual outcome, the issues raised, the problems of implementation and possible alternative approaches. The memorandum is based on the PCR, on the completion report prepared by the project unit in Bamako, on a project completion report and other documents provided by Groupe Huit, consultants, on the SAR, on project files, legal documents and other relevant material. It also drew from interviews with Bank staff members, with representatives of the Malian government, and the urban project team as well as with consultants who worked on the project. 4. The Audit agrees in most respects with the PCR. It does, however, stress the difficulties involved in implementation and evaluation of an exceedingly complex project. 5. Following standard OED procedures, copies of the draft PPAR were sent to the Borrower for coments. No comments were, however, received. - 11~ - PROJECT PERFORMANCE AUDIT REPGRT MALI FIRST URBAN DEVELOPMENT PROJECT (CREDIT 943-MLI) BASIC DATA SHEET EY PROJECT DATA Appraisal Estimate Actual or Current Estimte ---------------------------------------------------------------------- US$ Millions US$ Millions Total Project Cost 15.3 13.8 Credit Amount 12.0 12.0 Disbursed Amount 12.0 Date Physical Components Completed 3016183 06/30186 Proportion Completed by above Date 702 1002 Proportion of Time Overrun Economic Rate of Return 222 172 CUMULATIVE ESTIMATED AND ACTUAL DISBURSEMENTS (US$ Million) FY IDA 1980 1981 1982 1983 1984 1985 1986 1987 Appraisal Est. 1.2 3.8 9.1 12.0 Actual - 2.0 4.8 7.5 8.7 10.5 i1.7 12.0 Actual as Z of Estimate 0.0 52 53 62 - iii - OTHER PROJECT DATA Original Plan Revision Actual and Estimated Actual ------------------------------------------------------------------------. First Mention in Files -- 09/76 Issues Paper -- 11/13/78 Decision Memorandum -- 12/19/78 Negotiations 03115/79 05/07/79 Board Approval 04/07/79 06/26/79 Credit Agreement Date ** 11107/79 Effectiveness Date 02/15/80 06/03/80 Closing Date 12/31/83 12/31/84 . 12/31/85 12/31/86 12/30/86 Follow-up Project: Second Urban Development Project (Cr. 1677-MLI) MISSION DATA No. of No. of man Date-of Item Mo./Year Persons Days Weeks Report Identification 10/76 3 5 3 10/16/76 Preparation 1 03/77 1 8 1.6 04/07/77 Preparation 2 06/77 1 14 2.8 07/04/77 Preparation 3 11/77 1 3 0.6 12/08/77 Preparation 4 02/78 2 10 4.0 04/17/78 Preparation 5 04/78 2 8 3.2 06/01/78 Preappraisal 07/78 5 6 6.0 07/28/78 Appraisal 09/26 6 18. 21.6 11/02/78 Decision Memo 11/78 1 12/19/78 Post Appraisal 02/79 2 10 4.0 03/21/79 Supervision 1 08/79 2 8 3.2 09/25/79 Supervision 2 10179 1 4 0.8 11/16/79 Supervision 3 02/80 1 8 1.6 03/21/80 Supervision 4 02/81 2 13 5.2 03/05181 Supervision 5 06/81 2 14 5.6 08/07/81 Supervision 6 11/81 4 13 10.4 12130/81 Supervision 7 06/82 4 10 8.0 07/14/82 Supervision 8 10182 1 6 1.2 11/17/82 Supervision 9 04/83 3 10 6.0 05/20/83 Supervision 10 11/83 2 9 3.6 12/16/83 Supervision 11 02/84 2 10 4.0 03/23/84 supervision 12 07/84 2 6 2.4 08/24/84 Supervision 13 11/84 2 6 2.4 11/30/84 Supervisiun 14 03/85 3 7 4.2 04/29/85 Supervision 15 05/85 3 2 1.2 07/12/85 Supervision 16 05/86 2 9 3.6 06/15/86 - iv - COUNTRY EXCEANGE RATES Name of ,.rrency: antil 8184 Malian Franc (MF) 1/ from 8/8& CFAF Franc (CPAF) Rate Fluctuations: Years 1979 1;80 1981 1982 1983 1984 1985 1986 First Quarter 205 255 300 331 428 512 356 Second Quarter 220 274 346 374 426 469 469 353 Third Quarter 213 300 343 404 422 423 344 Fourth Quarter 216 278 356 406 450 395 329 Aanual Average 227 295 343 392 442 450 345 1980-86 Average 356 I/ Fixed Parity: 1 CEAF-- 2MF. - v - PROJECT PERFORMANCE AUDIT REPORT MALI FIRST URBAN DEVELOPMENT PROJECT (CREDIT 943-MLI) EVALUATION SUMMARY Introduction 1. The Mali first urban development project, approved in July 1979, was one nf the Bank Group's first attempts to assist the growing urban sector in the Sahelian region. Until then, Bank experience had concen- trated on the rural se--tor, and the Mali urban project, similar to other early urban projects, was characterized by a multiplicity of objectives and components, designed, essentially, to increase understanding of a field where few precedents existed. Bank strategy called for the introduction of low-cost alternatives for urban services and development as a means of making cities financially independent of the national government. Conse- quently, institutional strengthening and decentralization were stressed at a time and rhythm which the government was not, perhaps, totally prepared to accept. Objectives 2. The project combined a civil works component with social and eco- nomic programs. The first included a relatively large program of serviced lots and neighborhood upgrading, plus improvements throughout the city and the provision of equipment, a depot and technical assistance to the munici- pal maintenance department to improve garbage collection and road and drain care. The second added promotion of the informal sector through the construction of markets, development of local materials and a credit pro- gram for craftsmen and small businesses. Health and education were also included through the construction of a school and health centers. 3. Project replicability was to be guaranteed through the creation of a land management agency to carry out shelter programs similar to the ones financed under the operation. Urban maintenance services were to be fi- nanced by the Municipality, through a newly-established property tax, to be managed by another agency created under the project. Public water stand- pipes were to be financed through user charges. The project also included funding for a series of studies designed to identify further projpcts and planning issues for Bamako. Implementation Experience 4. Physical works were completed on schedule and within the projected budget. Savings on civil works contracts and the reallocation of funds from social programs permitted the extension of the shelter programs to include additional works for upgrading and serviced lots as well as the purchase of extra equipment for urban services and maintenance. A septic - vi - tank cleaning service was introduced and amplified as a result. At the same time, the credit programs for craftsmen and for home construction and improvement were dropped from the project as was the health component. The program to promote local building materials proved largely unsuccessful and was not continued in the follow-on project. Results 5. Studies were completed satisfactorily, including the design and negotiation of a second urban project, a ten year investment plan for the city of Bamako and a new land tenure law. In addition, the project spon- sored three seminars on themes of municipal finance, solid waste collection and city cleaning and the preparation and execution of subprojects. 6. Financial programs met with mixed results. Cost recovery for serviced lot projects has been highly successful, but largely because bene- ficiaries were better-off economically than anticipated, allowing for short-term repayment. On the other hand, the project has been unable to recover the costs of the neighborhood upgrading program from beneficiaries. The application of direct charges for water at public standpipes and septic tank clean-outs has worked well and guarantees the continuation of these programs on a city-wide basis. The property tax was duly established and applied, but has proved insufficient as a means of financing urban mainte- nance services. While municipal revenues have been bolstered by the intro- duction of other taxes, one of the main lessons of the project was the neez for a general financial office within the Municipality and an analytic study of Banmako's revenues. 7. Institutional goals remained elusive although project replicabil- ity has advanced significantly nonetheless. Establishment of the proposed land management agency proved more complicated than anticipated. Proposed reforms in real estate and land tenure codes were rejected, and a cadastral mapping of the city required more resources than originally calculated. Many of the functizns of the proposed land management agency, however, have been taken over by the Municipality and one of the agencies created by the project. At the same time, the project created two other municipal agen- cies which have little reason for remaining in existence, and, at this point, represent a financial burden on the Municipality. Sustainability 8. Overall project objectives were achieved and its benefits appear to be generally sustainable. Its poverty impact, however, has been some- what less than hoped for as the serviced lot project is slow to be occupied and beneficiaries appear to come from higher-income families. At the same time, the serviced lot program generated employment and stimulated the creation of new markets, indirectly benefitting low-income families. The upgrading and service components had a high impact, visible throughout the city. The serviced-lots and neighborhood upgrading programs succeeded as demonstration projects and similar programs have subsequently been devel- oped by the Municipality. - vii - Findings and Lessons 9. The Audit supports most of the findings of the PCR, but underlines that problems in cost recovery at.J institutional development resulted from project design and from inexperience with urban development in the context of the Sahelian countries at the time. Many of the projects' flaws were dealt with by the second project, which was seen by Bank ana local staff as a continuation of Bank strategy for the urban sector. 10. Institutional development suffered f:om the introduction of sever- al new agencies at once, withoat sufficient forethought as to their carry- ing costs. Furthermore, the major institutional component depended on the adoption of certain reforms of the real estate laws proposed by the Bank. In retrospect, it was ill-advised to tie an important part of the project to an uncertainty. Technical assistance played a key role in the institu- tional development. 11. Private sector participation in municipal service programs, namely management of public standpipes and septic tank clean-outs, has been very successful. 12. Local materials and traditional technologies (mainly stabilized earth and adobe) turned out to be more expensive than cement block for home construction. Appropriate technologies may have been better applied to the public service components, where sophisticated garbage collection systems proved problematie in many respects. 13. The massive serviced lot project proved a successful means for guiding urban development and expansion. However, it failed as a poverty alleviation measure, in part because the system of applying different prices to similar lots has encouraged resale at higher prices, and in part because the reduction of repayment time to five and seven years made debt service unaffordable to many families below the poverty line. Also, the mere fact that beneficiaries were required to document the--'.r salaries lim- ited the beneficiary group to higher-income earners from the fotmal sector. Finally, the fact that even these minimally serviced lots represent a con- siderably higher standard than most Bamakan residents enjoy made their eventual occupation by higher-income families inevitable. 14. The Bank push for small-sized lots, while based on a reasonable design to lower costs, contributed to environmental problems and may have raised costs in the end. The designers overlooked Malian traditional liv- ing patterns--and the fact that groundwater wells and septic tanks are used in the same plot. 15. Effective municipal revenue generation systems depend very much on local tradition. Attempts to introduce new mechanisms should be considered carefully. Collection costs should be analyzed and the costs of initiating a new system should not be overlooked. - viii - 16. "Christmas tree* projects depend on effective and flexible super- vision which can support successful components and discard unsuccessful ones. When the 'Christmas tree" project is used as a pathfinder to detect an effective strategy, then its lessons must be followed up or its purpose is lost. - 1 - PROJECT PERFORMANCE AUDIT MEMORANDUM MALI FIRST URBAN DEVELOPMENT PROJECT (CREDIT 943-MLI) I. SECTOR BACKGROUND 1.01 Mali is among the world's poorest countries, with an estimated per capita income of US$180/year. Since independence in 1963, Mali's public investment and its hopes for economic development have concentrated on agriculture. But years of heavy drought, from 1978- 1986, resulted in decreasing agricultural productivity and saw increased compen3atory bor- rowing, galloping deficits in the internal budget and balance of payments as the Government attempted to recharge the rural economy during a time of international inflation. Not surprisingly, non-farm production grew rela- tively rapidly during the past decade, at an average annual rate of 4.7% in real terms versus 2.6% for the agricultural sector. By 1986, over half of GNP was estimated to be generated within urban areas. 1.02 Mali remains a predominantly agricultural country with the rural population representing over 80% of the total. Nevertheless, trends over the past twenty five years show urban areas growing faster than the rural sector. The drought years set off waves of migration from countryside to city. While the country's population grew slowly at 1.3% per year from 1976 to 1987, the urban sector has developed considerably faster, at 3.4% during the same time period. Bamako remains the country's fastest growing city with an average annual rate of 4.2%. 1.03 What constitutes an appropriate urban policy for Mali and for Sahelian countries in general, has been a subject of debate in the Bank for the past fifteen years. During the late seventies, cities in the Sahel were viewed as "non-generative" parasites which robbed valuable resources from the productive rural sector. A second view of the urban sector e- merged after the drought years, when cities were perceived as buffers and poles of recovery against a failing national economy. (See Bank Staff Work- ing Paper No. 315, "Urban Growth and Economic Developmert in the Sahel" published in January 1979.) Six years later, the urban economy was consid- ered the primary locomotive for national development. At this writing, the pendulum appears to have swung again in the other direction, the main vir- tue of cities in the region, nowadays, being their capacity to consume agricultural goods and, thus, to increase demand for rural sector produc- tion. 1.04 The changing views towards urban development in Sahelian countries reflects the Bank's evolving understanding of the links between the urban and national economies. Bank strategy in the early 1980's emphasized: (i) the development of self-sustaining urban institutions to reduce the compe- - 2 - tition for resources with the rural sector; and (ii) investments in secon- dary cities, to develop important links with rural economies and to serve as generative markets. Thus, the first urban project in Mali was conceived as laying the groundwork in the capital, Bamako, and in several secondary cities for two subsequent projects, one in Bamako and another in five in- termediate cities. As Bank views of the urban sector and national economic development evolved, however, the secondary city project (originally the most important of the three) was set aside, to be later dropped altogether, making Bamako, the main beneficiary of the Bank's urban lending program in the country. 1.05 Bamako's population grew to around half a million by the early 1980s, sparking government concern over the rapidly expanding squatter settlements, deteriorating infrastructure and the city's generally rural appearance. The city had followed a somewhat haphazard growth pattern vith little systematic provision for property rights or public services, as the following observations illustrate: (i) Families could obtain occupation permits for lots from any one of several sources, the municipal lands office, the Ministry of the Interior, the Ministry of Public Works or even the police sta- tion. The first was the most frequently approached and it handed out provisional occupancy titles for CFAF 101 (about US$0.35). The lack of a single registry was further complicated by the ab- sence of a comprehensive system of naming and numbering streets. (ii) The Municipality depended on two drainage mains, which, like the overall system of stormwater drainage, were inherited at indepen- dence and had not been renewed or maintained since. (iii) About half the population received treated water either through household connections or public standpipes, delivered free of charge. The remainder of the population relied on water from ground wells, and an additional 30% supplemented their consumption with groundwater wells. At the same time, most families used septic tanks which filtered into the water tables, leaving the quality of the well-water in considerable doubt. (tv) The majority of the city's population relied on septic tanks with no clean-out system. The common practice of dumping their con- tents into the streets, further clogged the stormwater drains, with unpleasant hygienic and adverse environmental consequences. 1.06 The situation was aggravated by the Municipality's lack of insti- tutions capable of dealing with its problems. During the late 1970s, the Government of Mali established a National Commission for Administrative Reforms (CNRA) which ordered structural changes affecting city management and culminated in the decentralization decree of 1978 which granted Bamako regional status. Thereafter, Bamako should have planned and executed its own development projects, in addition to its traditional responsibilities: - 3 - maintenance of streets and public buildings, environmental sanitation (gar- bage collection, stormwater drainage, sewage removal and fumigation or pest control), fire prevention and regulation of land development and construc- tion. The city's meager resources and weak administration, however, re- duced the Municipality's capacity to such a point that its responsibilities 'have not been performed effectively, if not abandoned altogether" (SAR, para. 1.18). II. THE PROJECT A. Project Preparation 2.01 After an initial contact in 1976, project identification and de- velopment took the form of a two year dialogue which involved compromises on both sides. The Malian Government, frustrated by Bamako's deteriorating infrastructure and swelling population, wanted to invest urgently in civil works. The Bank, on the other hand, appears to have been more concerned with rationalizing future urban growth. A Back-to-Office Report after an early meeting sums up the different visions with a description of the Gov- ernment of Mali's interests: "issues such as land development for Bamako, use of the old airport site, employment centers to the south of the Niger River and policies for the peripheral settlements around the capital..." The same report urged that these questions "should be resolved orily after more fundamental policy decisions are taken concerning the nature and cost of services to be provided by the public sector." (BTOR, September 14, 1976). 2.02 As discussions continued, the issue of civil works versus institu- tionai development became clouded by a series of additional items intro- duced by the Bank including low-cost shelter solutions, local materials use, business loans for craftsmen, cost recovery and health, all areas in which the Bank had previous experience. Municipal strengthening, on the other hand, represented a relatively new idea. which defied clear defini- tion. Malian Government resistance to certain Bank proposals led to a year's delay in the feasibility study and caused the first effective time overrun. The major issues dealt with during the first phase of preparation are described in the following paragraphs. 2.03 Shelter. To a Government request for a program of housing for civil servants, the Bank countered with a program of serviced lots as a reception area for urban migrants and an urban upgrading program. Cost recovery proved a major hurdle in the case of serviced lots. In the past, the Malian Government had applied only token charges for the right to set- tle on unoccupied land and it did not accept the policy change easily. Service levels and construction standards were another point of discussion. The Malian Government expressed a preference for square lots of 500 and 600 sq. meters. Bank missions insisted on reducing lot sizes, first to an average of 240 m2, then iarther to 160 m2 and using rectangular forms and cluster-layouts. 2.04 While Malian resistance to reduced lot sizes may have appeared capricious to Bank planners, it was, in fact, based upon sound environment- al and economic concerns. The vast majority of the city's residents rely on water drawn from groundwells in their own backyards (para. 1.05 (iii)). At the same time, most use septic tanks with some form of ground seepage. With small lots the likelihood of the contaminatiot. of ground water is increased. The implicit environmental hazard was not considered in the feasibility study or in discussions during project preparation. By the same token, the smaller lot sizes implied a larger number of roadways, a factor which, of itself, raised the level of the development above that of the rest of the city. 2.05 Similar misunderstandings crose in the discussion of an urban upgrading component. Government negotiators hoped for complete renovation of informal neighborhoods which would have involved eradicating half of the existing housing. The Bank financed a study of three alternative levels of upgrading to support their case for low-cost upgrading. The selection of an appropriate neighborhood also required lengthy discussions and persua- sion. After rejecting at least five high and middle-income neighborhoods proposed by the Government, the consultants involved in the feasibility study, finally approved Magnambougou for an upgrading program and an adja- cent serviced lot development. 2.06 Land Tenure. Outright land ownership was rare in Bamako, although several agencies distributed provisional occupancy titles. The Bank re- garded this informal situation as a hindrance to future land management and rational urban growth, not to mention the implcit threat to private invest- ment and individual security. Considering that Bamako's development re- quired a municipal land management agency which could continue to prepare and implement serviced lot and upgrading projects, the Bank proposed to establish an independent and self-sufficient agency within the Municipal- ity. CAFOBA, in turn, depended on the establishment of a new real estate code which would have permitted families to obtain clear title to undevel- oped land. Although the Government vent along with the scheme, it did not appear to share the Bank's enthusiasm. The Bank's proposal would have reduced the powers enjoyed by the two Ministries which dispensed land titles. At the same time, the Mal.an Government did not look on private property as particularly desirable or necessary. 2.07 Informal Sector and Local Materials Components. Although not particularly related, these two components were linked together in the project. The theme of "small is beautiful" was adapted and applied in many Bank-financed urban projects of the 1970s in the form of support for craftsmen and traditional building techniques. Both components were pro- posed by Bank negotiators and accepted by the local government. Oddly enough, appropriate technologies were not considered in other components, such as sanitation. 2.08 Secondary Cities and Other Possible Components. The Malian Gov- ernment considered drainage a top priority, proposing sewerage systems for Mopti and Kayes and some water supply investments for Gao, together with a garbage collection system and rehabilitation of the central market for Kayes. During project preparation, the needs analysis shifted. Water supply became the focus for Kayes, as well as Gao, and a system for garbage collection was assigned to Mopti. Although cost recovery and relvenue mo- bilization for urban services were considered high priority in Bamako, they were not included in the secondary city components. Transportation, men- tioned frequently by the Bank during project preparation meetings as a high-priority need for the growing city of Bamako, was not considered par- ticularly urgent by the Government of Mali. Health facilities, mentioned as an item of "highest priority" during project preparation discussions, were similarly discarded. 2.09 A consulting group began the feasibility study in December 1977, six months later than initially planned. Institutional strengthening and revenue generation for the Municipality of Bamako, both recurrent themes of major importance to the Bank, were treated very sketchily in the terms of reference and in the resulting feasibility study, which focused on the civil works components, and, in particular, on defining the serviced lot and upgrading projects. The Malian Urban Project Unit has subsequently expressed its diss&tisfaction with the consultants' work and with several aspects of the feasiuility study, including: inadequate design of garbage collection system and drain clean-out schemes, inappropriate drainage de- sign for serviced lot and upgrading areas, neglect of sanitation questions and insufficient consideration of the need for a land use agency. Underly- ing such specific concerns, the project unit resented what it considered to be the consultants' lack of communication, collaboration or participation with local experts. The consultants group, for its part, in stating, "the Government didn't know what it wanted", has suggested that it was left with the problem of designing a project where neither the Bank, nor the Malian Government had given clear indications of a project purpose or direction. 2.10 The consultants recommended that the project be managed by exist- ing Government agencies. The Bank, on the other hand, preferred the crea- tion of a separate project management unit and ieveral new municipal insti- tutions, and established them as conditions for negotiations a* Appraisal. In November 1978, the proposed new institutions (of which there were many) ceased to be conditions for implementation, but rather were programmed for phasing into the project at large. B. Project Objectives 2.11 The differences between the Malian and Bank versions of project objectives are telling, particularly because the Malian objectives appear to follow the Bank's urban/country development strategy (as set out in the Sahelian Cities Working Paper) more closely than do the Bank's objectives. According to the PCR prepared by the Malian project unit: The urban development strategy devised by international donors, specifically the World Bank, foresaw: (a) con- taining the uncontrolled growth of the principal cities and developing economic and institutional ties between - 6 - the urban and rural sectors; (b) introducing new methods for furnishing low-cost urban services and for effective cost recovery; (c) strengthening national and local institutions; and (d) stimulating local participation in urban development programs. In contrast the SAR lists the following as general and specific project objectives: The main objective of the project is to assist the Gov- ernment in providing affordable urban services...to low- income urban residents. To insure replicability and maintenance of project components, cost recovery will be major goal, especially for shelter and municipal ser- vices... The specific objectives of the project will be to; (a) provide shelter and community facilities to meet about half of the growth in demand for low-income groups; (b) improve health and sanitary conditions; (c) stimulate productive employment (in the informal sec- tor); (d) provide essential water and sanitation ser- vices to secondary cities, and; (e) establish and strengthen institutions in the urban sector. 2.12 The Bank urban/country development strategy had emphasized high- impact projects, rural/urban linkages in primary and secondary cities, and reduction of national investments in urban development through the intro- duction of low-cost services and the mobilization of local resources. Aside from the attempt to strengthen urban institutions, the Bank objec- tives primarily defined a poverty alleviation project. 2.13 These divergences suggest that the following goals may not have been fully shared by the Malian Government, but rather may have befn large- ly imposed on negotiators by the Bank: (i) The Bank emphasis on shelter and services for low-income popula- tions contrasts with the Malian concern for containing uncon- trolled urban growth, i.e. for providing new neighborhoods of higher standards. (ii) The employment generation component. In retrospect, the Project Unit credits the project with generating a substantial amount of employment through the civil works, through on-going maintenance programs, and indirectly through the home construction generated by the upgrading and serviced site programs. In contrast, the programs for loans to small businesses and promotion of local materials program did not seem very convincing. (iii) The urban land management and development agency appeared crucial to the Bank, in order to provide a rational framework to meet the demands of Bamako's growing population. The Malian Government considered urban development more a question of adequate infra- structure than of real estate codes and property laws. -7- C. Project Description 2.14 The PCR provides a general description of the project components (PCR, paras. 2.03-2.06). Annex 1 provides a general outline of project investments. The following paragraphs offer some additional information regarding specific project goals and mechanisms: (i) The civil works programs, represented close to 55% of total proj- ect cost and included the following: (a) in conjunction with the shelter components; construction of 2,600 new serviced lots; improvement, grading and/or paving of about 9 kms. of roadway, with drainage ditches where necessary; construction of 20 public standpipes and public exterior lighting on main streets; construction of a school; a market and a health center; and off-site construction asso- ciated with water supply, drainage and electricity; (b) in conjunction with the city-wide sanitation program: con- struction of new roads; the repair of 120 existing public standpipes and installation of 20 new ones throughout the city; construction and/or improvement of a drainage main and drainage ditches and ten public toilet blocks; the purchase of vehicles and equipment for the provision of sanitation services; and construction of a maintenance park and work- shop. (c) in secondary cities: purchase of garbage and vacuum trucks and construction of a public toilet block in Mopti, comple- mentary water works in Gao and Kayes; and (d) in Bamako, construction of three additional markets. (ii) The loans program which represented about 9.5% of the total proj- ect cost and would have made about 5,000 credits available (450 for craftsmen and the remaining 4,500 to residents of Magnambougou's new serviced lots and upgraded neighborhood). (iii) The local materials component which aimed essentially at training and promotion through construction of some model buildings in stone and adobe, for US$190,000 or 1.6% of the total project cost. 2.15 Cost Recovery and Institutional Strengthening. The project aimed to recover full costs of the civil works and credit components, which rep- resented about 60% of total project costs. Recognizing that such goals represented new directions and tasks for the Municipality of Bamako (and, indeed, for the Government of Mali), the project proposed to create the following new institutions for the express purpose of mobilizing the rev- enues needed to recover project costs and to sustain or replicate similar programs in the future: The Housing Bureau (DSLD) to manage loans for home construction and lot charges in newly serviced areas and upgraded sections of Magnambougou; an assessment office (IDID) to assess and collect the proposed property tax which would sustain the public maintenance services; - 8 - and the property registry (IDD) which would charge for land occupation and titles, and later meld into the proposed CAFOBA, land management agency (para. 2.06). In addition, the Municipal Road Department was to be re- organized into a Municipal Technical Services Department (DSTD). The proj- ect design did not indicate by which institutions, or by what means, costs were to be recovered in secondary cities (about 5Z of the total project costs). About 7Z of the project was designated for Institutional Assis- tance, for the acquisition of vehicles, and equipment and to help defray operating costs. 2.16 Technical Assistance represented about 22% of total project cost and was geared in rather general terms to helping the new agencies and institutions become established and begin functioning. Notably, no techni- cal assistance was foreseen for Mopti's sanitation component, as occurred with Bamako, nor for institutional development in the secondary cities. Studies were focused mainly on the preparation of long term planning and two additional projects for Bank financing. In addition, an expert in real estate law was contracted to review the existing property codes and propose a new law consistent with the creation of a land management agency (CAFOBA) and with urban land development. 2.17 The project was to be managed by a special unit which coordinated the activities of the various participating agencies and maintained rela- tions with the Municipality of Bamako, under the supervision of the Ministry of the Interior and the Ministry of Public Works, which took re- sponsibility for most of the construction projects. III. PROJECT IMPLEMENTATION AND ACHIEVEMENTS A. Time Delays and Overruns 3.01 Like most of the urban operations of its generation, the first Malian urban project was characterized by time overruns, resulting from an overly optimistic implementation schedule. Project preparation took a year and a half longer than foreseen at the outset. The project required a year from credit approval to effectiveness and seven years to complete, as op- posed to the four years estimated at Appraisal. The disbursement schedule presented at Appraisal implicitly assumed that the Municipality of Bamako could double its investment capacity on a yearly basis, thus achieving an increase of 500% in the space of three years. In point of fact, the proj- ect did double Bamako's investment capacity, but only once and it main- tained a more or less corstant rate over seven years. 3.02 Time extensions in certain components can be traced t3 insuffi- cient preparation and/or errors in the feasibility study. In particular, the garbage collection system lacked definition, as did the beneficiary selection system for Magnambougou's serviced lots and the scheme for recov- ering upgrading costs in Magnambougou (para. 4.16). In addition, the site chosen for the DSTD warehouse and garage required i -filling. The terms of reference for the systems analyst were inadequa. and were rtodified during implementation, resulting in a substantial extension of services. - 9 - 3.03 Establishment of the various implementing agencies, which consti- tuted the major condition of effectiveness, proved more difficult than anticipated at Appraisal. The credit was declared effective in June 1980, more than six months after approval by the Board, on fulfiilment of the following conditions: (i) creation of the project implementation unit (PUM); (ii) creation of regional directorates to function within the Municipality of Bamako; (iii) establishment of the Bamako Housing Bureau and the DSTD; and (iv) the establishment of a property tax to defray the costs of sanitary maintenance (roads, drains, septic tanks and garbage collection). 3.04 Certain delays can be traced to lack of support from the Malian Government or, to the project's dependence on a variety of different agen- cies or events which could not always be coordinated in a timely fashion: (i) The rhythm of project implementation was seriously slowed in its second year because of a shortage of counterpart funding from the Central Government. The impasse was finally resolved by using revenues from project cost recovery accumulated in an Urban Devel- opment Fund (Fond d'Amenagement Urbain). Fortunately for the project, the fund had worked well, recovery levels were high and the statutes gave control over it to the PUM director and the mayor (para. 4.22). (ii) The attempt to develop a land management agency (CAFOBA) and a new real estate code along lines suggested by the Bank was frustrated by the Government of Mali's presentation of its own real estate code in a version which virtually contradicted the Bank's propos- al. The legal expert who had been hired under the project to review the existing law was consequently requested to draft the proposed legislation in lieu of recommending changes. 3.05 The devaluation of the dollar and the reallocation of resources within the project is mentioned as a reason for the project's extension. In fact, while some additional contracts were let out under the project, taking advantage of the increased purchasing power of the Malian franc (for a short period), a number of components were also eliminated from the proj- ect. The Audit considers that the time elapsed was more a consequence of the nature of the project than of currency fluctuations. B. Revisions in Project Scope 3.06 As the tables in Annex 1 show, the project as implemented remained essentially consistent with its design at Appraisal. Civil works repre- sented 56% of total project costs, as compared with 55Z designated at Appraisal. The most significant changes in investment came with the elimi- nation of the credit programs for home construction and small craft enter- prises and the consequent increases for institutional assistance and studies. The former components represented about 9? of the total project cost, and were dropped largely for lack of government experience with fi- nancing programs. Changes within each component are described in the fol- lowing section. - 10 - 3.07 A major change in project concept resulted from the failure to pass the real estate code and property laws recommended by the Bank (para. 2.16). Without the supporting legislation, the Municipality had no power to develop and to sell land, and CAFOBA lort any effective reason for com- ing into existence. Nonetheless, some of CAFOBA's functions were assumed by other municipal agencies. C. Project Achievements and Continuity 1. Civil Works 3.08 New Serviced Lots at Magnambougou. The SAR foresaw a total of 2,600 lots, housing approximately 22,000 persons. In fact, the overall size of the project was enlarged and the number of lots increased by al- most 40Z in spite of the fact that lot sizes wei:e somewhat larger thAn planned at Appraisal. The number of beneficiaries, however, cannot, as yet, be said to have increased by the same proportion because the majority of the new lots have yet to be occupied. The studies made available to the Audit make no attempt to quantify the degree of occupation of Magnambougou, nor to explain it. The fact that all lots were sold within a year of being put on the market certainly indicates the existence of an effective demand. There are several possible explanations for the time lag between lot sales and occupations (i) During implementation, the pricing system was change4 from the fixed price plan stipulated in the SAR (Table V-1) to one which charged families according to their income levels. The results may well be encouraging speculation and discouraging occupancy. Families earning between 12,500 and 47,500 CFAF (between the 30th and 80th percentile of income distribution) paid 1,125 CFAF per square meter. Families earning above that amount, and below 125,000 CFAF/month, (80th to 90th percentile) paid 3,100 CFAF per square meter, while the third range of families (greater than 125,000 CFAF/month and above 90th percentile) paid 4,800 CFAF per square meter. Two more categories existed for commercial use at still higher prices per square meter. The pricing system held no matter what the size, shape or location of the lots, except for the commercial lots, which represented less than 20% of the total, and were considered prime real estate, being located on the main road. For the 70Z of the beneficiaries who bought land at the lowest prices, the temptation to sell at the higher prices paid by their neighbors may act as a deterrent to building. The invest- ment is far more valuable for its resale value than for occupancy. The subsidy can be converted easily to a capital gain, once the lot is fully paid off. (ii) By the same token, according to the project unit, the sales value of lots at Magnambougou has already increased threefold since completion of construction. As a result, it has become ineffi- cient for families in lower and middle income brackets to occupy the land. Certain families may well be waiting to pay off their lots in order to resell. - 11 - (iii) The fact that no one knows exactly how many separate lots exist in Magnambougou at the moment emphasizes the current flux as lots change hands (PCR, para. 4.07). The project was originally de- signed with large lots so that families could subdivide and resell half-sized lots if they had difficulties with payments. In fact, the reverse has occurred; owners tend to purchase adjacent lots from their neighbors in order to combine and increase the size of their holdings. This process in itself may act as a further deterrent to building, like the two described above, insofar as it encourages resale rather than occupation. (iv) Another change during implementation reduced payment terms from twenty to a maximum of seven years. This system has obvious im- plications for the beneficiaries' income groups (para. 4.16), but it also signifies that monthly payments are relatively high. Many families may be waiting to pay off the cost of the land before commencing construction. (v) The exterior street lighting on Magnambougou's main streets sug- gests to residents that it is but a matter of time before house- hold electrical connections become available. Families may be waiting for electrical installations to become possible, either to build or to resell under the circumstances. (vi) Local tradition may also explain the lack of residents at Magnambougou. The project director observed that Malian families tend to remain at home during their parents' lifetime. Under normal circum3tances, all but one of a family's children will move out on the death of the parents. The case of families purchasing a lot for future security and taking theii time in construction was indeed confirmed by interviews on the project site. (vii) The elimination of the construction loans component and the conse- quent lack of financing suggests that families may not be able to pay off their lots and find money to build housing all at once. (viii) Bamako's population growth rate from 1976 to 1986 was somewhat less than anticipated, amounting to a 4.2Z annual increase, rather than the 8% calculated at Appraisal (SAR, para. 1.01). 3.09 The upgrading program was carried out as planned. It benefitted approximately 2,250 households, significantly more than the 1,770 foreseen at hppraisal because the area of works was enlarged in the course of imple- mentation. Also, the component was expanded to include more works: an ex- tension of the paved roads, a stormwater drain system along the primary road to Magnambougou, inter-block roads and a bridge over a tributary of the Niger River which surrounds the site and cuts the neighborhood off from its surroundings. The process introduced two important elements into Bamako's urban development: security of tenure and community participa- tion. As described by the project officer in 1984 (from an article in "Designing in Islamic Architectures - 5'). - 12 - Work was accomplished in three stages. First, all the roads were widened by a minimum of 13 feet to facilitate traffic flow and the eventual installation of a secon- dary-road network. Then the site plan. street map and land survey were made. The technical team indicated the boundaries for each plot and assigned a family to each, and indicated the structures that had to be demolished and those that could be retained. Finally the inhabi- tants themselves then realigned the streets as indicated and built their own houses on the plots assigned to them. 3.10 The project also resulted in the assignment of street names and numbers, not only in Magnambougou, but throughout the city. Although the benefits of urban nomenclature are difficult to quantify, they amount to a city-wide improvement of certain significance, which was foreseen neither in the SAR, nor in the feasibility study. 3.11 Improvement of Sanitary Conditions. This component was increased by about 47Z as a result of the reallocation of funds in the course of project implementation. The only sub-component not addressed was the con- struction of public toilets, which amounted to around 2Z of the original project. The toilets, scheduled to be installed in markets and public buildings, were abandoned because the cost of bringing in off-site water proved too high. Only one toilet bank was built, at the public school in Magnambougou. On the other hand, an important septic-tank clean-out compo- nent was added, with a much wider impact than the proposed public toilet project, and at a lower cost. Details of the program for the improvement of sanitary conditions are described below: (i) The DSTD Maintenance Shop proved a more ambitious undertaking than originally envisioned, costing over twice the dollar amount esti- mated. Construction was prolonged because the lot required infill and occupation was delayed because public works required payments from the Municipality before hooking up utilities. The original building program was also expanded to include a security wall, a ramp, truck-wash area and storage for bicycles and motorcycles. (ii) Garbage Collection. The project financed the purchase of garbage collection equipment (9 dumpster trucks and 190 dumpsters of dif- fering sizes) as well as technical assistance and operating costs for two years, or until the Municipality had mobilized revenues to pay for garbage collection on its own. In fact, the final defini- tion of a garba. e collection system took longer than that. The feasibility study offered only a generalized diagnostic study and plan of action, drawing from guidelines used in developed coun- tries and studies done in the late 1960s and early 1970s in Bamako. The dumpster system took time to catch on and produced unforeseen problems: garbage accumulation in the areas sur- rounding the dumpsters, or the loading time required by additional small dumpsters. The first problem was resolved by posting a - 13 - 'garbage guard' for short periods of time at each dumpster who trained families to dispose of their garbage in, and not nearby, the dumpster. The second has been left for the Second Urban Project which included a complete study of the city's wastes and an alternate collection system based on larger sized dum sters (7 m3). At present, DSTD claims to collect 1,000 to 1,200 m/day, of which 500 m3 is collected with equipment financed under the first urban project. (iii) Drainage Ditch Clean-Out and Construction. With the clean-out and repair of 73 kms. of priority drainage ditches, completed in the first year of the project (PCR, para. 4.16), the project unit fulfilled what they understood to be the total project goal for drainage clean-out. Although the SAR mentions clean-out and re- pairing of some 250 kms of driinage ditches, the feasibility study, based on a WHO/PNUD project of 1974, specified a short run (0-5 years) program of cleaning out 70 kms. of priority drains and the long-run establishment of a maintenance system which would cover the city-wide stornwater drain system. Cost estimates were based on 70 kms not on 250. In addition, the project excavated and rehabilitated a total of 15 kms of city stormwater drainage mains, built the off-site stormwater drainage system for Magnambougou, and installed concrete lining in stormwater drains as part of the urban upgrading program, fulfilling the program specified at Appraisal (para. 21). At present DSTD manages to clean about 50 meters of drain/day, which means a yearly average around 12 kms, or a city-wide clean out once every twenty years. Neither the SAR, nor the feasibility studies specified maintenance goals for DSTD. (iv) The SAR does not mention the septic tank clean-out program (attri- butable, at least in part, to semantics: the french "drenage', which tefers to sewage systems as well as storm water drains, is insufficiently translated as stormwater drainage only.) Nonethe- less, the component is very significant: the majority of Bamako's familie3 rely on un-leached septic tanks which require emptying about every three months. Before the provision of septic tank clean-out services, families generally dumped waste materials from septic tanks into the streets or street drains. Using unallocated funds, the project financed the purchase 4nd activa- tion of four vacuum pump trucks which responded to requests from individuals and cover about 36Z of the demand in Bamako, A ser- vice not foreseen in the Appraisal Report, but essential to the sanitary maintenance of the city. Since DSTD began offering a septic tank clean-out service private sector contractors have also begun to compete in the same field. As a result, the maintenance needs of Bamako's septic tanks are estimated to be covered by the combined efforts of the public and private sectors. DSTD charges for its septic tank clean-outs and its revenues have increased, as a result from 1.12 billion to 5.66 billion CFAF in four years (1980-1984). - 14 - (v) The public standpipe program was also expanded to provide a total of 180 operating standpipes, between new installations and re- pairs, as opposed to the 118 programed at Appraisal (PCR, para. 4.18). Like the septic tank clean out system, the standpipe pro- gram is recovered by direct charges to the users. 3.12 Informal Sector Support. The program of credits for artisans was cancelled at the request of the Malian Government, which saw the program as too complicated. Four existing markets in low-income neighborhoods were enlarged and upgraded with paving and drainage. The proposed individual stalls (SAR, para. 2.21) for craftsmen were not built as they have little relation to local tradition. Bamako's markets function out-of-doors with open stalls sheltered from the sun. Unfortunately, the consultants' de- signs placed concrete parasols at such a height as to be ineffective in providing shade except at 12 noon. 3.13 The local materials program is described in the PCR (para. 4.23). Aside from the construction of a school in Magnambougou, (even at costs estimated to be four times that of conventional construction), it provided an additional service of designing house types and plans for beneficiaries of the serviced lot program at Magnambougou. 2. Institutional Development 3.14 Land Policy, IDD and CAFOBA. While the proposed land management agency, CAFOBA was never established, one of its divisions not only came into existence under the project, but has come to manage many of the func- tions originally envisioned for CAFOBA (PCR, para. 4.12). IDrJ (L'Inspec- tion Domaniale du District de Bamako), was created in 1979 with two major functions: (i) to assign lots and pass titles to beneficiaries of the serviced lot and urban upgrading programs; and (ii) to put some order into the archives of tte former "Section Domaniale du Governat" which contained information about the former French neighborhoods under the colony. IDD was to develop and maintain a city-wide simplified cadastral study as part of CAFOBA. Such a study, however, was easier described than done: As Bamako had neither street names nor numbers, but rather innumerable devel- opments and neighborhoods with the same name--not to mention persons with the same name--the task was by no means a simple one. Furthermore, the project had not included sufficient funds to undertake the topographic surveys and transpositions required to transform aerial photographs into cadastral (or street) maps (which were later included in the second proj- ect.) IDD, nonetheless, managed to map, name and number streets and iden- tify the city's major neighborhoods, in addition to bringing the occupancy papers up to date for the majority of the city's households. 3.15 At the same time, IDD proved enormously successful at providing itself with an income-generating mechanism, by offering and charging for the following services: registering and granting permits for land occu- pancy, land transfers, land regularization and registry of deeds. While not acting entirely as a land development agency, IDD has initiated and made possible further urban development projects for the Municipality (para. 4.24). - 15 - 3.16 Created at the same time as IDD, the Municipal Tax Office, IDID (L'Inspection Divisionnaire des Impots du District) had the mission of generating resources to pay for road maintenance and garbage collection principally through the establishment and assessment of a general property tax based on estimated rental values. Although IDID was created irt 1979, and the property tax followed in 1981, collections did not begin until 1984 when the Municipality hired tax collectors. IDID works with a staff of 15 assessors, who p acess information from the communes (sulbdistricts of Bamako) and deternine the value of the taxes due in each commune. Under the project, IDID was able to furnish its office and purchase fifteen motor bikes. It also received technical assistance in the area of finance duriig the first three years of project implementation and assistance in computerizing data through the "systemization" component. Although IDID has initiated other new taxes, notably on consumption in the night clubs and taverns, its operating expenses according to its Director, are not covered by the revenues it generates. 3.17 DSTD, formerly the Municipal Road Maintenance Department, required a certain amount of time - get started. A supervision mission in 1983 which focused on its problems at that time, traced DSTD's inefficiencies to problems of: (i) a work force which never came to work; (ii) the use of the machines for purposes other than municipal maintenance; (iii) the out- right misuse of equipment; and (iv) the absence of a planning or time/effi- ciency division combined with the practice of not amortizing equipment. Since that time the project Director and project unit have intervened to reduce and replace over half of DSTD's employees and to attempt to put the Department on a sounder footing. In addition, it appears that DSTD rents out its equipment to the the local communes of Bamako in the afternoon, for their own maintenance purposes. It delivers household garbage and septic tank contents to farmers (for organic fertilizer). Although dr.'ored by the PUM for not covering costs of amortization, maintenance or ..asurance (rental charges cover merely the chauffeur and gas), these activities do represent an interesting, and perhaps efficient, use of the equipment. 3.18 The Municipal Housing Bureau, DSLD (Direction du Service du Logement du District) was created to manage the payment collections of Magnambougou, serviced lots, upgraded neighborhoods and the construction loans program. Had CAFOBA been effectively established, the DSLD would have become its integral housing and loans office. As it is, however, the office manages the credit programs for the Municipality's smaller housing projects, which amount to less than one hundred units, in addition to Magnambougou. Despite the rapid recovery of costs of the serviced lots at Magnambougou, the Municipal Housing Bureau's overall performance appears to be dwindling as its mission comes to completion. 3. Resource Mobilization 3.19 Cost Recovery (Shelter Projects). Although the success of cost recovery in Magnambougou can be rivalled by very few other Bank experiences (70% repayment in five years), it must be considered qualified in view of the following factors: - 16 - (i) Neither the project unit, nor the Municipal Housing Bureau can quantify what percentage of project costs are actually being re- covered (nor, in consequence, how much of the project is subsi- dized, or how much of a profit is being realized). The accounting system (or lack thereof) neither permits them to determine how much is actually outstanding, nor what were the original shelter project costs. Aside from the complicated system of cross-subsi- dized pricing (para. 3.08 (i)), the fact that a single contract paid for both upgrading and serviced lot development makes it impossible to calculate the cost per beneficiary for each compo- nent. Finally, the three year time delay between completion of physical works and receipt of payments were not taken into account in calculating project costs, so the interest rates applied do not necessarily correspond to real interest rates plus inflation. (ii) Cost recovery for upgrading appears to have failed. The project assumed that beneficiaries would repay project costs through costs associated with legalizing their lot titles after general upgrad- ing had taken place. In fact, less than five percent of the households have requested titles. The project Director (formerly Director of the Upgrading Program) considers that a more timely payment schedule would have been more effective. As it was, project costs/m2 were determined only after the civil works had been completed for several months, when families no longer felt any incentive to pay. 3.20 Service Charges. The introduction of private water salesmen at the standpipes (see PCR) appears to be working adequately. The PCR does not mention the charge system for the septic tank cleaning service, also introduced into the project through DSTD, and subsequently copied by two private competitors in Bamako. Unfortunately, the lack of adequate ac- counting makes it difficult to determine whether or not service' costs are being adequately recovered. 3.21 New Taxes. The evolution of the City of Bamako's income from taxes and charges appears below. The contributions from the Bank project are seen in the garbage collection tax (taxe de voierie et d'enlevement d'ordures municipale - TVEOM), in the septic tank and drain clean-out charges (droit de vidange et de curage) and in IDD's assorted charges (permis d'occuper, intervention de 1'inspaction dominial de Bamako.) These have contributed in total to between 13Z (1985) and 56Z (1988) of total municipal tax revenues. Again, the apparent success of the revenue genera- tion component z;ust be considered qualified for several reasons: (i) Of the three newly introduced taxes, only the septic tank and drain clean-outs are increasing in absolute terms, maintaining the same growth rate as that of overall municipal income. Understand- ably enough, the charges imposed by IDD, impressive as they are, vary and do not constitute a reliable source of income. Revenues from the TVEOM, on the other hand, are falling off, essentially because of the high cost of collection. According to the IDID - 17 - Director, the fifteen motorbikes used by his agents are now in disrepair, but the office does not generate enough income to per- mit replacement or repairs. (ii) Revenues through TVEOM grew until 1986, but remain far below the level forecast, covering neither DSTD's annual gasoline bill, nor even the cost of their own collections. The Director of IDID explains the TVEOM's problem as threefold: (1) only about 10% of the city's buildings (those built in cement and the adobe build- ings which are rented out) are actually included in the taxrolls; (2) of the calculated taxes due (5% of estimated rental value for non-industrial properties and 2Z for factories and industries), less than half actually gets collected, and; (3) of the amount collected an unspecified portion gets mixed into the city's gener- al revenues, because the same tax agents collect revenues from several taxes, often without differentiating between the sources. (iii) IDD's revenues come basically from three sources: the charge for occupancy permits, which has been increased slightly, the nominal title charge (US$0.50/lot) and from a recently applied road charge, which assigns the cost of access roads to the residents of adjacent properties. Recent municipal legislation stipulates that no lot may be assigned without an access road, and, consequently, obliges residents to pay a considerably higher initial cost for their occupancy permits than the nominal charge, previously applied. EVOLUTION OF MUNICIPAL REVENUES FROM TAXES AND SERVICE CHARGES (in '000 CFAF) 1983 1984 1985 1986 1987 1988 Totals 673.3 1,017.5 1,133.8 1,157.5 1,100.2 1,502.4 Total from Project 25.9 126.4 148.9 139.0 148.5 86.9 Percentage 4.0 12.0 13.0 12.0 13.5 6.0 TVEOM 0 0 0 24.2 12.7 22.3 IDD 23.9 120.7 143.5 109.5 125.0 54.6 Septic tank clean-outs 1.9 5.7 5.5 5.3 10.8 10.1 Source: Bamako Bureau of Tax Collection. 3.22 The Audit did not review programs in Mopti, Kayes and Gao. They are reviewed in the PCR. - 18 - 4. Technical Assistance 3.23 Technical assistance was offered in three basic areas: in project preparation and studies, in institutional strengthening for the newly cre- ated agencies and in project management. The last, assistance to the proj- ect unit, seems to have been the most successful. Considering the number and novelty of the new institutions created under the project, the techni- cal assistance component could be considered insufficient or misdirected. Had DSLD, IDID, and even IDD benefitted from the assistance of a financial expert they might have been able to better organize municipal finances, as well as their own, instead of leaving matters up to the second urban proj- ect. Similarly, DSTD could doubtless have profited from an expert in sani- tary engineering (none appears to have contributed to the feasibility study) and from an expert in management or business administration. Final- ly, the systems analyst was hired without clear terms of reference. While the project unit has been happy with his work in general the Audit feels that his contribution could have been more usaful with a clearer mandate. As it is, the pr-ject unit and several of the institutions have profited from storing textual data on the micro computers the unit has purchased, but no financial or spread sheet functions have been used to date to facil- itate operations in the agencies concerned with budgets, tax collection, cadasters etc. Certain of these problems can be attributed to weaknesses in the original feasibility study. 3.24 In general, project staff expressed satisfaction with the long term technical assistance. At the same time, they mentioned problems which arose in connection with the technical advisors hired through consulting agencies. (The PUM director relates the case of advisors paid salaries by agencies which were considerably less than the amounts their contracts had declared. This caused a certain amount of grief both for the project and the consultants in question.) 3.25 While some observers have criticized the long-term consultants for appropriating the project, for earning too much money, etc., the increase in technical assistance expenditure appears to be due to: (i) prolongation of the contract for the advisor to the PUM Director, recommended in view of the death of one of the project directors, the grave illness of the first advisor and the continuity and expert support offered by the particular consultant; (ii) the additional costs for the local materials advisor; (iii) the increase in the program of one systems analyst, originally con- templated for two years, to include the introduction of the computer-sys- tem, including the purchase and installation of three micro computers and the systematization of several project components; (iv) expenses involved in developing and negotiating the second urban project; and (v) three semi- nars on urban problems which led to certain improvements in the second urban project (PCR, para. 4.30). - 19 - IV. POINTS OF INTEREST A. Physical Impact 4.01 The serviced lot and upgrading projects at Magnambougou comprised extremely sizeable undertakings in any case, but for a first project they could have been overwhelming. Magnambougou signified an increase of ap- proximately 7% in the total number of urbanized lots in the city, account- ing for two years of population growth. Together with the upgrading compo- nent, the project improved housing conditions for about 10% of Bamako's families directly. In terms of area, .the new lots and the upgraded areas at Magnambougou signified an increase of 92 of urbanized land in the metro- politan area of Bamako. In addition, the project has, in effect, opened up the south bank of the Niger River as a new urtan sub-center. The off-site infrastructure installations permit urbanization of another 150 to 200 hectares in the near future. 4.02 The shelter project produced a curious result as far as land use densities are concerned. While the Bank urged Bamako's planners to accept reduced lot sizes of 160 square meters (para. 2.03) and even foresaw the possibility of subsequent subdivision into lots of 80 square meters, events at Magnambougou have demonstrated a preference for larger lot sizes. Resi- dents are buying up adjacent lots, thus doubling their land holdings before building houses. At the same time, the advantages of larger lots have become apparent to Bamako's urban planners, notwithstanding the traditional arguments equating reduced infrastructure costs with smaller house lots. The primordial argument is environmental (para. 2.04). A second argument refers to family size and tradition proposing that larger lots can house move people. Insofar as the Magnambougou program brought these issues to light, producing a subsequent backlash against the small lots it provides, it can be considered partially responsible for a recent decree prohibiting the occupancy or development of lots smaller than 300 square meters. The impact of this decree, albeit unanticipated by project planners, will be significant to Bamako residents and to its future growth patterns. 4.03 The most important impact of the shelter programs in the first urban project will be felt in the future. As a demonstration project Magnambougou has indeed succeeded in convincing municipal authorities of the possibilities inherent in urban development through planning extended residential and reception areas. The second urban project, currently un- derway, (para. 4.28) will continue such efforts on an even greater scale increasing Bamako's total (urbanized) area by over 252 and increasing the number of lots overall by close to 202. In addition, the Government of Mali has initiated serviced site programs on its own. Faladie, recently completed a total of 400 lots. At the same time, the importance of up- grading existing areas has been recognized. The second project will di- rectly improve the existing housing for another 25! of Bamako's residents. 4.04 Sanitation Project. The impact of sanitation programs: garbage collection, septic tank cleaning, drain clean-outs have not been adequately measured in terms of the city's health. Service today has diminished com- - 20 - pared with levels existing three years ago, but the Audit found no indica- tion that garbage collection or drain clean-outs had ceased or even been reduced significantly (PCR, para. 4.15). Septic tank cleaning and water supply through public standpipes have kept up with the city's growing popu- lation. However, the relative impact of urban maintenance services will continue to diminish as the city has grown by about 200,000 people (37%) since the first project was designed. 4.05 With reference to garbage collection, the feasibility study esti- mated that Bamako in 1978 produced 273,000 m3 of garbage/year. Simple interpolation implies that, in 1987, Bamako shouid produce some 450,000 m3 of garbage/year, given the same parameters. At present, DSTD claims to collect 1,000 to 1,200 m3/day, of which 500 m3 is collected with equipment financed under the first urban project, thus substantiating DSTD's claim that it collects 60-70% of the city's garbage production, a clear improve- ment over the 32? coverage calculated by the feasibility study. The suc- cessful use of supervisors at garbage collection points to educate families should not be underestimated. The dumpster sites visited by the Audit appeared to be clean and well-maintained, indicating that families have responded to the collection service. 4.06 The drain maintenance program proved very timely and providential during last year's floods. Water levels rose significantly in certain parts of the city for a matter of days and produced several deaths. It is estimated that without the drain clean-out program, the waters would have stagnated indefinitely, requiring major pumping, and the number of deaths would have been much higher. B. Institutional Impact 4.07 While the project aimed to create and strengthen agencies which could further its urban development goals, its most important institutional impact came in terms of municipal strengthening. Even though it never set a full-fledged land development agency in place, it established agencies through which the Municipality has begun to organize and plan land develop- ment for Bamako (para. 4.03). At the same time, the project has created certain problems for the Municipality which must clarify the future func- tions of such agenLies: (i) IDD, IDID and DSLD, were to have become part of the proposed land development agency, CAFOBA. Without CAFOBA, the usefulness of IDD and DSLD was, in theory, limited to the development of the first urban project. IDD, due to the dynamism and creativity of its Director created an indispensable service, that of municipal re- gistrar, and has managed to provide itself with significant reven- ues as well. IDD has recently increased its role to that of land developer, taking on the Faladie project. In the long run, the Municipality may be able to carry out the functions that were to be assigned to the independent agency, CAFOBA, through the aus- pices of IDD. - 21 - (ii) The same success has not characterized IDID and DSLD. Unless IDID can increase its revenues to provide for its own costs and gener- ate a significant surplus for the Municipality, it has little justification for existing, and could better be absorbed by the municipal or communal tax collection bureaus. Similarly, DSLD's purpose effectively ended after the first project, with the assignment of lots at Mangambougou. The second project does not assign roles to either DSLD or IDID, despite its repetition of shelter projects and revenue mobilization components. The fact that these two agencies, created by the first project, have out- lived their purpose and now represent an overall cost to the Municipality, must be considered an instance of negative impact. (iii) Ironically enough, the strongest institution to be created by the first project is the project implementation unit itself (PUM), which has proved immensely capable as a general coordinator, an efficient supervisor and implementor, effective at developing new projects and relying on a well-managed project fund (fond d'amenagement urbain). Nonetheless PUM itself has no clear future. 4.08 To a large extent, the first project served to clarify key weak- nesses in Bamako's institutional framework. The organization of a tax collection office, a technical unit for the Municipality of Bamako and a financial management and budgetary planning office, not to mention the further assistance of DSTD in management and waste disposal, were all needs which became apparent during the course of the first project. 4.09 Technical assistance appears to have had a direct bearing on the project's institutional impact. It is noteworthy that the weaker institu- tions created during the project (IDID, DSLD) are those which received little or no technical assistance, and only in program assistance rather than training. The second project emphasizes technical assistance in the form of specific training as a means of institutional strengthening. C. Policy Impact 1. Urban Development 4.10 The project can take credit for several important policy changes in terms of municipal management. For the first time, the organization of property registers became part of municipal affairs, as did the city's assumption of responsibility in urban upgrading and the development of new residential sites. Overall, the project has resulted in the Municipality's current policy of actively fostering urban growth and the improvement of existing neighborhoods. 4.11 Planning policies, as reflected in the ten year investment plan, PDI (Programme Decenial d'Investissement; PCR, para. 4.30), show a slight, but marked, change from those existing at the time of project development. Planning appears to have been moved from a strictly physical/zoning focus, - 22 - to a wider focus on resource mobilization, implementation of serviced sites and providing services to deficient areas. Maintenance and service deliv- ery, however, are still not considered to be a priority for investment. 2. Revenue Mobilization 4.12 The project has also led to the current government policy of charging for basic services, such as water supply, septic tank clean-outs, the registering of deeds, etc. While the Municipality is still reluctant to charge directly for land occupancy or purchase, it has introduced a new tenet requiring that new residents pay the costs of road construction and infrastructure installation. 4.13 Government has also accepted the principle of full cost recovery on land development programs. In doing so, however, it has applied strong- er measures than were considered in project development. Financing terms for the serviced lots at Magnambougou were changed from 20 year to maximum seven year loans (para. 3.08 (iv)). The Government has argued that long term loans are unheard of in Mali and that families would lose interest in paying after a certain time. 4.14 At the same time, the project had little impact on the Municipal- ity's policy of concentrating taxes on a small percentage of its residents. Similar to the municipal tax revenues, the TVEOM presumably finances ser- vices to the entire population citywide, but relies on contributions from only a small percentage. 3. Real Estate and Property Codes 4.15 Although the first project was unable to create a new land tenure law, a new code, proposed in 1986, opened the wqy for subsequent legisla- tion which should permit the majority of Bamako's residents to obtain clear title to their lots. This would represent a significant policy change since title has been the privilege of a small minority until the present. Under the second project, an estimated 15,000 families will be granted full registered titles. D. Poverty Impact 4.16 The serviced lot program at Magnambougou, described in the SAR as a poverty alleviation program (para 5.08, 6.02 and table 2), benefitted a relatively high-income group within Bamako. While no complete income dis- tribution studies of Magnambougou's residents are available. 1985 data indicate that only 17% of the original buyers were below the urban poverty threshold, falling between the 30th and 50th percentile of income distribu- tion, compared with the 85? foreseen in this category at Appraisal. On the other hand, a full 23% of lots were sold to families who declared their earnings to be above the 80th percentile of income distribution. The dis- appointing results of Magnambougou in terms of poverty impact are traceable to the conditions of selection of beneficiaries: - 23 - (i) As nuted, repayment periods were decreased during implementation from twenty years to seven. Monthly payments were correspondingly higher, and less accessible to low-income families. (ii) The number of lots under 170 square meters represented only 65% of the total. Even at the lowest price per square meter, the next range of lot sizes (320 square meters) required a payment of more than 252 of monthly income for any family earning below the 50th income percentile. (iii) All applicants for new lots were required to present documentation in evidence of their salaries. In effect, this limited applicants to persons working in the formal sector, which, according to Bank estimates in 1979, represented only 6.3% of the urban population. (iv) The project unit reported 20,000 applicants when the serviced lots program in Mangambougou was first announced. When the above re- quirements were published the number fell to 3,400. According to the project staff, the worst hurdle was the last (salaried em- ployee). The vast majority of applicants simply had no means to prove their incomes, leaving only those employees of the formal sector. It should come as no great surprise that formal sector employees tend to be: (a) among higher income groups and; (b) government employees or civil servants. Roughly two-thirds of formal sector employment is attributed to the public sector in a Bank study dated 1979. 4.17 In contrast, the upgrading component at Magnambougou did have a significant impact on the lower-income groups of Bamako. At least sixty percent of the beneficiary families are below the poverty line and repre- ser.t about 10Z of Bamako's below-median-income population. The city-wide sanitation program has also benefitted families in lower income groups, in particular those who depend on public standpipes for their water supply and who rely on the septic tank clean-out and garbage collection services. 4.18 The project has created a large number of jobs, both directly ad indirectly. Homes under construction at Magnambougou are two and three stories high, 300-400 square meters of construction, with complete furnish- ings. Sooner or later, some 3,500 such homes will be constructed, each one generating about a year's employment for a team of workers. Also, sponta- neous markets have emerged at the site of Magnambougou's serv?ced lots. Lower-income traders come to do business in the newly developed neighbor- hood. In the final analysis, the Magnambougou development is indeed bene- fitting the poorer families of Bamako. 4.19 New tenure policies set in motion through the project have set a precedent in legalizing occupancy and registering lots which has had a direct and positive impact on Bamako's poorer families. As a result, although spontaneous settlements continued to increase between 1976 and 1986, they are estimated currently to contain only 302 of Bamako's fami- - 24 - lies, 5Z more than in 1976, and far less than the 502 predicted for 1986 (SAR, para. 1.06). Furthermore, the city cannot be said to suffer from peripheral shantytowns. The large lot sizes (para. 4.02) and widely ex- tended services have signified that irregular and spontaneous settlements are provided with adequate drainage, sanitation and public water supply. E. Other Points of Interest 4.20 Local Materials Component. The notion of a special program to promote the use of local materials in Bamako seems like something of a contradiction in terms considering the long tradition of building with stabilized earth systems in Mali and the fact that the majority of the city's structures are already of mud and wattle, adobe, or stabilized earth. Consequently, the use of local materials hardly required introduc- tion as a new technology and one could assume that the residents of Bamako know the costs, benefits, payoffs and drawbacks of using traditional build- ing materials versus cement better than experts from outside. In general, it appears that earth structures are cheaper in the short run, but require very costly annual maintenance. The local materials program proposed to combine earth with cement in such a way as to reduce or eliminate the main- tenance. In so doing, it also reduced the economies originally associ- ated with the use of local materials. In fact, the original study support- ing a proposed local materials program did not suggest that the use of local materials would lower building <: sts, only that it might reduce the need to import cement. When costs of building in local materials cama to four times those of building in cement block (the case of the school for Magnambougou), it became clear to the experts what families in Bamako appeared to have known all along. 4.21 Land tenure. The project emphasized secure land tenure for bene- ficiaries of the serviced lot and upgrading programs, and, in addition, it attempted to make tenure a reality for all of Bamako's residents through the proposed reform in the real estate code. In fact, the IDD (para. 3.14) did reorganize and register over 50,000 occupancy permits throughout the city. To a certain extent, the occupancy permits appear to have fulfilled the same function as legal titles, at least in the minds of the residents. The fact that families came of their own volition, requested and paid for the permits suggests that they took them quite seriously. Ironically, the residents of Magnambougou appear to respect the occupancy permits as much as full titles, to the detriment of the project. Either full title means less to the residents of Bamako than was supposed, or the residency permits mean more. 4.22 Project Costs/Financing. From its outset, the project suffered financial constraints in the form of lack of adequate counterpart funding. As a result, the Bank agreed to change the financing plan so that IDA dis- bursements covered a total of 90% of project costs rather than the 80Z foreseen at project appraisal (para. 3.04 (i)). 4.23 The Economic Reevaluation, similar to the evaluation presented in the SAR, is based on assumed rental values. A simpler reevaluation could - 25 - be done based on the investment, versus real appreciation in sales values of land at Magnambougou. According to project staff, lot prices have in- creased by 300Z for unimproved lots. Like the evaluation presented in the SAR and PCR, the exercise is questionable because a higher rate of return would result from higher income beneficiaries who could afford higher lot prices. Furthermore, the analysis does not take certain important factors into account, such as: (i) The fact that the majority of new lots at Magnambougou are not occupied which renders the notion of a rental value meaningless since there is nothing to rent. (ii) The evaluation is equally meaningless unless compared to some other neighborhood which was not subject to the same improvements. (iii) The evaluation is applied only to the shelter component, which represented but 32Z of total project investment. An evaluation of the municipal sanitation and revenue generation programs would be more to the point, particularly since they represent attempts to make the Municipality economically self-sufficient. 4.24 Sustainability and Replicability. In management terms, the proj- ect established models and agencies for implementing the shelter and urban service components which have been reused successfully in the second proj- ect. Economic sustainability or replicability has been uneven. In gener- al, the project components iovered by direct user charges have achieved adequate cost recovery. The serviced lot project at Magnambougou, in par- ticular, has proved its replicability already. Not only did the revenues from Magnambougou serve as counterpart funds at a crucial moment in the first urban project, but the funds are being applied to the financing of similar activities in the Second Urban operation. Furthermore, the model is being used by the Municipality in its own development projects. The cost recovery at Magnambougou was eased by two elements: the creation of a special fund, the "Fond d'Amenagement Urbainw (Urban Development Fund), unde. the joint auspices of the project manager and the mayor of Bamako, and the short repayment periods introduced during implementation. The modified loan terms worked to the detriment of the poverty alleviation objectives (para. 4.16), but clearly benefitted those of cost recovery. 4.25 The introduction of user service charges for water at public standpipes and for septic tank clean-out services appears to be working well. Unfortunately, the Municipality's accounting does not maintain sep- arate records, so it is impossible to know if, in fact, the revenues indeed cover service costs fully. 4.26 The system of charging for services indirectly, through property taxes, has been less successful. The property tax, cr TVEOM, expressly designed to cover the costs of garbage collection and road and drain main- tenance, has not developed a tax base broad enough to produce significant revenues (para. 3.21). The charges for registry services and occupancy permits, developed by the IDD succeeded in increasing revenues substantial- - 26 - ly from 1985-1987. IDD's base for charging, however, depends on occasional services and cannot provide the sustained income necessary to pay for urban technical services. IDD's revenues dropped by 672 in 1988, following a gradual decline of about 10% per annum since it first began charging foL services in 1985. In other words, IDD generates sufficient revenues to pay for its own services (something IDID has not managed to achieve), but it cannot be expected to continue to maintain the DSTD as well. In general terms, the system of charging for urban services indirectly is weak by virtue of the fact that the charges have little relation to the services offered, and those who pay do not represent the beneficiaries of the ser- vices. A more successful mechanism might copy elements of the direct charge system applied to the water standpipes and septic tank clean-outs. 4.27 The upgrading program at Magnambougou has resulted in a virtual subsidy for beneficiary families (para. 3.19 (ii)). The second project makes no promises, nor any mention, of cost recovery mechanisms for its proposed upgrading component. In view of the difficulties in the first project, it may attempt to broaden the property tax base and to recover upgrading costs through increases in assessed property values, rather than charge beneficiaries directly for upgrading costs. 4.28 Second Urban Development Project. The second project was ap- praised and initiated on schedule. While it picks up several of the activ- ities initiated in the first project, in particular implementation of the land tenure law and maintenance taxes developed under the first project, as well as financial management of the revenues generated by the first proj- ect, it shows a somewhat different focus from the first operation, reflect- ing subsequent changes in Bank policy with regard to a transportation com- ponent, no attempts at employment generation and a much more detailed tech- nical assistance program aimed at training local municipal functionaries (rather than supplementing them as occurred in the first project). The second project concentrates on Bamako exclusively and goes much farther than tae first in terms of expanding the developed area of the city. Un- like the first, however, the emphasis is on secure tenure and low-level services, extended to the city as a whole. V. ROLE OF THE BANK 5.01 Bank influence is clearly responsible for the project's size, composition and complexity. In the words of the first project officer, "It is what we in the Bank call a 'Christmas tree', not because it bears any gifts, but because it has so many things attached to it.01 The "Christmas tree" design appears to have been applied in Mali not only because it was a popular approach to urban projects at the time, but also because it recon- ciled the traditional objectives of urban projects (i.e., poverty allevia- tion and planned growth and development), with the new objectives set forth by the proposed strategy for Sahelian cities, (i.e., bolstering urban self- sufficiency to reduce, or to end, their dependence on national resources). 1/ (From "Magnambougou, Upgrading Project,' in "Designing for Islamic Cultures-50). - 27 - 5.02 In view of the complexities involved, the Bank took a flexible and supportive position throughout supervision, admitting, in effect, that the project represented a pilot and learning experience for the Bank as well as for Mali. On the positive side, the Bank accepted the elimination of certain components which were clearly too difficult to implement such as the credit program and the local materials program which was not continued in the second project. At the same time, it permitted the reallocation of funds into areas which proved to be particularly fruitful such as the de- velopment of IDD and the computerizing of the cities' tax and cadastral bases. 5.03 On the negative side, the Bank ceded perhaps too easily to the changes in the loan terms for Magnambougou serviced lots and on the issue of differential land prices, (para. 3.08) which effectively turned the component from a low-income to a high-income housing program which may lend itself more to speculation than construction. The project also proved too complicated for effective supervision in certain areas, such as the secon- dary cities (PCR, para. 4.19). In hindsight, the 1roject could also have bei.efitted from additional assistance in the area of municipal finance and management of the DSTD. 5.04 Project staff in Mali have expressed their satisfaction with Bank supervision, support and concern given to the first project. The project was extremely complicated, due mainly to the design inputs from the Bank (para. 2.02) and its success depended on careful supervision. VI. LESSONS TO BE LEARNED A. Implications for Proiect Design 6.01 The Mali first urban project provides a clear example of how dif- ficult institutional creation and strengthening can be. Several lessons can be drawn from the project in this respect: (i) The creation of CAFOBA and the disappearance of three newly cre- ated agencies (to be absorbed by CAFOBA) depended on legal reforms over which the project had limited control. In retrospect, it appears to have been ill-advised to tie the project to such an uncertainty. (ii) It is noteworthy that the institutions which can be said to have advanced significantly under this project made use of technical assistance, and those which floundered had only limited assis- tance. (iii) This project could be said to have suffered from an overdose of institutional creation. In designing new institutions, operating costs and benefits should be estimated, and several alternative scenarios for the institutions' future growth or demise should be analyzed to help the local government, which will ultimately be expected either to maintain or to somehow eliminate the new insti- tutions. - 28 - (iv) A final lesson has to do with the inadequacy of the criteria used for evaluating and appraising projects. No handy means exist to measure advances in institutional development, as is the case with physical projects and civil works, unless explicit goals are set forth in project design. B. Implications for Physical Works 6.02 The project reveals some interesting relations between public and private sector operations. The septic tank cleaning service is considered a success city-wide, now that both DSTD and private contractors compete to serve the same market. The same approach might also have been applied to garbage collection and to drainage and street maintenance. It is notable, for example, that the project unit has been more successful at implementing physical works, through the use of outside contractors than the DSTD. 6.03 The project engineer has suggested that the Bank and the consul- tants who prepared the feasibility study accepted the notion of a sophisti- cated garbage collection system without considering the problems inherent in educating the community, in mechanical breakdowns and associated mainte- nance costs. (DSTD estimates that 1/3 to 1/2 of its vehicles are generally out of commission. The service is not diminished, they say, because the remaining vehicles work overtime, which leads in turn to a shorter vehicle lifetime.) A more effective solution might have entailed use of simpler technologies, growing out of local traditions. 6.04 The failure of the local materials component to reduce costs offers a lesson which has, in fact, been repeated in several countries during the past decade, that traditional and appropriate technologies have frequently been discarded for very good reasons. C. Implications for Revenue Generating Programs 6.05 The experience of this project makes it clear that effective rev- enue generation systems frequently depend on local traditions. Custom in Bamako appears to accept the notion of paying for goods or services, hence the acceptance of user charges at public standpipes, for septic tank clean- outs and for deed registry etc. At the same time, the city has tradition- ally resisted generalized taxes, explaining the disappointing results of the TVEOM and of the urban upgrading charges. Effective studies of costs and implications of proposed revenue generation mechanisms could help to avoid such problems in futu,e projects, or at least to estimate the costs involved before embarking on new programs. D. Implications for Urban Development 6.06 Serviced lot projects do not appear to reach lower income groups or to stem spontaneous settlements. Urban upgrading in low-income neigh- borhoods, or the widespread application of measures such as regularizing land tenure or providing public water standpipes, appear to be much more effective ways of reaching the urban poor. In cities like Bamako, overall - 29 - service levels are so low that serviced lots projects will end up being occupied by higher-income groups as a matter of course. The second project has drawn from this experience and is providing 9,500 lots on basically unserviced land. 6.07 Apropos of the serviced lot projects, the attempt to differentiate prices according to income levels has not been analyzed in the posterior evaluations, but it appears to have led to speculation and it has not pro- duced, in any case, the desired effect of increasing the number of low- income beneficiaries. 6.08 The Bank has long supported urban development through serviced lots projects for low-income families. In the first urban project in Mali these goals appear to have been somewhat incompatible. Serviced lot proj- ects are necessarily low-impact, compared to other investments. In conse- quence, it may not suit government agenda to reserve them for low-income groups, as appears to have been the case in Mali. In terms of opening up new areas for urban growth, the Government's approach makes a certain amount of sense. The location of higher-income neighborhoods in new areas provides jobs and spin-off settlements which benefit lower-income groups as well. E. Implications for Municipal Strengthening 6.09 The first urban project va.. not entirely successful in attempting to create new institutions with new functions (IDD, IDID, DSLD, etc.) with- out first considering the Municipality's overall institutional structure. Experience in the financial area, in particular, indicates that a review of the Municipality's entire financial services should have been undertaken before launching into the creation of any new taxes. F. Implications for Urban/National Economic Development Strategies 6.10 The Bank's evolving viewpoints concerning the relation between urban growth and national development may have made it difficult to develop a project which enjoyed genuine government support. The first Bank strat- egy aimed to reduce government spending on the urban sector; the second strategy aimed to increase it insofar as the urban sector generated econom- ic growth for the country as a whele. The project focused on poverty alle- viation and institutional development as measures to foster economic inde- pendence and economic growth. The Malian Government, on the other hand, appears to have been interested in urban development in terms of new resi- dences, new roads which were wide, well-paved and highly-illuminated. Such a viewpoint did not necessarily have to conflict with Bank strategies had the project concentrated on road-building, service delivery, and en- forcing measures to make them economically self-sufficient. But the Bank's attempt to tie urban development to poverty alleviation through the crea- tion of low cost shelter, serviced lots, new technologies, and small-busi- ness loans for low-income families resulted in conflicts throughout project implementation. This is not to imply that the Bank must accept all country loan requests without trying to introduce new ideas. But the project might have been more successful if a basic common ground had been sought at the beginning, as opposed to application of the "Christmas tree" formula. - 30 - VII. CONCLUSIONS 7.01 The project can be considered successful in terms of meeting and surpassing the physical goals established at Appraisal specifically in terms of the serviced lot and upgrading programs at Magnambougou, the civil works introduced in other parts of the city, construction of standpipes and introduction of public maintenance services. It had a significant impact in terms of improving living standards for Bamako's low-income residents. 7.02 The project's secondary objectives, those of insuring replicabil- ity and the maintenance of the project components through institutional strengthening proved more elusive. While cost recovery worked very well in the serviced lot program, its success seems linked to the fact that benefi- ciaries represent a high-income group on the whole. The same success was not achieved in the upgrading programs (which are notorious for the prob- lems they present in cost recovery). While the system of charging for water at public standpipes and septic tank clean-outs seems to be working well and guarantees the replicability of these components, the taxes de- signed to finance garbage collection and road and drain maintenance neither cover the service costs, nor are directed necessarily at the users. Simi- larly, even though new institutions were duly created under the project, as stipulated by the project design, it is not clear that they all contribute to the strengthening of the Municipality. 7.03 On the whole, project implementation went smoothly. The Audit finds the project was justified in extending the date of completion and in reallocating funds. Finally, the project worked well as part of a urban development strategy insofar as it shed light on the real strengths and weaknesses of the Municipality of Bamako and led to the development of the second project. - 31 - ANNEX 1 MALI URBAN I PROJECT COMPARISON OF INITIAL AND FINAL EXPENDITURES (in '000 US$) Component Appraisal Completion (base cost) (final costs) 1. Shelter 5,000 40% 4,820 31.5% 1.1 Upgrading Magnambougou 1.2 Serviced Lots 1.3 Off-site infrastructure breakdowns 1.4 Health Center not available 1.5 Schools 1.6 Construction Loans 2. Sanitary Conditions 2,050 16.5% 2.976 19.5% 2.1 Garbage Collection 980 8% 1.528 9.5% 2.2 Street Drains 170 1.4% } 211 1.12 2.3 Water Standpipes 40 0.3% 2.4 Septic Tank Cleaning n.a n.a 129 0.85% 2.5 Public Toilets 260 2.1% 48.9 0.322 2.6 Main Drainage Outlets 120. 0.9% 284.7 1.8% 2.7 Maintenance Depot 480. 4Z 1,038 6% 3. Informal Sector 530 4.2% 24.5 1.6% 3.1 Markets 260 2Z 213 1.42 3.2 Loans Program 80 0.7% 0 0% 3.3 Local Materials 190 1.6% 31 0.2% 4. Secondary Cities 610 4.92 704 4.6% 4.1 Mopti 4.2 Gao 4.3 Kayes 5. Consultants 2,670 21.5Z 3,426 22.4% 6. Studies 940 7.5% 1,912 12.5% 7. Institutional Assistance 590 4.7% 1.117 7.3% Totals 12.400 99.3 15,300 99.4 - 32 - ANNEX 2 MALI Page 1 of 3 PHOTOGRAPHS 1. Primary infrastructure installations on the road to Magnambougou. (Highway, street lighting, electrical cables.) The serviced lot area can be seen on the right in the distance. Settlements along the road have been stimulated by the Magnambougou project. 2A 4I, 2. The main drainage canals from the Magriambougou site -33- ANNEX 2 Page 2 of 3 3. A spontaneous market on the edge of Magnambougou. 4. Magnambougou's serviced lots seen in the background 34 ANNEX 2 Page 3 of 3 5. An extreme example of high-cost housing going up on Magnambougou's serviced lot site. 6. The data center, financed under the project's technical assistance c-omponent. - 35 - REPUBLIC OF MALI FIRST URBAN DEVELOPMENT PROJECT (Cr. 943-MLI) PROJECT COMPLETION REPORT August 31, 1988 Infrastructure Division Sahelian Department Africa Region - 36 - REPUBLIC OF MALI FIRST URBAN DEVELOPMENT PROJECT (Credit 94S-MLI) PROJECT COMPLETION REPORT I. INTRODUCTION A. Project Background and Objectives 1.01 In 1978, when the First Urban Development Project was prepared, Mali already ranked among the least developed countries, with a GDP per capita of the order of US$120. With a principally agriculture-based economy, Mali (like its Sahel neighbors) had been repeatedly hit by severe droughts through the 70s. The population of 6.6 million was growing at an average annual rate of 2.62 and about 17% of them were living in urban areas. Growth of this urban population evolved around 7Z a year. Around 40Z of all urban residents were concentrated in Bamako, the capital city. Most of the urban population lived under precarious conditions, suffering from rampant underemployment, inadequate housing in areas with little, if any, infrastructure, and poor sanitary conditions. Avefage 1978 income level of urban households was about CFAF 18,500 (MF 37,000), equal to US$84. 1.02 The First Urban Project in Mali focused on a number of key Issues raised by the rapid growth of the country's urban centers considered to be one of the bases for Mali's future economic development. In Bamako, it attempted to halt the rapid deterioration of urban conditions through a series of activities aimed at upgrading basic urban services and providing appropriate responses to the needs of the low-income urban population in the areas of housing, water supply, solid waste collection, primary education and health care.- In three secondary centers, it was to provide and improve essential urban servicess solid waste collection and water supply. 1.03 The present report reviews the history of the project, examines the results achieved through its implementation and highlights its successes, its problems and, where appropriate, its failures. It is based on the information contained in: (a) the Project Appraisal Report (no. 2444-MLI) of May 21, 1979; (b) the quarterly progress reports of the Mali Urban Project Unit (PUH); (c) the Back-to-Office and Aide-memoires of the World Bank supervision missions; (d) the reports on studies and consultant missions carried out during the course of the Project; and (e) data collected by a PCR preparation mission which visited Bamako in July 1987. For the convenience of comparisons and intelligibility, all monetary figures are expressed in CFAF, the Malian currency since August 1984. The previously used Malian Franc (MF) had a fixed parity of 1 CFAF - 2 MF. - 37 - II. PROJECT IDENTIFICATION, PREPARATION AND APPRAISAL A. Identification and Preparation of the Project 2.01 The First Urban Development Project originated in a request made by the Government of Mali in early Fall of 1976 for the preparation of an utban development program. In-response to the Government's request, an identification mission was carried out in October of the same year. 2.0i In parallel to a first preparation mission in March 1977, a household survey was conducted in Bamako to assess the living conditions of the population of the capital. The terms of reference of a project feasibility study were drawn up at the same time. This study was financed under a PPF*of a total amount of US$480,000. and carried out b.:tween December 1977 and July 1978. Four other preparation missions were undertaken between June 1977 and April 1978. Preappraisal took place in July 1978. a. Appraisal 2.03 The Project was appraised in September 1978. Its objectives were, Accorditg to the BAR, "to assist the Government in providing affordable urban services, particularly shelter, water supply, garbage collection, drainage, schools and health care to low income urban residents'. .To ensure rdplicability and maintenance of project components, cost recovery was to be "a major goal, especially for shelter and mnicipal services'. Specific objectives of the project were "to: (a) provide shelter services and related community facilities to meet about half of the growth demand for low income groups in Bemako; (b) improve health and sanitary conditions in Bamako; (c) stimulate productive employment through financial assistance to the informal sector and the promotion of local building materials; (d) provide essential water and sanitation services to secondary cities; and (e) establish and strengthen institutions in the urban sector.* To achieve these objectives, the project comprised the following components: (a) Shelter Improvement (Bamako): Upgrading of the low-income settlement of Magnabougou and construction of serviced plots on an adjacent site, c3nstruction of educational and health care facilities, and provision of construction credits to plot allottees; (b) Improvement of Sanitary Conditions (Bamako): Provision of vehicles and equipment as well as construction of a maintenance P facility to the Bamako District Technical Service Department (DSTD) for solid waste collection, maintenance of drainage ditches, repair of existing and installation of new public standpipes, and construction of public toilets; (c) Assistance to the Informal Sector (Bamako)t Construction of neighborhood markets, provision of credits to artisans and promotion of local building materi4s; - 38 - (d) Essential Urban Services in Secondary Cities (Gao, Kayes, Mopti): Supply of equipment for sewage disposal and solid waste collection to the municipality of Mopti. and extension of water supply systems in Gao and Kayes; (e) Technical Assistance for Project Implementation: Technical assistance to the Project Unit (PUM) and project supported Bamako District Services; (f) Studies: Consultant services for: (i) feasibility study, (ii) ten-year investment programming study, and (iii) preparatory studies for a second project; and (g) Assistance to Institutionst Supply of vehicles and equipment to: (i) the Project Unit, (ii) the Land Management Agency <CAFOBA), and (iii) the District Tax Inspectorate (IDID). 2.04 Total project cost was evaluated at CFAF 3.36 billion or US$15.3 million--of which an estimated US$7.3 million in foreign exchange (see Annex 1). The IDA credit of US$12.0 million covered 100Z of foreign exchange *costs and 702 of local currency costs (excluding taxes). Subsidiary credit agreements were concluded with the Bamako District (CFAF 1350 million or US$6.1 million) and the Municipality of Mopti (CFAF 52.5 million or US$240,000). The balance, mainly funding for non-recoverable costs (off-site infrastructures, healthcare centers, schools) as well as for the Kayes and Gao subprojects, was provided as grants. Funding for the technical assistance and studies, as well as the Government contributions to the Bamako District and Mopti subprojects were passed on as grants. 2.05 Overall responsibility for project implementation was entrusted to the Ministry of Interior. In view of the weakness of this ministry and the lack of familiarity of its staff with the implementation of technical projects, a specific Project Unit was established. This unit had to monitor and coordinate the activities of the various executing agencies responsible for the implementation of the subprojects: the Bamako District, the Municipality of Mopti, the Regional Water and Energy Directorates in Kayes and Gao (Ministry of Industry), and the Directorate of Urban Planning (Ministry of Public Works) (see Annex 2). 2.06 The project implementation was expected to take four years (from mid-1979 to mid-1983), with credit closure scheduled for December 31, 1983. - 39 - III. PROJECT IMPLEMENTATION A. Credit Effectiveness and Proiect Startup 3.01 The Credit was approved by the Board on June 26, 1979. and the Credit Agreement was signed on November 7, 1979. The following were conditions of effectiveness of the credit: (a) conclusion of two subsidiary credit agreements, between the Government and (M) the Bamako District and (ii) the Municipality of Mopti; and (b) opening of a Special Account for the Project Unit, with an initial deposit of CFAF 50 million (MF 100 million). 3.02 The Credit entered into effect on June 30, 1980, i.e., a full year after its approval by the Board. The reason for the delayed effectiveness was the Government's incapacity to provide the funding for the initial deposit on the Special Account. B. Revisions 3.03 During the period.of project implementation, the value of the US$ appreciated considerably and the US$/CFAP(MF)*exchange rate rose from CFAF 220 in 1979 to CFAF 512 et its peak in early 1985. The average rate over the total implementation period was CFAF 356, 622 higher than the appraisal level. The rise in the value of the US$ increased the CEAF equivalent of the credit amount from CEAF 2.6 b.Ilion to CFAF 4.6 billion by end of 1983. This substantial increase was beneficial insofar it allowed to expand various project components and to develop new activities. Its adverse effects were that it contributed to a significant lengthening of the project implementation time from four to seven years, and also severely strained the Government's capacity to mobilize counterpart funding. To alleviate the persistent shortfall of such funding the Credit Agreement was amended in 1984, with an increase in IDA funding for construction works (Category 1), from 70Z to 902. 3.04 The following new activities were undertaken as a result of the increase in available financing: - provision of supplemental solid waste collection and drainage maintenance equipment for the Bamako District Technical Service Department; S- preparation of final design and bidding documents for the Second Urban Project; - prolongation of technical assistance contracts to the end of 1986; - organization of three training seminars for Mali professionals: -40- - Financial management of municipalities (Oct. 1984) - City cleaning and garbage collection (Nov. 1984) - Preparation of urban projects (Sept. 1985); - preparation of a new National Land Tenure legislation (Code Domanial et Foncier). 3.05 The increase in available funding notwithstanding a number of project activities had to be cancelled for a variety of reasons; - te Construction Credits (CFAF 302 million) and Artisan Credits (CFAF 21.5 million) components because of implementation difficulties; - the Bamako Land Development Fund (Caisse Foncibre de Bamako-- CAPOBA) owing to persisting disagreement between the Government and the World Bank on the operational objectives and the legal framework of the proposed component; - the Gao water supply counterpart for which between project appraisal and credit effectiveness, the Mali Government secured financing fr9m the French FAC (Fonds d'Aide et de Coopdration). pI 41 - IV. PROJECT IMPLEMENTATION PERFORMANCE 4.01 The First Urban Project was a relatively complex and, in some respects, overly ambitious undertaking. While the implementation of the physical works components was generally carried out expeditiously, the large number of project components, their diversity, both physical and institutional, and the resulting profusion of implementing agencies overtaxed in many respects the absorptive capacity of the then--and still now-- excessively weak institutional environment. Besides, in hindsight, it is not unreasonable to conclude that the Government's actual commitment to the Project's institutional objectives wae significantly weaker than it had been presumed during project preparation. It is, however, difficult to formulate a global judgement, and in particular, it would be unfair, by underscoring the project's negative results, to generate the impression of an all-out failure. 4.02 One of the undeniably positive aspects of the project was that for the first time in Mali, an attempt was made to tackle a series of critical urban development problems in a comprehensive and systematic way. Indeed, the project tried to address simultaneously problems in the areas of: - housing conditions and urban land management through the implementation of neighborhood upgrading and sites and services programs; and the development of corresponding sector institutions; - improvement of sanitary conditions through the construction of drainage infrastructure and water supply systems, and support to solid waste collection services; - infe .lL sector development through the provision of credit lines for aL.isans, the construction of markets, as well as.support to the building industry through the promotion of local construction materialsi and - local resource mobilization through the creation of new and the reform of existing taxes and the strengthening of the administration entities in charge of their management. 4.03 Essentially to make good some delays incurred in plot allocations and to allow for the implementation of the complementary work programs made possible by the increase in available funding (see 3.03 supra), the closing date, initially set for December 31, 1983, was extended three times and the proj ct was actually completed on December 31, 1986, while the last disbursement took place on March 10, 1987. A. Improvement of Shelter Conditions 4.04 . Operations in the housing and land development sector formed the most important component of the project (over 40Z of base cost), and included in terms of physical works: - 42 - - the rehabilitation of the Magnambougou village, a squatter settlement of 12,000 inhabitants; - the implementation at the same location of a sites and services operation covering an area of 100 has and comprising a total of 2,600 plots; - the construction of corresponding off-site infrastructure; - the construction of schools and health care facilities; and in terms of institutionalifinancial objectives: - the provision of housing credits for beneficiaries*of the Magnambougou operations; - the strengthening of the Bamako District's Housing Directorate (Direction du Service du Logement du Listrict - DSLD), responsible for the management of the Magnambougou operations and their possible subsequent extensions. - the establishment of a Land Management Agency (CAFOBA) for the Greater Bamako area. 4.05 The implementation of the physical parts of this program was largely successful. The institutional achievements remained modest, however. The housing credit scheme proved to be non-operational and had to be cancelled. Likewise, the establishment encountered severe difficulties and was eventually abandoned. 4.06 Both the Magnambougou village rehabilitation and sites and services operations were implemented in a timely fashion and in accordance with agreed standards. As a result of internal reallocation of funds, the sites and services operation could be significantly expanded during implementation, and the number cf plots brought from initially 2,600 plots to finally around 3,600 plots. Th& village rehabilitation program in turn benefitted from a complementary assistance from FAC and could also be expanded. 4.07 At the implementation start-up of the sites and services operation, it was decided that around 85% of all plots should be of-a size not exceeding 160 sq.m. However, such small-size plots did not match the actual demand and as sales progressed, number of plots were merged. As of 1985, the proportion of small-size plots had descended to about two-thirds. Reportedly, the merging process is still ongoing, bct no precise data on the current status are available at this point. 4.08 Over 952 of the plots were sold on a lease-purchase basis with payment periods ranging from three to five years with a two-year grace period. About 5% of the plots destined for commercial occupation were sold directly. Plot allocation and .sales suffered a delay of about two years due to the Government's slowness in transferring title rights for the Magnambougou area - 43 - to the Bamako District. At the close of the project,, 3,015 deeds of sale had been registered and a total of 1,117 acquisitions had been completed. 4.09 The initial target set for the distribution of new plots was that 852 of the benefitting households would have an income inferior to the median income estimated at CFAF 19,250/month in 1979. This target was reasonably met, since according to a 1985 consultant report. 77% of the beneficiaries had a declared income of less than CFAF 47,500/month, the estimated median income for that year. The same report also noted, however, that the objective of extending the access to plot ownership to all segments of the low-income - population had been achieved only marginally, since 90Z of all purchases were civil servants or private formal sector employees. 4.10 The actual occupation of the Magnambougou sites and services a:ea took place at a noticeably slower than anticipated pace. At appraisal, it had been--somewhat over-optimistically--estimated that the majority of plots would be occupied at the end of a three-year period. A sample survey carried out in early 1986, a few months before project completion revealed that construction of houses had begun on only one third of the plots and that on less than 5Z it was nearing completion. (Construction startups and completions are said to have accelerated since 1986, but no reliable supporting data are available at this point.). 4.11 Though no in-depth inquiry has been made, the principal reason for the slow progress of site occupation appears to have been the financial difficulties experienced by many of the owners. Construction costs proved to be markedly higher than estimated initially. Based on surveys of the traditional construction industry sector undertaken during project preparation, it had been presumed at appraisal that most houses would be built with local materials (adobe bricks) and under self-help arrangements at a unit cost not exceeding CFAF 250,000. In reality, however, use of local materials was spurned by the owners for, according to a 1985 survey, more sociological than economic reasons. Moreover, it appears to have been on various occasions openly discouraged by the Malian administration. For instance, pre-authorized building permit plans, a scheme devised under the project to cut down on the usually lengthy and cumbersome permit approval procedures, called for investments of over CFAF 5 million and advised use of imported, so-called modern building materials. To expedite the construction process, the provision of housing credits to plot purchasers had been included in the project. The credit system never became operational as no agreement on its management procedures could be reached with the project authorities. 4.12 Relatively mixed were the results in regard to the institutional objectives of the housing and land development component. The administrative management of plot allocation and cost recovery for the Magnambougou operations was ensured fairly efficiently by DSLD, the Bamako District entity strengthened to this effect. Efforts to rationalize the market for urban land through the establishment of land management for the Greater Bamako area failed. According to the Credit Agreement, the Bamako Land Management Agency (Caisse Fonci6re de Bamako - CAFOBA) was to be created in a three-step process and completed on June 30, 1982. No agreement could be reached with the Malian authorities on the'precise modus operandi of CAFOBA and its implementation was -44- eventually withdrawn from the project. The lack of s.ccess in the creation of CAFOBA was at least partially offset by an unforeseen expansion of the activities of the District's Directorate of Domain's (IDD), and the preparation of a new land tenre law. Initially, intended primarily to support DSLD in the legal aspects of the management of the Magnambougou operations, IDD extended rapidly the scope of its role. It undertook an extensive reorganization of the District's entire land tenure administration by computerizing land title and occupancy permit registers. In this connection, it was also instrumental in creating new fiscal resources for the Bamako District (see (4.26 infra). The preparation of a new land te*nure law (Code Domanial et Foncier) was begun in 1982, and the law officially adopted by the Government on July 12. 1986. B. Improvement of sanitary Conditions 4.13 The project included components intended to help improve the citywide sanitary conditions in Bamako as well as--after the withdrawal of the Gao component--in the two secondary cities of Kayes and Mopti. 4.14 In Bamako, the improvement program included strengthening the solid waste collection services, clearing and repairing existing drainage ditches and building new stormwater channels, building public toilets, installing and repairing public standpipes, and constructing a maintenance facility for the District Technical Service Department (DSTD). 4.15 The results of the efforts undertaken under the project to improve the citywide sanitary conditions must be considered as rather poor. Thanks to the supply of an important lot of equipment and technical assistance, the project was successful in improving temporarily the performance of DSTD's solid waste collection services, as between 1980 and 1985, the volume of daily collected waste doubled and the coverage rate rose from 30% to 50%. However, most of the equipment was worn out at the end of the project, but, because of a shortfall of funds (see infra 4.24) could only partially be replaced. 4.16 No tangible and lasting results were achieved in the area of drainage maintenance. While the project foresaw the maintenance and repair of about 250 kma of ditches, a first campaign in 1981 allowed the clearance of about 73 kms. Subsequent campaigns managed directly by DSTD, failed to gain momentum, and over the following three years, only a modest total of 12 kms of ditches received maintenance attention. A last but largely unsuccessful attempt to improve the performance of DSTD in this area was made in 1986 through the purchase of complementary equipment and the financing of operating expenses from project funds. A stormwater outlet of 2.7 kms was built in 1981182 in the Niarela neighborhood by DSTD force account. Maintenance * stopped in 1984 with the breakdown of the bulldozer provided by the project to this effect. 4.17 The construction of a new maintenance facility for DSTD was an undertaking fraught with problems from the outset. Inadequate evaluation of site conditions resulted in the need for important additional landfill entailing delays and cost overruns. The liquidation of the lead construction enterprise due to the death of its owner created additional problems. - 45 - Primarily because of several defects on works executed by the electrical contractor, the facility was still not operational at the close of the project. (It got eventually completed in 1987 under the Second Urban Project.). 4.18 The repair of existing standpipes and the construction of new ones throughout Bamako was a fairly successful operation. About 40 new standpipes were installed and around 140 existing ones repaired. The most salient feature of this operation was, however, that it led the Malian authorities, both administrative and political, to accept the principle of direct cost recovery of water consumption at public distribution points. Accordingly, the operation of standpipes was farmed out to individual operators ensuring water distribution at an agreed upon tariff. 4.19 The interventions in secondary cities included the supply of solid waste collection equipment to the Municipality of Mopti and the extension of the water supply network in Kayes. Insufficient concertations between project management in Bamako and the Municipality of Mopti resulted in the purchase of equipment which upon arrival in that city proved to be unusable under the local conditions (e.g., the container trucks were too large for most of the streets), and consequently remained idle for several years. The construction of the water supply network extension in Kayes suffered delays and cost overruns due to technical insufficiencies of the design studies as well as the need to relocate unauthorized housing developments that had sprung up in the meantime on the construction-site. It was otherwise satisfactorily completed. C. Support to Informal Sector 4.20 Support to the informal sector was to be provided uAder three programs: (a) the construction of four neighborhood markets; (b) the provision of construction and equipment credits for artisans; and (c) the promotion of local construction materials. Aside from the construction of markets, none of these components brought tangible results. 4.21 The construction of four neighborhood markets was satisfactorily completed. Contrary to initial plans, the construction of stalls for traders and artisans did not take place, however. 4.22 The provision of credits to artisans for the construction of stalls and shops (50 loans at CFAF 135,000) and acquisition of equipment (400 loans at CFAF 27,500) was shelved principally because of implementation difficulties and redundancies with cimilar programs carried out simultaneously under other Bank projects. (Industrial Sector Development Project - Cr. 986-MLI.). A 4.23 The promotion of local construction materials, especially adobe bricks through the introduction of improved manufacturing techniques turned into a by and large fruitless exercise. Recruitment of suitable experts for which assistance was obtained form UNCES, proved to be difficult. Moreover, the contract of the finally hired expert was terminated before expiration. Aside from some test-walls, the only other demonstration object built was a 24 classroom school at Magnambougou, which was completed with assistance from - 46 - ADAOUA, a local NGO specialized in using local materials, but at a cost reportedly exceeding by far that of a construction with conventional materials. D. Local Resource Mobilization 4.24 An important objective of the project was to ensure an effective cost recovery not only for the Magnambougou rehabilitation of sites and services operations, but also for the delivery of basic urban services (solid waste collection and drainage maintenance). To this end, it was agreed to create a specific municipal maintenance tax (taxe de voirie et d'enl6vement des ordures - TVEO). The management of this tax was entrusted to a newly- created entity, the District Tax Inspectorate (Inspection Divisionnaire des ImpOts du District - IDID). TVEO was to be assessed on the rental value of all residential and commercial properties located within the District and expected to yield, even assuming a recovery rate of only 60, sufficient revenues fcr funding of future operating and capital costs of solid waste collection and drainage maintenance. 4.25 Actual results fell far from expectations. Taxation was limited to about one-fifth of the total properties. As shown below, no revenue was produced in 1982 and 1983, and recovery rates for the following years oscillated between 152 and 262. Proposals made by IDID to readjust other existing taxes or to improve their recovery conditions were objected by the Government and, in particular, the MiAistry of Finances. 4.26 As mentioned earlier (para 4.12 supra), a main achievement of IDD was the reorganization of the District's land title and occupancy permit administration. This reorganization also provided the opportunity for a thorough revision of all service fees related to these transactions and made them become the second most important source of revenue of the Bamako District. 4.27 In percentage of the total District income, the contribution of IDID and IDD resources evolved as follows: 1980 1981 1982 1983 1984 1985 1986 (in CFAF million) 1. Total Distr. Income 488.3 607.7 917.9 608.5 761.7 1022.4 1036.7 2. IDID Receipts - - - - 16.8 34.8 33.5 3. IDD Receipts 2.5 27.5 87.6 25.9 120.7 143.5 109.4 2 + 3 in I of 1 0.4 4.4 9.5 4.1 18.0 17.0 13.0 E. Technical Assistance. and Training 4.28 Technical assistance was provided under the project to the Project Unit (PUM) as well as to IDD and DSTD, two of the four District Services supported by the project. Turnover was relatively moderate. The actual total number of man/months financed under the project was 330m/m compared to 220m/m expected at appraisal. The reason for this difference lies mainly in the - 47 - extension of the project implementation period from three to seven years (para 2.03 supra). 4.29 Aside from on-the-job training of Halian counterparts by the technical assistance experts, specific training activities were organized by the project in form of seminars for administrative and technical staff of Malian municipalities: - Financial management of municipalities (October 1984) - City cleaning and solid waste collection (November 1984) - Preparation and execution of subprojects (September 1985). Finally, the project also funded the attendance by Malian project staff to various overseas training seminars or individual training courses.. F. Studies 4.30 Aside from the PPF funded project preparation studies, the project financed the preparation of three other major studies: * New Land Tenure Law (see 3.04 supra) - Ten-year Investment Study for Bamako - Second Urban Project.. G. Procurement 4.31 All contracts were awarded in conformity with Bank/IDA procedures either by ICB in the case of large civil works and consulting services or by LCB for smaller works and equipment and supplies purchase contracts. Procurement management by the Project Unit was efficient and expeditious. A total of 60 contracts were awarded of which about 20 were complementary contracts (avenants). Out of a total of 16 civil work contracts, the three largest ones were awarded to internal contractors, and 13 were awarded to local formal (3), and informal (10) enterprises. H. Costs 4.32 The comparison between appraisal estimates and actual costs is shown in Annex 1. When measured in US$, actual costs were only 90Z of appraisal estimates. Because of the depreciation of the local currencies (FH and CFAF in 1984) against the dollar, however, projects costs in CFAF exceeded appraisal estimates by about 45%. I. Disbursements 4.33 Actual disbursements against the Credit by quarter compared with appraisal estimates are given in Annex 3. Disbursements took place over eight years compared to the appraisal estimate of four years. Actual disbursements from the various categories of the credit compared to appraisal estimates are also shown in Annex 4. - 48 - J. Performance of Consultants and Contractors 4.34 The consulting firm which carried out the feasibility study was competent and produced reports of high quality. The same firm subsequently carried out the ten-year investment study and the preparatory studies for the Second Urban Development Project. The overall performance of the foreign experts recruited under the technical assistance component was mixed and, in some cases, the effectiveness of the experts was affected by personal conflicts. 4.35 The overall performance of the contractors and suppliers was satisfactory. Major problems occurred only in connection with the construction of the DSTD maintenance facility. Following the death of its owner, the lead construction company was liquidated and had to be substituted by another company. Defective installation of electric equipment resulted in legal suit against the contractor. K. Reports and Audits 4.36 Progress reports were submitted in a usually timely manner. 4.37 An independent auditor was retained by the Project Unit and audited project accounts were prepared regularly through 1986*. Aside from some difficulties encountered at the beginning--which led to the hiring of L professional accountant--project accounting was carried out satisfactorily throughout the entire life of the project. - 49 - V. INSTITUTIONAL, FINANCIAL AND ECONOMIC PERFORMANCE A. Institutional Aspects 5.01 One of the main objectives of the project was to strengthen the Bamako District's overall urban management capacity, both technical and financial. Project support to achieve this goal included provision of technical assistance, supply of equipment, and training to four Directorates (IDID, IDD, DSLD, and DSTD). It also included, in the case of DSTD, construction of facilities. Despite some undeniably positive aspects, it must be admitted that by and large, hardly any of these institutional objectives could be achieved. The project's impact on the improvement of the District's financial resources, though not negligible, remained far from the targets (and in fact resulted from interventions that had not been envisaged by the project). While the management of the Magnambougou operation was carried out satisfactorily, the shelving of the CAFOBA component prevented the land development process initiated by the project to gain momentum and to expand beyond the confines of the project area. The failure to attain any lasting improvement of the resource basis for basic urban services made it impossible to develop the quality of DSTD's service delivery, and thus jeopardized the efforts undertaken to achieve a more than ephemeral change in citywide sanitary conditions. 5.02 There are several reasons for the project's relatively poor performance in regard to its institutional objectives. A large part of it must be attributed to the novelty of some of the concepts which the project attempted to introduce. Another important factor was that the diversity of components and the ensuing multiplicity of executing agencies, severely overtaxed the Mali administration's absorptive capacity, particularly in respect to inter-agency coordination. Last, as noted earlier, the Government's actual political commitment to some of the project's institutional objectives, such as the strengthening of the Bamako District's resource base, was significantly weaker than it had been anticipated at appraisal. 5.03 Whether or not the very implementation arrangements contributed to the institutional shortfalls is a question that would warrant to be examined more closely. As outlined earlier, the Bamako District, although assuming under the project agreement legal and financial responsibility for almost half of the project, did not directly participate in the coordination and supervision of the project. Instead, these tasks were entrusted to a Project Unit (PUM) that operated on a largely autonomous basis and reported directly to t4e Minister of Interior, the Bamako District's tutelary authority. While under the then prevailing circumstances this triangular arrangement was probably the best possible sclution to guarantee a satisfactory project implementation, its trade-off was that it created a non-negligible potential for personal rivalries, let alone conflicts between PUM and other authorities. - 50 - B. Financial Performance 1. Counterpart Funding 5.04 The timely availability of counterpart funding proved to be a permt-int problem over the entire implementation perioi. The problem was rendt-ed even more acute by the continuous appreciation of the USS and hence, the increase of available IDA funding in CTAF equivalent. The difficulties were partially overcome by a readjustment of the IDA contribution rates that took place in July 1984 and increased the coverage of category I expenditures by the Credit from 702 to 90Z. Another significant portion of the counterpart funding needs was met by drawings on the Urban Development Fund (Fonds Special d'Amdnagement Urbain), a fund created under the project for the purpose of ensuring the replicability of land development operations (see 5.07 infra). The total amount of such funding diverted from its original purpose and the reimbursement of which by the Government, CFAF 170 million, was still outstanding at the end of the project. (This practice was continued under the Second Urban Project and by end of June 1987, the total amount of counterpart funding withdrawn from the Urban Development Fund was CEAF 270 million.). 2. Cost Recovery 5.05 At appraisal, it was estimated that around 60 of the project cost would be recovered either through municipal taxes (in the case of urban services) or cash repayments (in the case of housing and land development operations). 5.06 Effective cost recovery for urban services proved to be an illusive objective. A municipal tax for solid waste collection and drainage maintenance (TVEO), specifically created under the project, was expected to create revenues sufficient to cover both capital and operating costs of these services. Results were disappointing, however. Although created in 1981, it did not become operational until 1985 at rates not exceeding 252. Compared to the combined actual annual expenditure by DSTD for road and drainage maintenance and solid waste collection, TVEO revenue represented only around 5Z. Even if compared to the sole cost of solid waste collection, it represented about 40Z of total annual operating expenditure, at best (see Annex 5). 5.07 Cost recovery for the Magnambougou sites and services operations takes place through direct cash repayments by purchasers. At the end of 1986, out of total amount to be recovered of CPAF 1313 million, CFAF 677, or 512, had already been recovered. The entirety of the credits is expected to be recovered by 1991. Cost recovery for the rehabilitation operation was notieeably less successful. In late 1985, only about CFAF 12.5 million out of total estimated at CFAF 280 million had been recovered. The main reason for this shortfall was that a majority of the occupants felt no need to obtain a land title, as they thought that their tenure had been secured enough by the rehabilitation works themselves. 5.08 Proceeds of "?st recovery were deposited in a special account, the *Fonds d'Amdnagement Urbain. As mentioned earlier, withdrawals from this 51 - account were used repeatedly to substitute for counterpart funding due by the Mali Government. While using the account for the purposes of prefinancing the Government project contributions was explicitly authorized by the legal order (Arrft4 interministdriel) 4008/MF/CAB of December 7, 1983, it was also understood to be a temporary solution pending the corresponding payments from the Government. C. Economic Justification 5.09 In reevaluating the economic rate of return for the Magnambougou. rehabilitation and sites and services operation, the same methodology used at appraisal was utilized for the analysis. The actual costs of site preparation, on-site and off-site infrastructure, building materials and labor expenditure for houses, were included in the analysis. Also included in the cost stream were the future maintenance costs estimated at 2.5Z of the investment costs (estimated at 3Z at appraisal). Costs and benefits were considered over the 1980-2004 period (see Annex 6). 5.10 Benefits were based on imputed rental values of the units. The average monthly rental value was estimated at 1.42 of the total investment for the sites and services operations and at CPAP 3000 per room (value 1987) for the rehabilitation component. On the basis of this assumption, the economic rate of return was calculated to be 172 compared to 222 at appraisal. By component, the rates were 26? for the rehabilitation operation, and 15% for the sites and services operation compared to 262 or 102 respectively at appraisal. The decrease of the ERR for the sites and services operation is mainly due to the relative low elasticity ef rents over investments in the low and middle income segment of the housing market. - 52 - VI. THE ROLE OF THE BANK 6.01 Bank supervision was relatively continious and only one change in the lead project officer occurred during the life of the project. During implementation, a total of 16 supervision missions visited Bamako. The number of staff-weeks per year spent in supervision is shown below: 1979 1080 1981 1982 1983 1984 '1985 1986 4 1.6 21.2 9.2 9.6 8.8 5.4 3.6 VII. CONCLUSIONS 7.01 For a first intervention in an institutionally weak environment, the First Urban Development Project was a fairly ambitious operation. While it was successful in introducing new concepts for land management and cost recovery for sites and services, it failed to achieve major breakthroughs in the area of urban service management and improvement of sanitary conditions. 7.02 Overall implementation was in general expeditious and the lengthening of the implementation time from four to seven years stemmed to a large extent from the increase in available funding that resulted from the .US dollar appreciation against the local currencies. ref.mlipcr.wp August 31, 1988 - 53 - ANNEX 1 REPUBLIC OF MALI FIRST URBAN DEVELOPMENT PROJECT (Cr. 948-MLI) Project Costs: Appraisal Estimate, and Actual Costs - (CFAF Million -------- Component: Appraisal Actual Difference x A. Shelter Improvement Basako 1398.5 1387.2 -31.3 (2) 8. Sanitary Conditions Improvement Bamako SS.6 886.3 +320.7 67 C. Informal Sector Assistance Bamako 162.3 166.6 +14.8 9 D. Secondary Cities Services 157.7 215.7 +6.7 37 E. Technical Assistence 694.4 1201.4 +607.0 78 F. Studies 23.5 612.7 +374.2 157 0. Institutional Support 153.9 395.0 +241.1 168 Total 3380.9 4844.9 *1484.0 44 (US$ Million Components Appraisal Actual Difference x A. Shelter Improvement Sasako 6.3 4.0 -2.3 -37 B. Sanitary Conditlos Improvement Bamako 2.9 2.6 -0.3 -10 C. Informal Sector Assistance Bamako 0.6 0.4 -0.2 -33 0. Secondary Cities Services 0.74 0.7 - - E. Technical Assistance 3.1 8.3 +0.2 +8 F. Studies 1.0 1.6 +0.6 +6 0. Institutional Support 0.7 1.2 +0.5 .7 Total 15.3 13.3 -1.5 ref. projcost.wp - 54 - ANNEX 2 REPUBLIC OF MALI FIRST URBAN DEVELOPMENT PROJECT (Cr. 943-MLI) PROJECT COMPLETION REPORT Implementation Arrangements Z of Total Executing Agency Component Proiect Cost Minister of Interior Technical Assistance Project Unit. 23.0 District of Bamako Regional Directorate of Supervision of on-site Habitat infrastrucure works, schools. health facilities and markets. 29.3 Regional Directorate of Water Off-site water, Resources and Energy electricity and drainage, main drainage outlets of Niarela. 5.2 District Inspectorate (IDID) Establishment and District Directorate of Domain management of new municipal taxes, land registration. establishment of Land Management Agency. 3.3 (i) District Technical Reorganization of Services Directorate (Voirie) District services, (DSTD) garbage collection. clearing and repair of drainage ditches, repair of standpipes. public lavatories and off-site roads. 15.6 (ii) District Housing Pr-selection of Directorate (DSLD) beneficiaries, recovery of plot charges, building loana and credit to artisans. 9.5 Municipality of Mopti Sewage and garbage collection. 1.7 Directorate of Water Resources Water supply, Gao. and Energa (th Region) 1.0 Directorate of Water Resources Water supply, Kayes. 2.2 nd Energy (1st Region) National Directorate of Project preparation, Habitat- Ministry of Public investment programing Works study and second project study; and development of local building materials. 9.2 Total 100.0 ref. agency.dp - 55 - ANNEX 8 REPUSLIC OF MALI FIRST URAN DEVELOPMENT PROJECT (Cr. 943-MLI) PROJECT COMPLETION REPORT Estimated and Actual Querterly Disbursemens (in USS) IDA FY QUARTER APPRAISAL ESTIMATES ACTUAL quarterly Cumulative N Quarterly Cumulative x 1930 Seond 880 680 6 Third 250 880 7 Fourth 200 1,130 10 610 610 6 1981 First 420 .1,600 13 282 892 7 Second 600 2,200 18 798 1,690 14 Third 750 2,950 25 136 1,878 16 Fourth 900 3,850 32 119 1,997 17 1982 First 1,100 4,950 41 189 2,166 18 Second 1,800 6,250 62 538 2,724 23 Third 1,500 7,750 65 1,6 7 4,391 37 Fourth 1,400 9,150 76 449 4,840 40 1983 First 1,200 10,360 88 328 8,168 43 Second 900 11,250 94 841 6,009 50 Third 450 11,700 98 600 6,609 55 Fourth 300 12,00 100 857 7,466 62 1984 First 298 7,764 6s Second 276 8,C3s 67 Third 856 8,392 70 Fourth 275 8,867 72 1985 First 657 9,224 77 Second 402 9,626 80 Third 447 10,078 84 Fourth 488 10,511 88 1986 Fl rt 446 10,957 91 Second 842 11,299 94 Third 347 11,648 97 Fourth 106 11,752 98 1987 First 138 11,885 99 Second 8 11,88 99 Third 112 12,000 100 ref. disb.dp - 56 - ANNEX 4 REPUBLIC OF MALI FIRST URBAN DEVELOPMENT PROJECT (Cr. 943-MLI) PROJECT COMPLETION REPORT Estimated and Actual Disbursements by Category (in '000 US$) Amount Credit Amount Category Agreement Disbursed Difference 1. Civil Works 1A Bamako District 3,200 3,681 + 481 1B Mopti 20 - - 20 IC Gao/layes 300 334 + 34 2. Equipment, Vehicles and Supplies 2A Bamako District 2,160 .2,564 +404 2B Mopti 200 153 - 47 2C Gao/Kayes 100 - - 100 3. gousing and Artisan Credits 1,000 18 - 982 4. Operating Costs Project Unit 40 215 + 175 5. Technical Assistance 3,000 4,730 + 1,730 6. PPF Refunding 480 305 - 175 7. Unallocated 11500 - 1.500 Total 12,000 12,000 0 ref. categ.dp AI - 57- ANNEX 6 REPUBLIC OF MALI FIRST URBAN DEVELOPMENT PROJECT (Cr. 943-MLI) PROJECT COMPLETION REPORT Basak, District 1978-88 - Evolution of Income and Expenditure (in current CFAF million) 1978 1979 1980 1981 1982 1988 1984 198 1986 A. Income 1. Operating (ordinary) Income Business taxes 483 667 702 Other fiscal revenues 52 47 44 Market taxes 39 46 6 Other economic service fees 58 51 68 Solid waste taxes * 16 24 Housing and land management fees 120 180 124 Other social service fees 9 9 8 Transfers - 5 3 Subtotal 762 1,022 1,036 2. Investment (extre-ordinary) Income Sale District properties - * Loans Subsidies - - Transfers *255 128 121 Subtotal 256 128 121 Total Income 428 399 49 607 871 662 1.019 1.160 1.167 B. Expenditure 1. Operating (Ordinary) Expenditure General services 121 190 197 Street/Drainage Maintenance 272 421 629 Other economic services 80 25 11 Solid waste collection 89 75 as Other social services 120 128 127 Transfers 55 17 o Subtotal 688 852 1,120 2. Investment (extra-ordinary) Expenditure Construe.ions - - 3 Vghicles and equipment - 20 74 Iefrastructure - 99 130 Loan reimbursement 64 - - -Deferred payments - 146 - Subtotal 268 Total Expenditure 482 551 574 451 62 891 638 1.116 1,629 Net amount surplus (deficit) (64) (162) 76 15 209 (229) 43 32 (468) Net cumulatif surplus (deficit) (82) (184) (109) 46 255 26 69 101 (365) ref. exped.dp 口軍口口爾口口口鄴神啊口口唱甲口 . 覲汗矓“藝““涔谿環輟誹i汗誹州谿輾菲‘ l；排粩’：矓縱l：謝讓騙矓訐i】；排州；矓矓嶽i 馴h變物吃鬥鬥究叩斤他雙門雙究劉，究吃髡吃究吃凡究叩吃禿弋禿髡究究雙究：I斤吃門髡彎門門彎勿髡變鳥他然究他叩叩他 邵；&&&&,&“實”懿零““鸚：露“”鵝雜響萬；邵””華”寫‘&“誠神響：: 劊；．瀾．潤目 妝曆劉里l三口髡更究究雙究劇門雙雙究究開開究雙究開‘潤鰓門門倘吃鬥吃‘豐他勿鬥 ．調鄴。酌劍」認萬：’魚豐寫潤＃:：荔馮轟屆轟禺轟轟潤認寫寫日：萬：界萬：: d】。•■之！d日面！11」」 。〔活！三―A編I豐I觔朧計 ’劊鑼到籰‘，二‘,.－二‘, &&！劉“&,：震雖望龍．&,l蘇龍騙議‘}‘蘿‘排蕪 【；觀蘿菲：“龍”l，召’菲暹，“菲“l，易翁“:‘莊離 .：豐杷I,．寫豐蔥寫必啊莖言斗 面’一娥l一a＿州＿－一•，- g徑伯造日·af召召 謹廈勵，蠱黑露屆里旦華號暴要劉擊離齋 墓呈一’g薑專言莖審言 匹“箠呈至莖鎚廷“婁藝審終呈賽二“名響響醒賽韭無“賽藝藝呈要藝莖參醒讓藝話匹“莖賽莖賽莖蠱話“輪獎獎賽參旦至呈認 && 一→一一一一一一一一一～一→→•．■■■－」■一一一一一一－一 一留閱戶，一一→一― /’森． n明■月■国■■■■■■国国■■－～一→ー＿＿、＿ 阜昌曲D 139Bb 一．・ ー→―→ーーーーーーー→ー→→→加→→曲