Ecuador - National Low Income Housing Project

Document of The World Bank FOR OFFICIAL USE ONLY Report No. 9239 PROJECT COMPLETION REPORT ECUADOR NATIONAL LOW-INCOME HOUSING PROJECT (LOAN 2135-EC) DECEMBER 14, 1990 Infrastructure and Energy Division Country Department IV Latin America and the Caribbean Regional Office This document has a restr.cted distribution and may be used by recipients only ia the performance of their official duties. Its contents may not otherwise be disclosed without World Bank authorization.  Fiscal Year January 1 - December 31 Weights and Measures Metric System Glossary of Abbreviations BEDE - Banco de Desarrollo del Ecuador (Ecuadorian Development Bank) BEV - Banco Ecuatoriano de la Vivienda (Ecuadorian Housing Bank) GTZ - German Technical Assistance Agency JNV - Junta Nacional de la Vivienda (National Housing Board) UIAT - Technical Support and Innovation Unit  TF WORLD SANK. FOR OFFICIAL USE ONLY Was"ton. DC. 20433 U.S A -office of Diletnct-GeVl Opewa:htia Ivalvattas December 26, 1990 MEMORANDUd TO THE EXECJTIVE DIRECTORS AND THE PRESIDENT SUBJECT: Project Completion Report on Ecuador National Low-Income Housing Proiect (Loan 2135-EC) Attached, for information, is a copy of a report entitled "Project Completion Report on Ecuador - National Low-Income Housing Project (Loan 2135- EC)" prepared by the Latin America and the Caribbean Regional Office with Part II contributed by the Borrower. No audit of this project has been made by the Operations Evaluation Department at this time. Attachment The sacument Xa a inug dsMbutUm ad MaY be Und by re man oesnly m Iu Of t'her oflL dum. is emaanw maytNr notobtteim be dmld withut WoWd usk a*tt_n.  FOR OFFICIAL USE ONLY PROJECT COMPLETION REPORT ECUADOR NATIONAL LOW-INCOME HOUSING PROJECT (Loan 2135-EC) Table of Contents Page No PREFACE ..................... i EVALUATION SUMMWIARY..................... ii PART I : PROJECT REVIEW FROM BANK'S PERSFCTIVE .......1 A. Project Identity . ................................. 1 B. Project Background . . .............................. 1 C. Project Objectives and Description .... 2 D. Project Design and Organization .... 3 E. Project Implementation . . ..........................5 F. Project Results ......................... . .. .. . 7 G. Project Sustainability .......................... . . 9 H. Bank Performance ............. 9 I. Borrower Performance . . ............................ 10 J. Project Relationship .......................... . 10 K. Consulting Services . ............ ................. 11 L. Project Documentation and Data .... 11 PART II : PROJECT REVIEW FROM BORROWER'S PERSPECTIVE........ 12 A. Project Design. . . . . ...... . . ... ...... 13 B. Project Execution . .14 C. Evaluation of the Bank's Performance .14 PART III: STATISTICAL INFORMATION .....16 1. Related Bank Loans .......................... . 16 2. Project Timetable ... ........................ 17 3. Loan Disbursements ........................... 18 4. Project Implementation ..... .. ............... . 20 5. Project Costs and Financing . . ..................... 21 6. Project Results . . ......................... 22 7. Loan Covenants . . ......................... 24 8. Use of Bank Resources . . ......................... 26 This document has a restricted distribution and may be used by recipients only in the performance of their official duties. Its contents may not otherwise be disclosed without World Bank authorization.  PROJECT CONPLEtION REPORT ECUADOR NATIONAL LOW-INCOME HOUSING PROJECT (Loan 2135-EC) PREFACE This is the Project Completion Report (PCR) for the National Low- Income Housing Project in Ecuador, for which Loan 2135-EC in the amount of US$35.7 million was approved on May 4, 1982. The final loan closing was June 30, 1988, one year behind schedule, and the last disbursement was in July 1989. Total disbursements were US$33.7 million and US$1.9 million were cancelled. The PCR was prepared by the Infrastructure and Energy Operations Division of the Latin America and the Caribbean Region, Country Department IV (Preface, Evaluation Summary, Parts I and III). Following the old guidelines for preparation of Project Completion Reports (PCR), the Borrower prepared a draft PCR in July 14, 1989, with detailed cost tables. Subsequently, at the Bank's request, the Borrower prepared Part II of the PCR. Preparation of this PCR was started during the Bank's final supervision mission of the project in April 1988, and is based, inter alia, on the Staff Appraisal Report; the Loan, Guarantee, and Project Agreements; supervision reports; correspondence between the Bank and the Borrower; and internal Bank memoranda.  - ii - PROJECT COMPLETION REPORT ECUADOR NATIONAL LOW-INCOME HOUSING PROJECT (Loan 2135-EC) EVALUATION SUMMARY Introduction The most innovative aspect of the chapter on housing of the 1980-1984 National Development Plan in Ecuador was its emphasis on increasing the supply of affordable housing for lower income groups while minimizing public subsidies. The Bank's stLategy in the sector was to help the Government develop policies and institutions to meet such objectives improving the capacity of formal sector institutions to supply shelter for the poor. With Bank support in the form of an Urban Development Project in Guayaquil (Loan 1776-EC), the Government began considering development of low-income housing a top priority, and gave BEV/JNV (Ecuadorian Housing Bank/National Housing Board) primary responsibility for implementing housing programs. The National Low-Cost Housing Project (Loan 2135-EC), with the aim of continuing this process, became effective May 4, 1983. Under its terms the Central Bank and BEV/JNV were responsible for implementation. Objectives The principal objective of the project was to support the Government's policy of providing housing affordable to low-income families by strenghtening BEV/JNV %nd demonstrating several innovative approaches to low-cost housing on a national scale. I.plementation Experience As a consequence of the combined effects of (a) Government changes in the earliest stages of the project, (b) inadequate technical management (c) protracted bureaucratic procedures, and (d) natural disasters, project implementation was adversely affected (paras 12-13). Results The project has succeeded in demonstrating the feasibility of low- cost shelter alternatives and introducing concepts of affordable standards, cost recovery, and replicability criteria into publicly-supported housing programs. It provided BEV/JNV technical assistance for improving its operational efficiency and implementing appropriate savings and financing mechanisms consistent with beneficiary conditions (paras 23-25). It also generated a positive spinoff effect in that the main objective of the follow-up project (Loan 2898) was to assist BEV/JNV in building on the achievements and expanding the goals established under this project. Sustainability BEV/JNV's improved performance during the later stages of the project suggests a general level of success in maintaining financial  - iii - viability and expanding institutional capacity (para 30). The project has succeeded in its innovative approach to bringing building codes down to more realistic levels, demonstrating the technical and financial viability of low-cost housing solutions tiat could be implemented by both public and private developers (para 31). Finding and Lessons Learned The core stated objective of the project was to demonstrate the feasibility of providing the poor with adequate housing and basic urban services. This was successfully achieved through the establishing of criteria and standards for selection and design of basic low-income family houses, while at the same time setting the stage for the introduction, under the follow-up Project, of innovative financing solutions including development of a long-term mortgage instrument which should ensure sound financial performance of BEV as well as shelter affordability. The clarity of project rationale was exemplified by its successful execution despite the difficulties confronted at the beginning. This underlines the importance of: (i) providing adequate and close supervision and timely and appropriate recommendations at all levels; (ii) applying lessons from one project to the design of the next; and (iii) continuing both technical and financial assistance to BEV/JNV to consolidate gains made under this project and also to expand upon its ac!-ievements. On the other hand, the project's imolementation delays indicate the overriding importance of ensuring: (i) further strengthening of the financial, administrative and productive efficiency of B!V/JNV (para 36); (ii) appropriate institutional and operational systems to improve implementation capacity; and (iii) participation in project preparation and design on the part of those who will implement it.  ECUADOR NATIONAL LOW-INCOME HOUSING PROJECT (LOAN 2135-EC) PROJECT COMPLETION REPORT PART I: PROJECT REVIEW FROM BANK'S PERSPECTIVE A. Project Identity Project Name National Low-Income Housing ProJect Loan Number 2135-EC RVP Unit LA4 - Infrastructure and Energy Division (LA4IE) Country Ecuador Sector Urban Subsector Housing B. Project Background 1. Sector Development Objectives. Ecuador has made substantial efforts to meet the growing requirements for urban housing during tile past twenty years. To demonstrate the feasibility of providing the poor with more adequate housing and basic urban services, the Bank helped the Municipality in the preparation of the Guayaquil Urban Development Project during 1978 and 1979 (Loan 1776-EC), and agreed to finance a study of the feasibility of a more ambitious program through the Second Technical Assistance Loan (S-006-EC). With the inauguration of a new Government Administration in August 1979, urban development and housing policies at the national level were sharply reoriented to benefit lower income people. At the same time, implicit and explicit subsidies that were formerly channelled into housing were reviewed more closely with the aim of rationalizing them. The activities ef all the institutions involved in the sector were to be somewhat reoriented and designed to support the simultaneous shift of Government shelter policy to be more equity-based and on a more solid financial basis. It also called for an increase in construction of more affordable homes. As a result, the Government initiated several programs to meet the shelter needs of the poor. Concepts and technologies which had been included on a demonstration basis in the Guayaquil Urban Development Project were now announced as national policy. Consequently, the Bank decided to support the cnange by identifying a project on the national level almost entirely limited to the strongest institutions in the sector (BEV/JNV) and to the simplest component (new housing). 2. Policy Context. Ecuador's urban population of about 3.8 million at the time of appraisal (1981)--442 of the country's total population--has more than tripled over the last 30 years. Its then rapid population growth, with an annual growth rate of 3.42 (now about 2.9Z), and rural. migration brought the urban population growth rate to 4.5Z. About 45,000 new households are formed annually. The overall housing deficit is estimated at about 430,000 units. In the absence of corrective measures by  the turn of the century, 9 million people will be living in urban areas. Such urban growth will place a formidable burden on Ecuador's siystems for providing shelter and urban services. 3. Linkages Between Project, Sector and Macro Policy Objectives. Although links were not explicitly described in project documentation, the project concept fit well into the Government's strategy of pursuing economic development by fostering a climate favorable to private investment ana improving the living conditions of the lower-income segments of the population. On the other hand, it supported the Bank's strategy of creating a better physical and social infrastructure base, strengthening implementation agencies, and improving living conditions for the urban and rural pour. C. Project Objectives and Description 4. Project Objectives. The project was designed to initiate a large- scale residential development for low-income families at the national level, directly benefiting approximately 14,800 families in Quito the capital of Ecuador, and nine secondary cities. About SOZ of project investments were expected to benefit disectly families with incomes below the absolute poverty level; in fact, most of the estimated 81,000 beneficiaries were individuals who had previously not owned housing. Its prime institutional and policy development objectives were to: (a) support the Government's policy of providing affordable housing to low-income families; (b) strengthen the institutional, technical, and financial capabilities of BEV (Ecuadorian Housing Bank) and JNV (National Housing Board), and support reorientation of their programs toward lower-income clients; and (c) support the expansion of low-income housing programs in Ecuador's secondary cities. 5. PLoject Description. The Guayaquil Urban Development Project (Loan 1776-EC) began a process aimed at presenting the Government with low- cost shelter alternatives and reorienting the basic approach for shelter investment. The second project (Loan 2135-EC) had a more direct focus on the operations of BEV/JNV and was designed to achieve greater institutional and operational reforms in these agencies while at the same time focussing on low-income beneficiaries and cost-recovery policies. These objectives were further expanded and consolidated through Loan 2898-EC (Second National Low-Income Housing Project) in an effort to continue the process of rationalizing the public housing investment program by providing appropriate solations consistent with maximum cost-recovery and replicability criteria. 6. Project Components. The project included four major components: (a) New housing (approximately 602 of total project costs--US$100.4 million), through the provision of about 8,200 shelters, comprising service-d lots and two types of basic housing units, together with basic infrastructure, in the ten project cities; (b) Urban upgrading (approximately 52 of total project costs), through the provision of legal land tenure and basic infrastructure to  some 2,400 households in existing low-income settlements in Machala and other cities; (c) Community facilities (about 8Z of total costs), providing a total of 10 schools, 7 health centers, 9 day-care centers, 14 community centers, and one police station in the areas benefiting from the project. Planning for these facilities were coordinated with the sectoral planning agencies: (d) Home improvement loans (about 20Z of total costs), designed to provide about 7,500 small loans--from a minimum of about US$170 to a maximum of about US$3,300--to residents of identified low-income areas. These loans were supposed to be at BEV's standard interest rate (152) and most would have a 5-year term. Beneficiaries would utilize these loans to build new houses on the serviced lots provided under the new housing component or to make improvements in existing houses in established low-income neighborhoods; (e) The project also included funds for project management (7? of total costs) i.e., salaries and equipment, and some technical assistance (2Z) of total costs) for the implementing institutions, i.e., consulting services, training, and improvements of information systems. Amendments: (f) In December 11, 1984, the Bank agreed to expand the scope of the project to include subloans for housing and reconstruction work for about 5,500 families affected by floods. Originally covering 10 cities, the project was expanded to encompass 15 cities plus adjoining townships. Cg) To assist in the reconstruction and rehabilitation of residential houses and municipal buildings damaged by the 1987 earthquakes in Ecuador, an amendment to the legal agreements was approved in August 31, 1987. The reconstruction program represented about 14X of the total loan (US$5 million). The new components comprised: (a) reconstruction and rehabilitation of municipal buildings in the towns of Quijos, Chaco, and Borja; (b) reconstruction and rehabilitation of residential houses in the provinces of Napo, Pichincha, Imbabura, Carchi, and Pastaza, including extending grants and subloans and technical assistance to beneficiaries. D. Project Design and Organization 7. Based on the information available to the appraisal mission of April 1982, the Bank was fully aware of the importance of limiting the institutional scope of the project to the strongest institutions in the sector (BEV/JNV), and to. the simplest, most advanced component (new housing), which with the institutions already had considerable experience. At appraisal, the proposed project was kept administratively simple to avoid initial delays during the finalization of legal agreements such as  the Guayaquil project (Loan 1776-EC) had suffered. Also, while the Guayaquil project was prepared in seven months, with major inputs from Bank staff, the National project was prepared by the executing institutions over a period of 15 months (between March 1980 and July 1981). Although institutional development was considered among the important aspects of the project, BEV/JNV were cvaluated as strong enough institutions both technically and administratively, to implement it. The most serious deficiencies identified at appraisal were: (a) JNV's capability to implement the relatively new approaches to low-cost housing the project involved (self-help construction, home improvement loans, and urban upgrading); and (b) BEV's reorientation to serve the ?oor could have a negative impact on its ability to attract saving from middle-income families and at the same time confront some difficulties in collecting from low-income borrowers.l 8. The physical components of the project were to be implemented by JNV, while the financial aspects were handled by BEV. 9. The appraisal mission acknowledged three main weaknesses in both institutions: (a) inadequate planning; (b) lack of coordination; and (c) inadequate implementation of their policy-making role--especially true of JNV, which had difficulty fulfilling its anticipated role of setting national housing policy. However, the mission considered the measures being implemented to overcome such difficulties as appropriate, which then were to be reinforced during project execution. 10. The project had a somewhat innovative approach in terms of its aim to demonstrate the feasibility of low-cost housing solutions that could later be replicated by both public and private developers. The new housir; component had maximum cost limits and also included income-level requirements designed to ensure that low-income beneficiaries fell between the 20th and 50th percentile of household income distribution. The credits for home improvement were open to all residents below the 50th percentile income level, and they had limits of SI. 180,000 and SI. 360,000 (US$1,200 and US$2,400) for signature and mortgage loans, respectively. In terms of a.hievements, Loan 1776-EC was instrumental in getting the Government to ccept appropriate reductions in norms and standarus for low-income beneficiaries and began the process of focussing on cost recovery and institutional reforms. Loan 2135-EC continued the process by focussing on BEV/JNV. At the same time, it was expected that BEV would be able to support a future reorientation of the savings and loan associations while JNV would in future assume some role in strengthening municipalities of the secondary cities. The Second National Low-Income Housing Project (Loan 2898-EC), continued the process by focussing on BEV/JNV's entire operations and developing some of their innovative approaches, including other lending agencies and private developers. 1/ A technical support unit was established in 1984 for BEV/JNV with the help of the German Technical Assistance Agency (GTZ). This unit coordinated technical cooper&tion activities funded by the Bank and USAID as well. A gradual improvement of BEV/JNV over the subsequent five years was expected to result.  11. Interest Rates. During preparation and appraisal, it became clear that BEV's financial position was in danger of deteriorating unless the interest rates it charged on its loans (then 7 to 22 percent) were raised. BEV's administration demonstrated at that time that It was commiLted to a policy of full cost recovery by raising interest rates to 15 percent, then the prevailing rate of inflation. This new rate compared favorably to the rate of interest it paid its borrowers (9 percent), and was considered by the Bank to be an appropriate basis to go ahead with the loan. While it was expected at that time that the inflation rate would continue to be stable at about 15 percent, BEV and the Bank agreed on an annual review of interest rates, with adjustments as needed, so as to ensure a financial return of credit operations and housing sales which would be "not negative" in real terms. The Bank and BEV did not, however, anticipate the important increases in inflation experienced over the life of the project which eventually required development of new financial instruments (paras 28 and 29). E. Project Implementation 1Z. The project was originally to be implemented over a four and a half year period from early 1982 to June 30, 1986, and the proposed closing date for the Loan (amounting to US$35.7 million) was June 30, 1987; however, the implementation process experienced a number of delays that extended the closing date by one year to June 30 1988. 13. Project implementation was adversely affected by factors both exogenous and endogenous to BEV/JNV. The former included: (a) institutional changes in the earliest stages of the project; (b) protracted national procurement procedures; (c) floods in the winter of 1982/1983, and earthquake damages in 1987; (d) delays by local municipalities in approving infrastructure designs; and (e) some delays in land acquisition. Among the endogenous factors were: (a) lack of coordination between BEV and JNV; (b) delays in the production of adequate final designs by JNV; (c) protracted bureaucratic procedures within BEV/JNV, particularly for the hiring of consultants; and (d) inadequate technical management of the project. 14. Although Loan 2135-EC was approved on May 4, 1982, and signed December 9, 1982, it only became effective on May 26, 1983, due to delays in compliance with three important covenants in the Loan and Project Agreements requiring that: (a) by December 31, 1982, BEV furnish the Bank with a draft statement of financial policy to be adopted by BEV's Board by March 31, 1983; (b) JNV and BEV enter into agreements to define the functions, responsibilities, and procedures in the implementation of the project; and (c) JNV provide the Bank with a report on its construction program for the project period and an evaluation of sites for construction of new houses under the project (Part III, Table 7). 15. Several deficiencies in project planning and management and more specifically, lack of coordination within the rousing sector persisted until changes were made in BEV's managerial hierarchy in 1985. JNV's bottleneck in the technical decision-making, which caused serious difficulties in project implementation, was solved by the middle of 1985  -6- with the appointment of a Project Manager and a Senior Planning Assistant, in compliance with section 3.07 of the Loan Covenants (Part III, Table 7). 16. During December 1984, the Bank agreed to expand the scope of the project to include subloans to beneficiaries in areas affected by floods (para 6.f). The project area was further expanded by an amendment made in July 1987 to address earthquake damage in Ecuador that occurred March 5 and 6, 1987 (para 6.g). At that time, the closing date for components related to earthquake reconstruction works was extended to June 1988. 17. By January 1988 all scheduled works of the loan were underway, but the needs of the earthquake reconstruction project and unseasonal rains in the project areas caused some implementation delays for BEV. Consequently, it had about US$5.5 million in loan funds undisbursed. Because of delays in the development of a new mortgage system to be used for the emergency component, the conditions of Loan 2135-EC had to be adjusted to suit the special circumstances of these beneficiaries2 a one-year extension of the non-earthquake reconstruction components (i.e., new units and home improvement loans) was requested and obtained by BEV. This result was a common closing date for the entire loan instead of an existing partial extension of the reconstruction components only. 18. Interinstitutional Coordination. Interinstitutional coordination was the responsibility of BEVIJNV, which also aasumed overall implementation coordination for reconstruction and rehabilitation works, which were expected to be completed by December 31, '988. BEV managed the housing reconstruction activities while JNV had primary responsibility for the municipal reconstruction works. Emergency works in the municipal building -ere never implemented because BEVIJNV could not finalize agreement with the local authorities. 19. At the time of the earthquake, the Government was facing financial constraints which were aggravated by the destruction of the petroleum pipeline. BEV's major sources of funds were saving deposits (332) and retained earning (34Z), but its rapid growth during the 1970s also depended on generous transfers from the Government's petroleum revenues. After Government transfers were reduced, BEV had to finance continued growth by borrowing, and the need for full cost recovery became more pressing and obvious. 20. New Housing and Urban Upgrading. By the end of 1987, 812 of programmed investment was executed. Notwithstanding some initial difficulties with the Urban Upgrading component, new housing and urban 2/ The proposed reconstruction component was to serve a generally poorer segment of the population than the original project. Therefore, management approval was sought to reduce the interest rate (that evolved from 152 at appraisal to 182 in 1985, and again to 192 between 1986/87) for needy beneficitries from 192 to 9Z. However, this was not approved because it was believed that NGO's were dealing with needy cases and interest-rate subsidies should not be supported by the Bank.  upgrading were successful elements in the project in terms of implementation. 21. Housing Improvement Loans. Housing improvement loans were a very successful element in the project that benefitted from the previous experience of BEV through Loan 1776-EC. Ninety-seven percent of total investment was completed in December 1987. 22. Community Facilities. As in a previous operation, (i.e., Urban Development Project in Guayaquil, Loan 1776-EC), the schools and health centers components were not implemented due to lack of funds by the respective Ministries. Both the Ministry of Education and Health faced serious budgetary constraints that prevented them from financing that area of the project a.s it had been initially contemplated. Consequently, BEVIJNV agreed to lend the funds out of its own resources to the respective Ministries for these works. On June 3, 1986, BEV approved a loan of S/. 63 million to the Ministry of Education for the construction of 11 schools. By December 1988, however, both this works and the health centers were still in the negotiation process. F. Project Results 23. The project succeeded in strengthening the managerial and overall organizational capabilities of BEV/JNV. Both institutions increased their efficiency during the implementation process; it is expected that this institutional revitalization will be enhanced thrcugh different organizational commitments. There were however certain internal control weaknesses in BEV, particularly in the areas of coordination, communications, and planning, which were corrected through institutional reforms by the end of 1984. 24. The new-housing component proved to be one of the most successful components together with the home improvement loans. Both of them were efficiently implemented, with only minor delays after changes in BEV management in 1985, when project expenditures almost tripled compared to 1984.3 25. The project had a positive spinoff effect in that new housing operations were implemented within BEV's investment program. In addition, these operations complement prcspective systemwide housing finance measures such as increased savings mobilization and mortgage loan instruments better adapted to high and variable inflation. 26. Pro3ect Costs and Financing. As a result of implementation delays and much higher local inflation than forecast at appraisal, project costs in local currency terms rose to about twice original estimates. This represented about twice the originally estimated counterpart funding requirement (in local currency terms) and would have imposed a considerable burden on the fiscal resources at a time of budgetary austerity, since BEV would have been unable to fund such a cost increase from its internal resources. As a consequence of this local cost overrun, the Bank agreed to 3/ By mid-1985, the project was almost two years behind schedule.  -8- consequence of this local cost overrun, the Bank agreed to increase its percentage participation in the project, effective July 1, 1985, from about 352 to about 602 of tocal costs, and to finance under the project additional selected expenditures by BEV/JNV to facilitate project execution. This amendment did not increase the total amount of the loan. In addition, the Bank agreed to extend the dates for compliance of several Loan covenants (Part III, Table 7). Increased cost-sharing was accomplished by raising the disbursement ratio for civil works, home improvement loans, and local expenditures for services and equipment from 42X to 602. 27. BEV's major source of counterpart funds was BEDE, the Development Bank, and no major difficulties were experienced. BEV's access to funding improved from 1986. It negotiated a budgetary allocation of SI. 7.3 billion (US$68 million), which represented nearly a fivefold increase comparpd to 1985. Despite heightened austerity in Ecuador due to low oil prices, BEV did not have counterpart funding difficulty, since housing remained a high priority. 28. During project implementation, interest rates were adjusted twice, but not sufficiently to meet the levels oi inflation experienced over the period. The inflation rate surged in 1983 to 48 percent, coinciding with the collapse of oil prices and the beginning of the debt crisis, then declined to 30 percent in 1984. In 1985, interest rates charge on BEV's loans were increased to 18 percent (inflation was 28 percent); the interest rate prevailing in 1986 and 1987 was 19 percent, as compared to inflation of 23 percent and 30 percent in those years. During this period, BEV's lending rates were, however, more closely linked to the rates paid to BEV's depositors. In 1987, for example, BEV paid 21 percent on saving deposits. 29. During the course of project implementation, it became clear to the Bank, BEV and the Government that the current system of fixed-rate mortgages was no longer appropriate, and that innovative financing solutions would be required to prevent further erosion of the real value of BEV's portfolio while ensuring shelter affordability. Focus shifted to development of a long-term mortgage instrument bearing market-related variable interest rates and featuring partial capitalization of interest. Introduction of this instrument was a main feature of the follow-up project now under execution, the Second Low-Income Housing Project (Loan 2898-EC). 30. Overall, BEV's performance reflected good managerial procedures. Delinquency in mortgage loans waa about 4Z; however, signature loans showed a much higher delinquency rate (10%). Because of an increase in the minimum salary, the limits on the maximum sales prices were subsequently raised.4 Consequently, BEV exceeded the affordability target in Quito by about 382 (78% of the shelters were in the lower range instead of the 40? defined at appraisal). 4/ Maximum sales prices were defined to meet the affordability targets according to monthly family income level. House pricing, as per the ApDraisal Report and Loan Agreement, was based on cost with no subsidies.  -9- G. Project Sustainability 31. The new-housing component reached its objectives establishing criteria and standards for selection and design of basic low-income family houses. This was an opportunity for BEV/JNV to consolidate lessons learned from the past as well as to improve their planning and implementation capabilities in dealing with large-scale housing programs. 32. Cost recovery was estimated as seven points lower than the 90? foreseen at appraisal. The ex-ante overall rate of return for the three main components of the project with quantifiable benefits was 18? and for each of them separately: 17? for new housing, 21? for home improvement loans, and 41? for urban upgrading; whereas BEV's ex-post EER was 20? assuming the number of beneficiaries as a function of the total number of housing finished between 1985 and 1987 (Part !II, Table 6B). 33. Insofar as the project has succeeded in its innovative approach to low-cost housing, bringing building codes down to more realistic levels, while at the same time strengthening BEV/JNV management and organization. The next step should be to consolidate BEV/JNV's reorientation, strengthening their potential to become the central institution in supporting the reorientation of the savings and loan associations. Loan 2898-EC, the Second National Low-Income Housing Project, through its Technical Assistance component, was expected to address this issue. 34. Both the Ministry of Public Health and the Ministry of Education showed certain weaknesses in implementing the community component, and had difficulty promoting sustainable policies to ensure construction, operation, and maintenance of community facilities included in the project. B. Bank Performance 35. Project targets were overambitious in some respects, given the organizational difficulties of project institutions, particularly during the first half -of the implementation process: final designs were not available on time or offered inadequate and erroneous solutions, which required redesigns and site adjustments. From this point of view the appraisal estimation of implementation deadlines was unrealistic. Modification to the Loan and Project Agreement aimed at improving BEV/JNV's implementation capacity had to be introduced. 36. Insufficient attention was given at appraisal to diagnosing the financial difficulties of the Ministries of Education and Health, which led to unsuccessful results in the completion of the community facilities component. Given the past experience of the Bank in dealing with this particular problem (i.e., Loan 1776-EC), implementation and financial constraints on the part of these Ministries could have been foreseen at appraisal. 37. Notwithstanding the difficulties confronted at the beginning in the execution and managerial procedures, the Bank's staff through close supervision and timely and appropriate recommendations was able to prevent  - 10 - fur.her deterioration and largely improved project implementation. In that respect. project supervision teams operated as technical assistance support groups. I. Borrower Performance 38. BEV/JNV's execution and managerial weakness during project preparation and implementation (para 10 above) were caused mainly by: (i) some difficulties in the appointment of high level staff in managerial positions; (ii) lack of coordination (and at times rivalry) between BEV and JNV. BEV/JNV was, however, able to overcome these difficulties under strong leadership from the top management of both institutions, during the latter part of the project period, when 81Z of the implementation of its component was achieved; (iii) shortage of qualified personnel due to low JNV salaries, which resulted in inadequate capacity to resolve technical problems in a timely and effective manner. This was partially overcome through the Technical Assistance components which helped BEV/JNV with project administration procedures; (iv) protracted bureaucratic procedures within BEV/JNV; (v) lack of funds by the Ministries of Education and Health to implement the community facilities component; (vi) difficulties in procurement procedures that did not comply with provisions of the loan agreement. 39. The German Technical Assistance Agency (GTZ) formed a Technical Support and Innovation Unit (UIAT) to advise BEV/JNV on institutional improvements. This helped to correct a number of deficiencies, particularly the centralization of decision-making. 40. Two new posts were created at the suggestion of the Bank aimed at circumventing a bottleneck in the technical decision-making, which in fact had a positive influence on the implementation process. These appointments were a senior-level Project Manager within the Director-General Department of JNV, and of a Senior Planning Assistant within the Planning Directorate of BEV/JNV. 41. Overall, BEV/JNV's -liance with legal conditions of the Loan governing affordability and limitations were met. However, increasing inflation rates did not permit BEV to meet lending terms (Section 2.08, Project Agreement). The interest rates charged by BEV on housing loans remained below inflation rates (para 28). J. Project Relationship 42. The success of the dialogue wit} .ne Government can be attributed to at least two factors. First, the Bank essentially supported the Government's program of supplying affordable housing for lower income groups while minimizing public subsidies; and second, the dialogue, from the Bank's view point, took into account that the Govzrnment was committed to developing policies and institutions to meet such objectives and provided technical assistance for improving BEV/JNV operational and financial efficiency.  K. Consulting Services 43. Excellent work was performed by consultants in advising on project implementation and administration as well as in their evaluation. However, the foreign consultants deployed in the establishment of a financial management information system, project administration and urban upgrading, performed less than satisfactory work. L. Project Documentation and Data 44. Original documentation for the project was adequate. However, the changing conditions of the country during the implementation period hampered the scope of its usefulness. 45. This report relies mainly on the legal documents, the Staff Appraisal Report, aide memoires and supervision reports, which provided the basic information for the preparation of Part I and Part III.  - 12 - PROJECT COMPLETION REPORT ECUADOR NATIONAL LOW-INCOME HOUSING PROJECT (Loan 2135-EC) PART II: PROJECT REVIEW FROH BORROWER'S PERSPECTIVE5 Loan Agreement IBRD 2135-EC concluded on December 9, 1982, partially financed the National Low-Cost Housing Project in an amount of US$35.7 million. This program served basically to initiate urban development projects and large-scale housing schemes for low-income households. In addition, a significant benefit of the program, which was also intended to strengthen the institutional, technical, and financial capacity of the Banco Ecuatoriano de la Vivienda and Junta Nacional de la Vivienda (BEV/JNV) was that it made it possible to apply a systematic policy of expanding subsidized housing and community development by means of appropriate channeling of public investment. The Government, being aware of the purposes of the program, has ensured that the proceeds of this loan have been used in accordance with the execution schedules established. Some 952 of the funds were invested, on the basis of which the BEV and the JNV were able to obtain a new loan, IBRD Loan 2898-EC, "Second Low-Cost Housing Project." The housing in question was built in 10 cities, which were selected on the basis of variables such as demographic growth, low level of development, and site availability. The initial planning for the project envisaged construction of 8,200 dwellings, and upon its conclusion a total of 7,880 had been built. Execution of the project suffered significant delays to begin with, due mainly to difficulties in acquiring the sites, obtaining the approval of the municipalities involved for execution of the works, protracted and complicated procedures for contracting the works, etc. To these problems were added the need for the executing agencies, the BEV and JNV, to simultaneously undertake new programs such as incremental building, housing upgrading, and urban development. 51 This is an unedited version of Part II as written by the Ministry of Finance and Public Credit, with comments from BEV and JNV.  -13- Following completion of this project it can be stated that the executing agencies have achieved the level of organization and administrative improvement considered essential for implementation of the goals agreed with IBRD. In addition, as is known, significant changes occurred in the country's economic and social circumstances during execution of the program. These changes impacted its implementation, to the extent that they also affected the objectives and parameters on which the initial planning for meeting the demand for low-cost housing was based. These and other difficulties resulted in the closing date for disbursements from the loan, originally set at June 30, 1987, being extended to June 30, 1988, which meant that the BEV/JNV needed a further US$4 million for executing the projects. In conclusion, it can be stated that notwithstanding the complications that arose. the housing program in question had a very positive impact. The World Bank's readiness to assist Ecuador in accomplishing the set objectives deserves special mention. Finally, with regard to the project completion data contained in Parts I and III prepared by IBRD, this information is considered important and of great value for analysis and lessons to be drawn concerning the performance of the project administrative units. The report setting out the evaluation and comments of the BEV and JNV regarding their areas of competence is as follows: A. Project Design In evaluating project IBRD 2135-EC, the following should be noted regarding the design of the project: The project was reasonably designed both as regards its size and as to the resources needed for its execution, except for certain components such as community infrastructure and reconstruction of housing damaged by the earthquake, execution of which did not meet expectations for the following reasons: i. In the case of community infrastructure, it was impossible to implement agreements among the participating institutions either because of the limited operating capacity of certain executing agencies or else 'ecause political considerations placed a constraint cn the degree of execution of this component. The allocation for reconstruction subloans was not fully used because the target population also received grants and preferential loans from nonreimbursable funds provided by the German Government and AID.  - 14 - ii. The schedule for utilization of the loan funds assumed ideal circumstances and failed to take into consideration factors that affected its implementation in many ways. This observation is borne out by comparison of the closing date for disbursements set in the project appraisal (1987) and the date as of which 94.7Z of the external resources was actually disbursed, namely June 1989. This slippage in use of the loan funds was due to the following reasons: (a) The project preparation assumed an operating capacity that the executing agencies did not in fact possess, so that the programming proved too optimistic. (b) The programming failed to take into account the time required for bidding for the work contracts due to tne requirements of Ecuadorian legislation; the fact is that in some cases. especially in the execution of housing programs, this process took more than a year. B. Project Execution Regarding the execution of the project, the following should be noted: The experience gained with the execution of the Guayaquil Urban Development Project financed with loan IBRD 1776-EC enabled the BEV and JNV to set up appropriate mechanisms for execution of project IBRD 2135-EC, in which the organic structure of the two institutions was a significant factor. When execution of the National Low-Cost Housing Project was started. the BEV and JNV had a functional organic structure suited to the project's requirements; as a result, the housing upgrading loan program was implemented successfully and housing production by the JNV was in accordance with the targets set. The fine tuning of the BEV's and JNV's operating structure, together with the progressive change in the housing policy targeting middle and low-income families, definitely made it possible for the project targets and objectives to be met. C. Evaluation of the Bank's Performance i. The periodic supervision missions for loans IBRD 1776-EC and 2135-EC were fruitful in that they diagnosed the different problems affecting the normal progress of the projects and produced the most appropriate solutions for utilization of virtually the full amount of the external credit funds. ii. The technical assistance received by the BEV and JNV enabled the gradual change in administrative, financial, institutional and organizational spheres that led to the establishment of an operating structure compatible with the requirements for execution of the project.  -15 - iii. Disbursements were processed promptly and capably by the World Bank's Disbursement Division, thereby contributing to the utilization of 952 of the loan funds. iv. Although the BEV and JNV had technical assistance permanently available to them, the World Bank nevertheless did not assign sufficient importance to staff training, particularly as regards courses of longer than two months' duration; this negatively affected the institutions' ability to utilize 100% of the resources allocated for technical assistance.  PROJECT COMPLETION REPORT ECUADOR NATIONAL LOW-INCOME HOUSING PROJECT (Loan 2135-EC) PART III 1. Related Bank Loans Loan Year of Title Purpose Approval Status Comments 5-008-EC Pro-investment studies; 08/1977 Completed USS17.5 million Loan effective institutional development in March 1978. Closing delayed assistance to FONAPRE and the by 8 years. Despite initial Ecuadorian consulting industry difficulties all components were implemented. Good progress in institutional development of FONAPRE. 1778-EC Initiate large-scale programs of 12/11/79 Completed USS26.8 million Loan effective in employment generation and October 1980. Closing delayed by residential development for low- 2 years. income families in Guayaquil. 2898-EC Increase production of low-income 1/12/88 Ongoing USSeO.0 million loan effective in housing, and further strengthening September 1988. BEV/JNV administrative and productive capabilities while fostering private participation In low-income housing production.