Document of The World Bank

Transcription

1 Public Disclosure Authorized Public Disclosure Authorized Public Disclosure Authorized Public Disclosure Authorized Infrastructure Operations Division Southern Africa Department Africa Region Document of The World Bank FOR OFFICIAL USE ONLY PROJECT COMPLETION REPORT ZIMBABWE URBAN DEVELOPMENT PROJECT (LOAN 2445-ZIM) DECEMBER 23, 1994 Report No 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.

2 CURRENCY EOUIVALENTS Currency Unit : Zimbabwe dollar (Z$) EXCHANGE RATES Appraisal Year Average I 1US$ = Z$ Intervening Year Average I 1US$ = Z$ Completion Year Average IUS$ = Z$ GLOSSARY CDC LA MPCNH MFEPD MLGRUD MT OJT BS PCMU UTU ZAAT ZESA ZTSB ZUPCO Commonwealth Development Corporation Loan Agreement Ministry of Public Construction and National Housing Ministry of Finance, Economic Planning and Development Ministry of Local Government, Rural and Urban Development Ministry of Transport On-the-Job Training Building Societies (Central African Building Society, Beverley Building Society and Founders Building Society) Program Coordination and Monitoring Unit Urban Transport Unit Zimbabwe Association of Accounting Technicians Zimbabwe Electricity Supply Authority Zimbabwe Traffic Safety Board Zimbabwe United Passenger Company FISCAL YEAR July 1 - June 30

3 THE WORLD BANK Washington, D.C U.S.A. FOR OFFICIAL USE ONLY Office of Director-General Operations Evaluation December 23, 1994 MEMORANDUM TO THE EXECUTIVE DIRECTORS AND THE PRESIDENT SUBJECT: Project Completion Report on Zimbabwe Urban Development Project (Loan 2445-ZIM) Attached is the Project Completion Report on Zimbabwe - Urban Development Project (Loan 2445-ZIM) prepared by the Infrastructure Operations Division, Southern Africa Department of the Africa Regional Office. Borrower contributions were requested, but have not been received. The project focused on reform in the housing sector by pioneering the entry of existing private sector intermediaries in low-income housing finance. In addition, the project sought to strengthen infrastructure service delivery, including urban transport, through local governments. Progress towards the project's physical objectives was very good, as the number of plots serviced exceeded the targeted number by 85%. About 11,000 houses had been completed (and occupied) or were under construction at the time of loan closing -- delayed about two years to allow the use of additional proceeds created by exchange rate fluctuations. Institutional development of local governments (through training and technical assistance) was implemented only to a limited degree because of lack of interest from local governments. The project experience has provided a promising example of the involvement of experienced private sector intermediary institutions in low-income housing finance. The project outcome is rated as satisfactory, and its institutional development impact as modest. The sustainability of project benefits is rated as likely. The PCR is of satisfactory quality, and provides a comprehensive discussion of the problems and the achievements of the project as well as useful lessons for future operations. The project may be audited. Attachment This document has a restrictedistribution and may be used by recipients only in the performance of their official duties. Its codntents may not otherwise be disclosed without World Bank authorization.l

4

5 FOR OFFICIAL USE ONLY PROJECT COMPLEION REPORT ZIMABIWE URBAN DEVELOPMENT PROJECT (LOAN 2445ZIM) TABLE OF CONTENTS Page No. PREFACE... EVALUATION SUMMARY... i ii PART I PROJECT REVIEW FROM BANK'S PERSPECTIVE.1 Project Identity.1 Background.1 Project Objectives and Description. 2 Project Design and Organization. 3 Project Implementation. 5 Results. 9 Project Sustainability.10 Bank Performance.10 Borrower Performance.11 Relationship Between Bank and Borrower.11 Consultants Services.11 Project Documentation and Data.12 PART II. PROJECT REVIEW FROM BORROWERS' PERSPECTIVE PART m. STATISTICAL INFORMATION Related Bank Loans/Credit Project Timetable Cumulative Estimated and Actual Disbursements Project Implementation Project Costs and Financing Project Results: Direct Benefits Economic Impact Studies Status of Covenants Use of Bank Resources 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.l

6

7 i PROJECT COMPLETION REPORT ZIMBABWE URBAN DEVELOPMENT PROJECT (LOAN 2445-ZIM) PREFACE 1. This is the Project Completion Report (PCR) for the Urban Development Project for which Loan 2445 (US$ 43.0 million) was approved on June 19, 1984 and declared effective on October 7, Due to currency devaluations and to less than anticipated costs, the loan funds were sufficient to greatly increase the amount of work financed under the loan. However, an undisbursed balance of US$6,532,887 remained after two extensions of the Closing Date (September 30, 1993). This amount was cancelled and the Loan closed on January 28, The PCR was prepared by the Infrastructure Operations Division of the Southern Africa Department (Preface, Evaluation Summary, Parts I and IE), and by the Borrower (Part II). 3. Preparation of this PCR was based, inter alia, on the Staff Appraisal Report, the Loan Agreement and related documents; supervision reports; consultants' reports; correspondence between the Bank and the Borrower; and internal Bank memoranda.

8

9 -.1 - PROJECT COMPLETION REPORT ZIMBABVE URBAN DEVELOPMENT PROJECT (WLAN 2445ZIM) EVALUATION SUMMARY Project Objectives 1. The primary objective of the project was to promote Zimbabwe's financial and institutional capacity to supply affordable housing through reform of the housing delivery system and the housing mortgage market mainly by concentrating low-income housing activities of local authorities on the supply of serviced land and by introducing private sector financing of such housing through existing financial intermediaries. In addition, the project was to strengthen local government institutions for the supply of urban infrastructure and to support aspects of urban development (especially transportation) where new policy directives were being formulated. 2. Major project components as originally conceived, included: - the development of about 11,349 residential plots in Harare and three other towns with related infrastructure and community facilities; - the provision by three thrift institutions of long-term mortgage finance for construction of houses on these plots; - institutional development of various organs of local government throughout Zimbabwe; - policy and institutional development related to urban transport throughout Zimbabwe and the construction of a bus depot in Harare; and - development of facilities for management of the project. 3. Most of the project elements were the responsibility of the Ministry of Local Government, Rural and Urban Development (MLGRUD) while others fell to the Ministry of Public Construction and National Housing (MPCNH) and the Ministry of Transport (MT). A Steering Committee was to provide policy guidance in the work of various participating organizations and a Project Coordinating and Monitoring Unit was to coordinate administration of the loan.

10 Implementation Experience 4. The development of the original group of sites went according to schedule and generally at or below projected costs. The additional proceeds of the Bank loan created by exchange rate fluctuations were applied to help finance an additional group of plots (roughly doubling the original number to nearly 21,000). The execution of the second group was slower than the first because designs and other engineering documentation were not available and because plots could not be sold (as quickly as they were produced) owing to the lack of available mortgage finance. The availability of mortgage finance was subject to macroeconomic constraints in the financial sector created by a regulatory framework that differentiates interest rates between financial institutions. The assignment of plots and the construction of houses lagged behind original projections, especially when deposits for low-income housing at the thrift institutions tapered off because of changes towards more competitive interest saving rates offered to investors by the banking sector. 5. By the time of loan closure, 15,500 plots had been serviced, 5,000 more plots were in the course of being prepared, and about 11,000 houses had been completed or were under construction. The implementation process continued after loan closure but the pace of development was governed by the availability of housing finance. Generally the buyers appear to have had low incomes when they applied to the local authorities to be put on the housing waiting lists. 6. Elements of the project relating to development of local government institutions were implemented only to a limited extent primarily because the Zimbabwe authorities were reluctant to use external loan funds for this purpose. Some of the project elements related to urban transport were implemented, including policy analyses and the acquisition of equipment. 7. Throughout project implementation disbursements lagged behind appraisal estimates, and fell short of what the Borrower was eligible to receive. This occurred because local authorities were slow in reporting actual expenditures to the executing Ministries, resulting in a lag and eventual shortfall in disbursements from the Bank. Results 8. The project provided the vehicle for introducing new housing delivery systems to Zimbabwe for the low-income groups through production of serviced sites by local authorities and provision of mortgage finance to low-income households from local financial institutions. The main mode of construction selected by plot allottees was aided self-help housing, i.e., owner managed building with assistance of small contractors. The project demonstrated unequivocally the viability of involving local thrift institutions in the process of meeting low-income houses needs. 9. The project was successful in forging a link between public and private sector operations. It strengthened local government by shifting mortgage financing responsibilities to local building societies, thereby releasing public funds for other uses and simultaneously reducing fiduciary risks borne by local authorities, as well as extending the volume of funds available to low-income households for housing.

11 - iv - Sustainability 10. The project has proven to be sustainable as evidenced by the good mortgage loan repayment history maintained by project beneficiaries. Local authorities' efforts to maintain infrastructure services have managed to keep up with the demand created by the newly services housing suburbs. The replicability of the project approach is being demonstrated under a followup Bank-financed project, although the volume of mortgage lending has proven to be sensitive to changes in the interest rate structure of Zimbabwe's financial sector. Conclusions and Lessons L4arned 11. The main lessons learned from the project experience are: (a) (b) (c) (d) (e) In countries with a well developed financial sector, such as Zimbabwe, the ability to use World Bank loan funds to mobilize domestic resources produces pronounced economic and financial benefits. It is estimated that every dollar of Bank finance for land servicing leveraged three dollars of local finance for home construction, thereby producing a multiplier effect for the housing sector and permitting more effective use of Bank loan funds; The project provides an excellent example of a case where the traditional public sector role in housing is reduced from that of total provision to the more limited one of servicing residential land, leaving the financing and actual contraction of dwelling units to the private sector, i.e., building societies and the beneficiary households. The project demonstrates how the first steps in creating an enabling environment in the housing sector can be taken in a country with adequate local human, fiscal and physical resources; At times the Government and the Bank took opposing sides on matters of policy, as for example in the case of the most appropriate mode of construction of dwelling units. The issue was resolved by allowing beneficiaries to make the decision from a menu of approved methods proposed by both parties. The lesson illustrates that, given a choice, policy options should be made inclusive, not exclusive, and that building diversity into a policy framework can make it more robust; The Project Coordination and Monitoring Unit was not well integrated with the key line Ministries. This theoretically gave it independence to exercise its functions. However, in the Zimbabwe context with two strong and competitive lead Ministries and relatively efficient local authorities, the Unit did not receive the regard it required to exercise coordination and monitoring functions. One of the consequences was the lack of systematic monitoring needed to provide concrete information on the achievement of the Project. This supports the case of limiting the use of special project units in Bank-financed projects, if not eliminating them totally, in the interest of strengthening line Ministries; The need to expand the scope of the project because of fast implementation progress ahead of schedule suggests (in hindsight) the need for a more programmatic approach to lending in Zimbabwe's urban sector. In fact, the follow-up project (Urban Sector and Regional Development Project Loan 3079-

12 ZIM) did adopt a programmatic approach which grants the Borrower more flexibility in the design of the investments in accordance with agreed criteria; (f) (g) (h) The institutional strengthening component of the Project was not implemented in the broad comprehensive manner in which it was designed. The reluctance of the Borrower to implement the component was a function of both the hesitation to use external loan funds for 'soft" types expenditure and the concern about the ideological thrust of World Bank training efforts. There was clearly a lack of ownership of the ID component on the part of the Borrower and consequently the component was not implemented with the resources intended at the time of project appraisal; and The introduction of domestically mobilized savings into the low-income housing sector subjected the project to the cyclical vagaries of Zimbabwe's financial sector, exacerbated by the regulations governing the financial intermediaries in the country. The sustainabiity of this approach to housing sector finance will be strengthened with improvements in the regulatory framework of the financial sector. This emphasizes the need for successful long-term macroeconomic management at the national level. The implementation of the project was occasionally hampered by inadequate capacity to survey land, prepare legal documents, register titles and other administrative tasks. This emphasizes the need to strengthen administrative and legal support services in future projects.

13 PROJECT COMPLETON REPORT ZimBABVE URBAN DEVELOPMENT PROJECT WLOAN 2445-ZTM PART I. PROJECT REVIEW FROM BANK'S PERSPECTIVE 1. Project Identity Project Name Loan Number PAR Unit Country Sector Sub-sector Urban Development Project 2445-ZIM Southern Africa Department Infrastructure Operations Division Zimbabwe Infrastructure Urban 2. Background 2.1 During the period preceding the establishment of majority rule in Zimbabwe, migration into the country's cities was severely restricted. As a result Zimbabwe had a low rate of urbanization when the new Government was launched in the early 1980's, with only 23% of the population living in towns of 15,000 inhabitants or more. Based on the experience of other African countries after they achieved self- Government, the Government and the Bank foresaw a surge of new migrants into Zimbabwe's cities. 2.2 Although the urban population was less than one fourth of the population total, non-agricultural activities (largely concentrated in urban areas) produced 85% of GDP. While the new Government had established, with Bank support, programs to increase output and productivity in the rural areas and to give priority to rural development, the Government also wished to maximize the capacity of urban areas to absorb additional population. Urban growth thus would provide a means of absorbing excess people to relieve distress in the rural areas. The newly urbanized population was expected increasingly to be less affluent than in the past and to require correspondingly greater urban services. 2.3 Prior to national self-rule, Zimbabwe's cities and towns had a history of administrative independence from the central Government. They were financially sound and adhered to high standards in the provision of urban infrastructure and services for all income segments. The central Government had concentrated its work in urban areas on primary education and on some aspects of health care and was the main source of finance for low-income housing. The critical issues after self-rule thus were expected to be (i) the maintenance of municipal services in the face of rapid losses of personnel to emigration and to the local private sector and (ii) the continued provision of finance for housing for lowincome citizens in the face of many competing demands on the Government's capacity to raise capital. Prior to self rule private mortgage financing for housing was confined to high-income Borrowers and the provision of private funds for low-income housing would require expanding traditional building society lending to low-income persons.

14 Initiatives thus were conceived to assist urbanization through (i) provision of additional urban infrastructure; (ii) strengthening of governmental administration in the urban sector; and (iii) development of private finance for low-income housing. These themes were developed in detail in an Urban Sector Report for Zimbabwe (Report No ZIM) and provided the basis for the Urban Development Project. 2.5 Initial contacts between the Government and the Bank concerning this project occurred in November 1981 and the project was formally identified in March/April The project was developed and brought to the Board in the subsequent two years, with formal approval occurring on June 19, (Table 2) 3. Project Objectives and Description 3.1 The project objectives were to (a) Promote the financial and institutional capacity to supply affordable housing and related services in Zimbabwe through reform of the housing delivery system and the housing mortgage markets by - concentrating low-income housing activities of local authorities on the supply of serviced land; - introducing private sector financing for low-income housing through existing private sector financial intermediaries; - establishing new procedures for financing and construction of low-income housing; - maintaining standards for low-income housing to ensure continuity of the reformed process; and - strengthening urban management capacities of local Governments; (b) (c) Strengthen the financial and technical capacity of local Governments to provide essential urban infrastructure and services; and Support other areas of urban development (especially transportation) where new policy directions were being formulated. 3.2 The project's components as approved by the Board were: (a) (b) Provision of serviced residential land (11,349 plots in four cities) and related community facilities; Provision of loans by three local building societies (BS) to selected occupants of the lots for the construction of low-income housing. This part of the project would be financed in part by the Commonwealth Development Corporation, (CDC);

15 - 3 - (c) Institutional development, specifically for: (i) (ii) (iii) Strengthening the capacities for urban management of all local Governments in Zimbabwe; Support for setting up the Zimbabwe Association of Accounting Technicians (ZAAT); and Strengthening the Works or Community Services Departments of the four town Governments in project implementation and coordination. (d) Actions related to urban transport as follows: (i) (ii) (iii) (iv) (v) Establishment of the Urban Transport Unit (UTU) within the Ministry of Local Government and Town Planning (MLGTP); Acquisition of road safety equipment for the Zimbabwe Traffic Safety Board (ZTSB) in the Ministry of Transport (MT); Carrying out studies for MT of the equipment needed for control of motor vehicles and acquisition of such equipment including staff training in its use; Acquisition by Harare and Bulawayo city Governments of equipment to assist in traffic data collection and analysis; Construction of a central bus station in Harare; and (e) Establishment of a Project Coordination and Monitoring Unit (PCMU) in the MLGTP and of Working Committees in each of the four towns to assist in implementation of the project. 4. Project Design and Organization 4.1 Detailed discussion of the content of the project began in early The four project towns were selected and they agreed to prepare programs for housing provision for the next 3-5 years. Housing sites in each town were identified and concept plans, detailed layouts and detailed engineering were agreed to be made for these sites. The three BS undertook to mobilize domestic resources to finance mortgage loans to low-income Borrowers who would build on these sites. There would also be programs to help those owners who wished to build their houses on their own. Training needs were identified both in the municipal Government and among private contractors and detailed training programs were to be developed. Finally, the Government was to propose components to address Zimbabwe's urban transport problems. 4.2 Further preparation of the project extended over nearly 2 years to appraisal in October The project basically retained its original configuration throughout. However, finance from Commonwealth Development Corporation (CDC) was added in the form of a Z$ 10.4 million long term loan to the BS to provide seed capital for their entry into low-income mortgage financing.

16 4.3 The principal issues encountered in development of the project were the following: (a) (b) (c) (d) The size of the Bank's loan. The loan amount increased from US$25 million at project identification to US$43 million at project appraisal. The increase was intended to induce a substantial policy impact and to give recognition to the urgent need for additional housing. The role of the BS. The Zimbabwe authorities questioned the participation by the BS in the project and the extent to which their mortgage finance should supplant the existing governmental rental and ownership housing schemes. Eventually BS were made the sole vehicles for delivery of housing finance under the project. The standards for project housing. The MPCNH wanted the houses to be built under the project to have a minimum size of four rooms with a toilet/bathroom while the Bank advocated one room with a toilet/bathroom. The matter was resolved by an arrangement whereby the buyer of each plot could not occupy the plot until one room plus toilet/bathroom was completed and had to built the three additional rooms within 18 months. However, the project was targeted at low-income beneficiaries by requiring that no plot household should have an income of over Z$400 per month and that 70% of plot owners should not have incomes over ZS200 per month. The method of constructing houses. The MPCNH wanted all houses built under the sponsorship to be put up by building brigades (construction units staffed by municipal employees). The Bank, concerned about the inefficiency of public sector construction units, wanted plot owners to be able to choose whomever they wished to build their houses for them. It was agreed that plot developers could select from three approved modes of construction: (i) municipal brigades; (ii) cooperatives; or (iii) aided self-help (owner-builder).' In Masvingo houses were constructed by a local sub-contractor and then allocated to eligible households. 4.4 The project was presented to the Board in mid-1984 and was to become effective six months later. The course of implementation and disbursement foreseen at the start is shown in Table 2 and 3. The main targets envisioned for the project are presented in Table 4 and the project cost estimates and sources of finance are given in Table 5. Project results are foreseen at the start are summarized in Table 6. Project completion was to be December 31, 1990 and loan closing September 30, Procurement under the project was to be by international competitive bidding (ICB) for all contracts worth more than US$50,000 equivalent except for certain infrastructure works specifically agreed to be executed by force account. In a project financed by the CDC in Gweru (not part of the World Bank financed project), the use of council direct labor, with the local authority acting as the developer, did not prove to be an efficient and cost effective means of delivering housing, because of the difficulty of providing adequate supervision and of motivating direct labor whose jobs depended on spinning out the work.

17 4.6 Project design and organization were amended on several occasions: (a) (b) (c) In August 1985, Schedule 4 of the LA was amended to allow certain additional works by Bulawayo City Council to be done by force account. In February 1988, the Description of the Project was amended to increase the number of the serviced units from 11,350 to 23,000. This change was made when the production of the original group of units was completed earlier than scheduled and the proceeds of the Bank loan increased because of devaluation of the Zimbabwe dollar. In March 1989, the Description of the Project was amended to allow for the acquisition of buses (up to US$2,500,000 equivalent) and spares by ZUPCO (another US$2,500,000 equivalent), and Section 3.01 (b) (i) was amended to require the Government to make a loan agreement with ZUPCO for the corresponding amount. 5. Project Implementation General 5.1 Implementation occurred in two overall phases. The first phase spanned the years during which the project was implemented according to its original design. That design had a target production of 11,349 plots to be completed by the middle of 1987, with the bulk of housing construction costs on these plots to be incurred by the middle of The original design also contemplated over Z$7 million for technical assistance and training and about Z$2 million for equipment and vehicles. 5.2 The second phase began in mid-1987 when the Government reported that the production of plots (though not their electricity supply) was substantially complete, despite only 15% of the loan having been disbursed. The large discrepancy between work progress and loan disbursement was attributed to: (a) (b) (c) gains in the loan account due to exchange rate fluctuations; refusal by the Bank to finance certain expenses because of misprocurement; and lower than expected price escalation. 5.3 In view of the continuing severe housing shortage, the Government and the Bank agreed to apply the unused parts of the loan toward roughly doubling the number of plots to be produced under the project. However, no new project completion date was set nor did the parties determine a new set of project costs. Also, unlike in the first phase, the Government and the Bank did not formally agree at the outset on specific sites in each town nor on the associated off-site infrastructure and community facilities. The second phase adopted a sector-loan approach. 5.4 As it evolved, the second phase included the following major work: In Harare: Budiriro V with 3653 plots

18 In Bulawayo: Nketa IX with 2506 plots Ncema Water Supply Pipeline Engawini Substation In Mutare: Chikinga II with 2700 plots In Masvingo: Rujeko IB with 770 plots These schemes provided a total of 9629 plots. An additional major item during the second phase was the supply of buses and spares to ZUPCO, which required a separate amendment in A study of rural passenger transport was also included in the revised project design. Timing and Extent of Completion of Implementation 5.5 The project was designed so that the production of plots would be closely followed by their allocation to the owners and the subsequent construction of houses. The economics of the project was dependent on living space becoming available within a short time after the land was subdivided and provided with infrastructure. Table 4 in Part III summarizes planned and actual site development (plots available for allotment) and building construction (core houses completed) in each of the two phases. 5.6 Plot development under the first phase took only three years and was nearly in accordance with plans, while progress under the second phase proceeded more slowly. The main reasons for slower execution of the second phase were that planning and preparatory work had to be completed and that work was held up owing to lack of construction materials. The main reason for the delay was the difficulty the Building Societies had in attracting sufficient deposits for making loans owing to better return to investors in alternative saving schemes. 5.7 While the basic development of plots under the first phase was finished in 1987, the installation of electricity supply took more time. ZESA (whose participation was not envisaged under the original project) took over this function which had formerly been executed by the local Governments and it charged for installation at levels above what many lot owners were willing to pay. Also, ZESA standards disfavored provision of electricity to small houses. As a result, as of mid-1989, only about one half to three quarters of the plots under the first phase had electric supply. Some on-site infrastructure (notably street lights) and some off-site works and the community facilities also were completed after At the time of loan closing, the status of the major items included in the second phase was as follows: Budirio V 60% complete Nketa IX Completed December 1990 Chikanga II Completed December 1990 Rujeko IB Completed March 1992 Ncema Water Supply Pipeline 80% complete Rural Passenger Transport Study Not started It was agreed with GOZ that the reminder of the costs of Budirio V, Ncema Water Supply Pipeline and the Rural Passenger Transport Study would be financed under Loan 3079-ZIM.

19 The pace of implementation varied considerably among the executing agencies and particularly among the four project towns. The reasons were the differences in practices and quality of staff, effective use of technical assistance, and the inherent technical difficulties of developing some of the project sites. Where direct imports were involved (e.g. ZESA and ZUPCO), disbursements in foreign currency were made directly by the Bank. Implementation was also affected by the periodic lack of availability of municipal budgetary funds. Project towns were expected to cover all costs of the project from their own resources supplemented with funds on-lent from central Government. The towns were reluctant to finance training and consultancy expenses with use of loan funds. Some project elements were implemented only to a limited extent. Most important among these was the on-the-job training (OJT) program for local Government employees (part C1 of the Description of the Project). The primary reason for this was MLGRUD's reluctance to use loan funds for this purpose. It became apparent during project execution that because the design of this component occurred at a time when Zimbabwe was becoming familiar with the World Bank, the Government had not yet developed a policy towards using Bank loan funds for technical assistance. Once project implementation began, GOZ adopted this restrictive policy Procurement of civil works by contract generally went smoothly with awards going mainly to large firms established in Zimbabwe. Two contracts in Mutare under the first phase worth nearly Z$3 million were disallowed for bank financing because the authorities did not conform to the procurement requirement of the LA. The Government opted to award contracts to the bidder requiring the lowest amount of foreign exchange rather than to the bidder offering the lowest evaluated overall bid. Another contract in Mutare (original value of Z$8 million) for servicing of the Chikanga II subdivision under the second phase resulted in the allowance of a claim of about Z$1.8 million. This claim arose when the contractor interrupted work because the employer did not provide the foreign exchange for essential imports of equipment and supplies. Equipment, Vehicles and Spares 5.11 Whereas the loan originally envisioned a very modest amount of financing for this category of goods (US$1.5 million equivalent plus contingencies), the amount disbursed under this category by the time of loan closure was over US$20 million equivalent, all procured by ICB except for proprietary spares. The additional amount is represented principally by three items: buses and spares for ZUPCO (nearly US$5 million equivalent), electrical equipment and supplies for ZESA (about US$2.3 million equivalent) and the supply of 840 mm pipe and related equipment for the Ncema scheme. Much of the electricity supply under the original project was to have been carried out as force account by each of the towns. Its execution by ZESA provides a partial explanation of the reduction in civil works and the increase in equipment procurement. Consultant Services and Technical Assistance 5.12 This category of financing under the loan was projected to amount to US$7.5 million equivalent plus contingencies and was to cover 80% of the cost for engineering design and supervision, transport technical assistance and training, local Government training and other institutional development technical assistance. The final amount used under the loan for these purposes was less than US$1 million equivalent. This is explained by greater reliance on town staff for design and supervision, smaller implementation of the urban transport studies effort, and major reductions in institutional development training and consultancies. There also was the use of bilaterally financed technical assistance in substitution of what had been intended for loan financing. The ultimate impact of this was a lower

20 capacity at all levels of the government to implement similar projects than might have been achieved with the fuller use of project resources for technical assistance and training. Establishment Costs of ZAAT 5.13 This category of the list of goods was budgeted for US$170,000 equivalent to cover 80% equivalent expenses of ZAAT during 1984 and 1985 and was to be disbursed against statements of expenditures. The category was expanded to cover establishment costs as well of PCMU, MT and ZTSB and by loan closure, had absorbed about US$700,000 equivalent, more than half to PCMU (with disbursements continuing through 1993) and about US$100,000 (Z$247,000) equivalent for ZAAT. Urban Transport and Traffic Management 5.14 An Urban Transport Unit (UTU) was established in MLGRUD in the first year of project implementation and was provided with foreign advisors. The UTU prepared a series of technical papers on transport policy issues and road safety. It also assisted the Government in acquiring a majority share of ZUPCO. UTU staff were trained overseas under the project The Zimbabwe Traffic Safety Board (ZTSB) benefitted in the early years of project implementation from loan financing through contributions to establishment costs, for some equipment acquisition and for consultancy services. Also four senior staff of ZTSB were sent to the Eastern and Southern Africa Management Institute for training. A computerized traffic accident analysis package was acquired in Harare Central Bus Terminal was completed in Housing Finance and Construction 5.16 The provision of mortgage loans by the BS was part of the Definition of the Project and the allocation of plots to eligible beneficiaries and the provision of finance and construction of housing was the major thrust of the project. Allocation procedures followed the existing rules of the four towns using housing waiting lists which required local residence and employment. These rules were modified in the LA by reducing the qualifying household income from not more than Z$450 to Z$400 per month and requiring that 70% of the allottees should have incomes under Z$200. The limits were related to Zimbabwe's poverty level. The income limits were to be reviewed annually by the Borrower, the Bank and CDC but, while some discussions took place on this topic in the early years of implementation, the limits were never formally amended. Ninety percent of all beneficiary households allocated plots in the first phase of Harare's program earned less than Z$400 per month. Similarly, 80%, 86% and 67% of households in Bulawayo, Mutare and Masvingo respectively earned less than Z$400 per month. Household incomes among allottees in the second phase of each town's program were slightly higher: e.g. in Bulawayo 28% of allottees earned less than Z$400 but all earned less than Z$750. No towns achieved the lowest income target of 70% of plots to households earning less than Z$200 per month. In the first phase it was 1% in Harare in 1993, 2% in Bulawayo in 1991, 6% in Mutare in 1993, 9%in Masvingo in Bulawayo's second phase had I % of allottees with incomes under Z$200 in 1993, in retrospect, possibly an unrealistic target. This finding is common among site and service schemes in the Africa region: these programs rarely meet the homeownership needs of the bottom quintile of the income distribution. However, the poorest fifth of the urban population were the major beneficiaries of the rental opportunities created under the project which permitted partial subletting of owner-occupied dwellings.

21 The provision of mortgages for house construction was essential to most, though not all, buyers of plots under the first phase of the project; a small number, varying in different communities and ranging to nearly 20% in some, financed their houses on their own. The very low number of beneficiaries earning under Z$200 may have been due to the requirement that applicants be salaried workers in the formal sector In all towns but Masvingo, construction of houses was done by small contractors engaged by the owner, with some technical help from the BS and the towns; in Masvingo, which lacked a developed construction sector, the town authorities themselves engaged contractors to put up groups of houses on behalf of the owners. Once construction started, houses were completed quickly (typically within six months) and construction was mostly brought to the full house size right away, with core houses constituting only 20% of the units built. As indicated in para 5.6, house construction nevertheless continued to lag considerably behind the number of available plots, primarily due to the lack of mortgage credit. 6. Results Achievement of Project Objectives 6.1 The project succeeded in introducing Zimbabwe to a new system for the production of low income housing. Features of this system included the owner/builder concept and the provision of mortgage finance through domestic thrift institutions. Several variations on this theme were developed. Bulawayo, for example, provided a majority of its new housing though the construction of shell houses which are sold under tenant purchase arrangements. These houses of 50 m 2 have four rooms and an ablution unit but no doors, windows, electric wiring or other fittings which the purchaser adds gradually as he can afford them. As project elements which were directed at improvement of urban management were implemented to only a minor extent, the project had a correspondingly limited impact in this area. The project made limited contributions toward policy formulation for urban transport and toward improving urban transport operations. The project contributed to the transfer of a controlling interest in ZUPCO to the Government, to reequipping ZUPCO and to improving its schedule. These actions helped to improve ZUPCO's operations but its legal status remained to be sorted out. Finally, the project resulted in the introduction of systems for improved traffic safety, motor vehicle control and traffic management. Achievement of Physical Targets 6.2 The physical targets for the production of sites were largely achieved under the first phase but, due to slower production achievements under the second phase, the output of sites for the project as a whole fell about 20% under target. Most on-site and off-site infrastructure was built, though some of it behind schedule. Electric supply infrastructure remained low in relation to the completed units. As already suggested, the volume of consultancy and technical assistance was much below projections while equipment acquisition much above it. See Table 4 in Part IHI.

22 Project Sustainability 7.1 Sustainability issues in this project are raised in two contexts: the operation of the low-income housing production system introduced under the project and other governmental policies and procedures revised under the project. The new housing production system was introduced with limited outside help and this operation is sustainable with staff available in Zimbabwe, supplemented by modest external technical support. The continuation of this system under a follow-on Bank project testifies to the replicability of the systems. However, the mobilization of domestic capital to finance these systems has become a problem and threatens their continued application. To increase the ability of BS to attract savings requires a change in macro-economic policy, a matter beyond the confines of the project itself. There is also a distributional constraint, i.e., at any given time building societies will find it more profitable to lend scarce funds to more remunerative, less risky clientele than the low income residential market. The mid to late 1980s may have been a unique period when all factors came together, not the least that the building societies had ample liquidity. There is no guarantee that this will happen again even if real and nominal interest rates fall significantly. 7.2 The other governmental policies and procedures which were revised under the project were of more limited scope and their continued viability is therefore more difficult to establish. The UTU was merged into the regular governmental system and its know-how was thus permanently incorporated. Improved motor vehicle and traffic safety procedures were closely tied to computerized procedures which are themselves easy to sustain but which may become technologically outdated after some years. The OJT training systems were virtually not implemented and their sustainability thus does not arise. The future operations of ZAAT are in doubt because it is not able to raise sufficient income to meet operating expenses, much less service the debt incurred under the Project. 7.3 GOZ's interest in seeking alternative to hard Bank money for capacity building elements is understandable. But in actual fact, nothing was put in its place, the legacy of which has impacted the implementation of the follow-on project, Urban II. Neither PCMU nor the eligible local authorities (ELAs) themselves recognized early enough the needs for outside technical assistance and training. ELAs that have consultants have not managed them particularly well. ELAs that have not and have weak internal capacity have been seriously handicapped in their ability to benefit from Urban II. 8. Bank's Performance 8.1 The Bank allocated significant staff resources to prepare the project, which was one of the early loans in a new borrowing country, based on a system novel to that country. The Bank prudently limited the complexity of the project by concentrating it on three main concerns: production of housing, strengthening local Government institutions related to housing, and developing aspects of urban (and national) transport policy and administration. An extraneous element relating to job creation was dropped. 8.2 The Bank ensured that preparation of the works (designs and cost estimates) was made before loan approval and arranged for effective coordinating mechanisms at both the central and town Government levels. The rapid and timely execution of most works in the original project attests to the value of these preparations. By contrast, the weaker preparatory efforts related to the second phase partly explain its more troubled but still successful implementation.

23 - I1-8.3 The project also was materially helped by long-term continuity of senior Bank staff assigned to preparation and appraisal. This continuity continued despite several organizational changes including the 1987 reorganization. The Bank's supervision effort was adequate, with missions fielded over ten years. The missions were appropriately staffed, technically strong and provided good staff continuity. 9. Borrower's Performance 9.1 No single characterization of Borrower performance can be made with the large number of Borrower organizations which participated in the project and which ranged from central to local Government, from public to private institutions and from technical to financial and policy groups. There were significant unresolved tensions among them touching on such vital project issues as the use of external loans for housing, the participation of private sector organizations in low-income housing and the use of non-governmental technical forces to assist in project design and execution. The resolution of such conflicts was the primary concern of the Borrower though the Bank had the opportunity to influence this process and did to some extent. Coordinating mechanisms, particularly towns' coordinating committees, maintained interest and organizational discipline, typically under the active leadership of each town's public works authorities. These authorities were accustomed to executing works which was the central feature of the project and they applied their skills effectively to this end. Similarly, the BS were able effectively to adopt their normal loan processing to the requirements of low-income Borrowers. 9.2 The performance of Borrower entities who had to deal institutional elements of the project was more uneven. The OJT element of the project was hardly implemented at all and other institutional initiatives were only partly put into effect. Policy formulation regarding housing and urban transport was difficult to achieve. The Steering Committee, which was to review and monitor project execution, did not meet very often and exerted only limited influence. Finally, PCMU, rarely fully staffed, has had problems in gaining full cooperation of various agencies in providing adequate reporting on the project and in generating the documentation to support disbursements. The weakness of PCMU continued to have a negative impact on the start-up of the follow-on investment. 9.3 The Borrower's performance with regard to covenants is summarized in Table 7. There were significant shortfalls in this performance particularly with regard to the employment of consultants (few were used) and to the enforcement of the income guidelines for property ownership. Finally, financial reports were usually submitted substantially later than the loan specified. 10. Relations Between the Bank and the Borrower 10.1 Relations between the Bank and the Zimbabwe authorities were open and frank. The few serious disagreements between the parties took time to resolve (for example, the minimum size of standard houses). However, by Project completion there was general agreement on the appropriate principles for an effective low-income housing policy framework. 11. Consultants Services 11.1 The project provided for a large volume of consulting services (initially over the equivalent in US$ of 7.5 million) but only a small fraction of it was used. Primarily this was due to the fact that the

24 Borrower was able to persuade the Bank that it did not need the consultants listed in Schedule 6 of the LA. The largest single consultancy which was not implemented was the one relating to OJT. Also, no consultants were appointed to assist in the formulation and monitoring of housing programs which affected the information available on trends in the housing sector. The towns made much more limited use of consultants in the design and supervision of the works programs than was foreseen at appraisal. As a result, documentation on project implementation was somewhat limited and, in the second phase, design work was delayed Consultancy assigmnents which were carried out under the project were generally executed competently. Most consultants were Zimbabwe firms who worked on aspects of civil engineering. Foreign firms were employed to work on transport, road safety equipment and on motor vehicle administration. 12. Project Documentation and Data 12.1 The original documentation for the project was adequate, providing a useful framework for both Bank staff and the Borrower during implementation. Staffing delays in the PCMU and frequent turnover of key personnel contributed to incomplete reporting of project progress and achievements. 13. The main lessons learned from the project experience are: (a) (b) (c) (d) In countries with a well developed financial sector, such as Zimbabwe, the ability to use World Bank loan funds to mobilize domestic resources produces pronounced economic and financial benefits. It is estimated that every dollar of Bank finance for site servicing leveraged three dollars of local finance for home construction, thereby producing a multiplier effect for the housing sector and permitting more effective use of Bank loan funds; The project provides an excellent example of a case where the traditional public sector role in housing sector is reduced from that of total provision to the more limited one of servicing residential land, leaving the financing and actual contraction of dwelling units to the private sector, i.e., building societies and the beneficiary households. The project demonstrates how the first steps in creating an enabling environment in the housing sector can be taken in a country with adequate local human, fiscal and physical resources; At times the Government and the Bank took opposing sides on matters of policy, as for example in the case of the most appropriate mode of construction of dwelling units. The issue was resolved by allowing beneficiaries to make the decision from a menu of approved methods proposed by both parties. The lesson illustrate that, given a choice, policy options should be made inclusive, not exclusive, and that building diversity into a policy framework can make it more robust; The Project Coordination and Monitoring Unit was not well integrated with the key line Ministries. This theoretically gave it independence to exercise its functions. However, in the Zimbabwe context with two strong and competitive lead Ministries and relatively efficient local authorities, the Unit did not receive the regard it required to exercise

25 coordination and monitoring functions. One of the consequences was the lack of systematic monitoring needed to provide concrete information on the achievement of the Project. This supports the case of limiting the use of special project units in Bankfinanced projects, if not eliminating them totally in the interest of strengthening line Ministries; (e) (f) (g) (h) The need to expand the scope of the project because of fast implementation progress ahead of schedule suggests (in hindsight) the need for a more programmatic approach to lending in Zimbabwe's urban sector. In fact, the follow-up project did adopt a programmatic approach which grants the Borrower more flexibility in the design of the investments in accordance with agreed criteria; and The institutional strengthening component of the Project was not implemented in the broad comprehensive manner in which it was designed. The reluctance of the Borrower to implement the component was a function of both the hesitation to use external loan funds for 'soft' types expenditure and the concern about the ideological thrust of World Bank training efforts. There was clearly a lack of ownership of the ID component on the part of the Borrower and consequently the component was not implemented with the resources intended at the time of project appraisal. The introduction of domestically mobilized savings into the low-income housing sector subjected the project to the cyclical vagaries of Zimbabwe's financial sector, exacerbated by the regulations governing the financial intermediaries in the country. The sustainability of this approach to housing sector finance will be strengthened with improvements in the regulatory framework of the financial sector. This emphasizes the need for successful long-term macroeconomic management at the national level. The implementation of the project was occasionally hampered by inadequate capacity to survey land, prepare legal documents, register titles and other administrative tasks. This emphasizes the need to strengthen administrative and legal support services in future projects.