Cameroon - Urban Development Project

Document of The World Bank FOR OFFICIAL USE ONLY Report No. 8831 PROJECT PERFO1MANCE AUDIT REPORT CAMEROON FIRST URBAN PROJECT (LOAN 2244-CM) JUNE 25, 1990 Operations Evaliation Department This document has a restricted distribution and may be used by recipients only in the performance of their officid duties. Its contents may not otherwise be disclosed without World Bank atherization. ABOREVIATIONS AND ACRe.JYMS ARAN Agence de Restructurati4n et d'Aménagement de Nylon ARYNO Agence pour la Restructuration de Yaoundé Nord-Ouest BEAC Banque des Etats de l'Afrique Centrale BLAF Burau de Liaison Jes Affaires Foncières CAPME Centre d'Assistance aux Pet1te* et Moyennes Entreprises CFC Crédit Foncier du Cameroun CPN Calse Populaire de Nylon DDA Coopération Suisse au Développement et à l'AIde Humanitalre FEICOI Fonds Spécial d'Equipement et d'Intervention Interdormunal MAETUR Mission d'Aménagement et d'Equip.ment des Terrains Urbains et Ruraux MINUH Ministère de l'Urbanisme et de l'Habitat NOO Non-govern»ental Organizatlon OPC Organisation, Pi lotage, Contr8le de Travaux PiE Petites et Moyennes Entreprises RFP Request for Proposai ROW Right of Way SIC Société Immobilière du Caomaroun SNEC Société Nationale des Eaux du Cameroun SONEL Société Nationale d'Electricit6 UCPY Union des Caisses Populaires de Yaoundé FOR OFFICIAL USE ONLY THE WORLD BANK Washington. D.C 20433 U.S.A. OM1Ce as DMIrect-CenWra opatwis eviatiti June 25, 1990 MEMORANDUM TO THE EXECUTIVE DIRECTORS AND THE PRESIDENT SUBJECT: Project Performance Audit Report on Cameroon - First Urban Development Project (Loan 2244-CM) Attached, for information, is a copy of a report entitled "Project Performance Audit Report on Cameroon - First Urban Development Project (Loan 2244-CM)" prepared by the Operations Evaluation Department. Attachment This document has a restricted distribution and may be umed by recipents only in the performance of their official duties. Its contents may not otherwise be disclosed withoat World Bank authoriation. FOR OFFICIAL USE ONLY PROJECT PERFORMANCE AUDIT REPORT CAMEROON FIRST URBAN PROJECT (LOAN 2244-CM) TABLE OF CONTENTS Page No. Preface ......................................................... i Basic Data Sheet ................................................ ii Evaluation Summary .............................................. v PROJECT PERFORMANCE AUDIT I. PROJECT BACKGROUND ........................................ 1 A. The Economic, Demographic and Political Context ....... 1 B. The Urban Sector ...................................... 1 C. Bark Involvement and Strategy ......................... 2 II. THE PROJECT ............................................... 3 A. Origin, Development and Objectives .................... 3 B. Project Components and Costs .......................... 4 C. Costs, Financing and Administration ................... 4 III. PROJECT IMPLEMENTATION AND ACHIEVEMENTS ................... 6 A. Time Delays ................................ ......... 6 B. Cost Overruns ......................................... 7 C. Achievements and Continuity ........................... 11 IV. POINTS OF INTEREST ........................................ 17 A. Cost Recovery ......................................... 17 B. Community Participation and "Animation" ................ 19 C. Credit Foncier - Home Construction Finance ............ 20 D. Support for Small Businesses and Artisans ............. 22 E. Economic Rate of Return ............................... 23 F. The Second Urban Project ............................... 24 V. PROJECT IMPACT ............................................ 24 A. Physical Impact ....................................... 24 B. Environmental Impact .................................. 25 C. Institutional Impact .................................. 25 D. Policy Impact ......................................... 27 E. Poverty Impact ........................................ 28 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 (Cont'd) VI. ROLE OF THE BANK .......................................... 28 VII. LESSONS LEARNED ........................................... 28 VIII. CONCLUSIONS ............................................... 30 ANNEXES 1. Diagrams .................................................... 31 2. Swiss Calculations ................................. ........ 33 3. Photographs ................................................. 34 PROJECT COMPLETION REPORT I. BACKGROUND AND PROJECT DESCRIPTION ............ ........... 43 II. PROJECT IMPLEMENTATION AND COST ........................... 46 III. PROJECT RESULTS AND SUSTAINABILITY ........................ 52 IV. BANK, BORROWER AND CONSULTANT PERFORMANCE ................. 56 V. LESSONS LEARNED .................................. ........ 60 i PROJECT PERFORMANCE AUDIT REPORT CAMEROON FIRST URBAN PROJECT (LOAN 2244-CM) PREFACE 1. This is a Project Performance Audit Report (PPAR) on the First Urban Development Project, partially financed by Loan 2244-CM in the amount of US$ 20 million to the Republic of Cameroon. The loan was approved on March 15, 1983 and became effective on May 3, 1984. The project closed on June 30, 1988 as expected at appraisal and an unutilized loan balance of less than US$ 200,000 was cancelled on January 25, 1989. 2. The PPAR consists of the Project Performance Audit prepared by the Operations Evaluation Department (OED) and a Project Completion Report prepared by the Infrastructure Operations Division of the Occidental and Central Africa Department (AFlIN). The PPAR is based on the attached PCR, the Staff Appraisal Report, project legal documents, project correspondence files, a transcript of the Executive Directors' discussions at Board presentation, interviews with Bank staff and other releva.t materials. An OED mission visited Cameroon in February 1990 to discuss the project with local and national government officials. 1.sair kind cooperation and valuable assistance in the preparation of this report is gratefully acknowledged. 3. The PCR provides a good account and assessment of the project experience and indicates its successes and failures. In order to give a wider perspective, OED has attempted to provide additional information about the project, its history, its outcome, the issues it raised and the problems encountered during implementation. 4. Following standard OED procedures, copies of the draft PPAR were sent to the Government. No comments, however, were received. PROJECT PERFORMANCE AUDIT REPORT CAMEROON FIRST URBAN PROJECT (LOAN 2244-CM) BASIC DATA SHEET KEY PROJECT DATA Appraisal Actual or Item Expectation Current Estimate Total Project Cost (USS million) 54.7 74.0 overrun (W) * 355 IBRD Loan Amount (US$ million) 20.0 20.0 Disbursed -- 19.8 Cancelled -- 0.02 Repaid -- 0.47 outstanding - 19.33 Date Physical Cojvnents Completed 06/30/88 06/30/90 Proportion Completed by Above Date (W) 70 100 Proportion of Tim Overrun (W) -- 40 Economic Rate of Return (W) 18 24 Flancial Performance ** Acceptable Institutional Performance ** Acceptable CUMULATIVE ESTIMATED AND ACTUAL DISBURSEMENTS (US$ million) itet FY83 FY84 FY85 FY86 FY87 FY88 FY89 Appraisal Estimate 0.3 3.7 11.6 17.0 19.4 20.0 ** Actual 0 0.4 1.7 3.0 8.6 17.8 19.8 Actual as % of Estimate 0 9 15 18 43 89 99 OTHER PROJECT DATA Item Original Amended Actual Government's Application 08/24/76 First Mention to Timetable 08/01/77 Negotiations 10/1979 06/1981 03/04/82 Board Approval 11/1979 10/1981 03/15/83 Loan Agreement Date 01/1980 06/15/83 09/15/83 Effectiveness Date 12/15/83 05/03/84 Closing Date 06/30/88 06/30/88 Dorrower Republic of Cameroon Executing Agencies Ministry of Urban Developaint and Housing, MAETUR and CFC Fiscal Year of Borrower July 01 -- June 30 Follow-on Project Name Second Urban Project Loan number 2999-CM Amount (US$ million) 146 Loan Agreement Date January 25, 1989 La Including US$13.1 N additional financing provided under Second Urban Project (La. 2999-CM) / US$166,945.90 cancelled on 01/25/89 MISSION DATA Mission No.of No. of Date Item Composition Mo./Year Weeks Persons StaffWeeks of Report Identification 1 UP/EC 10/76 2 2 4 02/02/77 Identification 2 UP 06/77 2 1 2 08/01/77 Identification 3 UP 03/78 1 1 1 05/01/78 Preparation 1 UP 07/78 1 1 1 09/14/78 Preparation 2 UP/2EC q**79 1 3 3 04/04/79 Preparation 3 2EC */79 3 2 6 06/07/79 Preparation 4 UP/EC : 79 2 2 4 09/17/79 Preparation 5 UP/EN 10/79 2 2 4 11/19/79 Preparation 6 UP 12/79 1 1 1 01/15/80 Preparation 7 UP/ME/FA/EC/TE/INST 02/80 2 6 12 03/03/80 Preparation 8 UP 05/80 1 i 1 06/30/80 Pre-appraisal UP/FA/LC/INST/C / 07/80 3 8 24 08/20/80 Appraisal UP/FA/EC/CE/TE 12/80 3 5 15 12/19/80 Post-appraisal UP/EC 02/81 1 2 2 03/04/81 Post-appraisal UP/FA/TR 0581 2 3 3 06/02/81 Post-appraisal UP 09/81 2 1 2 10/05/81 Post-negotiation/ Supervision 1 / UP/FA/ME/EC/INST 04/82 2 5 10 05/03/82 Post-negotiation/ UP/ME/2EC/FA/EN 12/82 2 6 12 02/04/83 Supervision 2 /b Supervision 3 UP/EC/ME 06/83 2 3 6 06/29/83 Supervision 4 UP/ME 10/83 2 2 4 01/13/84 Supervision 5 / P//FA/C/CA/INSTE/ 02/84 2 8 16 03/23/84 Supervision 6 UP/ME/EC 07/84 1 3 3 08/15/84 Supervision 7 UP/EC 10/84 1 2 2 12/05/84 Supervision 8 A ME/EM/TE/DE 01/85 2 4 8 02/08/85 Supervision 9 /c ME/ENG/EC/TE/DE 04/85 3 5 15 05/07/85 Supervision 10 A ME/ENC/T/DE 07/85 3 4 12 09/06/85 Supervision 11 c ME/ENG/DE 10/85 2 3 6 11/05/85 Supervision 12 A TR 12/85 1 1 1 01/13/86 Supervision 13 A ME/UP/ENG 06/86 2 3 6 07/31/86 Supervision 14 A UP/ENG/DE 02/87 1 3 3 04/01/87 Supervision 15 A UP/INST/ENG/DR/2RE 05/87 1 6 6 07/17/87 Supervision 16 A UP/EC/RE 11/87 1 3 3 12/17/87 Supervision 17 jd UP/RE/INST 06/88 1 3 3 09/06/88 Completion / UP/EC 10/88 3 1 3 12/27/88 / UP - Urban Planner ME * Municipal Engineer SS * Soils Specialist EC * Economist TE - Traffic and Transport Engineer ID - Loan Officer FA - Financial Analyst INST - Institutional Expert AR = Architect EN - Energy CE - Civil Engineer TR * Training CA * Cadastre DE * Drainage Engineer RE * Road Engineer A Sector mission. / Preparation mission, Second Urban Project Ld Pre-negotiation mission, Second Urban Project L Project launch, Second Urban Project - iv - CURENCY EXCHANGE RATES Name of Currency (Abbreviation): (FCFA) Year Appraisal Exchange Rate US$1 - 270 CPA Intervening Years Average: Disbursement 2% 1984 Excbange R&te US$1 = 437 CFA Disbursement 7% 1985 Exchange Rate US$1 - 449 CFA Disbursement 8% 1986 Exchange Rate US$1 - 346 CFA Disbursement 27% 1987 Exchange Rate US$1 - 300 CFA Disbursement 56% 1988 Exchange Rate (E) US$1 = 300 CFA Weighted Average Rate (E) US$1 = 317 CFA Completion Year Average Exchange Rate (E) US$1 = 300 CFA (E) Estimated. STAFF INPUT (Staff Weeks) Fiscal Year: FY80 FY81 FY82 FY83 FY84 FY85 FY86 FY87 FY88 FY89 FY90 Total Preappraisal 104.9 -- -- 104.9 Appraisal -- 50.9 24.0 -- ** -- ** -- -- ** -- 74.9 Negotiations -- -- 22.6 3.4 -- -- -- -- -- -- 26.0 Supervision -- -- -- 4.0 12.3 22.2 10.0 14.5 7.9 6.5 ** 77.4 Other -- -- -- - -- -* - -- -- 5.2 1.1 6.3 Total 104.9 50.9 46.6 7.4 12.3 22.2 10.0 14.5 7.9 11.7 1.1 289.5 V PROJECT PERFORMANCE AUDIT REPORT CAMEROON FIRST URBAN PROJECT (LOAN 2244-CM) EVALUATION SUMMARY Introducti2n 1. The first urban project in Cameroin came at the end of a period of sustained economic growth, but its implementation coincided with the crisis which followed the collapse of oil prices in the mid 1980's. Project preparation was drawn out from 1977 to 1983 as the Bank worked along with other bi-lateral agencies and the Government of Cameroon to develop policy instruments which would complement the Government's focus on high and middle-income housing with programs for low-income families. The Zirst project aimed to introduce a replicable model for urban development. I is marked, however, by the change in the country's economic outlook which produt-wd a shift from the social interest policies to a focus on productive investments. As a result, the project's scale was reduced during the final stages of preparation. Works proposed for the coun:ry's capital, Yaound6, were postponed and the serviced lot components were eliminated, leaving as the project's centerpiece the upgrading of a 600 hectare squatter settlement in Cameroon's largest city, DouAla. Objectives 2. The project represented a bold attempt to introduce several new concepts including upgrading of squatter neighborhoods, land regularization, resettlement on unserviced plots, home construction financing, cst recovery and support for small businesses. Hefty and troubling cost overruns, resulting from the problematic terrain of the project area, Nylon, and inadequate engineering studies led to dramatic cutbacks in the proposed infrastructure. Once the civil works 1 :ogram had been redefined, through a renegotiation in 1987, the operation moved ahead rapidly, completing disbursement in record time (five years as opposed to the Bank average of 9.5 years for urban projects), although several components were left to be completed under a second project. Cost overruns continue to occur and a true economic evaluation of the project will not be possible until final works are completed and paid for. At that point, the cost/benefit ratio may look less sanguine than at present. 3. The first project included studies leading to the development of a second urban project, which appeared to absorb most of the time and interest of the Bank and the Government during supervision missions given the shift in the focus of urban policy. The second project is geared toward the development of primary urban infrastructure and municipal revenues on a national scale. It includes, nonetheless, certain priority works for Nylon in Douala. vi Implementation Experience 4. The project has had a marked positive impact on the urban environment of Nylon and on the city as a whole and has, in addition, stimulated economic growth in Nylon and strengthened community development. Cost recovery appears to be going well. Delays in resettlement of expropriated families and in issuing land titles, together with the failure to generate a viable home financing system or a program to support small businesses, represent the main shortcomings in an otherwise successful operation. 5. The project established a special executing agency in Nylon, ARAN (Agence de Restructuration et d'Am6nagement de Nylon). ARAN's position as technically independent, but actually under the aegis of two other agencies, led to considerable complications during implementation. In addition to its own identity problems (PCR para. 1.12), ARAN suffered from the imposed collaboration with other agencies for small business promotion and assistance and, more ihportantly, for home loans, plot acquisition financing, and, thus, for cost recovery of on-site civil works. Despite its isolation, ARAN managed the project effectively, although it has had periodic difficulties rerulting from its distance and consequent lack of support from national government. 6. Cooperation with the Swiss technical assistance agency, DDA, led to a fortunate, if occasionally disjointed: collaboration. DDA financed community facilities in Nylon and filled in where technical assistance was lacking. While the Bank focused on hardware components, the Swiss were able to work with the executing agency, ARAN, in areas of cost recovery, resettlement, land titling and home construction finance. Results 7. The existence of a strong community organization in Nylon, "Animation," made the area an attractive locus for a first project. Indeed, the collaboration between ARAN and Animation has been key in the success of components such as cost recovery and resettlement, and has led to additional programs focusing on small-business development, infrastructure construction through mutual assistance and the local credit union as a source of financing and savings generation. The experience of ARAN and Animation lends support to arguments for community part' -pation in urban development projects. 8. Among other aborted attempts to involve national agencies, the plan to establish CFC, the national home credit institution, as a financial intermediary for home construction loans and for cost recovery failed. CFC proved unprepared and/or unwilling to work with the informal sector and to collaborate with ARAN in project cost recovery as the Appraisal had specified. In retrospect, it was a mistake to assign cost recovery to a separate agency. Under the circumstances, ARAN has managed to develop a viable cost recovery system on its own, and is supporting the local credit union, the Caisse Populaire de Nylon (CPN), as an alternative home financing agent. vii Sustainability 9. Although a primary objective of the first project was to establish a replicable model for urban upgrading, to date no such replication has taken place. Nevertheless, the interest generated by the events in Nylon and the adoption of certain policy measures which support the legitimacy of squatter settlements suggest that urban upgrading has gained a small but sure foothold in Ganeroon's urban sector. Findings and Lessons 10. The project has been important for institutional development within Cameroon's urban sectGr insofar as it brought various agencies together and set a new precedent for coordinating efforts in the sector. In addition, it set in place a new lir. of communication from the community up to the Ministry. 11. Studies financed under the first project led to the preparation and negotiation of a second urban project in 1988. In addition to the Ussons learned about tite design of an effective management structure, the first project identified areas of institutional and policy weakness, and led to a reassessment of institutions in the sector without which the second project could not have moved forward. For example, both the rejection of the complicated institutional set-up of the first project, including a local executing agency, and the fact that the municipalities were given an important role in the urban development process in the follow-on project resulted from lessons learned in the first operation. 12. Complex barely begins to describe this project which vas apprairad twice, renegotiated in mid-stream and continues to be implemented even after disbursements have been completed. The PCR, completed ir, 1989, gives a very thorough account of events and clarifies the outcome of the principal components. The Audit has thus been able to focus on additional points of interest and to complement the PCR with information which has subsequently become available. PROJECT PERFORMANCE AUDIT REPORT CAMEROON FIRST URBAN PROJECT (LOAN 2244-CM) I. PROJECT BACKGROUND A. The Economic. Demographic and Political Context 1.01 During the fifteen years prior to appraisal in 1981, Cameroon was characterized by steady economic growth in an atmosphere of social stability and gradual national unification. Real annual GDP growth increased from an average of 4.2% between 1966 and 1971 to 4.5% during the following five years, and then to between 7% and 8% from 1976 to 1981. From 1966 to 1980 per capita GDP rose by an annual rate of 3.1% Donations from the French Government averaged around US$ 45 million a year during the 1960's. Thereafter, French foreign aid continued, increasing to about US$ 90 million per year during the 1970's, but grants diminished and most assistance came in the form of loans. 1.02 From 1978 to 1982 petroleum gained importance as a major export, moving from a 1% share of total exports to about 63% and replacing cocoa and coffee as the nation's leading export product and a major component of GDP. Government's concerns, echoed in dialogues with the Bank, centered on finding ways to convert the windfall of oil revenues into productive investment. Unfortunately, the economy's sudden boom was followed by a general crisis when oil prices fell in the mid 1980's. Despite its efforts to consolidate agriculture and industry, the Government could not contain the repercussions of the crisis which marked the project during its implementation and afterwards. 1.03 With a population of 8.9 million in 1983 and promising economic potential in many areas, tte Government of Cameroon had focused on agriculture (livestock and forestry), agro-industry, trade and commerce, port facilities and a transport network also serving neighboring landlocked Chad and the Central African Republic. B. The Urban Sector 1.04. At the time of appraisal, Cameroon ranked as the fourth most urbanized country in West Africa. Rapid urban growth suggested that half of the population would be living in cities by 1990, a prediction which has been borne out. At the same time, problems burgeoned along with the urban population, the majority of which were squatters on poorly drained and poorly serviced periurban land. Forty five percent of the residents had no access to piped water. Fewer than 10% held legal title to their land. 2 1.05 Nowhere was the incrasing urban squalor more evident than in Douala and Yaound6, Cameroon's two largest cities. Although the secondary urban centers were growing almost twice as fast, Douala and Yaound6, already housed more persons than all the secondary cities combined, a full 38% of the national urban population and 12% of the country's population as a whole. As Douala and Yaound6 continued to expand, the existing squatter se 'ements of Nylon (PCR para. 1.06) and the Briqueterie lacked infrastructure , -t suffered chronic envi-onmental problems: residual flooding, no drainage and insufficient garbage collection, resulting in the accumulation of solid waste and serious health hazards. 1.06 Jes,-ite these conditions, Douala and Yaoundd enjoyed certain advantage. t- -. >sitive prospects. Urban poverty was less pronounced than in other cou.....- 9c-. Both cities enjoyed high levels of formal employment: three- fourths of the work force in iaoundd and two-thirds in Douala were reported to be salaried employees. Although 17% of Douala's households and 25% of those in Yaound6 were estimated to earn below the respective poverty levels (higher than in other African countries), these percentages are markedly lower than those found in Cameroon's neighbors, or in the country's secondary cities. A high rate of internal mobility characterized both Douala and Yaound6 -- from 1972 to 1982 over 70% of their households moved at least once -- indicating that the cities' formation was still, to a certain degree, in flux. 1.07 In institutional terms, the urban sector in Cameroon was largely dominated by centcalized agencies, following the French model. Urban and regional policy, planning and housing functions, as well as land registry and land management, rested with the Ministry of Urban Development and Housing (MINUH), more specifically with its Division of Urbanism and Housing (DUH), formed in 1976. Three parastatal agencies were developed to assist MINUH: (i) SIC (Socift6 Immobili&re du Cameroun), the national housing agency, traditionally dedicated to housing for middle-income families and civil servants; (ii) MAETUR (Mission d'Amdnagement et d'Equipement des Terrains Urbains et Ruraux), created in 1977 and charged with developing residential land, but similarly geared to upper-income families, and; (iii) CFC (Cr6dit Foncier du Cameroun), established in the 1960's, but reborn in 1976 when a funding mechanism through a payroll tax was developed. The latter was responsible for financing MAETUR and SIC operations and generating affordable home financing. The municipal councils, presided over by a delegate from the national government, in turn, had limited budgets and functions. Infrastructure was and is the business of national parastatal companies, while state land was held and managed by the national government. C. Bank Involvement and Strategy 1.08 By 1982, the Bank had financed 41 projects in Cameroon. Transport and agriculture each accounted for about 40 percent of Bank commitments, although the period from 1975 to 1982 witnessed a growing tendency for the lending program to be broadened into areas such as industrial development, technical assistance, education and water supply. The recommendation to the President in 1983 notes that Cameroon's disbursement rate had been highly efficient in the past, but that recent problems in the central procurement agency had caused setbacks in the country's performance vis-&-vis Bank operations. 3 1.09 Bank efforts in Cameroon supported three main areas of development: (i) international trade and transportation -- strengthening and extending the trunk road and rail systems and improving the port of Douala; (ii) raising agricultural output and exports; and (iii) improving education. The Bank strategy in the early 1980's was to continue along the same basic lines, but to strengthen other infrastructure, in particular telecommunications, energy, health, water supply/sewage and urban. The sectoral diversification of Bank lending was seen as a reflection of the growing complexity of the country's development. 1.10 Although an urban sector mission visited Cameroon in 1975, project identification came some time later. In the interim, the Bank offered technical assistance and opened an important dialogue involving the sector, focusing on development of a national housing and urban growth policy. II. THE PROJECT A. Origin, Development and Objectives 2.01 Given their predominance in the sector (paras. 1.05 and 1.06) Yaound6 and Douala emerged as the primary targets for an urban project in the context of planned growth and squatter upgrading. A first project memorandum appeared in February 1977, mentioning the two cities and Nylon, Douala's most prominent squatter settlement -- well-known because of its community organization -- as areas of action and proposing an appraisal date for 1979. 2.02 By 1979, the Cameroon Government had only recently established the MINUH-DUH and a first feasibility study was contracted in July 1979. The Bank resident mission provided an advisor to the Director of Urbanism and Habitat in MINUH from 1979-1981 to help develop a national housing policy. Other bi- lateral agencies were also starting up activities in the shelter subsector in Cameroon, emphasizing low-cost solutions for low-income families. The Government, accordingly, redirected its planning efforts towards urban upgrading and serviced lots. 2.03 Given the focus of policy development, the project began to take form with a double objective: that of raising infrastructure standards for low-income families (PCR 1.08) and that of setting in place a sustainable and replicable program for dealing with sub-standard settlements in both existing and new developments. The project underscored two other key policy issues: granting tenure to squatters and cost recovery. The Government enunciated its goals as follows: "to develop the means of improvement which will permit a progressive evolution of (such) zones leading, ultimately, to publicly-acceptable living conditions. Moreover, the proposals seek to change the marginality and inaccessibility which characterize the zone(s) and touch every aspect of the population's life." 1 1 P.D.U. evaluation, p.6. 4 2.04 The proposed project included serviced and unserviced lots and squatter upgrading for both Douala and Yaound6. From 1979 to 1981 other components found their way into the project in accordance with the prevalent fashion for multi-faceted operations: home improvement loans, community facilities, market construction, small business support, urban transport, energy and municipal services. 2.05 With appraisal set for late 1980, the Bank concentrated on reducing projected costs, which ranged between US$ 100 and 120 million, to an acceptable US$ 74 million, and postponed negotiations until completion of the consultants' feasibility studies. When the Mayor of Yaoundd did not appear at pre-appraisal mission meetings in July, 1980, the upgrading and municipal support components for that city were dropped. With a total project cost of US$ 94 million, the Bank invited the Government of Cameroon to negotiate. 2.06 The concurrent reclassification of Cameroon from an IDA to an IBRD country turned the tables by giving the Government of Cameroon motives for further reducing project costs. As the rCR recounts (para. 1.08), negotiations resulted in elimination of serviced lot components for both cities, leaving only a skeletal program of community services for Yaound6 and making the upgrading of Nylon in Douala the project's centerpiece. B. Project Components and Costs 2.07 As the Table below shows, 74% of project investment covered infrastructure works in Douala. While certain measures benefitted the entire city, the bulk was destined to upgrading, primary roads, drainage, water supply and electricity for Nylon, which, at the time, covered some 600 hectares and had a population estimated at 90,000. Another 3.7% of the total investment was allocated to finance secondary and tertiary infrastructure -- including house connections as warranted -- additional standpipes and home improvement loans to a pilot upgrading zone of 50 hectares with a population of 20,000, within Nylon. In addition, the Swiss technical assistance agency, DDA, was to finance community facilities in Nylon, specifically a major market and secondary markets, community centers and health posts. Bank financing also covered the development of serviced lots for resettlement of families uprooted by upgrading activities, support for small businesses and technical assistance and studies focused mainly on the development of a second project. C. Costs. Financing and Administration 2.08 Total project costs came to US$ 54.7 million, of which the Bank loan covered US$ 20 million, or 36%. The DDA loan accounted for US$ 5.5 million, or 10%, excluding grants for Swiss technical assistance. The Government of Cameroon was to finance US$ 29 million, or 54% of the total investment. 2.09 Chart 2-1, taken from the SAR (para. 3.05) shows the flow of funds and the various agencies designated to implement the project. A special land development agency for Nylon, ARAN (Agence de Restructuration et d'Amdnagement de Nylon) was established as an extension of MAETUR. ARAN\MAETUR, technically independent, but actually under the aegis of MINUH-DUH, was responsible for the 5 C良織TZ。1 CAMEROON URBA可nRV次了np闕雙可？,oh?”戶， ，。。．。二，。，’。．。八“uov。‘Urn匕卹T FROJECT FLO可OF FUNDS 匹三二一一J·．- 屆，l一It細••細•織．••、． l一＿11戶．&．•悶•••••j州饑化L--se一一一＇T細細自．＇•開“州” l&-L,11一•1自細自• &L-eesees-wese一一一」、、瀾C，己L•••••••••■••．--＇劇馴森開口口細•鳥口口歸 I、、L．一＿八】州口•豳•細．購•騙•畸 、、、I智馴闢口向觔•伯開“取 、、、日一闢卹細甩細細馴•O開馴” 、、、屆。必口開鬨向物•個瞌•• 、，I口1 00口婦C細細向口細•”口• 、，―日•細”騙濘馴細’ 么•■■■■••■．•工口•細細細鬨 1__＿口I屆•自．••網•，．闕細畸 必二二」【’&&& 物口細•口自一認讓論 6 bulk of the project. In addition to its own identity problems (PCR para. 1. 12), ARAN suffered as a result of the imposed collaboration of other agencies, notably CAPME (Centre d'Assistance aux Petites et Moyermes Entreprises), for small business promotion and assistance and, more importantly, CFC, for home loans, plot acquisition financing, and, thus, cost recovery of on-site civil works. (SAR para 3.05) 2.10 Community facilities, with the exception of the major (Madagascar) market, technical assistance and primary infrastructure, were to be financed by the Government of Cameroon as an outright grant. The secondary infrastructure in the pilot upgrading zone and the home improvement loans, estimated at a total of US$ 2.02 million, were to be recovered by CFC from direct project beneficiaries. The construction costs of the Madagascar market were to be recovered from vendors through rents for market stalls. III. PROJECT IMPLEMENTATION AND ACHIEVEMENTS A. Time DelM 3.01 Although the project had advanced considerably during preparation, with the dredging of the Mgoua River (PCR para. 2.02) and identification of the first resettlement area, Dibom II, once negotiations were completed, a certain sluggishness set in, amply described in the PCR (paras. 2.02 - 2.06). Compliance with loan conditions proved elusive, in particular the definition of CFC's role in managing the home improvement loans program in terms acceptable to the Bank. CFC's subsequent performance turned out to be disappointing (paras. 4.13-4.18) and the general eagerness to have the relevant covenants agreed may have led to acceptance of an ultimately unworkable format. 3.02 The repercussions of the delays in loan effectiveness were passed on to project implementation. The engineering studies for the general upgrading of Nylon actually began in 1984, more than a year after negotiations. The studies for the pilot zone were contracted out a full two years after their scheduled completion. Because the SDAU (Schema Directeur d'Amdnagement et d'Urbanisme) presented a radically new urban road plan (PCR para. 2.07) in late 1984, new terms of reference were developed. Furthermore, ARAN\KAETUR and MINUH were unfamiliar with Bank procurement requirements QCR paras. 2.04, 4.02 and 4.12), but the lengthy internal procedures imposed by the complicated relations between implementing agencies also played a role (PCR paras. 4.08-4.10) in the observed delays. After road construction had begLu, in the fall of 1987, additional stumbling blocks became apparent (PCR paras. 2.13-2.15), while cost overruns (paras. 3.06-3.14) eventually led to an amendment of the project in early 1988. 3.03 Although the key project data chart shows a time overrun of 40%, the project remained stubbornly on course; in a surprising about face, ARAN managed to invest 56% of the total project in a single year. By the scheduled date of project completion almost 90% of the amount projected at appraisal had been disbursed, belying the pessimistic supervision mission reports which predicted that the project would have to be extended until 1992 and making a very positive 7 showing over the worldwide average of nine and a half years for Bank urban projects. 3.04 In retrospect, the project officer at the time (of the great leap forward) attributes the slowdowns to blockages in funding from the Central Government as a result of the fall in oil prices in the mid 1980's. He also notes that the Douala project was small in comparison to MINUH's other operations, suggesting that it was the prospect of a larger second project, on a nationwide scale (paras. 4.28-4.29) and which would give more prominence to MINUH, that actually spurred on implementation. 3.05 Amendment of the project (para. 3.02) meant that it was considered closed when funds ran out even though the financing for certain activities foreseen under the first operation was relegated to the second project. Thus, in a sense, the first project remains incomplete (paras. 3.15-3.18) and the supervision reports' predictions of the need for an extension until 1992 may, in practice, prove correct. B. Cost Overruns 3.06 From any angle, the project proved far moze costly than originally foreseen. The PCR calculates a 35% cost overrun based on the addition of US$ 13.1 million financed through the second urban project and an increase of US$ 2.7 million of DDA funds. The DDA evaluation shows an even higher figure -- closer to 50% -- taking into accr.tnt the final costs of community facilities (the Madagascar Market came in at 200% over the original estimate) and the extension 2 of technical assistance to ARAN (some 160% of the appraisal estimate). The Audit calculated a 56% overrun in the costs for Nylon upgrading to date (Table 3-1), although this figure lacks some important inputs, notably land costs for Ndogpassi and ARAN's operational costs throughout. 3.07 As the Table below indicates, the main cost overrun stems from infrastructure, associated studies and extended technical assistance. The original estimates for road and drainage costs were already considered high at the time of appraisal "because of the unique situation in Nylon... It is low- lying, it drains a large part of the surrounding area and is often subject to flooding." (SAR para. 2.04). With base costs of US$ 220,296 per kilometer (for a 12 meter wide road) for major roads and US$ 359,411/km for drains, Bank engineers had li-tle reason to assume that cost estimates were low, particularly when the first engineering study came in 1984 at half the Appraisal estimate. 3.08 According to final construction costs, as amended in late 1989, base road costs, however, have risen to US$ 1,961,009/km, for a 12 meter wide road. How could the Bank and project management be caught so unaware ? According to the project officer at the time, the road designs as developed in the feasibility and subsequent engineering studies were simply inadequate for the hydraulic conditions at Nylon and for the performance required of a major road (PCR para. 2.13). The PCR suggests that the after-effects of the dredging of the Mgoua River and the subsequent rise in the water table changed hydraulic conditions 2 Dossiers Sectoriels #5 para 1.1.1 8 at Nylon, (PCR para 2.13), an explanation supported by the supervising engineers at ARAN. Project !Costs at Apraisal and Comletion (US Dollars '000) ARRraisal Cpletion A. Development of NXlon 3 Land Acquisition 0.49 0.8% 0.95 1.1% Mgoua Drainage 6.95 12.7% 6.95 8.1% Water 1.95 3.5% 0.85 1.0% Roads 9.96 18.0% 30.0 35.0% Drainage 9.25 16.8% 1.3 1.5% Solid Waste 0.3 0.5% 0.0 0.0% Resettlements 1.27 2.3% 15.55 18.2% Electricity 0.77 1.4% 0.58 0.7% Market 4.55 8.3% 6.8 8.0% Small Business Project 0.96 1.8% 0.5 0.6% Community Facilities 2.91 5.3% 6.7 7.8% Housing Loans 1.15 2.1% 0.83 1.01 Engineering Studies inc. 18.31 21.4% TOTAL br-ON 40.55 74.1% 70.56 82.5% B. Yaondn Community Facilities 1.82 3.3% Market 1.58 2.9% Small Business Project 0.86 1.6% TOTAL YAOUNDE 4.26 7.8% 3 The US$ 49,000 for land acquisition refers to the cost of Dibom II. The cost of additional lots, Ndogpassi 1 and 2 are not available. Completion water, roads, drainage and electricity costs are from the contractor's records. Resettlement costs on completion include US$ 10.34 million for expropriation and US$ 3 million for land development for Ndogpassi 2 under the second project (SAR Annex 2-4), US$ 2.2 million in development costs for Ndogpassi 1 and studies for an alternative site (1'Aeroport). Completion market, small business project and community facilities costs are drawn from DDA's final report. Housing loans include the FFCA 20 million disbursed by CFC outside the project, but does not include CPN loans or loans made by private banks. Engineering study costs are from PDU 2. 9 C. Studies and Technical Assistance ' Project Execution 5.99 10.9% 7.6 8.8% Studies 4.28 7.8% 7.4 8.6% TOTAL TECH. ASSISTANCE 10.26 18.8% 15.0 17.5% TOTAL 54.7 100.0% 85.56 100.0% 3.09 Furthermore, under the French construction system, it is common for the contractor to produce the final design of works. In this case, the contractor who won the contract on the basis of the engineering design (albeit at a cost about 40% over the SAR estimate) proposed important changes in specifications which resulted in a revised cost of approximately US$ 600,000 per kilometer (again based on a 12 meter road). Two independent consultants reviewed and confirmed the validity of the contractor's proposed specifications and costs. The increased cost, however, implied a dramatic reduction in works. A proposal to cut back on the roads (para. 3.15) and eliminate water, electricity and a major portion of the d-ainage (paras. 3.15-3.18) was accepted in principle during project revision in 1987 (PCR para. 2.14), although ARAN insisted on substituting primary water and electricity infrastructure (para. 3.17) for additional road works, including a round-about linking the East-West Boulevard to the Boulevard d'Aviati6n (See Diagrams in Annex 1). 3.10 The final road costs are taken from the revised construction contract which indicates an addition of some US$ 7.8 million, ' bringing road costs close to ten times the figure used at appraisal. The amendment to the contract came after construction costs in Cameroon increased due to the Government's stated policy of paying only 25% of a contract on completion, with the remaining 75% payable over nine years with interest. This may account for the notable increase, although its impact should have been limited since the contractor is paid directly by the Bank on completion of works. The supervising engineers accept these cost increases, explaining that they resulted from the need to increase the foundations or base materials underlying the roads. 3.11 It is notable that the cost of water piping, electricity and drains should have augmented in approximately the same proportion as that of the roads. The supervising engineers considered that the installation of services required the previous construction of the road; the completed roads must be opened and later rebuilt to lay water piping, for example. Had the full costs of civil 4 Costs from MINUH-Cellule PDU2 including technical assistance in sanitation (48 months), in cadastres (42 months), in civil engineering (36 months), in institutional development (4 months), in economics (30 months), special assistance to ARAN (36 months), supervising engineers, audit and equipment for ARAN and for MINUH. s The amended contract became available only one month prior to the Audit and, according to the supervizing engineers, has yet to be reviewed by the Bank. 10 works been known earlier, alternative solutions, such as leaving the secondary and tertiary roads unpaved and laying infrastructure under sidewalks, might have been considered. The situation is complex and belies a certain lack of control which resulted from a combination of circumstances: the project officer most closely involved was taken sick shortly after the project was amended; current construction technically falls outside of both the first and the second projects; and, the supervising engineers, who are not particularly sympathetic to project objectives (para. 5.04), are presented as the on-site representatives of the Bank, ' thus precluding the need for further consultation between ARAN and the Bank on construction issues. 3.12 The Bank has financed several upgrading projects in low-lying urban areas. Given the dramatic cost increases in the present project, a comparison of costs and solutions applied could be helpful for planning and supervising such operations in the future. Specifically, the first urban project in Ecuador financed housing in the Guasmo area in Guayaquil, and the second urban project in Colombia financed upgrading in the Southeast Zone of Cartagena. According to the PCR for the latter, actual costs ran up to 240% of SAR estimates. Road costs, however, were not exaggerated, first because an overall infill solution was applied, rather than the drainage system proposed for Nylon, and second because secondary roads were left unpaved with sidewalks and curb drains built in concrete to protect infrastructure. 3.13 Designs for community facilities developed by DDA architects, in turn, also ran over original cost estimates to a point where MINUH felt they were unjustifiable. The DDA and ARAN countered with the argument that the state had yet to do anything for Nylon. It was further argued that since the Government was undertaking somethinf important, it was necessary that the people see that it was of good quality. 3.14 Cost overruns led to important changes in project financing, affecting, in particular, the residents of Nylon who have been carrying the project's administrative costs for the past four years. The increases in costs hold even greater implications for the families of Nylon in the future. As costs rose, so did the amounts and percentages of cost recovery (para. 4.02) which was designed to finance future upgrading. The resulting development levy jumped from US$ 233 per family to US$ 650 per family, on average. If costs of the homes which the relocated families were required to build -- estimated at between US$ 3,520 and US$ 30,000 (PCR para. 2.19) -- are included as a logical part of total costs, then the beneficiary's direct contribution to total project costs becomes considerable. 6 The supervising engineers stated to the Audit mission that they work directly for the Bank. At the same time, they complain that their payments are several months overdue. 7 Dossiers Sectoriels para 3.2.4 11 C. Achievements and Continuity 1. Civil Works 3.15 Roads and Drains. The SAR called for some 17 kilometers of secondary roads (12 meters wide), three kilometers of primary roads (30 meters wide - the "North-South Boulevard") and a total of 16 kilometers of tertiary roads in the pilot zone. Diagrams 1 and 2 in annex show the original road plan projected in the SAR for Nylon and for the pilot upgrading zone. The roads which have been completed as of the Audit are blackened in and roads to be constructed under the second project are hatched. The diagrams make the differences between the original plan and what has been implemented to date quite clear. Although the roads are solidly constructed and, as the PCR points out (paras. 2.09 and 2.10), were crucial to Nylon's development as a working part of Douala, they are concentrated in one sector and presently lack the vital connections to the downtown and the circur-erential highway (1'Axe Lourd). 3.16 Approximately US$ 15 million has been set aside in the s3cond project for complementary construction. Once completed, however, the road network could be in jeopardy because of lack of maintenance. The contractor's guarantee expires in July 1990 and the city of Douala has not: yet accepted responsibility for maintenance. The rest of Nylon's road network will be financed through collections from residents. Given the evident high cost of primary roads, the future of Boulevard Este remains in doubt and, in any event, will be difficult to finance without outside help. 3.17 Tertiary road construction in the pilot zone is about half completed, but should be taken care of under the second project. According to its long- term plan, ARAN hopes to complete Nylon upgrading, including construction of secondary and tertiary roads, by 1996. The program has begun well, but is facing serious financial problems (para. 4.07). 3.18 The primary drainage (14 kilometers in earth and 3 kiliveters in reinforced concrete) has been contracted out to the same company in charge of roads. Engineering studies have been completed and works should be financed under the second project. The number and length of drains has been greatly reduced, but secondary drains were installed along major roads and community efforts in collaboration with ARAN promise to assist 4n completion of drains along secondary and tertiary roads. 3.19 Water and Electricity. The SAR calls for 9.4 kilometers of primary electrical lines to permit public street lighting and, eventually, private domestic hook-ups. It also included 10 kilometers of primary water mains and the installation of 40 public standpipes throughout Nylon. Finally, some 39 garbage dumpsters were to be purchased for the area. By the time of the project's official completion, however, nothing had been done in these areas and the contractor, together with the supervising engineers, recommended concentrating all efforts in road construction. 3.20 ARAN, nevertheless, has concluded contracts to install electrical transformers and mains along the North-South Boulevard and Road R8, as shown in Diagram 1. The same contract includes a general water main, ten fire hydrants 12 and the necessary piping to allow eventual individual hook-ups in the pilot upgrading zone. The public standpipes were eliminated after a virtual veto from the water company which is already suffering serious financial problems and maintains that it cannot afford to offer free water from standpipes. Institutional problems still threaten the continuity of Nylon's infrastructure and the electric company has not agreed to a cost recovery scheme which would finance future investment in Nylon. 3.21 Garbage Collection. With the reduction in road construction, the garbage dumpsters had little purpose and were eliminated, but the city has placed containers at strategic locations along the new roads and the community has organized a collection system which serves unpaved roads and unimproved neighborhoods (para. 4.11). Final garbage disposal and maintenance of the new roads and drains will depend, however, on relations with the municipality of Douala, which are presently elusive. Although personal relations between ARAN officials and the city's leadership were essentially cordial throughout implementation, the municipality was never given a clear role in the project. A recent impasse reveals the fragility of this arrangement as garbage dumpsters have not been cleaned in over a month (photo 1). 3.22 Resettlement Sites. An initial resettlement area of 30 ha was developed in DIBOM II, South-East of Nylon. Dibom II, with 923 fully serviced plots (individual electric and water hook-ups and dischargeable septic-tanks, street drainage, unpaved -- but banked and curbed -- streets with sidewalks), was completed in 1984. Although slightly less than half of the plots appear to be currently occupied, virtually all of the remainder have houses under construction and the neighborhood already exudes a vibrant air (photo 2). The high quality of the early construction w would lead one to believe that Nylon's evicted residents either fooled the census-takers about their low incomes or that they subsequently sold out to wealthier families (photo 3). 3.23 A second resettlement area was opened in 1987 in the Ndogpassi III area, where a 279-ha site able to accommodate over 3,000 plots was incorporated into State lands by Presidential Decree. In 1987, ARAN developed a first of 34 ha at Ndogpassi III/1 with 1,032 plots. The development of a second site at Ndogpassi 111/2 with 1,350 plots was scheduled to start in 1988 but the works were postponed due to the Government financial difficulties. The high standards introduced in Dibom II were discontinued at Ndogpassi, to the frustration of many resettled families who were charged the same rate per square meter as at Dibom II (Photo 4). With a simply paved road and staked out lots, Ndogpassi, in fact, followed the model of unserviced plots proposed in the first project proposals. The process of eviction and relocation is described in greater detail in paras. 3.25-3.32 below. 3.24 Community Facilities. Agreements between the Swiss Technical Assistance Agencies (DDA) and the Government of Cameroon called for the construction of twenty community facilities, listed in Annex 2 (and in PCR paras. 2.20 and 2.21) in addition to the Madagascar Market. High costs and prolonged * This was the consequence of an obligation imposed by the first model house plans as explained in the PCR paras. 2.19 and 3.10. 13 implementation led to alterations in the overall p-.gramming of these facilities, including the elimination of a cemetery and a park, two secondary markets, a secondary health post, a community center for Ndogpassi and the second phase of Ndogpassi elementary school. The Madagascar Market, completed in the first semester of 1989, currently houses ARAN offices and the CPN, but its formal inauguration has been postponed indefinitely since negotiations with the municipality have stumbled (para. 3.21). Given its size and facilities (see photos 5, 6 & 7), the Madagascar Market, once open, could easily replace the city's central market as Douala's principal commercial center 2. Relocation 3.?5 Forced evictions and resettlement were not new to the residents of Nylon, which itself was first occupied by families %'to had been uprooted when the neighboring settlement of New Bell was developed in the early 1950's. This may account for the high degree of compliance on the part of the evacuated population. Starting in early 1983, ARAN began to conduct socio-economic surveys of the households to be displaced using maps produced through photogram.etric techniques. 3.26 MINUR established a scheme for varying compensation and indemnization according to the legal status of the families involved. For the majority who had no legal title and no building permit, MINUH gave no compensation, but a letter of attribution, or right to settle on a newly developed plot in a resettlement site. These families, however, had to pay the cost of infrastructure (roads, electrical and water hook-ups and dischargeable septic tanks in the case of Dibom II) and associated financing charges. ARAN argued, logically enough, that families who remained at Nylon would have to pay for infrastructure, so the resettled families should assume the same costs for an improved living situation. Families who remained at Nylon, however, did not lose their homes in the bargain. The second project's provision of indemnization for all families, to a total of US$ 10 million, may be a mixed blessing insofar as it gives due compensation to expropriated families, but accentuates the plight of the first group who were removed from their homes without such support. 3.27 In theory, the resettlement areas were to offer services better or equal to those of Nylon. However, the DDA evaluation notes that many families never made it to the resettlement areas. Between one a. I three years elapsed between the moment of eviction and the time when lots became available for resettlement and the initial conditions were tough; once assigned plots, families had four months to pay for them in full, the standard MAETUR requirement which presumed the availability of credit. Fewer than 5% of the displaced families actually received loans (para. 3.36). Futthermore, families were allowed to move on to their assigned lots only after they had built according to pre-approved plans with construction costs ranging from US$ 18,000 to US$ 30,000 (photo 3). For families paying rent in temporary lodging, it was extremely difficult to 9 With DDA assistance, ARAN and Animation used local small enterprises as sub-contractors in market construction. 14 manage the various payments, which may explain why the Audit could only confirm that some 10% of the evicted families actually live in resettlement areas outside of Dibom II. 3.28 Since project initiation, ARAN has taken steps to ease the costs of relocation. Most importantly, ARAN cbtained an exemption from MAETUR allowing up to one year for plot payment and it is currently seeking a further extension of up to five years. Insofar as ARAN has respected families' rights to resettlement plots even when they do not pay within the time limits, a de facto extension already exists. Moreover, ARAN now offers families the option of moving on to half of the plot when half of the cost is paid. New, lower cost housing models emphasizing progressive development were approved in 1989 (PCR para. 2.19) (photos 8, 9 & 10) and CPN has made close to 30 home construction loans (para. 4.17). Finally, agreements were reached whereby commercial property-owners were not evicted from their roadfront lots, 10 but instead became co-owners of the plots in back of their former properties. Thus, continuity of the commercial street front was guaranteed and a significant number of evictions were avoided. Temporary housing is now available in Nylon for about 200 families (photos 11 & 12). A flexible attitude and working relations with the community (paras. 4.07-4.10) have helped ARAN to resolve problems as they have emerged. 3.29 To an unfortunate extent, the Bank skirted the entire relocation issue by not financing any resettlement areas. According to the SAR (para. 4.19), "MAETUR/ARAN will... be responsible for... the compensation and resettlement of those current residents displaced by the reblocking and the installation of infrastructure and community facilities. The Government has also acquired and developed 30 hectares for resettlement requirements in Nylon and additional land is available as needed. This has been confirmed at negotiations." 1 Early Bank supervision reports considered the resettlement component as 100% complete once the lots in Dibom II were allocated. However, they did not appear to be concerned that Dibom II remained initially largely unoccupied or that it could not accommodate even half the number of displaced families, estimated at 2,000 it, the feasibility study (and an item in the 1980 issues paper), and at 3,000 by 1984. 3.30 The 1,032 plots of Ndogpassi III/1 only became available in 1987 and the occupation of this site proved slower than expected. Ndogpassi will remain a thorn in ARAN's side for some time to come. Still in a very rural state, its access roads are as much potholes as paving (photo 4), while residents wait for the extension of the water and electricity distribution networks. Ndogpassi is adjacent to Nylon but some ten kilometers from the center of Douala, and the recent construction of a handsome primary school at the site eases travel for children (photo 13). The development of Ndogpassi 111/2 was delayed as a result 1o Plot owners along the East-West Boulevard were evicted and their properties resold at commercial rates. 11 This was apparently not the case, however, as a mid-1984 supervision mission reported that Dibom II was still awaiting installation of septic tanks and plot allocation. 15 of Government financial difficulties. The proposed settlement area for the remaining families has yet to be cleared and provided with access ways, even though 1,600 families have already been notified of their eviction. At the moment, Ndogpassi 111/2 is a forest (photo 14), although its development is assured through financing under the second urban project. 3.31 As a result of the delays in providing resettlement sites, dislodged families looked for abandoned lands on which to settle. This situation was worsened by the absence of formal land development operations in Douala, and uncontrolled settlements have grown up around the airport recalling the beginnings of Nylon itself. A DDA-financed study undertaken in one squatter settlement on the "axe lourd Yaound6-Douala" in 1987 revealed that 700 families displaced from Nylon had temporarily moved into the area. 3.32 The second urban project initially included the development of 1,000 lots in Ndogpassi (SAR Annex 2-1, para. 14 (d)), as, at the time of appraisal (November 1987), it was still expected that the Government would finance the development of an additional 1,350 new plots at Ndogpassi 111/2. It could be argued that the growing number of displaced families resulted in large part from the increased standards applied to the road construction (PCR naras. 2.10-2.11) and that the Bank should have considered the additional relocations as part of its increased financing of road costs. When it became obvious that the Government would not finance Ndogpassi 111/2, the Bank did so and agreed to make financing available for the development of sufficient resettlement plots. It likewise requested ARAN to update the list of families' rights to resettlement plots and to ease relocation conditions (para. 3.28). In any case, the point may be marginal since present estimates of the final number of displaced families, after complete reblocking of Nylon's 13 neighborhoods, run to approximately 6,500, a full 30% of the original population. Furthermore, ARAN has reportedly modified its upgrading strategy to ensure that all future evicted families will have access to a resettlement plot and receive payment of due compensation. 3. Land Titling 3.33 Although the SAR indicates that, compared with the rest of the project, land regularization is "a much slower process" (SAR para. 2.63), it could not foresee the problems which would ultimately arise when ARAN, focusing on the pilot area and the resettlement zones, found itself utterly without established procedures or legal precedents and was left to develop original survey techniques and to coax the collaboration of various ministries (PCR para. 2.17). The Bank included financing for an expert in cadastres, but to be located at MINUH. DDA assistance filled the gap by helping out at BLAF (Bureau de Liaison des Affaires Fonci6res). "Report of Jean-Claude Bolay to DDA, 1989, pg. 8. 16 3.34 As of the visit of the Audit mission, BLAF had obtained legal land titles for only fourteen families in the pilot upgrading zone, out of a total of some 2,000. The titles have yet to be turned over to the residents. BLAF is waiting for an appropriate moment, such as the inauguration of the Madagascar Market, to present the first titles with due ceremony. To qualify for a title deed, residents must first cancel the upgrading development levy and pay additional costs running about 120,000 FCFA (US$ 428), to cover, among other things, survey and mapping and various officia' papers and stamps. This comes to an average of about 5,000 FCFA (US$ 18.87) per square meter. 3.35 Issued titles, moreover, are conditional upon the residents' building and occupying a dwelling meeting municipal building standards within two years. The PCR (para. 3.10) points out the implicit contradictions which had to be overcome to get this far; legal title depends on an architect-designed, legally sanctioned construction, with a building permit which can only be granted for legally-registered and titled plots. In order to avoid future disappointments, BIAF has obtained certificates from the municipality which confirm that houses on titled lots already, in fact, conform to municipal standards. Even after home construction is approved, title holders must wait another three years before they can sell. Such complications suggest some of the problems which ARAN and BLAF had to deal with in establishing a titling procedure at all, and, more fundamentally, in convincing other authorities that it made sense to legitimize tenure of Nylon residents, or of those in any squatter neighborhood. 4. Home Improvement Loans 3.36 The home construction loans foreseen at the time of Appraisal and in the Loan Agreement were not implemented as such for a variety of reasons. CFC was supposed to provide some 450,000,000 FCFA from its own funds for loans to the residents of the pilot upgrading zone for service hook-up, home construction and improvement and plot purchase (SAR para 2.07; Loan Agreement Schedule 2, Part A (ii) and Part D, CFC Project Agreement, Section 2.01). At the same time, Central Government was to loan CFC some 600,000,000 FCFA. As the President of CFC recalls, works in the upgrading zone were still stalled by the time of project completion and CFC had not been able to make loans to the residents partly because the stipulated upgrading had not taken place and partly because the financial crisis kept Central Government from providing CFC with its advance. Indeed, CFC has not received its standard budget appropriation, based on a national payroll tax, since 1986. 3.37 Even though the home loans for the upgrading zone never materialized, ARAN sought ways to develop a credit line to help evicted families to build in resettlement areas which eventually resulted in 120 loans through CFC and 30 from CPN. The complications which arose suggest that CFC had little interest in risking a credit program in Nylon (PCR para. 3.06). Regrettably, CFC did not keep separate accounts for its loans to Nylon residents and there is, thus, no way to analyze cost recovery or loan performance. Both programs are explained more fully in Section IV C. below and in the PCR (paras. 3.00-3.10). 17 5. Small Business Support 3.38 Agreements with CAPHE to open a small business support center in Nylon were never confirmed, although the assistance to the CPN -- financing of a bookkeeper and a general manager -- was in place from 1983-1984. These activities were to be complemented by a line of credit for small business under a separate Bank loan (SAR para 2.10). This loan, however, was, in fact, geared to businesses larger than those found in Nylon (para. 4.20). During the course of implementation, moreover, the intermediary agency, BCD, closed, so results of Bank efforts in this area were minimal. Nonetheless, ARAN and Animation have been active in identifying, organizing and assisting small businesses in Nylon. The DDA has sent several consultants and hopes to set-up a fully-fledged artisan and business support program with the completion of the Artisan Center scheduled for 1992 (paras. 4.21-4.24). 6. Studies and Technical Assistancp 3.39 Studies to develop a second project were completed satisfactorily (PCR para. 4.17) and led to the negotiations of a second project in 1988. Technical assistance, concentrated on MINUN, was increased by about 50% beyond appraisal estimates (see Table, page 8). In particular, the economic advisor remained in place for an extra six months, while the cadastre expert's contract was extended from 24 to 42 months. In addition, specialists in the following areas worked with MINUH: sanitation and drainage - 48 months; civil engineering - 36 months; and institutional analysis - 4 months. While the assistance to ARAN appears to have been minimal, but useful, adding manpower support to the project team, MINUH's staff report that the consultants tended to work in isolation, leaving little useable knowledge or techniques behind. IV. POINTS OF INTEREST A. Cost Recovery 4.01 The SAR (para. 5.02) specified that the costs of the secondary and tertiary infrastructure installed as part of pilot neighborhood upgrading would be recovered from the residents. In addition, the investment in the market was to be recovered through rents charged to vendors. The SAR, however, left the question of cost recovery in resettlement areas open and gave only vague and even contradictory indications as to how overall cost recovery should be managed. Table 5 of the SAR indicates that final beneficiaries would repay improvements at terms of 4 years with 2 yearr grace at 5% interest. 4.02 CFC was expected to offer 90% financing to final beneficiaries over 15 years with one year of grace "at an average rate of 4.5%." The SAR (para. 3.05) stipulates that CFC would be the financial intermediary .nd would assume all foreign exchange risk which was reaffirmed in the Loan Agreement. Since CFC's lending component was calculated to cover only half the residents of the pilot upgrading zone, the SAR left open the question of how the remainder of the beneficiaries were to be financed. And neither the SAR, nor the Loan Agreement mention any condition or form of cost recovery from the resettled families. 18 4.03 Despite its irresolute mandate, ARAN took cost recovery very seriously. Judging from both the terms of reference and subsequent supervision mission reports during the first four years of implementation, ARAN got little guidance from the Bank in this area, and covenants with CFC did not mention its role in debt collection. Left to itself, ARAN, nonetheless, developed an effective system which aims to carry the full cost of upgrading all of Nylon to completion through 1996 (PCR para. 3.16). This implied the recalculation of charges to residents in terms of future costs, including studies, legal and administrative expenses, rather than just completed investments in civil works as implied in the SAR. The project evaluation financed by DDA sets the amount to be recovered from project beneficiaries at 16% of the total cost, is considerably higher than the 5% stipulated by the SAR. 4.04 In principle, ARAN based debt service charges (the "development levy") on MAETUR's financing terms which called for 100% repayment in four months (para. 3.27). Even after MAETUR granted an exemption, the new terms for Nylon remained tight: 20% in four months and 80% within a year at 5% interest. For residents of the pilot upgrading zone, whose monthly household income averages 90,000 FCFA (US$ 340), 1 the monthly payments on a 100 s Iuare meter plot would come to 41,600 FCFA, or close to half of family income. ARAN has, however, taken its cue from MAETUR in allowing families to pay as they can (para. 3.28). Residents of resettled areas , who are not permitted to move on to their plots before completing payments, have a clear incentive to meet ARAN's requirements and have made an impressive contribution thusfar. 4.05 After five years, the costs of Dibom II are 89% reimbursed. Recovery of Ndogpassi has reached 48% after three years. ARAN enjoys less leverage in the upgrading areas where families already reside and where improvements, by and large, are yet to be seen. After two years, ARAN has recovered only 23% of the projected costs from the first pilot upgrading zone and only 1.6% for the second upgrading zone, Bilongue, after one year. ARAN does not calculate financial costs beyond the first year, so the real cost recovery rate is actually less than these amounts ARAN has currently proposed new financial terms to MINUH and to MAETUR which would allow families to pay over five years at 5% interest, possibly with a threat of eviction for non-payment. Based on its work with the community, ARAN feels assured that the application of an easier, but regulated, payment system would be acceptable to residents and would speed up cost recovery. 4.06 Given the current economic crisis and ARAN's tough payment schedules, it is noteworthy how much families have paid in development levies, particularly 1 But it is not clear whether or not the additional US$ 13 million from the second project is included in the total. Average reported by BLAF. Other estimates from Bank sources set median household income for all of Nylon somewhere between 30,000 and 40,000 FCFA per month. i The development levy is calculated according to the size of the beneficiary's plot. Cost per square meter also varies with plot location, increasing as plots get closer to the main roads. 19 since ARAN uses no regular collection system aside from community meetings and promotion through Animation. At ARAN, only three people work part time in the collection and cost recovery department. 4.07 Insofar as residents pay on the basis of plans, and not for completed works, cost recovery and, indeed, the project's continuity depend on ARAN's credibility which is currently in jeopardy. Since 1986, the Government of Cameroon has withheld appropriations for ARAN's operating budget and ARAN has used its income from cost recovery payments to finance project administration. Consequently, it has not accumulated enough funds to complete services in Ndogpassi (para. 3.29), nor to begin works in the second upgrading zone, Bilongue. Upgrading in the pilot zone is also lagging and, in all likelihood, these factors combine to discourage residents from meeting cost recovery requirements. The sale of plots at Ndogpassi 11/2 should produce a new source of income for ARAN and the agency's director is also working on a plan to reduce staffing and, consequently, its operating costs. Nonetheless, the continued success of cost recovery operations at Nylon will depend on continued visible progress. 4.08 In addition to neighborhood upgrading, revenues from the Madagascar Market were expected to add to cost recovery funds for reinvestmenc in Nylon. This scheme, however, has run into two snags. First, the municipality claims that it, and not ARAN, should dispose of the income from the market for city- wide needs. Second, the operating and investment costs were considerably higher than originally foreseen, making it difficult to attract vendors, particularly from Nylon. ARAN was forewarned. A rather outspoken consultant's report in 1981 stated that "it would be a fantasy to imagine that any municipal market could be a profitable venture, let alone amortize the investment in its own construction. It has become clear that the World Bank's assumptions are completely unrealistic and partially wrong... It is not clear where the World Bank got its rate schedule, but it is wrong.. .The operating costs were not taken into account by the Bank." " At this point, ARAN is not counting on income from the Madagascar Market. B. Community Participation and "Animation" 4.09 Besides its physical achievements, the project generated a series of community development projects at the neighborhood level, owing to the mutual cooperation between ARAN, the newly formed project administration unit, and Animation, the existing community organization, which made Nylon such an attractive place to work in the first place. Animation grew out of the joint efforts of Sister Marie Roumy, a French nun who went to live in Nylon in 1973, and the community leaders who were Nylon's founding fathers. Originally formed as a civil defense network in an area with no police protection or public lighting and which was cut off from the rest of the city, Animation evolved into a mutual assistance association that undertook major infill works when Nylon was little more than a swamp. 1 Dolf, Benedickt, Mission Report of June 1981. 20 4.10 Relations between ARAN and Animation passed through a rocky period at first as ARAN moved into direct works and relegated the community to an observer status. In 1986, a DDA consultant's survey showed that few residents understood what ARAN was and that those who did thought badly of it. A hefty percentage thought that a "Mr. Laran" was going to evict all of Nylon and keep the land for himself. Relations between the community and ARAN became smoother once the issue of expropriation was settled (paras. 3.25-3.30). ARAN's flexibility respect. Five meetinghouses for Animation were refurbished under Swiss-financed project components. 4.11 The peaceable acceptance of the expropriations and the progress of cost recovery are due, in large measure, to ARAN's present good relations with the community of Nylon, which in turn stem from ARAN's collaboration with Animation. Community members help un censuses and measure plots for the calculation of cost recovery charges (para. 4.04). In 1987, under ARAN's auspices, the leaders of Animation, presided by Soeur Marie, became directors of the CPN. In addition, Animation established a youth organization, Coopje, which has close to 1,000 members and collaborates in neighborhood upgrading projects such as trash recycling, garbage collection and infills. A women's association containing about 500 housewives is also involved in community improvement, handicrafts and small businesses. Residents of upgraded areas are currently building storm water drains along new roads. ARAN donates the cement and the technical advice, while Animation furnishes cement blocks and labor. 4.12 Tf, as Animation leaders mention, the project has laid a certain groundwork for community-managed upgrading, the project has likewise enjoyed a multiplier effect thanks to Animation. The individual investments in home improvement, as well as the neighborhood efforts, bespeak a confidence in the project and in ARAN which stems, at least in part, from its association with Animation. Community development, as opposed to physical upgrading, could be measured in terms of the number of families who have chosen to remain in Nylon versus those who opted to cash in on their appreciated property values by selling and moving elsewhere. It is, therefore, telling that Nylon's population has almost doubled since the project was first initiated, suggesting that new families are moving in, while the older residents are profiting from upgrading, but staying put. ARAN and Animation appear to have hit upon an inspiring form of collaboration whereby ARAN installs the infrastructure and Animation insures the community's continuity. C. Crddit Foncier - Home Construction Finance 4.13 The issue of home improvement loans and the SAR's undefined role for CFC has already been mentioned (para. 4.02). With its steady income from a payroll tax, CFC subsidized interest rates and had little desire to branch into unknown areas. The Bank, however, openly hoping to convert CFC into an independent home finance operation, funded a study aimed at identifying ways for CFC to expand its activities, specifically in the area of low-income housing. (PCR para. 3.09) 4.14 The study, presented in 1982, focused on the evident high capacity to save among low-income Cameroonians, the viability of a savings and loan system and the need to eliminate CFC's practice of differential interest rates. This 21 study pointed out the inherent weaknesses involved in the multiplier/credit system based on the French contract-savings plan ("dpargne-logement") and proposed a simpler savings and loan approach modelled on the British "building societies". One month after receiving the Bank consultant's report (April 1982), CFC proceeded to establish a home-savings program which ran precisely counter to his recommendations, requiring regular fixed deposits and offering automatic loans for three times the amount saved. As the consultant predicted, CFC's saving for housing program failed within two years and the Bank abandoned further efforts to work out a home-financing program through CFC. 4.15 Two years later, the same consultant reappeared in Douala, this time at the invitation of the Swiss technical assistance agency to work with ARAN and at the grass roots level. He analyzed the savings capacity of the local population through the "tontines" (an important traditional form of group savings), the local credit union, CPN (Caisse Populaire de Nylon), and the CFC to develop a three tier system to generate savings at the lower levels (tontines and CPN) and to guarantee loans from CFC. The consultant pointed out that the CFC approach, typical of housing finance institutions in many developing countries, effectively inherited the biases and outlook of the "Socidt6s Immobili6res" -- parastatal construction companies -- focusing on construction projects to resolve the housing problem of low-income families. This problem, however, in reality, had very little to do with a lack of construction capacity, but, rather, involved the limited access of low-income people to legalized urban land and credit. CFC confirmed the consultant's suspicions by not appearing at meetings organized with ARAN and failing to provide standard information about its operations. The consultant's report conta..ns a dignified chapter of blank pages in the place of CFC's contribution. 4.16 Undaunted, the DDA funded two more studies focused on the CPN. The first, a rather discouraging evaluation of the CPN by the Union of Caisses Populaires de Yaoundd cited the agency's disorganization and lack of credibility. The second, carried out in 1986, analyzed CPN in terms of the recommendations of the original housing finance study. The three-tiered savings and loan system remained unworkable because the upper two tiers lacked interest, in the case of CFC, and ability, in the case of CPN. The study did note that CPN had increased its membership to 428, but despite technical assistance financed by the Bank (para. 3.38), its record-keeping had not improved, it was not legally incorporated, its portfolio was extremely limited and it was not well-known in Nylon. 4.17 Since it merged interests with Animation in 1987 (para. 4.11), however, CPN has taken a new lease on life. Membership has doubled annually and the savings portfolio has grown accordingly. The CPN has already made about thirty home construction loans for around 3,000,000 FCFA each and is making small business loans as well. DDA has agreed to provide an equivalent of one million Swiss Francs to the CPN. The new national director of CFC has expressed interest in working with CPN through a two or three tiered guarantee system, although the local director of CFC professes to know nothing about projects in Nylon. So the Bank's original hope for a savings and loan program catering to low- income families may yet materialize. 22 4.18 In retrospect, the point stressed in the original housing finance study has been borne out. The issues which clouded formal sector financing agencies do not appear to be tied directly to the question of land titles or the salaries of the borrowers; rather they involve a viable guarantee. Both CFC and commercial banks have made substantial loans to residents in Dibom II and Ndogpassi, where legal tenure has yet to established. However, they chose their clientele carefully and assumed that the houses to be financed would have a high resale value (PCR paras. 3.06 and 3.08). It remains to be seen whether or not CPN will be able to proviA- satisfactory guarantees to CFC. D. Support for Small Businesses and Artisans 4.19 The project components designed to assist small businesses and artisans were rather unfocused and did not have much effect on Nylon (and, in the case of the Swiss financed components, have yet to take place.) Nonetheless, the project has resulted in several advances in small business development. 4.20 The Bank had proposed a three-pronged action, one of these being to help organize CPN in the eventuality that it could offer small business loans (paras. 3.38 and 4.16). CPN may yet become an effective source of finance for Nylon's entrepreneurs, but it will not be due to the Bank's support. Bank financing also covered eight man-months of technical assistance to CAPME, to develop training programs in coordination with the BCD (Banque Cameroonaise de Developpement) business loans project. CAPME, however, never signed the agreement officializing its participation in the project, although it did develop training programs in Nylon, particularly in conjunction with contractors who worked on construction of the Madagascar Market. The BCD program was really oriented to enterprises of a larger scale than those that generally exist at Nylon, offering loans from 1.5 million FCFA (US$ 5,555) to 10 million FCFA (US$37,000). During the three years prior to BCD's demise, it extended 118 such loans, but none is recorded as having benefitted Nylon residents. 4.21 DDA, working on the Madagascar Market and a proposed artisan center, was inevitably drawn into the realm of business support. By obligating the contractor of the Madagascar Market to sub-contract out to small local firms, DDA and ARAN helped promote the local small construction industry and evidently improved construction time and costs. The market itself is supposed to benefit Nylon's merchants and entrepreneurs, but negotiations with the city are currently stumbling over, among others, the issue of what percentage of vendors should come from Nylon. ARAN is pushing for a 60-40 balance, while the city is arguing for a 50-50 split. In any event, the deciding factor will be the rents charged to vendors (para. 4.08). An attempt to form an all-Nylon "self-managingh market failed, apparently to the relief of the vendors who prefer that the city of Douala run the market. 4.22 Even so, the mere presence of the market and the construction of its access roads appear to have stimulated local commerce. A consultant's report in 1981 counted some 500 vending stands in all of Nylon. The number of merchants in the old market place plus those along the roadways would now easily reach five times the earlier total. 23 4.23 The DDA consultant who visited Nylon in the early 1980's recommended against building an artisan center, noting that previous attempts to organize artisans into cooperatives had failed and that the groups involved -- artisans and their supporters --could not agree on what exactly the center should consist of, despite a general feeling that such a facility would be a good thing. The artisan center appears to be no better defined today, but its construction will go ahead shortly, based on a flexible floor plan whereby partitions can be shifted to accommodate different uses over time. 4.24 Local efforts to promote artisans and entrepreneurs began in the late 1970's through the combined efforts of Animation and PDI (PanAfrican Development Institute). Several small businesses were founded along import substitution lines, notably: Confijunyl, which produces candies and juices; Siropcam, manufacturing fruit juices; Chipscam, a maker of plantain chips and a needlecraft group. COOPAN, an umbrella organization serving the artisans of Nylon, had a sales outlet for members, but appears to be in decline at the moment. PDI and Animation ran small, but fairly successful, youth training programs in metalwork, carpentry and handicrafts, but floundered when it came to organizing their graduates into cooperative business ventures. Given the general rejection of cooperative businesses, Animation appears to be focusing on individuals. Concurrent with the project, the youth cooperative, Coopje, has helped get young people started on, among other enterprises, home poultry, cigarette sales, carting and popsicle vending. New hopes are pinned on CPN as a promoter of local businesses. 4.25 The particular efforts of Animation and ARAN notwithstanding, Nylon's physical development has been, perhaps, the project's greatest support to local business. The new connection to Douala has stimulated Nylon's artisans and has generated considerable commercial development along the main roads. Individual electric and water connections - - not to mention public street lighting - - should do even more to increase the productive and commercial capacities of Nylon's entrepreneurs. E. Economic Rate of Return 4.26 The SAR presents an economic justification of the project based on a projected increase in Nylon's rental values and on appreciation of land costs as a result of infrastructure investments in Nylon, leading to an overall ERR of 18%. The PCR (para. 3.18) reports a tentative completion ERR of 24%. Indeed, the project evaluation by DDA notes that the average land price in Nylon has jumped from 500 FCFA/m2 in 1983 to 1000 FCFA/m2 in 1988 (the exchange rate has gone down by 2% over the same period). Commercial land runs from 5,000 FCFA/m2 to 30,000 FCFA/m2 and land in the pilot action zone now sells for 2,500-9,000 FCFA/m2. Rents have increased by 50% since 1983. The fact that Nylon's population is now estimated at 220,000 indicates that the area's carrying capacity has increased by about 245%. 4.27 On the other hand, final project costs within Nylon will probably exceed the original investment estimates by about the same proportion. If land values increased so dramatically with the constructton of roads and infrastructure in a limited part of Nylon, it is difficult to believe that they will continue to increase at the same rate when Nylon as a whole is upgraded. 24 According to the present tally, some 73% of the project funds are invested in primary roads, engineering studies and technical assistance. Were the ERR revised to measure total investment in terms of overall benefits to Nylon (or to Douala in terms of land appreciation), the results might have been less sanguine. The Swiss evaluation adds a further note of caution, "it must be emphasized that the distribution of (project) benefits has not been particularly even and that the total o Peration will present a serious negative outcome for the dislodged families." To this, the Audit adds that a true ERR will depend on final project costs. F. The Second Urban Project 4.28 The Government requested the Bank to consider a second urban project focusing essentially on the upgrading of urban transport and drainage infrastructure in Douala. Engineering studies, financed under the first urban project, began in early 1985 and appraisal of an infrastructure project for Douala took place in February 1986. Negotiations were scheduled first in July 1986, then in October 1986, but never took place. The dialogue continued between the Government and the Bank, however, on a low-key basis, due partly to the illness of the project officer and partly to the Bank reorganization. In September 1987, the Government requested that the project be broadened in scope to complete elements of the first urban project, to undertake policy actions in line with its strategy for the urban sector including resource mobilization. Aside from its focus -- away from urban upgrading and low-income settlements - - the second project, which was appraised in November 1987 for a total cost of US$222.4 million, differs substantially from the first in its implementation scheme. A project development unit (PDU-2) based in MINUH is responsible for overall administration and the municipalities manage implementation of subprojects, with financing from CFC. 4.29 The first project led to the realization that municipalities must become involved in urban development projects and that MINUH should be the single project overseer. The second urban project also includes about US$20 million to complete works initiated in Nylon under the first project. Negotiations proceeded smoothly, once the Douala infrastructure project was redesigned to cover a range of municipalities. Despite the project's initial concept, the Government of Cameroon was not interested in concentrating investments in a single area. V. PROJECT IMPACT A. Physical Impacg 5.01 The project has had a major positive impact not only on Nylon, but also on the structure of the city of Douala. The new road networks link the project area to the city center, increasing considerably the area of influence of downtown Douala. Furthermore, they offer the promise of public services (water, electricity, drainage and garbage collection) to the 220,000 residents ' Dossiers Sectoriels 25 of Nylon. At the other extreme, Ndogpassi, located 15 kilometers from Douala, should, over time, become a new growth pole. The project's dynamic impact (PCR para. 3.01), however, should be qualified. The expropriations have also led to the formation of new squatter settlements and sites which will require future upgrading (paras. 3.30 and 5.13). B. Environmental Impact 5.02 The dredg.ng of the Mgoua River in 1982 evidently improved Nylon's environmental conditions, reducing the residual impact of annual floods. Since that time, however, the silting up of the Mgoua and the construction of secondary roads under the project, coupled with the completion of a major highway (1'Axe Lourd) which converted Nylon into a virtual basin, have complicated the area's drainage (PCR para. 3.04). In 1988, ARAN was obliged to undertake a study to find a solution for the .new puddling which occurred and has subsequently organized community works to help drain house lots. Nonetheless, according to an independent evaluation done by DDA, "at least one third of the lots in the pilot zone run the risk of having poorer drainage after the completion of works than before." is 5.03 While engineering supervision consultants note acidly that there is no solution to Nylon's drainage and add that "the project would have done better to raze the area completely," residents can be seen undertaking individual inf ill projects to raise the level of their house plots by up to 1.5 meters. The community itself is building and maintaining secondary street drains (para. 4.11). Nylon's residents are, evidently, not ready to give up and the project appears to have sparked both communal and individual efforts to deal directly with environmental problems. 5.04 As the PCR (para. 3.03) notes, the lack of piped water continues to threaten Nylon's environment. With only six public standpipes, Nylon's residents draw water from individual and communal wells on the same plots where latrines discharge black and gray waters, thereby presenting a major health and sanitation hazard. However, DDA evaluators note that 10% of Nylon's residents now have piped water connections, as opposed to a mere 1% in 1980. 19 Again, it is to be hoped that residents will themselves eventually install the solutions which the project itself could not implement. C. Institutional Impact 5.05 When analyzing institutional development, it is difficult to distinguish the input of the first project from that of the second or from the force of circumstance. In any case, the advent of the former appears to have served both as a first step and as a filter. Aside from those noted below, certain institutions, such as CAPME and BCD, dropped out of the urban development sector altogether. Others, which were not taken into account in project design, such as the municipality of Douala and the State Land Bureau, proved too 18 Dossiers Sectoriels 19 Dossiers Sectoriels #9 p.3. 26 important to be overlooked and have been brought into more active roles in urban development. Indeed, the most important changes to be noted concern the evolution of inter-institutional coordination among the agencies involved in the first project and the growing awareness of an urban sector interest beyond that of individual institutions. 5.06 MINH was charged with overseeing the first project and preparing a second one. As noted in the PCR (para. 1.12), however, MINUH tended to eschew its role as coordinator. Its involvement with ARAN was somewhat erratic and the project coordinator operated quite apart from the rest of MINUH staff, as well as from other agencies. In retrospect, the technical assistance offered through the project was not seen as particularly helpful to MINUH over the long term (para. 3.39). MINUH did, nonetheless, carry out studies for the second project, appraised in 1987, and its urban development unit subsequently appears to have taken on new life.2 Implementation of a large and complex program appears to be going smoothly. In the sense that the first project prepared MINUH for a follow-on operation, including the flushing out of inefficient management systems, its impact must be considered positive. 5.07 The project gave MAEglt its first real taste of land development for low-income families and pointed out the limits of the agency's procedures and standards for the target population. Under the second project, MAETUR appears better equipped both to meet the ieeds of the low-income population and to take on a broader role in urban land development. It is not clear, however, whether MAETUR should be involved in upgrading projects or whether these should be left to the cities themselves. As the PCR points out (para. 1.12), moreover, relations between ARAN and MAETUR were often sticky and the positive impact on ARAN was not also transferred to MAETUR. 5.08 While the project appears to have failed in its attempts to develop GC (paras. 4.13-1.17) as a major credit institution for low-income families, the present director of CFC, notably influenced by the agency's involvement in the second project, has expressed clear interest in augmenting its savings portfolio and in working out a loans program to serve the informal sector. Interestingly, the director has requested copies of the Bank consultants' reports written under the auspices of the first project. Although CFC's interest may be largely a temporary result of its present economic crunch, the first project may still prove to have had an important impact on CFC. 5.09 At the local level, the project doubtless strengthened Animation and set CPN on stronger ground. The most marked institutional development, however, occurred at AAN, which seems quite capable of completing the upgrading of Nylon on its own, barring unforseen economic problems. The question remains, however, as it often does in urban projects of this sort, what will become of ARAN after 1.996? 2o The change may also be attributed to a difference in personalities. Furthermore, while the former director of the PDU lacked support from auxiliary staff, his successor has relied on a more open management style. 27 D. Poliev Impact 5.10 Urban policies in Cameroon appear to have gone a full circle. In preparing the first project, the Bank attempted to concentrate MINUH's efforts on informal settlements and urban growth. But as the Bank itself changed its vision of urban development, it seems also to have led MINUH to shift focus. In the words of the director of PDU2, "we are interested in productive investments now rather than social investments." MINUH may be reaffirming the country's original outlook. 5.11 At the same time, as the PCR (para. 2.05) observes, the first project broke ground in several key areas such as cost recovery, legalization of tenure and the installation of services and infrastructure in squatter settlements. It also pointed out the need for policy changes to facilitate future urban development, the most important of which include: a) The -ole of municipalities. The dissocittion of Douala from the first project led to implementation problems and jeopardizes the project's future. By contrast, municipalities play a key role as developers in the second project; b) Land titling. The problems with transferring land from the national government complicated and prolonged the land regularization process considerably. The State Lana Bureau is currently studying a new procedure which will give development agencies the power to control and sell government land; and, c) Resettlement. It has become clear that urban development projects, like neighborhood upgrading, must be accompanied by resettlement and site planning programs. The second project includes components to develop settlement plots in tandem with upgrading and adds financing for adequat%.- indemnization of families evicted for urban improvement projects. This constitutes a de facto recognition of squatters' rights to land. 5.12 The first project's involvement with one quite particular community has opened a new line of communication which can influence MINUH policies directly. Changes in financing terms, in expropriation procedures, in building standards and in land regularization practices offer some examples of policy reform which grew out of the experience at Nylon and developed directly from the dialogue with Nylon residents. This reversal of the previously established pattern represents an important policy shift in and of itself. 5.13 While upgrading projects were not included in the second project, events of the last two years suggest that they may well prove as productive as other municipal development activities. If its cost recovery continues on schedule, ARAN will be able to present a strong case to MINUH for including upgrading projects in future municipal development programs. Even if MINUH chooses not to support upgrading projects, the experience at Nylon has, nevertheless, sparked local interest. The Mayor of the adjacent township of New Bell, moreover, has invited ARAN and Animation to advise on the feasibility of an upgrading program for this community as well. 28 5.14 Douala's agreement to reduce building standards (PCR para. 5.08) and its approval of construction on unregistered plots represents an important precedent, in line with the recognition of squatters' rights (para. 5.11c) The fact that such constructions have already been financed by formal sector institutions is a hopeful advance in credit policy. E. Poverty Impact 5.15 The residents of Nylon were identified as the poorest and most needy in Douala in terms of income levels and standards of living, respectively. The project had a noted positive impact on both. This success remains qualified, however, by the less positive situation of certain of the expropriated families, as well as of those adversely affected by continuing drainage problems at Nylon. VI. ROLE OF THE BANK 6.01 The Bank clearly played a key role in developing government policy and support for urban upgrading prior to project appraisal. Judging from internal reports, however, once negotiations were completed, the Bank concentrated on preparing the second project and, unfortunately, tended to leave the first operation largely on its own. Bank missions also became involved in the problems of highway costs, sending in two different consulting groups, and engaged in a long dialogue over alternative construction schemes for drainage canals, the Bank pressuring for a lower-cost, labor- intensive construction system which was eventually selected. But the "software" components of the project were virtually ignored (para. 3.31, PCR para. 4.12). 6.02 The Bank did use its influence to have resettlement sites transferred to MAETUR, but it could have offered considerable assistance with cost recovery, the housing loans program and budget transfers to ARAN and to CFC. The Bank's direct lines of payment to the supervising engineers and its evident concern for technical components may have led project directors to feel that civil works were beyond their control, while the remaining components held little interest for the Bank. The coordination with Swiss missions after April 1988 (PCR para. 4.14), albeit late in coming, was welcomed as evidence of renewed Bank concern for all project components. VII. LESSONS LEARNED 7.01 The Audit concurs with the findings of the PCR (paras. 5.01-5.06), adding the observations in the following paragraphs. 7.02 The first project is typical of the "Christmas tree" approach, prevalent among Bank-supported urban operations in the late 1970's and early 1980's, which focused on a particular site with a multitude of loosely-related components. Experience in Cameroon tends to confirm that such projects can be very difficult to administer locally and defy the supervision capabilities of the Bank. Specific projects, moreover, run the risk of incurring resentment at a national level. Most successful pilot projects take place in the country's 29 capital city and part of Nylon's administrative problems can be attributed to its location elsewhere. 7.03 As to the multiple components, here as elsewhere, the smaller elements tend to be eclipsed and key issues related to them tend to be overlooked by Bank supervision missions unless they have a significant impact on disbursements. Indeed, the project's success in resettlement, land regularization, cost recovery and small business support are due in large part to ARAN's dynamic director, who stayed with the project eight years and to alliances forged with the community, through Animation and the Swiss technical assistance program. 7.04 In retrospect, it was a mistake to give ARAN responsibility for project implementation, while farming out cost recovery to another agency (CFC). The second project has opted for a more streamlined model of administration whereby the municipalities are charged with both execution and cost recovery. 7.05 The project demonstrates, however, that urban upgrading can result in efficient cost recovery and, thus, be replicable. If ARAN fails to carry out its upgrading program for all of Nylon, it will be because of insufficient government backing, rather than lack of local revenues. The experience at Nylon also points to the need to include adequate programs for resettlement and new settlement in conjunction with upgrading and municipal development projects in general. 7.06 The project likewise illustrates the difficulties faced by an executing agency working in isolation. ARAN had to manage with only limited support from MAETUR, MINUH and the Bank. Had the municipality been enlisted in project implementation, the project might have run more smoothly and its future might have been better assured (PCR para. 5.04). Again, the second project has attempted to avoid this problem by establishing a stronger relationship with the municipalities. 7.07 The importance of defining clear and workable systems of financing and cost recovery should be stressed. This includes prior agreement with a viable financial intermediary. It also means developing affordable projects and payment schemes, a lesson which ARAN learned and applied in the course of implementation. Problems with financing and cost recovery have plagued Bank urban projects in the past and will probably continue to do so in the future. Successful experiences, therefore, should be well documented and effective technical assistance 1,p:r)moted in these areas. 7.08 Inasmuch ar* the project could have benefitted from closer assistance with regard to certain problem areas occurring during implementation -- notably cost recovery, land titling and resettlement -- it might be advisable to leave unallocated funds for unforeseen technical assistance needs, particularly in pilot projects. The fact that technical assistance centered on MINUH, rather than ARAN, did little to help the situation. Under the circumstances, DDA's assistance and flexibility were invaluable. 7.09 The impa-,t of a first project may not be felt immediately. Several lessons of this experience are just beginning to be reviewed at the national 30 level and their results will probably not be fully observable until a th:rd project is implemented or until local initiatives, such as that at New Bell, take place. 7.10 The project points out the advantages of collaboration with community organizations and co-financiers working on site. In many instances, DDA made up for the Bank's weaknesses in supervision. The efforts of ARAN and Animation in the areas of small business support and community development, however, suggest that such activities may be better suited for management by donors other than the Bank. 7.11 The troublesome cost overruns experienced in this project suggest that additional emphasis is needed on the review of cost calculations, consideration of alternative solutions and better control of construction activities and costs. VIII. CONCLUSIONS 8.01 The project represented a bold attempt to introduce several new concepts including the upgrading of squatter neighborhoods, land regularization, resettlement on unserviced plots, home construction financing, cost recovery and support for small businesses. The problematic terrain at the project site, Nylon, led to hefty and troubling cost overruns and a dramatic reduction in the civil works program. Compared with Bank urban experience world-wide, however, the project completed disbursement in record time (five years as opposed to 9.5 years), although several components were left to be completed under a follow- on operation. 8.02 The project has had a marked positive impact on the urban environment of Nylon and on the city of Douala as a whole. In addition, it has st-Imulated economic growth in Nylon and strengthened community development. Unlike many other Bank-financed urban projects, cost recovery appears to be going well. Delays in the resettlement of expropriated families and in issuing land titles, together with the failure to generate a viable home financing system and a program to support small businesses, represent the main shortcomings in an otherwise successful project. 8.03 Although a primary objective of the first project was to establish a replicable model for urban upgrading and, to date, no such replication has taken place, the policy changes and interest generated by the experience at Nylon suggest that urban upgrading has gained a small, but secure, foothold among official urban development activities in Cameroon. 8.04 Studies financed under the first project led to the preparation and negotiation of a second urban operation in 1988. The first project identified and took the initial steps to address areas of institutional and policy weakness, without which the second project could not have moved forward. 神： •．••■．闕．“••••，’一一－一一一一·一一”-..，一綢•網•，－一－一·一一一～-－一一一一甲一－·一－.-“•一．．·一－·－一一一一一一一－一‘ 馴 33 ANNEX 2 S V 1 S S CAL C-U-LAT 1 ON-S 1988 1989 1990 199 1 1992 et. - SOLDE Cout ant 1T 2T 3 4T IT 21 3Y 4T IT ZT 3T 4T 0 E P E M S-E 1 ECOLE DE BILONGUE 312 147 97 54 14 CENTRE DE SOBOUM 280 40 75 90 65 10 MAISON SOCIALE DE C.C.C. 26 5 5 il 5 MAISON SOCIALE D'OYACK 26 5 16 5 DISPENSAIRE DE BILONGUE 100 is 35 30 is 5 JEUX D'ENFANTS DE BILONGUE 20 4 8 6 2 ESPACE COLLECTIF DE NKOLMINTAC 222 40 50 52 60 20 JEUX D'ENFANTS DIOYACK 20 4 5 8 3 0 MAISON SOCIALE DE KDOGPASSI 0 MARCHE DE NDOGPASSI CENTRE DE JEUNESSE DE TERGAL 200 10 25 45 65 35 20 COMPLEXE SPORTIF DE DIBON Il 280 50 60 60 70 40 JEUX D'ENFANTS DE BRAZZAVILLE 20 3 9 6 2 ECOLE DE NDOGPASSI 111/2 DISPENSAIRE DE NOGGPASSI 100 20 20 25 25 10 CENTRE DE SANTE DE C.C.C. CENTRE COMMERCIAL DE OISON Il ESPACE COLLECTIF DE NYLON CIMETIERE DE OISON Il CENTRE ARTISANAL DE MADAGASCAR 100 30 40 10 5 0 1M 1-47 -fi 54 54 'fl0 T36 T98 M 29 97 o 0 ETUDES ET CONTROLES 99 _M 2 1 2 o TOTAL EQUIPEMENTS COLLECTIFS 1805 235 99 55 56 82 112 137 199 235 222 237 109 22 5 01 0 MARCHE MADAGASCAR (SOLDE) 599 _ 108 12 56 23 MONTANT TOTAL ....... 2404 235 607 67 112 82 112 137 199 235 2221 23-11 109 22 5 0 23 REPORT .............. 554 -53 nà 111 129 17 207 10_8_ 153 31 14 50 28 23 23 F 1 M A M C E M E M T S S U 1 S S E 2019 789 258 327 280 220 145 0 CAMEROUN 385 20 _ -Ab. -0 1 luui - 100 - 100 - - - - 0 MONTANT TOTAL ....... 2404 789 01 343 100 01 3271 1001 2801 1001 2201 . 145 01 01 01 0 SOLDE A REPORTER ..... 554 -531 M_ 1111 1-1 -- 171 2071 1081 1531 -311-141 so ul 23r 231 0 ANNEX 3 PHOTOGRAPHS Page 1 of 7 Photo I Problems with public service maintenance--like garbage collec- tion--stem from the runicipalityTs distance from the project Photo 2. Dibom II under construction 35 ANNEX 3 Page 2 of 7 Photo 3. The first house models were beyond the means of most Ny1on req*-ents Photo 4. Ndogpassi is more of wilderness than a settlement 36 .NNEX 3 Page 3 of 7 Phote Photo 6. 37 ANNEX 3 ?age of 7 W-P Photos 5, 6 7. The Madagascar Market Photo 8. Differunt house models at Dibon II 38 ANNEX 3 Page 5 of 7 Photo 9 Progressive houses financed by CPN Photo 10. Another CPN housing loan .~~> .~ .' . --- r -- Photos 11 & 12. e c.r us e left side of te <treet -r 1- V right side ,f7 戶祈叫 兀、－．兀，B：丫：一”一 & &”必 作 嫵癱 :&.&to’→．一：！兀d n..&r？二e七亡：雙。註t 11一戶1吃七 &：三。：：、：；一f 41 PROJECT COMPLETION REPORT CAMEROON FIRST URBAN PROJECT (IDAN 2244-CM) October 5, 1989 Occidental and Central Africa Department Infrastructure Operations Division 42 43 REPUBLIC OF CAMEROON URBAN DEVELOPMENT PROJECT, LOAN 2244-CM PROJECT COMPLETION REPORT I. BACKGROUND AND PROJECT DESCRIPTION Sector Background and Project Identification 1.01 In 1980 when the project was appraised, about a third, or 2.6 million people, of Cameroon's 8.4 million population lived in urban areas. In 1987 when a follow-up Second Urban Project was appraised, 40Z of the 10.5 million population were urban residents. The steady rate of urbanization is a major factor contributing to strong growth of GDP in non-oil sectors during this period. Urban-based activities generated over two-thirds of new employment opportunities and real growth in non-oil GDP. Cameroon's unusually well-developed and dense network of urban settlements has allowed, through the development of the internal food market, for increased agricultural production and the related shift from subsistence into market-oriented agriculture. Besides its polycentric structure, the urban network is notable for the market importance of the secondary centers, which play as large a role in absorbing urban migrants as the two main cities put together. 1.02 Douala, the site of this project, is the country's main port and industrial center and also the regional economic center for the provinces of the western region, Cameroon's agricultural and commercial heart and its most densely settled region. Douala is also the largest city, with about 600,000 in 1980 and about 842,000 in 1986. Located at the mouth of the Wouri River on'the Gulf of Guinea, it has been a major trade center since the arrival of the Portuguese in the sixteenth century. Its port now handles virtually all import- export traffic for Cameroon as well as goods en route to or from landlocked Chad and the Central African Republic. The second-largest city, Yaoundd, the capital, is located centrally and has served as the administrative center of Cameroon since 1915. These two cities contain about a third of the country's urban residents. Douala will pass the one million mark within the next few years and Yaound4 will pass it by 1995. 1.03 The identification of the First Urban Project grew out of an extended and fruitful consultation concerning urban and housing policy between the Bank and Government, initiated in 1975 by Government's request for a sector mission. Important aspects of the secter mission's findings were reflected in the urban and housing sections of the Fourth Plan (1976-80) and the dialogue continued through project preparation missions assisted by the Resident Mission, and by an advisor to the Ministbre de l'Urbanisme et de 1'Habitat (MINUB), provided from 1979-81 througi the First Technical Assistance Project (Credit 673-CM). 1.04 The project was initially conceived in 1978 as a three-pronged effort to: initiate urban upgrading,operations, institute a sites and services program, and strengthen the institutions on which the future of urban development depends. Both upgrading and sites and services involve provision of basic road, drainage, water, education and health facilities in urban areas, along with land tenure and housing credit arrangements which encourage private investment in permanent housing stock. Upgrading involved the introduction of these measures in existing uncontrolled and usually illegal settlement areas, while sites and services programs opened new areas. Both types of infrastructure programs were receiving 44 Bank Group support from the early 1970s in a number of countries as the most practical means of addressing the need to extend basic infrastructure networks in the face of rapid urbanization, and to provide for the basic needs of the urban poverty group. 1.05 The sites and services approach formed the basis for Government creation in 1978 of a new Agency for Urban and Rural Land Development (Mission d'AmOnagement et d'Equipement des Terrains Urbains et Ruraux or MAETUR), under the aegis of MINUH. The Government was prepared to envisage a long-range, multi- city program for sites and services but wanted to proceed more cautiously regarding upgrading efforts. The Bank considered that priority should be given to upgrading operations as the main mechanism for clearing the urban infrastruc- ture backlog as well as to sites and services programs, providing that appropriate action be taken to eliminate two major constraints: lack of provision of sites for households below the 25th income percentile, and the discontinuation of heavily subsidized housing loans (interest rates below savings deposit rates). 1.06 The Nylon area of Douala was clearly identified by both partners as the highest priority area to upgrade from all points of view: technical, social, and political. Nylon was the largest slum and the worst squatter area in the country. Housing 90,000 residents in 1980, a sixth of the city's population and a quarter of its poor, Nylon was located on marshy ground along the airport road and directly adjacent to already developed areas, whose runoff it received. Unlike its namesake, Nylon never dried out. First established in the 1950s as an extension area for residents of next-door New Bell, it was for many years designated as a "green area' by the urban master plan. Bypassed by infrastructure extensions, the residents of Nylon, noted for their cohesive social structure, formed an effective and well-organized system of thirteen neighborhood action committees. Beginning in 1970, these committees mobilized work groups equipped with hand tools who filled the swampy land and constructed and maintained makeshift bridges, drainage canals, and dirt roads. 1.07 While the solidarity of the community has been and continues to be a great strength, there was a limit to what it could achieve. In 1980 the area was served by a single water standpipe, a single low-tension electric line, and lacked street lighting. It was inaccessible to buses, garbage vehicles or trucks and deficient in health and education facilities. Vendors were plentiful but there was no market since these are built and run by the city government, and in any case trucks could not have gotten through. Project Obiectives and Description 1.08 The project's main objectives were to raise urban infrastructure standards to an envir 'tally acceptable minimum, and to strengthen the institutional base and % - -op appropriate mechanisms for a continuing upgrading program in Cameroon. The project was designed to attain these objectives by supporting the folloving activities: (a) construction of primary infrastructure for the whole 600-ha Nylon area, and to raise the environmental sanitation conditions and road access to an acceptable minimum. This included the provision of major roads, drainage, dredging of the Mgoua River, water supply, electricity and street lighting, as well as the development and servicing of 45 resettlement areas to accommodate the households displaced by infrastructure construction (56Z of total base cost); (b) complete upgrading of a first neighborhood of 50 ha, including the provision of secondary and tertiary infrastructure, reblocking of plots, land tenure regularization, housing credit, and cost recovery mechanisms (3% of total base cost); (c) community facilities for Nylon and the resettlement areas, including a major market, community centers, public health facilities and primary schools financed by Swiss bilateral assistance (15? of base cost); (d) community facilities in Yaound4 Northwest, an under-serviced densely settled low-income area of the capital city (8% of total base cost); and (e) institution building consisting of support for the urban development institutions responsible for carrying out the project, support to strengthen MINUH, in particular its land administration sections, and financing for preparation cf a Second Urban Project and other urban studies (192 of total base cost). 1.09 In addition to the community facilities program, Swiss bilateral aid (Coopdration Suisse au Ddveloppement et & l'Aide Humanitaire -- DDA) also provided a wide range of support activities for the existing community associations, including three full-time project unit staff members, tools and supplies for ongoing neighborhood self-help infrastructure maintenance, and a health and nutrition education program. The community facilities program was extended twice to include another group of community facilities and further support. This support included training for project unit staff and for small- scale contractors involved in the construction sector, design and establishment of an artisan center in the market, proposals for market management, and management assistance to the Nylon residents' housing credit association. Project Design and Oraanization 1.10 The above project components were designed over a two-year period culminating in a November 1980 appraisal of a much larger project to be financed by a US$40 M IDA credit, twice the amount of the loan eventually approved. The project size was reduced during negotiations in February 1982 at the Government's request by the withdrawal of a large component originally intended to support the launching of a sites and services program and the technical assistance associated with this component. This cut was requested because of the increase in the cost of local financing associated with the shift of the project from the last IDA credit to Cameroon to the first IBRD loan. 1.11 The institutional setup originally agreed at appraisal, to create a project-specific independent agency to carry out the upgrading of Nylon, was eventually retained, despite strong pressure from the Bank to establish a program-oriented one. Once the decrees to create the agency reached the Presidency, the President ruled in summer 1981 that the agency should be created within MAETUR, itself newly created, in order to discourage proliferation of parastatal institutions. This option had earlier been considered and rejected due to the risk of overburdening MAETUR, which was engaged in launching government's first land development efforts. A compromise was sought which would 46 separate the Nylon agency's management and budget from those of the MAETUR land development arm. The setup finally approved left administrative and accounting functions with the central MAETUR office in Yaound4 and operational respon- sibility with the director and staff of the Nylon agency in Douala, named ARAN (Agence de Restructuration et d'Am6nagement de Nylon). 1.12 The subsequent experience with this institutional arrangement was not satisfactory. Poor working relations between the Douala ARAN office and the Yaoundd MAETUR office, which had to approve every administrative and technical action, remained a constant source of friction and delay throughout the project's life. Lack of delegation of authority from Yaoundd along with rigidity and detail-ridden procedures rooted in the legal tradition were to blame for much of this friction. A Steering Committee made up of Ministry, municipal and community representatives was to have played an important role in setting work programs and budgets; this arrangement also did not work: the Committee met only once. The project's financial arrangements, designed at appraisal to strengthen two inexperienced financial intermediaries, were complex, and made the drafting of the legal documentation required for effectiveness quite time consuming. II. PROJECT IMPLEMENTATION AND COST Overview 2.01 The most critical variances between planned and actual project implementation were a four-year delay in start-up of civil works, and a consequent extension from four to seven years of project technical assistance. These delays were due in part to factors known at appraisal but which could not be accurately estimated, in particular the slowness of procurement and land administration procedures in Cameroon. Also, some factors were not foreseen: in particular the need to redesign basic infrastructure in response to the 1985 master plan revision, and the 1986187 economic crisis plus the fall of the dollar. However, new land administration procedures and institutional arrangements worked out during the first project made it possible to plan more realistically for implementation of the Second Urban Project. Start-up Activities 2.02 Government's commitment to the project was clearly demonstrated by the fact that even before negotiations, it undertook about US$4 H in preparatory civil works adjacent to Nylon. A Bank urban sector mission which visited Cameroon in spring 1982 found that works to prepare a first resettlement site to accommodate roughly half of the households to be displaced and to dredge the Mgoua River which drained Nylon, had already been completed under Cameroonian financing. Retroactive financing for these works was originally intended to be included in the project, but dropped at Government's request in order to reduce the size of the proposed loan. 2.03 During the two-year period between negotiations and effectiveness (February 1982--May 1984), activity was stalled on three fronts: institutional, technical andland availability. Institutionally, the responsibilities of MINUB and of the kJnistry,\5 Plan for the overall supervision of the First Urban Project were not clearly defined. Then during the 4rafting in Cameroon of legal documentation required for effectiveness, it became necessary to incorporate the 47 concerns of agencies who had not participated in negotiations. At issue was not the project concept, but rather the project's complex financial arrangements and the procedures for repayment and disbursement of sub-loans. 2.04 Other delays resulted from the requirement that the selection of consultants for design and engineering contracts comply with the Bank guidelines. This requirement acccunted for about four of the six months delay between initial consultant selection and start of design work (December 1983--June 1984). 2.05 On the land availability front, a search was going on for a second resettlement site. The first resettlement site (30 ha) called Dibom II which was part of State lands at the time of negotiations could accommodate only about a third of the households to be resettled. At negotiations, Government had assured the Bank that additional land was available for extension of the resettlement area as needed. In fact the situation turned out to be much more complex. A second site called Ndogpassi I which was set aside for public use in spring 1985 turned out to be unusable because 80 of the 180 ha was titled and would have required compensation, and 35 ha were occupied by squatters. Another site called Ndogpassi III, 3 km east of Nylon, was designated and integrated into State lands in fall 1985. This 279-ha site was large enough (over 3,000 plots) to accommodate the remaining households to be relocated from the project and also those anticipated to require resettlement as part of the Second Urban Project which was then under preparation (see map). 2.06 The two years between negotiations and effectiveness were also a time of intense activity for sector work and identification of follow-on projects. Three interdisciplinary sector missions visited Cameroon during this period, and an Urban Sector Review Report (4540-CM) was issued July 5, 1984. The supervision mission which coincided with the last of these (February 1984) also formally identified a US$200 M Douala Infrastructure Project focussing on major road and drainage improvements in Douala, as well as a Cadastre (land administration) Project and a Secondary Cities (municipal development) Project, in line with the recommendations of the sector work. The First Urban Project included financing for the preparation of a Second Urban Project, and supervision of the feasibility study for this project took up a large share of project supervision staff time. At the same time, the prospect of this follow-on project acted as an incentive to resolve satisfactorily certain complex issues, particularly land tenure arrangements, which arose during implementation of the First Project. Other start-up activities during this period were the recruitment of technical assistance staff and training of Cameroonian counterpart staff. Road and Drainage Infrastructure Design 2.07 Project civil works components accounting for about 60Z of project costs were grouped in two large contracts, one covering the basic primary road and drainage network in the whole of Nylon and the other covering secondary infrastructure in the 50-ha pilot upgrading zone. In early 1985, at the same time that adequate drainage design to cope with severe geotechnic problems was becoming a major concern, Government decided to change the basic structure of the road and drainage network in Nylon to reflect the new Urban Master Plan for greater Douala. 2.08 The above change in the basic structure of the road and drainage network was a fundamental one, involving a shift in perspective from viewing Nylon as a quartier, albeit a large and important one, to viewing it as a major 48 commercial center. For one thing, in the five years separating engineering design from appraisal, Nylon had outgrown its earlier boundaries. Described in the appraisal report as a 600-ha area with 90,000 population, Nylon had grown to 700 ha with 120,000 by the spring of 1984. 2.09 In addition, the time had come in the growth of Douala that the four- lane road east-vest approach to the city had to be extended, and Nylon happened to be the logical location for a crucial link. Such a revision would have been impossible to avoid at this point in time. The same Bank and MINUH teams who were supervising the First Urban Project were simultaneously involved in preparing the Douala Infrastructure Project which was intended to identify those road and drainage investments critical to the efficient functioning of the city as a whole. 2.10 The design consultants for the project were instructed to quickly develop preliminary design solutions taking into account the transformation of the north-south access road into a four-lane highway and the east-vest highway into a six-lane highway. Thirty meter rights-of-way for there segments had been included in the design in anticipation of future needs but initial project cost had not provided for construction of such wide roads. The cost increase implied by this type of design upgrade could now easily be accomLodated, it was thought, since the Bank loan was denominated in US dollars and the dollar had risen from 270 CFAF at the time of negotiations to 400 CFAF. The expected date of bidding documents was postponed from December 1984 to March 1985 to allow for the redesign. 2.11 Final design and final bidding documents were in fact not completed until late 1985, following a detailed review and comments on preliminary versions by a spring Bank mission. No sooner were the bidding documents satisfactorily revised than the terms of reference for on-site supervision (Organization, Pilotage, Contr6le des Travaux (OPC)) had to be rejected. Revised versions of the terms of reference for OPC bounced back and forth between the Bank, ARAN, MAETUR and the Ministry of Public Contracts, the Government procurement ministry, from mid-1985 to mid-1986; the Marchds Publics then proceeded to call for local bidding against Bank procurement guidelines. A short list agreeable to all parties was finalized in August and the Bank was informed of OPC bid evaluations in November. The selection process for the consultants in charge of on-site supervision created additional delays. The preparation of a request for proposals in compliance with Bank guidelines took over one year and Bank approval of the OPC contract was not given until receipt of the bid evaluation report (February 1987), final TOR (March 1987) and the detailed contract (April 1987). 2.12 Bids for civil works in Nylon were opened in April 1986 and the bid evaluation was reviewed with the Bank mission in June 1986. The Bank was informed in September of the July decision of the Procurement Commission concerning the selection of a civil works contractor. Following several months of Bank-ARAN-MAETUR disagreement concerning important details of the winning bid proposal, a Bank mission visited Cameroon in February 1987 and resolved these difficulties. A preliminary engineering report was approved by the Bank in late April, rights-of-way were cleared beginning in early 1987 and road construction began in fall 1987. 2.13 The construction contractors soon ran into two unanticipated problems: geotechnical difficulties requiring the construction of special road bases impervious to water which led to substantial increases in costs, and errors in 49 the quantity survey. At the same time, Cameroon had to face a deteriorating economic and financial environment. Oil production had begun to drop in FY86, but Government spending had continued to increase, drawing on accumulated oil proceeds. By late 1986, the economic crisis due to the fall in oil prices, combined with the weakening of the US dollar, was belatedly making itself felt at the project level. 2.14 Due to these technical and financing constraints, it was agreed following a mission in late 1987 that the Bank would increase its share of the project financing. It was also agreed that the scope of the First Urban Project would be limited to components already committed, by reallocating to these components uncommitted funds which had been reserved for the training program, the Yaound4-Northwest community facilities, non-essential technical assistance, and small components for health education and artisan assistance. From this point on, efforts focussed on the completion of primary roads and drainage, and of secondary infrastructure in the pilot upgrading zone on which construction had begun in fall 1987, and on improvement of the land tenure situation. 2.15 The most important design changes at this stage were to accept a reduced configuration for the east-west highway which provides the main access road to the area, rather than the six-lane highway which under earlier economic conditions would have been affordable, and to realign the north-south road to become a link in a new ring road, while retaining its four-lane configuration. At the time of negotiations for the Second Urban Project in July 1988, which now consisted of a scaled-down Douala Infrastructure Project combined with priority works in Yaound4 and secondary cities, an additional US$13 M was added to the Second Urban Project to complete the upgrading of the Nylon area. All works in Nylon initiated under the First Urban Project are expected to be completed by June 1990. Improvement of the Land Tenure Situation 2.16 The vast majority of Nylon residents had developed their plots outside of all legal procedures for land acquisition, subdivision, development and registration. As a result, they were considered illegal squatters by the urban authorities, and were refused the right to any secure land tenure which in turn limited their willingness to invest in the improvement of their housing. One of the project's major objectives was to develop appropriate procedures and mechanisms to regularize land tenure in uncontrolled settlements. The delivery of secure land tenure would be conditionai to the payment of an infrastructure servicing fee in order to allow for the financial replicability of the operation. Mechanisms to improve the land tenure situation were developed for two types of situationst (a) the upgrading of uncontrolled settlements; and (b) the development of resettlement areas to accommodate households displaced as a result of the clearing of the rights-of-way required for the provision of priority infrastructure in h1lon. 2.17 ARAN, through its land management bureau, Bureau de Liaison des Affaires Foncibres (BLAF), and with the support of DDA, developed detailed procedures for the restructuring and land regularization of a pilot area of 50 ha. All plot occupants were identified by a Land Commission assisted by the neighborhood associations. Aerial photographs were used to restructure the plots layout in order to allow for minimal infrastructure servicing. After the incorporation of the corresponding area into State lands, the subdivision plan was officially approved and individual plots were demarcated. Files were 50 prepared by BLAF for the beneficiaries of the 1703 plots as well as for the 407 households which had to be displaced as a result of the upgrading operation. The first land titles were delivered in mid 1989 after payment of an infrastruc- ture servicing fee averaging CFAF 4500 per square meter. 2.18 The clearing of the rights-of-uay of the primary infrastructure and community facilities planned as part of Nylon upgrading led to the displacement of a total of roughly 3,000 households among the 20,000 households which originally occupied the site. BLAF, together with neighborhood associations, conducted surveys of the households to be displaced allowing for the expropria- tion commission to determine corresponding compensation rights. The vast majority of the households concerned were illegal squatters and did not have any official right to compensation. However, MINUH agreed to grant each displaced household the right to a resettlement plot against pay-Aent of its infrastructure servicing costs. 2.19 An initial 30-ha site, called Dibom II (para. 2.05), was prepared in 1984. All of its 923 plots were allocated, and 865 of them were paid in full with house construction completed or actively underway. A second resettlement area was opened in 1987 in the Ndogpassi III area, where a 279-ha site able to accommodate over 3,000 plots was incorporated into State lands. Plots occupancy and house construction in Ndogpassi III is slower than expected. Among the first 1,032 plots, only 933 had been allocated and 345 paid in full by mid 1988. ARAN commissioned a non-governmental organization (NGO) to analyze problems encountered by resettled households. The major conclusions and recommendations of this study are that the delay for plot payment should be extended from four months to two years and that the standard housing designs prepared by ARAN to accelerate the delivery of building permits led to construction costs of over CFAF 5 million each which was out of reach of the majority of the households. Consequently, the NGO was commissioned, under DDA financing, to design and supervise the construction of core housing units for a cost of CFAF 2 million. These recommendations, which were already included in the Resettlement Action Plan of the Second Urban Project, will be integrated into ARAN's future operations. 2.20 Community facilities in Nylon were covered by the Government's agreement with DDA. This agreement provided for the construction of the Madagascar Market as well as 13 other community facilities in Nylon as originally appraised by the Bank. While construction on the market had begun in 1984 and was completed in early 1989, only one of the other community facilities had been started, a primary school for the Ndogpassi III resettlement area miles away from the nearest existing school. The delay in building community facilities was a consequence of ARAN's decision to carry out all the design, contracting and supervision functions for community facilities in-house and to develop new designs rather than adapt existing ministerial building prototypes. The new designs appear more innovative and functional but took longer to design and construct, and also proved to be more expensive. 2.21 Despite these problems, DDA and Cameroon Governments agreed to additional financing in 1987 to complete the remaining primary school, two health facilities, and eight sociallsports facilities, and to further extend the program to include a community center, three kindergartens, and an artisan center in the Madagascar market; in Ndogpassi III, an additional school, market and dispensary; and in Dibom II, a market and cemetery. ARAN has recently agreed to delegate the design function to local architects. Most community facilities 51 are in the design stage and discussions center around cost reduction measures. A five-year schedule agreed with the DDA in spring 1989 shows construction starting on the Soboum community center in fall 1989 as the Bilongue school is being completed, with nine more facilities starting up during winter 1989/90 and the remaining nine facilities coming on stream over the 1991-94 period. 2.22 Following the creation in spring 1986 of the Agence pour la Restructuration de Yaound4 Nord-Ouest (ARYNO), a HAETUR agency responsible for upgrading operations in Yaound4, preliminary proposals for community facilities in Northwest Yaoundd were first presented in 1987. The area had substantially changed since appraisal as a result of major road works by both central and municipal governments and subsequent private building construction. The sites previously considered for these facilities were no longer available and the proposed location would have required considerable demolition. The costs of the proposed facilities were also much higher than available financing. As noted in para. 2.14, this component was postponed and the uncommitted project funds allocated to the completion of priority works in Nylon. Central and municipal governments are in the meantime proceeding to construct the needed facilities as financing permits. Institution Building 2.23 Five technical assistants and experts were in place from 1984 until 1988, with contract extensions being covered either by the reallocation of project funds agreed in November 1987 or by Government funds. The main achievements of the project's institutional development component, along with an urban sector training component financed under the Second Technical Assistance Project (Credit 1168-CH), are the support provided to ARAN, creation of new units within the MINUH to deal with urban infrastructure maintenance and with training, supervision of the engineering studies for the priority infrastructure project in Douala and preparation of an action program to improve Cameroon's land cadastre. Proiect Cost (Table 1) 2.24 The project costs amount indicated at appraisal was US$54.3 M, or 14.8 billion CFAF at the 1983 exchange rate (US$1 - 270 CFA), of which 51% was the foreign exchange component. The Loan amount of US$20.0 M was to be applied to finance civil works (50%), materials and equipment (60%) and consultants and studies (70?). The Loan was to finance 36Z of total project costs, the Government of Cameroon, 542, and the DDA, 10%. A February 18, 1988 amendment to the Loan Agreement modified the Bank financing share to 85? for works and 100% for consultants. Total project cost at closing (30 June 1988) was estimated at CFA 22.2 billion, or US$74.0 M at the then current rate of exchange (US$1 - 300 CFA). This corresponds to a 35% cost overrun for all project works, including those to be completed with funding from the Second Urban Project. 2.25 Of the US$19.8 M disbursed under the First Urban Project, US$13.5 M went for civil works in Nylon, compared to the US$11.9 M planned, and US$6.0 M was disbursed for institution building, compared to US$5.6 M planned (Table 2). The final allocation of funds to the two main components thus generally reflects the original intent of the project. The main change has been in the share of funding assumed by the Bank, which was to have been 36? but became 72? following the adjustment in financing arrangements made in November 1987 in response to 52 the economic crisis and the provision of additional funds under the Second Urban Project. III. PROJECT RESULTS AND SUSTAINABILITY Impact on the Urban Environment 3.01 Physically, the project has resulted in the construction of a primary road and drainage network over a 600-ha site in Nylon which for the first time fully incorporates this area into the urban fabric of Douala. This infrastructure, along with the completion early in 1989 of a huge modern market, has transformed Nylon from a swamp into one of the city's most dynamic commercial zones. The Nylon market, Marcho Madagascar, is actually more easily accessible to many city residents than is the main Lagos market in downtown Douala. 3.02 Equally important, two new housing sites, Dibom II and Ndogpassi III, have been developed, subdivided, allocated, and occupied. Over two thousand plots have been completed to resettle households. Individual land titles for those Dibom residents whose permanent structures are in place have been approved by the provincial cadastral office, and their actual delivery began in July 1989. 3.03 The most disappointing shortfalls from a physical standpoint are the difficulties in securing the systematic provision of basic urban services and facilitis. Provision of water and electricity networks was delayed because of financial difficulties with the utilities parastatals responsible for these works, limiting their capacity to service their clientele. This in turn has delayed installation of standpipes and street lighting by the municipality. As a result, a temporary water line, financed under the project, had to be installed in 1988 to serve Nylon proper, including Dibom II. Other difficulties result from delays in adjusting the contract with the private contractor in charge of garbage collection to include Nylon in its routes. These matters should be resolved in the near future. 3.04 Following the initial dredging of the Mgoua River in 1982, the water table in Nylon had fallen from a centimeter or two beneath the surface to as much as a meter. Since then, it has been slowly rising as the Mgoua gradually silted again. Dredging is again required; however, appropriate financing mechanisms between central and municipal government have not yet been approved. Until the Mgoua is dredged and the water table lowered, runoff from the new highways in Nylon will adversely affect those living alongside. The lack of adequate dredging would be likely to create adverse conditions for some of the plots in the pilot upgrading zone which would be less well drained than they were before the new highway construction (para. 2.09). Policy Impact 3.05 New mechanisis and procedures have been worked out for resettling households displaced by public works, for carrying out public works improvements, for recovering infrastructure servicing costs, and for regularizing land tenure in existing illegal or irregular settlements (urban upgrading). The development of these mechanisms and procedures extended over much of the project's life because the situations produced by the project had never been addressed before, administratively, legally, or financially. Procedures were developed to 53 streamline land rights assessment and registration within irregular settlements, and to resettle and compensate households in such settlements displaced from infrastructure rights-of-way. Cost recovery procedures were developed and appear to be working well which should permit replicability of upgrading operations. The experience gained by the project executing agency, ARAN, is a valuable asset created through the project and probably the project's most important achievement. 3.06 Less satisfactory were arrangements made for housing credit. Cameroon's housing credit bank, CFC (Credit Foncier du Cameroon) had agreed to loan about CFAF 220 million to 600 households for plot purchase and home improvement in the pilot neighborhood. This provision left out of account the needs of the over 2,000 resettled families who needed financial support to accelerate the house construction process. By fall 1985 only 60 loans had been approved; by spring 1987, this had risen to 121 loans, but the average amounts were much larger than had been envisaged. Although the overall amount loaned by CFC exceeded that agreed at appraisal, the end result is far from that intended. 3.07 A major barrier which was anticipated and provided for at appraisal was the CFC's policy of lending only to salaried workers, which constitute the upper income segments of Nylon. CFC was to work out an arrangement with the Nylon neighborhood savings association (Caisse Populaire de Nylon) to encourage non-salaried workers to start CFC savings accounts in order to qualify for loans. This arrangement proved difficult to conclude because of residents' reluctance and unfamiliarity with formal sector financial institutions. CFC then worked out an agreement to lend to non-salaried household heads provided the Caisse Populaire de Nylon constituted a guarantee fund. 3.08 Another important housing finance issue resolved during the project concerns CFC's insistence over most of the project lifetime that housing construction laans would be conditional on receipt of individual title. CFC in fact suspendad all 121 loans made to salaried wage earners in late 1986 because borrowers had no title. Since receipt of title is conditional on erecting a permanent structure under the law, the CFC condition developed into an impasse for resettlement plot allottees. The project provided the need and incentive for CFC to agree, in the resettlement case, that a plot allocation award certificate and receipt for payment in full of plot costs would constitute sufficient documentation for a housing loan application. 3.09 However, the issue of avalable funds remains. By the time the Caisse Populaire guarantee fund and loan conditionality arrangements had been agreed on, and CFC had agreed to extend its formal agreement with MAETUR and reserve CPAF 150 million a year for three years to resettlement plot allottees, CFC had no more funds to lend. The CFC financial crisis was due both to worsening economic conditions, slowness in identifying other sources of funding than the payroll tax, a generally inappropriate interest rate structure, and other problems of the country's and region's banking and financial system. The interest rate problem and structural problems with the Cameroonian finance system had been a major concern prior to negotiations; they had been resolved at that point by agreeing to undertake a study and reassessment of the financial sector. This was done, but rigidities of regulations established by the Central Bank for States of Central Africa (BEAC) have continued to be a stumbling block. 54 3.10 Another policy problem not adequately anticipated was the municipal- ity's insistence on applying building code standards to new structures going up on the resettlement plots which far exceeded the means of most Nylon residents. The building code issue was resolved in 1989 for both the First and Second Urban Projectss allottees who have paid half the plot charge may build a temporary structure; once they are fully paid up, a permanent structure, required to receive permanent land title, may be erected. Urban Strategy issues and Institution Building 3.11 The First Urban Project proved instrumental in assisting Government in clarifying its strategy for the urban sector, to be implemented as part of the Sixth Development Plan (1986-91). The priority objectives of this revised urban strategy include: (a) the rationalization of public investment in the urban sector; (b) the strengthening of municipalities; (c) the improvement of urban land management; and (d) the strengthening of the role of the private sector in housing and land development. 3.12 The Second Urban Project, initially appraised in February 1986, was revised at Government's request in November 1987 to broaden its geographic scope and include priority works in Yaound6 and other cities and to support the revised urban strategy. In addition to urban infrastructure improvements, this project includes a program to improve urban resources mobilization and the implementation of a policy action plan to consolidate the Government's urban strategy. As a first step in the policy action plan, the Government has created an inter-ministerial Urban Development Steering Committee to coordinate urban development policies and implement the urban strategy. Other actions include the strengthening of urban institutions (MINUH and municipalities) and the rehabilitation of parapublic enterprises in the urban sector. In addition, a Project Special Unit was created by Presidential Decree to be directly responsible for project administration coordination and inform the Steering Committee on the status and progress of the project's components and policy actions. 3.13 Implementation experience under this project has also put project executing agencies and the Bank in a good position to design and undertake efiective rehabilitation of the parastatals in this sector within the larger effort of the Structural Adjustment Loan approved in June 1989. Cost Reco%very and Sustainability 3.14 The cost recovery principles defined at negotiations were carried out as agreed. Primary infractructure and community facilities costs amounting to about 80% of project costs, not normally charged directly to beneficiaries, were borne by Government. The cost of the market is to be charged to users, and standpipes, street lighting, and garbage depots are to be provided at the expense of the city government. The cost of secondary and tertiary infrastructure in the pilot zone and the two resettlement sites is recovered from the residents of the zone in the first case and from plot allottees in the second case; in both cases the cost varis according to location. Commercially zoned property along primary roads is priced at two to four times the actual servicing costs, while property along internal footpaths is priced to sell at less than cost to those with lower incomes. For Dibom II, no distinction was made between plots according to location and the cost of secondary and tertiary infrastructure and 55 of latrine installation is being recovered through plot charges which are equal for all allottees (average cost) 3.15 Cost recovery for Dibom II has gone well since plot allocation in 1984, with allottees 52Z paid up after a year, 68% after two years and 85% after three years. It has gone more slowly for Ndogpassi III where only 27? of the amounts due had been received after a year. This is due to the fact that plot allocation in early 1987 coincided with the economic crisis. Cost recovery is going even more slowly in the pilot upgrading zone, where only about 6Z of the total amounts due were collected in the first year. This is due to the fact that households remaining in the upgrading zone already occupy their plot and are paying primarily in order to receive a title at some time in the future. They had little incentive to pay as long as they could see that those who have already constructed permanent houses in the resettlement areas suffered abnormal delays in receiving their title documents. With the streamlining of land titles delivery currently underway, this situation should improve significantly. 3.16 ARAN has prepared a 10-year program to achieve the complete upgrading of all twelve neighborhoods constituting the whole Nylon area. Based upon the experience of the First Urban Project, new mechanisms were developed to better adjust infrastructure levels of service with actual cost recovery in order to allow for the replicability and sustainability of the program. ARAN has already extended the program of upgrading operations to additional zones not included in the initial project: construction of secondary infrastructure in these zones is to commence following surveying, recovery of costs from households and sale of commercial plots bordering the new roads and is to be self-financing. Preparations in these zones, covering 50 ha each, had reached the following stages in early 1989: design and engineering studies had begun for the second zone, the subdivision plan was being reviewed for the third zone, and site surveying was completed for the fourth zone. ARAN's intention is to complete the secondary infrastructure for the remaining 11 neighborhoods in Nylon by the late 1990s. 3.17 From a financial point of view as well, ARAN appears to have achieved considerable success in its development; construction volume in FY88 exceeded that of all previous years combined and was higher than that of MAETUR's own land development programs, which have declined during the economic crisis. Cost recovery funds from upgrading and resettlement plot charges are estimated in an early 1989 budgeting exercise to cover ARAN's administrative budget and all works in the upgrading zones. All conditions appear to be met to fulfill the objective of the replicability of the upgrading operations. Economic and Poverty Impact 3.18 The ex-post economic rate of return for the upgrading of Nylon has been recalculated based on land values in 1988. It reaches 24Z as compared to an expected 182 in the staff appraisal report. All direct and tangible benefits enumerated in the project economic analysis will have been realized by the completion of project works in mid 1990. Economic rates of return will be recalculated after project works are completed. It is expected that they will far exceed those appearing in the staff appraisal report, since actual increases in the rental value of property in Nylon have outstripped expectations. As income for many households fell during this same period, the current economic crisis has increased the pressure to transfer plots. For this reason, DDA is 56 redirecting the focus of its follow-on efforts to a more coordinated and longer- term effort to organize new income generating activities for Nylon residents. 3.19 At the time of appraisal, about 552 of Nylon area residents were estimated to have incomes below the poverty threshold for Douala, leading to an estimate of about 502 of total costs with direct benefits to the urban poverty group. Because effective economic monitoring of the beneficiary population is currently being mounted with DDA assistance, no information is yet available concerning changes in the resident population due to legal or illegal transfer of land rights. Earlier DDA-supp-rted efforts to set up a monitoring and evaluation system failed because it was too complex. 3.20 Upgrading can leave the intended beneficiaries no better off in terms of housing, and lead to the proliferation rather than the reduction of illegal settlements, if the windfall realized by any resale of land or rights to higher income groups cannot readily be reinvested by these beneficiaries in serviced residential land. The Bank's willingness to support upgrading in this project was based on an assumption that MAETUR would be supplying at least all but the lowest 10 of the market for serviced land during the upgrading project's lifetime, this was indeed one reason why the project as originally appraised included both upgrading and serviced sites programs. MAETUR's programs have been slower starting without external assistance, and they have provided little land affordable below the 50th income percentile. Therefore it is possible that some original beneficiaries have chosen to depart for new uncontrolled settlements. Mitigating against unfavorable economic and land market circumstances are an unusually strong sense of community solidarity observed in Nylon for the last 20 years, and the close and continuing involvement of community representatives in the project implementation process, through liaison effected by two ARAN staff members, one Bank- and one DDA-financed. The issue of redirecting MAETUR's operations towards lower-income groups is being addressed as part of its rehabilitation action plan to be implemented under the Second Urban Project. IV. BANK, BORROWER AND CONSULTANT PERFORMANCE Bank Performance 4.01 The Bank's main strength throughout the project lay in correctly identifying major issues and in pressing for their resolution. This applies to the debate about institutional arrangements for the upgrading agency and about housing finance which preceded negotiations; the technical, procurement and resettlement site issues preceding start-up of civil works, and the land administration issues raised throughout the project. In retrospect, the Bank wanted the project to be executed as conceived; the main case where the Bank position did not in the end prevail, that is, regarding the institutional arrangements for the upgrading agency, the resultant compromise handicapped project execution. The lack of resolution (to this day) of the housing finance issue (para. 3.09), discussed at length before negotiations, remains a handicap to many project beneficiaries unable to qualify for housing credit. 4.02 The Bank's main weaknesses were, at the design stage, failing to simplify the project financial arrangements, underestimating physical contingencies for civil works, and overlooking the need for periodic redredging of the Kgoua River. Given the low-lying, swampy nature of the site and the fact 57 that Douala's annual rainfall is one of the highest in the world, on the order of 3-4 m, the contingency estimates in the appraisal report were overly optimistic. The lack of accurate topographical ant soils data led to an underestimation of total project cost. Following the appraisal of the project, the Bank began to require progressively more advanced stages of design and engineering prior to project appraisal in response to similar problems in many other projects. 4.03 Another approach to coping with technical uncertainty suggested by the Borrower in its Completion Report (see Preface), which would also mitigate the vulnerability of project implementation to currency fluctuations, is to divide project works into discrete subproject phases which could be finsaced sequentially on the basis of available financing during the subperiod. In addition, the Borrower suggested the introduction of the concept of an acceptable exchange rate band; movement outside this band would automatically require some form of formal adjustment in project design and financi-&. Such a proposal is not acceptable considering the Bank's existing lending guidelines. Borrower Performance 4.04 The Government's greatest strength throughout the project's life has been its demonstrated commitment to overall project and sector objectives, and willingness to assume financial responsibility. Government committed about US$4.0 M prior to Loan effectiveness for initial dredging of the Mgoua and preparation of the first resettlement site. Until early 1987, when the economy was in rapid decline, counterpart funding was furnished as required. During the economic crisis, Government arranged with contractors to keep the project going after loan funds were exhausted in 1988, until arrangements could be made for follow-on financing under the Second U,ban Project. Cost recovery from residents of the upgrading zones begun under the project is now continuing as the project execution agency extends upgrading to the other zones of Nylon. Land transfer procedures and the delivery building permits were modified to provide for orderly and fair development of resettlement zones and tenure regularization. 4.05 The Douala municipal government participated financially in the construction of the Madagascar Market, for which it has agreed to pick up a hefty construction cost overrun (around CFAF 3 million). However, the city has not yet provided the garbage collection, street lighting and standpipes for which it is responsible. 4.06 The main shortcoming of Government performance was the slowness of its political and administrative processes, a phenomenon which affects all projects in Cameroon. Two factors can explain this situation: (a) top decision-makera remain uninvolved at preparatory stages and are not prepared to take a timely decision; they must first seek advice from those who have been involved and arrive at a consensus; and (b) responsibility for technical, legal and administrative matters are not sufficiently separated and participants often get involved in areas beyond their expertise. Management deficiencies stemming from these factors also reduced the effectiveness of technical assistance, which was frequently diverted to solve problems not directly related to project matters. 58 4.07 Two areas in particular suffered from the general delay phenomenon: selection of a resettlement site and housing credit. While the particular political sensitivity of land matters may have enhanced the delay factors, it is difficult to believj that the proposal of Ndogpassi I as the second resettlement site could have been made in ignorance of the many encumbrancqs of this site. A suitable site was then proposed (Ndogpassi III), but six months had already been lost. It is also regrettable that new arrangements for housing credit to non-salaried household heads and resolving the CFC requirement for title prior to loan approval were not given higher priority by Cameroonian authorities. 4.08 Apart from general delay factors, Government's position on the institutional arrangements at the outset of the project had a negative impact on project implementation. Government's insistence that the upgrading project execution agency be established within MAETUR rather than as an independent agency may have been reasonable in principle, but the specifics of the relations between ARAN and the MAETUR Yacund4 office had the result of handicapping ARAN's functioning. 4.09 The problem of poor relations between ARAN's Douala office and the HAETUR central administration in Yaoundd was a recurrent theme in supervision reports. A management consultant was asked to study the situation and recommend ways of alleviating it in early 1985. His conclusions were that while some procedures could be simplified, the time-consuming nature of the administrative procedures was irremediably imbedded in the project institutional setup and the requirements of the legal system. It was agreed that additional staff in ARAN's Douala office could help, and they were recruited in late 1985: an additional engineer, three technicians, three topographers, and a secretary. Arrangements were also made to computerize ARAN's accounting system. 4.10 Besides absorbing energy and time, the friction between ARAN and MAETUR impacted negatively on project monitoring and audit. ARAN would forward its quarterly reports to MINUH which would hold them. It may be that MINUH's intention was to consolidate them with its own reports on technical assistance and training and forwarding them to the Bank; however, this was never done. The solution eventually adopted was for each agency to forward its reports directly to the Bank. In a similar vein, the First Project audit was rejected by the Bank because it covered only ARAN's accounts and contained no consolidated project accounts. The accounts for 1985/86 were successfully audited in June 1987. 4.11 Taking a longer view, it is also regrettable that at negotiations, the Government chose to drop the sites and services component and that no alternative financing was found. As a result, MAETUR programs were slow to get underway and did not follow the policy of responding to all segments of demand, including households below the 50th income percentile. The economic crisis has brought MAETUR programs virtually to a standstill. As a result, most of the population, including Nylon residents, with or without resettlement plots who desire or need to set up housekeeping elsewhere, have little choice but to squat illegally as long as there are no low-cost plots coming on the marke.: for sale, in a variety of urban locations. 59 Proiect Relationship 4.12 The necessity to follow Bank procurement guidelines was more than just a recurrent theme in Bank supervision reports and a constant source of friction between the Bank and Borrower teams. Both teams poured so much energy into resolving this problem that little was left for other important aspects of project management and implementation. As a result, non-civil works aspects of the project were neglected, the housing credit issue dragged on, coordination with DDA languished. On the surface, the friction over procurement could be interpreted as due to poor understanding of procurement procedures. The Bank repeatedly had to remind the project agency to send TOR and full contract documentation for its review and to specify selection criteria in TOR. Some kind of local assistance in procurement could be helpful in this case, as will the drafting of standard contract clauses being carried out under the Second Urban Project. Relationship with Swiss Aid 4.13 Following the preparation stages of the project, the Bank and DDA efforts in the Nylon area were implemented virtually independently. The Swiss Government wrote the Bank in summer 1986 regretting the infrequence of contact and proposed regular meetings; they also requested that the Bank inform them on a routine basis of the timing of Bank missions. In its own Project Completion Report (para. 4.03), the Government expresses its regret and surprise at the Bank's apparent lack of interest in the housing credit, employment generation and community development aspects of the project. Both this report and the DDA- financed report on project replicability stressed the necessity of community participation in upgrading operations, at least one of whose objectives is to improve living conditions for the resident population. In particular, the Government praised the financial and technical flexibility of DDA in comparison to the Bank, which allowed it to adapt more readily to changing needs. 4.14 In April 1988, the Bank took up the suggestion made earlier by DDA. It announced its upcoming mission to finalize investments required to complete the First Urban Project and invited DDA to join the mission in Cameroon. DDA replied immediately, agreeing to synchronize missions and outlining their primary concernst secondary and tertiary roads and drainage, housing finance and income generation; housing improvement, house construction assistance, support to artisans and small enterprises, and waste collection. From this point on, close coordination has been maintained. Consultants' Performance 4.15 Consultants' performance was generally good with the exception of the Nylon civil works design and engineering. Of particularly good quality was the technical documentation for Douala roads and drainage improvements, financed under the First Urban Project, in preparation of the Second Urban Project. This information will prove essential for the implementation of infrastructure improvements over the entire city. 4.16 Concerning the civil works design and engineering for the Nylon civil works, Bank supervision teams devoted considerable effort to require that drain lining technologies appropriate to the environment be considered, among other shortcomings in the drainage design. It was not until the civil works contractor began work that errors in the quantity survey for roads design were discovered. 60 In the Borrower's view, the underestimate of project costs at appraisal encouraged the design engineering consultant to underestimate quantities at the quantity survey stage. It is also true that technical standards were being substantially revised upwards during the design stage, due to the need to make the project road and drainage network compatible with the newly issued Master Plan, and this created a difficult situation for all parties. V. LESSONS LEARNED 3.01 This was an innovative project, and one which was ambitious in seeking to remedy the deep-rooted and longstanding problems presented by the Nylon area. It is not surprising that there were unusual resistances and obstacles to be overcome, even without vnfavorable economic circumstances. It is also not surprising that the project was so rapidly overtaken by events during the numerous delays it suffered, given the inherently dynamic nature of urban milieux generally and that of Douala, one of West Africa's major port cities, in particular. What is surprising is that the project succeeded as well as it did, and paved the way for the Second Urban Project, which is tackling directly structural areas that had resisted earlier attempts to address: land management and urban institutional environment. Benefits of Cofinancing 5.02 The project demonstrates one effective means of integrating "hard* and 'soft' elements with the same project, that is, for the Bank to focus its efforts primarily on the first, while cofinanciers better adapted to working closer to the ground handle the second. In this project, staff within a single project unit managed both Bank supported infrastructure investments and grant cofinanced activities oriented to human and community development. 5.03 This arrangement proved particularly useful as a way of providing an umbrella for community and NGO participation. House construction in the resettlement sites benefitted from low-cost designs and construction methods worked out with the help of the French NGO Architectes Sans Frontibres. House renovation in the pilot upgrading zone has been receiving assistance from a group organized by the University of Geneva's School of Architecture. Artisan training involved two local NGOs. Pit latrine construction was carried out by small local contractors, who also carried out about half of the works to construct the Nylon market. As mentioned above, DDA found that creating work for small contractors and artisans does not form a sound foundation for a longer term effort to improve economic opportunities in such a community. They are currently developing a strategy which takes a broader view of the zone's economic development within a citywide context. Payoff from Project/Sector/Country Work Coordination 5.04 The project history demonstrates the kind of payoff which can come from close association of sector and project work, though this means some hard decisions by Bank missions on the inevitable tradeoffs between supervision, preparation and sector work. The aide-memoire of the July 1988 urban mission is a good example, covering a wide range of issues concernir, the First and Second Projects, infrastructure in Yaound4 and secondary ci.ies, municipal assistance, and urban sector action plan and strategy. In the sector strategy 61 itself, the redefinition of responsibilities for infrastructure as between central and local government grew directly out of the experience with the First Urban Project. Central government financing should be limited to primary infrastructure, while city governments and private developers should be given more scope for action and their responsibilities broadened to include financing and execution of secondary and tertiary infrastructure. 5.05 The sector strategy also reflects project experience in defining other strategic objectives -- opening up of the land and housing finance markets -- and emphasizes the need for strong cointry economic support in order to resolve stubborn structural problems such as those blocking housing credit. It will be very important to future operations in this sector that the housing credit agency (CFC) along with the land development agency (MAETUR) and other institutions involved in housing and industrial development have partiLipated in the current parapublic rehabilitation program being carried out with Bank support. 5.06 The experience gained by ARAN through project implementation will allow for the continuation of a program of replicable urban upgrading operatior.s. In addition, this project proved instrumental in assisting Government in the clarification of its strategy for the urban sector which will be consolidated with the support provided by the Second Urban Project. Issues to be Resolved under Future Projects 5.07 As obvious as it seems, it is important to restate a verity amply demonstrated by this projects a first project in a new sector is not likely to be smooth-sailing. On the other hand, the payoff can be greater than expected: the years-long working out of new detailed procedures for compensation and transfer of title, applicable to resettlement of residents of rights of way or of illegally settled urban areas, will mean that the Bank's Second Urban Project, among other projects, should be spared such delays. 5.08 The Second Urban Project will also profit from lessons learned in the First Project about infrastructure cost recovery and housing construct'on on resettlement plots. While most households had little difficulty coming up with the roughly US$1,000 equivalent to pay for basic infrastructure costs of resettlement plots, unrealistically high building standards imposed by the city government and a drying up of housing credit meant that, until recently, those allocated resettlemeat plots were hindered in their efforts to set up new permanent homes. This situation was resolved by agreeing with the city government that an exception to formal city building standards should be made for resettled families. 5.09 Conflicts regarding the application of Bank procurement guidelines consumed many valuable weeks of Bank staff time that could have more usefully been devoted to other aspects of project implementation. To remedy the shortcomings of tendering practices, Government has created a new body which should be fully in operation in early 1990. As part of the Second Urban Project, efforts are now underway to draft standard administrative clauses for works and equipment contracts which can accelerate approval by Cameroon and the Bank. Improved dialogue on procurement will be essential if Bank missions are to properly supervise important institution-building aspects of the Second Urban Project. 62 5.10 Better coordination is required with water and electricity parastatals to persuade them to install their services in parallel with road and drainage works. Provision should be made for financing periodic redredging of the Mgoua River, as well as for garbage collection by the city of Douala. The Urban Development Interministerial Committee created in the framework of the Second Urban Project will ensure improved coordination between the various urban agencies. 5.11 More appropriate design standards should be selected in designing future community facilities. This would have made it easier for local contractors to build, speed up construction and lower costs. 5.12 A lesson learned, but still unimplemented, involves identifying better arrangements for housing credit (paras. 3.06 and 4.07) to unsalaried household heads not eligible for loans from the government parastatal. The arrangement would involve setting up a guarantee fund by the Nylon neighborhood savings organization, for which the Swiss are prepared to provide backing. Unfortunately, the parastatal's current lack of funds has rendered this agreement moot. 63 Table 1 REPUBLIC OF CAMEROON URBAN DEVELOPMENT PROJECT, LOAN 2244-CM PROJECT COMPLETION REPORT Prolect Cost and Financing (US$ M) * ----- Appraisal L! --------- ---------------Actual Lb *------ Project Component Total Cameroon IBRD Swies-DDA Total Cameroon IBRD SwIes-ODA Other NYLON -- Preparatory Works 2.6 2.6 -- - 2.5 2.5 - - -- - Resettlement Areas 4.8 4.8 -- - 8.5 8.5 -- -- Roads and Drainage 28.8 11.9 11.9 -- 32.6 0.0 18.5 - 18.1 ** Market and Comm. Fec. 7.0 1.6 - 5.5 11.0 5.2 - 5.8 - YAOUNDE NORTH-WEST -- Market and Comm. Fec. 8.5 1.8 1.7 - 8.5 8.5 -- - - INSTITUTIONAL DEVELOPMENT - Support to ARAN 4.3 2.2 2.1 -- 6.2 4.5 1.3 2.4 -- - Support to MINU 6.8 8.5 3.3 - 16.9 6.4 4.5 -- - -- Other 1.5 0.8 0.7 - 1.6 1.8 0.2 * -- FRONT END FEE 9.8 - 6.8 0.8 - 6.3 Total 643 28.8 0. . 74.0 2.9 19..1. /a USS a 270 CFAF (Appraisal Estimate) b US$ 817 CFAF (Project Average) Is To be financed under the Second Urban Project AFlIN October 8, 1989 A:T1.pcr (S) 64 Table 2 REPUBLIC OF CAMEROON URBAN DEVELOPMENT PROJECT, LOAN 2244-CM PROJECT COMPLETION REPORT Loan Disbursement by Expenditure Category (USS1,0M) --Original Financing Plan-- ---Revised Financing Plan--- ------ Total-------- from 65/03f4 to 92/18(88 from 62/118/88 to 91/25/89 from 0S/98/84 to 01/26/89 Expenditure Category IBRD IBRD Gov't. IBR) IBRD Gov'$t. IBRD Gov't. Share Diab. Share Total Share D1s. Share Total Dieb. Share Total 1A. Works Nylon BO 4,851 4,851 8,702 85 9,111 1,068 19,719 18,462 5,959 19,421 2A. Equip. Nylon 601 184 89 228 61 7 5 12 141 04 285 28. Equip. Yaound6 691 47 81 78 60 - - - 47 81 78 $A. Consult. Nylon 701 611 282 878 1ow 79w -- 706 1,311 262 1,578 3C. Consult. CAPME 791 8 8 11 101 - - - 8 8 11 SD. Consult. MINUH 70% 8,190 1,867 4,557 100m 1,879 - 1,879 4,69 1,867 5,988 FRONT END FEE - 296 -- 298 298 -- 298 Total S.S87 S18 47 1.9 1.S 2.810 AF1IN October 8, 1989 A:T2.pcr (SS) 65 Figure 1 REPUBLIQUE OF CAMEROON URBAN DEVELOPMENT PROJECT, LOAN 2244-CM PROJECT COMPLETION REPORT Map of Nylon and Ndogpassi r5= I . . . ' . . . A-- At* %* . . *"*aw* kQe 4xe