Honduras - Municipal Development Pilot Project

Document of The World Bank FOR OFF[CIAL USE ONLY Report No. 14042 PERFORMANCE AUDIT REPORT HONDURAS MUNICIPAL DEVELOPMENT PILOT PROJECT (LOAN- 2583-HO) MARCH 10, 1995 Operations Evaluation Department This document has a restricted distribution and may be used by recipients only in the performance of their official duties. Its contents may not otherwise be disclosed without World Bank authorization. Currency Equivalents (annual averages) Currency Unit = Lempiras (L) US$1.00 = L. 6.12 Abbreviations and Acronyms BANMA Banco Municipal Aut6nomo (Autonomous Municipal Bank) CELCADEL Regional Training System for Local Municipal Development in Latin America CONSUPLANE Consejo Superior de Planificaci6n Econ6mica (Superior Council for Economic Planning) ENE Empresa Nacional de Energfa Elctrica (National Electricity Company) ESW Economic Sector Work IDB Inter-American Development Bank MCI Municipal Credit Institution MDPP Municipal Development Pilot Project SHCP Secretarfa de Hacienda y Cr6dito Pdblico (Secretariat of Finance and Public Credit) SOI Secretarfa del Interior (Secretariat of Interior) UNDP United Nations Development Programme USAID United States Agency for International Development Fiscal Year Government: January 1 - December 31 FOR OFFICIAL USE ONLY THE WORLD BANK Washington, D.C. 20433 U.S.A. Office of Director-General March 10, 1994 Operations Evaluation MEMORANDUM TO THE EXECUTIVE DIRECTORS AND THE PRESIDENT SUBJECT: Performance Audit Report on Honduras Municipal Development Pilot Project (Loan 2583-HO) Attached is the Performance Audit Report on Honduras - Municipal Development Pilot Project (Loan 2583-HO) prepared by the Operations Evaluation Department. At the time of the loan, projections for the year 2000 indicated that 55 percent of Honduran citizens would be living in the cities, with the bulk expected to occupy new areas in and around the Central District. In spite of its already apparent weaknesses, the Bank and the Government of Honduras saw the Autonomous Municipal Bank (BANMA) as the best institution to help the urban sector face growing demands for infrastructure. The project attempted to strengthen both BANMA and the municipalities through technical assistance in improved administration and financial management. Project incentives focussed on BANMA, and not the largest municipalities, where established patterns for handling payments and collections needed to change radically. Local mayors did not share the pilot project's priorities. Technical assistance was peripheral and not central-as it ought to be in a project attempting to cause administrative changes with sweeping social impacts. BANMA did not prove to be a sustainable institution: the project achieved limited results for reasons partially related to project design but also because necessary legal and incentive structures and country preconditions for project success were simply not there. During the course of the project, BANMA's financial situation steadily worsened-it is no longer delivering banking services to any municipality. The Executive Branch has requested that the institution be closed but a vote for closure has not yet been scheduled by the Congress. The Project Completion Report provides a brief but full account of the project experience; it is informative on the achievements and on the compliance with covenants. The Performance Audit Report elaborates on the achievement of institutional objectives, on project design and on technical assistance. It is unlikely that there will be any sustainable benefits for either BANMA or the municipalities as a result of the loan. The Municipal Development Pilot Project's performance as a pilot intervention was poor. The Audit rates the overall project outcome as unsatisfactory, its sustainability as unlikely, and its institutional development as negligible. Attachment This docunent has a restricted distribution and may be used by recipients only in the performance of their official duties. Its contents may not otherwise be disclosed without World Bank authorization.  FOR OFFICIAL USE ONLY Contents Preface .................................................................. .3 Basic Data Sheet .......................................................... 5 Evaluation Summary ....................................................... 7 1. Objectives and Methodology .......................................... 11 The Audit ....................................................... 11 2. Background ..................................................... 13 The Project Context and Urban Development ............................ 13 Project Expectations concerning BANMA's Relationships with the Municipalities .............................................. 14 3. The Project ...................................................... 15 Issues in Project Preparation .......................................... 15 Project Objectives and Appraisal ....................................... 15 Financial Arrangements ............................................. 16 4. Project Implementation and Results .................................... 17 Physical Implementation Experience .................................... 17 Analysis of Selected Components ....................................... 18 5. Conclusions and Lessons Learned ...................................... 23 Project Rating .................................................... 23 Identifying Preconditions of Project Success .............................. 23 Policy Reform a Precondition ......................................... 23 Achieving Preconditions ............................................. 23 Limited Municipal Financial Capacity .................................... 24 Intermediation Value ...... ... 24 Lack of Coordination with other International Organizations .................. 24 Importance of T/A in Municipal Development Projects ...................... 24 Poor Assessment of Training Needs ..................................... 25 Supporting Sectors between Projects .................................... 25 Preconditions to Decentralization ...................................... 26 Annex 1 - Comments from the Borrower ....................................... 27 This report was prepared by Alcira Kreimer (Task Manager) and Ronald Parker (Consultant) who audited the project in July-August 1994. E. Tweddle provided administrative support. 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.  3 Preface 1. This is a Performance Audit Report (PAR) on the Municipal Development Pilot Project, involving a World Bank Loan amounting to US$6.9 million equivalent to the Government of Honduras. The loan was approved on June 13, 1985. The pilot project loan covered in this report closed on December 31, 1990. US$2.2 million was disbursed under the project, with the last disbursement made on January 1, 1992. 2. The PAR is based on the Project Completion Report (PCR) prepared by the Latin America and Caribbean Regional Office and issued on December 6, 1993 (Report 12582), the loan documents, and study of the project files. There was no Staff Appraisal Report (SAR). An OED Audit mission visited Honduras during August, 1994. The excellent cooperation and valuable assistance provided by the Unidad de An6lisis de Polfticas Econ6micas (UDAPE) and the Autonomous Municipal Bank (BANMA) in the preparation of this report is gratefully acknowledged. 3. The PCR provides a brief but full account of the project experience; it is informative on the achievements and on the compliance with covenants. The PAR elaborates on the achievement of institutional objectives and on project design. 4. Following standard OED procedures, the draft PAR was sent to the relevant Government officials and agencies concerned for comments. Comments received from Banco Municipal Autonomo are attached to the report as Annex 1.  5 Basic Data Sheet Municipal Development Pilot Project (LOAN 2583-HO) Loan Position (Amount in US$ million) As of June 30, 1993 Appraisal Actual/ Actual % of Estimate Reestimate Appraisal Estimate Total Project Cost 8.24 2.65 32 Credit Amount 6.9 2.22 32 Canceled - 4.68 68 Cumulative Estimated and Actual Disbursements Appraisal Estimate (US$M) 6.9 Actual (US$M) 2.22 Actual as % of Estimate 32.1 Date of Final Disbursement 09/26/89 Project Dates Original Actual Identification 05/84 05/84 Appraisal 10/84 10/84 Negotiations 04/85 04/85 Board Approval 06/13/85 06/13/85 Loan Signature 07/31/85 07/31/85 Loan Effectiveness 08/85 11/08/85 Loan Closing 09/30/88 12/31/90 Project Completion 03/88 12/31/90 6 Staff Inputs (staff weeks) FY85 FY86 FY7 FY88 FY89 FY90 FY91 FY92 FY93 Total Preappraisal 13.9 - - - - 13.9 Appraisal 35.1 - - - 35.1 Negotiations 1.4 - - 1.4 Supervision - 10.4 9.7 12.4 6.9 1.9 5.7 .7 10.0 57.7 Other 9.8 0.4 - - - - - - 10.2 Total 60.20 10.8 9.7 12.4 6.9 1.9 5.7 .7 10.0 118.3 Mission Data Stage of Project Montlh No. of No. of Performance Cycle year persons weeks Rating Through Appraisal 09/83 1 1 d 11/83 2 1.2 d 05194 2 2 d 08/84 3 5 d 10/84 4 10 d 02/85 2 5 d 06/85 1 1 d Supervlslon 09/85 1 6 d 03/86 2 2.2 2 12/86 2 2 2 05/87 1 1 d 09/87 1 2 - 11/88 2 2 2 05/90 1 .4 d 10/90 1 .4 3 02/91 1 .4 4 01/93 2 2 - Other Project Data Borrower. Government of Honduras Executing Agency: Banco Municipal Autonomo Follow-up Project: Loan 2421-Ho, Water Supply and Drainage, for US$19.6 million. Approved on May 22, 1984, closed on June 30, 1993. 7 Evaluation Summary Audit Focus 1. This Audit evaluates the Municipal Development Pilot Project (MDPP) implemented in 1985-90. Two factors guided the choice of areas to study: firstly those project effects that were central to the original objectives of this project, including the progress which the executing agency and the municipalities have made in debt repayment, infrastructure investments and financial administration, and secondly the achievement of project objectives due to policy reforms and other events which occurred after the loan closed. In all three cities that executed subprojects under the loan, semi-structured interviews, secondary data review, direct observation and walking/riding transects were conducted during the audit process. Project Overview 2. In response to its debt crisis and in order to make municipal operations more transparent, Honduras created a municipal bank which was intended to replace grant funding with loans, thereby decreasing demands on the national treasury. The purpose of the MDPP was to assist the Government to strengthen the Autonomous Municipal Bank (BANMA). BANMA, in turn, was supposed to strengthen the municipalities and enable them to provide improved physical infrastructure through improved revenue administration, budgetary management and financial and investment planning. An additional project objective was to strengthen CONSUPLANE, the Superior National Council for Economic Planning (which has evolved into SECPLAN). At the time of project identification, "institution building" was a main objective. Before the establishment of BANMA, some donors such as USAID had worked directly with the municipalities. However, according to donor agency officials, early municipal strengthening endeavors produced little of lasting value because the administrative capacity of the municipalities was minimal. 3. The Borrower was the Republic of Honduras. Loan 2583-HO was for the amount of US$6.9 million. Implementation was to be the responsibility of the Autonomous Municipal Bank (BANMA) which was to provide loans for infrastructure to municipalities that developed credible revenue strategies and financial plans. The project was approved by the Board on June 13, 1985 and conditions of effectiveness were substantially met by November 8, 1985. The loan was closed two years and three months behind schedule. The outstanding balance of US$4.69 million was canceled on May 29, 1991. 4. During the course of the project, BANMA's financial situation (which the loan was designed to improve) steadily worsened. Audit mission requests for the report "Financial Situation of the Autonomous Municipal Bank" were denied. During the Audit visit to Honduras BANMA was not delivering banking services to any municipality. The Economic Council (Gabinete Economico) has recommended that the institution be closed. However, in order to close a semi-autonomous institution, like BANMA, the Constitution 8 requires a request from the Executive Branch and a two-thirds majority vote in Congress. At the time of the Audit mission Congress had not yet scheduled a vote on the issue. Sustainability 5. BANMA did not prove to be a sustainable institution, but in the period following the loan the municipalities have made tremendous progress in lowering their operating costs for services and they are approaching full cost recovery for infrastructure investments in many neighborhoods (para.4.10). The municipalities are much more autonomous now then they were before the project, and the mayors are making numerous moves to stabilize municipal finance (para. 4.13). For example, the mayor of Tegucigalpa has reduced the number of municipal employees from 4865 to 1676 in a period of barely a year, raised tax compliance levels by 6,000 families, and made significant reductions in the municipal debt. Among the issues which limited municipal autonomy during the period of the loan were: (i) the requirement that municipal budgets be approved by the central government; (ii) the requirement that all expenditures over Lempiras 100 (about US$50 at the time) have central government approval; (iii) the requirement that all capital investments be approved by central government; and (iv) the lack of a formal system for resource transfers from central government to the municipalities against a backdrop of frequent transfers. A new municipal law (1990) is expected to increase local autonomy and to change the unsustainable municipal pattern of borrowing. The new law fundamentally changes the legal relationship between the national government and cities by removing the requirements listed above and it allocates 5% of the national tax revenues directly to the cities while cutting off their access to public banks and subsidized credit, making them more financially independent. Findings and Lessons 6. Project Ratings. It is unlikely that there will be sustainable benefits for either BANMA or the municipalities as a result of the project. The Audit rates the overall project outcome as unsatisfactory, its sustainability as unlikely, and its institutional development as negligible. These ratings are identical to those based on the PCR. 7. Identifying and Obtaining the Preconditions of Project Success. The lessons learned in this project indicate that more care will be needed identifying limiting factors and providing in-country follow-up. Projects that require large numbers of people and agencies to come to a consensus are slow yet the loan was processed very quickly, almost as if the changes which had to take place could be accomplished by a stroke of the pen. Project results might have been better had conditions of negotiation (or even appraisal) been established for the municipalities in areas such as purchasing shares and opening accounts in BANMA or making a first loan payment. 8. Policy Reform a Precondition. While it may be extreme to say, as the PCR does, that the institutions that worked in the project were the problem and not the solution, the GOH's experience with BANMA was an important learning experience and perhaps even a catalyst for broader political reform. The GOH attempted to promote decentralization by creating yet another centralized institution. The BANMA project achieved limited results for reasons partially related to project design (note PCR's criticisms in paras. 4.02 9 and 4.03) but also because the necessary incentive structures and legal and policy preconditions for project success were simply not there. Major policy and (almost) cultural changes were expected, without fully recognizing that the prevailing tax/fee collection and debt repayment status served strongly established interests in important ways. 9. Limited Municipal Capacity. A municipal financial intermediary might have been more successful in a country where the municipal civil service was more established and, perhaps, funded by a sales tax. As long as mayors had the legal option of not collecting taxes (as one mayor said) "those who wanted to, paid, and those who didn't did not have to." The largest municipalities, where established patterns and practices needed to change radically, did not share the pilot project's priorities. 10. Intermediation Value. The underlying problem which the MDPP attempted to correct was the financial viability of BANMA. The MDPP assumed that by providing computers, data bases, technical assistance and training the delinquent municipalities would proceed expeditiously to collect their accounts receivable and pay off BANMA. The assumption that the mayors were not collecting their accounts because essential information was inaccessible on paper and/or stored in file cabinets seems flawed and it was not substantiated by the Audit. For the municipalities there was simply no value- added from having BANMA as a financial intermediary. 11. Weak Technical Assistance. Technical assistance was peripheral and not central-as it ought to be in a project attempting to cause sweeping changes. BANMA should have taken a more flexible approach, developing service packages based on the complex of needs and problems which the various municipalities were experiencing. The training should have taken place within the municipalities themselves, and it should have supported the equipment purchased under the loan. 12. Preconditions to Decentralization. It seems reasonable to assume (especially given the unsatisfactory results of this project) that there are certain structural preconditions which must be present if decentralization is to be sustainable. Among those suggested by the MDPP and subsequent political reform in Honduras are: * Municipalities must have a degree of freedom in terms of day-to-day management. Where local officials cannot even spend $50 without direct authorization from the central authority, technical assistance to municipalities may not result in innovation well-tailored to local circumstances; * A municipality needs to have the autonomy to set its own budget and to reallocate within it; * Mayors must be legally obligated to collect taxes due and to set fees for local services; 10 * The decision to work only with municipalities of more than 80,000 population in the pilot project is also questionable, given that administrative practices in the smaller cities were likely to be easier to change, and that the Presidents Report (para. 37) noted that the larger cities would be likely to have problems covering their debts. 11 1. Objectives and Methodology The Audit 1.1 This Audit evaluates the Municipal Development Pilot Project (MDPP) imple- mented in 1985-1990. Two factors were taken into account in the choice of areas to study: firstly those project effects that were central to the original objectives of this project (the progress which the executing agency and the municipalities have made in debt repayment, infrastructure investments and financial administration), and secondly the achievement of project objectives due to policy reforms and other events which occurred after the loan closed. In all three cities that executed sub-projects under the loan, semi- structured interviews, secondary data review, direct observation and walking/riding transects were conducted during the audit process.  13 2. Background 2.1 The Project Context and Urban Development. Municipalities in Honduras traditionally build and maintain local roads and drainage and provide other services including water supply, solid waste management, sanitation, meat inspection/slaughtering, public markets and bus terminals. Historically, when investment capital has been required for infrastructure projects, municipalities depended on grants from the central government. 2.2 In response to its debt crisis and in order to make municipal operations more transparent, Honduras created a municipal bank, the Autonomous Municipal Bank, BANMA, which was intended to replace grant funding with loans, thereby decreasing demands on the national treasury. Since its creation, the cities have borrowed from BANMA, but the principal financial objective was never achieved. Mayors knew that should they fail to meet their obligations, the national government would have three alternatives: (i) to condone their debts; (ii) to cover their immediate debt service obligations in order to keep BANMA solvent, or (iii) to allow the arrears to go uncovered, thereby de-capitalizing the financial intermediary. Since the early 1980s, the latter has been the most frequently used modality. During the course of the Municipal Development Project, BANMA's financial situation (which the loan was designed to improve) steadily worsened. While the municipalities have always levied user fees on municipal services, until the mid-90s the amounts collected were usually significantly below their economic cost, with the cash flow used to cover ongoing municipal operating expenses and to subsidize other activities (principally school building construction/maintenance and the local operations of the Red Cross). 2.3 The high degree of dependency on central government decisions exhibited by the nation's cities is largely explained by the fact that mayors were usually appointed by the President (as governors still are), and there were few incentives to encourage responsible financial management at the local level. Among the issues which limited municipal autonomy during the period of the loan were: (i) the requirement that municipal budgets needed to be approved by the central government; (ii) the requirement that all expenditures over Lempiras 100 (about US$50 at the time) needed central government approval; (iii) the requirement that all capital investments be approved by central government; and (iv) the lack of a formal system for resource transfers from central government to the municipalities against a backdrop of frequent transfers. 2.4 A new municipal law (1990) is expected to increase local autonomy and to change the unsustainable pattern of municipal borrowing behavior. The new law fundamentally changes the legal relationship between the national government and cities by removing the requirements listed above and it allocates 5% of the national tax revenues directly to the cities while cutting off their access to public banks and subsidized credit, making them more financially independent. 2.5 There are indications that some elected mayors are using their new source of guaranteed income to over-borrow from private sector banks, leading some observers to 14 wonder if the cities will not mortgage their future income to such a degree that they will soon need infusions of public monies to cover capital expenditures once again. 2.6 The importance of finding the proper incentives for the sustainable financing of municipal infrastructure in Honduras is closely tied to the nation's rapidly changing demographics. In the period prior to project identification the urban population had tripled within a 20 year period, placing tremendous demands on services and creating marginal neighborhoods where collection of fees was difficult-a situation exacerbated by the manual maintenance of records and files and limited clerical staffs within the municipalities. Before there was a municipal bank, some donors such as USAID had worked directly with the municipalities. According to USAID's officials interviewed by the Audit mission, early municipal strengthening endeavors produced little of lasting value because the administrative capacity of the municipalities was minimal. 2.7 At the time of project identification it was felt by both the Borrower and Bank staff that institution building was a major priority: by creating a new bureaucracy which would support the municipalities with ongoing technical assistance and supervised credit, the central government would wean them of their appetite for grants. The Government established the Banco Municipal Autonomo (BANMA) to help the various municipalities finance infrastructure. The Inter-American Development Bank, IDB, provided municipal development funding during BANMA's early years. At the time of appraisal BANMA had financed 170 loans in about 60 municipalities. The bulk of its portfolio was in arrears, most of it owed by the largest city in the country, Tegucigalpa. 2.8 Project Expectations concerning BANMA's Relationship with the Municipalities. The municipalities were both BANMA's owners and its debtors. The law which created BANMA required each municipality to buy shares in the new bank for an amount equal to 5% of its annual operating budget. Only a few ever fulfilled this requirement. Additionally, municipal governments were prohibited from using any other public or private bank for any purpose whatsoever: all accounts with municipal monies had to be in BANMA, although only a small number of municipalities complied with this provision. In order to secure municipal deposits, and to make the use of its facilities more convenient, BANMA branches were opened in the larger cities. Through the provision of advanced information processing technology and training the Bank expected that the various municipalities would collect what was owed to them and in turn begin paying off their loans, thereby restoring BANMA's financial position and ensuring its survival as an institution. Small investments in infrastructure were included in the pilot project for municipalities that prepared financial plans detailing how their debts would be amortized. 15 3. The Project 3.1 Issues in Project Preparation. At the time of the loan the urban areas contained 32% of the national population. Projections for the year 2000 indicated that 55% of the nation's citizens would be living in the cities, with the bulk expected to occupy new areas in and around the Central District comprising Tegucigalpa and Comayaguela. In spite of its already apparent weaknesses, the Bank and the GOH saw BANMA as the best institution to help the urban sector face growing demands for infrastructure presented by rapid urban growth. However, because BANMA did not have a good track record in terms of collecting on its loans, and because the municipalities made only grudging and limited use of its services, the prospects for turning BANMA's financial situation around were not deemed auspicious. In fact, there were serious doubts as to whether the project should be done at all. However, Bank staff involved with the project felt that the pilot project could be beneficial to Bank-country relations and macroeconomic policy dialogue. 3.2 A second issue had to do with the scale of the project. The main question was whether the project should be a modest pilot project which would prepare the way for a broad-based urban investment project, or should be commensurate with the scale of the problem from the outset. The incentives created for BANMA in the pilot endeavor were that if it could achieve its financial targets (reducing arrears on its portfolio to not more than 10% while maintaining cash collection on current loans at 75% or better) the Bank would proceed to prepare a comprehensive investment program using BANMA as the institutional base for its implementation. There is an additional question which was not fully explored during project design because of the strong support expressed by the Borrower for BANMA: where was the political will supposed to come from (to collect debts and raise user fees) at the municipal level when no incentives had been built in? Or, inversely, why were the municipalities to believe that the flow of funds from central government was really going to stop at this particular moment given their desperate needs for infrastructure and the small incomes which existing fees provided for them? Project Objectives and Appraisal 3.3 The purpose of the MDPP was to assist the Government in strengthening the Autonomous Municipal Bank so that it could support the municipalities and enable them to provide physical infrastructure through improved revenue administration, budgetary management and financial and investment planning. An additional project objective was to strengthen CONSUPLANE, the Superior National Council for Economic Planning (which has evolved into SECPLAN). 3.4 Appraisal was originally scheduled for October 1984. Two post-appraisal missions in February and June of 1985 were required. The institutional development goals of the project were centered on bringing BANMA into financial balance. Additional goals for BANMA included improved organizational planning (updating an organizational manual and developing better personnel management policies); improving financial policies relative to the efficient e.:tting of interest rates; improved accounting and auditing practices and the establishment of a functional management information system. 16 Secondary institutional development goals were developed for the municipalities and CONSUPLANE. Municipal objectives included improved billing and collection procedures and cost recovery for major public services. Strengthening CONSUPLANE was to be done through technical assistance as well as through comprehensive planning. The preparation of a municipal sector financial strategy for infrastructure was part of the comprehensive planning approach. Financial Arrangements 3.5 Total project cost was originally estimated at US$8.24 million (including contingencies). The US$6.9 million loan had a term of 20 years and covered 84% of total project costs, financing the foreign cost of the pilot investments and the total cost of professional services. BANMA would lend US$ 1.1 million (14% of total project cost) from its own resources to cover local costs for pilot investments, and the municipalities would contribute US$240,000 (about 2% of project cost). US$6.1 million of the Bank loan would be relent at the same terms to BANMA. Of that amount, US$5.7 million would be onlent to selected municipalities for technical assistance and pilot investments, US$0.4 million would be used for professional services and to purchase data processing equipment. The borrower retained US$0.13 million which was allocated to CONSUPLANE to cover the cost of preparing an urban-municipal development plan. US$0.67 million were to be allocated as necessary for contingencies. About US$500,000 of the loan proceeds financed retroactive contracting costs. 3.6 BANMA's General Manager was in charge of project implementation. A steering committee composed of representatives of CONSUPLANE, SHCP, and BANMA provided policy guidance and liaison with other government agencies. BANMA's Directorate of Technical Services was responsible for assessing the feasibility of municipal investment and repayment proposals. In each municipality a Project Coordinator was appointed to monitor expenditures and consultants. The borrower was the Government of Honduras. 17 4. Project Implementation and Results Physical Implementation Experience 4.1 The Princial Objectives Were Not Achieved. The Bank loan did assist one municipality with infrastructure building, and provided equipment for solid waste collection and office furniture in all three cities that executed sub-projects before it closed after disbursing 32% of the appraisal estimate. However, its more important institutional development and financial objectives went unmet: in spite of their improved information processing capacities the municipalities did not choose to pay back the loans and BANMA's arrears grew accordingly. Audit mission requests for the report "Financial Situation of the Autonomous Municipal Bank" were denied. During the Audit visit to Honduras BANMA was not delivering banking services to any of its former clients. As the result of uncollected loans from the larger municipalities, BANMA no longer has any capital to lend, almost all of its staff have been dismissed, and it is now marginal to ongoing municipal development. Its only activity is to act as an institutional umbrella for a Project Implementation Unit working on an IDB project. 4.2 Political and Macroeconomic Environment. The PCR for this project notes that with the benefit of hindsight it is clear that BANMA and CONSUPLANE were incorrectly perceived as potentially being part of the solution when in reality their roles were part of the problem faced by the sector (infrastructure finance) which this project was attempting to correct. The PCR states that municipal projects worldwide since the 1970s that focussed on increasing property tax revenues have not resulted in sustainable increases of such revenues. And it observes that when such projects shifted their emphasis to financial intermediaries-municipal credit institutions such as BANMA-insufficient attention was paid to the political environment (particularly relationships between central government and the municipalities) and the potential of Honduran municipalities for increasing revenues and providing infrastructure, given the legal and institutional framework in which they operated. The PCR also notes that the macroeconomic situation was a major determinant of project failure-during the implementation period the country moved in and out of suspension on loan disbursements, and during these periods all physical works and contracting were effectively halted. The final result of multiple interruptions was that by the time uninterrupted disbursements resumed, all chance of completing the investment program had been lost. 4.3 Donor Coordination Problems. The MDPP reveals a long history of donor coordination problems. The Bank took over the project after USAID tired of funding its operation without observable results. While the Bank was trying to exert pressure on BANMA, IDB offered softer money and less conditionality (PCR para. 5.3). After the Bank's unsuccessful experience with BANMA, both USAID and the UNDP began to support the municipal sector. The lack of a common donor front made it hard for BANMA and national and municipal decision-makers to confront their problems. 18 4.4 Quality of the PCR. The PCR is of excellent quality and the Audit supports its conclusions. Rather than go over ground already covered in the PCR, the Audit focusses on the impacts of the project which have become apparent in the interval since the PCR was written, as well as the links between the current municipal development policy and the GOH's experience with BANMA. Some project objectives were achieved even as BANMA was ceasing operations. Analysis of Components 4.5 The amounts disbursed under this loan and the number of sub-projects were so small that their impacts have been reviewed individually rather than discussed under a more general heading. This has been done to provide background for the lessons learned which will be discussed subsequently. 4.6 BANMA as a Financial Intermediary. For the municipalities there was no value- added from having BANMA as a financial intermediary. Before the loan was closed by mutual accord between the Borrower and the Bank, the project financed the construction of a few very small infrastructure schemes which will be described below (paras. 4.10-4.12). Its technical assistance components directed at the municipalities (such as its training program in the maintenance of municipal infrastructure) are almost universally recognized as being ineffective. Other donors visited categorized the training as overly top-down, the Presidential Commission on Government Modernization observed that BANMA's technical assistance did not take the beneficiaries' needs into account, and the Honduran Association of Municipalities described its interventions as overly authoritarian. Nor did it make allowances for the vast differences in administrative and financial capacity which characterize Honduran municipalities. 4.7 BANMA-financed Training: A One-shot Approach. Training was designed by BANMA staff. The municipalities were invited to send staff to Tegucigalpa to receive courses that had been prepared without their input. And technical assistance was not designed to be ongoing-a one-shot approach can be problematic where the turnover in municipal employees following changes in mayors is almost total. 4.8 Technical Assistance Was Not Well Targeted. In terms of the technical assistance provided to BANMA by consultants, (according to the current CEO) the personnel manuals developed for BANMA were of adequate quality, but (according to one former employee) they were originally designed by the consultants for another agency and never modified sufficiently to take BANMA's situation into account. One consultant involved in the process noted that a principal recommendation involved a significant staff reduction but that after the report was presented staff numbers tripled. Additionally, the personnel qualification recommendations were not followed after the manual was completed, and key positions were filled by people without the required technical skills and background. 4.9 BANMA: The Beginning of the End? Municipal income did increase in some municipalities because of cadastral work funded by the project, but this did not translate into significant municipal payments on new and past-due commitments to BANMA, in spite of the debt repayment plans developed. BANMA's end is near but it is not a done deal. In order to close a semi-autonomous institution, the Constitution requires a request 19 from the Executive Branch and a two-thirds majority vote in Congress. The Economic Council (Gabinete Economico) has presented the opinion of the Executive to Congress saying that the Institution should be closed, but Congress has not yet scheduled a vote on the issue. The Presidential Commission for the Modernization of the State continues to argue for a continuing but greatly reduced role for BANMA as a receptor of donor funds and underwriter of municipal bond issues. 4.10 Municipal Financial Improvements Were Only Marginally a Result of Project Activities. The three cities that executed sub-projects under the loan were visited by the Audit mission. El Progreso and San Pedro Sula have made significant financial and administrative progress but it is only partially due to project activities. In San Pedro, the municipality was able to purchase a large IBM System 36 computer with fourteen satellite terminals. With this system it has been able to track the following fees and taxes: real estate, personal property, sales tax, public services (fire, street cleaning, environment), pavement, as well as notes and credits due. In addition, all municipal accounts are kept on the system. The system manager complained that in the five years the IBM equipment was rented before it was acquired under the Bank loan, it had become outmoded; he also noted that faster and more powerful machines could have been purchased more economically. Both cities estimate that they are now collecting about 50% of what they should be taking in, but this is a significant improvement over the pre-project situation. Mayors in the two cities attribute the improvement more to their willingness to take the political heat (begin insisting on debt repayment) than to any other administrative change-a provision in the new law expressly prohibits mayors from forgiving any payment owed to the municipality, and mayors may now be jailed if they cannot demonstrate that they actively attempted to collect from all debtors. The municipalities are much more autonomous now than they were before the project, and the mayors are making numerous moves to stabilize municipal finance. For example, the mayor of Tegucigalpa has reduced the number of municipal employees from 4865 to 1676 in a period of barely a year, raised compliance levels by inscribing 6,000 new properties on the tax rolls (bringing cadastres up to about 120,000 out of a potential total of 200,000), and made significant reductions in the municipal debt. It is hard to see a connection between project technical assistance activities and subsequent municipal policy reforms, however. 4.11 Cadastre Was Not a Prority. In Tegucigalpa, cadastral work had been neglected until a new administration entered last year. Out of 29 sectors in the Central District only 3 sectors have cadastres which are up-to-date. The computers that will be necessary to make rapid progress only arrived during the Audit mission visit, previously information had been kept manually. The mayor commented that the computer obtained under the loan had been disposed of because it had been inadequate to handle the volume of the city's data processing. The Director of Tegucigalpa's Cadastral Office estimates there are 80,000 unregistered plots in the city at this time. The Cadastre equipment purchased in San Pedro (US$8,000) and Tegucigalpa (US$6,000) was primarily office furniture, file cabinets and adding machines-items which contribute primarily to office comfort and convenience. 4.12 Problems with Municipal Sanitation. San Pedro purchased a fleet of 15 garbage trucks under the loan, 14 are well-maintained and in excellent condition and the remaining vehicle is reparable. El Progreso purchased 2 trucks to use for the same purpose. A 20 recent AID funded study estimated that only 1% of the city's production was being collected with that equipment (primarily because they are dump trucks, routinely assigned to road and school construction and maintenance activities). In an attempt to improve this situation two tractors have been assigned collection routes and the municipality currently estimates coverage at about 30%. Tegucigalpa purchased 18 garbage trucks under the loan: 10 remain in service and 8 are out of service but are reparable. While this equipment is highly rated by those who work with it in Tegucigalpa, it is of Argentine manufacture and the need to order spare parts from Argentina has led to long down- times, a situation which has been criticized by the city's newspapers. 4.13 Sustainable Infrastructure Finance. In terms of achieving the project's secondary (municipal) objectives, the mayors interviewed maintain that a 1987 law (Contribucion por Mejoras) has been the single most important development in recent years. Under the new provisions, neighborhood plebiscites vote on proposed infrastructure improvements. If 75% approve the project, the municipality is authorized to levy a special fee only on the area impacted and all families are required to contribute to full cost recovery. This approach is being used in all three cities, apparently with highly satisfactory results. Tegucigalpa uses the law's provisions to work in the more affluent areas and subsidizes improvements in poor zones. In terms of the municipal debt to BANMA, the mayor of San Pedro noted that the value of the city's shares in BANMA and the amounts that it had in its bank accounts when BANMA ceased offering banking services exceed what the city owes. In his opinion, it would be hard for him to make any additional payments on the loans before recovering the funds owed to the municipality. The Mayor of Tegucigalpa said that he faces many municipal needs far more urgent than covering debts to a moribund bank, and the sooner the central government comes to terms with the situation the better. The Mayor of El Progreso noted that he would like to pay off the loans, but building a new sewer system is his current highest priority, and it will be some years before this infrastructure is paid for. 4.14 CONSUPLANE: A Negligible Impact. Because of high rates of staff turnover, the Audit found only one person currently working in SECPLAN-the institution into which CONSUPLANE evolved-who had been in a division of CONSUPLANE working with the municipalities at the time of the loan. This individual did not recall the municipal development plan prepared under the Bank loan. It can be inferred that, since no one can recall its provisions, the plan is not greatly influencing that organization's ongoing municipal work. The Audit is forced to conclude that its impact was negligible. Nevertheless, SECPLAN is a more important institution than its predecessor. Consequently, institutional strengthening as a result of the study process is likely to have had a positive impact, but not one that was measurable by the Audit. 4.15 Advances in Governance. The Government's experience with the BANMA project ultimately led to major changes in governance and, in the three cities visited, the country is advancing significantly in public financial administration. The new municipal law (see para. 2.4) will help to overcome the weak financial capacity of the sector. The new municipal law also provides for greatly enhanced public participation in government as well. It requires mayors to hold five Cabildos Abiertos (open town meetings) yearly and it creates municipal Development Councils manned by local notables appointed by the mayors to advise on infrastructure investments. Administrative reforms that both reduce 21 operating costs and contribute to full cost recovery of infrastructure investments have taken place since the project closed within the municipalities that implemented project subcomponents (see paras 4.10 and 4.13). 4.16 Compliance with Operational Directives. Administrative, legal, and procedural requirements were complied with insofar as all of these were required for Board approval and loan effectiveness. 4.17 A Mired Record in Compliance with Covenants. Apart from the standard covenants of a typical loan/credit agreement there were few specific covenants applied to the credit and these fell into the following categories: a) Covenants specifying studies and action plans. These were all complied with. b) Covenant requiring the development of a training program on infrastructure maintenance was complied with. c) Covenant regarding the reduction of arrears. Provisions regarding the reduction of arrears in BANMA's portfolio and cash collection on currently due loans were not complied with. 4.18 The latter proved to be the most important, and failure to comply with its provisions was the principal reason for the project's unsatisfactory outcome.  23 5. Conclusions and Lessons Learned. 5.1 Project Rating. The PCR rates this operation as unsatisfactory largely due to its failure to meet its objectives regarding the strengthening of the Autonomous Municipal Bank. This Audit concurs with that evaluation and the reasoning behind it. The project outcome warrants an unsatisfactory rating. It is unlikely that there will be any sustainable benefits for either BANMA or the municipalities as a result of the loan. The MDPP's performance as a pilot intervention was poor and institutional development as a result of its activities is negligible. 5.2 Identifying Preconditions of Project Success. If the Bank is to give assistance to the Honduran municipal sector in the future, care will be needed in identifying limiting factors and providing in-country follow-up. One Bank staffer asked "why would you want a municipal bank in an environment where municipalities have found they cannot repay?" Institutional development is often planned as a prerequisite to policy reform. The experience of this project seems to suggest that the absence of a minimum reform platform can thwart institutional development. When this project began BANMA was in no shape to help anyone, yet the technical assistance components were largely an add- on-and what the country needed above all was a new framework for municipalities-to national government relationships. 5.3 Policy Reform a Precondition. While it may be extreme to say, as the PCR does, that the institutions that worked in the project were the problem and not the solution, the GOH's experience with BANMA was an important learning experience and perhaps even a catalyst for broader political reform. The GOH attempted to promote decentralization by creating yet another centralized institution. The BANMA project achieved limited results for reasons partially related to project design (note PCR's criticisms in paras. 4.02 and 4.03) but also because the necessary incentive structures and legal and policy preconditions for project success were simply not there. Major policy and (almost) cultural changes were expected, without fully recognizing that the prevailing tax/fee collection and debt repayment status served strongly established interests in important ways. The assumption that the mayors were not collecting their accounts due solely because essential information was inaccessible on paper and/or stored in file cabinets seems flawed and it was not substantiated by the Audit process. A more credible explanation is that the political cost of collecting those receivables was too high, especially given the likelihood that it was the most powerful families and companies that owed the most. Years of experience had shown that there were no significant costs related to not paying off BANMA loans, and the mayors were not legally required to collect. From a municipal perspective, improving BANMA's ability to collect debts (which it had not bothered to collect very actively before) actually created a new and troubling problem for the mayors. Finally, within the municipalities, the computerization of accounting/bookkeeping operations would eliminate numerous jobs that had traditionally gone to supporters of the mayor. 5.4 Achieving Preconditions. Changing the patterns of municipal governance and finance required a new consensus in numerous areas of action by a large number of 24 agencies. The loan was processed very quickly, however, almost as if the changes which had take to place could be accomplished by a stroke of the pen. Project results might have been better had conditions of negotiation (or even appraisal) been established for the municipalities (in areas such as purchasing shares in BANMA, opening the legally required bank accounts there, making a first loan payment, etc.). 5.5 Limited Municipal Financial Capacity. A municipal financial intermediary might have been more successful in a country where the municipal civil service was more established and, perhaps, funded by a sales tax-as long as mayors had the legal option of not collecting taxes (as one mayor said) "those who wanted to, paid, and those who didn't did not have to." The largest municipalities, where established patterns and practices were needed to change radically, did not share the pilot project's priorities. Tegucigalpa was not about to pay its debts, for example, no matter how much assistance it was given. During the loan period its mayors did not pay because of their extremely close relationship with the central government. San Pedro did not pay its debts during the loan period because it had been obligated to buy shares in an institution that the mayors knew was not going to last precisely because Tegucigalpa would never pay and the money was going to run out. 5.6 Intermediation Value. In Honduras the underlying problem which the MDPP attempted to correct was the financial viability of BANMA. Most municipalities were not paying off their loans. Neither were they collecting the fees owed to them. The MDPP assumed that by providing computers, data bases, technical assistance and training the delinquent municipalities would proceed expeditiously to collect their accounts receivable and pay off BANMA (President's Report para. 38). There were no reported instances of other Government actions in support of BANMA's collection effort. For the municipalities there was no value-added from having BANMA as a financial intermediary. BANMA did not provide them with additional funding or any technical assistance that they felt a need for. Even from BANMA's perspective, the incentives were somewhat perverse. Without redefining BANMA's functions, the result of the municipal strengthening reforms envisaged by the MDPP, had they been successful, would have been to weaken the political and financial power of BANMA. For example, by increasing its client's municipal financial and administrative capacity e.g. in the identification of municipal capital investments, BANMA made itself more marginal to the development process. 5.7 Lack of Coordination with other International Organizations. The lack of targeted incentives in the MDPP and of coordination with the IDB proved to be problematic, especially considering that IDB was offering BANMA a loan at a lower rate of interest to continue what was essentially the same program, albeit targeting (mostly) smaller municipalities. 5.8 Importance of T/A in Municipal Development Projects. The MDPP was as much a technical assistance project as it was a municipal development project. Given the track record of BANMA and the nature of its credit holdings, the MDPP was a high risk project from the outset. Was the time frame for implementation too short? Not really. Was the level of resources sufficient to make an impact? Perhaps the project should have provided more on-site technical assistance and financed less infrastructure if the MDPP was really 25 to be a pilot endeavor. One Bank staffer noted that all the good that came out of the project was accomplished before the first disbursement (with the preparation of the various municipal financial plans). In a sense, the technical assistance provided was peripheral and not central. In the municipality of San Pedro Sula which received computers, map-making software and digital plotters, no technical assistance was given in these areas. 5.9 Poor Assessment of Training Needs. While BANMA did provide training, it used a blueprint approach for course development which resulted in classes that were not relevant to what most municipal employees were actually doing. A town like El Progreso with 200,000 inhabitants which uses two tractors to pick up the garbage does not need the same administrative training curriculum in any activity as Tegucigalpa which manages a fleet of 53 specialized vehicles for the same purpose. BANMA offered the same basket of services to about 70 municipalities without taking into account differences between them. Such aspects as implementation capacity, repayment record, tax base, population or relationship to the central government were not factored into the equation in certain significant ways. A more tailored approach would have made the MDPP more functional. San Pedro Sula (a hotbed of entrepreneurial spirit) was fairly good about paying its debts, Tegucigalpa was not, knowing that the central government would come to its rescue rather than face service interruptions. On the other hand, many smaller municipalities were municipalities in name only. BANMA should have taken more of a flexible approach, developing service packages based on the complex of needs and problems which the various municipalities were experiencing. Municipal officials complained that training sessions were usually held in Tegucigalpa. The training would have been more effective had it taken place within the municipalities themselves, and it should have more directly supported the equipment purchased under the loan. 5.10 Supporting Sectors between Projects. In order to achieve the MDPP's objectives, policy and cultural changes affecting the way the municipalities relate to both local citizens and the central government would have been required. While such a radical change in practice may have been necessary to get municipal finance on a sustainable footing, project activities were clearly not sufficient to do the job. The Regional Training System for Local Municipal Development in Latin America (CELCADEL), is an EDI Latin America project which provides follow-up even when there is no open loan. In countries facing similar problems, and where the Borrower and implementing agency are receptive, the Bank may want consider new ways to provide ongoing technical assistance before or between loans. Several former directors of BANMA suggested that more intensive guidance would have been welcome. Helping even a few Honduran municipalities to collect from the public in a timely fashion and to begin paying off debts (which can easily amount to a multiple of their potential annual income) to municipal banks represents a sea change. The Honduran case suggests that, at a minimum, both political and legal frameworks needed to change before the municipalities could begin to approach a sustainable financial footing. Changes so fundamental may go beyond the scope of most Bank municipal loans: the amount of time during which the Bank needs to support governments as they modify their legal and institutional framework and otherwise side-step towards decentralization exceeds the time available in the normal project cycle, especially when policy reform and institutional strengthening cannot be simultaneous. Municipal/central government relationships are extremely difficult to change during the 26 time available within one project cycle if they also involve shifts in local government/local citizen relationships, a constraint which is further complicated given the turnover on both the Bank and Government side which normally occurs. 5.11 Preconditions to Decentralization. It seems reasonable to assume (especially given the unsatisfactory results of this project) that there are certain structural preconditions which must be present if decentralization is to be sustainable. Among those suggested by the MDPP and subsequent political reform in Honduras are: * Municipalities must have a degree of freedom in terms of day-to-day management. Where local officials cannot even spend $50 without direct authorization from central authority, technical assistance to municipalities may not result in innovation well-tailored to local circumstances; * A municipality needs to have the autonomy to set its own budget and to reallocate within it; * Mayors must be legally obligated to collect taxes due and to set fees for local services; * The decision to work only with municipalities of more than 80,000 population in the pilot project is also questionable, given that administrative practices in the smaller cities were likely to be easier to change, and that the Presidents Report (para. 37) noted that the larger cities would be likely to have problems covering their debts. 27 Annex 1 Comments from the Borrower (Translation from Spanish original) BANCO MUNICIPAL AUTONOMO Tegucigalpa, M.D.C. December 28, 1994 Mr. Yves Albouy Chief, Infrastructure and Energy Division Operations Evaluation Department The World Bank Washington, D.C. Dear Mr. Albouy: This is in reference to your letter dated November 14, 1994, where you request comments regarding the Municipal Development Pilot Project (Loan 2583- HO). Regarding the subject, I would like to inform you that we have analyzed the preliminary documents of the evaluation of the project in detail, and we consider acceptable your conclusions, consequently, we have no comments to modify the Project Performance Audit Report. Sincerely, P.M. LEONCIO MEJIA MORALES CHIEF, FINANCIAL ADMINISTRATION cc: Ministry of Finance and Public Credit