Argentina - First Housing Sector Project

Documet of The World Bank FOR OFFICLAL USE ONLY Report No. 13194 PROJECT COMPLETION REPORT ARGENTINA FIRST HOUSING SECTOR (LOAN 2997-AR) JUNE 22, 1994 Trade, Finance and Private Sector Development Country Department IV Latin America and the Caribbean Region 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 witbout World Bank authorization.  CURRENCY EOUrVALENTS Currency Unit = Argentinean Peso Exchange Rate I US Dollar = I Peso (Since April 1, 1991, the exchange rate has been established by law) GOVERNMENT OF ARGENTINA FISCAL YEAR January 1 - December 31 ABBREVIATIONS AND ACRONYMS BHN - National Mortgage Bank (Banco Hipotecario Nacional) FONAVI - National Housing Fund (Fondo Nacional de la Vivienda) IPV - Provincial Housing Institute (Instituto Provincial de la Vivienda) SVOA - Secretariat of Housing and Environmental Management (Secretaria de Vivienda y Ordenamiento Ambiental)  FOR OFFICIAL USE ONLY THE WORLD BANK Washington, D.C. 20433 U.S.A iffice of Director-General Operations Evaluation June 22, 1994 MEMORANDUM TO THE EXECUTIVE DIRECTORS AND THE PRESIDENT SUBJECT: Project Completion Report on Argentina First Housing Sector Project (Loan 2997-AR) Attached is the 'Project Completion Report on Argentina - First Housing Sector Project (Loan 2997-AR)" prepared by the Latin America and the Caribbean Region. Part II was prepared by the Borrower. This was the first Bank assistance to the housing sector in Argentina. The US$300 million loan was designed to help the government reform sector policies and to reduce the deficit in low-cost housing. All but less than US$22 million were cancelled at the Borrower's request. Project risks were correctly identified but underestimated, and thus were inadequately reflected in project design. Volatile economic conditions and insufficient government commitment hindered reform. Constant staff changes in executing ministries and agencies impeded implementation. The inherent limitation of central government management of dispersed housing investments was another complicating factor. The project was not restructured because of the lack of ownership by borrowers and executing agencies as well as the different policy introduced by the new government. The PCR gives a satisfactory account of the circumstances which led to the cancellation and allocates fairly responsibilities for the failure. Overall, the project outcome is rated as unsatisfactory, its sustainability as unlikely, and its institutional impact as negligible. The project may be audited together with other large housing projects in Latin America. Robert Picciotto by H. Eberhard K8pp Attachment 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.   FOR OFFICIAL USE ONLY ARGENTINA PROJECT COMPLEION REPORT FIRST HOUSING SECTOR (LOAN 2997-AR) TABLE OF CONTENTS Page No. Preface ............................... i Evaluation Summary .............u ;-iv PART I PROJECT REVIEW FROM BANK'S PERSPECTIVE ....1.... A. Project Identity ............ .............................. 1 B. Background ............................................ 1 C. Project Objectives and Description ........................... 2 D. Project Preparation, Design and Organization ................... 3 E. Project Implementation ................................... 4 F. Lessons Learned From the Implementation Process .... .......... 8 G. Project Results and Sustainability ........................... 10 H. Bank Performance ...................................... 10 I. Borrower Performance .................................. 10 J. Project Relationships .................................... 10 K. Consulting Services .11 L Project Documentation and Data ........................... 11 PART II PROJECT REVIEW FROM BORROWER'S PERSPECTIVE ....... 12 PART m STATSTnCAL ENFORMATION ............................. 15 Table 1. Related Bank Loans ................................... 15 Table 2. Project Timetable ..................................... 15 Table 3. Cumulative Estimated and Actual Loan Disbursement .... ...... 15 Table 4. Project Implementation ................................. 15 Table SA. Project Financing ................. ................... 15 Table 6. Project Results ....................................... 16 Table 7. Status of Major Loan Covenants .......................... 16 Table & Use of Bank Resources ................................. 18 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.   - i - ARGENTINA PROJECT COMPLETION REPORT FIRST HOUSING SECTOR (LOAN 2997-AR) PREFACE This is the Project Completion Report (PCR) for the First Housing Sector Project in Argentina, for which Loan 2997-AR in the amount of US$300 million was approved on October 8, 1988, and became effective on March 3, 1989. The project's performance was not satisfactory and the loan was cancelled (US$278.3 million, 92.8 percent) in June 1992 at the request of the Borrower. This PCR was prepared by the Trade, Finance and Private Sector Development Division, Department IV of the Latin America and the Caribbean Regional Office (LA4TF) and is based, inter Aug, on the President's and the Appraisal Reports, the Loan Agreement, supervision reports, correspondence between the Bank and the Borrower, final project analyses prepared by the Borrower, and internal Bank memoranda. The Govermment of Argentina submitted Part II, Project Review from Borrower's Perspective on August 23, 1993.  I  - ii - ARGENTINA PROJECT COMPLETION REPORT FIRST HOUSING SECTOR (LOAN 2997-AR) EVALUATION SUIMMARY A. Introduction 1. The Housing Project was conceived by a newly redemocratized govemment to correct worsening social conditions while it sought simultaneously to stabilize its economy. Part of its impetus came from the expanding shortage of affordable low-income housing, lack of private building activity and the disappearance of the capital market for housing finance. Another stimulus was a desire to raise the productivity of public expenditures and to seek broad gauged sectoral reforms. This interest focused particularly on social programs seen as poorly justified, inefficiently managed and only marginally helping the intended beneficiaries. 2. This was the context within which the project was designed: seeking to strike a balance between some short-term social benefits and the need for lasting policy and administrative changes. The loan was ambitious and complex. It was intended as a funding structure for activities that would assist the Government in strengthening its housing management for significantly reoriented operations, in an increasingly unstable economic and political setting. In the event, the intended contributions were not realized as, for the reasons indicated below, the project's poor performance led to cancellation of the loan some 39 months after its effectiveness. B. Objectives 3. The overall objective of the project, the first Bank-assisted housing operation in Argentina, was to help government reform sector policies and reduce the housing deficit for low income families. The main purposes of the reforms were to increase the productivity of the National Housing Fund (FONAVI) programs, improve their finances and targeting, and to reduce subsidies. The project was to finance construction of low-cost housing and contained a package of policy improvements.  - iii - C. Implementation Experience 4. The RBnk correctly identified the main risks of the operarion as including the difficulties of altering FONAVI's financial management, the weaknesses of Federal Government officials, and institutional deficiencies throughout the housing sector. However, these risks were underestimated, and the absorptive capacity of the Government overestimated. They were thus inadequately addressed in project design. The project suffered delavs and obstacles from the outset. It also was consistently buffeted by extremely volatile economic and political conditions, inappropriate circumstances for trying to introduce broad policy and institutional changes. These alone would have weakened the prospects for success of the project. Unforrunately however, there was also insufficient government commitment to the goals of the exercise. 5. During its implementation. there were constant personnel changes in the Ministry of Health and Social Action and its Secretariat of Housing and Environmental Management (SVOA), the Government agency responsible for defining and executing housing policy, including adrministration of FONAVI. This instability figured centrally in the project authorities' failure to take the necessary actions to implement the operation and comply with loan covenants. In particular. the failure to employ the required financial manager and other members of the Advisory Committee seriously affected the management of FONAVI funds. This faulty management contributed as weUl to poor SVOA control of the Provincial Housing Institutes (IPVs), the project's executing agencies. It also obstructed the enforcement of SVOA's agreements with the provinces and municipalities on eligibility conditions for access to loan funds. The technical assistance programs largely designed to strengthen SVOA were not well administered. 6. The Menem administration which took over in mid-1989 seemed to share the Bank's assessment of the project's difficulties. However, noncompliance with loan covenants continued, and it became clear that there were unlikely to be any significant improvements in the project entities or their interests in carrying out the operation more adequately. With no remedial action, the Borrower agreed in July 1991, not to apply for additional disbursements unil major loan covenants were complied with. The Government requested cancellation of the loan in March 1992, with less than 8 percent of its proceeds disbursed. D. Results and Sustainability 7. Effectively, the project objectives were not achieved. E. Fmdines and Lessons Learned The operation was unsatisfactory, mainly for the following reasons: (a) There was an under-estimation of the weaknesses of the institutional structure, the influences of political circumstances,  - iv - and of the prospects for further deterioration of Argentina's economic situation. (b) Provincial interest in the project progressively waned, especially as there proved to be insufficient incentives to encourage them to comply with the loan conditions. Perhaps this was because the amount of funds which they could have received from the loan were relatively small compared to the magnitude of regular FONAVI resources estimated to be provided during the project implementation period. Throughout much of this period moreover, the recurrent bouts of high inflation eroded the value of FONAVI lending to the extent that it became financially unattractive. Further, there were increased provincial fiscal problems at the time. (c) The Federal Government's centralized administrative system proved unable to govern a substantial number of individual investments in this large, regionally diverse country. However, FONAVI's inability to control the IPVs was not a unique circumstance. The increased autonomy of the provinces then under Argentina's resurgent federalism, and the progressive decentralization of most public services to them have stringently limited the possibilities of successfully implementing nationwide projects, especially those involving all provinces. (d) FONAVI's problems and this project experience also demonstrated that the state had, in many respects, lost its capacity to effectively direct day-to-day operations, and needed more drastic medicine than intemal management improvement campaigns. They also showed that the public administration had become increasingly unable to apply equitable rules for dealing with the diverse favored beneficiaries of many programs, and how entrenched these groups had become. 8. The project, with the benefit of hindsight, must be judged as too ambitious an undertaking. The restructuring of entrenched, highly inefficient Argentine Government institutions in and of itself would not have been easy in any circumstance. But the approach taken in designing the project was greatly complicated by the uncertainty of the economic environment.   ARGENTINA FIRST HOUSING SECTOR (LOAN 2997-ARG) PROJECT COMPLETION REPORT PART I: PROJECT REVIEW FROM BANK'S PERSPECTIVE A. Project Identity Name: Housing Sector I Loan Number: 2997-AR RVP Unit: Latin America and the Caribbean Country: Argentina Sector: Infrastructure/Urban Subsector: Housing B. Back2round 1.1 The Housing Project was conceived as a newly redemocratized government's attempt to correct worsening social conditions while seeking simultaneously to stabilize the economy. It originated from the efforts of the Alfonsin administration to respond to the economic crisis it inherited when civilian rule was reinstated in 1983. After some setbacks, the Government seemed to be effectively taking command. The 1985 change in the economic team, the shift from gradualism to "shock" treatments, and some initially successful reform efforts provided a positive backdrop to the administration's decision to introduce broad sectoral and administrative changes. At the time, there was impetus for the project stemming from the expanding housing shortage, lack of private building activity and the disappearance of the capital market for housing finance. In addition, there was the hope for some economic growth impact from increased housing construction works (which amounted to some 6 percent of GDP and provided about 4 percent of employment in the previous decade). Other stimuli were the desire to raise the productivity of public expenditures and to help trim the public sector deficit. This interest focused particularly on social sector programs seen as poorly justified, inefficiently managed and of little impact on intended beneficiaries. 1.2 In the light of these factors, in conjunction with requests for other structural reforms, the Argentine Government asked for a housing sector project in September 1987, with great urgency. The unmet housing demand at the time was estimated to be as large as 2.5 million units. Demand for new housing was increasing at 220,000 units per year, but only about 80,000  -2- were built in 1986. The need was especially severe among the lower-income population. The authorities therefore sought to have more houses built for a larger number of families in the shortest possible time, and more cheaply. This was the context within which the project was designed: seeking to strike a balance between some short-term social benefits and the need to achieve lasting policy and administrative changes. The loan was ambitious and complex in terms, for example, of efforts to obtain reforms on numerous fronts, as well as the multiplicity of institutions involved. It was intended as a funding structure for activities that would assist the Government in strengthening its housing management for significantly reoriented operations, in an increasingly unstable economic and political setting. This relied on the belief that the answer to the great institutional needs in this (as in other) areas could derive from technical assistance to the key government agencies. 1.3 The appraisal found that low-income housing was financed mainly by the Housing Secretariat of the Ministry of Health and Social Action (SVOA), drawing on the proceeds of a 5 percent earmarked tax on salaries and levies on the self-employed. SVOA allocated monies from the National Housing Fund (FONAVI) to provincial housing institutes (IPVs) for the latter's execution of these programs through private construction. It was found that FONAVI was inefficiently managed and had targeted beneficiaries poorly. Its houses were costly (about US$20,000 each) and took too long to complete (around three years). Only about 10 percent of their costs were recovered. Both FONAVI's salary tax and mortgage payments had been eroding rapidly with the country's high inflation. 1.4 The project was the Bank's first involvement with public sector housing in Argentina. Its preparation drew on some reconnaissance on the sector and, in particular, the deficiencies of existing housing policies that were identified under the Technical Assistance Project for Social Sector Management (Ln. 2712), approved on July 26, 1988. A IJNDP housing study (July 1987) also had identified some sector problems. In addition, the project was included in the US$1.2 billion package of four loans approved (including amnendments to a previous loan) in November 1988, in order to support the new "Plan Primavem" structural reform program. C. Project Objectives and Description 1.5 The main objectives of the project were reforms in sector policies and reductions in the housing deficit by rationalizing the use of FONAVI funds. The project was to finance construction of low-cost housing and contained a package of policy improvements which included: * the establishment of registries of potential beneficiaries in order to facilitate targeting subsidies and selecting beneficiaries; * new housing designs to lower unit costs;  -3- * the relaxation of standards that excessively restricted innovative, more economic housing designs; * improved, standard bidding procedures; * a new mortgage repayment system to improve cost recovery; * the promotion of savings; * transparent subsidies restricted to low income groups; and * new rules for allocating FONAVI funds based on such factors as effective demand, IPVs' repayment records and efficiency. 1.6 The proposed loan was to finance the construction of about 39,000 two-bedroom houses, 20,500 three-bedroom houses, and the improvement and rehabilitation of about 40,000 existing houses. Up to 15 percent of base project costs were reserved for improving and rehabilitating substandard housing. Efforts were to be made to cut construction periods by two thirds. Uniform contracts were to be used for more transparent procurement. And the Loan, about 1-1/2 years after effectiveness, was amended to introduce even less expensive house types for low income groups, and to give FONAVI more flexibility to providing other housing models (the latter without Bank financing). 1.7 Loan proceeds together with FONAVI resources and the beneficiaries' down payments would finance the units to be constructed or imnproved. FONAVI was to assume the cross-currency and the US dollar exchange risks. The Loan funds were to be made available to the IPVs in local currency following agreed allocation criteria spelled out in agreements defining loan recovery and other obligations. The project housing credits would be repaid at an annual real interest rate of 8.7 percent over up to 25 years, with adjustments for inflation. The repayment amounts would vary depending on the beneficiaries' income levels. This means-based subsidy, in contrast with the previous over 95 percent levels, would average 37 percent. No subsidies would be granted to families with incomes above US$280 per month. Only the lowest 40 percent income earners would qualify for FONAVI house financing. The loan included funds for technical assistance to carry out studies and implement recommendations in several areas required to make SVOA and the IPVs more efficient. Loan components: Housing construction loans, US$250 million Home improvement loans, US$40 million Technical assistance and auditing services, US$10 million D. Project Preparation, Design and Organization 1.8 In preparing the project, the Bank correctly identified its main risks as including the difficulties of altering FONAVI's financial management, the weaknesses of Federal Government officials, and institutional deficiencies throughout the housing sector. However, these risks were underestimated,  - 4 - and the absorptive capacity of the Government overestimated. They were thus inadequately addressed in project design. Given the Alfonsin administration's activity then in restructuring much of the administrative apparatus, and the exceedingly low government salary scales (and motivation), it is not surprising that the project suffered delays and weak integration from the outset. However, the persistent discontinuity within the Social Action Ministry and among the responsible project officers was unexpected. 1.9 Basically, the project approach was to try to strengthen the existing public sector institutions without insulating housing programs from the public sector weaknesses. Questions were raised during project formulation about whether SVOA and the IPVs had the capacity to carry out all the envisaged policy reforms. The reply was that the project would provide substantial technical assistance to strengthen them. In fact, however, the TA itself was never sufficiently mobilized, let alone brought to bear on the housing institutions' ills. The project design therefore failed to give adequate attention to the question of whether even an arnple body of consultants would suffice to overcome fundamental problems of inadequate national expertise. 1.10 Another of its shortcomings (even after considerable debate about the scope and scale of the operation) was that the project size was too ambitious. Moreover, its scheme for centrlly administered operations was faulty in the face of the Federal Govenmment's waning capabilities and the provinces growing responsibilities and resources. 1.11 Because of FONAVI's poor financial performance and comparable provincial problems, the Loan provisions constructively introduced a wide array of would-be safeguards on these matters. These included new mortgage instruments and other measures to increase loan recoveries, the requirement that FONAVI revenue generation be double the size of the Bank loan repayment obligation, and linking FONAVI allocations to provincial financial improvements. In addition, release of the Loan's second tranche was hinged to progress in these improvements. These safeguards however proved inadequate to offset the effects of the hyperinflation and disarray in public finances, let alone the breakdown in government operations and discipline, during the project period. 1 .12 The preparation team aptly identified the danger of lack of continuity in the Government's willingness to carry out the necessary sector reforms as a major risk in advancing the Loan. It appeared, however, to have put undue faith in the initial enthusiasm of the Government to carry out the project. E. Project Implementation 1.13 During project implementation, SVOA experienced repeated changes in its administration, and as a result did not undertake the actions needed to implement the project and comply with loan conditions. SVOA's weaknesses  -5- reflected operational constraints which were commonplace in the social sectors: significant politicization of the administrative corps, excessively rigid regulations, poor management and lower morale, and inadequate funds. The personnel turnover became so severe that in mid-1989, the composition of the project unit was completely changed, and its very existence was threatened. The lack of stability in the personnel ranks of the executing agency was, moreover, exacerbated by SVOA's failure to hire the covenanted financial manager and the Advisory Committee he was charged with leading, despite persistent Bank requests during the course of supervision. 1.14 As the manager's post was effectively vacant throughout the project period, the intended financial management improvements were not undertaken. This may have contributed to SVOA's disbursement of US$8.5 million of project funds for ineligible expenses on one occasion. Also, without the manager, SVOA failed to complete the covenanted development of the 1989-1992 financial program. 1.15 The Advisory Committee was intended to monitor project implementation, oversee the provincial entities, and clear ex-ante all procurement actions, among other tasks. Its complete, sustained staffmg was therefore vital to the conduct of the project. But SVOA's five changes of administration during project implementation, along with a dozen Secretariat resolutions affecting the Committee's authority, prevented the latter from ever consolidating its role and responsibilities. It was not surprising therefore that there was no speedy resolution of the project issues which arose about borrowers' income limits, group savings goals, relating dwelling types to incomes rather than family sizes, subsidies, repayment terms, etc. 1.16 With the lack of the envisaged improved controls over the IPVs, there were major shortcomings in the provinces' implementation of the project- inspired rule changes concerning significant housing system operations. Govermment audit reports showed that neither SVOA nor a single IPV were in condition to proceed adequately with the implementation of the project during 1988-92. All of the foregoing deficiencies contributed to recurrent delays in project execution and lack of compliance with loan covenants. 1.17 Another significant problem was the provinces and municipalities' failure to provide the required contributions to and debt regularization with FONAVI, which comprised a main obstacle to their participation in the project. The IPVs declared their willingness to comply with, and the provinces and SVOA duly signed, agreements for the FONAVI debt regularization. However, actual provincial payments to the Fund ultimately depended on the governing provincial and municipal officers, who then faced serious financial difficulties. Their effects emerged in connection with the Loan Agreement condition calling for the recovery of provincial loan portfolios on FONAVI's "old" program. The requirement called for the reduction of payment arrears under this prior program to 40 percent in 1988, and to 35 percent later, as eligibility criteria for the project funds. In the  -6- event, there was only erratic fulfillment of this condition. Almost three years after Loan effectiveness, only 10 potentially eligible jurisdictions had met the 40 percent target (Cordoba, Chubut, Entre Rios, La Pampa, Misiones, Neuquen, Santa Cruz, Santa Fe, Tierra del Fuego, and Federal Capital). Even fewer, in essence less than 10 percent of the provinces, met the 35 percent target. 1.18 Many individual consultants were hired to carry out the TA components as well as to perform numerous line functions. With the almost constant SVOA personnel changes, the consultants themselves were also frequently coming and going. Thei work was not monitored, and virtually no results were obtained. In September 1990, the Bank requested an assessment of the consultants' performance, which SVOA failed to provide. Finally in March, 1991, at the Bank's request, SVOA cancelled all outstanding consulting contracts. It likewise agreed to employ firms for the technical assistance work instead of individual consultants in order to avoid further segmentation of tasks and to facilitate monitoring results. Moreover, it was decided that the Advisory Committee was to select the firms and supervise the implementation of the technical assistance program. The streamlined Committee was to consist of four consultants, the most important of whom would be the financial manager, along with three SVOA staff members. Notwithstanding these agreements however, since the Advisory Committee was never effectively manned for any lasting period, these arrangements were left unrealized. 1.19 Alongside these difficulties, project implementation was severely affected by Argentina's inflationary currents and other macroeconomic ills at the time. Input price changes occurred so rapidly that FONAVI borrowers repeatedly received progressively smaller loans than requested, and their declining value often disrupted the planned completion of housing repairs and improvements. These forces and borrowers' declining incomes resulted in substantially reduced FONAVI collections at the same time that its resources also lost their purchasing power. Further, these diminishing new FONAVI resources had other results which also served to compromise the project's aims. A May 1990 supervision mission found that the authorities, in order to stimulate economic activity, were sustaining FONAVI's "transition program" in all provinces despite agreements to the contrary. SVOA's action thus reduced provincial incentives to meet the new eligibility criteria for access to the Loan-supported funds, obstructing the initiation of reformed sector rules. In addition, the limited supply of counterpart funding also discouraged provincial participation, for which in June 1990, the Bank agreed to increase temporarily its cost sharing proportion from 50 percent to 70 percent. 1.20 In mid-1989, the precipitous collapse of the Alfonsin government rised questions about the future of the project. The successor Menem administration seemed sympathetic to the Bank's concems about the difficulties in project implementation. However, the project situation failed to improve while the new economic team began to prepare the way for  -7 - drastic reforms. The March 1990 suspension of Housing Bank retail activities effectively terminated this institution's direction of the home improvement loan component, with less than US$414,000 of the US$40 million Loan proceeds disbursed. Noncompliance with the loan covenants continued. It in fact worsened when, without consultation with the Bank, the new authorities in August 1991, moved to transfer FONAVI's wage tax revenues to benefit the social security system instead, a breach of the Loan conditions. 1.21 A supervision mission in mid-1991, determined that most of the agreements reached several months earlier had not been carried out, e.g., there was still insufficient progress in hiring the financial manager and several other members of the Advisory Committee. Further, because of the aforementioned factors, it became increasingly clear that the IPVs' incapacities and lack of accountability were mounting. Government audit reports in 1989-90 revealed the severity of the situation. In reviewing SVOA's books, the Tribunal de Cuentas (the national public sector accounting agency) abstained from issuing an opinion, based on the lack of an auditable base. The Sindicatura General de Empresas Publicas (Government's auditor of Federal and provincial enterprises) issued an unfavorable opinion after auditing all the IPVs. Its summary report indicated that the provincial agencies' information systems were insufficient, especially in their lack of registers and controls to provide adequate data. 1.22 Therefore,in July 1991, the Government agreed with a Bank supervision mission to not make further applications for disbursements until major loan covenants were complied with. During these discussions, it was recognized that, under the existing institutional framework, it would be extremely difficult to address all the prevailing issues. Accordingly, the Bank suggested the possibility of restructuring and downsizing the project in order to reduce the problems to more manageable dimensions, e.g., limiting its activities to three or more of the better managed provinces, along with stronger Federal controls. However, the Govermment failed to act concretely to either comply fully with the Loan conditions or propose an alternative. Therefore, the Bank concluded that there were not adequate conditions for implementing the project, and that none were likely to evolve soon. 1.23 The Bank accordingly suggested to the Ministry of Economy that the Government request cancellation of the Loan, in whose absence actions would be initiated for formally suspending disbursements. On December 12, 1991, based on the numerous covenants that had not been complied with (ref. Section m, para. 7), the Bank suspended the Loan. It also informed the Government that the cancellation of all undisbursed funds would proceed on January 13, 1992 unless the Bank then received satisfactory evidence of compliance with the Loan Agreement obligations. 1.24 It required some ten months to conclude these arrangements, initially delayed at the request of the Minister of Economy, and in response to the  public protestations which resulted from the suspension of disbursements. On January 13, 1992, the cancellation was postponed for ninety days. On March 3, 1992, the cancellation was agreed upon in meetings in Argentina. This decision was confirmed on June 9, 1992, when the Minister of Economy wrote the Bank to request the cancellation of US$270 million of the Loan. However, at the time, there was US$15 million outstanding in the project's special account, which needed to be reimbursed, failing evidence of disbursements for eligible expenditures. The resolution of this issue took until November 9, 1992, when the Minister of Economy wrote to confirmn the respective amounts to be reimbursed to the Bank and cancelled. On December 2nd, the reimbursements (which increased to US$278,250,880.14 in the final reckoning) were received and the Loan was closed. Consequently, only US$21.7 million was disbursed, of which 93 percent went for housing works with the remainder used for technical assistance and audit services. In this connection, the administrative (non-legal) process of trying to cancel the Loan showed that the Bank's machinery for these purposes is inadequate. F. Lessons Learned From the Implementation Process 1.25 The Bank was never able to make the envisaged impact on the sector. The operation was unsatisfactory, mainly for the following reasons: (a) There was an under-estimation in project design of the weaknesses of the institutional structure, which the conditions of the Loan proved inadequate to remedy. However, even if these conditions had been more rigorous, it is unlikely that they could have overcome the influences of political circumstances and the further deterioration of the economy. (b) Provincial interest in the project progressively waned, especially as there proved to be insufficient incentives to encourage them to comply with the loan conditions. Perhaps this was because the amount of funds which they could have received from the loan were relatively small compared to the magnitude of regular FONAVI resources estimated to be provided during the project implementation period. Throughout much of this period moreover, the recurrent bouts of high inflation eroded the value of FONAVI lending to the extent that it became financially unattractive. Further, there were increased provincial fiscal problems at the time, which also contributed to diminished interest in participating in the scheme. (c) The Federal Government's centralized administrative system proved unable to govern a substantial number of individual investments in this large, regionally diverse country. However, FONAVI's inability to control the IPVs was not a unique  9- circumstance. Argentina's resurgent federalism, the provinces' greater financial autonomy after the January 1988 Coparticipation Law, and the progressive decentralization of most public services to them have stringently limited the possibilities of successfully implementing national projects, especially those involving all provinces. (d) FONAVI's problems and this project experience also demonstrated that the state had, in many respects, lost its capacity to effectively direct day-to-day operations, and needed more drastic medicine than internal management improvement campaigns. They also showed that the public administration had become increasingly unable to apply equitable rules for dealing with the diverse favored beneficiaries of many programs, and how entrenched these groups had become. 1.26 The project, with the benefit of hindsight, was too ambitious an undertaking, particularly for the first Bank involvement in a large sector which is also highly susceptible to political manipulation. The restructuring of entrenched, highly inefficient Argentine Government institutions in and of itself would not have been easy in any circumstance. Confining the project to fewer provinces and project components would have been better. But the approach taken in designing the project was greatly complicated by the uncertainty of the economic environment, and the difficulties of assessing how its conceivable changes would affect the course of the operation. Unfortunately, the risks of launching this project eventually proved greater than the possibilities of its success. 1.27 In this connection incidentally, it bears noting that, during project preparation, the Bank sought to obtain agreement that the ultimate sector policy objective should be to get the Government out of the business of building houses, to be achieved by stimulating the development of a self- sustaining finance system based on voluntary savings. The project experience in fact demonstrated that direct mortgage lending operations are exceedingly difficult for a government agency to administer, particularly given the pressures for dubious subsidies, uneconomic cost recovery provisions and inadequate savings mobilization. Prior to negotiations however, the Bank accepted that the Government Policy Letter should not include a definite timetable for public sector disengagement from direct construction. The Letter instead conditioned the transfer of responsibility to the private sector on the emergence of sufficient stability in the financial system to permit such financing. In retrospect, the disappointing experience of the project confirms the view that, to the extent possible, Government's role in the housing sector should be kept less operational and restricted to enabling and regulating the market, while also providing justifiable subsidies through direct budget support. However, subsequent developments in the country confirm the validity of the Argentine authorities' opinions then that i  - 10 - precise deadline for phasing out public housing financing needed to await even more fundamental financial sector reforns. 1.28 A general lesson from this experience is that sector reform projects can be undone by the inconsistencies between pursuing rapid radical changes in policies and operating methods while seeking simultaneously to meet unmet, urgent social needs. G. Project Results and Sustainability 1.29 None of the project objectives were achieved. H. Bank Perfornance 1.30 The Bank's work was mainly flawed in the design of the project for the reasons indicated above, as well as in replicating the "early generation" of urban projects. As a result, the project had too many components and implementing agencies, and excessively weak coordination arrangements for the nature of Argentina's problem. In addition, the project's targets were over-ambitious. After Loan approval in late 1988, there were semi-annual supervision missions throughout 1989-91, which consumed almost 23 staff weeks of Bank resources p.a. on average, followed by the clean-up work in FY92. In all, 87.2 staff weeks were spent on supervision. In this period, partly because of changes in organizational responsibilities within LA4, three different task managers were responsible. During supervision, the Bank could have been more persistent at the initial stages but the turnover in task managers does not appear to have been a significant factor. I. Borrower Performance 1.31 The Government's advocacy of the project and commitment to achieve its objectives implied its agreement with the planned reforms in sector policy and operations. Clearly though, it failed to implement them. Little was done to instill or strengthen all of the project agencies' authorities and capacities. In particular, the results showed that SVOA was not equipped to manage the project, and probably should not have been selected for this role. But once chosen, after SVOA's weak performance became so manifest, the managerial role might either have been reassigned to an institution with more of a stake in the project's success--or SVOA better equipped to carry it out. Most of the project participating agencies in fact never had enough autonomy. Nor was enough done to define and carry out an effective division between central and local governments. However, all these deficiencies pale by comparison with the devastation wrought by the rapidly deteriorating economic and political conditions in Argentina at the time, which did not permit the introduction and maintenance of the envisaged reforms.  - II - J. Proiect Relationships 1.32 Bank relationships with the Government began well with the mutual enthusiasm for the project. However, there were problems beginning with the initial implementation delays, and these continued with the lack of continuity and commitment by the responsible government agencies. These were minor factors though compared to problems caused by the inadequate institutional framework and the inauspicious circumstances for the endeavor. K. Consulting Services 1.33 The use of consultants in this project to assist SVOA to carry out routine program functions was necessary to maintain these operations at all, given the disarray in the public sector at the time. However, the heavy reliance on them as a short-term solution proved to be weak, if not self- defeating. As detailed in para. 19 above, the consultants failed to constitute an adequate talent base and fount of institutional strength. They themselves in fact became one of the project's main weaknesses, especially in the use of individual consultants. The Borrower's report on the project (Part II of this PCR) states that the consultants, rather than helping to nurture and reinforce the Government staff, became a truly independent entity and progressively tended to absorb routine tasks from the permanent personnel. L. Project Documentation and Data 1.34 Documentation for the project is generally very good. In particular, the Staff Appraisal Report was comprehensive, as well as incisive in its recognition of the project's immediate and broader risks. During the period of project execution, there were some problems involving consultants' terms of references and delays in audit reports. 1.35 This PCR relied mainly on the Staff Appraisal and President's Reports, the legal documents, aide memoires and supervision reports. It was facilitated by the Government's views on the project experience. Project databases provided information for the preparation of Part I and Part m.  - 12- PART II: REVIEW OF THIE PROJECT FROM THE BORROWER'S PERSPECTIVE A. Background 2.1 The First Housing Sector Project (Loan 2992-AR) was instituted, in part, by representatives of the democratic government that came into power in 1983. The objectives of this loan were to address the many structural deficiencies in the housing sector and to implement the conditions of Law No. 21.581. This law established the Fondo Nacional de la Vivienda [National Housing Fund] (FONAVI), which had a mandate to build housing for poor populations in jurisdictions throughout Argentina. 2.2 The definition of legal and administrative arrangements were initiated during the end of 1983, and were elaborated in a 1987 sectoral study. The latter diagnosed the operational and administrative shortcomings of the housing sector which, with UNDP assistance, helped establish a project design. The Loan Agreement was approved by the Bank in October 1988 and was signed by the Government of Argentina and the Bank on November 29, 1988. The Loan became effective on March 3, 1989. B. Obiectives 2.3 The main objectives of this loan was the reorganization of the housing sector. The Loan attempted to factor in political factors, program management, contractual agreements, utilization of technology, application of standards, cost structures, implementation schedules, dispute resolution mechanisms, and returns on investment. Its goals were to increase produc- tivity and the number,of permanent dwellings, and to reduce government subsidies. C. Project Developments 2.4 The problems that surfaced during the implementation of the Project were due to underestimations of the weaknesses and operating difficulties of government organizations (The Housing Secretariat and Provincial Institutions). Compliance with the basic conditions of the Loan Agreement (which FONAVI consented to, acting as a representative of the Provincial and Municipal Governments) was not achieved due to these jurisdictions' difficulties in revenue collection. This situation was due, in part, to the deterioration of economic conditions in Argentina, which impacted on project operations. Other factors were the continuous change of leadership of government organizations, the frequent rotation of consultants, and the inability of technical assistance programs to strengthen institutional structures. This last menioned difficulty was related to the suspension of disbursements in July 1991 and the request by the National Government to cancel the Loan in March 1992. Thus, investments from the available Loan  - 13- funds were extremely limited. As a result of these varied problems, the objectives of this Loan were not realized. Lessons Learned 2.5 Project operations were not successful due to the following reasons: (a) Underestimation of the weaknesses of the institutional structures; (b) Lack of foresight before (and during) the deterioration of economic conditions and the erosion of revenues caused by inflationary and hyperinflationary pressures; (c) Decreasing interest and participation in the Project and the lack of compliance with the stipulated conditions by the Provincial governments. Compared to FONAVI's available funds, the Loan funds were severely limited; and (d) Difficulties in implementing the necessary reforms in the sector. Radical changes were needed in the political and operational models in response to urgent social demands and the need to ameliorate the housing deficit. Evaluation of Bank Performance 2.6 Successive Bank technical assistance and supervisory missions helped to formulate a reasonable project design. Nonetheless, the latter had some flaws and burdensome requirements that generated some criticism during the implementation period. Problems became apparent then emanating from excessively rigorous technical, social, financial and administrative guidelines. These guidelines did not take into account the difficulties of a federal government system under which the 24 provincial governments often operate at cross-purposes with national priorities. In addition, some proposals tended to require corrections and adjustments to the Loan Agreement in order to improve the operational viability of the project. The Bank was unreceptive to these changes, primarily because of the constant change of Government managing authorities. 2.7 Bank Missions did generate the following positive results: the improvement of bidding and contract awarding procedures for public housing; the standardization of bidding and contract awarding procedures; improvements in the pricing system; and a shortening of the implementation period. These outcomes impacted favorably on the Housing Fund's other operations.  - 14 - Evaluation of the Borrower's Performance 2.8 It is evident that serious defects existed in the implementation of the project's operations, which were due to intrinsic institutional deficiencies that were overlooked in the project's design. Some difficulties arose in connection with the Housing Secretariat's supervisory role. Also, the discontinuation of operations by the Housing Fund did not prevent the jurisdictions from continuing to provide 'transition" programs (contrary to the provisions under the terms of the Loan Agreement). 2.9 Numerous changes in the composition of the Project Advisory Committee did not allow it to perform its envisaged role. Also, uncertainty as to the appointment of the Project Financial Manager affected the operations of the National and Provincial administering agencies, as noted in the audits of the Tribunal de Cuentas de la Naci6n (The National Accounts Court) and the Sindicatura General de Empresas Publicas (The General Union of Public Enterprises). These difficulties were exacerbated by the deterioration of economic conditions in Argentina which made it harder to implement and deepen the proposed reforms in the housing sector.  - 15 - Part III STATISTICAL INFORMATION 1. Related Bank Loans N.A. 2. Proiect T-metable - Identification 12/03/87 - Preparation 11/03/87 - Appraisal Mission 04/18/88 - Loan Negotiations 08/09/88 - Board Approval 10/27/88 - Loan Effectiveness 03/03/89 - Loan Cancellation 06/24/92 - Loan Closing 06/30/93 3. Loan Disbursements Cumulative Estimated and Actual Disbursements (US$ '000) Appraisal Estimate 300,000 Actual 21,749 Actual as % of Estimate 7.3 Date of Final Disbursement 03/07/91 4. Proiect Implementation Appraisal Actual Indicators N.A. 5. Proiect Bank Financinz US$ (1) Civil Works 20,119,715.16 (2) Improvement Loans 413,411.86 (3) Auditing 299,484.12 (4) Technical Assistance 940,284.00 (5) Currency adjustment (23.775.28) Total 21,749,119.86  - 16 - 6. Proiecr Results N.A. A. Direct Benefits N.A. B. Economic Imract N.A. C. Financial Impact N.A. D. Studies None. 7. Status of Major Loan Covenants LOAN AGREEMENT SECTION DESCRIPTION COMMENTS 3.05 Completion date for Could not be carried out studies under Part C (ii) before the members of of the project; six months the Advisory Committe from the Effective Date in were hired. They had respect of Provinces and not been hired. within twelve months from the Effective Date in respect of agreed number of municipalities. 3.06(b) Section 3.06(a) asked for Although signed, the SVOA to enter into tax agreements were not compliance agreements with being monitored or Provinces and with enforced by SVOA. municipalities. Paragraph (b) of such Section requests SVOA to exercise its rights under each such agreement so as to protect the interest of the Borrower and the Bank and to accomplish the purposes of the loan.  - 17- 3.08 Aggregate amount payable This issue could only be under Transitional Loans resolved after the which were in arrears should starus of FONAVI's not in FY 89 and FY 90 portfolio had been exceed 40% of Transitional determined by means of Loans portfolio of all the studies referred IPVs. In FY 91 and all to in section 3.04(a) FYs thereafter should not of the Loan Agreement, exceed 35 %. that could be carried out before the members of the Advisory Committee were hired. 4.01(a) Maintenance of separate records Not in compliance. and accounts (for SVOA, IPVs, and Financial Agents). 4.01(c) Keep proper records on SOEs Not in compliance. and allow Bank access to them. 5.01(a) Carry out FONAVI operations in Members of the accordance with sound Advisory Committee practices under the had not yet been supervision of qualified staff hired. Performance in adequate numbers. unsatisfactory. 5.03(a) Appointment of four senior Not done. staff, with qualifications and experience acceptable to Bank, for the organizational units of SVOA implementing the project. 5.03(b) Secondment of IPVs staff Not done. to SVOA, from time to time and as necessarv for purposes of efficiently carrying out the project. 5.04 Establishmnent and Members of the maintenance of Advisory Advisory Committee Committee, to assist SVOA had not yet been in monitoring the project hired. and to coordinate its activities of a technical nature among the Borrower, the Bank. IPVs and the consultants referred to in Section 3.04(c). Such  - 18- Committee would consist of four consultants, with qualifications, experience and terms of reference acceptable to the Bank, and at least three other members, with qualifications and experience acceptable to the Bank. 8. Use of Bank Resources A. Staff Input Proiect Cvcle Total Staff Weeks Through Appraisal 39.6 Appraisal through Board Approval 33.2 Board Approval through Effectiveness 5.4 Supervision FY89 21.3 FY90 20.1 FY91 27.3 FY92 18.5 Sub Total Supervision 87.2 TOTAL 165.4 B. Missions Stage of Date Number of Number of Project Cycle Days Persons Pre-appraisal 02/16/88 22 5 Appraisal 04/29/88 Supervision 11/30/88 14 3 07/31/89 5 2 05/02/90 6 2 09/03/90 11 3 02/25/91 11 2 07/21/91 10 2 09/30/92 5 1 09/06/93 4 1  - 19- C. Costs Proiect Cycle Total $ Cost of Staff Inputs Through Appraisal 76,150.80 Appraisal through Board Approval 63,843.60 Board Approval through Effectiveness 10,384.20 Supervision 167,685.60