Tanzania - National Sites and Services Project

Document of The World Bank FOR OMCIAL USE ONLY Report No. 4941 COMPLETION REPORT TANZANIA - FIRST NATIONAL SITES AND SERVICES PROJECT (CREDIT 495-TA) February 15, 1984 Eastern Africa Regional Office This document has a restricted distribution nd may be used by recipients only in the performace of their official duties. Its contents may not otherwise be disclosed without World Bank authorization.  FOR OMCIAL USE ONLY TANZANIA: FIRST SITE AND SERVICES PROJECT PROJECT COMPLETION REPORT ABBREVIATIONS AND ACRONYMS Ardhi Ministry of Lands, Housing and Urban Development BRU Building Research Unit CIDA Canadian International Development Agency LRSC Land Rent and Service Charge PMO Prime Minister's Office ELDO Regional Land Development Office RDD Regional Development Directors TANESCO Tanzania Electric Supply Company THEB Tanzania Housing Bank UNDP United Nations Development Program This document has a resaicted distibution and may be used by repients only in the performance of their offcial duties. Its cDntents may not otherwise be discosed without World Bank authorization.  TANZANIA: FIRST SITE AND SERVICES PROJECT PROJECT COMPLETION REPORT TABLE OF CONTENTS Page No. PREFACE ............................... (i) BASIC DATA SHEET ....................... ....................... (ii) HIGHLIGHTS (S-mary) ..................................... .................. (viii) I. INTRODUCTION ................................ 1 A. Sector Background ..... ................. I B. The Project ........................................... 2 II. PROJECT IDENTIFICATION, PREPARATION AND APPRAISAL ......... 4 A. Identification ....... 4 B. Preparation ...... 4 C. Appraisal ..... 4 D. Conclusions ...... 5 III. IMPLEMENTATION .................... 6 A. Effectiveness and Start-Up ........ 6 B. Progress of Physical Works ...... 6 C. Progress of Non-Physical Program ..... ... 11 D. Project Costs ........ 13 E. Procurement ....... 15 F-. Disbursements -o..................oooo. 15 G. Unit Costs .... ................. 18 H. Conclusions .. ..... 18 IV. OPERATING PERFORMANCE ....... 19 A. Role and Performance of Consultants .... ... 19 B. Performance of Contractors ......................* ... 19 C. Performance of Standards Used ........ 20 D. Compliance-with Covenants .. . .... 22 E. Financial Performance and Project Replicability ....... 23 F. Conclusions ... o.................................... o 25 V. PROJECT BENEFITS ..................... o ................. 26 A. Beneficiaries ........... .............. -. .. ..... .... 26 B. Economic Reevaluation ................... oo.o .......... 27 C. Conclusions .......................................... 28  Page No. VI. INSTITUTIONAL PERFORMANCE AND DEVELOPMENT ................... 29 A. The Ministry of Lands, Rousing and Urban Development..... 29 B. The Tanzanian Housing Bank .............................. 30 C. Coordination ........................................... 31 D. Conclusions ............................................... 32 VII. ROLE AND PERFORMANCE OF TEE BANK ........................... 34 VIII. LESSONS LEARNED .................... ......................... 35 A. Civil Works ................................................ 35 B. Cost Recovery ............................................ 35 C. The Community Development Input .... ..oo..eo ....o35 D. The Institutional Framework ..... ........................ 36 E. Institution Building ........... ......................... 36 F. The Macro-Economic Framework . .......................... 36 Annex: Borrower Comments ........................................ 37 Photographs ..................................................... 42 MAPS IBRD 12572 (PCR) IBRD 12573 (PCR) This report was prepared by Ms. Faye Johnson (EAPWU), with assistance from Ms. Donna Haldane.  -1- PROJECT COMPLETION REPORT TANZANIA: FIRST NATIONAL SITES AND SERVICES PROJECT (Credit 495-TA) PREFACE Credit 495-TA for the First National Sites and Services Project was signed in July, 1974. The Credit, for US$8.5 million, was fully disbursed in October, 1981. During implementation, the project's content was revised and the closing date was postponed. This Project Completion Report (PCR) was prepared by the Water and Urban Division of the Eastern Africa Projects Department. It is based on a review of files and documents, interviews with Bank and Borrower staff and a visit to Tanzania in September, 1981. In accordance with the revised procedures for the project performance audit reporting, this Completion Report was read by the Operations Evaluation Department (OED) but the project was not audited by OED staff. The draft Completion Report was sent to the Borrower and its agencies. Their comments have been reflected in the PCR and are reproduced as an annex.  -ii- PROJECT COMPLETION REPORT TANZANIA: First National Sites and Services Project BASIC DATA SHEET KEY PROJECT DATA Actual as Actual or % of Appraisal Estimated Appraisal Estimate Actual Estimate Project Costs (US$million) 14.5 1/ 15.2 106 2/ Credit Amount (US$ million) 8.5 8.5 100 Date Board Approval n.a. 7/2/74 n.a. Date Effectiveness n.a. 10/03/74 n.a. Date Physical Components Completed 6/30/77 9/30/80 3 years delayed Proportion then completed (%) - 98 98 Closing Date 12/31/78 12/31/80 2 years delayed Institutional Performance Satisfac- Less than tory satisfac- tory. Economic Rate of Return (%) 10.4% to 16% to 33% n.a. 14.5% Beneficiaries 194,000 231,180 119 1/ The total project cost in the Staff Appraisal Report (SAR) differed from the amount implied by the Credit amount and disbursement percentages. The total project cost has therefore been revised downwards from 16.7 US$ million equivalent to 14.5 US$million equivalent. The Tanzanian Housing Bank's (THB) home loan disbursements were erroneously stated as being TSh.45,000,000 in the SAR. According to thc Development Credit Agreement, US$ equivalent of 2,000,000 was provided for THB loans. The Association was to finance 50% of the amount disbursed, hence the TSh. amount that was to have been disbursed by the THB should have been TSh.28,560,000 (US$4,000,000 equivalent). 2, The project scope and content were changed during implementation. Although the total amount actually spent on the project increased by only 6%, the cost per serviced s-te increased 15% (from TSh.3,020 to TSh.3,480). The number of units was therefore trimmed back from 10,600 planned to 8,500 actual. Upgrading costs per unit, on the other hand, came in 15% below that expected (from TSh.1,810 per house to only TSH.1,550) due to there being more houses than expected in the upgraded areas. Fully 14,600 houses benefitted, compared to 8,800 estimated at appraisal.  -iii- CUHMLATIVE DISBURSEKMTS FY75 FY76 FY77 FY78 FY79 FY80 FY81 FY82 Appraisal Estimate (US$.illion) 1.6 4.1 7.3 8.5 Actual (US$million) .9 2.2 4.2 5.2 5.9 6.4 8.2 8.5 Actual as Z of estimate 57 54 57 61 69 76 96 100 Date of final disbursement 10/23/81  -iv- PROJECT COMPLETION REPORT TANZANIA: First National Sites and Services Project BASIC DATA SHEET OTHER PROJECT DATA Borrower: Government of Tanzania Executing Agency: Ministry of Lands, Housing and Urban Development Fiscal Year: July 1 to June 30 Name of Currency (abbreviation): Tanzanian Shillings (TSh) Currency Exchange Rate: Appraisal Year Average: US$1 - 7.13 Intervening Years Average: US$1 - 8.03 Completion Year Average: US$1 - 8.29 Follow-on Project: Name: Tanzania: The Second National Sites and Services Project Credit Number: 732TA Credit Amount (US$million): 12.Om Date Board Approval: 12/3/77  -rI PiwEa m Ear TANZANA First National Sites ad Service Project MN IIDICAMRS &tual as Z Appraisal of Appaisal Estimate Actual Estiate I. EESIDENI TAL UNITS SEWICED Newly-serviced sites (sites & services): ar es Sualam 7,450 6,338 1/ 85 nza 2,300 1528Z/ 66 Mteya 85D 618 / 73 Subtotal 10,600 8,484 80 Upgraded Areas4 Dar eS Salasm 7,600 12,225 4/ 161 Thya 1,200 2:409 V/ 20D Subtotal 8,800 14,634 166 Total no. of rsidential units serviced: 19,400 23,118 119 II. NEW RES 1DI UNllS ORE DJ 10,600 4,106 / 39 IaI. CMl4lY FACTT:mm CONl DC Cmmity 9dtion Cters Dar es Salam 9 7 6u 78 MMa 1 1 100 Mteya 1 1 100 Urban Health Ceters DaresSalas 1 1 100 Dispefsaries Dar es SaLsm 5 4 6/ 80 Mherza 1 1 100 Ibeya 1 1 100 Large Markets Dar esSalami 2 2 100 hernza 1 - 6/ 0  -vi- Actual as Z Appraisal of Appisal Estimte Actual Estimte MLoor Mrakets Dar esSalaa 6 5 6/ 83 amza I 1 100 Mbeya 1 1 100 Total ruber of comai1ty facilities 30 25 83 IV. TAINI OF ECHNICAL PRSONNEL NS V. WaJSE IOANS Small Construction Loans 7,400 1,313 18 Improvement Ynm 4,400 1,50D 34 TnroL tLa,80 7/ 2,813 8/ 24 _/ Total as of 3/83. Eigrty-thwee plots in Sina and 20 plots in Kijitnyama were xt developed due to the alignnt of a wter mn tely 150 plots in Sinza and 200 plots In Kijitoyama were oDt developed because of unforeseen tecimical proble. The development of 881 plots in M[koche was deferred in order to avoid cost overruns. 2/ The development of 498 plots ms deferred to accomondate cost overrus. S/ Seven plots found to be unsuitable for developmeat; secticx of 243 plots squatted upon and 4 43 intermingled with squatters. 4/ Population increased as a result of densification during the iuplemtation period. 31 To habitable stage. Data for Dar es Salaam is as of 2/82 4threas data for Nxbya and Hmza is as of 8/81. Between 8/81 and 2/82, the nuber of units occupIed In Dar es Salaam increased by 38% hence figure for Nbeya and MEmnza (345) is Likely to be below the actual. The data Is being upiated by the Towa Cauncils. 6/ IS conmwnty education centers, one dispensary, one large maiket and one minor micket deleted from comnmity facllities component to avoid coet overnmms. 7/ IHB's house loan disbursements were erroneously stated as being TSh.45,00O,000 ILn the Staff Appraisal Report. Accordlg to the Dyeopmmat Credit Agremaent, US$ equivaeLmt 2,000,000 uss provided for THB house loans. The Aswciation was to finance 50Z of the am=unt disbursed, hence the TSh. amonit that ms to have bemi disbursed by the THB should have been TSh.28,560,000 (US$4,000,000 equidvalet). The fwids provided for the THB in the project therefore 'as only enaough to cover about 60% of the mwbier of loms indicated above as the appraisal estimate. 8, As of 12/31/81. T coatime to be granted frm funds from the second project to plot beneficiaries uWder the first.  PROJECT COMPLETION REPORT TANZANIA: First National Sites and Services Project Mission Data Date Number Person days Specializations Performance Type of Mission Mb/Yr/ Persons in field Representedl/ Rating2/ Trend3/ Problem4/ Reconnaissance 2/71 3 6 agj - - - Preparation 11/72 2 4 bf - - - Preparation 2/73 1 18 b Appraisal 8/73 4 22 ebfj Subtotal 50 SPN 1 6/74 1 2 b t 2 10/74 2 14 bc 2 1 fm 3 2/75 2 14 bd 2 1 fm 4 7/75 2 25 bd 1 1 m 5* 10/75 1 19 bd 1 1 mf 6* 3/76 1 17 d 1 1 mf 7* 7/76 3 57 aad 2 2 pmf 8* 9/76 2 36 da 2 1 pef 9* 6177 1 5 d 2 1 uf 10* 9/77 2 10 da 2 1 PM 11* 1/78 2 12 ad 2 1 pf 12* 5/78 3 14 di 2 2 pm 13* 10/78 3 21 acd 2 2 PM 14* 3/79 4 52 afcd 2 3 pfm 15* 7/79 4 36 fcdj 2 3 pm 16* 10/79 2 22 af 2 3 mpf 17* 2/80 2 28 ff 2 2 pfm 18* 6/80 3 42 fej 3 1 mf 19* 11/80 1 16 fi 3 2 mf 20* 3/81 2 54 fi 2 1 f 21* 9/81 2 36 fe 2 1 mtf Subtotal 532 :7 -77 person days in field p.a. TOTAL 582 t/ a - Economics; b - Engineering; c - Hanagement; d - Architecture/planning; e - City and regional planning; f - Urban affairs; g - Training; h - Legal; i - Finance; j - Departmental management. 2/ 1 - Problem free or minor problems; 2 - Moderate problems; 3 - Major problems. 1/ - ' Improving; 2 - Stationary; 3 - Deteriorating. t/ f - Financial; m = Managerial; t - Technical; P - Political. In addition to the supervision Inputs outlined above, the Bank sent out Housing, Land Financing, Economic and ManAgement consultants to assist with the resolution of problems, to design and put various systems into operation and to conduct training seminars. * Joint with identification, preparation and supervision of second project.  -viii- PROJECT COMPLETION REPORT TANZANIA: First National Sites and Services Project HIGHLIGHTS The first National Sites and Services Project was one of the first such projects to be funded by the Association and represented the first demonstration phase of Tanzania's program to deliver housing and services to the country's low-income majority. The project included the development of newly serviced plots in three urban centers; improvement to basic infrastructure and additional services to existing squatter settle- ments in two of the three urban centers; community facilities; consultants' services for the design and detailed engineering of basic infrastructure, project supervision and the preparation of a follow-up project; expansion of physical facilities for a town planning training program; monitoring and evaluation; a pilot nutrition project; training for technical staff of the Ministry of Lands, Housing and Urban Development (Ardhi); financing for house construction and improvement loans; and equipment and vehicles. The implementation period proved to be a very difficult one as the economy deteriorated following the increase in international prices in 1974, the collapse of the East African Community in early 1977 and the war with Uganda in 1978-79. There were widespread shortages; delays in the import of equipment and spare parts and rapid inflation. The environment was a very difficult one to operate within. The Government also changed the institutional structure. Urban councils were recreated in 1978, after having being abolished in 1972. The difficulties in the institutional and economic environment and the newness of the approach resulted in a two-year delay in completion. The project as completed was alro somewhat changed in scope and content. It is perhaps noteworthy and not atypical of first generation urban projects that the civil works or physical component went more smooth- ly than the 'software' elements. Greater difficulties were encountered in developing an adequate community development support network, initiating cost recovery, and in sustaining an adequate level of infrastructure maintenance. Cost recovery was perhaps the primary issue of project implemen- tation. The mid-stream shift after appriisal from one vehicle (direct plot charges) to another (property taxes) inevitably involved a second round of political approvals. Once secured, it took much longer than expected to complete the subsequent mechanics of house registration, tax roll preparation, and the hiring and training of collections staff. Thus a three year lag occurred between completion of services in upgraded areas and start-up of cost recovery.  -ix- A principal objective of the project was the establishment of a cadre of professionals capable of implementing a program based on modest standards and a self-help approach to shelter and services. The project achieved this. It did not, however, develop a functional shelter delivery system. Upgrading was successfuly done by one agency-Ardhi. Serviced sites, however, require the cooperative (-systemic") efforts of several institutioas for completion and fragmented responsibilities made it difficult to pinpoint and resolve bottlenecks. Low civil service salaries and high staff turnover were also endemic problems in Tanzania. The resulting morale problems were exacer- bated by a rising general level of corruption which parallelled the country's economic deterioration. All this made the task of project management exceptionally difficult-even had the institutional structure been more workable from the start. The Tanzania Housing Bank (THB) proved a particularly disap- pointing intermediary. The THEB suffered generaliy, like many Tanzanian in- stitutions, from too-rapid expansion and weak financial management. Staff turnover was unusually high, loan procedures remained cumbersome, outreach was weak, reporting unsatisfactory, and financial performance poor. Interest rates were quite negative in real terms and discouraged lending to the poor while contributing to the agency's decapitalization. THEB did, however, successfully initiate lending to lowest income groups. Fully 5OZ of all loans made by 1978 were to families below the established TSh. 750 monthly income ceiling. Upgrading of squatter dwellings was initiated for the first time in the Housing Bank's history-no small accomplishment in itself. Such policy breakthroughs are not easily achieved. Treasury's performance in collecting charges for serviced plots and particularly upgrading areas has been a bright spot of institutional competency. The task has been successfully managed from its outset in 1980 and the agexcy seems somehow to have overcome the pervasive headaches created by overall country conditions. Collection administration startup costs were about 10% of total collections . In sum, an important start has been made on institutional development. Greater accountability is needed in the shelter delivery side and a longer time frame than one project is necessary to achieve significant progress. Implementation capacity remains thinly stretched and lending operations must be carefully matched to this capacity. Greater attention is needed in future project preparation and appraisal to the institutional aspect.  Over 200,000 people benefited from the project-a substantial increase over that estimated at appraisal-85Z of whom lived in upgraded areas. Based on THB's figures, perhaps half the serviced site recipients were within the targetted population of TShs.750 and below. Thus, while there was some poaching of plots by higher income groups, over 70X of proj- ect beneficiaries appear to be within the targetted population. The proj- ect did therefore reach the poor. Health and educational levels in squatter areas certainly improved. Though this, of course, is hard to quantify. Rates of return offer more quantifiable benefits, and actually exceeded expectations at appraisal. Serviced sites show a return of between 16-20X--versus an estimated 10.4%, and upgrading fully 33Z-over double the expected 14.5%. These figures largely reflect the increasing scarcity value of shelter in Tanzania. It is our conclusion that the project achieved its basic demon- stration impact and successfully helped Tanzania translate policy into action. Upgrading was more successful than serviced plots. Weaknesses in the institutional delivery system is the key problem in plot development, specifically in plot allocation, house loans, and community development followup. Cost recovery, though hard-won, has proven workable. It is perhaps surprising that achievement of the project's basic objectives did not founder on the institutional weaknesses and increasing adversity of country conditions. The changes made during implementation, however, did not compromise the principal goals of demonstrating a workable approach to providing houses and services to the urban poor and creating some capacity to undertake similar projects in the future. The first project has made an important beginning. Perhaps one of its major achievements has been as a diagnostic tool giving us a practical working knowledge of the Tanzanian urban environment, and helping pinpoint some of the underlying problems needing attention in subsequent phases of urban lending. Aspects of the project implementation experience that may be of special interest are: - the difficulties and time needed to have changes to regulations and established procedures effected (PCR, paras. 2.09, 3.10, 3.12, 3.13, 4.17, 4.19, 4.20, 4.24-4.27); - the importance of a programmatic approach extending beyond the project period (PCR, para. 3.19, 4.02, 4.19, 4.20, 6.16, 8.03); - the need for clear assignment of accountability (PCR, paras. 6.12, 8.06).  TANZANIA: FIRST SITE AND SERVICES PROJECT PROJECT COMPLETION REPORT I. INTRODUCTION A. Sector Background 1.01 At the time of project identification in 1972 although the absolute number of urban dwellers in Tanzania was relatively small, about 900,000 or approximately 6.5Z of a total population of 14.0 million, the rate of urbanization was steadily climbing. Between 1957 and 1967, the ten principal regional centers grew at an average annual rate of 6.5%, which was nearly two and one-half times the national rate. It was foreseen that, if the existing trends continued, Tanzania's urban population would double in 11 years and that Dar es Salaam would reach 1.0 million by 1983. A by-product of these high rates of urbanization was a severe nationwide housing shortage, particularly among the nearly 80% of the urban households in Dar es Salaam then earning less than TSh.500 per month. 1.02 In 1969, over 40% of the housing containing an estimated 65Z of Dar es Salaam's population was located in squatter areas. Municipal services were practically non existent due to insufficient funds and equipment and a shortage of trained manpower. Over 80% of the households were tenants, of which approximately 90% lived in single rooms. Only about 25% of Dar es Salaam's households had piped water to their plots, another 38% had access to water from kiosks and the rest (37%) had no access to piped water. Nearly 73% of the houses were without electricity. The situation in the other urban centers, although not as well quantified, was similar. 1.03 In order to cope with the abovementioned problems the Government of Tanzania had adopted a decentralization policy to divert growth away from Dar es Salaam into the other urban centers and rural areas. The decentralization policy involved (a) the abolition of urban councils, (b) the relocation of the government machinery from Dar es Salaam to regional administrative centers-over 40% of the establishment had recently been transferred; (c) the development of Ujamaa villages;l! and (d) the promotion of alternative urban growth centers--Arusha, Moshi, Tanga, Dodoma, Tabora, Mwanza, Mtwara, Nbeya and Morogoro had been so designated. 1.04 As of 1972, it also adopted the sites and services and squatter upgrading approach to housing. A spec-'al parastatal of the Treasury, the Tanzania Housing Bank (THB) had been established in the same year to support the Government's new housing initiatives. The THB was to mobilize savings and external resources for housing development and to promote housing development by making technical and financial assistance available for sites and services and squatter upgrading projects, housing cooperatives, Ujamaa villages, and other owner-occupied housing schemes under the proposed program. I/ Societal units consisting of groups of families who farm the land collectively and share in the proceeds according to the amount of work contributed.  1.05 The second five-year plan (1970-1975) called for 5,000 serviced plots to be provided, but in 1972 only 600 plots had been provided. The initial idea of low standards had become modified during implementation and the focus had shifted from the original idea of services only to the development of completed housing units. Following a mid-plan review, the officials in Tanzania felt that a good demonstration project was needed to establish the approach and standards for the program. Tanzania lacked the funds for their program and therefore approached IDA for assistance in October 1972. B. The Project 1.06 The project was to be implemented over a three-year period, and was to be the first phase in the Government's nationwide low-cost housing program. As appraised, it consisted of: (a) Provision of about 10,600 serviced sites for low income housing located as follows: (i) nearly 7,450 plots in three sites in Dar es Salaam, (ii) about 2,300 plots in Mwanza, and (iii) about 850 plots in Mbeya. Criteria used for the selection of project sites included the availability of a reasonably large parcel of land in conformity with the master plan; a good location in relation to job opportunities, transportation and off-site infrastructure; and, good natural drainage. Due to the magnitude of the problem there, about 70% of the plots were to be in Dar es Salaam. The typical plot size was to be 288m2, which although large compared to previous Bank/IDA projects, was appropriate in the Tanzania context where multi-family occupancy and gardening are standard. (b) Upgrading of existing squatter settlements accommodating nearly 9,000 families in Dar es Salaam and Mbeya. Simple layouts were emphasized for economic construction and servicing and minimizing compensation payments. The size of the project and the number of regional centers included were determined by the estimated capacity of Ardhi, THB and the regional administrative offices; (c) Provision of eleven community education centers, one health center, and eleven markets; (d) Financing for low income housing loans through the THB; (e) Consultant services for the design and detailed engineering of basic infrastructure and community facilities, preparation of a follow-up project and overall project supervision; (f) A training program for technical staff of the Ministry of Lands, Housing and Urban Development (Ardhi), to be carried out by consultants; (g) Expansion of the physical facilities of Ardhi Institute for its town planning program;  - 3 - (h) A pilot nutrition project to improve the nutritional awareness of households in sites and services areas; and (i) Equipment and vehicles. 1.07 Total costs were estimated to be approximately TSh.119.6m (USS16.7m equivalent). The IDA Credit of US$8.5 million was to finance the foreign exchange costs and 28% of local costs. The principal objectives of the project were to (i) introduce and demonstrate an approach to housing and servicing the urban poor; (ii) build up the institutional capacity for implementing similar projects throughout the country; (iii) open up access to credit for low income families to build low cost housing; and, (iv) increase the housing stock and service infrastructure for the low income population in project cities. 1.08 The project was set up for an implementation period of three years. The special unit, initially known as the Sites and Services Directorate, which had been established in Ardhi for preparation of the project, was also to be responsible for its implementation. It was headed by expatriate advisors who were funded by United Nations Development Program (UNDP) and Canadian International Development Agency (CIDA). Other staff members were to be transferred from the Office of the Commissioner for Housing and consultants were to be used to fill the remaining gaps, which were then mainly in the areas of engineering and design.  - 4 - II. PROJECT IDENTIFICATION, PREPARATION AND APPRAISAL A. Identification 2.01 The project idea arose from the visit of an economic mission to Tanzania in September 1971. The mission's report highlighted Tanzania's increasing requirements for infrastructure and housing. This was followed by a reconnaissance mission in February 1972 which found Government policies regarding low-cost urban housing to be suitable for Baak-financing. The Government's official request for assistance which, was sent to the Bank in October 1972, was therefore favorably considered and the formal process of project definition got underway. B. Preparation 2.02 Preparation was carried out by the newly established Sites and Services Directorate in Ardhi with the help of an engineering consulting firm. In the period between the first mission in February 1972 and appraisal in August 1973 there were only two preparation missions. 2.03 Before the visit of the appraisal mission the consultants prepared preliminary engineering designs and cost estimates for the physical components. Preparation of the rest of the project, including TEB's lending program, institutional and financial aspects, as well as technical assistance was far less advanced. These components were being handled by two expatriates funded by UNDP and CIDA, in the Sites and Services Directorate of Ardhi. Most of the work on these components was done during the appraisal mission with the assistance of the Bank staff. C. Appraisal 2.04 Prior to the Appraisal mission's departure, concern centered on four issues: (a) the political acceptability of the site and service and upgrading concepts, (b) the unsuitability of the existing building code and need to revise it, (c) the buildup of Ardhi's staff to manage the project, and (d) the ground rent structure for recovering project costs. These issues were discussed in the Bank prior to the mission's departure and departmental positions agreed. With the benefit of hindsight these still appear to have been the right issues. In addition we should probably have directed greater energies to reviewing the institutional framework for managing the project and to the probable availability of building materials. It should be noted that since this was one of the Bank's first urban projects, considerable effort was devoted to working out the basic physical details and design-matters which are now fairly standard, particularly in second and third generation projects. 2.05 Appraisal took place from August 9 to September 10, 1973. The mission consisted of three Bank staff and one consultant who collectively spent 16 staff weeks on the task. The appraisal concentrated on working out a suitable physical composition of the project and obtaining government agreement on the preceding four issues. The original proposal which called for the development of 19,000 new plots and infrastructural improvements to  - 5 - 5,000 existing squatter plots over a four year period was also scaled down to what was thought to be a more manageable size as the Bank felt that the proposed target was too high from the point of view of demand and implemen- tation capacity. 2.06 Three of the four pre-appraisal issues subsequently became the subject of covenants in the legal documents. These concerned house stan- dards, project unit staffing and cost recovery. The fourth (political acceptability of the serviced site approach) was resolved by the Govern- ment's willingness to proceed with processing of the project. 2.07 Negotiations were completed in April 1974 without any major changes to the project. D. Conclusions 2.08 Since there was no prior experience with this type of project in Tanzania anti it was also a new area of lending for the Bank, preparation was based on the state of the art at the time and on the best estimate of Government implementation capabilities. The preparation process therefore did not adequately address all aspects of the project. The principal shortcomings were (i) the insufficient attention paid to the preparation of institutional and management aspects of the project; and (ii) the failure to recognize the need for technical and advisory assistance to plot benefi- ciaries. The serious deterioration in Tanzania's overall economy, following the global recession in 1974 was also not anticipated but would have been hard to predict. This led to widespread shortages, price escalations and increased uncertainty which made the environment a very difficult one to work in. The measures taken to address these constraints were therefore less adequate than originally anticipated. The shortage of staff in Ardhi also proved to be more difficult to redress than anticipated at appraisal and this compounded the problems outlined above. 2.09 Although the Government agreed to the initially modest physical standards, changes in staff and the institutional framework resulted in difficulties in maintaining these standards. Further steps might have been taken to amend the building regulations during preparation, or the amendments might have been made part and parcel of the implementation process. This, however, would have been quite time-consuming and would have delayed appraisal by at least a year. 2.10 In general, the preparation process might have been a more deli- berative one, given the newness of the concepts and weaknesses in the institutional structure.  - 6 - III. IMPLEMENTATION A. Effectiveness and Start-Up 3.01 The consultants for preliminary design for the infrastructural element of the project were appointed from February 1973 and prequalifica- tion of contractors started in May. As part of the preparatory work a pro- curement workshop, sponsored by CIDA, but with participation by the Bank, was held in August and September 1973. By the time of loan approval there- fore the unit established in Ardhi was ready to engage contractors. There were no special conditions set for effectiveness, which took place in October 1974. Nevertheless, a two to three month delay occurred because of difficulties in procurement, as is discussed in futher detail below. (Section IIIE). B. Progress of Physical Works 3.02 The physical components of the project can be divided into three main categories: infrastructure works; commiunity facilities; and, house improvement and construction. Progress with each category is outlined below: 3.03 Infrastructure. Infrastructure contracts covering roadworks, drainage and the installation of the water distribution network had to be scaled down to avoid cost overruns and because of unforeseen technical pro- blems. Changes made to avoid cost overruns were: (i) the deferral of the development of 881 plots in Mikocheni, Dar es Salaam and the reduction of Nyakato in Mwanza by 498 plots; and, (ii) the lowering of the standards of some project roads. The scale of sites and services development was also reduced by 350 plots, all of which were in Dar es Salaam,because of adverse soil conditions; 103 other plots could not be developed due to the align- ment of a water main in Dar es Salaam; and, 286 plots in Mbeya were squatted upon before they could be developed. Altogether, 20% fewer plots than planned were actually serviced. These changes resulted in 3 to 6 months delay in contract completions as the quantities in some contracts had to be reduced, and some others had to be retendered. 3.04 Work in upgraded areas proved to be more difficult than anticipa- ted and this resulted in approximately six months delay (Chart 1). 3.05 The electrification contract was to have been implemented by Tanzania Electric Supply Company Limited (TANESCO), a parastatal body. Work was delayed initially by insufficient progress with other infrastruc- ture work which had to be done first; and later, by closure of the border with Kenya, in 1977, which increased the difficulty of procuring poles. Finally, TANESCO diverted light fixtures ordered for this project when there was felt to be a greater need elsewhere since occupancy levels in the sites and services areas was still low. When Ardhi finally insisted that TANESCO reorder and install the fixtures, a dispute over responsibility for  TANZANIA FIRST NATIONAL SITES AND SERVICES PROJECT Implementation of Infrastructure and Community Facilities 1974 1975 1976 1977 1978 1979 1980 1981 SITE -.- - - - --- - - - - SITE ~~~~1 |2 |3 |4 1 | 2 | 3 |4 11 2 3 4 1 12 |3 |4 1 |2 |3 |4 1 |2 |3 |4 1 |2 |3 |4 1 |2 |3 |4 INFRASTRUCTURE/I COMMUNITY FACILITIES SINZA Dar Es Salaam Infrastructure , _ _, . Community Facilities Wn =- . 0 -- . 0d e S 101 e- ss -M KIJITONYAMA Dar Es Salaam Infrastructure 1 f = - - 1 MArtinEzZ