Mexico - Water Supply and Sewerage Projects

Document of The World Bank FOR OMCAL USE ONLY Report No. 12145 PERFORMANCE ARDIT REPORT INEXICO MEDIUM CITIES WATER SUPPLY AND SEWERAGE PROJECT (LOAN 1186-ME) SECOND MEDIUM-SIZE CITIES WATER SI. PLY AND SEWERAGE PROJECT (LOAN 1913-ME) AND THIRD 1EDIUM-SIZE CITIES WATER SUPPLY AND SEWERAGE PROJECT (LOAN 2281-ME) JUNE 30, 1993 MICROGRAPHICS Report No: 12145 Type: PPAR 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. EXCHANGE RATES Currency: Mexican Peso (P) 1972 US$1 = 12.50 Pesos 1980 " a 25.60 * 1973-80 a = 18.00 " 1982-90 a - 507.00 " 1990 a - 2,267.50 " GLOSSARY OF ACRONYMS BANOBRAS National Bank of Public Works and Services (Banco Nacional de Obras y Servicios PGblicos) CEAPAS State Commission of Potable Water and Seweraje of Sinaloa (Comisi6n Estatal de Agua Potable y Alcantarillado de Sinaloa) CNA National Water Commission (Comisi6n Nacional del Agua, dentro de SARH) DIGESAPA Potable Water and Sewerage Headjuarters within SEDUS (Direcci6n General de Sistemas de Agua Potable y Alcantarillado, en SEDUE) FIFAPA Investment Fund for Water Supply and Sewerage Works (Fondo de Inversionea Financieras para Aqua Potable y Alcantarillado) INTA Mexican Institute for Water Technology within SARK (Instituto Mexicano para la Tecnologia del Agua, dentro de SARH) MCMA Mexico City Metropolitan Area (Area Ketropolitana de la Ciudad do M6xico) PIU Project Implementation Unit (Unidad Je Implementaci6n del Proyect:, PNDU National Plan of Urban Development (Plan Nacional de Desarrollo Urbano) PNAPYA National Program for Water and Sewerage (Programa Nacional de Agua Potable y Alcantarillado) SAHOP Secretariat for Human settlements and Public Works (Secretaria de Asentamientos Humanos y Obras PGblicas) SARH Secretariat of Agriculture and Hydraulic Resources (Secretaria de Agricultura y Recursos HidraGlicos) SEDESOL Secretariat for Social Development (Secretaria de Desarrollo Social) SEDUE Secretariat of Urban Development and Ecology (Secretaria do Desarrollo Urbano y Ecologia) SHCP Treasury and Public Credit Secretariat (Secretaria do Hacienda y Cr6dito PGblico) SARH Secretariat of Hydraulic Resources (Secretaria de Recursos HidraGlicos) SSSP Planning and Budget Secretariat within SHCP) (Subsecretaria de Planeaci6n y Presupuesto, dentro do SHCP) FISCAL YEAR OF BORROWER January 1 - December 31 FOR OFFICIAL USE ONLY THE WORLD BANK Washington, D.C. 20433 U.S.A. Office of Director-General Operatione Ivaluation June 30, 1993 MEMCRARDUM TO THE EXECUTIVE DIRECTORS AND THE PRESIDENT SUBJECTs Performance Audit Report on Mexico Medium Cities Water Supply and Sewerage Project (Loan 1186-ME) Second Medium-Size Cities Water Supply and Sewerage Project (Loan 1913-ME) Third MediumSize Cities Water Supply and Sewerage Prolect (Lnn 2281-ME) Attached is the report entitled "Performance Audit Report on Mexico - Medium Cities water Supply and Sewerage Project (Loan 1186-ME), Second Medium- Size Cities Water Supply and Sewerage Project (Loan 1913-ME) and Third Medium- Size Cities Water Supply and Sewerage Project (Loan 2281-ME)" prepared by the Operations Evaluation Department. The physical elements of the projects were implemented, albeit with significant delays and changes. However, institutional and sector policy improvement efforts were undermined by frequent organizational and administration changes. Furthermore, the important objective of ensuring the sector's financial viability by replacing government grant financing with systematic cost recovery was not achieved. A number of cities withdrew from the projects given the availability of alternative (grant) funding. The performance of the first two projects are rated as unsatisfactory and that of the third project as satisfactory. The sustainability of all three projects are rated as uncertain. Institutional development, in the first two projects was partial and, in the third, substantial. These ratings agree with the ratings based on the review of the PCRs. There are two overriding lessons to be drawn from this series of projects: 1) Optimistic assumptions regarding presumed government commitment are no substitute for clear and binding agreements and 2) beneficiary agencies must share in the planning, design and setting of objectives for development projects involving them. Attachment This docunent has a restricted distribution and may be used by recipients only in the performance o their offkcfat duties. its contents my not otherwise be disctosed without Wortd Bank authorization. FOR OFFICIAL USE. ONLY PERg1MANCE AUDIT REPORT MEXICO MEDIUM CITIES WATER SUPPLY AND SEWERAGE PROJECT (LOAN 1186-ME) SECOND MEDIUM-SIZE CITIES WATER SUPPLY AND SEWERAGE PROJECT (LOAN 1913-HE) AND THIRD MEDIUM-SIZE CITIES WATER SUPPLY AND SEWERAGE PROJECT (LOAN 2281-ME) TABLE OF CONTENTS Page No. Preface *.....**** ...... ...... Basic Data Sheets . . . . . . . . . . . . . . . . . . . . . . . . . . Evaluation Summary . . . . . . . . . . . . . . . . . . . . . . . . . ix I. PROJECT SUMMARY. . . . . . . . . . . . . . . . . . . . . . . . 1 A. Background ....... ............. . 1 B. Bank Lending . . . . . . . . . . . . . . . . . . . . . . 2 C. Objectives . . . . . . . . . . . . . . . . . . . . . . . 4 D. Sector Organization . . . . . . . . . . . . . . . . . . . 5 1. Past Organizational Arrangements . . . . . . . . . . 5 2. Present Organizational Arrangements . . . . . . *. . 6 II. THE PROJECTS AND THEIR OBJECTIVES . . . . . . . . . . . . . . 8 A. Medium Cities Financing Mechanism (FIFAPA) . . . . . . . 8 MEDIUM CITIES WATER SUPPLY AND SEWERAGE PROJECT (LOAN 1186-ME) . 9 A. The Project and Its Objectives. . . . . . . . . . . . . . 9 B. Implementation . . . . . . . . . . . . . . . . . . . . . . 9 C. Project Costs and Disbursements . . * . . . . . . . 10 D. Procurement . . . . . . . . . . . . . . . . . . . . . . . 11 E. Performance of Contractors and Consultants . . . . . . . . 11 F. Results and Sustainability . .... . .. . . . . . . . 11 SECOND MEDIUM-SIZE CITIES WATER SUPPLY AND SEWERAGE PROJECT (LOAN 1913-ME) . . . . . . . . . . . . . . . . . . . . . . . . . 12 A. The Project and its Objectives . . . . . . . . . . . . . . 12 B. Implementation . . . . . . . . . . . . . . . . . . . . . . 12 C. Project Costs and Disbursements . . . . . . . 13 D. Procurement . . . . . . . . . . . . . . . . . . . . . . . 13 E. Performance of Contractors and Consultants . . . . . . . . 14 F. Results and Sustainability . . . ..... . . . . 14 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 (Cont.) Page No. THIRD MEDIUM-SIZE CITIES WATER SUPPLY AND SEWERAGE PLOJECT (LOAN 2281-ME) . . . . . . . . . . . . . . . . . . . . . . . . . 15 A. The Project and its Objectives . . . . . . . . . . . . . . 15 B. Implementation . . . . . . . . . . . . . . . . . . . . . . 16 C. Project Costs and Disbursements . . . . . . . . . . . . . 16 D. Procurement . . . . . . . . . . . . . . . . . . . . . . . 17 B. Performance of Contractors and Consultants . . . . . . . . 17 F. Results and Sustainability . . ........ ...... 17 III. PERCORMANCE OF THE BANK AND THE BORROWER . . . . . . . . . . . 18 A. Performance of the Bank . . . . . . . . . . ..... . 18 B. Performance of the Borrow and the Executing Agency . . . 19 C. Project Relationships . . . . . . . . . . . . . . . . . . 20 D. Institutional Development . . . . . . . . . . . . . . . . 20 B. Financial Aspects .. . . . . . . . . . . . .*. . . . . . 21 P. Economic and Environmental Assessments . . . . . . 23 G. Project Impact on Urban Poverty and Women . . . . . . . . 23 IV. CONCLUSIONS AND LESSONSALEARND . . . . . . . . . . . . . . . . 24 ANNEX 1 Current Organization of the Water Supply and Sanitation Sector in Mexico . . . . . . . . . . . . . . . . . . . . . . . . 28 ANNEX 2 Coments from the Borrower ................... 29 L PERFORMANCE AUDIT REPORT MEXICO MEDIUM CITIES WATER SUPPLY AND SEWERAGE PROJECT (LOAN 1186-ME) SECOND MEDIUM-SIZE CITIES WATER SUPPLY AND SEWERAGE PROJECT (LOAN 1913-ME) AND THIRD MEDIUM-SIZE CITIES WATER SUPPLY AND SEWERAGE PROJECT (LOAN 2281-ME) PREFACE 1. This report presents the results of the Performance Audit of the three Medium Size Cities Water Supply and Sewerage Projects. Loans for these projects were made as follows: (i) Medium -ities Water Supply and Sewerage Project (Loan 1186-ME) was approved December 18, 1975, became effective on April 26, 1976 and closed 3 years late on January 14, 1984. Of the loan amount of US$40 million, 99.5% was disbursed and the remainder, US$0.21 million, was canceled. The loan represented 40% of the total estimated project cost at appraisal of US$100 million. (ii) Second Medium Cities Water SuRRlv and Sewerace Proeot (Loan 1913- MN) was approved October 4, 1980, became effective on June 23, 1981 and closed 2 years late on November 30, 1986. Of the loan amount of US$125 million, 60.6% was disbursed and the remainder, US$49.27 million, was canceled. The loan represented 39.3% of the total estimated project cost at appraisal of US$318 million. (iii) ThirdMedium Cities Water Sumely and Seweraae Proiect (Loan 2281-ME) was approved May 17, 1983, became effective on February 27, 1984, and closed on December 31, 1991 with only a 6-month delay. Of the loan amount of US$100.3 million, 95% was disbursed and the remainder, US$5 million, was canceled. The loan represented 52% of the total estimated project cost at appraisal of US$193 million. 2. All three loans financed mostly physical improvements in various medium- sized cities in support of the Government's efforts to decentralize urban development and reduce the growth of Mexico City. The loans included funds for training and institutional development at central, st.1 .e and local level. Bank funds were onlent through a central sector finance mechanisms the Investment Fund for Water Supply and Sewerage Works (FIAPA). This fund was established and managed by the National Bank for Public Works and Services (BANOBRAS) with the intention of eventually funding all urban water supply and sewerage investments on a loan basis, and so eliminating Mexican municipalities' traditional dependence on Government grants. 3. The Performance Audit Report (PAR) is based on the Project completion Report (PCRj prepared by the Latin America and the Caribbean Regional Office and submitted to the Board on December 21, 1992 (Report No. 11448). ORD has reviewed 11 the PCR, the Staff Appraisal and the President Ia Report, the legal documents and Bank projects' files as well as the transcripts of the Minutes of the meetings of the Executive Directors at which the loans were approved and discussed with involved Bank staff. An OED mission visited Mexico in September 1992 and discussed project experience with government and project city officials and visited some of the state and municipal authorities who participated in the projects. The mission also visited some of the facilities financed by the loans. Discussions, particularly about the earlier projects, were not as extensive as would have been desirable, due to the length of time since project implementation and, more importantly, the many changes in executing agencies and units responsible for project implementation within executing agencies. This made it difficult to find staff with an accurate historical perspective of events. 4. The audit finds that the PCR gives a concise and thorough account of a rather complicated project experience and draws appropriate conclusions. In order to compare project results with project design, the PCR used project feasibility study data as a base. Because some of these feasibility studies were completed after appraisal, conclusions do not precisely reflect data on which the Erecutive Directors based their decision to approve loans 1913-ME and 2281-ME. This results in discrepancies between the projected and actual number of people served; however, the incremental service provided as a result of the investments made for those cities which were incladed in the appraisal, and the per capita construction costs, are roughly as projected in the appraisal report, and the overall conclusions are not affected. The PAR further analyzes some aspects of the PCR's findings and provides independent comments and conclusions on the various aspects of the projects. S. Following standard OED procedures, copies of the draft PAR were sent to the Government and the borrower for comments on April 9, 1993. Comments received from BANOBRAS are reflected in, and attached to the report as Annex 2. iii, MMMANCE AUDIT WSPORT MEXICO MEDIUM CITIES WATER SUPPLY AND SEWERAGE PROJECT JLOAN 1186-NEI BASIC DATA 1EET E,Y PROJECT DATA Actual or Actual as % Appraisal Current of Appraisal item Expectation Istimate Estimate Total Project Cost (US$ m) 100 76.5 76.5 Loan Amount (US$ m) 40.0 39.79 99.48 Economic Rate of Return N/A N/A N/A Institutional Performances Unsatisfactory Q;MULATIVE ESTIMATED AND ACTUAL DISBURSEMENTS Appraisal Estimate (US$M) 40.0 Actual (US$M) 39.79 Actual as % of Appraisal (t) 99.48 Date of Final Disbursement: August 31, 1987 PROJECT DATES oriainal Actual Identification 01/74 Preparation 07/74 Appraisal 11/74 06/75 Post Appraisal Negotiations 11/10/75 Board Approval 12/18/75 Loan Signature 01/13/76 Loan Effectiveness 04/15/76 04/26/76 Loan closing 12/31/80 01/14/82 Project Completion 06/30/80 01/14/84 IV fIsui DAT Date No. øf No. of Mth/Xear DaXs Persons Identification 1/74 25 5 Preparatlon 2/74 2.5 3 Preappralsal I 8/74 4 1 vreappralgal II 9/74 45 4 Preappraisal III 12/74 10 3 Preappraisal IV 3/75 37.5 3 Appraleal 6/75 60 3 Supervision 5/76 5 2 Supervision 10/76 15 2 Supervision 5/77 20 2 Supervision 12/77 15 2 Supervision 4/78 10 2 Supervision 10/78 20 2 Supervision 3/79 5 2 Supervision 6/79 15 2 Supervision 2/80 5 2 Supervision 10/80 5 2 Supervision 4/81 10 3 SupervisiL. 9/81 7.5 2 Supervision 2/82 10 5 Supervision 7/82 10 3 Supervision 10/82 30 5 Supervision 06/83 15 4 Supervision 9-10/83 5 2 Supervision 3/84 5 2 Supervision 5-06/84 5 3 rTAFF INPTS (Staf f-Weeks) Preapprai1 5.0 1.1 2.8 24.1 Appralsal 1.1 21.1 18.1 wegotiatao 4.8 Supervision .5 3.1 33.8 26.9 25.4 10.6 11.5 17.3 28.2 13.2 1.3 5.7 4.2 1.2 15.8 OLher .9 2.0 1.7 10.8 To~l 6.1 2.5 4.8 46.9 36.9 33.8 26.9 23.4 10.6 11.5 17.3 28.2 13.2 1.5 5.7 4.1 1.2 15.8 V PERPOMNCE AlIT ,LFgZ SECOND MEDIUM-SIZE CITIES WATER SUPPLY AND SEWERAGE PROJECT (LOAN 1913-ME) BASIC DATA SHEET KEY PROJECT DATA Actual or Actual as 2 Appraisal Current of Appraisal Item Exeetation Estimate Eatmate Total Project Cost (US$ m) 297 115.89 39.02 Loan Amount (US$ m) 125 75.73 60.58 Economic Rate of Return NIA N/A N/A Institutional Performances Unsatisfactory CUMULATIVE ESTIMATED AND ACTUAL DISBURSEMENTS Appraisal Estimate (US$ m) 125 Actual (US$ m) 75.73 Actual as 2 of Appraisal (1) 60.58 Date of Final Disbursement: August 31, 1987 PROJECT DATES Original Actual Identification 03/79 Preparation 05179 Appraisal 03/10/80 Post Appraisal 06/80 Negotiations 08/80 09/11/80 Board Approval 11/80 10/04/80 Loan Signature 01/23/81 Loan Effectiveness 06/23/81 Loan Closing 12/31184 11/30/86 Project Completion 06/30/84 11/30/86 vi STAFFR PT (Staff-Weeks) 7v7 F78 P-$ wo mT0 p8. pyio m m e 27$ wo VT8 o me WI 77 Proappraisal 1.3 .3 4.4 33.7 APPraisal 37.3 9.9 Nosotiation~ 11.3 Supervision 9.0 16.6 33.0 19.3 10.7 9.4 19.0 .3 1.2 1.3 Other .3 .3 .9 4.0 6.0 Total .3 1.7 .3 5.3 77.2 36.0 16.6 33.0 19.3 10.7 9.4 19.0 .3 1.2 1.3 vii PERFORNCE AUDIT REPORT MEXICO THIRD MEDIUM-SIZE CITIES WATER SUPPLY AND SEWERAGE PROJECT (LOAN 2281-ME) PASIC DATA SHEET KEY PROJECT DATA Actual or Actual as 2 Appraisal Current ef Appraisal item Expectation Estimate Estimate Total Prfject Cost (US$ a) 192.80 128.57 66.68 Loan Amount (US$ a) 100.30 95.30 95.01 Economic Rate of Return N/A N/A N/A Institutional Performances Satisfactory CQMULATIVE ESTIMATED AND ACTUAL DISBURSEMENTS Appraisal Estimate 100.30 Actual (US$ a) 95.30 Actual as % of Appraisal (%) 95.01 Date of Final Disbursement: March 13, 1991 PROJECT DATES Original Actual Identification 04/81 Preparation 07/81 Appraisal 03/82 Post Appraisal 02/83 Negotiations 04/07/83 Board Approval 15/17/83 Loan Signature 09/25/83 Loan Effectiveness 02/27/84 Loan Closing 12/30/90 12/31/90 Project Completion 12/31/89 12/31/90 viii STAFF INPUTS (Staff-Weeks) Wl 22n Un MI ML MO PY» O8 M I M M 02 ?"f PYO n88 VY89 O 9l MI Proapprasal 7.8 18.3 Appreal 38.0 22.3 %egotatlwa 7.8 Suprvontm 3.3 1.6 10.2 17.0 14.6 23.4 28.1 12.2 3.3 3.3 1.4 Othor .4 11.3 Total 8.4 39.7 43.2 10.2 17.0 14.6 23.4 28.1 12.2 3.3 3.3 1.4 OTHER PROJECT DATA Borrovers BANOBRAS Executing Agencies: Government of Mexico ix PERFORMANCE AUDIT REPORT MEXICO MEDIUM CITIES WATER SUPPLY AND SEWERAGE PROJECT (LOAN 1186-ME) SECOND MEDIUM-SIZE CITIES WATER SUPPLY AND SEWERAGE PROJECT (LOAN 1913-ME) AND THIRD MEDIUM-SIZE CITIES WATER SUPPLY AND SEWERAGE PROJECT (LOAN 2281-ME) EVALUATION SUMMARY Introduction 1. The provision of adequate water health problem among the lover-income supply in Mexico is difficult and groupse expensive. The principle reasons are very uneven distribution of water 3. The Bank made its first loan in resources, the varied climates, and the sector in 1973 (Loan 909-ME: the topography, coupled with the Mexico City Water Supply Project) to population pattern which has resulted help finance investments to expand in 751 of the population and 80% of Mexico City's bulk water supply industry being located above 500 a capacity. In addition to the above sea level, while some 85% of projects being audited, the Bank is surface water resources are located financing a follow-up project "Water below that elevation. Mexico made Supply and Sanitation Sector Project rapid progress in improving its (Loan 3271-ME)" and has also citizens' access to water and supported a project intended to sewerage services during the 1970s, Improve water supply and sanitation but progress has slowed since. and generate income-producing Coverage for water supply and activities for women: "Water, Women sewerage increased from 49% and 37%, and Development (Loan 3101-ME). respectively, in 1970, to 701 and 49% in 1980, and to 751 and 58% in 1990. The Proiects-and their Objectives 2. The water and sewerage sector 4. All three projects were has been adversely affected by the designed to achieve similar slowdown of the Mexican economy which objectives, which can be summarized began in the second half of the as: 1970s. Even the short-lived oil boom in the late 1970s had a negative 0 Improving Service Coverage and impact, because it resulted in an Standards, excessive dependence on the Goverment transfers for financing with particular emphasis on water and sewerage infrastructure, improving services to neglected and this continued to the 1990s. As lower-income groups, toimprove a consequence, the sector has barely environmental health kept up with population increases, conditions. Each project had and water-and sanitation-related hlseases present a serious public Irops x its own specific numerical Those funds originally included for targets for coverage expansion, training in the initial medium cities project and for health education in o Improving Cost Recovery, the third medium-size cities projects were reallocated to materials and through adequate tariffs and construction, either because the eventual achievement of components were not being Implemented financial viability of service or because the projects were revised providers to accelerate disbursement, o Enhancing Sustainability, 7. The composition of the projects changed substantially during project through institutional implementation, especially in the improvements assisted by case of the second and third decentralized technical projects. Cities which had assistance and local artonomy originally planned to participate in sector operations decided instead to take advantage of Government grant financing, and other o Establishment of a Unified cities joined. As a consequence, System of Sector Financing, appraisal projections are not comparable with service coverage with clear methods of resource actually achieved. overall, figures allocation, rules of show that targets achieved are eligibility and priorities for roughly comparable (adjusting for implementation project changes and loan cancellations) to those which were o Improving th, Environment, anticipated at the timeof appraisal, based on projected investments. through water pollution control Imlementation Exerience projects, policies and monitoring, and by the use, for 8. The Bank attempted to bring all projects, of sound about fundamental changes in a highly environmental assessments centralized sector through projects 5. The third medium-size cities which amounted annually to only about project also included a health 8Z of total sector investments. It education component with the attempted to accomplish this objective of improving public health, objective through a modified version and particularly the health of those of a sector loans working with an newly served, to be administered by inexperienced executing agency and CEAPAS. But with the delegation of borrower, and with little direct responsibility for water supply and contact with beneficiary regional or sewezage to municipalities, CEAPAS municipal water supply and sewerage was dissolved and the component organizations. Staff resources eliminated. devoted to project supervision would have been suff icient to work with the 6. All of the funds disbursed (US$ borrower and executing agency, but million 39.79, 75.73 and 95.3 they were inadequate to supervise respectively, for the three loans) individual sub-projects with the were used to pay for materials, intensity customary in other Bank construction and consultant services. projects (which was warranted with central organizations new to working xi with the Bank, especially while competition from grant funding which attempting to bring about fundamental remained more attractive to changes in the sector). The Bank was municipalities long used to the aware of the risks, but clearly notion that water supply wad sewerage unprepared for the magnitude of the services yore the responsibility of problems encountered. the National Government. Nevertheless* by the time the third 9. The tasks of the Bank and its project was being implemented, the counterpart organizations were made borrower had sufficient experience more difficult by events beyond their with managing the fund to have control, such as the economic developed effective management tools, depressions which occurred in the general and sub-loan policies and second half of the 1970s and again in agreements. Undoubtedly, the facts the early 1980s, and the oil boom in that the borrower had remained the between. The austerity programs, same agency throughout the particularly the one begun in 1982, Implementation period, and that its severely restricted counterpart staff remained with the projects for funds, bringing the projects to a lengthy periods, were the reasons for virtual halt, while the generous this competence. grant funding during the economic expansion raised municipal 12. To the problems imposed on the expectations for continued grant sector from the outside were added financing to unrealistic levels. The those it imposed on itself, which the Mexico City earthquake in 1985 responsible institutions could have further aggravated the financial avoided or resolved. Some of the difficulties of the sector, with basic problems were the frequent funds being diverted to jurisdictional disputes between reconstruction efforts. executing agency and borrower$ the frequent change in executing 10. One of the fundamental problems agencies, and the even more frequent affecting the implementation of all organizational changes in the units three projects was the lack of a firm within the executing agency commitment by the government to phase responsible for project out grant financing. There appears implementation. Coupled with the to have been an agreement in lack of staff and consultants principle with the Bank, but experienced in least-cost analysis of subsequent actions indicate that the projects, and the relatively low Government found it politically level of institutional competence of inexpedient to implement this most sub-project operating agreement. In practice, this made institutions, these organizational the projects' funds a resource of problems delayed project last resort, and placed a heavy implementation. burden on the borrower in its attempt to promote the idea of cost recovery 13. The project chronology (see and financial viability. Basic Data Sheets, page vi) illustrates some of the impact of 11. The Investment Fund for Water these problems. As shown in the Supply and Sewerage Works (FIFAPA), portion recording the lending which the borrower (BANOBRAS) process, the time between estimated implemented with the objective of completion and actual completion of eventually making it the single Project 1 is four years, of Project 2 financier of the sectorm, had is two years and of Project 3 is one xii year. This shows clearly that as water supply and basic health care staff competence increased and (they carry water from whatever jurisdictional disputes decreased, source to the household, and take progress accelerated. The table also care of infants and the sick in the shows, for example, that the change family), they benefit greatly from from SANOP to SEDUE in 1982 and the improvement in water and sanitation abolition of the Project services and the resulting Implementation Unit (PIU) by SEDUE in improvements in the health of their 1985 was followed by serious families. implementation problems. In fact, the problems associated with the 16. The Bank recognized this lack latter event led the Bank to deny a of women's participation in project request for further extensions to development and implementation. Loan 1913-HE, which led to the Rather than to attempt adding further cancellation of some US$49 million in requirements to institutions already undisbursed funds. The earlier challenged beyond their capacity by period of difficulties led to ongoing project requirements, the restructuring the second loan and to Bank developed a separate activity: financing materials of the second the "Women, Water and Development loan with funds from the first in Project". The project concept was to order to expedite closing of the combine improvements in water supply first loan. and sanitation with the generation of income-producing opportunities for Impact on Poverty Groupe and Women women* This approach was deemed particularly attractive because the 14. The SARs of the second and provision of water and sanitation third project specifically discuss services clearly eliminates a lot of the impact of the projects on urban the time women spend on carrying poverty groups, reporting that from water and improves their living 392 to 90Z of the urban poor in the environment, and thus frees time for cities participating will benefit more productive endeavors. For from the projects. Actual data is not reasons beyond the control of the available to assess whether this Bank (explained in paragraphs 106 and expectation was met. It is reasonable 107 of this report), the loan, which to conclude, however, that a was approved by the Board on June 22, significant proportion of the 1989, never became effective. projects' beneficiaries belong to the urban poor, because not only did the 17. Nevertheless, the -ialogue, appraisal identify specific locations started with project preparction and inhabited by urban poor to be served, continuing with subsequent attempts but with service coverage already at to modify the project so it could be the level of 65% or more in most implemented despite changing cities at the start of the projects, conditions and priorities, has it is clear that the majority of resulted in the adoption of policies those not then served belonged to the promoted by the project. In fact, low-income groups. the policies of SOLIDARIDAD, the financier of projects with high 15. The projects' impact on women social priorities, are those first is not discussed in any of the SARs, developed under the proposed "Water, nor in the PCR. To the extent that Women and Development Project". women of the urban poverty groups Conversations with officials at tend to be the families' providers of various levels and in various xiii institutions familiar with the policies. The third projects in intended project and SOLIDARIDAD- particular$ reveals some hopeful financed projects credit the signs for future progress$ especially development work of associated with the effective management of services the project with much of the by specific entities and the innovation implemented today by delegation of tariff-setting SOLIDARIDAD. %uthority by some state legislatures to the operating authority. Water Results and Sustainability tariffs have been significantly increased by the beneficiary 18. The projects more or less organizations as a result of the accomplished their physical targets, projects. although not necessarily as originally proposed, nor at the 20. The sustainability of the locations indicated in the appraisal investments made under the first and reports. A total of 227,000 water second loan is less certain than that supply and 130,000 sewerage of those under the third loan, connections, and the necessary although there are effective supporting infrastructure, were operating institutions among those constructed as a result of the loans, who participated in the first two and an estimated total of 1,870,000 loans. Technical assistance by the people are now being served by these borrower and the executing agency, connections (there is some double- particularly by CN&9 is considerably counting, as some people are included improved from the earlier period. in both water and sewerage service). The cities selected to participate in This compares with original estimates the third project after the of about 2,500,000 people. With loan elimination of CEAPAS appear to be disbursements amounting to US$ 215 reasonably effective. All have million, rather than US$265 million raised tariffs very significantly and total loan amount, the proportion of should be able to meet at least population served (75%) is roughly operating and maintenance expenses proportional to the percentage of the (which has become a requirement for loans disbursed (81%). national financial support). As a consequence, the chances for 19. The institutional and financial sustainability are higher for the obiectives of the first two loans investments financed under this were not met, and the third loan was project than for those of the first only marginally more successful. and second projects. This was largely the result of the Government's initial lack of strong 2gonomic and Environmental Aspects commitment to sector reform. Only during the last project did the 21. No attempt has been made to Government begin to formulate and recalculate the SAR Internal then implement policies which Financial Rates of Return. The clarified conditions of Government absence of adequate investment and participation in water supply and operating data, and the my chan3es sewerage investments, instituted a in participating water agencies, unified financing mechanism with the would have made the exercise same norms and conditions for all difficult and any comparison with the borrowers, and established the original figures meaningless. organizational framework and sector strategies to implement the new xiv 22. Environmental benefits can long experience of the need for and undoubtedly be ascribed to the impact of community participation in projects to the extent that this sectorp the Bank did not manage accomplishing the physical targets to have operating institutions increased the number of people participate in project development, supplied with potable water, or nor is there any strong evidence that benefiting from the removal of the borrower and executing agency wastewater. However, there is no were alerted to this need. information available to evaluate the Similarly, the Bank was unable to impact of the increased wastewater convince the executing agency to discharges on receiving waters this abandon the "hard-hat d,isign-and- is of particular concern since some build approach", in favor of of the sub-projects did not include including institutional development adequate wastewater treatment and training as an Important and facilities (or had none at all). essential part of its activities. There is a real possibility that the Fortunately, vith the establishment benefits of additional water supply of CR&, this problem seems to have are diminished by increased pollution been resolved. of surface vg:ers from the discharge of additional wastewater. 25. The Audit confirms the projects' performance ratings based The Role of the Bank on the findings of the PCRs. Accordingly the first two projects 23. Beyond a doubt the most are rated Unatisfactgry and the significant results to which the third as satisfactory. Bank's participation in the sector Sustainability of all three projects has contributed are the new financial is judged uncertain and their policies and the organizational institutional development structure the Government has achievements as partial, established for the sector. The Bank clearly has been very lenient in the Lessons enforcement of its contractual remedies for inadequate loan 26. The lessons arising from these implementation. But because of its three loans can be summarized as: continuous dialogue and support for the sector in difficult 0 Optimistic assumptions about circumstances, the Bank's suggestions Government participation and for rationalizing financial support support, no matter how rational mechanisms and institutional and sensible the anticipated development were finally accepted and actions, are no substitute for implemented. While it would be easy precise and binding agreements. to claim too much credit, there is no doubt that even the unsatisfactory 0 Government policies (grant first two medium cities projects have financing) which do not support contributed to the long-term project policies (cost recovery substantial improvement of the sector and loan financing) prevent as a whole. successful project Implementation. 24. The Bank was less successful in the implementation of the projects, 0 Implementing agencies, and and particularly, in ensuring the particularly local beneficiary quality of sub-projects. Despite its Xv authorities, must share the objectives being promoted and participate in determining the means to implement them. 0 An initial effort at using an inexperienced executing agancy (in terms of investment and project analysis and institutional development) to develop and supervise a large number of sub-projects requires the Bank to provide much greater support and technical assistance than was provided for these loane. PERFORMANCE AUDIT REPORT MEXICO MEDIUM CITIES WATER SUPPLY AND SEWERAGE PROJECT (LOAN 1186-ME) SECOND MEDIUM-SIZE CITIES WATER SUPPLY AND SEWERAGE PROJECT (LOAN 1913-ME) AND THIRD MEDIUM-SIZE CITIES WATER SUPPLY AND SEWERAGE PROJECT (LOAN 2281-ME) I. PROJECT SUMMARY A. Background 1. Mexico, a country of 762,605 square miles and a 1990 population of 82.5 million, forms the southern part of the North American sub-continent. Its climate varies from the hot arid north-central area to the subtropical hot and humid south-east. The country's topography varies widely. The principal features are the north central arid plains, bounded east and west by the Sierra Madre mountain chains, which join to form the mountainous central highlands the southern tropical lowlands; the coastal lowlands of the Pacific Ocean and of the Gulf of Mexico on the Caribbean Sea; and the arid Baja California peninsula. 2. During the 1990s Mexico's population growth has slowed considerably; it is now 21, compared to the 3.6% which prevailed during the 1950s and 1960s. It is the government's intention to reduce this rate further, to 1% by the year 2000. Some 751 of the population lives in urban areas, about 251 in the Mexico City Metropolitan Area (MNA). 3. The water resources of Mexico are very unevenly distributed. Average rainfall ranges from 100 m/year in the dry zones of the North to 5,000 mm/year in the tropical lowlands of the South. About 85% of the surface water resources are located at an elevation of 500 meters or below, while fully 751 of the population and 80% of industry are located above that elevation. These conditions have forced Mexico to develop a sophisticated system of water resource management, including the establishment in 1976 of a National Water Plan based on a UNDP/World Bank-assisted water resource study. Some 801 of exploited water is used in agriculture, 11% for municipal supplies, 71 in industry, and 21 for various other uses. About 71 of all irrigation water consists wholly or in part of treated wastewater. 4. Mexico's economy has been characterized by unusual turbulence over the past two decades (PCR paras. 2.15 -2.17). Up to 1970, the economy grew rather rapidly, but then it began to slow, with major recessions occurring in 1976 and 1982 (with a short-lived oil boom between those yesrs). In 1982 the governmant instituted an economic stabilization program which has brought inflation under control; the country's economy is now growing again, although slowly. In 1983, the constitution was amended to make municipalities responsible for infrastructure service, and so reduce their previous almost complete dependence on federal government funding. Implementation of this policy has been slow, and subsidies remain high, although they are now based on a more rational and more transparent system of eligibility. 2 5. As a consequence of the economic downturn of the 1980s, and of the somewhat haphazard implementation of the decentralization mandated by the constitutional amendment, recent progress in increasing water supply and sanitation service coverage has been slow compared to the rapid expansion achieved in the 1970s. This is illustrated in Table 1.1 below. During the 1980s, progress more or less matched population growth, but the backlog of unmet demand did not decrease. TABLE 1. 1: WATER SUPPLY AND SANITATION COVERAGE' (population in millions) Total Water Supply Sanitation Year Population Population % Population 2 1970 48.1 23.8 49 17.6 37 1980 67.4 46.9 70 32.9 49 1990 81.2 60.6 75 47.2 58 1990* 82.5 62.3 75 49.6 60 1991* 84.1 64.8 77 51.4 61 2000* 100 94 94 82 82 6. The lack of progress in extending water supply and sanitation coverage has been aggravated by problems with poor water quality. Disinfection is generally inadequate; many small systems lack functioning disinfection facilities. Largely as a result of this deficiency, gastro-intestinal disease is still one of the principal causes of morbidity and infant mortality. Although infant mortality declined from 55 per 1000 live births in 1976 to 35 per 1000 in 1986, water-related diseases still cause approximately 75 deaths per 100,000 among children between the ages of 1 and 4 years. The poorer population groups suffer particularly from an absence of adequate water and sanitation services; ominously, the number of cases of diarrhea reported for children under five in these groups rose from 334,000 in 1974 to 4,700,000 in 1986 (it must be noted, however, that it seems likely that this increase is primarily attributable to better reporting of a persistently high level of morbidity, rather than to a sudden increase in the number of cases). Comparable data for 1990 and 1991 indicate that cases have dropped to 3,479,000 and 2,993,000, respectively, as a consequence of a major campaign to improve water disinfection promoted by the Secretariat for Environmental Health. B. Bank Lending 7. The Bank has supported the water supply and sanitation sector in Mexico since 1973, when it made the first water supply loan. So far a total of six loans has been made, of which one is still ongoing. 3 (i) Mexico City Water Supply Project (Loan 909-HE), approved 1973, closed 1981, with 4h years' overrun. The US$90 million loan was to help finance a project estimated to cost US$155 million, the objective of which was to build bulk water supply facilities producing a total of 15.5 mo/sec. The project included various studies. Due to the urgency of increasing Mexico City's inadequate water supply, the project was launched with more than the usual uncertainties, and the eventual bulk water production was only 13 a?/sec. The actual cost (without studies) of the project was US$138 million, and US$ 11.8 million of * ' loan was canceled for various reasons, including non- compliance w.h the procurement rules of the Bank. (ii) Performance: The technical work of the executing agency, the Comisi6n de Agua del Valle de Mexico, was adequate, but institutional and financial performance was generally less satisfactory, although improving over time. The borrower was Nacional Financiera S A., whose performance was satisfactory. 8. The second, third and fourth loans are the subject of this audit memorandum: (i) Medium Cities Water Supply and Sewerage Project (Loan 1186-MS), approved in 1975, closed in 1984, with a 4 year delay. (ii) Second Medium Cities Water Supplv and Sewerage Proiect (Loan 1913-HE), approved in 1980, closed 2 years late in 1986. (iii) Third Medium Cities Water Supply and Sewerage Project (Loan 2281-HE), approved in 1983, closed in 1991 with only a 4-month delay. 9. The fifth loan that the Bank made in support of the sector, "The Women, Water and Development Project*, was an innovative project which proved impossible to implement in the institutional and policy environment which then existed (this environment also contributed to the lack of success on overall sector improvements experienced by the Medium Cities projects). The project was abandoned, prior to effectiveness, with the mutual consent of Mexico and the Bank. (i) Women, Water and Development Prolect (Loan 3101-ME), approved in 1989. At the request of the borrower, the date of effectiveness (April 26, 1991) was allowed to lapse and the loan never became effective. The loan of US$20 million was intended to support water and sanitation improvements and to provide productive opportunities for women (who were the major beneficiaries of the water and sanitation component, through the reduction in the time required for carrying water to the household and for the maintenance of sanitary conditions in the absence of facilities for the disposal of human wastes). (ii) Performance: Although Mexican sector staff familiar with the project agreed with it and still praises the concept, the loan was eventually canceled because the Government created a new concessionaire funding 4 arrangement "SOLIDARIDAD", providing grants for the same activities. SOLIDARIDAD is now implementing activities the project was designed to undertake, for which credit financing was no longer attractive once subsidies became available. 10. The project still being implemented is the Water Supply and Sar itation Sector Project: (1) Water Supply and Sanitation Sector Project (Loan 3271-ME), approved in 1990, with a proposed closing date of June 30, 1996. The project is expected to cost US$650.9 million, with a Bank loan of US$300 million. The project objectives are to assist the Government to implement the National Program for Water Supply and Sanitation: to carry out the sector reorganization, improve water and sanitation services (the beneficiaries are estimated to number about 3.2 million, mostly from lower-income groups), and promote pricing policies which progressively reduce the need for Government subsidies to the sector. (ii) Performance: The Government has carried out a comprehensive sector reorganization and has established clear sector financial policies which govern all investments, regardless of the executing agency. The process developed to implement the new policies, to evaluate projects and to help improve implementing agency competence appears to be functioning well. While it is too early to judge the overall impact, one measure of success is clearly very positive: by the end of 1992, 3h years ahead of schedule, disbursement will be completed. This is the first such success for the sector in Mexico. C. Objectives 11. The Bank's overall objectives in the sector today can be summarized as follows: (a) Development and Implementation of Sector Policies resulting in water and sanitation services sustainable over the long term with revenues derived from service charges; (b) Creation of Local (state or municipal) Water and Sanitation Organizations capable of providing adequate water and sanitation services, with a national system of technical support and training appropriate to assist local ortanizations as necessary. (c) Improving the Environment through water pollution control projects, policies and monitoring, and by the use for all projects of sound environmental assessments. 12. These objectives have changed little over the years, although emphasis and project-specific detailed objectives have varied in response to needs and the perceived receptiveness of the government and the borrower. For example, environmental considerations have become more important in the recent past. Although Mexico agreed in general with the principles inherent in these 5 objectives, the government did not establish appropriate criteria and policies until quite recently (as part of the latest sector reorganization preceding Loan 3271-HE), and its past support of these principles was not sufficient to bring about the institutional and cost recovery improvements that had been hoped for as a result of the Medium Cities projects. D. Sector Organization 1. Past Organizational Arrangements 13. In an effort to improve its effectiveness, the sector has undergone many reorganizations during the past two decades. Overall, the results have been less than spectacular, due either to inherent institutional inefficiencies, or to a lack of a sufficient number of staff of adequate competence (or, most likel?, to both). In 1988 the Bank undertook a review of operations in the sector , on which the following paragraphs (14 to 18) are based. 14. In 1948, the Secretariat of Hydraulic Resources (SRH) was charged with the planning and implementation of water supply and sanitation projects. SRH constructed facilities, which it then handed over for operation and maintenance to federal water supply and sanitation boards which it organized for this purpose. 15. At the time the Bank made its first water supply loan to Mexico (the Mexico City Water Supply Project: Loan 909-ME), Nacional Financiera was the financial agent charged with channeling external and local funds to the project. When the Medium Cities Water Supply and Sewerage Project (Loan 1186-ME) was approved, responsibility for financing the subprojects was given to the National Bank of Public Works and Services (BANOBRAS). Funding was to be provided to BANOBRAS by the Bank and the Government of Mexico. For some time prior to this loan and previously, the government had been providing grant funding to local authorities through a variety of channels, including BANOBRAS, and it continued to do so throughout the medium cities projects. 16. In 1976, water supply and sanitation activities were divided into "bulk" and "retail" functions. The former remained with SRH, now renamed the Secretariat of Agriculture and Hydraulic Resources (SARH), and responsibility for the latter was given to the Secretariat for Human Settlements and Public Works (SAHOP). 17. In 1980, in a first attempt at decentralization, more responsibility for the sector was given to state organisms, under the overall direction of SAHOP. Some states decentralized responsibility further, to municipalities. Unfortunately, no program for financing investments accompanied this decentralization, and communities clamored for subsidies for both investment and operation. I Mexico Urban Water Supply and Sewerage Sector Study: Institutional and Financial Constraints. Report No. 7109, June 30, 1988. Infrastructure and Energy Operations Division, Latin America and the Caribbean Region, the World Bank. 6 18. In 1982, sector responsibilities were transferred to the Secretariat of Urban Development and Ecology (SEDUE). SEDUE also became the executing agency for external loans to the sector. Under this arrangement, construction responsibilities were also decentralized. SARE continued its activities in development and implementation (including construction) of bulk supply facilities. 19. As noted above, Article 115 of the Constitution was modified and expanded in 1983 to establish the responsibility of municipalities for the provision of water supply and sanitation services, giving them authority over all aspects of the development, implementation and operation of the necessary facilities. 20. It is relatively easy to find staff familiar with the history of FIFAPA in BANOBRAS because reorganizations and staff changes within that organization have been less dramatic than elsewhere (which, of course, also explains why among the various organizations, BANOBRAS has performed best). Nevertheless, it was possible during the performance audit field visits to review FIFAPA's historical development with former SEDUE staff who had moved on to other positions within the water sector. 21. Discussions with former SEDUE staff reveal that: (i) investment decisions where guided more by political considerations than by project priorities and investment needs; (ii) project unit staff had relatively little influence on investment decisions, even though the loan conditions charged them with responsibility for project appraisal; and (iii) supervision of implementation was systematically neglected. Staff found themselves in a rather untenable situation, having to follow guidelines agreed upon by the CO and the Bank, obey superiors motivated by or responding to political considerations, and deal with prospective clients (municipalities) rejecting the substitution of loans for the heretofore free government funds. Some staff left the organization in frustration caused by their inability to effect changes and perform according to appropriate professional or ethical standards. 2. Present Organizational Arrangements 22. At the beginning of 1989, the National Water Commission (CRA) was created as an autonomous Jeconcentrated organization within SARH. The functions of CNA can be described as policy development in the water resource sector (including water supply and sanitation); definition of priorities; planning of developauuc; protection and administration of water resources; facilitating of decentralization; and coordination of federal, state and local administration of the sector. In particular, CRA had the responsibility for developing the National Program for Water Supply and Sewerage for the period 1989-94. 23. In 1992, the water supply and sanitation sector functions and environmental protection activities of SEDUE were assumed by the Secretariat for Social Development (SEDESOL). At the same time, the responsibilities of SARH, CNA and SEDESOL were defined in greater detail in a decree published on May 25, 1992, in the official gazette of the Government of Mexico. 24. As a result of tha 1983 devolution of responsibility to municipalities, institutional reforms are taking place at the state and local levels, assisted 7 by CNA, to implement the organizational structure shown in Figure 1. Undoubtedly the use of formal eligibility criteria for grant support by Government is encouraging this process. At the local level, there presently exists a mixture of organizational arrangements for the provision of water supply and sanitation services, ranging from municipal departments to autonomous municipal authorities. A recent phenomenon is the involvement of state organizations, to which municipalities contract part or all of their water and sanitation service functions. 25. At the state level, two kinds of organizations are developing. Where municipalities so prefer, a state water supply and sanitation authority assumes full responsibility for planning, implementation and operation of facilities, along the lines of the model successfully used for many years in Brazil. This solution is particularly useful for small communities, the s4ze of which would make independent operation inefficient. A second arrangement is based on a state organization which provides technical assistance and support in project development and construction (also for operations, if necessary), while leaving the actual operation of facilities to the local municipality. This solution provides most of the benefits of economies of scale, but preserves the autonomy of the municipality. 26. At the federal level, there are five major actors in the urban water supply and sanitation sector. CNA is responsible for technical support, as previously described. INTA, another autonomous organization within SARH, assists CNA in the design of systems and the implementation of the sector training program. BANOBRAS is responsible for channelling all funds to the urban water supply and sanitation sector, whatever the source of these funds. SEDESOL is responsible for setting overall policies and standards, and the Sub-Secretariat for Planning and Budgeting (SSPP) within the Secretariat of Finance and Public Credit (SHCP) participates in preparing, and approves, the annual investment program for the sector. In rural areas, SOLIDARIDAD works with communities to promote economic development, including the provision of potable water and sanitation facilities, and government funds are channeled through it. CIA provides technical assistance for rural water supply and sanitation projects. 8 II. THE PROJECTS AND THEIR OBJECTIVES A. Medium Cities Financing Mechanism (FIFAPA) 27. The first medium cities water supply and sewerage project was conceived as more than simply an investment project to expand and improve water supply and sanitation services in Mexico's medium-sized cities. The Bank's experience with the Mexico City Water Supply Project (Loan 909-ME) demonstrated the need for fundamental changes in the sector. Mexican authorities wanted to use a sector loan to accomplish these changes, but the Bank insisted on a more limited program due to the absence of an experienced sector institution to manage such an effort (PCR, para. 3.06). 28. The compromise was a program loan designed to finance improvements in a number of cities and technical assistance to improve institutions at every level. The objectives did not change significantly during the Implementation of the three medium cities projects (see paras. 40, 55 and 71 for project objectives, and para. 11 for overall sector objectives). To accomplish one of the sector objectives, the establishment of aUnified System of Sector Financing, Mexico and the Bank agreed to charge BANOBRAS with the formation and management of the "Investment Fund for Water Supply and Sewerage Works" (FIFAPA) (PCR, para. 3.08). 29. FIFAPA was designed as a revolving fund, financed initially by the Bank loan and Government contributions, and later additionally by interest and loan payments from subproject beneficiaries borrowing from FIFAPA. The intent of establishing FIFAPA was to make the sector ultimately self-sufficient financially by channelling all resources as loans through FIFAPA. Other objectives integral to the revolving fund concept were the use of clear project selection and funding priority criteria, the design of projects on the basis of least-cost solutions, the acceptance by beneficiaries of the concepts of cost recovery through tariffs, and of financing through loans rather than through grants. 30. Unfortunately, potential beneficiaries were not consulted in the formulation.of these objectives during the preparation of the first project, nor during the second or the third. With the Government continuing its grant- financing of water and sewerage investment, municipal agencies had little inclination, and few incentives, to participate in a financing scheme based on loan-financing. Rather than a bold new demonstration on how to overcome financial and institutional problems and create a financially self-sufficient sector, FIFAPA became a lender of last resort, at best. The institutional, financial and economic objectives set for FIFAPA were not reached in any of the three medium cities projects. An additional complication during the second and third projects was the high rate of inflation which had not been reflected in the feasibility studies, raising project costs. Furthermore, BANOBRAS subloan agreements incorporated fixed interest rates, and therefore had to be renegotiated with borrowers (previously used only to grant-financing) in order to reflect more realistic and much higher current interest rates. 31. Although at least the first two projects must be judged unsatisfactory when viewed in isolation (only their physical objectives can be said to have been met with a reasonable degree of success), nevertheless the resultant establishment 9 of FIFAPA had a positive impact on the sector. The institutional framework established as part of the first medium cities project remains in place, although the institutions themselves have changed considerably. This continuity of institutions, and the resultant continuing discussion and review of the need for rerorms in sector management and financing and institutional arrangements, has ultimately resulted in the implementation of detailed sector strategies and organizational arrangements (firmly established in the National Program for Water Supply and Sewerage) which meet Bank and Government criteria and which appear to be successful in improving the sector's performance. BANOBRAS' management of FIFAPA has steadily improved and is today a highlight of the sector's improved performance. MEDIUM CITIES WATER SUPPLY AND SEWERAGE PROJECT (LOAN 1186-MB) A. The Project and its Objectives 32. The project consisted of the extension of water supply and sewerage services to new customers, and improvements in service to existing customers, in eight cities: Ciudad Victoria; Jalapa; Mexicali; Morelia; Reynosa; Salamanca; Tampico-Cuidad Madero; and Tuxtla Cutierres. The physical works included construction of water and sewer mains, installation of new meters and repair of existing ones, and leak detection programs. The project also included a training program for sector staff connected to the project. 33. The borrower of the loan was BANOBRAS, which established, as part of the loan implementation, a revolving Investment Fund for Water Supply and Sewerage Works (FIFAPA), which was to on-lend funds from government equity contributions and the Bank loan. SARK was the executing agency, responsible for project preparation and implementation (tasks which at that time were a responsibility of the federal government). 34. The project had a number of objectives, conceived to achieve an overall upgrading of sector operations. The first was to improve service quantitatively and qualitatively, to achieve local financial self-sufficiency through tariff increases, and to develop stronger local operating institutions. The second was to develop, within SARN, the capacity to monitor subproject preparation, appraisal, design, execution and administration. This was needed because at that time the Mexican approach to the design of water supply and sewerage systems was based on conventional engineering techniques and emphasized short-term solutions to meet demands, with little regard to the evaluation of alternatives to determine the least-cost solution and to the definition of institutional development needs. The third major objective was, through the FIFAPA mechanism, to begin to replace government grants with loan financing, and to introduce, progressively, tariffs aimed at greater cost recovery. B. Imlementation 35. Loan 1186-ME became effective on April 26, 1976. The first sub-loan agreement was sigred in early 1977, and construction began in late 1977. Preparation of other subprojects was slow; two of the initial beneficiaries withdrew, refusing to raise tariffs (but one other joined); and eventually three subprojects that were not completed at loan closing were transferred to the 10 second project (Loan 1913-ME). The project started slowly because of internal institutional problems within SARH, which continued when responsibility was transferred to SAHOP, and lack of adequate coordination between these institutions and BANOBRAS. The Bank's early conclusion, that it would have been premature to consider a sector loan, was amply justified, as demonstrated by SARE's and SAHOP's inability to prepare subprojects adequately (PCR paras. 4.08 - 4.09). 36. By the end of 1978, after repeated suggestions by the Bank on possible means of improving project preparation, progress had indeed improvedl during 1979-1981 the project was being successfully implemented. As a consequence, the Bank agreed, initially, to an extension of the closing date to January 14, 1982, and later to January 14, 1984, four years beyond the original closing date. Regrettably, the improved rate of progress was short-lived. Just prior to a new administration taking office, an economic and financial crisis stopped the project completely: government equity funds for FIFAPA loans were not available, and SAHOP reduced its supervision activities pending the new government's takeover. In December 1982, project responsibility was transferred to SEDUE, and the problems (inadequate staffing, lack of coordination) experienced earlier with SARH, SAHOP and BANOBRAS, started anew (PCR para. 4.13). C. Proiect Costs and Disbursements 37. The appraisal report estimated the cost of the project as US$100.0 million. In 1981, based on then-current prices and exchange rates, and changes in project components, thr total cost of the revised project was estimated to be US$78.2 million. Finally, as of the closing date, the cost of the reconstituted project was US$76.5 million (see PCR Table 3.1, Part III). 38. The PCR provides a detailed description of project costs (Schedule 1, page 19) and disbursements (Table 4, page 54) and the various changes they represent. As a result of the many changes, deletions and additions, and the substitution of new borrowers for others who declined to participate, appraisal estimates and actual project costs are not directly comparable. Further complicating any such attempt are the diverging trends of the domestic price index (1977 - 1984 index 9.575) and the Peso exchange rate (1977 - 1984 index: 6.461), which resulted in an appreciation of the Peso vis-I-vis the dollar of 48.2%. For those project components which were included both in the original SAR and in the final project, the original SAR estimates were US$77.8 million, while the final actual costs were US$52.7 million (PCR Schedule 1, page 19). 39. As part of a Bank-wide "Special Action ProgramO, to assist Mexico's attempt to revitalize its economy, the Bank agreed to various modifications in project components and funding reallocations in order to disburse the remain:,ng funds by the closing date. This resulted in the reallocation of funds from one category to another, and also in an increase in the permitted disbursement percentage (from 40% to 1001 for equipment and from 40% to 702 for civil works) in order to accomplish more rapid disbursement of the remaining loan funds. Uncompleted portions of two subprojects at the closing date were transferred to the Second Medium Cities Project (Loan 1913-HE) and the financing of some equipment and materials of the Second Project was transferred to Loan 1186-HE to accelerate 11 disbursement. Eventually, the loan of US$40.0 million was disbursed by May 23, 1984, apart from a small amount, US$0.21 million, which was canceled. D. Procurement 40. In general, the procurement process was satisfactory, although some delays occurred in the preparation of documentation. Considerable training of BANOBRAS staff was carried out initially, but the documentation and review process was considered to be excessive by the staff of the borrower BANOBRAS and the Project Implementation Unit (PIU) of the executing agencies. Civil works were undertaken exclusively by Mexican contractors, a reflection of the well-developed and competent construction industry in Mexico. E. Performance of Contractors and Consultants 41. The performance of contractors was satisfactory. Consultants performed equally well in the design of facilities, but less well in the preparation of feasibility reports. A particular problem was the quality of economic and financial analysis of project alternatives, which also reflected the veakness of the PIU staff. As a consequence, Bank staff had to devote a considerable amount of time to improving subproject feasibility studies. A training program for project-connected sector staff, for which funding was available and which could have helped alleviate the problem, was not implemented until a late stage in the project. The delay was a result of the priorities which the PIU set for itself, which emphasized design and construction of facilities rather than training and institution-building. F. Results and Sustainability 42. The result of this loan can best be described as mixed. Its physical objectives were more or less accomplisheds the number of people served increased substantially. The PCR (Part III, Table 6.1) compares projections and actual population served. The table shows that coverage targets were largely met, but that indicator reflecting institutional improvements, namely unaccounted-for- water reductions, shows a lack of success, and financial indicators (tariffs), give a mixed picture. These indicators reflect the fact that, amongst other shortcomings, the technical assistance components of the project, such as training and studies to identify measures for institutional improvements were either not implemented, or not implemented in time (corresponding to the PIU's preference for construction, and disinclination to undertake institution-building and training activities). 43. The 1982 (design year) population of all cities participating in this project was lower than that estimated at appraisal. Actual populations for individual cities ranged from 92% to 772 of previous estimates. Coverages for water supply and sewerage similarly were below those projected, with those of five c.ties ranging from 652 to 952 of projections, and only the water supply coverage of one city exceeding the target (by a 2% margin). 44. Given the beneficiary organizations' apparent lack of success in improving institutional and financial deficiencies, there has to be serious doubt about the long-term sustainability of the physical improvements funded by the loan. 12 Subsequent lending operations attempted to build on the lessons drawn from this experience, and incorporated expanded conditions for institutional and financial improvements. SECOND MEDIUM-STRe CITIRS WATER SUPPLY AND SEWERAGE PROJECT (LOAN 1913-ME) A. The Project and its Objectives 45. The project was designed to support the Government's National Plan for Urban Development (PNDU) which, amongst other measures, promoted water supply and sanitation in urban centers other than Mexico City in an effort to decentralize urban growth and encourage industrial development in medium cities. Government policy for the sector being developed during the time of project preparation was to implement some of the recommendations contained in a 1978 Bank/PAHO sector report: reducing grants in favor of loan-financing; raising tariffs; and decentralizing responsibility for the provision of water and sanitation services to local autonomous agencies. 46. The cities included in the project, at the time of appraisal, were: Monclovia; Nogales; Tepicl Toluca; Ciudad Juarez; Matamoros; and Queretaro. At the time of appraisal, feasibility reports for the first four cities were available; the other three were submitted later. 47. In addition to supporting the general PNDU objectives, specific project objectives included strengthening the management and operating capacity of the newly-created autonomous water and sanitation agencies, by providing technical assistance (similar to the components earlier included in the first Medium Cities loan); expansion of water and sewerage services; implementation of a chlorination program to improve water quality for all the inhabitants of the project cities; and the provision of sewage treatment in those cities with pressing pollution problems or where there were downstream water users. 48. The most noticeable characteristic of the project preparation and appraisal process is the optimism exhibited by Bank staff. There is little reference in the SAR to the very serious implementation problems encountered during the first Medium Cities Project. The Second Project was appraised during a period when both borrower and executing agency had made substantial improvements in their coordination and the staffing of the PIU, and the "oil boom" engendered a spirit of optimism at all levels of government which seems to have infected the Bank staff with equal enthusiasm about the feasibility of implementing the government's rather ambitious plans for the sector. B. Implementation 49. Loan 1913-HE was approved October 4, 1980, but effectiveness was delayed until June 23, 1981, since Mexico's legal opinion on documents was not forthcoming until that date. Arrangements followed the pattern of Loan 1186-ME: thc borrower was BANOBRAS, the executing agency SAROP. The project started during a period when SAHOP was performing well and when coordination with BANOBRAS was adequate. However towards the end of 1982, SAROP, anticipating changes due to the impending change in administration, reduced its pruject activities (PCR paras. 4.13, 4.18), and was subsequently replaced as executing 13 agency by SEDUE when the new administration took office. The earlier problems of lack of staff in PIU and inadequate coordination with BANOBRAS resurfaced, although by 1984 progress seemed to have been restored (PCR para. 4.21). 50. The economic crisis affected project implementation as well. Counterpart funds were not available for several months during 1983, and the project scope was reduced. Several subprojects were eventually eliminated, and in 1986 uncompleted subprojects were transferred to the Third Project (PCR Schedule 2, page 24, provides information on the various modifications of sub-projects). The very rapid inflation during that period (PCR para. 4.20) affected project costs and, together vith higher interest rates, required time-consuming renegotiations of subloans between BANOBRAS and the beneficiaries. The depreciation (292) of the Peso vie-&-vis the US$ also resulted in lower dollar requirements (PCR para. 4.17). 51. In 1985, the IU was abolished and its functions distributed to various units of SEDUE. As a consequence, coordination with BANOBRAS again suffered, and supervision of subprojects effectively ceased (PCR para. 4.21). Neither BANOBRAS nor SEDUE was able to comply with the "Plan of Action" that they had earlier agreed with the Bank. The government stopped making counterpart funds for FIFAPA available, stopping project activities entirely for a few months. The continuation of grants to finance water supply and sanitation investments discouraged potential beneficiaries of medium cities projects from participating in PIFAPA. When the loan was finally closed, two years late (on November 30, 1986), total disbursement vas only US$75.72 million, and US$49.28 million was canceled (PCR para. 4.23). C. Project Costs and Disbursements 52. As described above, the project was modified considerably and repeatedly; a comparison of the original total cost estimates (US$297 million) with the actual total costs of the components eventually implemented (US$115.9 million) is therefore meaningless (PCR Table 3.2; Table in PCR Part IIs Borrower's review). For those components which were included both in the original SAR and in the final project, the estimated costs in the SAR (after adjustments in scope) were US$136.8 million, and the final costs were US$84.8 million (PCR Schedule 2, p. 24). 53. Disbursements are shown on PCR Table 4 (page 54). For the same reasons as project costs, total disbursements (US$75.7 million) are not comparable to projections at the time of appraisal (us$125 million). D. Procurement 54. In general, the procurement process was satisfactory, although some delays occurred in the preparation of documentation, as in Loan 1186-MS. Procurement was handled properly at first. In 1985, when SEDUE abolished the PIU, inexperienced staff changed procurement procedures, disregarding Bank requirements and Bank staff advice. Only after the Bank threatened to stop disbursements was agreement finally reached on bidding procedures for all Bank operations in Mexico. 14 E. Performance of Contractors and Consultants 55. The performance of contractors continued to be satisfactory, as it had been in the First Project. Consultants also continued in the same pattern, performing well in conventional tasks, such as the design of facilities, but less well in the preparation of feasibility reports. The quality of economic and financial analysis of project alternatives remained a problem, reflecting the weakness of the PIU staff. F. Results and Sustainability 56. The results of this Project, just like those of the First Medium Cities Project, can best be described as mixed. Physical objectives were more or less accomplished: the number of people served increased substantially, although it is not possible to compare the figures projected in the SAR with those actually achieved (the SAR targets were preliminary and were modified in the feasibility reports completed after appraisal. In addition, the project components eventually implemented varied a great deal from those appraised). The PCR compares feasibility report projections (Part III, Table 6.2) with the actual population served, considering only those components included in the appraisal report which were actually implemented. The table shows that coverage targets were largely met, but that indicators reflecting institutional improvements, namely unaccounted-for-water reductions, show a lack of success, and financial indicators (tariffs) give a mixed picture. These indicators reflect the fact that, amongst other shortcomings, the technical assistance components of the project, such as training and studies to identify measures for institutional improvements, were either not implemented, or not implemented in time. This was an inevitable consequence of the PIU's preference for construction, and its disinclination to undertake institution-building and training activities. 57. For its analysis, the PCR considers projections based on feasibility reports which were completed after the preparation of the SAR. The 1988 (design year) population of all cities originally participating in this project was lower than estimated at appraisal: actual populations for individual cities ranged from 402 to 962 of SAR estimates. The percentage coverage for water supply and sewerage achieved through the project was below that projected in the SAR, being as low as 252 of the projections in some cases. Neither the SAt projections nor the revised lower feasibility report projections were met. 58. Given the lack of success in improving institutional and financial deficiencies by the beneficiary organizations, there had to be serious doubt about the long-term sustainability of the physical improvements that were accomplished by the conclusion of the lending operation. The progress of institutional and financial improvements were particularly disappointing, leading the Bank eventually to deny further requests for postponing the closing date, which in turn led to the cancellation of about 40% of the loan amount. 15 THIRD MEDIUM-SIZE CITIES WATER SUPPLY AND SEWERAGE PROJECT (LOAN 2281-HE) A. The Project and its Obiectives 59. This project was considered essentially as a further extension of the support provided to the Government's National Plan for Urban Development begun with Loan 1913-HE. The Third Project was prepared during 1981 and early 1982, during a period when both the first two Medium Cities Projects were being implemented with a reasonable degree of success. When the Mexican economy took a turn for the worse, and lack of counterpart funding virtually halted progress of the first two projects during a few months in 1983, Bank staff concluded that the scarcity of Government funds would result in much greater dependence on credit funding through FIFAPA. On the basis of this assumption, rather than awaiting improvements, the Bank approved Loan 2281-ME on May 17, 1983, presumably as much as a demonstration of support for Mexico as in hopes that FIFAPA would finally, out of necessity, become the major funding mechanism for the sector. Due to slow processing by Mexico, the loan only became effective on February 27, 1984. 60. The project included as a beneficiary the Sinaloa State Water Board (CEAPAS) and through it regional companies which included Culiacan, Los Mochis, Topolobampo and Mazatlan, as well as medium cities in other states. Included in the project, at the time of appraisal, were Celaya, Colima, Cuidad Guzman and Tapachula. Later, La Paz, Cuidad Obregon, Pachuca, Acapulco, Manzanillo and Puerto Vallarta were added. The project also included 'stitutional support and a health education component for Sinaloa State to improve environmental health conditions there. When, as a result of the amendment to the national constitution's Article 115, the state abolished CEAPAS as an operating institution, the health education component was eliminated from the project. 61. The project objectives closely paralleled those of the Second Project. In addition, the project provided for the support of a newly-created state water company and a health education program for Sinaloa. As a refinement, the project eliminated beneficiary selection criteria but elaborated the criteria to be used in the selection of sub-projects. It also increased the FIFAPA on-lending rate which, as a result of inflation, had become unrealistic and resulted in depleting FIFAPA. 62. Just as in the case of the Second Project, the optimism exhibited by Bank staff during appraisal is noteworthy. No evidence was presented which would have justified the conclusion that FIFAPA would become the major financier of the sector under a policy of cost recovery, other than the belief of the appraisal team that the scarcity of funds would force the government to implement such a policy. As it happened, grant-financing continued virtually without change. The Bank continued its exclusive dependence on BANOBRAS and SEDUE for project implementation, and made no discernable attempt to ensure that beneficiaries would be consulted in project preparation. There again is little reference in the SAR about the very serious implementation problems encountered during the First and Second Projects, and no attempt was made to formulate an agreement between BANOBRAS and SEDUE to improve coordination. 16 B. Implementation 63. Loan 2281-HE became effective on February 27, 1984, but had to be modified as early as 1985, in order to reflect the latest Government actions. To comply with the amendment to Article 115 of the Mexican Constitution, CEAPAS lost its executive function and became a technical assistance agency, and the project component supporting the company was changed to make the regional companies the direct beneficiaries. In 1986, further modifications were made in order to allow loan funds to be used for the repair of earthquake damage in Mexico City. In 1987, by mutual agreement between BANOBRAS and the Bank, the uncompleted projects of Loan 1913-ME were transferred to Loan 2281-HE. 64. Arrangements followed the pattern of the previous two loans: the borrower was BANOBRAS, the executing agency SAHOP. The elimination of SEDUE's PIU and personnel changes at BANOBRAS had a detrimental effect on project implementation. The project was further retarded by the slow production of sub-project feasibility studies. After the closing of Loan 1913-ME and the cancellation of the undisbursed loan amount, project performance improved in 1987. In 1988, the performance of SEDUE improved as well (PCR paras. 4.29, 4.30). 65. Schedule 3 of the PCR (page 28) provides information on the various modifications, additions and deletions to the original project. The loan was closed on December 31, 1990, six months later than originally planned. The undisbursed amount of US$5 million was canceled. 66. During the implementation of the Third Project, the Bank concluded that the fundamental problems were due less to the performance of the central agencies, but rather to a lack of solid support from Government for the proposed FIFAPA financing process. As a consequence, the Bank and Mexican sector staff conducted a study which found, inter alla, that between 1981 and 198f, 85% of federal resources allocated to the sector were provided in the form of grants. The report made a number of recommendations which, although they were ever formally accepted by the Mexican Government, are now part of the National Program for Water Supply and Sewerage, and underlie the Bank's Sector Loan 3271-ME. 67. During the sector's "austerity" years from 1986 to 1989, annual investments ranged from 60 to 403 billion Pesos, of which Government grant funding represented from 28 to 51 percent. In 1990, total investment was 1,627 billion Pesos, of which grants comprised 71 percent, and in 1991 the estimated figures are 3,054 billion Pesos and 86 percent. Although the grant proportion remains discouragingly high, the new policies of the Government provide for clear and transparent guidelines on the allocation of grant funds, based on the socio- economic status of the beneficiaries. C. Project Costs and Disbursements 68. As described above, the project was modified considerably and on an ongoing basis, as shown on PCR Schedule 3 (page 28). A comparison of the original total cost estimates in the SAR (US$192.8 million) with the actual costs of projects eventually implemented (US$128.6 million, including the US$15.05 million transferred to relief operations after the earthquake) is therefore meaningless. For those components which were included both in the original SAR and in the 17 final project, the estimated costs were US$22.6 million, and the final costs US$15.4 million. 69. Table 4 (page 54) and Table 5.3 (page 57) of Part III of the PCR compare actual and forecast schedules of disbursements and the allocation of proceeds. For the same reasons as project costs, they are not comparable to projections at the time of appraisal (total forecast disbursementst $110.3 million; actual disbursements, including earthquake relief: US$95.3 million). D. Procurement 70. In general, the procurement process vas satisfactory, although some delays occurred in the preparation of documentation. Procurement was handled properly, Edthough Mexican staff considered the Bank's documentation requirements to be unduly onerous. E. Performance of Contractorg and Consultants 71. Paragraph 63 above comments on the performance of contractors, consultants and SEDUE staff during the implementation of the Second Project (1980-86). This of course overlapped the Third Project (1983-91) to a considerable extent, and the same comments therefore apply for that period and the remainder of the project until 1991. F. Results and Sustainability 72. The results of this Project, just like those of the preceding First and Second Projects, can only be described as mixed. Physical objectives were more or less accomplished (see PCR, Part II, Table 6.3), although, as in the first two projects, some cities originally included dropped out of the project and others were included as substitutes. This is basically a result of the elimination of CEAPAS from the project. Judging from available information, the organizations substituted are more effective than would probably have been true for some of the smaller local units formerly administered by CEAPAS. For example, performance of the Culiacan and the Los Mochis Water and Sewer Authorities are first-rate, but most of Sinaloa's municipal water and sewer agencies are unable to match their state capital's performance, despite efforts by the State to organize technical assistance programs with the help of the better performers to replace CRAPAS' efforts. However, these smaller operating units are not part of the project, and the performance of the substitutes can reasonably be expected to be better. 73. With the exception of Mazatlan, both the population and service coverage projections for this project were better than in the case of the preceding projects. The 1991 (design year) populations were close to those estimated in the SAR, and the maximum -ariation for either water or sewerage coverage was only 3% from the targets. For Mazatlan, however, the SAR projects the 1991 population for 1991 as 316,000, and the feasibility report used in the PCR analysis estimates 423,000, while the actual population was estimated to be 272,000. Even so, the project did not achieve its targets: some 150,000 people, rather than the projected 269,000 are served as a result of the project (552 of the total population, rather than the estimated 851). 18 74. The cities selected to participate in the project after CEAPAS was eliminated appear to be reasonably effective. All have raised tariffs very significantly and should be able to meet at least operating and maintenance expenses (now a requirement for national financial support). As a consequence, the chances for sustainability are higher for the investments financed under this project than for those of the first and second projects. III. PERFORMANCE OF THE BANK AND THE BORROWER A. Performance of the Bank 75. The most distinguishing feature of the Bank's performance has been the consistent support provided for the Mexican Government's efforts to improve sector performance over the duration of the three loans, from 1975 to 1990. This persistence in working for long-term goals, despite continuous setbacks and frequently only questionable support from the sector and even from the Government itself, was remarkable, to say the least. At the moment, the apparent success resulting from the latest organizational changes and the application of the decentralization and cost recovery policies appears to justify the Bank' dogged pursuit of long-term sector objectives in Mexico despite temporary setbacks. Whether this judgement is more realistic than the many which turned out to be far too optimistic throughout the implementation of the three projects remains to be seen, though indications are that this time, assisted by Sector Loan 3271-ME, real changes are taking place. 76. The basic problem with the Bank's performance has been its excessive over- optimism, for example in assessing the seriousness of potential risks, and an unrealistic trust in the sector's ability to make fundamental changes without explicit Government undertakings to support such changes. As late as the time of the appraisal of the Third Medium Cities Project, the Bank mistakenly assumed that, due to economic constraints, the government would reduce grant-financing and act firmly in support of its own (and the Bank's) declared objectives to adopt loan funding for infrastructure and to insist on cost recovery. Although this was a reasonable assumption, it was not backed by a firm Government commitment, and subsequently turned out to be wrong. 77. The project Women, Water and Development (Loan 3101-ME) is another example of over-optimism resulting in good intentions not reaching successful implementation, due to a lack of agreement with Government (as opposed to sector officials). The ideas that the project sought to implement have found universal acceptance in Mexico; in fact, they are now being implemented by the newly- created Program SOLIDARIDAD. Nevertheless, because there was no commitment by Government for the Project's attempt to finance socially-oriented investments through credits rather than grants, the loan in fact never became effe4tive and was canceled by mutual agreement between the Bank and Government. The SOLIDARIDAD activities, which are similar to those which had been proposed for implementation under the Project, are financed by grants from the Government. 78. The PCRs examine a number of problems of the Bank's performance in detail, and do so very well. Two of these problems deserve additional attention. The 19 f irst relates to the lack of Bank efforts to insist on che participation of beneficiaries in project development. The Bank's experience worldwide has clearly shown that the sector is usually unable to achieve permanent improvements without the active participation of beneficiaries. Not to include community participation as a fundamental requirement of these projects, especially at a time when the Government itself was attempting to decentralize the sector and empower mnicipalities to provide services, was clearly a serious deficiency. 79. The second deficiency in Bank performance relates to training and technical assstance to improve the operating capacity of local water authorities, and that of tie borrower and executing agency. The Bank included funds for these activities, but in fact training and technical assistance efforts remained inadequate, apparently due to lack of interest on the part of the executing agency. More forceful action by the Bank was warranted, for without increasing the competence (particularly of local agencies) their effective participation in project development was clearly not possible. 80. In addition to foreseeable problems, project implementation was also affected by unforeseeable difficulties, such as the 1987 earthquake and the serious economic difficulties which the country experienced at various times. The Bank demonstrated a great degree of flexibility in making appropriate adjustments in disbursements in order to help the Government overcome these difficulties, particularly by reallocating loan funds to Mexico City's earthquake reconstruction efforts. B. Performance of the Borrower and the Executing Agency 81. The greatest success clearly has been the management of FIFAPA by the borrower, BANOBRAS. The borrower clearly concentrated on establishing and maintaining an efficient funding mechanism for the sector. Unfortunately, the concentration on this aspect of FIFAPA led to reduced emphasis on supervising the financial and institutional performance of the beneficiary institutions. As a consequence, the borrower was less successful in helping local authorities to improve their financial management and achieve fiscal independence. Field mission conversation with the borrower's staff confirmed that the Government's continued provision of concessionary funds inhibited BANOBRAS from providing stronger leadership in changing the funding and financial practices of the sector. Understandably, supervision of beneficiaries' financial management in particular was therefore practically non-existent. Inadequate coordination and rivalries with the executing agency at times seriously affected the implementation of the Projects. 82. The executing agencies changed so often, and the PIU was reorganized so frequently, that the Projects suffered. Despite a lack of a sufficient number of adequately trained staff, the borrower did not take advantage of the funds available for training and technical assistance. As a result, Bank staff had to spend more time than anticipated in reviewing or revising feasibility and other project documentation, reducing the time they had available to work directly with beneficiaries. Discussion with former staff of the PIU in its various incarnations indicate that working conditions frequently were nothing less than chaotic, and staff morale low. Given the traditions of the sector, it is also not surprising that, despite agreements with the Bank and management confessions to 20 the contrary, decentralization was never highly popular and was not implemented with any consistent effort. 83. Neither the borrower nor the executing agency devoted adequate resources to assist and supervise beneficiaries. As a result, institutional improvements will have to be addressed in future operations if long-term sustainability of facilities financed under the Medium Cities loans is to be assured. For further discussion of this topic, see "Institutional Development" below. C. Project Relationshis 84. The Bank's relationship with the borrower, the executing agency and the local institutions has been excellent9 despite the many changes in the organization of the various agencies which acted as the executing agency. The influence of the Bank clearly resulted more from the consistent promotion for change than from any single action (including the cancellation of the undisbursed amount of the Second Medium Cities Loan). It is clear t1tat the Bank played a considerable part in influencing the sector to adopt more appropriate policies (loan- rather than grant-financing, enhanced cost recovery, etc.), even though the Medium Cities Projects themselves did not incorporate these policies with any consistency. The latest project, Sector Loan 3271-ME, however, is based on these principles, and the Government has adopted appropriate policies for the sactor as a whole. D. Institutional Development 85. Because the planning and implementation of the three projects overlapped considerably (see Project Chronology), it is unrealistic to attribute particular successes or failures in institution-building to one project or another. Instead, the series of projects should be regarded as a longer-term means for achieving necessary reforms in the institutional framework for the sector. 86. In this respect, the projects had some success, although far less than anticipated at the time when they were approved. The transition between the past and present institutional arrangements, described above, would almost certainly not have occurred without the projects. However, few of these benefits occurred during the actual project implementation period. 87. Institutional development under the first loan was disappointing. At the central level, the frequent reorganizations of the PIU and its host secretariats, coupled with lack of a sufficient number of appropriately-trained staff, prevented the PIU from discharging its functions effectively. The elimination of the PIU in 1985, during implementation of the second project, contrary to agreements with the Bank, demonstrated how little influence the Bank was able to exert on the executing agency. The lack of coordination and of collaboration between the borrower and executing agency was clearly detrimental to the timely implementation of subprojects. 88. During the first project, government policy was one of a highly-centralized approach to the financing and implementation of infrastructure projects. In keeping with this approach, the executing agency concentrated on design and construction of subprojects, without any concern about the participation of the 21 beneficiary. The PIU even neglected to use the funds available for a training program for sector personnel, considering its principal responsibility to be the construction of facilities. Similarly, the borrowers failed to involve the beneficiaries in discussions on project financing and cost recovery. As a consequence, beneficiaries did not have any feeling of "ownership" of the facilities that were built, and little incentive to raise revenues to pay for them. More generally, the various beneficiary organizations had little commitment to improving their institutional performance. 89. One of the objectives of the second project was to support the government's new policy: to decentralize the sector and to give responsibility for the provision of water and sanitation services (including their financing) to the municipalities. In implementing this policy, the previous federal water organizations (Juntas) were to become independent agencies. In 1983, Article 115 of the Mexican Constitution was appropriately amended to provide the legal basis for this chenge. 90. Unfortunately, implementation of this new policy was slowed, due firstly to the resistance of municipalities, who preferred to maintain the concessionary funding that government had previously provided, and secondly, to the lack of enthusiasm and skills of central agencies in creating the needed programs for this decentralization. Instead, they continued to design and implement the subprojects of the second project without concern for the participation of the beneficiary. Technical assistance to local institutions, for which financing would have been available under the loan, was not provided and in fact remained a matter of disagreement between the Bank and the executing agency, which did not consider such assistance of a high priority. The philosophy of building and handing over completed projects remai.d. 91. One casualty of the decentralization policy was a major institution- building component of the third project. This was dedicated to assisting CEAPAS, but it had to be canceled when the function of that Board was changed by state authorities from that of an operating organization to that of a technical assistance agency. 92. While institutional development generally remained disappointing at all levels, there was one exceptions the management of FIFAPA by BANOBRAS, which was adequate in most respects (see the next section). Apart from that, the first major step forward occurred in December 1989, when the Government established the National Water Commission (CNA) as an autonomous agency within the jurisdiction of SARE, to be responsible for all technical aspects of water resource development, including water supply and sanitation. As a consequence, there is now a single agency chargedwith the sector's development, including training and technical assistance. BANOBRAS remains responsible for funding sector investments. CHA's creation came too late to help improve the Medium Cities Projects, but appears to have had a highly beneficial impact on the performance of the Sector Loan 3271-ME. E. Financial Aspects 93. The Bank loan was intended to prc-ride support for the initial steps needed to change from the traditional centralized grant funding approach towards more 22 local autonomy, credit financing and cost recovery. Because government grant funding continued, however, prospective borrowere had easy access to alternate "free" fundst this was a disincentive to the implementation, at loccl level, of the proposed changes. At the time of project implementation, government commitment to the proposed changes vas clearly not sufficient to overcome the natural reluctance of local administrations to abandon sources of "free" funds. 94. The Bank did not have sufficient access to the beneficiaries to determine whether those who did eventually participate in the subprojects financed through the loan actually complied with its financial covenants. BANOBRAS limited its supervision to ensuring that loans were repaid. Information supplied to BANOBRAS by the beneficiaries was not sufficient to determine whether improvements in financial management and cost recovery were being accomplished. 95. The Bank strongly supported the government's declared policy for the sector: to adopt loan financing and to reduce dependence on grants. Unfortunately, government grant funding continued, accounting for some 85% of sector investment funds. This was obviously a strong disincentive, at local level, to the proposed change to loan financing and to imposing the tariff increases needed to recover operating and capital costs. 96. Information provided to central organizations by beneficiaries was insufficient to determine how well beneficiaries managed their affairs, and Bank missions were limited largely to the central level. However, spot checking by Bank staff revealed that financial covenants by beneficiaries were not being met. 97. As part of the Third Project, the Bank continued to give strong support to the government's official policy, of increasing loan financing and reducing dependence on grants. Unfortunately, as noted previously, this policy was not firmly enforced by the Government, and grant funding continues to finance the bulk of Mexico's water supply and sewerage investments. 98. As in the Second Project, BANOBRAS again limited its supervision to ensuring that loans were repaid. However, information which is available indicates that water tariffs, which were very low, have generally been increased dramatically. As a result, some of the stronger beneficiaries, such as the regional water company of Culiacan in Sinaloa State and the State Water Company of Queretaro, are now at least managing to pay for operation and maintenance. 23 F. Economi2 and Environmental Assessments 99. No attempt has been made in the PCR to determine the Internal Financial Rates of Return, or Economic Rates of Return. The absence of adequate investment and operating data would make such an exercise meaningless. 100. The SAR lists the environmental benefits customarily claimed for water supply and sanitation projects. To the extent that accomplishing the physical targets increased the number of people supplied with potable water, or benefiting from the removal of wastewater, these benefits would undoubtedly have arisen. However, there are reasons for questioning whether these benefits are real. Firstly, no information is available to evaluate the impact of the increased wastewater discharges on receiving waters; this is of particular concern since some of the subprojects did not include adequate wastewater treatment facilities (or had none at all). A second major concern is about the eustainability of the systems: clearly, long-term benefits from a non-functioning system are small or non-existent. 101. Evidence about health benefits is equally mixed. As noted above, the health statistics are not entirely reliable, and in any event specific information on any reduction in the incidence of gastro-intestinal diseases as a result of project activities is not available. In principle it is reasonable to assume that the health of people directly served by improved facilities would benefit. At the same time, doubts concerning the long-term sustainability of the systems, coupled with lack of balance between increased water supply and facilities for removal and safe disposal of wastewater, could lead one to discount claimed health benefits quite severely. G. Project Impact on Urban Poverty and Women 102. By increasing the number of people with access to water and sewerage, the projects undoubtedly increased the number of poor people receiving services. For ten of the thirteen cities for which comparative data is available, the appraisal reports list existing service coverage for water at 65% or more of the population. The SARs on the second and third projects state that the urban poor expected to be benefitting from the projects will rise from a low of 39% to a high of 90%. The first project does not provide similar data, and comparative data at project completion for any of the projects are not available due to the previously mentioned inadequate information system. 103. Nevertheless, it is reasonable to assume that a significant proportion, probably a very large proportion 6f the project benefits have accrued to the urban poor. There are two reasons for this conclusions first, the second and third projects specifically targeted these groups to the extent of identifying target areas and ensuring that expansion to these areas was included in the projects; and, secondly, the fact that existing service coverage was relatively high (65% or higher) signifies that middle- and upper-income groups were already reasonably well served, so that the major portion of newly-served beneficiaries would belong to the urban poor. 104. The projects' impact on women is not discussed in any of the SARs, nor in the PCR. To the extent that women of the urban poverty groups tend to be the 24 families' providers of water supply and basic health care (they carry water from whatever source to the household, and take care of infants and the sick in the family), they benefit from improvement in water and sanitation services and the resulting Improvements in health of their families. In the absence of any attempt to involve the beneficiary communities in a dialogue about project development and implementation, there was, of course, no attempt to involve women specifically as participants in the process. 105. The Bank recognized this lack of women's participation in project development and implementation. Rather than to attempt adding further requirements to institutions already challenged beyond their capacity by ongoing project requ!.rements, the Bank developed a separate activity: the "Women, Water and Development Project". The project concept is to combine improvements in water supply and sanitation with the generation of income-producing opportunities for women. This approach is particularly attractive because the provision of water and sanitation services clearly eliminates a lot of the time women spend on carrying water and improves their living environment, and thus frees time for more productive endeavors. 106. Water, Women and Development (Loan 3101-ME) was identified on March 7, 1988, and approved by the Board on June 22, 1989. Before the loan became effective, the new Government in Mexico had instituted a Program called SOLIDARIDAD, which had as its objective the improvement of living conditions and economic opportunities of the low-income groups. This program, which among its other activities sponsors ones similar to those to be provided by Loan 3101-ME, is funded by Government grant financing. The Bank loan thus became unattractive to the intended beneficiaries. After attempts to modify the loan, the Bank and Mexico eventually concluded that it was more efficient to terminate it and possibly identify a new operation. As a consequence, after postponing the date of effectiveness eleven times, the loan was allowed to lapse with the passing of the last date of effectiveness on April 26, 1991. 107. The effort invested in this unsuccessful operation has not been in vain. Today, SOLIDARIDAD is financing many activities incorporating the concepts developed as part of the preparation of Loan 3101-ME, and Mexican sector officials credit the Bank's participation with the development of an effective approach to improve water supply and economic activities of low income groups. III. CONCLUSIONS AND LESSONS LEARNED 108. The projects more or less accomplished their physical targets, although not necessarily as originally proposed, nor at the locations indicated in the appraisal reports. A total of 227,000 water supply and 130,000 sewerage connections, and the necessary supporting infrastructure, were constructed as a result of the loans, and an estimated total of 1,870,000 people are now being served by these connections (there is some double-counting, as some people are included in both water and sewerage service). This compares with original estimates of about 2,500,000. With loan disbursements amounting to US$215 million rather than US$265 million nominal loan amount, the proportion of population sex ted (75Z) is roughly proportional to the percentage of the loans 25 disbursed (811). Although this improvement in service coverage is a significant accomplishment, it only kept up with population increases, without meeting the target of reducing the percentage of the population without access to servise. 109. Because the loans amount to only 8% of total sector investments, the influence of the loans on overall sector policies has not been as substantial as had been hoped. Because of the large amounts of grant funds available, few municipalities were really interested in changing their approach, and cost recovery in the sector remained an elusive goal. Nevertheless, the steady and persistent effort by the Bank to promote sector financial viability eventually did result in a change in Government sector policies, accompanied by the needed organizational changes. There is reasonable hope that the most recent loan, Sector Loan 3271-HE, will consolidate these changes and bring about the reforms and improvements unsuccessfully sought during the three Medium Cities Projects. 110. The institutional and financial objectives of the first two loans were not met, and only marginally so during the third loan. This is largely the result of the initial lack of strong commitment to sector reform by the Government, which only during the last project began to formulate and then implement policies which clarified conditions of Government participation in water supply and sewerage investments, instituted a unified financing mechanism, FIFAPA, with the same norms and conditions for all borrowers, and established the organizational framework and sector strategies to implement the new policies. The third project, in particular, reveals some hopeful signs for future progress, such as the effective management of services by specific entities and the delegation of tariff setting authority by some state legislatures to the operating authority. Water tariffs have been significantly increased by the beneficiary organizations as a result of the projects. Some states appear to have adopted measures further rationalizing government financial support by limiting central government grant support to weak institutions and excluding stronger ones; one such example was revealed during a mission visit to Sinaloa state, where the Culiacan Water Authority is not receiving any of the grant funds made available by the central government, such funds going instead to weaker organizations. 111. Beyond a doubt the most significant results to which the Bank's participation in the sector has contributed are the new financial policies and the organizational structure which the Government has established for the sector. The Bank clearly has been very lenient in the enforcement of its contractual remedies for inadequate loan implementation. But because of its continuous dialogue and support for the sector in difficult circumstances, the Bank's suggestions for rationalizing financial support mechanisms and institutional development were finally accepted and implemented. While it would be easy to claim too much credit, there is no doubt that even the unsatisfactory first two medium cities projects have contributed to the long-term substantial improvement of the sector as a whole. The Audit's assessment of the overall performance of these projects confirms the ratings based on the findings of the PCRs, namely that the performace of Loans 1186-HE and 1913-HE were unsatisfactory and that of Loan 2281-HE was satisfactory. The sustainability of all three projects are judged unceratain and their institutional development achievement as partial. 112. The fundamental lesson of these loans is that optimistic assumptions about Government participation and support, no matter how rational and sensible the 26 anticipated actions are no substitute for precise and binding agreements. Without explicit agreement, sector institutions ultimately cannot implement even the best intentioned improvements. The Bank clearly did not realize the magnitude of the political difficulties involved in bringing about the fundamental changes it was seeking. 113. Another important lesson, actually part of the first, is that implementing agencies, and particularly local beneficiary authorities, must share the objectives being promoted and participate in determining the means to implement them. With the Government continuing its grant financing, there was no real incentive for beneficiaries to participate in projects which required them to assume financial responsibility for investments. 114. With the loans representing such a small part (8Z) of the sector's activities, the Medium Cities Projects should have been treated more as demonstration projects to inform both federal and local authorities of the advantages of implementing the proposed policies. Such an approach would have given priority to training and technical assistance, which was neglected due to the executing agency's priority on design and construction. 115. An initial effort at using an inexperienced executing agency (in terms of invegtment and project analysis and institutional development) to develop and supervise a large number of subprojects required the Bank to assume much of the supervision of subprojects, to train both the executing agency and the beneficiary local organizations. This was not done to a sufficient degree, not even when it was determined during supervision mission spot checks that both borrower and executing agency had not complied with their supervision obligations. The Bank underestimated the effort required to create the counterpart implementing capacity for these projects, even though the appraisals recognized the risk involved. 116. Significant progress has been made in Improving the sector's performance, as demonstrated by the Government's actions establishing the national water and sanitation program and the institutional reorganization of the sector to implement it. While these changes have not improved the implementation of the three medium size cities projects, they are clearly the result of the experience and lessons learned from the implementation of these projects and offer hope for further and substantial improvements. Nevertheless, an examination of present practices and mission conversations with those responsible for implementation of the new policies at central and state level indicates that it would be premature to conclude that all the obstacles have been overcome. 117. There is no doubt that the borrower and the executing agencies are competent and are the best-qualified institutions available to successfully guide the sector. Their staff is well-qualified, with technical and professional qualifications second to none. The caution in predicting future success arises from the fact that this staff has worked previously in precisely the centralized decision-making/financing environment it is now expected to abandon in favor of one stressing decentralization and financial self-sufficiency. It would be unfair to expect this transition to occur without difficulties. An example demonstrates potential problemst 27 The first sector loan which succeeded the third medium cities loan has accomplished rates of disbursements practically unheard of, disbursing 1002 of the loan amount in less than the predicted time. Nevertheless, CNA has been unable to provide evidence that the quality of economic and least-cost project analysis, which has always been one of the Bank's major and consistent concerns, has improved, or to demonstrate that appropriate training activities have been implemented to improve existing capacity in project analysis. The very real danger thus exists that speed in project implementation may result in faulty investments, due to deficiencies in adequate preparation and investment analysis. 118. Historically, Mexico's water sector and associated construction industry has a deserved reputation for quality and innovative construction. Decision- msking has traditionally been highly centralized. Investment efficiency, from the local perspective, has been unimportant because concessionary funding was customarily provided by central government. Overcoming these past practices and perceptions will require substantially greater efforts than are evident so far as a result of the first sector loan. The Bank*s dialogue with Mexico should concentrate on changing these past practices, and lending should be conditioned on real evidence of progress being made. 28 CURRENT ORGANIZATION OF TRE WATER SUPPLY AND SANITATION SECTOR IN MEXICO Source: Sector Loan SAR mure ORtSWA110N SCom NetaeUa wtver CelmaSem (enAl stat. VO?.? 00p'y and Setatom DeenteeaUz State Agen500 (UM 0meIeaaty eogasutas Aestan Opståam. De %mur ad e Opeestione Capetteettaa Loeptosem e0te syo~ ft~Iatigas A 00B C poup5.00 eol 1. Figuso aind are estimated, as ofDesir31 l tn data ar faam the Casu coauta~d 'a march of yar 1990 ma 1991 fig~Mos am fgem Ch oa ~ e Inm~no DY . Dum~ with the 1990 Coem, th d« o of Oacc * hduppy nludeb oly j-hou~ or ann~tte <yard oneOtion osti, and euCud~ tuggy (provIoney in~auded) sanimtio lades conmftion to a a~ system or to a epti tak (b~t not other ~maus ch as other fosto of oA-eito or off-iLte d18ea). - 29 - Pace 1 of 4 Comments from .the Borrower (Translation from Spanish original) United Mexican States Secretariat of Finance and Public Credit Public Credit Directorate International Finance Agencies Department Ref.s 393.111.1.182 Subjects Comments on the Performance Audit Report for the Medium Cities Water Supply projects (Loans 1186-MN, 1913-ME and 2281-ME) Mexico City, June 1, 1993 Mr. Yves Albouy Chief, Infrastructure and Energy Division Operations Evaluation Department The World Bank Dear Mr. Albouys In reply to your letter of April 9 requesting our comments on the Performance Audit Report prepared by OED on the above-referenced projects, we feel it is important for audit reports, as well as project completion reports, to be done in a more timely fashioo, since when so much time elapses information often gets lost and the officials involved change. Below we summarize our comments to the report, in particular the conclusions we feel must be taken into account in the preparation and execution of other projects The projects designed at the outset did not enjoy sufficient acceptance by the municipalities because federal subsidies were 1. Borrowers comments were taken into account and incorporated into Final Report. - 30 - anneg 2 Pae 2 of 4 also available. In order to expand the responsibilities of the municipalities in potable water supply, constitutional amendments were passed and changed the attitude of the local authorities. At present, owing to practices introduced in conjunction with the World Bank under the Water Sector Project (Loan 3271-ME), financing is done through a mix of funds, including federal, local (both state and municipal) resources, as well as credit from the National Bank of Public Works and services (BANOBRAS). In this way, the Government has reduced subsidies to the municipalities and is seeking total cost recovery and financial viability. Consequently, municipal use of the funds has meant that the proceeds of the new World Bank loan have been committed in their entirety and will even be disbursed before the scheduled date. Moreover, as some of the borrowed funds will have to be repaid plus interest, the municipalities agreed to ensure efficient operating entities. As a result, many of these companies have been self-financing and it is expected that in the future all such companies will be in the same position. With regard to the municipalities or final beneficiaries, once the creditworthiness of the inhabitants where the works are to be [illegible: ?done] is assessed, the needs for transfers from the federal government can be determined, as the same mix of resources is not used for all municipalities, in that lower-income areas may receive relatively larger benefits. Nevertheless, with a view to credit recovery, the municipalities are asked to meet minimal requirements in order to be eligible for credit. Once the loans are repaid, monitoring of cost recovery and of works maintenance becomes more important within the municipalities, if return on investment is to be maximized. With regard to the lower-income population that is not eligible for credit, the SOLIDARIDAD program seeks to provide basic water infrastructure. This has made it possible to increase federal transfers for water and sewerage works, without adversely affecting loan-financed projects. Project performance would be improved if realistic feasibility studies were used as a basis; the objectives of the earlier projects were not achieved in full because some of the studies were completed after the identification report. In addition, the beneficiaries are involved in project identification - formerly done by SEDUE, BANOBRAS and the World Bank, without consulting the pertinent municipalities. In other words, decision-making was - 31 - Annux2 Page 3 of.4 done at the central level, which meant that some localities that had been assessed subsequently refused to participate. The Mexican Government further agrees with the Bank's comment regarding the importance of having a greater number of water treatment plants, since broader coverage of potable water and sewerage services means an increase in wastewater, making it a priority to have an adequate treatment policy. Efforts by the Mexican authorities to give greater autonomy to the municipalities and provide better water service has meant constant improvements in the implementation of the various projects. With regard to the contents of the report, we agree in general with the data and information, but have certain comments that we would like to be taken into consideration: In the translation of the Executive Summary, it would be more correct to replace all references to "medium-sized cities" by "medium cities." Likewise, the word "sostenibilidad" in Spanish should be replaced by "sustentabilidad." Para. 2 of the reface. The wording at the end of the paragraph that reads "recurrir a donaciones gubernamentales, como es tradici6n on los municipios mexicanos" ("to rely on the Mexican municipality's traditional use of government grants") should be changed to "recurrir a transferencias gubernamentales, como o tradici6n en los municipios mexicanos* (*to rely on the Mexican municipality's traditional use of government transfers"). In para. 1 of the Evalation Summary, the data are not current. The end of the paragraph states that after the 1970s the pace of improvements in water services slowed. As of 1990, under the National Potable Water Plan, both coverage and quality of services has again risen. Para, 2. Instead of "Government grant financing for infrastructure," the text should read *Government transfers to finance...." Para. 8 reads "trabajando con un organismo do ejecuci6n sin experiencia y un ente prestatario y con escaso contacto..." ("working with an executing agency that was inexperienced and a borrower, and with little contact...") should read "trabajando con un organismo de ejecuci6n y un ente prestatario sin experiencia y con escaso contacto..." ("working with an inexperienced executing agency and borrower, and with little contact..."). Para. 9. The qualifier "of the oil boom raised municipal expectations [for continued grant financing) to unrealistic (but hard to abandon) levels" should be eliminated. - 32 - annex 2 Pag 4 ol PAra. 13. "As shown in the portion recording the lending process, the time between estimated completion and actual completion of Project 1 is four years, of Project 2 is two years, and of Project 3 is one year." Closing of the third project was delayed by only six months, and not one year as indicated in the report. Paras. 16 and 17. The end of para. 16 reads: wFor reasons beyond the control of the Bank (explained in paras. 103, 104 of this report), the loan, which was approved by the Board on June 22, 1989, never became effective.' It is precisely because of the SOLIDARIDAD Women*s Program, mentioned' in para. 17, which had the same objectives and enjoyed grant funding, that the municipalities were not interested in the IBRD loan, including those that committed during its preparation. Moreover, the paragraphs in question are 106 and 107, and not 103 and 104. In 2aras. 93 and 95 of the Proiect Summary, the last sentences in both paragraphs are identical: "Government commitment to the proposed reforms was clearly not sufficient to overcome the natural reluctance of local administration[s] to abandon sources of *free, funds." Halfway through paras. 93 and 95 the same statement is repeated: "prospective borrowers had easy access to alternate *free' funds." The end of 2ara. 94 and the beginning of 2aras. 96 and 99 are identical: "BANOBRAS limited its supervision to ensuring that loans were repaid. Information supplied to BANOBRAS by the beneficiaries was not sufficient to determine whether improvements in financial management and cost recovery were being accomplished." I hope that these comments will be taken into account in the final version of the report. Yours truly, /s/ Lorena Gabilondo Alpisar For the Director of International Finance Agencies Deputy Director of Social Projects and Human Resources