Colombia - First Urban Development Project

Document of The World Bank FOR OFFICIAL USE ONLY Report No. 13278 PERFORMANCE AUDIT REPORT COLOMBIA FIRST URBAN DEVELOPMENT PROJECT (LOAN 1558-CO) JUNE 30, 1994 Operations Evaluation Department 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. CURRENCY EQUIVALENTS Official Parallel Rate Rate Year ($b/US$) ($b/US$) 1979 25 N/A 1982 196 N/A 1983 500 N/A 1984 9,000 25,000 LIST OF ABBREVIATIONS AND ACRONYMS BP - Banco Popular CAIP - Integrated Preschool Care and Nutrition Center CDV - Centro de Desarrollo Vecinal (subsequently renamed - CIP - Centro de Integracion Popular) CFP - Corporacion Financiera Popular DNP - Departamento Nacional de Planeacion DRI - Integrated Rural Development Project ICBF - Instituto Colombiano de Bienestar Familiar ICCE - Instituto Colombiano de Construcciones Escolares ICT - Instituto de Credito Territorial INSFOPAL - Instituto Nacional de Fomento Municipal IPC - Programa de Integracion de Servicios y y Participacion Comunitaria MAC - Basic Module of Health Care MEN - Ministerio de Educacion Nacional MINSALUD - Ministerio de Salud Publica PAN - National Nutrition Program SENA - Servicio Nacional de Aprendizaje SIP - Secretaria de Integracion Popular FISCAL YEAR January 1 - December 31 FOR OFFICIAL USE ONLY THE WORLD BANK Washingon, D.C. 20433 U.S.A. Office of Director-General Operations Evaluation June 30, 1994 MEMORANDUM TO THE EXECUTIVE DIRECTORS AND THE PRESIDENT SUBJECT: Performance Audit Report on Colombia First Urban Development Project (Loan 1558-CO) Attached is the "Performance Audit Report on Colombia - First Urban Development Project (Loan 1558-CO)" prepared by the Operations Evaluation Department. The project was a response to a long period of urban growth/migration that changed Colombia from a predominately rural nation into one that was about 70% urban. A Government survey showed that lower-income urban populations were not receiving their fair share of health, schooling, child care, and other services, and noted that the homes of the poor in Colombia's cities lacked the most basic amenities. The purpose of this loan-the first Bank financed urban project in Colombia-was to provide social services, credit and infrastructure to large groups of low-income households in a highly coordinated manner within 23 intermediate size cities. In addition, parallel efforts at increasing employment and productivity were included in the design. The project was uniquely comprehensive, attempting to solve most urban problems in the target cities once and for all. Poor urban families received a carefully coordinated mix of services provided in large community centers constructed for that purpose. The PCR provides a brief but full account of the project experience; it is informative on the achievements and on the compliance with covenants. The PAR elaborates on the achievement of institutional objectives and on project design. The Audit focussed on two factors: first, participation of the communities in the management of the centers, and second, the way the community makes use of the center facilities. The PCR rated this project as unsatisfactory, largely due to doubts about its financial and institutional sustainability. In spite of numerous organizational changes the program has remained operational. While the future is largely unknown, it would be fair to say that the project has already exceeded the original expectations. Its recurrent costs for staff and center upkeep have proven to be surprisingly modest. And the project has begun delivering new services-notably in the areas of senior citizen activities and municipal recycling-which were not noted at the time of the PCR. Based on the success of the project in achieving its main objective and delivering a comprehensive basket of basic services in urban neighborhoods where they had previously been scarce, a rating upgrade is now warranted. Thus, this Audit rates the project as satisfactory. Robert Picciotto by H. Eberhard K6pp Attachment This document has a restricted distribution end my be used by recipients only in the performnce of their official duties. Its contents my not otherwise be disclosed without Morid Bank authorization. 6 FOR OFFICIAL USE ONLY PERFORMANCE AUDIT REPORT COLOMBIA FIRST URBAN DEVELOPMENT PROJECT (LOAN 1558-CO) TABLE OF CONTENTS RAN PREFACE .......................................... BASIC DATA SHEET .................................. EVALUATION SUMMARY............................... v 1. OBJECTIVES AND METHODOLOGY ..................... 1 The Audit as a Rapid Evaluation of Impact (REI) .............. 1 2. BACKGROUND. .................................... 1 The Project Context and Urban Development ................. ..I... 1 Institutional Arrangements. .................................. 3 3. THE PROJECT ..................................... 4 Issues in Project Preparation ........................... 4 Project Objectives and Appraisal ........................ 4 Financial Arrangements .............................. 5 Institutional Development ............................. 5 4. PROJECT IMPLEMENTATION AND RESULTS ............... 6 Physical Implementation Experience ...................... 6 Analysis of Selected Components ........................ 8 Project Results ................................... 10 Operation Experience ............................... 10 Compliance with Operational Directives and Covenants .......... 11 This report was prepared by Alcira Kreimer (Task Manager), and Ronald Parker (Consultant) who audited the project in January and February of 1994. Elizabeth Tweddle provided administrative assistance. This document has a restricted distribution and may be used by recipients only in the performance of their official duties. Its contents may not otherwise be disclosed without World Bank authorization. TABLE OF CONTENTS (Coat) Page No. 5. CONCLUSIONS ................................................. 12 Significant Impact in Spite of Over-ambitious Objectives ............... 12 Project Design, Points of Special Interest ........................... 13 Lessons Learned ............................................. 14 1 PERFORMANCE AUDIT REPORT COLOMBIA FIRST URBAN DEVELOPMENT PROJECT (LOAN 1558-CO) PREFACE This is a Performance Audit Report (PAR) on the First Urban Development Project, benefitting low-income and squatter households in 23 medium size cities, involving a World Bank Loan amounting to US$24.8 million equivalent to the Government of Colombia. The loan was approved on April 22, 1978; US$18 million was disbursed by August 1984, when US$6.8 million was reallocated to the Popayan Reconstruction Project. The closing date was extended for funds used in that project, while the urban development component covered in this report completed on December 31, 1984. The PAR is based on the Project Completion Report (PCR) prepared by the Infrastructure Operations Divisions of the Latin America and Caribbean Country Department III and issued on April 20, 1988 (the borrower did not provide Part II), the Staff Appraisal Report (SAR), the loan documents, and study of the project files. An OED mission visited Colombia during January and February of 1994. The excellent cooperation and valuable assistance provided by the Fondo de Solidaridad y Emergencia Social (Sub-direccion I.P.C.) and the Departamento de Planeacion Nacional in the collection of information for this report is gratefully acknowledged. The PCR provides a brief but full account of the project experience; it is informative on the achievements and on the compliance with covenants. The PAR elaborates on the achievement of institutional objectives and on project design. Following standard OED procedures, the draft PAR was sent to the Government and the Borrower for comments. However, no comments were received.  11 PERFORMANCE AUDIT REPORT COLOMBIA FIRST URBAN DEVELOPMENT PROJECT (LOAN 1558-CO) BASIC DATA SHEET KEY PROJECT DATA Appraisal Actual item Expectations Estimates Total Project Cost (US$ million) 62.0 65.9 Underrun or Ovenm (%)f_ 6.3 overrun Loan Amount (US$ million) 24.8 24.8 Disbursed 18.0 Reallocated' 6.8 Cancelled - Date Physical Components Completed 12/31/81 12/31/83 Proportion Completed by above date (%)' 30 95 Time Overrun (%)5 Financial Performance Moderate Problem Institutional Performance Moderate Problem Economic Rate of Return (%) 10 10 Shelter Components Water Supply Component 10 10 I' In relation to adjusted project cost of US$63.9 million cost overrun was 3%. b Loan balance allocated to Popayan Reconstruction Project - Loan 2379-CO. ? 95% based on adjusted physical targets and appraisal targets due to increase in water supply component under adjusted scope. ill CUMULATIVE ESTIMATED AND ACTUAL DISBURSEMENT (US$ million) FY79 FY80 FY81 FY82 FY83 FY84 FY85 Appraisal Estimate 7.1 13.5 19.7 24.8 Actual (US$ million) 0.7 1.9 5.1 10.6 14.6 18.0 Actual as S of Estinfe 5.2 9.6 20.6 42.7 58.9 72.6!' Date of Final Disbursement- July 1986. Canceled $0.3 at closing on 6/15/88. ' Excludes US$6.8 million allocated to Popayan Reconstruction Project - Loan 2379-CO. PROJECT DATES Item Original Plan Actual First Mention in File. 04/08/85 Appraisal 07/76 02/77 Negotiations 11/29/77 02/28/78 Board Approval 01/17/78 04/22/78 Loan Agreement Date 07/21/78 Effectiveness Date 12/01/78 Loan Completion 06/30/82 12/31/84 Closing Date 06/30/85 STAFF INPUTS (Staff-Weeks) FY FY FY FY FY FY FY FY FY FY FY FY FY 76 77 78 79 80 81 82 83 84 85 86 87 88 Preappraisal 16.7 38.0 Appraisal 58.1 16.0 Negotiation 5.1 Supervision .7 36.5 51.6 38.4 27.1 12.8 12.8 4.1 8.3 .1 .5 Other .3 3.1 14.1 TOTAL 17.0 99.1 35.9 36.5 51.6 38.4 27.1 12.8 12.8 4.1 8.3 .1 .5 iv MISSION DATA Stage of Month No. of Project Project Cycle Year Persons Staff Week Rating Identification 4/75 2 2 Preparation 9/75 3 3 Preparation 2/76 2 5 Preparation 5/76 4 7 Preparation 7/76 5 11 Preparation 12/76 2 4 Appraisal 2/77 7 21 Appraisal 5/77 5 7 Appraisal 9/77 1 1 Supervision I 7/78 4 11 1 Supervision II 11/78 2 4 2 Supervision III 1/79 3 5 2 Supervision IV 5/79 1 3 2 Supervision V 8/79 5 10 2 Supervision VI 10/79 1 1 Supervision VII 12/79 4 7 2 Supervision VIII 6/80 4 7 3 Supervision IX 10/80 4 6 3 Supervision X 4/81 4 7 3 Supervision XI 6/81 3 4 3 Supervision XII 1/82 2 3 2 Supervision XIII 4/82 2 2 2 Supervision XIV 12/82 2 3 2 Supervision XV 9/83 1 1 2 Supervision XVI 5/84 1 1 1 OTHER PROJECT DATA Borrower: Republic of Colombia Executing Agency: Secretaria de Integracion Popular Follow-up Project: Cartagena Urban Development (1694-CO)  V PERFORMANCE AUDIT REPORT COLOMBIA FIRST URBAN DEVELOPMENT PROJECT (LOAN 1558-CO) EVALUATION SUMMARY Introduction Objectives 1. As per Procedural Guidelines for 2. The project assessed in this report was a Operations Evaluations Staff dated May 1990, response to a long period of urban the main objective of this audit is to arrive at an growth/migration that changed Colombia from a overall evaluation of project performance predominately rural nation into one that was including its technical, economic, financial, social about 70% urban. The homes of the poor in and environmental aspects. In light of the fact Colombia's cities lacked basic amenities, and a that this loan was closed about nine years before Government survey showed that lower-income this assessment, the Audit also included a rapid urban populations were not receiving their fair evaluation of the social and physical share of health, schooling, child care, and other sustainabilitity of the project. Thus, this Audit services. The purpose of this loan-the first evaluates the impact of this urban development Bank financed urban project in Colombia-was project implemented in 22 cities since 1978 to provide social services, credit and through detailed observations in three cities. infrastructure to large groups of low-income The research undertaken for the Audit focusses households in a highly coordinated manner on two factors: first, participation of the within 23 intermediate size cities. In addition, communities in the management of the centers, parallel efforts at increasing employment and and second, the way the community makes use productivity were included in the design. Thus of the center facilities. In some instances, the project was uniquely comprehensive, current conditions were compared to the attempting to solve most urban problems in the photographic record maintained by the target cities once and for all. The objective was community centers. In all cities studied key to facilitate the economic and social integration informants, semi-structured interviews, of some 200,000 marginal families into the life secondary data review, direct observation and of their surrounding cities. The families would walking/riding transects were utilized during the receive a carefully coordinated mix of services audit process. provided in large community centers (referred to as CDVs) constructed for that purpose. vi 3. A principal objective of government Consequently, the Audit mission went to policies during the period was to require Colombia with very low expectations, and increased cooperation at the national, regional received the pleasant surprise that the project and local levels, which sometimes led to was still quite operational, possibly even multisectoral programs of this type. Loan 1558- achieving certain of its original objectives more CO, often referred to in the project files as "the effectively now than at any time during marginalized urban areas project," would require disbursement. the coordinated action of nine separate ministries and agencies. Careful citing of the Implementation Experience community centers, policy coordination at the national level, and service coordination at the 5. The Borrower was the Republic of local level within them once constructed were Colombia, and implementation was to be done expected to increase the effectiveness of through SIP, which was to coordinate work by community outreach in a manner which had not the Ministry of Education, Ministry of Health been possible through the work of nine single- and other project entities. The Bank loan was focus agencies working independently of each intended to cover 40% of estimated costs with other at dispersed and difficult to locate offices the other 60% to be covered as follows: in the city centers. In a sense, the project was Government (32%), Banco Popular (3%), ICT a pioneering attempt at decentralization, which (20%), ICBF (2%), and SENA (3%). The enjoyed some of the advantages of that project was declared effective December 12, orientation, even though it was severely 1978, with conditions of disbursement (primarily constrained because the bureaucratic milieu the development of the Area Implementation under which it operated remained highly Plans) being met about one year later in a little centralized. over half the cities. The Bank did not approve 100% of the Area Plans until 1982. Following 4. The marginal areas the project was loan effectiveness the project took five years to attempting to deal with included virtually all reach full completion, although the loan aspects of what continues to this day to be a remained open to allow for disbursements under difficult urban situation. All Government and the Popayan Reconstruction Project until 1986. Bank staff involved with the project knew from the outset that achieving comprehensive Results solutions to the problems of the marginalized areas in one shot was going to be difficult. The 6. As planned, the project did build 23 fact that such an ambitious project was community centers, 22 of which remain in active attempted at all is due to the fact that it was in operation. A very interesting conflict can be tune with the spirit of the time: it was in seen in the development of these centers. harmony with a policy speech made by Mr. Constructing large buildings costing US$430,000 McNamara; it mirrored Colombia's development in slum areas is not financially replicable. The goals, and respected the manner in which that project was hoping that these marginal nation had determined to address its most communities would somehow raise funds to intractable urban problem. Within the Bank, cover operating costs-which never happened. however, this operation was perceived as having On the other hand, the existence of the been problematic. Of all the people at impressive center buildings ultimately preserved headquarters in Washington contacted during the IPC program. Between 1986 and 1994 the the audit process (about 25), only two had centers have been located under five different anything positive to say about the project. Even institutional frameworks, passing through several within Colombia the fact that the centers periods of organizational limbo (some of more continue to operate is not widely known. than a year) without an operational budget and vii at times without staff salaries. Had less likely be mined for the bricks and other important buildings existed in the marginal construction materials they contain. areas, there is little doubt that the program would have been abandoned. Through the & Several of the centers are well equipped activities of this project 7,381 families received and located to play an important role in the case home improvement credits, and all of those of emergencies: the Pasto center is located near were allowed to borrow funds for title an active volcano that menaces a large sector of regularization if such was required. The project its target area; the Tumaco center can provide proved highly successful in delivering basic emergency services and shelter in the event of services to urban neighborhoods where they had the frequent fires, storms, earthquakes and high previously been scarce. Educational, nutritional, seas that menace the area; and the Pereira health, and child care facilities were built within center is located above a settlement along a the CDVs to appraisal standards or higher. river which has been subject to flooding in the These continue to function effectively. Health past. IPC staff in most cities serve on local services have been expanded to include dentistry disaster/emergency response committees. There in many centers. The vocational training has been a tremendous improvement in living component administered through SENA became conditions around many of the centers: home active early in the project and it provided skills improvement loans have generally resulted in a upgrading to residents of marginal significant upgrading of housing stock. In neighborhoods. There are serious questions as Pereira, a vast slum area that was mostly to whether a majority of graduates actually cardboard and wooden shacks when the center obtained jobs because of the training they opened now contains paved streets and received. No study was ever done that attractive brick houses connected to all evaluated the impact of the training on job- municipal services. seekers or counted the number that found work. These courses continue to be popular, however, Sustainability partly because many poor families can lower their own expenditures by, for example, learning 9. The data base from PCRs on completed how to make their own shoes and clothing. urban projects rates this operation as Vocational education has led to significant unsatisfactory, largely due to doubts about its micro-enterprise development. The financial and institutional sustainability. In the neighborhoods around the centers are filled with opinion of this Audit, the project now warrants small enterprises specializing in the very a satisfactory rating. In spite of numerous activities taught in the centers, and several organizational changes the program has cooperative endeavors were found at the centers remained operational. Its recurrent costs for visited. staff and center upkeep have proven to be surprisingly modest. Initial investments in high- 7. The identity of the ultimate operator of quality construction have kept costly repairs to the centers was not covered in the original plan a minimum. And the project has begun because it was hoped that the community would delivering new services-notably in the areas of take over that job. With hindsight this seems senior citizen activities and municipal naive: the neighborhoods surrounding the recycling-which were not noted at the time of centers are marginal in all senses of the word. the PCR. Clearly not all the project's objectives While it appears that the CDVs remain very were met entirely. This often happens, highly regarded by their communities, in the however, even among projects which receive a opinion of several IPC staffers, without the satisfactory rating. However, the PCR did note presence of paid watchmen the centers would that "project achievements reached or surpassed viii the revised targets set for almost all local capacity to solve local problems. The components." CDV centers were supposed to funnel assistance emanating from central agencies, eliminating 10. In the field the audit staff began to gaps and overlaps in the process. They were to question why the project should continue to be help the community to carry out a plan that rated poorly. The project proved highly would bring both economic and physical successful in delivering a comprehensive basket improvement. And they were to provide a of basic services in urban neighborhoods where mechanism to deliver rapid response to they had previously been scarce. Even the PCR emergencies and other spontaneous needs of noted that the project provided comprehensive the marginal communities. The project assistance to the urban poor in 23 cities and that continues to largely achieve these aims, the project had improved their living standards particularly as a result of improved coordination and productivity. In the last five supervision between the agencies which work out of the missions, project ratings showed a steady centers. This project was designed with major improvement (terminating with a rating of 1), flaws: it attempted to do too much at once, and perhaps indicating a shift in staff orientation the activities that were employed were unlikely from center construction to center operation. to solve the problem of marginality. At best, a Although its original objectives may have been project directed at the health, education, vaguely formulated, they have largely been employment, housing and infrastructure achieved. The centers continue to function, problems of the marginal areas could only services are being provided in a coordinated succeed in reducing them. Solving each of the fashion, CDVs can be considered to be more significant problems of Colombia's cities "beneficiaries' organizations," and the GOC has will involve several loans before the majority of continued to monitor the program and some observers will be persuaded that a solution is at improvements have been made, although more hand. consistent funding would doubtless have led to greater advances. The community centers and 12. Although the activities undertaken under the various projects run through them by the loan CO-1558 attained largely positive IPC have demonstrated remarkable institutional outcomes, there are three aspects of the loan's and stream-of-benefits sustainability and basic approach which deserve mention: (i) it therefore merit a rating upgrade to satisfactory failed to work within the reality of the at this time. governance system; (ii) it proved largely impossible to monitor; and (iii) it failed to Findings and Lessons achieve the form and quality of community participation anticipated. Interagency 11. It would be a shame if the only lesson coordination was never achieved in the manner which was learned from the experience with this anticipated and the management structure Put loan is that working with the most vulnerable together for the project proved inadequate to populations in the large and growing cities of the job of constructing the facilities while the developing world is difficult (and therefore implementing a highly complex program aimed bad for one's career). The world's cities have at a difficult sector of the population-hardly grown to unprecedented size and the risk of surprising since there was no pilot experience urban catastrophes has multiplied. For that upon which to build. The use of standardized reason alone, the lessons listed below should not plans not tailored to site and climatic conditions be interpreted as suggesting that urban resulted in construction delays, and poor programs are better off not addressing the construction supervision was largely responsible problems of the poor in a comprehensive for the collapse of one of the CDV buildings manner. The project was designed to increase (which had to be rebuilt subsequently). The ix monitoring problem proved daunting, and it was happily accept the responsibility of paying the complicated by high levels of staff turnover: expenses of the community centers; (iii) the dea Bank staff tended to spend missions with upper that one intervention could possibly solve most level staff from the project entities in Bogota of the problems in 23 large urban areas at the and, as a result, found it difficult to make the same time; (iv) the determination not to use time to visit project sites. existing local organizations (and to create new and "improved" local organizations just for the 13. Some of the conceptual leaps of faith purpose of this project); and (v) the notion that which the project design made are reflected in participating on one directive board would the following decisions: (i) constructing large empower slum dwellers enough to shift the and expensive buildings (without asking the balance of political power sufficiently to ensure beneficiaries if they wanted them) was marginal communities a significant voice in the determined to be the most appropriate modality running of the cities in the future. to promote community participation with slum dwellers; (ii) expecting the about-to-be- organized but impoverished neighborhoods to  PERFORMANCE AUDIT REPORT COLOMBIA FIRST URBAN DEVELOPMENT PROJECT (LOAN 1558-CO) I. OBJECTIVES AND METHODOLOGY The Audit as a Rapid Evaluation of Impact (REI) 1.01 As per Procedural Guidelines for Operations Evaluations Staff dated May 1990, the main objective of this audit is to arrive at overall evaluation of project performance including technical, economic, financial, social and environmental aspects. In light of the fact that this loan was closed about nine years before this assessment, the Audit also included a rapid evaluation of the social and physical sustainabilitity of the project. To this end, the Audit evaluates the impact of this urban development project implemented in 23 cities since 1978 through detailed observations in three cities. The research undertaken for the Audit focusses upon two factors, first, the participation of the communities in the management of the centers, and second, the utilization of project facilities by the communities they serve. In some instances current conditions were compared to the photographic record maintained by the community centers. In all cities studies key informants, semi-structured interviews, secondary data review, direct observation and walking/riding transects were utilized during the audit process. I. BACKGROUND The Project Context and Urban Development 2.01 The project assessed in this report was a response to a prolonged period of urban growth which had produced extensive slums largely populated by squatters. By the mid-1970s, urban migration had made Colombia about 70% urban-almost a mirror image of demographic patterns in the mid-30s when the population was more than 65% rural. The rapidly growing cities developed severe problems while the Government was preoccupied with promoting employment opportunities and achieving sustainable economic growth. Once the Government turned its attention to the cities, the most intractable problems facing the nation's urban areas had become those related to the situation of low-income neighborhoods. The years immediately prior to project identification were characterized by economic transition as well. During the latter half of the 60s and through the 70s, the traditional importance of the largely coffee-based agricultural economy declined as a result of Government's export promoting policies which made the national economy more industrial and oriented to international trade. Merchandise exports increased four-fold, while the relaxation of exchange controls and increased investment yielded a 6.5% annual average GDP growth between 1967-74. 2 2.02 The poor living in Colombia's cities lacked basic amenities: one-forth of the urban population did not have access to piped water and two-thirds lacked sewerage. Although education and health services had spread rapidly, and birth control had been a national success story, the quality of services aimed at the urban poor was far worse than their high costs seemed to justify. A study conducted by the National Planning Department revealed that the lower income neighborhoods were not receiving their fair share of health, schooling, child care, and other services. Before tackling these problems in the largest cities where they would be most entrenched, public expenditures were to be oriented towards investments and infrastructure in the intermediate cities. The nation was unusual in that growth was dispersed over a large number of intermediate cities, because of geography and the difficulty of travel between regions prior to the development of air travel. 2.03 This project was a pioneering attempt at decentralization, which enjoyed some of the advaitages of that orientation, even though it was severely constrained because the bureaucratic milieu under which it operated remained highly centralized. In the early 1970s, Colombian officials had begun grappling with the details of projects which could improve the delivery of services and income levels of the urban poor. A principal objective of government policies during the period was "to close the gap" (the title of the 1974-78 Development Plan). To redress observed imbalances, new relationships and increased cooperation were to be required at the national, regional and local areas, resulting in multisectoral programs such as this project, and a series of measures designed to promote more effective interagency coordination. Because each of the intermediate cities was so different, the idea of decentralization appealed to policy makers, although its achievement ultimately had to wait for constitutional changes which only took place in the 90s, long after the loan had closed. 2.04 The impact of this project was expected to reach beyond its target cities. Policy objectives reflected a strong concern with slowing the growth of the nation's four largest cities (which were not directly to be served by proposed project activities). It was thought that by generating employment, upgrading housing, regularizing land titles, and providing a range of urbanization and health-related services in intermediate size urban centers, this endeavor would reduce the level of migration from the rural areas to the four largest and most rapidly growing urban areas. Another aspect of urban marginality in Colombia was that the poor had experienced no proper access to credit markets. Loan 1558-CO (often referred to in the project files as "the marginalized urban areas project") was planned to include investments in 27 zones located in 23 cities and its achievement would require the coordinated action of nine separate ministries and agencies. Within each city, the project concentrated on the provision of multiple services through one large building to be constructed in areas easily accessible to the residents of one or two target neighborhoods. Careful citing of these community centers and service coordination within them once they were constructed were expected to increase the level of effectiveness of community outreach which had not been possible through the work of the nine single-focus agencies when they had worked independently of each other at dispersed and difficult to locate offices in the city centers. The notion of building smaller, separate offices for each agency within the marginal areas was considered but discarded because of the difficulties inherent in obtaining so many parcels of land within the target areas. 2.05 The national development plan for the period consisted of three complementary projects which attempted to deal with the problems of those who had not benefitted from decades of economic growth: (i) DRI, an integrated rural development project which aimed at increasing the productivity of the small farmer through the provision of credit and technical assistance, as well as social services components which were intended to upgrade the quality of rural life and remove the incentive to migrate to the urban centers; (ii) PAN, an integrated health and nutrition project which 3 included subsidized high protein food, nutrition education and the provision of primary health services; and (iii) IPC, the urban zone project which aimed to provide training and otherwise improve the standard of living in urban centers, particularly of the marginalized. 2.06 Loan 1558-CO was integral to the decentralization strategy that guided the urban policy of the time. The Bank had been supporting the Government's decentralization policy in the DRI, PAN and IPC undertakings. The purpose of this loan--the first Bank financed urban project in Colombia-- was to provide social services, credit and infrastructure to large groups of low-income households in a highly coordinated manner within the target 23 intermediate cities. In addition, parallel efforts at increasing employment and productivity were included in the project's design. Thus the project was uniquely ambitious and comprehensive, attempting to solve most urban problems once and for all. In this project, the Bank's aim was to facilitate the integration of some 200,000 beneficiary families into the life of their surrounding cities. The families would receive a carefully coordinated mix of services in large community centers (referred to as CDVs) specially constructed for that purpose. The notion of cost recovery was not at the forefront of Government efforts aimed at the poor during this period. Institutional Arrangements 2.07 Historically urban projects had been planned through the National Planning Department (DPN) and implemented by the Instituto de Credito Territorial (ICT) in coordination with the various municipalities. Enthusiasm for the IPC project ran high in DPN and ICT, and discussions between them as to how the marginalized urban areas project was to be implemented did not go smoothly. In August of 1976, a number of key people involved in project preparation resigned from the DPN. Around that time, the decision was made at the highest levels that the program would be executed through the newly created Popular Integration Secretariat (SIP), established in the President's office. Given that the SIP had almost no personnel at the outset, this arrangement was modified so that DNP could continue to provide technical assistance as needed during project preparation and implementation. Although no other international financing was involved in this project, IDB, CIDA and USAID were financing DRI and PAN components. 2.08 The Borrower was the Republic of Colombia and implementation was to be done through SIP, which was to coordinate work by the Ministry of Education, Ministry of Health and other project entities. The project was initially conceived as having the following components: a "work bank" to provide employment opportunities; worker training; credit and technical assistance to small and community-based enterprises; childcare; special pre-school and primary school programs for marginal areas; health; nutrition; adult education; cultural events; recreation; literacy promotion; community participation; legal assistance clinics; installation of water/sewerage (at least to minimal levels); community groups (especially created to channel credit and social services); and housing credit. A national council (Consejo Nacional de los Centros de Desarollo Vecinal) composed of representatives of SIP, DNP, Ministry of Health, Ministry of Education, ICBF, SENA, ICT, INSFOM and CFP was to handle inter-agency coordination. Local issues would be handled by a local council consisting of all the above plus municipal officials. In each urban zone (and CDVs often served more than one neighborhood) a local board (Junta Directora) was established. 4 III. THE PROJECT Issues i Project Preparation 3.01 The Government had little relevant experience working with marginalized urban areas upon which to build, and the Bank had not made a similar loan previously in the region. Nevertheless, latent management problems inherent in a program of this nature were recognized very early on-- some were clearly apparent already during project preparation. Supervision was going to be a problem because of the geographic separation of the 23 cities: they were located in 17 different departments. Even though the fact that the project would require more supervision than normal was noted, no mechanism that would allow this to happen was devised. Supervision by sampling was considered but rejected as unlikely to work since there was no reason to assume that conditions in the various cities were similar in terms of their likely impact on implementation. The success of this project was clearly going to depend on DNP and SIP being able to coordinate--something that had never happened before. Project preparation was slow for a variety of reasons. The resignations of key DNP staff members and the time required to make the new SIP a functioning organization slowed initial actions. Because many ministries/project entities were involved, lengthy discussions regarding administrative structure, coordination, overall project objectives, CDV building design, etc., had to take place. The project was largely based on a DNP study conducted 10 years prior to implementation which had ranked 58 cities according to the incidence of poverty, population growth rate and the need for services. Later preparation work concentrated on mechanisms that would be required for the delivery of services to the most needy 44 of those cities through the CDV community buildings. 3.02 During the preparation stages, work in the Cartagena area, originally ranked as high priority by the DNP study of marginalized urban areas, was appraised as a separate loan (Loan 1694-CO). Because of the unique physical environment near flooded and swamp areas around Tesca Lagoon-- which required additional engineering studies before land reclamation work could begin --the other 23 cities were processed separately. To offset the difficulties inherent in supervision and to remedy the outdated nature of the preliminary study, the development of area plans for each CDV that identified the works to be performed, prioritized them, scheduled their implementation, and budgeted expenditures in a manner that would ensure that the resources would actually reach the target beneficiaries, was a condition of disbursement. In requiring the preparation of these quite detailed plans, the limited implementation capacity of the newly created SIP may not have been fully taken into account. Financial implications and the difficulties inherent in supervising a project spread over such a large number of cities led the DNP to opt. for a three-phase approach: the IPC project was to begin with 23 cities, the second phase was to include the Cartagena Southeast Zone Rehabilitation Project (Loan 1694-CO) and the third phase was to consist of all the rest. Only the first two phases of the three phase IPC project were actually ever implemented. Project Objectives and Appraisal 3.03 The objectives of the project were: (i) to provide comprehensive assistance in several mutually supporting fields to about 550,000 urban poor in 27 zones of 23 cities (Armenia, Barranca Bermeja, Bucaramanga, Cartago, Cucuta [4 zones], Girardot, Ibague, Rio Hacha, Manizales, Moneria, Neiva, Pasto, Pereira [2 zones], Quibdo, Sabana Larga, San Andres, Santa Marta, Sincelejo, Tulua, 5 Tumaco, Turbo, Valledupar, and Villavicencio) in order to improve their living standards and productivity; (ii) to base the delivery of services increasingly on the beneficiaries' organizations; (iii) to strengthen the capability of the Government to coordinate the activities of the various executing agencies in order to ensure attainment of the project's objectives; and (iv) to enable the Government, through continuous monitoring and evaluation, to improve the program. The project included the construction of a community development center (including recreational space/parkland) in each of the 27 areas. The project was appraised on the basis of technical standards and a preliminary study of all intermediate urban areas, supplemented by more detailed pre-feasibility studies conducted in 6 project zones. Additionally, sectoral agencies presented prototype models of their proposed facilities. Several project preparation tasks were left for implementation. Appraisal was originally scheduled for August of 1976 but postponed until February 1977, largely because of the shift from DNP to SIP. Two post-appraisal missions in May and September of 1977 were required to complete preparation of the potable water component. The SAR never described in detail how this project would be monitored, and Bank staff were unable to visit even a third of the project sites during implementation. Financial Arrangements 3.04 The Bank loan was intended to cover 40% of estimated costs with the other 60% to be covered as follows: Government (32%), Banco Popular (3%), ICT (20%), ICBF (2%), and SENA (3%). Banco Popular entered into a contract with Corporacion Financiera Popular under which the former made a US$2,000,000 commitment to cover its contribution to the project. Although the Bank loan exceeded foreign exchange costs and covered US$10.3 million equivalent of local currency costs (excluding the contributions of beneficiaries of the credit program to the cost of the specific investments to be financed with such credits), in view of Colombia's progressive development policies and domestic resource mobilization efforts, local-cost financing by external agencies was deemed by the Bank to be justified. US$10.9, US$5.0 and US$2 million equivalent of Bank loan proceeds would be transferred by the Government to ICT, INSFOPAL, and CFP respectively, as equity contributions. Institutional Development 3.05 The equity contributions mentioned immediately above with respect to ICT, INSFOPAL and CFP not only enabled these agencies to participate in the project, but were intended to increase their capacity to participate in similar projects subsequently. Each of the CDV buildings was to house the following components: health post or center, pre-school, primary school, nutrition center, cafeteria, cooperative market, post office, police station, community hall, library, sports facilities, technical training classes, non-formal education center and community awareness/outreach administration. These would not only increase the capacities of the various ministries and agencies (SIP, MoH, MoE, ICBF, SENA, ICT, INSFOPAL and CFP) to deliver services to important but hard-to-reach target populations, but also improve communications between them. The active participation of the community within the coordinating structure at each CDV was intended to ensure that marginal groups would have an influence commensurate with their numbers, in effect the project was to create an organizational structure through which slum communities could channel their concerns, which would ensure an ongoing governmental commitment to deliver services at the levels contemplated by project designers in perpetuity. 6 4. PROJECT IMPLEMENTATION AND RESULTS Physical Implementation 4.01 The PCR for this project mentions that service coordination was never achieved in the form and quality anticipated and the management structure put together for the project proved inadequate. It might be asked what new organization could have been expected to handle the range of tasks (i.e., building community centers, coordinating service delivery with nine agencies, creating community organizations, and commencing credit, housing and infrastructure sub-components) that were required to initiate the full gamut of project activities. Not only was there no pilot service coordination experience upon which to build, but the limited capacity available was absorbed by the development of detailed Area Implementation Plans which had been mandated. In addition, the PCR notes that not only did the implementation schedule lag far behind appraisal projections but it also differed from the sequential order of completion per city (PCR, para. 3.06). The Loan Agreement was declared effective December 1, 1978, with conditions of disbursement (primarily the development of Area Implementation Plans) being met about one year later in a little over half the cities. The Bank did not approve 100% of the Area Plans until 1982. Following loan effectiveness the project took five years to reach full completion, although the loan remained open to allow for disbursements, through an amendment of the Loan Agreement dated August 2, 1984, under the Popayan Reconstruction Project until 1986. 4.02 Main concerns identified by the PCR include the following: (a) Financial Reports. The lack of implementative/administrative capacity within SIP was reflected in a lack of systematic reporting. Accounting deficiencies demonstrated poor financial management and audited accounts were only submitted to the Bank after considerable delay (PCR, para. 3.11). (b) Institutional Weaknesses. CFP had difficulty working in cities where it did not have offices: disbursements were less than expected for small scale loans, and several enterprises were refused loans because of lack of collateral (PCR, para. 3.13). ICT was unable to hire staff to assist the municipal governments within the land tenure regularization components (PCR, para. 1.14) and these objectives were not achieved. Additionally, the performance by INSFOPAL and Municipal Empresas was less than satisfactory (PCR para. 1.14). (c) Cost Recovery. In general the cost recovery objectives of this project were not met, and the lack of cost recovery data is an indication of the lack of ability of SIP to properly monitor this project (PCR, para. 4.06). (d) Evaluation. The evaluation of the project was subcontracted with Instituto SER. The performance of the evaluation was disappointing and baseline data which would allow the rigorous documentation of the project's impacts was never obtained (PCR, para. 3.15). 4.03 The use of standardized plans not tailored to site and climatic conditions resulted in construction delays, and poor construction supervision was largely responsible for the collapse of one of the CDV buildings, which had to be rebuilt. The marginal areas project made little progress of 7 note toward its main goals of integrating the delivery of services and promoting significant popular participation in its first three years after Board approval. The various project entities were centralized bureaucracies which usually prepared their budgets and workplans at least one year in advance. Although the national and local committees met, and discussed what needed to be done, the various representatives were powerless to change organizational decisions that were set in concrete. During the second and third year SIP began to make some progress with the construction of the community centers, and advances were noted in the vocational training effort and primary education. Infighting between DNP and ICT, budgetary shortfall and delays in hiring key staff at SIP contributed to the slow start. From the outset the project was plagued by high staff turnover: key staff resigned from DNP during the planning phases; implementation saw five different directors (coupled with changes in senior staff) at SIP; staff and management discontinuity reduced the effectiveness of INSFOPAL and ICT; and there were four task managers on the part of the Bank. Engineering work for water supply components was not begun until late 1979. 4.04 By June of 1980 only US$0.75 million had been disbursed, about 5% of the appraisal estimate-which led to a reformulation of budgeting and transfer procedures on the Government side. When the likelihood of reaching the 23 target cities seemed remote, a series of proposals and counter-proposals for the reformulation of the project were exchanged beginning in the second half of 1980, with no decision made until early 1982. Ultimately, the number of cities did not change, although four zones within cities covered under the original plan were dropped. The Borrower and the Bank were in accord with regard to all but one of the revisions which were adopted. There were other problems which stemmed from outdated information contained in the DPN study used for project preparation which was conducted 10 years prior to implementation. For example, when construction activities initiated, it was already clear that several cities, far from growing at a dangerous rate, were clearly losing population. Investments in those cities were not going have an impact on urban demographics. Nevertheless, they received their CDV. 4.05 Although the activities undertaken under Loan CO-1558 attained largely positive outcomes, there are three critical aspects of the loan's basic approach which deserve mention: (i) it failed to work within the reality of the governance system; (ii) it proved largely impossible to monitor, and (iii) it failed to achieve significant community participation in CDV operations. As was noted above, the coordinating mechanisms designed for the program proved to be nearly useless because most organizations were centralized and could not change their activities, even where local conditions were such that certain actions were proving to be counter-productive. However, when there were no more Bank funds being disbursed, the national and local coordinating bodies stopped meeting and some real coordination began to take place between the staff of the various entities engaged in community activities at the local level. There were many ways in which the project attempted to bypass the governance structure, assuming away political and institutional constraints. Some examples: sectoral agencies with conflicting agendas were expected to subsume their need for institutional visibility and to present a coordinated front; agencies that could not change their workplans were expected to coordinate; loans were to be given out by SIP but collection was to take place in the municipalities; the obstacles inherent in the regularization of land tenure for squatters were not fully anticipated; and Area Implementation Plans were required of SIP before disbursement, even though it was clear at the time that staffing levels and expertise were not at a level at which AlPs could be produced. The monitoring problem proved daunting, and it was complicated by high levels of staff turnover: Bank staff tended to spend missions discussing project issues with upper level staff from the project entities in Bogota and, as a result, found it difficult to make the time to visit project sites. 8 Additionally, the frequent change in task managers on both sides resulted in substantial losses of time as the involved parties became acquainted and re-visited certain issues. 4.06 Some of the conceptual leaps of faith which the project design made are reflected in the following decisions: (i) constructing large and expensive buildings (without asking the beneficiaries if they wanted them) was determined to be the most appropriate modality to promote community participation with slum dwellers; (ii) expecting the about-to-be-organized but impoverished neighborhoods to happily accept the responsibility of paying the expenses of the community centers; (iii) the idea that one intervention could possibly solve most of the problems in 23 large urban areas at the same time; (iv) the determination not to use existing local organizations (and to rely on SIP to create new local organizations just for the purpose of this project); and (v) the notion that participating on one directive board would empower slum dwellers enough to shift the balance of political power sufficiently to ensure marginal communities a significant voice in the running of the cities in the future. Actual physical achievements are summarized in Table 2 of the PCR and will not be repeated in the analysis below. Analysis of Selected Components 4.07 CDVs. Only nine of the 23 centers had their full complement of SIP staff by the completion date. Many of the centers have never been able to obtain legal title to the land upon which they are constructed. Most lots not deeded to the project belong to the municipalities or other public entities, so they are not at risk of being taken away, however. The identity of the ultimate operator of the centers was not covered in the original plan because it was hoped that the community organizations created by SIP would prove capable of that job. With hindsight this seems naive: the neighborhoods surrounding the centers are marginal in all senses of the word. While it appears that the CDVs remain very highly regarded by their communities, in the opinion of several IPC staffers, without the presence of paid watchmen the centers would likely be mined for the bricks and other construction materials they contain. When viable community organizations failed to materialize and SIP was discontinued, serious unresolved issues arose around which governmental agency/project entity would cover their long-term maintenance and operational costs. A very interesting conflict can be seen in the development of this project. Constructing large buildings costing US$430,000 in slum areas is not financially sustainable. The project, however, was hoping that these marginal communities would somehow raise funds to cover operating costs-which never happened. On the other hand, the existence of the impressive center buildings ultimately preserved the IPC program. Between 1986 and 1994 the centers have been located under five different institutional frameworks, passing through several periods of organizational limbo (some of more than a year) without an operational budget and at times without staff salaries. Had less important buildings existed in the marginal areas, there is little doubt that the program would have been abandoned. There were serious problems surrounding the construction of the centers. ICT convinced the Bank that reliable contractors would not participate in competitive bidding for works scattered among 23 cities, and physical works were handled through direct administration of local contractors. Contracts were signed even before the engineering drawings and soil analysis had been completed. Construction delays, structural defects, and a substantial repair process (at least partially the result of an inadequate design process) delayed the use of the community centers for their intended purposes. Work on 11 CDVs was held up because the Bank did not approve their Area Implementation Plans. The final two CDVs were not built until 1984. This component generally incurred significant cost overruns (10%), though buildings built later in the process utilized standard bidding documents and had fewer quality problems. 9 4.08 Tenure Regularization and Housing. Because of inadequate funding and a lack of regional offices and staff in several project areas, ICI proved unable to implement the sites and services and housing components in some project cities. Revisions were ultimately made that reduced these housing target figures. The appraisal had expected that ICT would lend to municipal authorities to allow them to cover the costs of issuing legal land titles to families in the marginal areas. During implementation ICT indicated that it was unable to lend to municipalities. None of the 31,000 property titles that were to be granted under this component were issued. However, those 7,381 families that received home improvement credits under another project component were allowed to borrow funds for title regularization if such was required. 4.09 Water. Delays affected the execution of water systems in 10 out of 21 cities where investments were made. Many squatter communities were located on steep land where it was difficult to lay pipes. Materials often did not arrive on time. Additionally, it was intended that water lines follow streets, and putting in the streets often involved getting neighbors to agree to moving houses off the land where the streets would run. The Government attached a high political priority to the construction of water systems, and investments in water were revised upwards. Once INSFOPAL improved modalities for the procurement of supplies and engineering support, implementation became more agile. The Borrower requested the introduction of a water plant into the project; the Bank did not approve because the financial feasibility of the endeavor seemed inadequately demonstrated. 4.10 Social services. The project proved highly successful in delivering basic services to urban neighborhoods where they had previously been scarce. Educational, nutritional, health, and child care facilities were built within the CDVs to appraisal standards or higher. These continue to function effectively. New and renovated classroom totals exceeded appraisal estimates by 94%, classroom equipment purchases exceeded estimates by 74%. One notable area where objectives were not attained was in health extension. Medical professionals did not fully support the use of paramedic personnel, which resulted in only 25% of the appraisal estimate of such staff being trained by MoH personnel and incorporated into the project Additionally, paramedics were never allowed to work within their communities as was originally planned: they were utilized within the CDVs under the direct control of professional staff. Health services have been expanded to include dentistry in many centers. The vocational training component administered through SENA became active early in the project and it provided skills upgrading to residents of marginal neighborhoods. There are serious questions as to whether a significant number of people actually obtained jobs because of the training they received. No study was ever done that evaluated the impact of the training on job-seekers or that counted the number of "graduates" that found work. These courses continue to be popular, however, partly because many poor families can lower their own expenditures by, for example, learning how to make their own shoes and clothing. Vocational education has led to significant micro-enterprise development The neighborhoods around the centers are filled with small enterprises specializing in the very activities taught in the centers, and several cooperative endeavors were found at the centers visited. 4.11 Community Participation. Although SIP promoted community participation generally, and was committed to community control of the CDV through its board of directors, community management of the program was never achieved as anticipated, largely because existing community organizations were bypassed. Project designs which do not take the concerns of marginal beneficiaries into account will not have the optimum impact on their problems. Both the Bank and the Government should allow greater participation in the design phase when working in low-income 10 urban areas in the future. Project officials admitted that the community never even knew that a community development center was coming until the construction of the building started. Several directors of CDVs believe that the project continues to pay a high price for not allowing the community to participate in project planning and the identification of proposed project activities. IPC staff insist that it was necessary to exclude the existing organizations, either because they were politically compromised, or they were committed to the interests of a group that did not coincide with the centers' service area. Community participation in project planning might have led to more involvement in later project activities, but SIP staff had limited their consultations to mayors, politicians and other high-ranking spokespersons. Community participation continues to be most apparent in sports, cultural and commercial activities subsidized or fully paid for by the centers--a far cry from what the project hoped to achieve, but not an insignificant accomplishment considering the conditions in which the project worked. Project Results 4.12 Sustainable improvements. The project achieved its housing and infrastructure aims, and continues to achieve its service goals--a result of improved coordination between the agencies which work out of the centers. The construction of the community center not only provided an agreeable site for service providers to work in dangerous neighborhoods, but also symbolized the government's concern with the problems of the poor. The project was designed to increase local capacity to solve local problems. The CDV centers were supposed to funnel assistance emanating from central agencies, eliminating gaps and overlaps in the process. They were to help the community to carry out a plan that would bring both economic and physical improvement. And they were to provide a mechanism to deliver rapid response to emergencies and other spontaneous needs of the marginal communities. During the visits to the field undertaken as part of the audit process it could be seen that the buildings remain in use, providing a centralized place to receive services. The agencies really are coordinating within the centers: IPC uses SENA buses to take senior citizen groups on field trips, various agencies make use of computers in the schools, IPC recruits trainees for SENA, the health clinics tend sick children from the schools and day-care centers, the school children use the IPC libraries, etc. As has already been noted, the hand-over of the centers to community control never did take place. However, the community clearly makes use of the centers for meetings, cultural and sports events and recreation. Several of the centers are well equipped and located to play an important role in the case of emergencies: the Pasto center is located near an active volcano that menaces a large sector of its target area; the Tumaco center can provide emergency services and shelter in the event of the frequent fires, storms, earthquakes and high seas that menace the area; and the Pereira center is located above a settlement along a river which has been subject to flooding in the past. There has been a tremendous improvement in living conditions around many of the centers. In Pereira a vast slum area that housed shacks when the center opened now contains paved streets and attractive brick houses connected to all municipal services. Home improvement loans have generally resulted in a significant upgrading of housing stock. It must be noted, however, that while the neighborhoods around the centers have solved many of their problems, new marginal neighborhoods have cropped up around the edges of the improved zones. Operation Experience 4.13 Several community groups complained of having submitted proposals for commercial activities to IPC/Bogota in response to expressed offerings, and then having to wait for years before receiving a response. The project has been short on funds for community activities and understaffed virtually 11 since its inception. During the first two years of implementation SIP was unable to produce timely, comprehensive and reliable reports at least partly because there was not enough staff. Since SIP has ceased to exist, IPC has been plagued by a budgetary shortfall which has made it difficult to keep a full complement of personnel in each center, and funds for cooperative activities and small enterprise development have never kept up with the number of projects presented. ICTr, the bank responsible for housing related loans, has been closed and its portfolio is being liquidated to provide funds for a housing subsidy program conducted by INURBE aimed at the urban poor. INSFOPAL was also closed down, and under the current decentralization policies the municipalities are now responsible for the construction and maintenance of the infrastructure within their geographic areas. In cities where health clinics existed within the target areas, health facilities were never build within the CDVs. During site visits, the beginnings of institutional disintegration could be seen with regard to SENA/IPC. SENA's need for visibility is not met within the CDV where the umbrella organization receives a large measure of credit for what is done. In Pasto, SENA closed its workshop and removed the equipment purchased under the Bank loan during the OED visit. Interviews with SENA officials revealed that "high technology" centers were going to be opened in the city center in order to capacitate industrial workers. This would be a return to the patterns and practices of the 70s, when the marginalized areas were excluded from the activities of many service providers. Protests from the community and IPC on the national level had been lodged, however, and the possibility that SENA would return to its facilities at the CDV center was still under discussion. The location of the project within the President's office has also contributed to a certain instability. Each new administration has a clear tendency to create a new initiative which will be clearly tied to the new executive in the public consciousness. The emergence of the Plan Nacional de Rehabilitacion to help former guerilla groups integrate themselves into society originally led to a de-emphasis of urban projects, although as this population has largely opted to live in the cities, urban concerns seem to be coming to the fore once again. Compliance with Operational Directives and Covenants 4.14 As for any project, a very large number of Bank Operational Directives (OD, Operational Manual Statements (OMS), Operational Policy Notes (OPN) and Guidelines were applicable to the preparation, appraisal, and implementation of the project. Compliance with certain "administrative," legal, country economic and procedural requirements can be taken for granted as these were required for Board approval and effectiveness. 4.15 Apart from the standard covenants of a typical loan/credit agreement there were few specific covenants applied to the credit and these fell into the following categories: (a) Covenant regarding the transfer of funds to the project entities. The PCR notes that the covenant requiring the prompt and non-reimbursable transfer of funds to ICT, INSFOPAL and CFP was not complied with for the first two years of implementation and then subject to insufficient and delayed financial transfers. (b) Covenant specifying procedures for the approval of subprojects. ICTr charged an interest rate which was lower than that specified in the Legal Agreement. The Empresas Municipales did not contribute their own funds to the INSFOPAL project component as required. (c) High staff turnovers and unfilled vacancies resulted in non-compliance with provisions that required full staff at the CDVs and at the SIP central office. The various councils that were 12 to ensure coordination at the national, municipal and neighborhood levels met only sporadically. As was noted previously, the Area Implementation Plans were prepared with significant delays. (d) Accounting records were established later than had been agreed, and there were instances, notably at ICT, where accounts were not kept in accordance with sound accounting principles. 5. CONCLUSIONS Significant Impact In Spite of Over-ambitious Objectives 5.01 The OED data base on completed urban projects rates this operation as unsatisfactory, largely due to doubts about its financial and institutional sustainability. In the opinion of this audit, the project now warrants a satisfactory rating. In spite of numerous organizational changes the program has remained operational. Its recurrent costs for staff and center upkeep have proven to be surprisingly modest. And the project has begun delivering new services--notably in the areas of senior citizen activities and municipal recycling-which were not noted at the time of the PCR. Thus, although its original objectives may have been vaguely formulated, the community centers and the various projects run through them by the IPC have demonstrated remarkable institutional and stream- of-benefits sustainability and therefore merit a rating upgrade at this time. 5.02 The marginal areas project was attempting to deal with virtually all aspects of what continues to this day to be a difficult urban situation. All Government and Bank staff involved with the project knew from the outset that achieving comprehensive solutions to all the problems of the marginalized ares in one shot was going to be difficult. The fact that such an ambitious project was attempted at all is due to the fact that the it was in tune with the spirit of the time: it was in harmony with a policy speech made by Mr. McNamara; it mirrored Colombia's development goals, and respected the manner in which that nation had determined to address its most intractable urban problem. This operation is widely perceived within the Bank as having been problematic. Within Colombia the fact that the centers continue to operate is not widely known. When the OED audit mission was making appointments in the field, one mayor contacted did not realize that there was a CDV in his city. Subsequently, however, he visited the installations and was impressed with the ongoing work. It would be a shame if the only lesson which was learned from the experience of this project is that working with the most vulnerable population in the large and growing cities of the developing world is difficult (and therefore bad for one's career). The world's cities have grown to unprecedented size and the risk of urban catastrophes has multiplied. For that reason alone, the lessons listed below should not be interpreted as suggesting that urban programs are better off not addressing the problems of the poor in a comprehensive manner. 5.03 This project was designed with two major flaws: it attempted to do too much at once, and the activities that were employed were unlikely to solve the problem of marginality. At best, a project directed at the health, education, employment, housing, and infrastructure, problems of the marginal areas could only succeed in reducing them. Solving each of the more significant problems of Colombia's cities will involve several loans before the majority of observers will be persuaded that a 13 solution is at hand. Reducing the vulnerability of marginal areas incrementally allows governments to "declare victory" more often. If the goal of a loan is to integrate slum areas into the fabric of the surrounding city, what task manager will wait around to celebrate the successful achievement of the loan objectives. Rather than assuming away political and institutional constraints, programs that deal with disempowered groups need to build in political visibility, in effect they must go out of their way to provide "ribbon cutting" opportunities for key individuals within the governance structure. When projects take a "Christmas Tree" approach, with a plethora of components designed to please almost everybody, implementers will concentrate on those outcomes which are easy to achieve, even at the expense of the main goals and objectives. Clearly, urban loans have to be more explicitly oriented towards outcomes tied to their fundamental purposes which can be counted, and under which success is possible to prove. Project Design, Points of Special Interest 5.04 This loan is widely recognized within the Bank as both ambitious and complex. It has been observed in this document that it took place in 27 sites and attempted to coordinate nine agencies. In fairness it must be admitted that the project was attempting to solve precisely this problem-it was known that there were too many agencies that were not coordinating, and the idea was that bringing them together in one center would resolve many coordination issues. Where the project might more accurately be described as having gone astray is in its one-shot approach-the assumption that it would solve the slum problem once and for all, integrating marginalized neighborhoods into the fabric of the surrounding city by providing basic services, even when there has been no meaningful participation of the target group in the planning stage. There was also an unrecognized assumption that population trends would be uni-directional, that the intermediate cities would continue to grow at past growth rates--even though we know that small localized events can have major impacts on settlement patterns. With a novel multi-sector project such as this, unanticipated local circumstances often cause changes in project activities. Since slum areas are generally concentrated, the project was designed so that the various services would be accessible to the slum dwellers in one large community center. When the area plans were being drawn, it was discovered that some of the neighborhoods that were to receive services were located at unreasonable distances from the service centers. It would have been reasonable to build more centers. In addition, the community centers required a large plot of land and not all cities had available such sites in close proximity to the target populations. Some cities were forced to abandon the idea of the large central structure and to construct several mini-centers. 5.05 The fact that population can shift very rapidly within cities, regions and countries underscores the importance of closely examining key assumptions during project design and maintaining flexibility during implementation. While the project intended to upgrade the level of services available in the intermediate cities, and thereby slow the growth of the largest cities, it was already clear during implementation that the demographic pattern had been misunderstood: rural people did not stop off in intermediate cities on their way to the big ones no matter how good services were, they went straight to the largest because they present better employment opportunities. The DPN study conducted 10 years prior to implementation might have been profitably revisited, if only briefly, for the purposes of a "reality check." 5.06 Although the entire nature of this project was decentralization and participation, the various institutions involved in the national and local governing boards made a series of decisions which caused the project to become ever more centralized and exclusionary (in terms of decision-making). 14 The rapid growth of urban areas around the world, and the inability of government agencies to improve the lot of a rapidly growing marginalized urban population when they work in uncoordinated ways indicates that the Bank either needs to decide to learn how to manage complex urban projects or decide to address the needs of marginalized populations through single-sector projects. Dilution of political and institutional commitment to project objectives over time are described in the files. Both in Colombia and in the Bank, an apparent loss of understanding and confusion as to the philosophy and central objectives of the project took place. Lessons Learned 5.07 Complexity. Because of the complexity of this project several unintended outcomes ensued: project entities focussed on easier to implement activities with the result that highly important objectives (such as service coordination, development of community organizations, regularization of land tenure, and participatory management) fell by the wayside. Because no incentives/disincentives were put in place for variations of organizational behavior in the various cities, physical works (notably the construction of the CDVs) and other easy to implement components tended to dominate the project. For example, while it is, no doubt, a good thing that 76% more primary school equipment was purchased than was appraised, to what degree does this acquisition counterbalance the fact that the land tenure component was dropped? Similarly 33% more children were attended at child welfare centers, but the paramedic outreach program was dropped. Certainly many worthwhile outputs were the direct result of this loan. However, the central goals of basing the delivery of services on beneficiaries' organizations and enabling squatters to participate more fully in the life of the cities were sacrificed. The ability of the Government to staff and equip primary schools was enhanced, but it was already well established. On the other hand, whether the project actually strengthened the capability of the Government to coordinate the activities of the various executing agencies beyond the day-to-day activities on the local level is questionable. The above analysis begs the question of what is the most appropriate approach to the range of problems facing marginal areas. Chronic urban problems are unlikely to be significantly ameliorated over the short term. If the time that can be allotted to such projects is limited, the answer would seem to be that tackling them a few at a time, under a thorough sequential plan is the better approach. First, provide a means to legal land tenure. Then build curbs, introduce water and sewerage. Finally, pave the streets. The high level of arrears may indicate that housing credit is not the best alternative for upgrading construction in neighborhoods with no history of borrowing for housing. In Latin American marginal neighborhoods, population tend to be relatively stable, and to build slowly when materials are available inexpensively, often using self-help. Several observers noted that poor communities rapidly improve their appearance along paved streets. Given the way in which people build in Colombia's marginal urban areas, a home's value may bear little relationship to its actual cost. While the comprehensive approach attempted by this project did achieve many significant objectives, the time originally allotted for their achievement was too short for the impacts of the project to be fully appreciated. 5.08 Working Within the Reality of the Governance Structure. Several instances where political and institutional realities were assumed away have already been given, both at the local and national levels. Perhaps the foremost among these involved the Popular Integration Secretariat (SIP) which, when it was established in the President's office, had almost no personnel, and no experience with coordinating the work of any of the project entities. Not enough allowances were made for the limited implementation capacity of the newly created SIP, and efforts to develop the institution had a limited impact because it was a Presidential Program, subsequently eliminated by a later chief 15 executive. The DNP may not have fared any better as a borrower, it is hard to see how a planning department could have run such a complex project over such a wide geographic area. Leadership is critical to provide the political leverage, knowledge and diplomacy necessary for prodding competing agencies, obtaining funds and policy changes and for directing complex programs. Most of the activities the CDVs continue to realize are owed to the personal attributes and initiative of the center directors and their staff. Area Implementation Plans were required of SIP before disbursement, even though it seemed clear at the time that staffing levels and expertise were not at a level at which AIPs could be produced. It may be that the scope, content and importance of these reports were blown up all out of proportion relative to what was anticipated in the original project design. 5.09 Monitoring. This project proved largely impossible to monitor. In the centers visited, numerous instances of employee longevity were found. Staff at the CDV centers believed that this continuity of personnel proved critical both for the work with the surrounding community, and to the projects continuing survival within the national bureaucracy. 5.10 Community Participation. Where project implementers have difficulty promoting participation they can, at the very least, keep the community informed as to what is going to happen. The neighborhoods around the centers may have been willing to pay small fees for specific services, but to think that hundreds of families would somehow contribute financially to the upkeep of a government building that the community had never requested was a conceptual leap of faith. Although it was hoped that the community would eventually take over and manage the centers, such management is more complicated by a large and expensive to maintain building. Physical works-the construction of the CDVs-dominated the concerns of project managers during the early project years, over-riding potentially more important community activities in terms of staff involvement while disbursements continued. As has been noted, the existence of the CDV buildings was a two-edged sword-they preserved the existence of IPC as an institution while inhibiting the achievement of the organization's participatory objectives. An impressive building tends to define the program as belonging to the service providers. A building built around a defined set of services freezes the form of the program according to its original conception, and to some degree reduces the possibility of new or different activities springing up in response to the changing needs of the intended users.