Morocco - Urban Development Projects

Document of The World Bank FOR OFFICIAL USE ONLY / f~)I'.Report No. 9 7 29 PROJECT PERFORMANCE AUDIT REPORT KINGDOM OF MOROCCO RABAT AND SECOND URBAN DEVELOPMENT PROJECTS (LOANS 1528-MOR AND 1944-MOR) JUNE 28, 1991 Operations Evaluation Department This document has a restricte 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. GLOSSARY OF ACRONYMS ANHI National Agency for Struggle against Insanitary Housing (of MOH) l"Agence Nationale de Lutte contre l'Habitat Insalubre] BCP Popular Central Bank (HQ of BPs) ["Banque Centrale Populaire"} BP (Regional) Popular Bank ("Banque Populaire"] CIH Real Estate and Hotel Credit Bank ("Cr6dit Immobilier et H8telier"] ERAC Regional Planning and Construction Agency ["Etablissement R4gional d'Am6nagement et de Construction"] 01 Government of Morocco HBM Low/middle cost housing ("Habitat & Bon Marchd"] IMF International Monetary Fund MHAT Ministry of Housing and Urban Planning (latterly MOH) l"Minist6re de l'Habitat et de l'Am6nagement du Territoire"] MOF Ministry of Finance ("Minist&re des Finances"] MOH Ministry of Housing (formerly MHAT) l"Minist6re de 1'Habitat"] PCR-R Project Completion Report of Rabat Urban Development Project PCR-II Proj. ct Completion Report of Second Urban Development Project SAR-R Staff Appraisal Report of Rabat Urban Development Project SAR-II Staff Appraisal Report of Second Urban Development Project UNDP United Nations Development Programme UNESCO United Nations Educational, Scientific and Cultural Organization USAID United States Agency for International Development FOR O0FCIAL USE ONLY THE WORLD BANK Washington, D.C. 20433 U.S.A. Oice of oirector-Cewal Opfaftms Evautidkn June 28, 1991 MEMORANDI21 TO THE --XECUTIVE DIRECTORS AND THE PRESIDENT SUBJECTs Project Performance Audit Report on the Kingdom of Morocco Rabat and Second Urban Development Projects (Loans 1528-HOR and 1944-MOR) Attached, for information, is a copy of a report entitled "Project Performance Audit Report on the Kingdom of Morocco - Rabat and Second Urban Development Projects (Loans 1528-MOR and 1944-NOR)" prepared by the Operations Evaluation Department. Attachment This document has a restricted distribution and may be used by recipients only in the performance of their ofcial duties. Its -ritents may not otherwise be disclosed without World Bank authorization. FOR OFFICIAL USE ONLY PROJECT PERFORMANCE AUDIT REPORT KINGDOM OF MOROCCO RABAT AND SECOND URBAN DEVELOPMENT PROJECTS (LOANS 1528-MOR AND 1944-MOR) TABLE OF CONTENTS Page No. Preface ............ . .. ......... .................... .. .... .. i Basic Data Sheets .... .... . ....... ......................... .. . 1 Evaluation Summary ... ..... ........................... xi I. SECTOR BACKGROUND ...................................... 1 A. 4ecroeconomic Performance .............. ............ 1 B. Moroeco's Urbanization ............................. ..... 2 C. Urban Sector Policy ......... ........ . 3 II. THE PROJECTS ..................... . .. . ... . .. .... ... 4 A. Issues in Project Preparation ................... 4 B. Project Objectives ......................... 8 C. Project Description ...........................o 9 III. PROJECT IMPLEMENTATION AND ACHIEVEMENTS ...................... 11 A. Factors Contributing to Execution Delays .................. 11 B. Financial Performance ................................... 13 C. Non-Use of Building Materials Loans ....................... 15 D. Project Physical Achievements ............................ 15 E. Target Beneficiaries and Poverty Impact ................... 16 F. Employment Creation ....................................... 17 G. Limited Role of Municipalities ............................ 18 H. Sector Policy Impact ...................................... 18 I. Project Environmental Impact ............................. 18 IV. POINTS OF SPECIAL INTEREST ................... 19 A. Major Contributions to Urban Physical Development ......... 19 B. Projects' Policy Impact ................................... 20 C. Renewal versus Upgrading .................................. 22 D. Divorce of Physical and Financial Aspects of Project Management ..................................... 23 E. Management of World Bank Loan in Morocco .................. 24 F. Limited Role of Municipalities in all Cities .............. 25 G. Small Preparation Effort Needed for Both Projects ......... 26 V. ROLE OF THE BANK ............................................. 27 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.) VI. MAJOR LESSONS TO BE LEARNED .................................. 27 A. Need to Integrate Physical and Financial Aspects of Projects ....................................... 27 B. Urban Projects can be Prepared Quickly .................... 28 C. Land Titling as a Pro-requisite for Cost Recovery ......... 28 D. The Need for Substantial Household Savings Project Success ................................... 28 E. Project Impact is Local .................................. 28 i PROJECT PERFORMANCE AUDIT REPORT KINGDOM OF MOROCCO RABAT AND SECOND URBAN DEVELOPMENT PROJECTS (LOAdS 1528-MOR AND 1944-MOR) PREFACE 1. This document is a Project Performance Audit report (PPAR) on the Rabat and Second Urban Development Projects in Morocco. The Rabat Urban Development Project was approved by the Board on June 14, 1977, to be partly financed by Loan 1528-MOR in the amount of US$18 0 million which was only 43.9% disbursed. Various amounts to the total of US$10.1 million were cancelled. The Second Urban Development Project was approved by the Board on January 22, 1981, to be partly financed by Loan 1944-MOR for US$36.0 million which was only 27.2% disbursed. Various amounts to the total of US$26.2 million were cancelled. 2. The PPAR consists of an Evaluation Summary and a Project Performance Audit Memorandum (PPAM) prepared by the Operations Evaluation Department (OD). Project (;mpletion Reports (PCRs) for the Rabat Project prepared under the old guidelines by the former Urban and Regional Development Projects Division of the Europe, Middle East and North Africa Region (EMENA) and for the Second Project by the Infrastructure Operations Division of 1ountry Department II (EM2IN) of the same Regional Office were previously sent to the Board as OED reports No. 6184 dated May 12, 1986 and No. 8687 dated May 23, 1990, respectively. Given that the two Bank PCRs were prepared using different formats, for comparative purposes it was decided to include Basic Data Sheets for both operations. This means that some information for the Second Urban Development Project is repeated from Part III of the respective PCR. 3. The PCRs provide accounts of the project experiences and indicate their successes and failures. In order to give a wider and more recent perspective, OED has attempted to provide additional information about the projects, their history, their eventual outcomes, the issues they raised and problems encountered during implementation. The PPAN is based upon the Bank's PCRs, the PCRs prepared by the Borrower's Ministry of Housing (MOH), the Staff Appraisal Reports (SARs), project files, legal files, transcripts of the Executive Directors' discussions at Board Presentation and other relevant materials. It also drew from interviews conducted with Bank operational staff previously involved with the projects and with Moroccan Government and local officials in Rabat, Meknis and Kenitra, project cities of the two operations. 4. The Audit agrees with the PCRs in most respects and looks more closely at the key issues raised. 5. The draft PPAR was sent to the Borrower for comments; however, none were received. iii PROJECT PERFORMANCE AUT REPORT KINGDOM OF MOROCCO RABAT URBAN DEVELOPMENT PROJECT (LOAN 1528-MOR) BASIC DATA SHEET KEY PROJECT DATA Appraisal Expectation Actual Total Project Cost (US$ million) 37.6 25.6 cost underrun (%) -32.0 (DH million) 165.5 177.8 cost overrun (%) 7.4 Loan Amount (US$ million) 18.0 18.0 Disbursed 18.0 7.9 Cancelled 0.0 10.1 Repaid 4.74 Outstanding 3.19 Date of Physical Completion 03/81 12/85 Time Overrun (%) 75.0 Economic Rate of Return (2) 20% 20.0% CUMULATIVE ESTIMATED AND ACTUAL DISBURSEMENTS (US$ million) 1978 1979 1980 1981 1982 1983 1984 1985 Appraisal estimate 0.8 5.6 13.5 17.9 18.0 Actual 0.0 0.0 1.2 2.1 2.8 5.4 7.1 7.9 Actual as 2 of estimate 0.0 0.0 8.9 11.7 15.6 30.0 39.4 43.9 iv PROJECT_DATOS A2raisal Estimate Revised Aagal First Mention in Files 07/72 Pre-Appraisal Mission 07/74 Appraisal Mission Departures: - Casablanca (aborted) 11/76 - Rabat 06/77 Negotiations 04/77 10/77 01/18 Board Approval 06/77 12/77 02/78 Loan Agreement Date 03/78 Effectiveness Date 06/78 11/78 Closing Date 03/82 03/83 12/84 OTHER PROJECT DATA ------------------------------------------------------------------------ Borrower Kingdom of Morocco Executing Agencies - Ministry of Housing and Urban Development (MON) - Ministry of Social Affairs & Handicrafts (MASA) - Ministry of National Education (MEN) - Ministry of Youth and Sports (MJS) - Ministry of Public Health (MSP) - Ministry of Labor and Vocational Training (MTFP) - Popular Central Bank (BCP) - Electricity and Water Company (RED) - Municipality of Rabat Fiscal Year of Borrower January 1 - December 31 Follow-on Project Second Urban Development Project (Loan 1994-MOR) EXCHANGE RATES Currency Unit: Dirham (DH) Annual Averages 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 3.9 4.3 6.0 7.1 8.8 10.1 9.1 8.4 8.2 8.5 Sourcess (1980-1982) official rates reported in: World Bank African Economic and Financial Data (1983-1989) period average market rates reported in International Financial Statistics V MISSION DATA NO. OF NO. OF STAFF D= OF MO/YEAR PERSONS WEEKS WEEKS EP Identification I 01/72 4 4 8 02/72 Identification II 07/72 1 2 2 07.72 Preparation I 06/74 1 2 2 07/74 Preparation II 11/74 3 2 6 01/75 Preparation III 07/75 2 1 2 07/75 Preparation IV 10/75 3 1 3 11/75 Preparation V 01/76 3 2 5 02/76 Preparation VI 05/76 5 3 15 07/76 Preparation VII 07/76 1 2 2 08/76 Preparation VIII 1/ 11/76 9 3 27 12/76 Sub-total 72 Preparation IX 02/77 1 1 1 03/77 Preparation X 03/77 1 1 1 04/77 Appraital 06/77 8 2.5 20 07/77 Spevsi-total 22 Supervision I 04/78 3 3 9 06/78 Supervision II 10/78 3 1 3 10/78 Supervision III 05/79 6 3 18 07/79 Supervision IV 10/79 4 3 12 11/79 Supervision V 04/80 3 1 3 06/80 Supervision VI 11/80 3 2 6 12/80 Supervision VII 04/81 1 1 1 05/81 Supervision VIII 10/81 2 2 4 11/81 Supervision IX 02/82 1 2 2 03/82 Supervision X 06/82 1 1 1 06/82 Supervision XI 10/82 2 1 2 11/82 Supervision XII 03/83 1 1 1 10/83 Supervision XIII 09/83 2 1 2 04/83 Supervision XIV 02/84 2 1 2 03/84 Supervision XV 10/84 1 1 1 10/84 Completion 02/85 2 1 2 05/86 Sub-total 69 I/ Aborted appraisal mission for Casablanca Urban Development Project vii PROJECT PERFORMANCE AUDIT REPORT KINGDOM OF MORDCCO SECOND URBAN DEVELOPMENT PROJECT (LOAN 1944-MOR) BASIC DATA SHEET KEY PROJECT DATA (See also Part III of Bank PCR) Appraisal Estimate Actual Total Project Cost (US$ million) 81.1 26.4 cost underrun (%) -67.4 (DH million) 308.0 231.7 cost underrun (2) -24.8 Loan Amount (US$ million) 36.0 Disbursed 36.0 9.8 Cancelled 0.0 26.2 Repaid 6.48 Outstanding 3.31 Date of Physical Completion 01/86 01/88 Time Overrun (2) 75.0 Economic Rate of Return (M) 182 13% CUMULATIVE ESTIMATED AND ACTUAL DISBURSEMENTS (US$ million) 1982 1983 1984 1985 1986 1987 1988 1989 Appraisal estimate 2.8 6.9 19.9 30.5 36.0 Actual 0.0 1.8 2.1 5.6 7.2 9.0 9.4 9.8 Actual as 2 of estimate 0.0 26.1 10.6 18.4 20.0 25.0 26.1 27.2 viii PROJECT DATES Appraisal Estimate Revised Actual In.tial Project Brief 03/79 Appraisal Mission Departure 04/80 Negotiations 10/80 Board Approval 01/1 Loan Agreement Date 04/81 Effectiveness Date 08/81 Closing Date 07/86 07/87 07/88 Project Completic,% 01/86 07/88 OTHER PROJECT DATA Borrower Kingdom of Morocco Executing Agencies - Ministry of Housing and Urban Development (MOH) - Ministry of National Education (MEN) - Ministry of Youth and Sports (MJS) - Ministry of Public Health (MSP) - Popular Central Bank (BCP) - Municipality of Meknbs - Municipality of Kenitra Fiscal Year of Borrower January 1 - December 31 Follow-on Projects Pilot Project for Communal Infrastracture Fund (Loan 2272-MOR) Housing Finance Project (Loan 2245-MOR) EXCHANGE RATES Currency Units Dirham (DH) Annual Averaies 1980 1981 1982 1983 12&4 1 1986 1987 1988 1989 3.9 4.3 6.0 7.1 8.8 10.1 9.1 8.4 8.2 8.5 Sourcess (1980-1982) official rates reported ins World Bank African Economic and Financial Data (1983-1989) period average market rates reported in International Financial Statistics ix MISSION DATA NO. OF NO. OF STA"F DATE OF NOMYEAR PERSONS WEEKS WEEKS REPORT Identification Preparation Appraisall/ 04!80 6 20 12/80 Sub-total 20 Supervision I 02/81 1 2 2 03/80 Supervision II 12/81 3 3 9 05/81 Supervision III 03/82 1 2 2 08/81 Supervision IV 06/82 1 1.5 1.5 10/81 Supervision V 11/82 2 1.5 3 06/82 Supervision VI 03/84 2 0.5 1 10/82 Supervision VII 10/84 1 1 1 03/83 SupervisirL VIII 03/85 1 1 1 08/83 Supervision IX 03/85 1 1 1 12/83 Supervision X 03/86 1 1 1 03/84 Supervision XI 07/86 1 1 1 08/84 Supervision XII 03/87 1 1 1 12/84 Completion 06/89 1 1 1 12/85 Sub-total 25.5 1/ PCR provides no breakdown of the missions up to appraisal. Identification and preparation missions were fielded together with supervision of the Rabat Project. There was one pro-appraisal and one appraisal misecon for the Second Project xi PROJECT PERFORMANCE AUDIT EPORT KINGDOM OF MOROCCO RABAT AND SECOND URBAN DEVELOPMENT PROJECTS (LOANS 1528-MOR AND 1944-MOR) EVALUATION SUM4ARY Introduction 1. The two operations bidonvilles in the majority of these audited here, which were executed at urban areas. Regarded as blemishes Rabat, the capital city of Morocco, on the urban fabric by the highest Meknas and Kenitra, together authorities and as a vestige of constitute the Bank's lending underdevelopment, GOM's traditional experience with physical urban policy had been to eradicate development projects in Morocco. bidonvilles and replace them with Follow-on projecto had been high standard, highly subsidized envisioned for otiter cities, but housing for which cost recovery was were not prepared due to GOM's largely impossible given unenthusiastic support for the affordability limits. bidonville (squatter settlement) upgrading approach espoused by the 4. The Bank's approach of Bank and the shift in the latter's upgrading, rather than eliminating, interest to the financial aspects of squatter settlements was not Morocco's urban sector. welcomed by GOM. Policy making for urban development was complex, 2. Implementotion took involving numerous sectoral place during a period of rapidly ministries, as well as the powerful changing economic fortunes for regional authority of the Governors Morocco. When the Rabat Project was who were personally appointed by the prepared, GOM's finances were King. Bank missions with little buoyant, but by com,letion public previous experience in the urban sector fiscal constraints due in sector or the country had difficulty p..rt to IMF and Bank-supported mapping out sectoral policy and adjustment programs had laid to rest clearly knowing what GOM's position hopes of continuing massive was at any given moment. The sector subsidized public investments in policy dialogue between the shelter. An export drive to Government and the Bank, moreover, stabilize the country's balance of was a protracted one, lasting five payments resulted in sharp years before appraisal of the first devaluations of the dirham and, operation in Rabat. In retrospect, consequently, a large unused surplus this was an important, but costly, of loan funds for both projects. schooling for the Bank in the ways of doing business in the urban 3. Morocco has a large sector in Morocco. urban system with major cities in various parts of the country. Rapid 5. Preparation of what growth fuelled by rural migration became the Rabat Project was a led to the formation of large traumatic experience for the Bank. xii Given the dimensions of Morocco's representations to that effect. In urban sector and the commensurate response, GOM communicated its size of its urban problems, Bank continuing support for the project management early on had hoped to to the Bank. With this green light, finance a very large project. preparation wae expedited and the Although unwelcome by management, Rabat Project was appraised eight reports of GOM's limited months after the debacle in implementation capabilities were Casablanca. Preparation of the regularly filed by field missions, Second Project, by contrast, was implying that a large operation was uneventful. In fact, it was not feasible. Nonetheless, the prepared very quickly and with a Bank's initial approach was to minimum of Bank staff deployment. prepare an urban development project in Casablanca, the nation's largest Obiectives and Project Description city and principal industrial and commercial center. 8. For a small operation, the objectives of the Rabat Project 6. Seven missions and 72 were very ambitious. They included staff weeks of mission time were the introduction of low-cost urban expended on preparation of this upgrading, infrastructure and operation which had to be abandoned housing concepts that were to during the appraisal mission when ensure affordability and permit the main project site's soil improved cost recovery. The conditions were found to be operation was to initiate country- inadequate for the proposed wide housing programs for low-income development. Surface rock, which groups and was intended as a had apparently gone unnoticed demonstration project to show how to despite a large feasibility study, tackle the most difficult of prevented the low-cost installation Morocco's urban problems. The of basic services. At this eleventh Second Project, as a follow-on hour, the Bank was forced to abandon operation, was to further the same the operation and the Casablanca aims. Urban Development Project ceased to be part of the lenling program. 9. Both operations were basically squatter upgrading and 7. Faced with the risk of sites and services projects. having no project at all, the Bank Bidonvilles in selected was fortunate when the Governor of neighborhoods were identified for Rabat reiterated GOM's interest in improvement. Upgrading involved preparing a smaller upgrading providing basic infrastructure and project in the capital city. Even re-housing displaced households this preparation, however, did not within the sites and services go smoothly. During the pre- schemes. Building materials loans appraisal mission, the Minister of were to have been provided to help Housing announced that the finance dwelling unit extensions and Government was no longer willing to improvements and freehold land title support the project since was to have been granted to project maintaining a slum in the nation's beneficiaries. In the Rabat capital was considered unacceptable. Project, ancillary components At this point, the Bank contemplated included employment creation withdrawing from the urban sector in programs and community facilities. Morocco altogether an,* made strong The main innovation in the design of xiii the Second Project was the inclusion amounts that were not needed. of a municipal services component Reimbursement requests were tardy which involved the construction of and the complex flow of funds maintenance depots, provision of resulted in considerable delays in equipment and technical assistance loan proceeds reaching contractors and the elaboration of studies for and suppliers. In spite of GOM's follow-on operations. financial difficulties, however, substantial counterpart funding was Implementation Exnerience provided for the projects. Another positive outcome was the significant 10. In contrast with its mobilization of private savings to preparation, execution of the Rabat pay for the expansion and Project was relatively improvement of dwelling units at straightforward. Compared with the project sites. projected three-year implementation period there were delays, but these 13. Despite initial plans, were relatively minor when compared building materials loans were not with the average performance of used in the projects. This was projects in the region. Those partly because the aforementioned delays that did occur in this and mobilization of private savings led the Second Project were caused by to a reduction in demand for these inadequate attention having been loans. On the supply side, the paid to land acquisition and titling Popular Central Bank (BCP), which problems. Some land still awaits was to have granted ths loans, was titling in Mekn&s ten years after unwilling to do so for a number of initial acquisition procedures reasons. In particular, it felt started. that the loans would be unsecured since beneficiaries did not possess 11. Implementation would land titles and risks were have been more rapid had the considered high. The 1% institutional arrangements for the administration fee permitted under Rabat Project not been so complex. the project, moreover, was seen as The newly created Ministry of insufficient to compensate the cost Housing was charged with overall of managing a portfolio consisting responsibility for project execution of numerous very small lending even though it was inadequately operations. staffed and there was no central project budget for it to control. Results This state of affairs, which encouraged conflict among 14. The physical participating agencies each of which achievements of the projects were had independent authority over its significant, benefitting appreciable component, was corrected under the proportions of the populations of Second Project when MOH's regional the participating cities: Rabat delegations were put in command. (5.6%); Meknbs (10.1%) and Kenitra (12.3%). In addition, except in 12. In both operations, Kenitra where the recent growth of there was a divorce between the slums has been the most rapid, the physical and financial aspects of operations contributed to a project management. Rapid substantial decrease in the share of devaluation of the dirham led only the total population living in belatedly to cancellation of loan bidonvilles. xiv 15. Most beneficiaries were neighborhoods and upgraded their low-income, although some higher living conditions more generally. income households have demonstrated their willingness to live on the 17. The projects' expected improved sites. Adequate ex-post policy impact was given high profile monitoring, however, needs to be during appraisal. Specific aims carried out in order to shed were to encourage the Moroccan greater light on a number of factors Government to favor bidonville including the reported practices of upgrading over eradication beneficiaries' sharing lots with throughout the country and to other households and lot re-sale by introduce less costly standards for the original allottees. On balance, housing programs in order to make the results of the employment them affordable to the poor and generation components were permit cost recovery. Although the disappointing. The keen interest respective PCRs conclude that the demonstrated by the Bank during operations have had a significant preparation appears to have policy impact, the Audit's review of evaporated during implementation. recent Government policy statements GOM's own lack of enthusiasm for and actions suggests that the this component helped to seal its changes have been much less fate. substantive than desired by the Bank since bidonville eradication remains 16. Of particular interest firmly on the agenda. Even though are the projects' contributions to the projects demonstrated an the physical environments of their approach that was less costly than respective cities. In Rabat, where the traditional urban renewal the main works were completed in solutions employed by GOM, they were 1985, a consolidated low-income not used as a blueprint for the rest residential area now stands in the of the nation's cities. The studies place of a former bidonville. At for additional projects, which were the Douar Doum site, the original to have been financed under the plot layout has been preserved and Second Project, for example, were attention to architectural details not carried out. has helped to produce scenes reminiscent of the ancient medinas 18. The creation of the of Moroccan cities. Although not as National Agency to Combat Unsanitary attractive, Douar Maadid has Houcing (ANHI - Agence Nationale de successfully drawn commercial Lutte contre 1'Habitat Insalubre) in activity to the point that it is now 1984 and its current program to "re- the fourth largest retail center in absorb" bidonvilles in Moroccan Rabat. Being more recent, the cities (through renewal) indicates project sites in Mekn4s and Kenitra the Government's continued desire to are not as complete, with many lots be rid of this urban phenomenon. In still awaiting full development. addition, a comprehensive census of Nevertheless, present indications bidonvilles was undertaken by MOH in are that they too will become 1989, among other reasons in order integrated areas in their respective to help detail a program for their cities. Beneficiary households elimination. More generally, it is interviewed by the Audit mission important not to underestimate the confirmed that project investments antipathy towards bidonvilles in brought improvements to their Morocco, even at the highest level. xv The current National Development Bank loan. Because of successive Plan (1988-92) includes the devaluations of the dirham, large US eradication of bidonvilles as one of dollar surpluses built up in both the principal actions proposed in loan accounts. By the time the the housing field. The Bank itself loans were closed, less than one- appears to have become less third of the total amount had been interested in this question with the used. GOM paid heavily in recent shift to housing finance as commitment fees as a result of the the main thrust of urban lending in slow cancellation of unneeded loan Morocco. resources. This occurred because those responsible for management of 19. Even though the Moroccan the loan were not fully briefed with Government formally agreed to respect to prospects for using up upgrading in the Second Project, the funds. Local managers had no what occurred in Meknbs was really direct concern with this since they urban renewal. Existing shacks were were not responsible for the demolished (as a condition of financial consequences of project residents continuing in the program) execution. In addition, requests and plot boundaries on the site were for loan reimbursements were made redrawn. This action, which was very slowly and the complicated flow recognized by local managers as of funds resulted in a lower Bank violating the spirit of the original share of total cost than foreseen at project, was not in compliance with appraisal. the Loan Agreement. Nonetheless, existing rights-of-way were felt to 22. Although there is be inadequate and a more regular presently much emphasis on municipal site geometry was considered development in Bank urban lending, desirable in and of itself. The this was not the case when the approach followed, however, was projects were undertaken. While a disruptive since it involved special municipal services component additional involuntary displacement, was included in the Second Project, increased costs and delayed project both projects were essentially execution. conceived as operations to be conducted by central government 20. Another point of special agencies. As a result, the role of interest is the aforementioned the municipalities was very limited. divorce between management of the physical and financial aspects of Sustainability the projects. One of the main victims of this divorce was cost 23. Although the projects recovery which was regarded by made important contributions to the project managers as just one more physical development of their problem to be dealt with and one respective cities, much greater that did not adversely affect short- attention should have been paid to run physical implementation. Aa a cost recovery and local maintenance result, insufficient attention was to ensure their sustainability. given to the necessary institutional Project implementation did not and administrative arrangements for generate a cash return to GOM, which cost recovery. has thus had to subsidize the operations. Local tax efforts have 21. This shortcoming also been insufficient to ensure that harmed financial management of the municipalities have the resources xvi necessary for proper maintenance of implementation. Both the upgraded areas. As a result, operations suffered from maintenance services are presently the divorce between the lacking, especially with respect to two. Increasing the streets, drains and solid waste emphasis on financial disposal. In the Saknia district of management is essential, Kenitra, for example, loose sandy but by itself not soil threatens to block the on-site sufficient. The drainage system altogether. Unless management of physical these issues are adequately dealt execution must be with, there is a risk that important improved at the same benefits achieved to date will be time, lest the divorce lost as project sites revert to continue with only a their pre-upgrading condition. change of perspective. (iii) Land titling is a findings and Lessons complex process in Morocco, but is an 24. Among the main findings essential pre-requisite and lessons of the Audit are the for cost recovery to following: proceed. It is necessary not only for (i) When the objective is banks, but for any clear and support is agency responsible for forthcoming from local cost recovery, to have a authorities, it is firm guarantee that the possible to prepare amounts due will be urban development paid. As in most other projects quickly. countries, the main Preparation of each of guarantee is property the projects, excluding which can only be the aborted Casablanca mortgaged if a potential operation, required some borrower Las full legal six months through to rights in the form of a appraisal. Most freehold title. If not, problems that arose possible sanctions, such during implementation, as foreclosure, have no could be dealt with by firm legal base and are supervision. The thus ineffective. The experience under these Region confirmed that operations accordingly land titling problems calls for a shift of persisted with the Bank staff resources follow-on project (Ln. from project preparation 2245-MOR), but that to supervision. CIH's special agreements with private developers (ii) There is a need to helped overcome some of better integrate the bottlenecks. management of the physical and financial (iv) In all three project aspects of project cities, most home preparation and improvements were xvii financed by private participate in financing savings. These were their own housing. especially important in view of the legal (v) Since they were not impediments to mortgage replicated elsewhere in lending due to the Morocco and due to the absence of land titles. difficulty of Romarkably, even ascertaining thb wider households whose income international influence remains low today were of these operations, it able to mobilize must be concluded that considerable savings. their impact has been In Morocco, personal largely restricted to savings in the form of the project cities jewelry, livestock and themselves. While there rural property appear to is potential for a play a significant role broader demonstratlon in the informal economy. effect, this will only In addition to be achieved if project traditional results are affordability analysis, systematically future operations should monitored, evaluated and review patterns of disseminated. The Bank accumulated wealth, as can play an important well as current income, role in the broader in order to determine dissemination of project the ability of low- experiences that it income families to finances. 1 PROJECT PERFORMANCE AUDIT REPORT KINGDOM OF MOROCCO RABAT AND SECOND URBAN DEVELOPMENT PROJECTS (LOANS 1528-HOR AND 1944-OR) I. SECTOR BACKGROUND 1.01 Particular interest in reviewing the Rabat Urban Development Project - Loan 1528-MOR (the "Rabat Project") and the Second Urban Development Project - Loan 1944-MOR (the "Second Project") together derives from their representing the entire Bank experience of financing physical urban projects in Morocco. During implementation of the Second Project, GOM and Bank priorities shifted towards sector lending, beginning with the Housing Finance Project (Loan 2245-MOR) with the Real Estate and Hotel Credit Bank (ClI). No further traditional style urban projects were prepared, even though it had been intended early on to develop similar operations in all major Moroccan cities. However, the experience of these two projects focussed Bank attention on the urban sector in Morocco for a period of some seventeen years (1972-1989) during which many changes occurred both in the national economy and within the sector itself. A. Macroeconomic Performance 1.02 The period from the first identification mission to completion of the Rabat Project (1972-1986) was one of bewildering change in Morocco's economic fortunes. As far as macroeconomic prospects were concerned, preparation began in 1974 on a euphoric note as the price of Morocco's main export, phosphates, quintupled in the wake of the 1973 oil shock. So favorable were the terms of trade that GOM upwardly revised the targets of the 1973-77 national development plan to provide for greater pblic capital expenditure, particularly for housing. Morocco's external debt at that time was a mere US$1.0 billion and the country enjoyed a sound balance of payments surplus. Precisely when the cost-cutting approach to urban development espoused by the Bank seemed least urgent to a government temporarily awash with money, GOM and the Bank began what proved to be protracted discussions to prepare the first urban development operations in the country. The fiscal bonanza, however, fed Bank hopes that a very large operation could be undertaken. 1.03 Morocco's economic performance worsened with the second oil shock of 1979 which was not as easily absorbed as the first since international phosphate prices did not react favorably. With a deteriorating balance of payments, GOM sought the assistance of the IMF, which approved an Extended Fund Facility for US$1.0 billion in 1980 in exchange for a tight stabilization program that included sharp cutbacks in government expenditures for housing and urban infrastructure. Like a number of other developing countries at the time, Morocco attempted to sustain high economic growth through external borrowing, a strategy which led to the rapid expansion of foreign debt--currently some US$18.8 billion- -and a serious balance of payments and foreign exchange crisis in March 1983. By that time, when the first two urban projects were being implemented, the enchantment with large physical urban development projects had been lost and the Bank shifted its support to loans for housing finance and municipal infrastructure. 2 1.04 Part of Morocco's macroeconomic stabilisation and adjustment involved an export drive. For this reason, there were drastic devaluations of the national currency, the dirham, during the 1980s. By the time the Rabat Project was closed in 1986, the average exchange rate to the US dollar was 9.1s1, representing a 571 decrease since 1980. This change was to have important repercussions for project financing since far fewer dollars were required to fund the predominantly local costs of the operations. How this was managed and GOM's slowness to cancel loan funds that were not used is treated as a point of special interest later in this report (paras. 4.23-4.26). 1.05 Thus, the Rabat Project, at its inception, was felt to be merely the first stage of what was expected to become a large public housing and urban development program that would eventually cover the entire country. Given the buoyant economy in 1977 when the operation was appraised, this seemed a feasible prospect since the subsequent drastic cutbacks in public investments could not have been foreseen at that time. 1.06 Even though the difficulties faced by the Moroccan economy had become more evident by the time the Second Project was appraised in 1980, this operation was couched in much the same terms as its predecessor. The development of a program for other cities, for whose studies the project provided f-nds, was still very much a priority at this stage. As the second project was executed, however, the setbacks became more severe. In 1980, Morocco's GNP per capita was US$879; by 1988, the most recent estimate, it had fallen to US$830. The debt service ratio of the country's overseas borrowing, in turn, had risen from 38.6% to 89.8% of GNP over the same period, strongly increasing external constraints on the economy. During this period, successive Standby Agreements were reached with the IMF and Bank loans for structural adjustment were provided. Among other conditionalities, stabilization measures included restraints on public spending, a factor which was to put hopes for a series of major urban development projects to rest once and for all. B. Morocco's Urbanization 1.07 Rapid urban growth provided the backdrop to the proposed sector operations in Morocco. The country has a well-developed urban system with important settlements spread throughout the country. The largest cities with their estimated 1990 populations are: Casablanca (3.2 million), Rabat-Sal6 (1.2 million), Fez (0.8 million), Marrakech (0.8 million) and Meknbs (0.7 million).' Today, more than 47% of the country's 24 million inhabitants live in urban areas as compared with only 35% when the Rabat Project was being prepared. With urban growth fuelled by migration from rural areas, Morocco's cities have had to cope with increasing pressures on their housing stocks. Accordingly, it is not surprising that the Government's main concern was with housing demands arising from this growth. Authorities at the highest level were especially concerned about the development of what are referred to in Morocco as "bidonvilles," unauthorized squatter settlements of precariously built shacks without basic Data for metropolitan areas and urban agglomerations from United Nations Centre for Human Settlements, Global Report on Human Settlements 1986, Oxford University Press, 1987. 3 *ervices. These settlements were particularly unwelcome in the central part of the country where the principal industrial and commercial center of Casablanca and the capital city of Rabat-Sal6 are located.2 The Government's initial aim was to eradicate these bidonvilles which weri considered to be a stain on the fabric of Moroccan cities, conveying a negative image of poverty and ugliness that GOM wished to remove. At worst, they were a focus of social unrest and disturbance. While the Moroccan Government wanted to eliminate these vestiges of underdevelopment altogether, the Bank argued that they should instead be upgraded so as to integrate them at relatively low-cost into modern urban society. This basic difference of views nourished a prolonged policy dialogue, whose outcome reflects less of a shift to the Bank's position than the respective PCRs suggest (see parae. 4.05-4.13 below). C. Urban Sector Policy 1.08 During the lengthy identification and preparation of what became the Rabat Project (1972-77), GOM remained consistently skeptical of the benefits of the low-cost standards and squatter upgrading approach recommended by the Bank. The lack of continuity of interlocutors on both sides, however, did not encourage clear communications for the policy dialogue. Thus, the actual degree of GOM commitment to abandoning what the Bank considered wasteful and costly practices of bidonville eradication was not always fully understood by the Bank. In retrospect, it appears as if there was much soul searching both within the Government and the Bank, with wide differences of opinion persisting over what the best policy would be. The resurgent interest of the central government and local authc.ities in bidonville eradication and the continued pursuit of high standard solutions for low-income housing made the policy dialogue and project preparation difficult for all concerned. The Government's propensity for urban renewal created particular problems for the Bank which contemplated suspending sector dialogue on three different occasions. 1.09 Responsibility for the formulation of urban policy in Morocco today is shared between the Ministry of Housing (MOB) and the Ministry of Interior (MOI) due to the latter's responsibility for liaison with the increasingly powerful municipalities. Even early on, however, it was a major challenge for Bank missions to understand where sector policy originated. This was especially complicated by the overlay of the authority of regional Governors on the actions of sectoral ministries within their jurisdictions. Governors, who are personally appointed by the King, wield considerable -ower on his behalf. As a result, the Governors' opinions frequently prevail over those of the local representatives of central ministries. The contribution of the Governor of Rabat, for example, was crucial to bringing the preparation of the Rabat Project to a successful conclusion (para. 2.11). 1.10 Also confusing for the Bank in its early sector dialogue was the occasion when the Minister of Housing took a position against the proposed Rabat Project that vae later construed by the Moroccan Government as not being official 2 Rabat and Sald are twin cities whose historic centers face one another across the River Bou Regreg. The functional urban area includes both centers, but the administrative capital is Rabat where project interventions were made. 4 policy (para. 2.14). In addition, senior officials worked out what were effectively provisional policy positions in informal discussion with Bank missions. Only much later would these be formalized as official policy or rejected by the competent political authority. In short, GOM's policy-making process was very complex and difficult for outsiders to understand. As a result, early Bank missions having little experience in the urban sector in general and in Morocco in particular had a difficult task mapping aut the Government's policy position at any one point in time. 1.11 In retrospect, the five year sector dialogue can be regarded as az important (bet costly) schooling of the Bank in the ways of conducting businoss in Morocco. The learning was put to good use during preparation of the Second Project when those policy issues which remained to be resolved were approached in a more pragmatic way. A change of Bank miosion staffing introduced a fresh approach to the issues that the Moroccan Government and the Bank had been wrestling with for nearly a decade. Bank missions recognized the Government's continuing desire to eradicate bidonvilles, but also understood that the ever tightening public sector fiscal situation would probably take care of the problem by itself. Rather than continuing to argue a matter of principle, lack of resources simply would not permit the Government to proceed with its ambitious plans to free all Moroccan cities of bidonvilles. II. THE PROJECTS A. Issues in Proiect Preparation 2.01 Preparation of what became the Rabat Project was a traumatic and costly experience for the Bank. From the first identification mission, more than five years elapsed before the operation was appraised, a protracted process with few parallels elsewhere in the sector. Given Morocco's comparatively lerge urban sector and comensurate problems, Bank management had hoped to finance a large project. The figure of US$500 million was mentioned by staff, but in the end the Bank had to content itself with a loan of US$18 million (of which only US$7.9 million was disbursed). GOM's limited implementation capability at that time was cited by Bank missions as the principal constraint on the size of the operation. 2.02 The initial project was meant to be located in Casablanca, but the rocky soil conditions of the main project site forced an eleventh hour change of venue to Rabat, where a scaled-down operation was agreed. Despite the considerable resources and effort expended, the consultant's report on the feasibility of the Casablanca operation proved unacceptable to the Bank, having overlooked, among other things, a soil survey of the proposed site. Even with the decision to scale down the operation and move it from Casablanca to Rabat, the project's tribulations were not over. During the pre-appraisal mission in Rabat, the Minister of Housing announced that the Government was no longer interested in pursuing an upgrading program, as the continuing existence of a bidonville in the nation's capital was an eyesore that would not be tolerated. Only after strong representations by top Bank management, did GOM take a position favorable to continuing project preparation through to appraisal. 2.03 Although the Audit agrees with the PCR that project preparation was too long (PCR-R Highlights para.(a)), both its own and the SAR's explanation of 5 how the operation came about are incomplete (PCR-R para. 2.04, and SAR-R para. 1.06).' Since there are many lessons to be learned from this experience, it is useful to give a fuller account here. Preparation of what came to be known as the Rabat Project, in short, went through two main phases: the protracted and unsuccessful effort to prepare an operation in Casablanca and the quick, but successful, work to prepare the Rabat Project itself. Preparation of the Second Project was straightforward and is dealt with further below. 1. The Aborted Casablanca Urban Development Project 2.04 Identification missions fielded as early as 1972 correctly concluded that housing vas the priority area for GOM and they soon set the tone for the ensuing dialogue. Moroccan housing standards were felt to be too high and costly. In a confrontational style that was to trouble the future dialogue, as early as 1974 the Bank contemplated suspending its work in the urban sector unless the Government agreed to lower standards in its housing projects.' Although technically well-founded from the standpoint of cost recovery and sustainability, the Bank's view was not communicated in a way that gained the sympathy of Moroccan officials. 2.05 After the initial hiatus, once the Government indicated that it would be willing to consider minimum cost solutions preparation missions began working on the Casablanca Urban Development Project. Casablanca was chosen because it was Morocco's largest city where the Bank hoped to finance a large operation as the first urban project in North Africa. In retrospect, this was an overambitious choice for the Bank, which up to that time possessed no urban sector experience in the country. Local authorities in metropolitan Casablanca, in turn, had their own ideas about urban development which included a strong pre- disposition towards bidonville eradication. This made the Bank/GOM dialogue even more complicated. Nevertheless, the Bank decided to pursue preparation of the operation. Up to this point, there was no mention of Rabat or any other city in Morocco. All preparation efforts were concentrated on Casablanca. 2.06 The first of seven preparation missions was fielded to Casablanca in 1974. By their own account, members of the Bank's small urban staff at that time 3 The SAR-R justifies the choice of Rabat, for example, by saying that "there would have been a considerable risk of time consuming preparation if another urban area had been chosen." (para. 1.06). It makes no mention of the preparation effort for the aborted Casablanca Urban Development Project which consumed 85% of staff mission resources for what later became the Rabat Project. The PCR-R makes only a passing reference to the intended operation in Casablanca (para. 2.03). A reader might thus be mistakenly left with the impression that Casablanca was only a minor component of the Rabat Project, whereas, in fact, four out of the five years of identification/preparation time were devoted exclusively to it. 4 This was the first of three such occasions prior to the appraisal of the Rabat Project. The other two were when the Casablanca operation was aborted in 1976 and when the Government temporarily retracted its intention to go ahead with the project during the pre-appraisal mission. 6 recognized that they had little experience in the sector and in North Africa in particular. According to one staff member, "there was much groping in the dark" and "mistakes were bound to be made". If there was a lack of experience, there was certainly no shortage of dedication and enthusiasm as mission teams were keen to set up a pioneering operation in Morocco. In addition, they felt challenged by management to prepare the first large Bank-financed urban operation in the region. Staff recall the experience as having been very difficult, but intellectually stimulating. 2.07 Despite the problems and uncertainties, the Bank was encouraged to continue working in Casablanca by the UNDP's willingness to finance a large feasibility study to serve as the basis for the project. Progress, according to preparation missi in reports, was very uneven. Some praised the "imaginative" work of the consultants, while others reported all kinds of difficulties including staffing problems. It is not clear, however, whether these reports reflect uneven performance of the preparation work or the varied opinions of the many Bank staff involved during this period. 2.08 In fact, discontinuity of Bank mission staff was itself a problem. Interlocuto, in Morocco recall receiving conflicting signals from different staff members about the appropriate design of an urban development project for Bank financing. As the planned appraisal .f the Casablanca Urban Development Project approached, Bank missions in the field felt very much left to their own devices. Contacts with Washington, when management guidance on project development might have been needed, were limited. With efficient international telecommunications, contacts between missions and headquarters are now much easier and more frequent than they were during this period. At that time, however, mission staff felt under considerable pressure to deliver a completed package to management. Bank staff involved in the project confirmed that presenting alternatives and lists of potential problems in Washington did not conform to the management style of the period. Mission reports relating the limited exeruting capabilities of GOM agencies were particularly unwelcome since they dampened hopes of approving a very large project. 2.09 The full appraisal mission of nine staff and consultants for the Casablanca Urban Development Project was fielded in November 1976. Instead of successfully culminating an effort of 72 staff weeks of missions to Morocco with an agreed project, the mission nearly ended with the Bank having no operation whatsoever. During a visit to the 180 ha site that was expected to accommodate 2,500 low-income sites and services plots, the mission discovered that so.i conditions were inappropriate for the proposed development. Surface rock would have made laying infrastructurea very costly and, thus, unaffordable to the targeted beneficiaries. In spits of all the preparatory work, this discovery came as a surprise to the Bank. 2.10 Faced with this unforeseen problem and a local authority that was, at best, lukewarm to the upgrading approach supported by the Bank, the appraisal mission was aborted. From that moment, the Casablanca Urban Development Project 7 ceased to exist.3 The immediate reaction to this debacle in the Bank was the desire to suspend all work in the urban sector in Morocco and, thus, to jettison an investment of 72 staff weeks of mission time. Under the circumstances, the saving grace of a much smaller operation in Rabat appeared just in time. 2. The Rabat Proiect 2.11 The Governor of Rabat had already expressed interest in participating in an urban sector operation, but encountered little enthusiasm on the part of the Bank which felt that consultant resources would be over-stretched to prepare a component in Rabat too. However, after the impasse in Casablanca, the Rabat proposal took on a new importance. The Bank was indeed fortunate in having an influential and enthusiastic interlocutor in the Governor of Rabat who was also sympathetic to the low-cost upgrading approach. 2.12 In stark contrast to the effort in Casablanca, preparation in Rabat was much quicker. This was due to a number of factors. The projec. was smaller and simpler. The Bank needed an operation as consolation for the resources spent upon Casablanca. As a result, Bank staff became actively involved with GOM personnel in preparing detailed project designs and documents. In addition, the proposal counted on the active support of the Governor. Even so, preparation was not without its problems. 2.13 During the pre-appraisal mission, the Bank was informed by the Minister of Housing that the Government had shifted its position regarding the project in Rabat. Informally, the mission was told that the King did not want an operation that purposefully retained a slum in the nation's capital. According to the Minister, the proposed project site of Douar Doum was "too valuable from a locational point of view to justify the continuing existence of bidonvilles". Part of the site is indeed a privileged location with a panoramic view over the Bou Regreg River valley. 2.14 Faced with what appeared to be a second major setback in less than five months, the Bank made strong suggestions that it would cancel its work in the urban sector (and more) unless previous agreements to develop the Rabat Project were honored. At the time, there was much introspection within the Bank about the advisability of continuinj :o work in the sector. Continuity was assured, however, after the Minister of Economic Affairs reaffirmed GOM's favorable position towards the project. The Minister of Housing's earlier statement was characterized as a personal view, not reflecting that of the Government. With such diverse opinions within the Government and the Bank, it is not surprising that disagreements should have arisen from time to time. As one interlocutor of the Audit mission remarked, "while we in the Government didn't entirely agree with them, Bank missions forced us to re-think our approach and to get more organized." 2.15 With resolution of this problem, project appraisal could quickly proceed. Altogether, only three missions involving 22 staff weeks over a seven S Today, a public sector middle-income housing scheme stands on the proposed project site. 8 month period were necessary to prepare and appraise the project. Speed was possible thanks to the considerable efforts made by both GOM and Bank staff. With a concentrated effort and a clear objective, which in this case included making up for time lost in Casablanca, project preparation can be greatly shortened and made less costly. The comparatively short time needed for preparation is one of the key lessons of the Rabat experience (para. 6.04). 3. The Second Project 2.16 After the earlier traumas, elaboration of the Second Project was a straightforward affair, requiring only 20 staff weeks of Bank missions. It was decided to proceed wich preparation of the second operation, however, before any implementation results of the Rabat Project were known. Preparation of the Second Project, in fact, began only five months after the loan for the Rabat Project had been made effective. After all the previous difficulties, Rabat and the Bank were both anxious to extend operations to other parts of Morocco. A change of staff on both sides also allowed a fresh, less combative approach to GOM/Bank relations. 2.17 At this point, in short, both the Moroccan Government and the Bank wished to develop similar operations in other cities. Preparation was expeditiously undertaken by combining it with supervision of the earlier loan. As a follow-on operation, preparation of the Second Project benefitted from the learning experie -ce of the Rabat Project. B. Project Objectives 2.18 Although the SAR-R recognized that the Rabat Project was only the first result of the Bank's sector policy dialoguz with the Government, its overall objective was very ambitious for such a small operation, namely: "to introduce low-cost urban up8rading, infrastructure and housing concepts which would ensurt affordability and incorporate improved cost recover,. This will enable GOM to initiate country-wide prt.rams to more effectively tackle already serious urhan housing problems and to meet social services and in rastructure requirements of urban low-income groups." t3AR-R para. 3.01) The operation was expected to be a demonstration project that would display "solutions to some of the most difficult problems in urban development," among which "technical concepts of and administrative arrangements" for upgrading low- income settlements were specifically mentioned (SAR-R para. 3.01) 2.19 At its inception, the Second Project was explicitly recognized as a follow-on operation to the Rabat Project. Thus, at appraisal, its objectives were to continue to support the same policies, but with greater emphasis on affordability through lower cost structures and the strengthening of municipal management (SAR-II para. 2.01). 9 2.20 Nearly three years separated the two appraisal missions. Perhape for this reason and the fact that no individual Bank staff member or consultant participated in both appraisals, there is a distinct change of focus between the first and second operations. In the Rabat Project, greater attention was given to social issues such as the rights of women. Ta the Second Project, in turn, there was more emphasis on physical design. Particular reference was made, for example, to the need to respect architectural and urban design traditions. Under the second operation, greater attention was also given to the need to strengthen municipal management capabilities in the two project cities. C. Project Description 2.21 In spite of the short time required for its preparation, the Rabat Project became a complex operation involving many small interventions, not all of which were fully implemented. According to the Loan Agreement (Schedule 2), the project was to have consisted of the followings Mi) Squatter Uparading: integrated upgrading in Douar Don, Maadid and Hajja, requiring the relocation of 900 households; upgrading would include infrastructure extension and rehabilitation, expansion of commercial facilities and self- help home improvement with the provision of construction materials and equipment. (ii) Sites and Services: plots would be developed on about 12 ha in an area known as "La Butte" and provided with contractor-built core units. (iii) Employment Generation: in oider to create 2,500 jobs: (i) about 11 hectares of land and credit would be provided for the construction of buildings, equipment purchase and working capital; (ii) technical assistance in marketing, production and management would be provided to the Economic Activities Unit; (iii) a vocational center with the capacity to train about 360 people annually would be established; and (iv) an educational center to provide functional literacy programs and skills to young women would be set up; this center would accommodate about 250 trainees, be combined with day care facilities for about 400 children and provide child care orientation for mothers. (iv) Community Service Equipment: vehicles would be provided to supplement the existing municipal fleet and better meet basic requirements for garbage disposal, fire fighting and ambulance services. (v) Technical Assistance: specific programs would be carried out tos (i) assist the Sous-Habitat Upgrading Unit of the Ministry of Housing and Urban Planning (MHAT) in the financial and technical analysis of urban upgrading projects (32 man- months); (ii) guide the Ministry of Finance (MOF) in preparing sector reviews and recommendations for housing finance reforms 10 (8 man-months)I (iii) provide general guidance to the Economic Activities Unit (30 man-months); (iv) develop vocational training capabilities (20 man-montha); and (v) guide the Municipality of Rabat in using and maintaining municipal equipment (20 man-months). 2.22 It is interesting to compare the above description with that of the Second Project. The latter gives physical details of the shelter components that were absent from the Rabat legal documents. In comparison, less attention is given to employment generation, while an important new component for the improvement of municipal services was added. According to the Loan Agreement (Schedule 2), the Second Project was to consist of the followingt (i) Slum Upirading: in Meknbs, the Bordj Moulay Omar, Ain Chebik and Sidi Baba squatter settlements, and, in Kenitra, the Saknia, Bouchtyne and Fath slums were to be improved. Upgrading was to include: (i) the definition of plot boundaries in accordance with existing settlement patterns, whith the land to be acquired by the Borrower and transferred to its occupants with an option to purchase after five yeares (ii) off and on-site vehicular roads would be built or improved; (iii) off-site sewer facilities would be extended and on-site sewer networks would be installed; (iv) the water supply system would be improved and expanded; (v) an electricity network would be built and street and pedestrian path lighting provided; (vi) in-fill plots with core units would optimize infrastructure use and accommodate displaced residents; (vii) building materials loans would be made available for home improvement and/or expansion; (viii) primary schools and a dispensary would be built or expanded; and, (ix) buildings for project administration and beneficiary assistance would be constructed. (ii) Sites and Services: projects would be executed on about 46 ha at Sidi Baba and Ain Chebik in Mekn&s and on about 26 ha at Saknia in Kenitra; the infrastructure, community facilities and building materials loans to be provided would be simil.r to those in the upgrading areas and five types of on-plot development would be undertaken: serviced plot only; serviced plot with perimeter walls; serviced plot with sanitary core; serviced plot with sanitary core plus one room; and, serviced plot with sanitary core, one room and a foundation for horizontal expansion. (iii) Eimlovment Generation: in Meknis, serviced plots would be developed for business activities with basic off and on-site street, water, sewer and electricity networks on about 2.5 ha in Bordj Moulay Omar and Sidi Baba; in Kenitra, 6 ha adjacent to the Saknia settlement would be subdivided and provided with similar infrastructure for business use. 11 (iv) Imrovement of Municipal Services: the Municipal Works Departments (MWDs) of Mekn&As and Kenitra would be reorganized into five functional unites administration, studies, civil works, workshops and maintenance; municipal depots, that would centralize all MWD activities, would be built and equipped; garbage collection and street and sever maintenance services would be improved; studies for improving the organization of the Accounting and Tax Departments would be carried out in Mekn6s and a sewerage master plan would be prepared for Kenitra. (v) Preparation of New Urban Development Proiects: studies for additional urban development projects elsewhere in the country would be undertaken. 2.23 In summary, the central feature of both operations was shelter upgrading, a fact which is brought out more clearly in the description of the Second Project. An important change between the two projects, however, which reflects shifting Bank priorities more generally, was the move away from employment generation toward municipal development. Regarding the choice of cities, finally, the almost accidental selection of Rabat has already been discussed above (paras. 2.11-2.12). As concerns the other sites, GOM appears to have informally allocated major cities among different donors. Thus, USAID was to work in Casablanca and Tanger, UNESCO in Fez and the Bank in Mekn6s and Kenitra.' The Bank was happy to concur with this choice since the latter cities fulfilled two essential conditions: they both had substantial bidonvilles and their local authorities were comparatively sympathetic to the low-income shelter approach advocated by the Bank. III. PROJECT IMPLEMENTATION AND ACHIEVEMENTS A. Factors Contributing to Execution Delays 3.01 Given the setbacks of the aborted Casablanca operation, COM's wavering support for the project in Rabat and the limited time in which the Rabat Project was prepared, the latter's implementation was remarkably trouble-free. There were delays compared with the planned schedule, but this was partly because of the unduly optimistic three-year execution schedule conceived at appraisal (SAR-R para. 5.01). The Audit agrees with the PCR-R that this was unrealistic for a first project of its kind (PCR-R para. 3.06). Judged not by that aim, but rather by the performance of similar projects elsewhere, the five and a half year implementation period was reasonable. 3.02 Implementation could have been faster, however, had more adequate attention been paid to a number of important issues. Foremost among these was the extremely complex question of land ownership and titling. The superposition of traditional and modern systems of land titling and the protracted processes of land expropriation and transfer in Morocco can leave ownership rights unclear 6 Local authorities in Marrakech were said to be uninterested in participating in this kind of program. 12 for many years.' For this first experience in the country, the Bank was satisfied with Government assurances that final acquisition would be expeditious, including expropriation if necessary (SAR-R para 3.05). The SAR was even led to conclude that "the probability of land acquisition problems is relatively low" (SAR-R para. 8.02). As events turned out, the land was only fully acquired toward project completion. While construction could proceed before final acquisition, freehold titles could not be given to beneficiaries as intended (SAR-R para. 3.06) because plot boundaries could not be delineated due to the lack of overall title. Moreover, GOM lacked the necessary surveyors to undertake this work in the difficult conditions of an irregular plot layout. The lack of individual title did not preclude beneficiaries from remaining on the site, but it did exclude them from access to formal financing since the property could not be mortgaged to secure loans. 3.03 Similar land problems were experienced in Meknbs under the Second Project, where one site has been awaiting final acquisition for more than ten years while disputed claims are investigated. Again, without formal title, a plot beneficiary cannot enter into formal financial arrangements with lending agencies. Banks generally and the Popular Central Bank (BCP) in particular were unwilling to lend without the mortgage guarantees that a secure land title could offer. Without a final title, moreover, there would be no cost recovery, as the projects were to demonstrate. An important lesson of the project, accordingly, was the adverse impact of beneficiaries' lack of land titles on cost recovery (paras. 6.05 - 6.06). The Region confirmed that these difficulties persisted and made some eligible programs under the follow-on project (Ln. 2245-MOR) impossible to finance. Through special agreements with private developers, however, CIH was able to remove some land titling bottlenecks. 3.04 During supervision of the Rabat Project and early supervision of the Second Project, Bank missions were primarily concerned with physical implementation given that these were the first operations in all three cities. Since physical execution could proceed without a final settlement of the land title issue, the latter received less attention than more pressing short-term problems such as the relocation of beneficiary households. The lessons of the experience are clear. In Morocco, it is important to concentrate project development on land already owned by local authorities or the central. government. This was the case with some of the project land in Kenitra, where individual land titling is most advanced at present. 3.05 Another aspect of these operations, especially the Rabat Project, which acted as a brake on their implementation was the complex institutional arrangements for their execution. In Rabat, there were nine executing agencies (seven ministries, BCP and the Municipality of Rabat), each with autonomy over its own budgetary allocation for the project. Had there been no overlapping responsibilities, execution would have been smoother. But the potential for disputes between sectoral ministries over many matters, such as the technical specifications of primary schools for example, was unlimited. Project start-up, moreover, occurred in the same year that MOH (then MHAT) was created. A newly ' For a detailed discussion of these issues, see PADCO, Morocco Urban Land Policy Study, Washington D.C. July 1987. 13 established ministry simply did not have the authority or experience to bring effective coordination to bear upon so many traditional areas of responsibility. Thanks in part to the first operation, MOH's experience (and authority) grew considerably to the point where its regional delegates were given responsibility for all implementation natters under the Second Project in Mekn&s and Kenitra. Although it had to deal with six executing agencies, MOH's clear coordination powers simplified decision making and sped up project implementation in the latter cities. 3.06 Another of MOH's problems in its first year of existence was in adequate staffing which had been explicitly recognized as a risk for the Rabat Project (SAR-R para. 8.02). Project start-up in Rabat was delayed as MOH struggled to put togeth3r its management team. This was not a surprise, but, in retrospect, there were few alternatives open to the Bank at the time. One of the few constant themes reported by successive preparation missions was the limited execution capabilities of many agencies, not just MON (SAR-R para. 4.01). When project management was facing difficulties, the Governor of Rabat was able to smooth decision making for the operation. It was recognized that the Rabat Project would provide on-the-job training for MOH (SAR-R para. 4.10). The operation succeeded in doing this not only for MOH, but also for the Bank itself. This was reflected in the more effective staffing of the Second Project. By that time, MOH's authority and experience had become more consolidated. B. Financial Performance 3.07 As far as cost performance was concerned, there was only a 7.4% cost overrun in dirhams at completion for the Rabat Project. In the Second Project, there was a 24.8% underrun in dirhams due mainly to the construction materials loans which did not materialize. However, there were significant cost underruns in US dollar terms for both operations because of successive devaluations of the Moroccan currency during the implementation period, a point that is further discussed below. 3.08 The emphasis placed on the supervision of physical implementation was appropriate for the Rabat Project as the first operation of its kind in the country. Local project units and Bank missions were staffed accordingly. The same emphasis continued during supervision of the Second Project. The entire period, however, -as one of major economic change and financial difficulties for the Moroccan Government during which public sector financing was cut under Standby Agreements with the IMF. The difficult economic and financial situation was unexpected and did not capture the priority attention of the local managers of either project in spite of counterpart funding problems, an unstable foreign exchange rate and growing external constraints which had direct repercussions on the performance of the operations. A divorce, thus, arose between the technical and financial aspects of project management, a subject that is dealt with as a point of special interest later in this report (paras. 4.19 - 4.22). 3.09 Substantial devaluation of the dirham occurred during the implementation of both projects. In 1980, only 3.9 dirhams were needed to buy one US dollar; by 1985, the figure had risen to 10.1 dirhams. This meant that available Bank loan funds far exceeded what was needed to pay the local costs of the operations. To avoid unnecessary commitment fees, the Borrower would 14 normally request the Bank to cancel the amounts that would not be used. Although only 44% of the Rabat project loan and 27% of the Second Project loan were eventually disbursed, requests for cancellations were tardy. This occurred because those responsible for project execution had little knowledge of the overall financial situation of the loans, while those responsible for loan management were unaware of future project implementation prospects. This resulted in the costly administration of the Bank loans, another point of special interest (paras. 4.23 - 4.26). 3.10 The devaluations may have induced a feeling among project managers that Bank resources had become unlimited and costless. Surprisingly, in view of GOM's tight fiscal constraints, the way the projects' financial aspects were managed did not even reflect the growing scarcity of internal resources. There were cases under the Second Project where delays of up to one year occurred between actual project expenditures and requests for reimbursement from the loan. The complex flow of funds resulted in further delays as loan disbursements were slow in finding their way to project suppliers and contractors. In Meknbs, it was reported in 1984 that some DH 18 million (US$2.0 million) had been paid to contractors for eligible expenditures under the project, yet no request for reimbursement from the Bank had been made. 3.11 Such inaction is inconsistent with what is normally required when there is a severe shortage of funds. On the other hand, it is consistent with the behavior of agencies that do not bear direct responsibility for the financial consequences of their actions, as was the case with the local project units. At the other extreme, actions typical of a financial squeeze, such as delaying payments to contractors, did occur. But such delays should have been discouraged so as to avoid calling the credibility of the operation into question in the eyes of the private sector. 3.12 Project counterpart funding represented a major difficulty during a period of fiscal constraint. Nevertheless, for the Second Project, GOM was able to provide 85.3% of expected (dirham) counterpart funding for a significantly reduced project that did not require finance for construction loans. This was a notable performance by a government facing financial duress. Altbough data on final project financing is not available from the Rabat Project PCR, it is likely that GOM exceeded its dirham commitments to the operation. 3.13 One positive financial result at the micro-level was the mobilization of private savings by the project. The majority of improvements and extensions made to project'supported dwellings derived from the application of private savings by beneficiaries. When asked by the Audit mission, beneficiaries universally indicated private savings or loans within the family as the source of the wealth invested in the dwelling unit. Due to the lack of land titles, most beneficiaries did not have access to formal construction financing (para. 3.02). However, even in the case of the sites and services in Rabat, where land titles had been awarded, demand for such financing was negligible. This suggests that reliance on self-finance through private savings is an essential ingredient of low-income housing projects in Morocco. This is one of the major lessons to be learned from the experience (para. 6.07). 15 C. Non-Use of Building Materials Loans 3.14 The provision of building materials loans to help beneficiaries finance home construction was highlighted by the Second Project in particular. As already mentioned, however, their use was preempted by the ipplication of private savings. At the same time, the local Popular Banks (BPs) that were to have provided such financing felt uneasy about the lack of guarantees due to the absence of firm land titles in the hands of potential borrowers. In actual fact, loan provision through the BPs was negligible. Whether this was due primarily to lack of demand or supply is still a much discussed issue. It was probably caused in part by both. 3.15 On the supply side, the local BPs in MeknAs and Kenitra were unwilling to provide what they considered to be unsecured and high risk loans. Also, a large number of very small operations would have resulted in relatively high administrative costs for the banks, much more than the 1% administration fee allowed under the project. In addition, BCP was unwilling to enter an operation where the costs were unknown and the loan amount was likely to prove inadequate. In short, BCP was unenthusiastic about the construction financing proposed by the project. Although BCP's conservative lending criteria were known early on to the Bank (SAR-R para. 4.16), the former's behavior was widely considered in Morocco to have been too rigorous, too much like that of a commercial bank. Whether this criticism was fair or whether it simply highlights yet another aspect of the divorce between the technical and financial aspects of the projects discussed elsewhere (paras. 4.19 - 4.22) is a matter for discussion. In hindsight, the construction loan component should have been prepared in closer collaboration with BCP. Further progress in sector finance had to await the Housing Finance Project (Loan 2245-MOR) with GOM's Real Estate and Hotel Credit Bank (CIH) approved in 1983, a project that has yet to be audited by OED. Whether this operation effectively built a bridge between technical and financial concerns or simply shifted support from one divorced party to the other will have to be determined by a future evaluation. D. Project Physical Achievements 3.16 By improving large low-income neighborhoods, both projects succeeded in making important contributions to the physical environments of their respective cities, a result that is treated as a point of special interest later in this report (pares. 4.01-4.04). The projects also made significant improvements to the overall housing situation in Rabat, Meknbs and Kenitra. The ideal way of measuring this would be to assess the condition of the housing stock in the respective cities before and after the projects. While the information needed to make this comparison is not complete, it is possible to draw some preliminary conclusions from the figures presented in Table 1. 3.17 The projects brought benefits to appreciable shares of the populations of Rabat (5.6%), Meknbs (10.1%), and Kenitra (12.32). In both Rabat and Meknis, the total bidonville population fell during the 1980s. Assuming no new bidonvilles were formed in this period, one might conclude that project upgrading contributed to 80% of this decrease in Rabat and 66% in Meknbs. In the latter city, bidonvilles have been eradicated almost altogether. The situation in Kenitra is somewhat different, however. Located in a region of recent 16 economic expansion, it has been one of Morocco's fastest growing cities and, as a resultv was still undergoing a process of rapid bidonville formation. Thus, despite the improved conditions for some 34,000 people through project upgrading, the total population living in bidonvilles grew by about 9,000 during the 1980s. Although the project by itself was not sufficient to halt the expans n of bidonvilles, presumably their share of the total population would have grown more rapidly in the absence of the Second Project. Table 1. SELECTED DATA ON PROJECT CITIES (in thousands) RAW MEKNES (1) Urban Population (1980) 725(*) 360 216 (1989) 1,159 454 391 (2) Bidonville Population (1980) 196(*) 70 65 (1989) 121 26 74 (3) Project Beneficiary Population Upgrading 60 29 34 Sites and Services 5 17 14 (4) Beneficiaries as share of Urban Population (1989) [%(3)/(1)] 5.6Z 10.1% 12.3 Sources: Populations (1980) Project SARs; (1989) Recensement G6ndrale des Bidonvilles. Project Beneficiaries: (Rabat) Project SAR; (Mekn6s, Kenitra) Project PCR. Notess 5.5 persons per household; (*) 1978 data. E. Target Beneficiaries and Poverty Impact 3.18 In accordance with Bank sector priorities, both projects were targeted on the urban poor. This was done by providing low standards of housing that higher income households were expected to find unacceptable rather than through the tedious vetting of family incomes to determine eligibility. For the conditions in Morocco and many other countries, this was a pragmatic approach since household earnings are very difficult to measure and verify where there are informal sources of income. Even though some families living in the project sites clearly do not belong to the urban poor, the majority appear to be low- income. Projects such as this cannot guarantee the rigorous spatial segregation of the poor if higher income households are willing to live among them. This was especially noticeable in some of the sites of the Meknbs component of the Second 17 Project. Detailed monitoring of current beneficiary income levels, moreover, still needs to be carried out. 3.19 Further monitoring and evaluation of project results could throw light on a number of important issues. For instance, there were many reported cases of beneficiaries selling their plots after having improved them. Also, renting part of the unit was regarded as common practice. In other cases, the original allottees allegedly shared expenses with other households, usually in the same extended family, so as to make the plots affordable. This practice, however, was not allowed in Kenitra. More data would be required to determine the actual extent of phenomena such as these. 3.20 Where there had to be explicit selection of beneficiaries, in turn, such as in the sites and services schemes, the process generally resulted in the choice of low-income beneficiaries. In the case of the sites and services households visited by the Audit mission, all were former residents of the upgraded areas. This suggests that the agreed criteria of giving priority to households displaced by project works were, in fact, applied. 3.21 Under the Rabat Project, finally, there was a special case of beneficiary targeting. The situation of young women in particular was to have * been improved through the provision of a vocational training and work center whose operation was thoroughly detailed in project documentation. Unfortunately, more attention was paid to how the center should operate in theory, than to the * practical issues of how it would be implemented and whether there was local support for it. As a result, the project's efforts to bring benefits to this group floundered when the component was cancelled because of inter-agency coordination problems and a lack of counterpart funds. F. Employment Creation 3.22 As with numerous urban projects in other countries, the employment generation components of these two operations did not succeed in creating many new work-places for the intended beneficiaries. Although 2,500 new jobs were to be created in Rabat (SAR-R para. 3.07) and another 2,270 in Mekn&9 and Kenitra (SAR-II para. 6.08), neither PCR provides an estimate of the actual number of jobs generated. The scaling down and cancellation of some of these components, especially in Rabat, however, probably meant that results fell far short of initial expectations even though some actions, such as the extension of credit to small-scale businesses by BCP in Rabat and the provision of 289 small plote for similar enterprises in Meknbs and Kenitra, were taken. 3.23 While there was much enthusiasm with respect to employment generation on the Bank's part during project preparation, the importance of these components diminished during supervision as the missions' attention turned to the larger upgrading and sites and services works. Furthermore, it was clear from discussions with Government officials that interest in Morocco was limited for this kind of intervention. The impression is that these components may have been included largely in order to make the projects conform to the Bank's general blueprint for urban development operations at the time. 18 G. Limited-Role of Municigalities 3.24 Both operations were conceived as central government interventions in the respective project cities. One of the main differences between the two, however, was the greater participation of the municipalities as executing agencies in the Second Project. Under the Rabat Project, the role of the municipality was never intended to be important. Today, in contrast, there is a growing awareness of the importance of local governments in Morocco under the orientation of the Ministry of the Interior - MOI. For that reason, the performance of the municipalities is treated in more detail as a point of special interest later in this report (paras. 4.27-4.30). 3.25 In spite of their limited participation in project implementation, maintenance of completed project works remains a responsibility of the corresponding local governments. Particularly with respect to solid waste collection and street cleaning, however, maintenance is wanting in all three cities. The municipal maintenance center financed under the second operation in Kenitra appears to have helped the project site very little. In spite of major earthworks having been completed, sandy soil continues to build up, blocking drains in the Saknia area. A similar center planned for Meknbs was cancelled from the project for reasons unknown to present local authorities. H. Sector Policy Impact 3.26 From the project objectives (paras. 2.18-2.20), both operations were meant to have a significant policy impact by introducing new shelter priorities in Morocco. Both PCRs highlight the policy achievements made (PCR-R para. 1.04 ard PCR-II Evaluation Summary para. 1). An evaluation of policy impact, however, requires a comparison of present policy with that prevailing prior to the Bank- financed projects. The Audit's review of GOM sector policy over the period reveals a consistent commitment to bidonville eradication which is still the case today. There thus appears to have been much less of a shift in the Government's position than the Bank had wished. Even if all of the relatively minor changes observed can be attributed to the projects, their policy impact appears to have been much more modest than the ambitious aims laid out at appraisal. Because of its importance, this matter is dealt with further as a point of special interest later in this report (paras. 4.05-4.13). 1. Project Environmental Impact 3.27 By providing sanitary housing conditions in areas that were previously deprived of any urban services whatsoever, both projects made significant improvements to the physical environments of their respective cities. If and when a more detailed evaluation of project results is undertaken, it should reveal improvements in the health conditions of beneficiaries living in the upgraded areas. The gains will only be sustained, however, if maintenance is improved. Unless blocked drains are cleared in Kenitra, for example, flooding and pollution could return the sites to their unsanitary pre-project conditions. 19 IV. POINTS OF SPECIAL INTEREST A. Major Contributions to Urban Physical Develo2ment 4.01 In all three cities, the projects contributed to a significant improvement in the urban environment. The Rabat Project, completed in 1985, is now being consolidated. This permits a fuller appreciation of its condition as a mature housing area for its inhabitants, as well as of its contribution to the city's commercial activity. The upgraded neighborhoods of Douar Dous, Masdid and Hajja are now thriving parts of the larger city. The extensive improvements made to dwellings by residents are now more or less complete, giving them the appearance of consolidated low-income residential and commercial neighborhoods. Also important in a nation that values its architectural heritage, much of the improved area incorporates the architectural traditions of North Africa so vividly expressed in the ancient medina of Fez. In particular, Douar Doum, which is in the more favorable location overlooking the river valley, is an aesthetically attractive residential area, reminiscent of Morocco's urban medina. 4.02 The architectural quality of Douar Maadid is perhaps not so high, but it has, nevertheless, become a vibrant commercial center having the fourth largest retail sales volume in Rabat. Unfortunately, the commercial success of Douar Maadid has led to excessive crowding and serious problems of traffic congestion that need to be addressed by the local authority of Rabat. Special mention should be made of the experimental sites and services scheme called "La Butte" within the same neighborhood. This scheme has been successfully developed by low-income beneficiaries and has blended into the urban landscape. There is now little difference between it and nearby upgraded areas. 4.03 Being more recent, both the upgrading and sites and services works of the Second Project in Meknbs are still far from complete with many empty plots remaining to be developed. Even so, there are clear indications that the neighborhoods of Bordj Moulay Omar, Ain Chebik and Sidi Baba will all be fully incorporated into the urba- fabric of Mekn&s. Unlike Rabat and Kenitra (see below), upgrading in Meknbs followed the original plot subdivisions only very roughly. In fact, as the result of changing plot boundaries, very little remains of the original area making it more like an urban renewal operation than a typical upgrading exercise, an aspect dealt with in further detail below (paras. 4.14-4.18). The upgrading and sites and services areas have, nevertheless, made an important contribution to Meknbs and Kenitra by allowing the inevitable urban expansion to proceed in an improved physical environment characterized by safe water supply, electricity and guaranteed access. 4.04 As already mentioned above, the projects also contributed to the reduction of bidonvilles in Rabat and Mekn&s and helped to curtail their growth in Kenitra (para. 3.18). The areas improved under the project were integrated into their respective cities in a form that was more acceptable to local authorities. Beneficiaries interviewed by the Audit mission in all three cities agreed that the housing provided as a result of the project represented an improvement over their previous living conditions. The positive rates of return re-calculated for both projects indicate that these benefits were also recognized by the market through the increased value of the improved lots. 20 B. Policy Impact 4.05 Although the projects made a valuable contribution to the cities in which they were undertaken, their aim was to have a much larger impact. Their broader objective was to demonstrate the feasibility of the low-cost upgrading and sites and services approach in an attempt to convince the Moroccan Government to abandon its traditional policy of alum demolition. For the projects to have been fully successful in terms of this objective, it would be necessary to verify that GOM no longer subscribes to bidonville eradication. For Rabat, the PCR concludes that the project was "instrumental in influencing the Government's policy of alum upgrading versus demolition" (PCR-R para. 1.04). The Second Project, in turn, is credited with having "helped the Kingdom of Morocco develop a new housing policy" (PCR-II Evaluation Summary para. 1). From the Audit's review of recent Government policy statements and actions, however, the changes have been much less substantive than hoped for. The eradication of bidonvilles remains firmly on the urban policy agenda in Morocco. The Bank now clearly recognizes this in the SAR for the Second Housing Finance Project (Ln 3122-MOR) in its reference to "the elimination of slum areas" (para. 2.09). 4.06 When preparation of the Rabat Project began in the early 1970s, GOM's policy was explicitly in favor of eradicating bidonvilles and building high standard housing. Medium to high income households tended to benefit most from this approach. The Bank's position in favor of upgrading and low-cost affordable standards was reluctantly accepted by the Government insofar as the latter agreed to finance the project jointly with the Bank. The policy dialogue during 1972- 1977, however, was a difficult one which nearly broke down at three distinct moments. As mentioned earlier, the last occurred during the pre-appraisal mission for the Rabat Project in 1977, when the Minister of Housing announced that the approach espoused by the Bank was no longer acceptable (paras. 2.13- 2.14). Purposefully retaining a slum was considered to be a blemish on the urban fabric of the nation's capital. The aesthetic argument is a very strong one in a country that values its architectural tradition. Self-help methods were considered the poor man's solution. Clearly, there was much reluctance within the Moroccan Government to agreeing to the project at all. 4.07 The Rabat Project, however, did successfully demonstrate that the proposed upgrading and sites and services solutions were less expensive than the approaches previously followed. The operation provided a blueprint for a less costly approach which would have been expected to appeal to a government under increasing financial constraints. The Second Project included studies for follow-on operations in other cities, but none were undertaken. GOM on the other hand, also suspended its plans for massive countrywide programs of bidonville eradication. But this latter stance, in the view of the Audit, was taken primarily because of the lack of funding for such a program rather than any change of philosophy with respect to urban policy. 4.08 As soon as resources again become available, GOM may resume pursuit of its long-term goal of ridding Moroccan cities of slums. The decision not to undertake a nationwide program of bidonville upgrading and the Government's recent policy actions and statements confirm its commitment to the eradication goal.. During implementation of the Second Project, preparation began for an informal housing ('habitat clandestin') project. Discussions broke down over the 21 issue of awarding titles to beneficiaries which GOM felt would legitimize and encourage this form of irregular settlement. 4.09 Other indications of the Government's desire to rid Moroccan cities of bidonvilles were the creation of the National Agency for Combat against Unsanitary Housing (ANHI - Agence Nationale de Lutte contre l'Habitat Insalubre) in 1984 and the establishment of seven regional planning and construction agencies (ERACs) under MOH. Together, these agencies presently constitute the executive arm of GOM's institutional arrangements for the sector. ANHI is the most directly concerned with bidonvilles. According to its legal statutes, its tasks include land acquisition to "reabsorb" bidonvilles and the construction of dwellings to rehouse former bidonvilles residents. Its current plans entail delivery of some 11,200 units annually, but funding has not yet been assured. Although ANHI has been successful in mobilizing pre-payments by potential beneficiaries during construction, the prices charged require central government budget subsidi:s in order for the program to continue. The principle of cost recovery, which was not well demonstrated by the Bank-financed projects, has yet to be fully incorporated into ANHI's program. Until it is, ANHI's activity is likely to continue to decline. The ERACs, in turn, serve mainly the middle income housing market and have recently increased the scale of their operations, some of which are indirectly funded through the Housing Finance Project (Loan . 2245-MOR) and its successor. 4.10 Still another indication of the Government's desire to come to terms * with squatter settlements is the recent comprehensive national census of bidonvillese. This survey found that a total of 1,064 bidonvilles existed in the country, home to some 890,920 people or 7.9% of the urban population in 1989. Although this is an important study aer se, its underlying purpose is to serve as a policy instrument. From the data it provides, each regional representation is able to design a program to eradicate bidonvilles within its area of jurisdiction. Work is underway on these programs and the limited availability of financing is the only major constraint on their execution. 4.11 It is important to stress that antipathy toward squatter settlements is much stronger in Morocco than in many other countries and is found at the highest levels. A key objection to bidonvilles is an aesthetic one. As the King of Morocco himself has suidt "Whether dealing with luxury, average or modest housing, their aesthetics must rigorously protect the beauty of our country." Indeed, Morocco must be one of the few countries in the world where architectural standards feature prominently in a national development plan, such is the pride in its urban design heritage whic s4.s, indeed, worth preserving. 4.12 Finally, clear indication of the Government'3 sector priorities comer from the current National Development Plan (1988-1992), which explicitly discards the hope that policy toward squatter settlements would follow the Bank's early prescription. In its section on economic infrastructure, the Plan's first proposed action in the field of urban housing is the "eradication of bidonvilles through the development of land management activities and conventional policy * Royaume du Maroc, Ministbre de 1'Habitat, Recensement Gdndral des Bidonvilles, Rabat, June 1989. 22 associating the efforts of the State, local municipalities, public entities and beneficiaries."' In this connection, MOH's regional delegations have the responsibility to determine program costs and to mobilize the necessary resources. 4.13 On the Bank's side, in turn, even before the projects reviewed in this document were completed, interest shifted to the financial aspects of Morocco's urban sector. The Region pointed out that the shift to housing finance lending instruments was made to provide a more effective means of encouraging programs designed to address the country's critical housing shortage and increase its accessibility and affordability, particularly for low income families. This, nevertheless, meant that the ensuing policy dialogue became less directly concerned with the option of bidonville upgrading versus demolition. One Bank staff member has characterized this as a shift from "charity to development." The assignment of staff with skills in the areas of economics and finance helped make this change possible. C. Renewal versus Upgrading 4.14 Linked to the issue of policy impact is the way in which the Bank's proposal to upgrade low-income settlements has been interpreted in Morocco. Although GOM appeared to have temporarily overcome its anathema towards bidonvilles by agreeing to improve them through the Bank-financed projects, the work undertaken -- especially in Meknbs -- did not correspond to the spirit of the upgrading approach espoused by the Bank. By redrawing plot boundaries (something that was done to a much lesser extent in Rabat and Kenitra) and reducing the densities of the areas, the community which emerged from the so- called upgrading bore little resemblance to what had been there before. Since little of the original settlement remains, this process might better be characterized as urban renewal. As far as early Bank-supported urban projects were concerned, squatter upgrading assumed that "households would be permitted to stay in their existing settlements and that, with only minimal demolition and relocation, infrastructure -- including water supply, sanitation, roads, footpaths, drainage, and electricity -- would be extended into the settlements.'o What was done in Meknbs clearly did not correspond to this description. 4.15 First, the existing units on the sites to be upgraded were entirely demolished. They were generally poorly built shacks, so low as to make it impossible to stand inside and in a very precarious condition that could not have been expected to become permanent shelter. GOM and the local authorities correctly felt it necessary to remove them. The continued participation of residents in the scheme, in fact, was made conditional upon their demolishing their shacks. According both to the local officials and beneficiaries interviewed by the Audit mission, residents who felt sure they would stay in the ' Royaume du Marce, ?remier Ministre, Plan d'Orientation pour le D6velogvement Economiaue at Sociale 1988-1992, Rabat, Section III. 10 World Bank, Learning by Doings World Bank Lending for Urban Development. 1972-1482, Washington D.C. 1983, p. 16 23 area were willing to demolish their shacks in return for security of tenure on a clearly demarcated plot. 4.16 While this did interrupt their living arrangements, demolition of the shacks did not bring as much disruption to the community as the fresh demarcation of plot boundaries. Like most equatter settlements worldwide, initial plot subdivisions followed an irregular pattern, even though residents were usually careful to preserve rights-of-way. In Rabat and Kenitra, upgrading more or less followed the original plot boundaries so as to minimize disruption and comply with the Loan Agreement. In MeknAs, local authorities attempted to redraw the site layout so that it would be more regular, in conformity with what they thought a modern urban development should look like. Furthermore, one of their objectives was to reduce the densities by removing "excess" households. Regular site geometry was considered an aim in itself, in addition to the need for greater circulation and open space. Implementing the new site layouts, however, did not comply with a key condition of the Loan Agreement for the Second Project which required that "plot boundaries ... be determined according to the present settlement pattern of the sites" (LA-II Schedule 2 A.(a)). More important than the formal omission, however, was the disruption caused by having to relocate a larger number of households and disputes between neighbors who had happily accepted their plot boundaries in the past. 4.17 Local officials in Mekn&s reported that, during the early stages of implementation the "spirit of the project," which meant disturbing as little as possible, was maintained. Later, however, they argued that the residents who expected to remain themselves wanted more regular lots and accepted MOR's view that the existing rights-of-vay were too narrow. The beneficiaries correctly felt that the value of their lots would increase if they were more regular and accessible. The change of approach was discussed with Bank missions, which reportedly agreed to the more regular layouts proposed, even though there was no formal amendment of the Loan Agreement to this effect. The change of approach is mentioned in the Borrower's PCR as one of the causes of implementation delays since much extra design work was required (PCR-II-MOH II.C.). 4.18 In Kenitra, in turn, and particularly in the neighborhood of Saknia, upgrading was easier than in Mekn&s since a grid street layout had been in place since 1945, much before the bidonville was established. Where this is not the case, flexibility such as that demonstrated by the Bank in Mekn4s will be required in order to find the best technical solution for each specific location. In all cases, however, involuntary displacement should be kept to a minimum to limit disruption and maintain control over costs. D. Divorce of Physical and Financial Asnects of Prolect Management 4.19 The poor cost recovery performance of both operations was due in large measure to the lack of legal past titles which precluded beneficiaries from borrowing money from financial intermediaries (parse. 3.14-3.15). This situation arose in large measure as a result of the limited attention given to the projects' financial aspects. Because of their innovative nature, GOM and the Bank gave primary emphasis during project preparation and implementation to their physical elements. Local project management and Bank missions were staffed accordingly. Meanwhile, the financial aspects were largely the concern of BCP, 24 an agency that should have been more closely involved in project preparation and supervision. One of the main victims of the divorce between technical and financial aspects was cost recovery. For the managers of the physical works, however, this was just one problem among many. Since it did not directly affect on-site construction, it did not receive priority attention. 4.20 both project documents make reference to the need for cost recovery. The appraisal report for the Rabat Project refers to the "principles" of cost recovery (SAR-R paras. 6.01-6.04) and mentions monthly installments to be paid by the upgrading and sites and services beneficiaries (SAR-R para. 6.03). Who was to pay whom, how much, and through which institutions, however, is not clear. Nor were institutional arrangements for cost recovery specified in connection with the Second Project, whose SAR is very economical in its treatment of the subject (SAR-II paras. 5.01-5.04). In addition, project supervision, which concentrated on physical development aspects did not give priority to cost recovery. Despite this, 75% of costs were expected to be recovered and the funds generated were to be "recycled into new projects through the national budget process" (SAR-Il para. 6.03). For this to have occurred, greater attention would have had to have been given to the administrative arrangements for the payments to be made and their channeling to the appropriate sector, as through a development fund for example. 4.21 Given this situation, it is not surprising that the projects were not replicable. The public sector financial squeeze following the first IMF Standby Agreement in 1979 ruled out continuation of the program in the absence of full cost recovery from its beneficiaries. Lacking this, only a dramatic improvement in public finances or a resumption of external grant funding could guarantee implementation of the GOM's ambitious shelter plans (such as through ANHI) for the low-income segment of the urban population (para. 4.09). 4.22 Even though financial questions were not highlighted in connection with the projects, financial acumen was demonstrated by local officials in several instances. Under the Second Project, the use of revenues from the sale of commercial plots to increase cash flow was well understood in both Meknbs and Kenitra. But performance, with 39% arrears (PCR-II-MOH), was disappointing and formal arrangements for cross subsidies to help finance the less profitable parts of the operation have apparently yet to be devised. One notable exception to the generally poor cost recovery performance, however, was the sale of sites and service plots at "La Butte" in Rabat. A delinquency rate of only 5% was reported. Nevertheless, better integration of the physical and financial aspects of project management is essential if exceptions such as these are to become the rule. E. Manaaement of World Bank Loans in Morocco 4.23 Of the US$54 million of the two loans combined, only US$17.7 was disbursed, less than a third of the amount intended. While reductions in project scope contributed marginally to this result, the main reasor. was the sharp successive devaluations of the dirham against the US dollar during the project implementation period. Due to tardy cancellations of unused loan funds, moreover, GOM had to pay heavy commitment fees on the amounts not disbursed. This was another result of the divorce between the physical and financial aspects 25 of the projects, only in this instance it was the financial managers who were in the dark about prospects for future physical development. 4.24 The Ministry of Finance, together with the Ministry of Budget and Plan, is responsible for management and servicing of World Bank loans to Morocco. In the case of the two urban development projects, MOF was aware of the very slow disbursements of the respective loans. As financial manager, however, it was not in a position to automatically know the causes of the delays or what the prospects for improvement were. Accordingly, it had to await the initiative of MOH, through local project management, to be informed that most of the loan would not be used. For its part, local management was little concerned about accumulation of a large US dollar surplus in the loan account since it was not responsible for commitment fees. To the contrary, in their eyes this seemed to offer the comforting security of an unlimited financial reserve, albeit one which could only be used if considerable extra counterpart funds were forthcoming. 4.25 The separation of technical and financial decision making, in short, meant that requests to cancel loan amounts were very slow to be made. The first cancellation (of US$11 million) was only made in 1985, even though supervision missions had initially drawn attention to the problem in 1983. The inertia in dealing with surplus loan funds was compounded by a feeling at the local level, that cancellation of part of the balance would mean actually losing money. As long as the divorce between the management of project technical and financial aspects continued, there was little response to the evolving financial situation at the local level. 4.26 In the case of the Rabat Project, the US$7.9 million loan amount actually disbursed appears to fall short of what might have been expected based on project cost at completion. At appraisal, the Bank loan was expected to finance 47.9% of total project costs, but the amount eventually disbursed corresponds to only 30.9% of final project costs. Had the appraisal average been maintained, US$12.3 million of the Bank loan would have been disbursed instead. At first, this result appears surprising in view of the fact that the lowest disbursement percentage in the Loan Agreement was 57% for civil works. One of the probable causes of the loss of some US$4.4 million in loan proceeds was the Government's slowness in processing requests for reimbursement. Being denominated in a fast devaluing currency, the US dollar value of reimbursements fell by as much as 20% over a six month period in some years. A similar phenomenon occurred in connection with the Second Project. Tighter control of project financial management could have reduce these losses. F. Limited Role of Municipalities in all Cities 4.27 With the increasing emphasis on municipal development in the Bank, it is of special interest to determine how the munici;alities participated in the two projects and with what results. To do so, however, it is important to recall the very weak position municipalities occupied in Morocco's centralized political structure, especially at the time the Rabat Project was being prepared. Municipalities presently have growing responsibilities in the country, but, even though they have guaranteed participation in the revenues of the new value added tax, they continue to depend on the central government through MOI. Even local taxes are still collected by central authorities who redistribute them to local 26 governments. As a result, the Rabat Project envisioned only a minor role for the municipality which was the beneficiary of 20 man-months of technical assistance for municipal maintenance. Owing to the long time elapsed since project completion, it is difficult to assess the impact of the component today. 4.28 The main feature distinguishing the Second Project from the earlier operation was a component designed to improve municipal services. Although this helped ensure more active municipal participation in project implementation, the Audit mission was struck by the lack of familiarity with the operation demonstrated by current local officials in Meknbs. In both Meknbs and Kenitra, moreover, maintenance of the project sites, which is a municipal responsibility, is wanting. Particularly in Kenitra, the sandy soil at the project site in Saknia is threatening to silt up the city's drainage system unless street cleaning and drain servicing are urgently undertaken. In Rabat, in turn, solid waste collection from the project areas of Douar Doum and Douar Maadid is inadequate in spite of the fact that the municipality assumed formal responsibility for this service. 4.29 These problems, notwithstanding, the projects did provide an important learning experience for the Bank with respect to the evolving role of municipalities in Morocco and how this might be consolidated in the future. The contribution of municipalities to future urban development in the country, accordingly, should be expected to increase. G. Small Preparation Effort Needed for Both Proiects 4.30 At first glance, when c spared with other early urban operations in Africa, the 92 staff weeks of mission time required to prepare the Rabat Project appears to have been an excessive use of Bank resources for a small operation.u However, as noted above, most of that effort was expended on preparation of the Casablanca Urban Development Project (paras. 2.04-2.10). Since the idea of a project in Rabat came only after the proposed operation in Casablanca was thwarted, it is reasonable to conclude that little of the earlier Bank effort benefitted Rabat. The move from Casablanca to Rabat involved a complete change of cast at the local level, so even the benefit of officials in Casablanca having worked with the Bank was lost. Also, previous policy dialogue to persuade the Moroccan Government to accept the Bank's upgrading approach brought little benefit to the Rabat Project which was almost immediately threatened with cancellation by M0H (paras. 2.13-2.15). 4.31 The Bank decided to cut considerable losses in leaving Casablanca. Assuming that very little of that effort could be salvaged for Rabat, comfort may be drawn from the fact that only 22 staff weeks were needed to prepare the latter operation over a period of barely six months. The aim of urgently replacing the Casablanca operation forced the pace of preparation and appraisal in Rabat. The preparation of the Second Project was also expeditious, requiring only 20 staff 1 Forty-two staff weeks to prepare C8te d'Ivoire: First Urban Development Project (US$44 million Loans 134711348 IVC approved December 1976) and 45 staff weeks to prepare Nigeria (Bauchi) Urban Development Project (US$13.7 million Loan 1767-UNI approved February 1980). 27 weeks through to appraisal. The experience of these two operations demonstrates that complex projects can be prepared quickly provided there is a clear objective and commitment by both the borrower and the Bank. V. ROLE OF THE BANK 5.01 Both projects were the outcome of a long and protracted dialogue between GOM and the Bank. With notable consistency, the Bank made its basic priorities for urban development evident to Moroccan authorities. Especially with regard to bidonville upgrading and low-cost housing standards, the approach advocated by the Bank was not always welcomed by COM. A difficult dialogue was made more complex by the frequent change in Bank mission staff, from whom COM received conflicting signals about how a project could be implemented to fulfil basic aims that were themselves well understood. Despite questioning in the Bank about the advisability of continuing to work in the urban sector in Morocco, staff persisted even after successive setbacks. When necessary, as with the threatened cancellation of the Rabat Project during pre-appraisal, the Bank took a firm position which was clearly communicated to the Moroccan Government. 5.02 Considerable Bank resources were spent on the ill-fated Casablanca Urban Development Project. The Bank and GOM pursued the operation in good faith, hoping that it would lead to a project. In retrospect, a more prompt assessment by the Bank of the operation's technical feasibility could probably have resulted in a much earlier change of course, thus avoiding a costly deployment of resources. Performance improved dramatically in the follow-on operations, however. Although some consider the time and resources spent trying to prepare the Casablanca operation as having been lost, they can be regarded as an important, albeit expensive, schooling for the Bank in Morocco's urban sector. 5.03 In spite of the difficulties, good relations were enjoyed between Bank urban staff and their interlocutors in Morocco. While not always agreeing with the approach recommended, GOM officials recognized that Bank missions helped to stimulate improved organization of the sector in the country. Frequent mention was also made of the dedication of individual Bank staff members to their particular tasks. VI. MAJOR LESSONS TO BE LEARNED A. Need to Integrate Physical and Financial Aspects of Projects 6.01 The early sector dialogue between the Moroccan Government and the Bank was largely about physical standards for low-income housing development. Financial issues were only to come to the fore much later. For this reason, it was consistent of the Bank to emphasize project design aspects in order to ensure adequate physical implementation. The staffing of Bank missions was appropriate for this purpose. 6.02 The experience of these operations, however, also demonstrates what can occur when financial aspects are not given adequate attention or are handled by agencies that are unaware of the project's physical development. Thus arose the divorce between the technical and physical aspects of project implementation mentioned above (paras. 4.19-4.22). Among the consequences of this divorce were 28 a disappointing cost recovery performance and a lethargic flow of funds which resulted in financial losses to the project. 6.03 The clear lesson is the need to better integrate the physical and financial aspects of urban projects, so that they are managed in tandem. Increasing the emphasis on financial management is essential, but by itself not sufficient. It is likewise important that improved financial management be accompanied by better management of physical execution, lest the divorce continue with only a change of perspective. B. Urban Proiects Can Be Prepared Quickly 6.04 When the objective is clear and the support of local authorities is forthcoming, it is possible to prepare urban development projects quickly. Both of the operations assessed here (excepting the aborted project in Casablanca) were prepared in some six months through to appraisal. Most problems that arose during implementation could be dealt with during supervision. The experience of these operations, accordingly, is a call for a shift of Bank staff resources from preparation to supervision. C. Land Titling as a Pre-requisite for Cost Recovery 6.05 Land titling is a complex process in Morocco, but is an essential pre-requisite for cost recovery to proceed. It is necessary not only for banks, but for any agency responsible for cost recovery, to have a firm guarantee that the amounts due will be paid. As in most other countries, the main guarantee is property which can only be mortgaged if a potential borrower has full legal rights in the form of a freehold title. If not, possible sanctions, such as foreclosure, have no firm legal base and are thus ineffective. The Region confirmed that land titling problems persisted in the follow-on project (Ln. 2245-MOR), but that CIR's special agreements with private developers helped to overcome some of the bottlenecks. D. The Need for Substantial Household Savings for Proiect Success 6.06 In all three project cities, most home improvements and expansions were financed by private savings. These were especially important in view of the legal impediments to formal lending due to the absence of land titles. Remarkably, even households whose income is low today were apparently able to mobilize considerable savings. In Morocco, personal savings in the form of jewelry, livestock and rural property play an important role in the informal economy. It is clear from the project experiences, however, reviewed that private savings were essential for their successful completion. In addition to affordability analysis, future operations should include a review of patterns of wealth, as well as income, in order to determine the extent to which low-income households can participate in financing their own housing. E. Proiect Impact is Local 6.07 At appraisal, project impact was expected to be felt both nationally, where there would be "long-ranging benefits for all urban areas in Morocco," and internationally. The Rabat Project in particular was intended to serve as model 29 for the North Africa region (SAR-R para. 8.01). Although there have been overseas visitors to the project in Rabat, the demonstration effect of these operations is difficult to assess. As previously discussed, moreover, their policy impact has been less than desired (paras. 4.05-4.13). What is not in doubt, however, is that they have had a favorable impact on the local physical environment (paras. 4.01-4.04.). While the potential exists for a broader demonstration impact, this will only be achieved if the experiences are actively monitored and evaluated and their results adequately disseminated. The Bank can and should play a direct role in the broader dissemination of project experiences that it helps to finance.