Mexico - Decentralization and Regional Development Project

Document of The World Bank FOR OFFICIAL USE ONLY Report No. 16229-ME IMPLEMENTATION COMPLETION REPORT MEXICO DECENTRALIZATION AND REGIONAL DEVELOPMENT PROJECT (LOAN 3310-ME) December 31, 1996 Mexico Country Department Latin America and the Caribbean Regional Office 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 (December 3, 1996) Currency Unit = Mexican New Peso (N$) US$1.00 = N$7.89 N$1 million = US$126,743 WEIGHTS AND MEASURES Metric System GOVERNMENT FISCAL YEAR January 1 - December 31 ABBREVIATIONS AND ACRONYMS CUD Convenio de Desarrollo Social (Social Development Agreement) GEF Global Environment Facility COPLADE Comisi6n de Planeaci6n para el Desarrollo Estatal (Commission for State Development Planning) DRD Decentralization and Regional Development PIDER Programa Integral para el Desarrollo Rural (Integrated Rural Development Program) SEDESOL Social Development Secretariat SAR Staff Appraisal Report Vice-President: S.J. Burki Director: 0. Lafourcade Manager, Sector Leadership Group: M. Baxter Staff Member: A. Silverman, Institutional Development Specialist  FOR OFFICIAL USE ONLY IMPLEMENTATION COMPLETION REPORT MEXICO DECENTRALIZATION AND REGIONAL DEVELOPMENT PROJECT (Loan 3310-ME) TABLE OF CONTENTS Page No. PREFACE ...................................................................................................................................i EVALUATION SUMMARY . .................... . ii-vi PART I. PROJECT IMPLEMENTATION ASSESSMENT . . 1 A. Background .1 B. Project Objectives 1....I Linkage with predecessor PIDER projects and sector studies .2 Complexity of project design .3 Inadequate institutional mandates and incentives .3 Municipio investment program .4 C. Achievement of Project Objectives .4 Overview .4 Municipal funds and Escuela digna .6 State investment program .8 Environmental protection .9 Poverty targeting .10 Monitoring system .10 Least cost solutions .10 D. Implementation Record and Major Factors Affecting the Project .11 Factors not subject to government control .11 Factors subject to govemrnment control .11 Factors subject to SEDESOL control .11 E. Project Sustainability. 12 F. Bank Performance .12 Preparation and appraisal .12 Implementation .13 G. Borrower performance .13 H. Assessment of outcome .14 I. Future operations .14 J. Key lessons learned .15 PART II. STATISTICAL TABLES .................................... 16 Appendix A. Borrower's contribution .................................... 30 This doeument has a restricted distribution and may be used by recipients only in the performance of their official duties. Its contents may not otlherwise be disclosed without World Bank authorization.  I  MEXICO DECENTRALIZATION AND REGIONAL DEVELOPMENT PROJECT (Loan 3310-ME) IMPLEMENTATION COMPLETION REPORT Preface This is the Implementation Completion Report (ICR) for the Decentralization and Regional Development Project in Mexico, for which loan 3310-ME in the amount of US$350 million equivalent was approved on June 26, 1991 and made effective on December 17, 1991. The loan was closed as planned on December 31, 1995. It was fully disbursed, with the last disbursement taking place on August 14, 1995. The ICR was prepared by Andrea Silverman of the Natural Resource Management and Rural Poverty Division (LA2NR), Country Department II, Latin America and Caribbean Region, with the assistance of Mark Cackler (LA2NR). It was cleared by Rudy Van Puymbroeck of the Legal Department, and reviewed by Michael Baxter, Chief, LA2NR, and Kreszentia Duer, Projects Advisor, LA2DR. It was finalized by the Sector Leadership Group for the Mexico Department. The ICR is based on findings of field visits, discussions with government officials and a review of the project file. The borrower commented on the draft report and prepared its own evaluation of the project, a summary of which is attached as Appendix A to this report. Among the other documents reviewed for this report were the Project Completion Report for the Mexico Integrated Rural Development Project -- PIDER III (Loan 2043- ME), December 11, 1990, and the Mexico Strategy Proposal for Regional/Rural Development in the Disadvantaged States, May 22, 1989.   MEXICO DECENTRALIZATION AND REGIONAL DEVELOPMENT PROJECT (Loan 3310-ME) Evaluation Summary Introduction 1. The Decentralization and Regional Development (DRD) Project (Loan 3310-ME) was designed to finance elements of a broad based multi-sector approach to address the problems of poverty in four poor states as part of the Government's poverty alleviation program, Solidarity. The Solidarity program, initiated by the Salinas administration in 1989, was a collection of programs that were implemented through a partnership of federal, state and local governments, and communities using matching grant resources.1 Solidarity's annual expenditures reached US$2.2 billion in 1994, the last year of the Salinas administration, with the DRD project financing specific elements of the program. In 1994, this DRD support amounted to 31% of Solidarity expenditures in the four project states, or approximately 7% of the total national program. Project Objectives 2. The objective of the DRD project was to increase access of poor and indigenous populations in the states of Chiapas, Guerrero, Hidalgo, and Oaxaca to basic infrastructure, social services, markets and technologies. The project was to: (i) assist the states to develop, execute and finance investment programs with a strong poverty alleviation focus; (ii) assist the federal and state planning and budgeting ministries in making more efficient and flexible project selection, environmental assessment, and monitoring and evaluation; and (iii) enable state institutions and municipio& to identify, prepare, build, operate, and maintain the investments in a more participatory manner. The US$350 million loan over four years (1991-1994) was used by the project executing agency, the Social Development Secretariat (SEDESOL), to finance four components: state and municipio rural investment; environmental protection; cultural site restoration; and institutional development. Elements of the Solidarity program have been continued by the current presidential administration under the name of Programa de Bien Estar. 2 "Municipios" in Mexico are the equivalent of districts or counties in other countries. They include rural as well as urban areas.  - ill - Implementation Experience and Results 3. Project implementation experience was generally positive and resulted in substantial achievement in each project component. Implementation was hindered, however, by an overly complex project design involving many different sector agencies in each of the four states, along with federal and state planning agencies and 840 municipios. Most of the project loan financing was dedicated to the state and municipio investment programs, with US$165 mnillion and US$153 million of loan disbursed in those two programs, respectively. 4. Municipio Investment Program. The Municipal Funds and Escuela Digna programs were well conceived and extremely successful in meeting project objectives. This program succeeded in: (i) creating or rehabilitating infrastructure in rural areas at low cost; (ii) providing rural municipios the opportunity to manage resources and to plan and implement (through community groups) small scale public works; (iii) establishing a new forum for citizen participation accessible to the rural poor; and (iv) creating and testing a successful model of community implementation of local small scale projects, providing a viable alternative to the traditional centralized implementation. The Municipio Investment Program financed the construction or reconstruction of an estimated 5,100 school buildings, rehabilitation and/or equipping of 18,000 school buildings (through the Escuela Digna program), rehabilitation of 2,500 rural roads, including construction of bridges and fords, and construction or rehabilitation of 4,500 water supply systems. The quality of physical infrastructure was adequate in most cases, although some recurrent problems were found in the water supply and income-generating investment, due to the technical and/or organizational complexity of these two sectors. Problems were discovered in approximately half of the projects in these two areas, accounting for approximately 10% of the Municipal Fund and Escuela Digna financing. 5. State Investment Program. The State Investment Program brought mixed results. Most successful was the school construction program, which built school buildings at two- thirds the cost of those federally contracted, followed by the rural water supply and electrification programs. The weakest components were rural roads and agriculture investments which were much less successful, mainly due to inadequate procedures for subproject identification and selection. Institutional capacity problems existed, especially in the roads and agriculture sectors, which were compounded SEDESOL's inaction in addressing issues of quality and application of subproject selection criteria, as weli as the lack of linkage of the states' Solidarity budgets to their compliance with subproject eligibility criteria. Physical achievements in the State Investment Program included construction, rehabilitation or extension of approximately 740 water supply systems, construction of 4,000 classrooms, and upgrading of 800 km of rural roads. In addition, one major success of the program was the successful piloting of a labor-intensive, low cost approach to rural road rehabilitation called "Puntos Criticos" in Hidalgo during the last year of the project, resulting in the rehabilitation of 400 km of road. 6. Environment Component. This component (US$16 million of loan funds disbursed) principally supported the protection of the Selva Lacandona tropical rainforest in Chiapas. Most successful activities of this component were: (i) demarcation of the reserve  - lv - and the creation of the park protection service, with many of these activities transferred to the Comunidad Lacandon by the last year of the project; and (ii) basic education and community medical programs, using trained community workers. Less successful were the agriculture activities of the program, especially those activities which attempted to implant technology "packages" which were not appropriate to the conditions of the area. By contrast, those productive investment activities which involved significant amounts of community organization and participation, combined with technical assistance, were successful. Limiting the success of this component was the delay in contracting the agro- ecological study of Selva Lacandona, which was still not complete by project closing. 7. Institutional Development Component. This component financed a combination of long and short term consultants to assist the State Planning Commissions (COPLADEs) in analyzing and addressing regional development issues. This technical assistance may have had a positive effect on state capacity in general, but was flawed in that insufficient expert assistance was provided to line agencies and to local, municipal governments where planning and technical improvements were most needed. 8. Cultural Sites Protection Component. This component (US$3 million of loan funds disbursed) was designed to preserve archeological sites by the National Institute of Anthropology and History (INAH). While limited in scope, it was extremely successful in meeting the twin goals of protecting cultural heritage and providing a source of short-term employment in poor regions. 9. Targeting to the Poor. The targeting of DRD project investments to the poor was in large part successful. Better off communities also benefited from investments, but analysis of the Municipal Funds showed that greater proportion of this investment in basic, small scale infrastructure was received by rural than urban areas, where poverty is concentrated, and that within rural areas, heavily indigenous communities received as much per capita as non-indigenous communities. 10. Monitoring System. The project succeeded in instituting a data base system for monitoring the investment portfolio at the state and national levels. This system allowed for monitoring the evolution of the project, as information became available regarding how states and municipios decided to invest project funds. Sustainabiity of the information system has been made more likely by the integration of the system within SEDESOL's normal administrative approval procedures. In addition, the national Solidarity program has adopted the system. It was installed in all eight of the Second DRD project (Loan 3790- ME) states in 1994, and is being expanded to all remaining states starting in 1996. 11. Sustainability. The sustainability of the new Municipal Fund and Escuela Digna program mechanism has been demonstrated by the continuation and strengthening of this program as an integral part of the resource transfer system among levels of government. Sustainability of the investments under the Municipal Fund and Escuela Digna program, and the state investment program, still needs to be studied, but it is anticipated that the sustainability rate will be significantly higher than for most rural development projects,  -v - given the simple design and significant community participation in the selection and construction of many of the works. 12. Bank Performance. The Bank's performance had both strengths and weaknesses. On the positive side, Bank staff worked closely with government to identify new institutional mechanisms for efficiently implementing the project by communities and municipios. During implementation, Bank staff worked with project implementing agency staff to review field experience and to revise and strengthen project guidelines for each component. On the negative side, the appraisal did not adequately identify many of the implementation difficulties mentioned above, resulting in significant performance problems in some components. 13. Borrower Performance. Borrower commitment to core elements of the project was strong, especially to the Municipal Funds and Escuela Digna subcomponent. These programs benefited by Government's general support to Solidarity, including the traininlg of Solidarity Committee members. The executing agency should also be commended for the creation and maintenance of a new information system which has yielded important technical, social and geographical information about project investments (see para. 10 above). Government performance could have been improved by it: (i) conducting more frequent field supervision to address technical quality and sustainability issues; (ii) focusing technical assistance to address performance problems in sector agencies and municipios; (iii) enforcing investment program guidelines and eligibility criteria; and (iv) providing greater financing to key project elements, including the agro-ecological study of the Selva Lacandona and efforts by state governments to pilot the Puntos Criticos road rehabilitation program. Summary of Findings, Future Operations, and Key Lessons Learned 14. Assessment of Outcomes. The outcome of the project is rated satisfactory, principally due to the successful Municipal Funds program, which created a new mechanism for planning and implementing poverty-targeted rural investment programs by municipios, with high levels of beneficiary and community participation. In addition, the Seiva Lacandona subcomponent made important advances in the protection of the Montes Azules Reserve by involving the local indigenous and other communities around the reserve in managing the protection program and participating in the delivery of improved basic health and education services and income-generating projects. 15. Future Operations. Follow-up on some of the most important investments and institutional development actions financed by the DRD project has been folded into the Second DRD project (Loan 3790-ME), which continued to support aspect of Solidarndad and its successor, Programa de Bienestar. Further needs to strengthen protection for the Selva Lacandona are being addressed by an existing GEF project, and by community development activities financed by Municipal Funds within the Second DRD project and by State government. In addition, the "Southern States Initiative" in Chiapas and Oaxaca is identifying ways to address general rural development issues in those states, and a possibe  - vi - Sustainable Development Project could apply lessons learned about agricultural development in poor, indigenous areas in Oaxaca. 16t Key Lessons Learned. Lessons learned include: (i) the kind of approach used in the Municipal Funds program of allowing rural communities to actively manage their own small scale development investments can be successful; (ii) because national government may not be willing to impose or enforce rules and guidelines for the use of funds transferred to lower level jurisdictions, such programs should be designed to ensure broad participation and transparency of resource use; (iii) participatory evaluations, involving different levels of government and community members, can be done effectively with limited resources; (iv) capacity building measures and incentive structures need to be integrated into the design of programs containing substantial community complexity and diversity of sector investments, and need to include also the local level; and (v) the need to ensure commitment of involved agencies and clear institutional mandates.   MEXICO DECENTRALIZATION AND REGIONAL DEVELOPMENT PROJECT (Loan 3310-ME) IMPLEMENTATION COMPLETION REPORT PART I. PROJECT IMPLEMENTATION ASSESSMENT A. Background 1. The Decentralization and Regional Development (DRD) Project (Loan 3310-ME) was designed to assist government's efforts to alleviate poverty in southern Mexico. Over a period of four years (1990-1994), the project loan financed approximately 15% of federal resources going to the new multi-sector regional development program, Solidarity, in the states of Chiapas, Guerrero, Hidalgo and Oaxaca. The Solidarity program was the centerpiece of the Salinas administration's poverty alleviation strategy, and was implemented through a partnership of federal, state and local governments, and communities using matching grant resources. 2. The Solidarity program, and the DRD project, was a direct outgrowth of the government's earlier regional development program, supported by a series of three Integrated Rural Development Projects (PIDER I, II and III, Loans 1110-ME, 1462-ME and 2043-ME, respectively) implemented between 1975 to 1988. These programs provided funds to support investments in poor regions in a number of states. For the implementation of the most recent of these projects, a new intergovernmental mechanism (the Convenio Unico de Desarrollo, CUD1) was established to allow investments to be jointly financed by state and federal governments. A state level inter-agency planning and development commission (COPLADE) was created in each state to review and approve regional development program investments, and provide inter-governmental and inter- agency coordination of all investments. When Solidarity was introduced in 1989, it continued to use the CUD and COPLADE mechanisms for coordinating joint federal and state poverty activities. B. Project Objectives 3. The central objective of the DRD project was to increase access of poor and indigenous populations in four of the poorest states of Mexico (Chiapas, Guerrero, Hidalgo, and Oaxaca) to basic infrastructure, social services, markets and technologies, through financing of specific elements of the Solidarity program and related institutional development technical assistance. The project was to: (i) assist the states to develop, 1 Renamed Convenio de Desarrollo Social in 1993.  - 2 - execute and finance investment programs with a strong poverty alleviation focus; (ii) assist the federal and state planning and budgeting ministries in making project selection, environmental assessment, and monitoring and evaluation more efficient, and financing more flexible; and (iii) enable state institutions and municipios? to identify, prepare, build, operate, and maintain the investments in a more participatory manner. These objectives were to be achieved with the assistance of a US$350 million loan over four years to finance four components: state and municipio rural investment; environmental protection; cultural site restoration; and institutional development. In addition, the project incorporated action plans for the decentralization of government services and fiscal strengthening of local government. Linkage with predecessor PIIDER projects and sector studies 4. The Project Completion Report (Report No. 9175) for PIDER III (1981-1988) indicated that while that project generally met its objectives, it suffered from several problems. These included: (i) low levels of total public investment in poor regions in spite of the project; (ii) inadequate participation of beneficiaries; (iii) lack of technical quality in the preparation and appraisal of projects; and (iv) inadequate project monitoring. Furthermore, a poverty sector study was conducted in 1988-89, ("Mexico Strategy Proposal for Regional/Rural Development in the Disadvantaged States", Report No. 7786- ME, May 22, 1989) concluded that the situation of poverty in Chiapas, Guerrero, Hidalgo and Oaxaca would not improve "unless a major effort is mounted in terms of public investment, technical assistance, and community leadership and mobilization." 5. The DRD project was designed to address the findings of the poverty study of the disadvantage states and to build on the learning of the PIDER projects. The project name "Decentralization and Regional Development" reflected its focus on the development of the poor rural south of Mexico through decentralized planning and implementation. In the project, expenditure programs were planned and coordinated at the state and municipal level and not by central government; and execution of project components were to be done by agencies at the state or lower level, moving the locus of execution closer to the intended beneficiaries. The project was also designed to specifically address the weaknesses of its PIDER III predecessor. First, to increase the flow of funds to the poor states, during project preparation government agreed to reform the federal revenue-sharing formula. Second, to increase participation of poor communities, the Solidarity program included a strengthened role for communities through the Solidarity Committees and the Municipal Funds program, which financed community-initiated proposals. To address the problem of inadequate implementation capacity, an institutional development component was designed to strengthen state level planning, plans were made for greater sector decentralization, and simple subproject selection criteria were developed for most sectors. Finally, insufficient monitoring in PIDER III, due in part to the cutbacks in government operations, was to be 2 "Municipios" in Mexico are the equivalent of districts or counties in other countries. They include rural as well as urban areas.  addressed by strengthening the project monitoring capacity of the implementing agency, the Social Development Secretariat (SEDESOL),3 at central and state levels. Complexity of project design 6. The objectives of the DRD project were consistent with priorities established in the Bank's Country Assistance Strategy, but were too far-reaching to be fully achievable given the limited capacity and commitment of involved agencies. The financing of a broad-based rural investment program at the state level, i.e., the "State Investment Program," created severe implementation problems as it involved state offices of five federal agencies and over 20 state agencies (at least five in each state) and six major areas of investment-rural water supply, drainage, rural roads, rural electrification, educational infrastructure, and agricultural development. Many of these sector agencies were weak in one or more areas, including planning, technical approaches, community participation, contracting, and operations and maintenance. 7. The project was further complicated by the inclusion of components not closely related to the central investment program, the components included: (i) an environment component broken into two subcomponents, one to strengthen state environment agencies and one to protect the Selva Lacandona tropical rainforest in Chiapas; (ii) a cultural sites component to preserve archeological sites by the National Institute of Anthropology and History (INAH); and (iii) as part of the institutional development component, technical assistance to evaluate and strengthen the Regional Solidarity Funds of the National Indigenous Institute (INI). The inclusion of these components created significant implementation challenges for both government and the Bank, as they involved an additional number of agencies and stakeholders, and in the case of the Selva Lacandona, required very difficult development issues to be addressed. Inadequate institutional mandates and incentives 8. Difficulties with the state investment programs were compounded by the fact that SEDESOL's federal and state offices did not have a clear institutional mandate to oversee the quality of investments or the application of subsector project selection criteria. They did not actively enforce the state's application of agreed selection criteria or quality standards, and occasionally presented for Bank approval and reimbursement subprojects that did not meet eligibility standards. In addition, the incentives for states to comply with project guidelines were weak; their Solidarity budgets were not made contingent on or linked to the eligibility of investments for DRD project financing. 9. As part of the decentralization and institutional strengthening objectives of the project, action plans for the decentralization and deconcentration of agriculture and rural 3 The original project implementation agency was the Secretaria de Programacion y Presupuesto (SPP- Programming and Budget Secretariat), which had previously been responsible for PIDER and PDR. SPP was re-organized in 1992 and the subsecretariat responsible for Solidarity and the DRD project was eventually transferred to the newly created Secretaria de Desarrollo Social (SEDESOL, Secretariat of Social Development).  -4 - roads functions to the states were included as supplemental letters to the legal agreement. These actions, however, were not under the purview of the project implementing agency (SEDESOL), but of federal and state line agencies, and were only partially achieved during the project. In retrospect, they would have been better included in sector policy dialogue and sector operations. Municipio investment program 10. In contrast with the over-complexity of the state investment program, the design of the muntcipio investment program (Municipal Funds and Escuela Digna) was well conceived and extremely successful in meeting project objectives. By its nature, the program required much less technical oversight and planning; community participation in investment selection effectively took the place of eligibility criteria; and both central and state governments showed their strong commitment to this program with budgetary and staff resources. C. Achievement of Project Objectives Overview 11. In the largest project component, state and municipio investment, the project supported two main groups of activities: (i) Municipal Funds and Escuela Digna, which financed small scale infrastructure investments selected and implemented by communities and municipal authorities; and (ii) State Investment Programs, which were selected by the State Planning Commissions, and carried out by responsible state-level line agencies. 12. Municipio investment program. The DRD project was able to achieve a large part of each of its objectives because of the successful and innovative Municipal Funds and the related program limited to school rehabilitation called Escuela Digna. In these programs, the project used new institutional and financing mechanisms for identifying, preparing, building, and maintaining rural infrastructure investments in a participatory manner. Initially conceived as a small subcomponent of the project, funding for the municipal investment program was more than tripled through two loan amendments, so that municipios received a total of US$305 million in project financing (including US$153 million, or 44% of the US$350 loan). In the Municipal Funds program, over 20,000 subprojects were financed at an average cost of less than US$13,000. In the Escuela Digna program, over 18,000 schools were rehabilitated or equipped at an average cost of approximately US$3,000. Since the end of the project, government has renewed and increased its commitment to these programs, now merged into one program, which continues to receive support in eight states under the Second DRD Project (Loan 3790- ME). 13. State investment program. In the state investment subcomponent, success varied by sector. Inadequate institutional capacity, especially in agriculture and rural roads, led to poor performance, and to significant reductions in the allocation of project resources to the state program in general and specifically to those two sectors. As many proposed rural  roads and agriculture subprojects were rejected for project financing for not meeting minimum criteria, increased financing was made available for the highly successful school construction and rehabilitation program. Even when the state and Municipal Funds road investments are taken together, they accounted for only 15% of project financing, compared to the 35% estimated at appraisal. Agriculture investments were limited in both the state and Municipal Funds programs, and only accounted for 7% compared to the 29% of project financing estimated in the SAR. The educational infrastructure program was the largest investment category in the state program, and, including the Municipal Funds and Escuela Digna program, accounted for 32% of project financing compared to the SAR estimate of 12%. The state water supply component financed the construction or rehabilitation of relatively simple rural water supply systems with an average cost of US$85,000 per system or US$100/water user. When combined with the much smaller scale investments made under Municipal Funds, water supply investments amounted to 18% of project financing, double the SAR estimate of 9%. 14. Environment and cultural sites protection components. The Selva Lacandona program achieved only partial success, due to: (i) problems of project design, which overly depended on government rather than community action; (ii) conflict in the area which began in January 1994; and (iii) limitations placed by government on the financing of elements of the program. The total loan disbursement was significantly below expectations, at only US$16 million of the budgeted US$41 million. This figure, however, is somewhat understated, as some expenditures for the program were not eligible for loan financing because of a lack of consistency with procurement guidelines. Limiting the success of this component was the delay in contracting the agro-ecological study of Selva Lacandona, which was still not complete by project closing. The cultural sites component, as originally designed was limited in scope (US$3 million). Nevertheless, it was successful in meeting twin goals of protecting cultural heritage and providing a source of short-term employment in the region. 15. Institutional development component. The success of this component is difficult to quantify. Support for the Regional Indigenous Funds through the provision of technical assistance and a program evaluation was somewhat successful, but success was limited by the fact that the evaluation was completed only near the end of the project, and to its results were not fully addressed. The results of other elements of the Institutional Development Component, the long-term technical assistance and training contract and other technical assistance and training activities of each state were positive, but did not fully meet expectations laid out in the SAR. 16. Effect of the DRD project on Mexico's poverty alleviation program. Despite improving the lives of millions of people, elements of the Solidarity program have suffered from various shortcomings. These include: (i) investments being selected in a non- transparent process and not adequately targeting the poor; (ii) limited community involvement in investments that were contracted or implemented by government, rather than being managed by communities; and (iii) technical problems and the lack of resources for the operation and maintenance of investments. The elements of Solidarity included in the DRD project, which financed only 31% of the Solidarity investments in the four project states and 7% of the Solidarity investments nationwide, had a much better record. The  - 6 - DRD project has had a substantial beneficial impact on Solidarity in general by demonstrating how Municipal Funds could work, leading to a major increase in funding for that program in all Mexican states. In addition, due to the positive experiences in states of Hidalgo and Oaxaca, the special Municipal Development Councils were strengthened and given more responsibilities in all the states of Mexico. The project operational manual for Municipal Funds was adopted in 1996 for use in all states, including the requirement for limiting investment in district capitals, formulas for allocating resources among municipios, and varying community contribution requirements depending on project type. In addition, the data base system developed under DRD is being instituted nationally to monitor the Solidarity Municipal Fund, a new program encompassing the old Municipal Funds. As a consequence, the implementation impact of DRD, which is considered satisfactory on its own terms, becomes even more noteworthy because of its beneficial impact on the government's poverty alleviation program overall. Municipal funds and Escuela digna 17. Overall accomplishments. The objective of the Municipal Fund and Escuela Digna component was to finance rural development investment targeted to the poor while strengthening planning and beneficiary participation at the municipal level. Qualitative and quantitative evaluations of the Municipal FundlEscuela Digna program were conducted jointly by SEDESOL, state governments and the Bank, as well as through an independent evaluation of Municipal Funds in Oaxaca. They confirm a number of accomplishments of the program, including: (a) Valuable infrastructure was created or rehabilitated in rural areas at low cost, especially school buildings, bridges, roads, and, to a lesser extent, water supply. (b) Rural municipios were given experience, many for the first time, in planning development activities, managing resources and implementing (through community groups) small scale public works. (c) A new forum for citizen participation was created, in which members of rural poor communities are given an opportunity to participate in local development decisions. (d) A successful model of community implementation of local small scale projects was established, providing a viable alternative to the traditional centralized implementation or contracting of these works and which often resulted in failure. 18. Community participation. While quality of implementation of Municipal Funds varied across states and municipios, in nearly all circumstances community participation was a major element of their success. Following program guidelines, nearly all municipios established special councils composed of members or representatives of rural communities to manage the program with the Municipal President. Subprojects selected by the municipios reflected a high degree of community commitment, demonstrated by the  minimum community contribution of 20 percent of subproject cost, including community participation in construction activities. In Hidalgo and Oaxaca, community Solidarity Committees were usually given complete responsibility for managing the investment budget, including purchasing materials, hiring skilled labor and supervision of the work. In Guerrero and Chiapas, municipal presidents were less willing to transfer resources to communities, preferring to control material purchases centrally; communities still would be responsible for managing implementation, but using materials provided to them. 19. Participatory field evaluations and Bank supervision provided information of how the participatory elements of Municipal Funds were working, and gave evidence that the more management was delegated to communities through the Solidarity Committees, the more likely that subproject selection reflected community interests, and that implementation would be done with higher quality and cost efficiency. Project supervision stressed the importance of following the community management model and of strengthening the role of the municipal council to ensure participatory decision-making. Of the four project states, however, significant efforts to strengthen these practices occurred only in the states of Hidalgo and Oaxaca, where they received strong support by state level officials. 20. Because of their achievements, the Municipal Funds were used as model for community and local government participation in development planning and implementation in several other countries. The Municipal Funds became the model for the successful redesign of the Brazil Northeast Regional Development Project (Loan 2761-BR). In Nicaragua, a similar fund is proposed as part of the Rural Municipalities Project (Credit 2918-NI). 21. Physical targets. The Municipal Funds and Escuela Digna program resulted in the construction or reconstruction of an estimated 5,100 school buildings, rehabilitation and/or equipping of 18,000 school buildings (through the Escuela Digna program), rehabilitation of 2,500 rural roads, including construction of bridges and fords, and construction or rehabilitation of 4,500 water supply systems. Supervision missions as well as an independent evaluation reported that the quality of schools was generally satisfactory, and that subproject costs were from one-half to two-thirds that of equivalent contracted works. Similarly, works by communities in simple road repair and bridge and ford construction were considered of good quality at extremely low cost. Water supply subprojects were much less successful as shown by an evaluation of Municipal Funds in Oaxaca, in which 43% of a sample of 30 water supply projects were not functioning because of technical design inadequacies or because project funding was only adequate for one stage of the needed construction. Success was also very limited in the small number of income generating projects financed by Municipal Funds, especially those requiring group organization such as corn mills (molinos de nixtamal) and bakeries. In spite of identification of needs for technical assistance during project supervision, little improvement was made in this area during the project life. Strengthening of technical assistance to communities and the conduct of annual physical performance reviews of subprojects became key activities agreed within the Second DRD project.  -8- State investment program 22. Investment planning. Only limited improvements of the state planning function were achieved during the project, principally because of differences between the Bank and government regarding effective rural development planning. While both agreed on the need to promote participation in the identification of development priorities, a major area of disagreement was regarding the use of selection criteria in order to ensure the efficient and effective targeting of development funds. While the project incorporated the use of eligibility criteria based on maximum cost per beneficiary, depending on the sector, commitment to this approach by both the national and state governments was weak. One measure of this was that careful Bank review of Statements of Expenditures was required throughout the project to ensure that ineligible subprojects were not financed. 23. Sector planning innovation -- the "Puntos Criticos" program. Significant progress and innovation was made in sector planning, however, with the piloting of the spot improvement (puntos criticos) rural roads rehabilitation program in Hidalgo during the final year of the project. This program used a model for cost-effective, labor intensive rural road rehabilitation which subsequently became the basis for a project component of the Second DRD project (Loan 3790-ME). In 1994, 400 km of roads in Hidalgo were rehabilitated at an average cost of US$7,000 per km. 24. Physical targets. Approximately 740 water supply systems were constructed, rehabilitated or extended under the state investment component, serving approximately 630,000 persons at an average cost of US$100 per person. While these systems were mainly contracted, the project also supported in its final years systems constructed by communities. The educational infrastructure investments of both the state and municipal funds investment programs were especially successful. School buildings were contracted by states at two-thirds the cost, on average, of those contracted by central government, with no difference in quality. In addition to the estimated 4500 classrooms built under the Municipal Funds, an additional 4,000 classrooms were constructed under the state program. Approximately 800 km of rural roads were upgraded (unpaved) during the project at an average cost of US$80,000 per km, with 76% of the investment occurring in Hidalgo and Oaxaca. The amount spend in this component (US$ 63 million total, of which US$ 31 million was reimbursed from the loan) was less than half the amount anticipated in the SAR, as much of the state rural road portfolio did not fit the project eligibility requirement of a maximum cost of US$500/inhabitant or did not meet contracting requirements as they were implemented by direct administration of the state road agency (especially in Guerrero). 25. Institutional development technical assistance and the state investment program. The institutional development component financed a combination of long and short term consultants hired by SEDESOL and the states to provide training and to conduct studies and analysis, focusing on regional development issues. This technical assistance may have had a positive effect on state capacity in general, but it had little apparent impact on improving project-related investment planning and implementation, which was centered in the sector line agencies and the municipios. The project design was flawed in that institutional development activities were linked to the centralized state level planning  -9- organizations (COPLADEs), and not to the line agencies or local governments where planning and technical improvements were most needed. Environmental protection 26. The Environment Component had two objectives: (i) environmental protection of the Selva Lacandona in Chiapas; and (ii) institutional strengthening of the state offices of the national environment agency (SEDESOL/INE)4 in order to improve the quality of environmental reviews of rural development investments. The first objective was addressed by the financing of local long-term consultants in each state office to assist in the development of review procedures, for a total of US$ 2 million. 27. Approximately US$ 14 million of loan financing was used for the Selva Lacandona Protection Program. This is substantially below the amount anticipated in the SAR of US$ 36 million, because some government expenditures were not eligible for reimbursement and because the implementation capacity of involved federal and state agencies had been over- estimated at appraisal. In addition, project expenditures declined substantially after the January 1994 Zapatista uprising as this prevented continuation of many program activities in nearly one-half of the Selva Lacandona area. 28. Activities supported by the program included: (i) demarcation of the reserve and of communal lands of the Lacandons; (ii) resolution of land tenure disputes within and around the reserve in order to promote greater recognition and observance of reserve boundaries; (iii) establishment of a park protection service, in order to prevent incursions into the reserve and prevent extraction of logs and other valuable flora and fauna through a road inspection system; (iv) establishment of community based approaches to the provision of basic medical and education services; and (v) provision of other social and economic investments to improve the quality of life of settlements around the reserve. 29. The most successful social programs were the basic education and community medical programs, each of which trained community workers to work as primary school teacher and community health workers, respectively. The continuation post-project of the basic education program looks likely because of the adoption of the program by the State Education Secretariat. The continuation of the medical program, however, is in doubt. Substantial progress was made in the demarcation of the reserve and the creation of the park protection service. During the final year of the program, implementation responsibility of significant parts of these activities was transferred to the Comunidad Lacandon. The agriculture activities of the program were less successful, especially those activities which attempted to implant technology "packages" which were not appropriate to the conditions of the area. The productive investment activities that were successful involved significant amounts of community organization and participation. These included the women's program (bakeries and community gardens), and export oriented programs with the organic coffee producer associations and the Comunidad Lacandon. 4 The first implementor of the this component was the Secretariat of Urban Development and Ecology (SEDUE). When SEDUE was merged with part of SPP in 1992, the environmental protection functions were taken over by the Instituto Nacional de Ecologia (INE-National Ecology Institute), within SEDESOL.  - 10 - Poverty targeting 30. The targeting of DRD project investments to the poor was in large part successful. Better off communities also benefited from investments, but analysis of the Municipal Funds showed that a greater proportion of the investment was received by rural than urban areas, where poverty is concentrated, and that within rural areas, heavily indigenous communities received as much per capita as non-indigenous communities. The nature of investments also reflected poverty targeting, as investment went to address basic needs of educational facilities, electrification, water supply, and roads and bridges in poor rural areas. Monitoring system 31. A major focus during project implementation was the institution at national and state levels of a data base system for monitoring the investment portfolio, including for Municipal FundslEscuela Digna, the state investment programs, and the Selva Lacandona program. This system allowed for monitoring project implementation as information became available regarding how states and municipios decided to invest project funds. It replaced a focus on prior reviews of annual investment programs (POAs, Programa Operativo Anual), which were very tentative and often did not reflect final investment decisions. The information system allowed the Bank and government to review compliance with eligibility criteria for the state programs and to monitor the nature and location of projects being implemented within the state programs and the Municipal Fund and Escuela Digna programs. Sustainability of the information system is not ensured, but has been made more likely by the integration of the system within SEDESOL's normal administrative approval procedures. By the last year of the project, all states were actively using the system for their own analytical purposes, as well as for data sharing and for obtaining expenditure approvals from the central SEDESOL office, and the system was even being introduced in states not participating in the project. Least cost solutions 32. A net present value (NPV) was not calculated for project investments at the time of preparation, given the difficulties of such calculations for the wide range of investments, many of them small scale "social" infrastructure such as water supply systems and classrooms. Instead, the eligibility criteria for investments used a least cost approach by establishing cost limits per beneficiary for investments, while requiring NPV calculations for agriculture investments over US$100,000. Community participation in the selection and implementation process for Municipal Funds and Escuela Digna program led not only to the selection of high priority investments given limitations of funding, but also to saving of significant government resources from the mobilization of community resources and from the cheaper cost of these works.  - 11 - D. Implementation Record and Major Factors Affecting the Project Factors not subject to government control 33. One significant factor affecting implementation not subject to control by the project agency was the initiation of Zapatista conflict in the Altos and Selva Lacandona regions of Chiapas in January 1994. From this time, financing of the Selva Lacandona component was limited to areas outside of the conflict zone, given the difficulty in project implementation and supervision. This meant that the process of demarcating the Montes Azules reserve in the Cafiadas region could not continue, and the education, health and productive projects were limited to two of the five subregions in the Selva (Comunidad Lacandon and Marquez de Comillas). This prevented gains in the protection and research programs from being consolidated, and held up the contracting of the agro-ecological study and the completion of the demarcation program. These efforts are continuing after the project, with government support. In addition, the conflict has created an opportunity for greater incursion into the reserve and further deforestation in the Cahadas region, reducing the potential impact of the project. Factors subject to government control 34. During the project, Government showed strong support for financing of Solidarity and within Solidarity, the Municipal FundslEscuela Digna program. Annual financing of that program was doubled from 1990 to 1994 in the four project states, leading government to request a reallocation of funds among loan categories in 1992. This allowed the project to be much more successful in achieving community participation and municipal strengthening objectives than it would have been otherwise. On the negative side, government's continuing policy to delay liberation of a large percentage of budgetary resources until the middle of the fiscal year, and little flexibility in carrying over unused budget from one year to the other, common to many projects and programs in Mexico, created severe problems for implementing agencies and municipios, especially since funds for construction projects were therefore not available until after the beginning of the rainy season. 35. Another problem, already mentioned above (para. 8), was the decision to manage the allocation of the Solidarity budget to states in such a way that states would have no financial incentive to invest in subprojects that met project eligibility criteria. States received no additional funds for implementing project eligible subprojects. Because of this, the project only financed a small percentage of the state Solidarity investment program, as government often selected investments regardless of whether they were project eligible. Factors subject to SEDESOL control 36. The project implementing agency, SEDESOL, did not directly execute the programs. They were implemented by state and federal agencies through contractors, and by municipios and communities. While SEDESOL did not directly control implementation, it had significant influence through its approvals of financing, and its funding of local level technical assistance and training. Project implementation would have been strengthened  - 12 - substantially had SEDESOL better enforced the Municipal Funds guidelines regarding: (i) the distribution of funds within municipios to avoid concentration of investment in the municipal capital; and (ii) the requirement that the Solidarity Committees be given full implementation responsibility, including for the management of funds. The quality of investments could also have been improved if SEDESOL had paid greater attention to monitoring the participation and technical aspects of the investments, and to assisting states address implementation problems identified through monitoring. E. Project Sustainability 37. The sustainability of the new Municipal Funds/Escuela Digna program mechanism has been demonstrated since project closing by government's substantial increase of resources going to the program. What began as a special program has become an integral part of the resource transfer system among levels of government, providing supplemental resources, especially to poor municipios, to assist their financing of priority poverty- targeted investments. Sustainability of the investments under the Municipal Funds/Escuela Digna program and the state investment program has not been studied. Nevertheless, given that most investments were for very small, technically simple works, they were selected and partially funded by communities themselves, and quality was in most cases adequate (except for the Municipal Funds water supply and income-generating investments), it is anticipated that the sustainability rate will be significantly high than for most rural development projects. F. Bank Performance Preparation and appraisal 38. The Bank's performance from identification through appraisal had both strengths and weaknesses. On the positive side, Bank staff worked closely with government to identify new institutional mechanisms for efficiently implementing the project by communities and municipios. The objective of increasing community participation was met in a highly collaborative design process, which resulted in the development of the Municipal Funds (Bank staff had initially proposed the establishment of a Community Fund to be managed by NGOs). In order to include the innovative Municipal Funds component, Bank staff worked with government to identify modifications to disbursement procedures within government in order to be able to advance money to the municipios, as well as to Bank procurement procedures to permit procurement of small works implemented by community committees. In addition, policy level discussions with Secretariat of Finance contributed to government action to increase resources to poor states through the reform of the revenue sharing formula. 39. On the negative side, the appraisal did not adequately identify the implementation difficulties mentioned above, especially: (i) lack of commitment by SEDESOL and state governments to the simplified, sector planning approaches upon which project investment selection was to depend; (ii) need for institutional strengthening of state agencies, especially  - 13 - in the use of appropriate technology for rural roads, and for the strengthened community organization for the Municipal FundslEscuela Digna program and rural water supply; and (iii) weak commitment to an active implementation quality monitoring function. Implementation 40. Due to its complexity and innovation, the project received a very high amount of supervision resources. In addition to the Task Manager based in Washington, the project was supervised by an experienced Bank staff transport engineer and water supply engineer consultant, both based in Mexico. Total amount of supervision staff weeks was 392 weeks, of which 139 were for the two engineers; 53 by environmental specialists (mainly for the Selva Lacandona subcomponent); 13 by social scientists (mainly for the Regional Funds and the Cultural Sites components); 12 by agriculturalists, and 41 by the task manager (Institutional Development Specialist). In addition, 12 staff weeks was devoted to audit (especially focusing on the Municipal Funds). The average number of staff weeks over the 3.5 years of implementation was extremely high at 98 staff weeks per year, of which at least 18 staff weeks per year was devoted to the environment and cultural preservation components and Regional Funds activities. Commitment of staff resources was highest during the first year, and decreased significantly during the 3.5 years. The state and municipal investment components (total loan disbursement US$318 million for over 28,000 subprojects) and the institutional development component (total loan disbursement US$13 million) required approximately 80 staff weeks per year over three and a half years. This amounts to an average of one staff week of supervision for every US$1.2 million in disbursement, which might be considered reasonable given the decentralized nature of the implementation of a large number (over 40,000) of small investments. G. Borrower Performance 41. The project and the Solidarity program received substantial attention by government during project implementation. SEDESOL maintained a sizable cadre of personnel in state regional offices to assist municipal officials and follow-up on progress. In support of the Solidarity program, SEDESOL initiated a number of innovative steps. One of these was the initiation of a program of "Vocales de Control y Vigilancia" to create special members (or vocales) within Solidarity Committees to act as an internal "watchdog" or auditor, resulting in the more transparent use of funds. As project implementation proceeded, SEDESOL and the project state governments, with assistance from the Bank supervision team, initiated a process of participatory field evaluation. These evaluations provided valuable data to both state and national governments on participation, and are being continued in the Second DRD project. In addition, SEDESOL and the state governments, working closely with the Bank, invested significantly in the creation and maintenance of a new monitoring system which is yielding important technical, social and geographical information about the project. Finally, the commitment of government to provide adequate budget for the project stands out at a time when many other development projects were severely underfunded. 42. Areas in which the Government could have done better, and which are to be addressed in the implementation of the Second DRD project (Loan 3790-ME), include: (i)  - 14 - conducting more frequent field supervision to address technical quality and sustainability issues; (ii) focusing technical assistance on performance problems in sector agencies and municipios; (iii) enforcing investment program guidelines and eligibility criteria; and, (iv) providing greater financing to key project elements, including the agro-ecological study of the Selva Lacandona and efforts by state governments to pilot the Puntos Criticos road rehabilitation program. H. Assessment of Outcome 43. The outcome of the project is satisfactory, principally due to the successful introduction of the Municipal Funds program, which created a new mechanism for planning and implementation of poverty-targeted rural investment programs through municipios, with high levels of beneficiary and community participation. In addition, the Selva Lacandona subcomponent made important advances in the protection of the Montes Azules Reserve by involving the local indigenous and other communities around the reserve in managing the protection program and participating in the delivery of improved basic health and education services and income-generating projects. The failure of certain aspects of the state investment program was used as a learning experience so that in the follow-on Second DRD project improved strategies were introduced, especially for rural roads. I. Future Operations 44. Follow-up on some of the most important investments and institutional development actions financed by the DRD project has been folded into the Second DRD project. In addition to what has already been mentioned above (para. 42), the second project has built on the lessons from the first by: (i) providing most of project assistance through the Municipal Funds mechanism, to ensure better targeting, participation, and cost efficiency of expenditures; (ii) limiting financing of state investment to only two sectors, water supply and roads; and (iii) within these two sectors, limiting project investment to those strategies that most clearly target the rural poor with cost effective technologies. 45. Further needs to strengthen protection for the Selva Lacandona are being addressed by an existing GEF project and by continued financing of Municipal Funds within the Second DRD project. In addition, the "Southern States Initiative" in Chiapas and Oaxaca is identifying ways to address general rural development issues in those states. One of the key elements of the Selva Lacandona program, the basic education program, is being continued by the State Education Secretariat. 46. One of the areas of failure of the DRD project was in its assistance to productive income-generating investments. New strategies to achieve this are being developed in the preparation of the Sustainable Development Project, which is targeted to poor, predominantly indigenous regions in Oaxaca.  - 15 - J. Key Lessons Learned 47. Lessons learned derive principally from the success of the Municipal Funds program, difficulties encountered project supervision, and the pilot experience with the rural road spot improvement program. They include: (a) Increased transparency, accountability, cost savings, and effective investment selection can be achieved through the simple approach used in the Municipal Funds program of allowing rural communities to actively manage their own small scale development investments. (b) Institutional complexity and diversity of sector investments, such as that contained in the DRD state investment program, makes the efficient and effective achievement of targets and institutional development goals extremely difficult, and is reliant on the high commitment of involved agencies and clear institutional mandates. (c) In a decentralized, local government and community driven project, reliable information on project investment plans may not be available or accessible. It makes more sense to focus resources on developing systems for timely post reviews of project implementation, rather than attempting to apply centralized, ex-ante controls. (d) In a decentralized project context, national government may not be willing to impose or enforce rules and guidelines for the use of funds transferred to lower level jurisdictions. Therefore, decentralized project designs should build in a system of local reviews and controls, with as much public accountability and transparency as possible. The role of national governments should be strengthened in the conduct of frequent evaluations to determine whether locally managed systems are achieving the development goals intended by the financing. (e) Based on the experience of participatory evaluations of Municipal Funds in one of the DRD states, it can be concluded that such evaluations can be done effectively with limited resources, and with full participation of different levels of government and community members. Such evaluations can have great, immediate impact as they provide credible and timely information to decision-makers. The techniques of rapid rural appraisal are well suited to these efforts.  - 16 - MEXICO DECENTRALIZATION AND REGIONAL DEVELOPMENT PROJECT (Loan 3310-ME) IMPLEMENTATION COMPLETION REPORT PART H. STATISTICAL ANNEXES Tables 1. Summary of Assessments 2. Related Bank Loans 3. Project Timetable 4. Loan Disbursements: Cumulative Estimated and Actual 5. Project Expenditures by Component and Subcomponent 6. Project Investment Expenditures by Sector: Combined State and Municipal Investment Programs 7. Project Investment Expenditures by Sector and by State: Combined State and Municipal Investment Programs 8. Project Investment and Targets by Sector: State Investment Program 9. Project Investment and Targets by Sector: Municipal Investment Program 10. Achievements of the Selva Lacandona Protection Program 11. Bank Resources: Staff Inputs 12. Project Financing by Loan, Government and Community  - 17 - MEXICO DECENTRALIZATION AND REGIONAL DEVELOPMENT PROJECT (Loan 3310-ME) PART II: STATISTICAL TABLES Table 1: Summary of Assessments Substantial Partial Negligible Not applicable A. Achievement of objectives Macroeconomic policies Sector policies Financial objectives Institutional development Physical objectives Poverty reduction Gender concerns Indigenous concerns Environmental objectives Public sector management Private sector development Likely Unlikely Uncertain B. Project sustainability Highly satisfactory Satisfactory Deficient C. Bank performance Identification Preparation assistance Appraisal Supervision D. Borrower performance Preparation Implementation Covenant compliance Operation (Not Applicable) Highly Highly satisfactory Satisfactory Unsatisfactory unsatisfactory E. Assessment of outcome  - 18 - MEXICO DECENTRALIZATION AND REGIONAL DEVELOPMENT PROJECT (Loan 3310-ME) Table 2: Related Bank Loans Year of Loan/credit title Purpose approval and Status amount Preceding and concurrent operations Integrated Rural Dev. Regionally planned rural 1981, for US$ Closed in Project, (PIDER III), development activities in seventeen 175 million. December 1988; Ln. 2043-ME poor micro-regions in four states, equiv. PCR completed. including agricultural development for smallholders, rural roads and other infrastructure. Chiapas Agricultural Integrated development for two 1985, for US$ Closed in Development Project, regions in Chiapas, financing flood 90 million December 1993, Ln. 2526-ME control and road works, on-farm equiv. PCR completed. works, soil conservation, production support services, and institutional development. Second Tropical Agricultural development of poor 1986, for US$ Closed in Agriculture Project micro-regions in the humid tropics 109 million December 1995, (PRODERITH II), Ln. through investment principally in equiv. after a 2658-ME large scale and on-farm drainage, cancellation of roads, and extension. US$ 20.7 million. ICR completed. Basic Health Case Project Improve basic health care service 1990, for US$ Implementation (PASSPA), Ln. 3272- and nutrition assistance to uninsured 180 million still underway, ME poor in the four same states as the equiv. closing 12/31/96. DRD project and the Federal District; strengthen implementation US$40.0 of sector reforms related to million has I________________________ decentralization to states. been cancelled. Primary Education Project, Improve quality and efficiency of 1991, for US$ Implementation Ln. 3407-ME primary education in the same four 250 million still underway, states as the DRD project, financing equiv. closing 6/30/97 materials, teacher training, facilities and institutional strengthening. US$37.4 million has been cancelled.  - 19 - MEXICO DECENTRALIZATION AND REGIONAL DEVELOPMENT PROJECT (Loan 3310-ME) Table 2: Related Bank Loans (Cont.) Following operation: Second Decentralization Building on the DRD project, 1994, for US$ Implementation of and Regional Development expands financing for Municipal 500 million Municipal Funds Project, Ln. 3790-ME Funds (community managed equiv. going well, but infrastructure investment), rural water supply and roads rehabilitation and rural water rural roads supply to a total of eight states (the lagging due to four DRD project states plus four lack of new ones). Increases focus begun in government DRD on community participation financing and management of investments. Sustainable Development This proposed project builds on the GOM yet to give DRD project experience, especially support for with income-generating activities in project. indigenous areas (i.e., the Indigenous Regional Funds and the Selva Lacandona program). It will finance sustainable agricultural development in indigenous poor areas of Oaxaca in particular, using approaches to technical assistance and investment designed to promote producers' participation in planning I and management of activities.  - 20 - MEXICO DECENTRALIZATION AND REGIONAL DEVELOPMENT PROJECT (Loan 3310-ME) Table 3: Project Timetable Steps in project cycle Date planned Date actual/latest estimate Identification N/A March 1988 Preparation N/A June 1988 - 1st Mission September 1988 - 2nd Mission July 1989 - 3rd Mission Appraisal February 1990 February 14, 1990 (departure) Negotiations January 1991 January 28, 1991 (start) Board presentation March 26, 1991 March 26, 1991 Signing N/A June 26, 1991 Effectiveness N/A December 17, 1991 Project completion June 30, 1995 June 30, 1995 Loan closing (latest estimate) December 31, 1995 December 31, 1995  - 21 - MEXICO DECENTRALIZATION AND REGIONAL DEVELOPMENT PROJECT (Loan 3310-ME) Table 4: Loan Disbursements: Cumulative Estimated and Actual (lS$ millions) l FY 1991 FY 1992 FY 1993 FY 1994 FY 1995 FY 1996 Appraisal 70 170 270 315 342.5 350 estunate Actual 0 75 189 258 348 350 Actual as 0 44 70 82 102 100 % of estimate Date of final disbursement: August 14, 1995  - 22 - MEXICO DECENTRALIZATION AND REGIONAL DEVELOPMENT PROJECT (Loan 3310-ME) Table 5: Project Expenditures bv Component and Subcomponent- Actual Projcct Expcnditurc Loan Expendidtur Projoct Component and Subcomponcnt Plannod % sActual (SAR) .(.-SS million) Total Inv. (USS million) (USS million) Investment Component State Investment 329 49 165 224 Educ. Infra. 104 16 52 27 Water Supply 63 9 31 22 Rural Roads 63 9 31 83 Electrification 50 8 25 19 Agriculture 32 S 16 68 Drainage 17 3 9 S Municipio Invest. (FMS, 305 46 153 50 Esc. Digna and other) Cultural Preservation Comp. 3 <1 3 2 Environment Component Institutional 2 <1 2 5 Strengthening Selva Lacandona 14 2 14 36 Institutional Development 13 2 13 18 TOTAL 666 100 350 3352 1Estimates based on information gathered during prior reviews of SOE applications. 2 Excludes USS 15 million in unallocated category on disbursement schedule.  - 23 - MEXICO DECENTRALIZATION AND REGIONAL DEVELOPMENT PROJECT (Loan 3310-ME) Table 6: Project Investment Expenditures by Sector: Combined State and Municipal Investment Pro rams . . . ... . ...... .2. ......... ..... SECTOR hvsuicw'' f (US$million) (US$ million) (US$ million) Ed. Infrastructure 104 46 54 204 32 Water Supply 63 51 114 18 Drainage 17 24 41 7 Rural Roads 63 33 _ 96 15 Electrification 50 10 60 Agriculture 32 12 - 44 7 Streets/Parks/ - 51 - 51 8 Street Lights Other - 24 - 24 4 TOTAL 329 251 54 634 100  - 24 - MEXICO DECENTRALIZATION AND REGIONAL DEVELOPMENT PROJECT (Loan 3310-ME) Table 7: Project Investment Expenditures by Sector and by State: Combined State and Municipal Investment Pr grams * SL CXTIR CitIAPAS GIJERllERO iflDAAIUJ OAXACA1 10 IAL .- ..t___ .. ._...: .._... .. ....I . ._.-..... il t.S~~~~~~IS IJS ISS l-55 I isS CY'f . . . . ....... ........... .... . . . .. .nl in ........ no ulun,,,il,,, ,,,,! .miie,,l l iA,l Ed Infrastructure 65 48 33 58 204 32 Water Supply 36 24 24 31 114 18 Drainage 9 6 15 11 41 6 Rural Roads 16 13 28 38 96 15 Electrification 21 17 14 7 59 9 Agriculture 2 7 15 20 45 7 Streets/Parks/ 16 13 11 12 51 8 Street Lights Other 3 7 11 3 24 4 TOTAL | 168 135 | 151 180 634 100  - 25 - MEXICO DECENTRALIZATION AND REGIONAL DEVELOPMENT PROJECT (Loan 3310-ME) I ;11:11 N.: .111ojtIl 111VSi1111t.iulil 111(i II Ig.1% I l S oir: S1:11CT InIveTulienit lrogrer.m ! $............ ....... .............. . ........ ..;... ..........A lEducation 32 28 6 38 104 US$15,000 4,000 Infrastructure classrooms Water Supply 27 17 9 10 63 US$100/ 630,000 beneficiary beneficiaries, l ~~~~~~~~~~~~~~~~~~~~~740 systems Drainage 5 3 6 3 17 NA NA Rural Roads 12 3 20 28 63 US$80,000/km 800 km rural constructed/re l ~~~~~~~~~~~hab. Electrification 19 14 13 3 49 NA NA Agriculture 1 5 14 13 33 NA NA TOTAL 96 70 68 95 329  - 26 - MEXICO DECENTRALIZATION AND REGIONAL DEVELOPMENT PROJECT (Loan 3310-ME) Table 9: Project Investment and Targets bv Sector: Municipal Investment Program * ., .1. I . *. ........................ *4iU I - I AX U~ ~~|I;5| 1| ; Ii MTtf IMIJ'fl i flAYCM l 'qfl|AI.sm I ' I * .1MA*. 11.. . ~~~~I .nilliura Education Infrastructure - 21 12 2 11 46 9,000 5,100 classrooms -Mun. Fund Education Infrastructure - 12 8 25 9 54 3,000 18,000 - Esc. Digna I classrooms Water Supply 9 7 15 21 52 12,000 4,500 systems Drainage 4 3 9 8 24 14,000 1,700 systems Rural Roads 4 10 8 10 32 13,000 2,500 roads, bridges or fords Electrification 2 3 1 4 10 14,000 NA Agriculture 1 2 1 7 11 10,000 NA Streets/Parks/ Street Lights 16 13 11 12 52 13,000 NA Other 3 7 11 3 24 TOTAL 72 65 83 85 305  - 27 - MEXICO DECENTRALIZATION AND REGIONAL DEVELOPMENT PROJECT (Loan 3310-ME) Table 10: Achievements of the Selva Lacandona Protection Program r . ... ...,,,,,,,, ..... ....... .. .,, ,... .... . . . . .. . . .. . . . . . ... . .. ......... ........ ..... .. .. .......... ..... . Selva Lacandona Technical Comnnitteel 1,564.70 Reserve Demarcation 2,889.86 462 boundary monuments Protection and Inspection 6,022.66 7 guard houses, 1 program Program Basic Education 17,149.05 3166 students Medical Services 2,802.57 Agricultural Investment 24,387.69 34 micro-firms, 9 nurseries 11 pig farms Other investment 9,839.28 2 rural roads, 3 storehouses, 35 schools Studies 4,725.18 6 studies Total 69,380.99 l/ This technical group assisted programn planning, conducted annual evaluations, and worked with communities to address major constrains, including land tenure issues.  - 28 - MEXICO DECENTRALIZATION AND REGIONAL DEVELOPMENT PROJECT (Loan 3310-ME) Table 11: Bank Resources: Staff Inputs ~~~~~~~~~~~~~~~~~~~~~~~~~~~.... ..................... _. ......... . . ....... ..... .... .. - :>- r............................. ......;a_..... . . . __._ ..... ;.. ....................... ..................c .... i~tai Staff Input. VWIII Time- Staff Cost, fees Roads .S ' _ .. ' ','. __ ... ...... . S_' __ .... ...................... ......... ' ................. ............................._._. ....................,. .,.'................,.......................... ............ frame- Weeks and ben. Res. Res. Engr Env. Spec. Soc. Audit Inst. Dev.' Agricul- Other Uniden- (US$) Engr. Cons. Scientist tare tified2 Identification - July '88 - 222 437,026 7 5 9 2 23 10 166 Appraisal Dec. '90 Neg. - Jan. '91 - 59 142,327 6 2 4 2 1 1 43 effectiveness Dec. '91 Supervision Jan. '92 - 342 845,433 81 58 53 13 12 41 12 16 56 June '95 Completion July '95- 6 6,124 2 March '96 Total | | 625 1,430,910 94 58 60 26 11 6 67 1 12 27 265 Supervision 98 l 2 l_ 2,472_l I / Institutional Development includes 30 staff weeks of Task Manager's time during supervision stage, from July 1992 to June 1995. 2 / Unidentified includes all of Task Manager time from identification through the first year and a half of supervision (June 1992).  - 29 - MEXICO DECENTRALIZATION AND REGIONAL DEVELOPMENT PROJECT (Loan 3310-ME) Table 12: Prect Financing by Loan, Government and Community Total Project Loan Estimated Government Financing Community' Cost Financing Financing ________ National State Municipal2 US$ million 735 350 189 122 5 69 Percent 100% 47.6% 25.7% 16.6% 0.7% 9.4% 1 / Value of community contribution is estimated, it consists priinarily of unskilled labor and local materials, as well as some cash contributions. 2/ Municipal government contribution limited to the state of Guerrero and Hidalgo (last year of project only) and to the Municipal Funds and Escuela Digna programs.  - 30- APPENDIX A Borrower's Contribution MEXICO DECENTRALIZATION AND REGIONAL DEVELOPMENT PROJECT (LOAN 3310-ME) Final Execution Report - Executive Summary Introduction 1. Within the strategy framework to combat extreme poverty proposed by the Federal Government through the Solidarity program, the Social Development Secretariat developed the Decentralization and Regional Development (DRD) Project to back up the efforts already made in the states of Chiapas, Guerrero, Hidalgo and Oaxaca, identified as those with the biggest social problems at a national level. 2. Due to the high priority for this project, the Mexican government asked for the financial support of the World Bank. The Bank loan of US$350 million was added to a national contribution of US$772 million of federal and state participation, and US$241 million from the beneficiaries, for a total project cost of US$1,363 million. 3. The loan was formalized on June 26, 1991 with the signature of loan number 3310- ME by which a installment of five years was agreed upon starting from 1990 and ending in 1995, but which due to quick implementation of the project actually concluded in 1994. 4. The request for financing to support the DRD Project originates from the relationship between the Mexican Government and the World Bank in the execution of diverse investment projects to promote rural and regional development. In this respect, the most important previous case was the financing of the Public Investment Program for Rural Development (Programa de Inversiones Puiblicas para el Desarrollo Rural). In its three different stages, this program, was characterized by the promotion of rural and multisectorial development in the less developed states of the country. Other cases included the joint efforts done through the National Plan for Rainfed Regions (Plan Nacional de Atencion a Zonas de Temporal) and the Investment Program for the Development of the Tropics (Programa de Inversiones para el Tr6pico Humedo). Note: This Executive Summary was prepared by the World Bank from SEDESOL's two volume Reporte Final de Ejecuci6n (no date).  - 31 - Project Objectives, Strategies, Policies and Goals 5. The Decentralization and Regional Development Project is part of the Solidarity Program, which has similar objectives and strategies. 6. The Project's main objective was to rise the living conditions of the poorest population of the four states, basically rural groups, through better access to basic services, improvement of productive activities, and access to better technologies. All of this was to be done in the context of improving the efficiency in the implementation and development of programs and activities from the state and municipal administrations, as well as vigorous promotion of the decentralization of programs, mostly in favor of municipalities and rural communities. As part of the specific complementary objectives of the project, it was proposed to help the preservation of the historical and archeological heritage of the states involved, and promote the strengthening of the state and municipal finances. 7. The project had the following four components: a) Infrastructure. Through this component, the social, support and productive infrastructure programs were financed, by which the poor population of the states involved would improve their living conditions. b) Restoration of Historical and Cultural Sites. In this component, activities oriented to the exploration and restoration of archeological and historical sites were promoted, selected by the National Institute of Anthropology and History and the state governments in the states of the project. c) Environmental Protection. This component developed activities for the preservation of the environment in the four states involved. Within this component, the Consideration and Development Program for the Lacandona Jungle, in the state of Chiapas, was included. d) Strengthening of State and Municipal Finances. This component was to support the development of activities to help the strengthening of state and municipal finances in order to allow the states and the municipalities a bigger financial capacity in order to strength their programs of poverty alleviation and promotion of development, and to result in less financial dependency on federal resources. 8. The investment made in the framework of the DRD during its execution represented 26.4% of the total budget of the Solidarity Program in the four states involved, equivalent to N$9,552.9 million, which confirms the high priority that the government gave to the four federal states which have the highest levels of poverty. 9. The largest expenditure of the Solidarity Program was in Hidalgo, which absorbed 31.9% of the N$1,611.2 million spent during 1990-94 in the state; in Guerrero, Oaxaca and Chiapas, the equivalent was 23% and 27% of the budgets (N$2,185.0, N$2,858.5 and N$2,898.2 million expenditures).  - 32- 10. The distribution of the resources of the project within the four states slightly favored Oaxaca with N$723.8 million (29% of total DRD investment); Chiapas had N$688.8 million; Guerrero N$598.0, and Hidalgo, N$513.2 million (27%, 24% and 20%, respectively). 11. The DRD received an important boost, specially during the first three years of its executions since the investments made in 1990 for N$230.2 million grew 47% in 1991 and in 1992 they were increased 86% in relation to the previous year. There was an annual growth rate of 3.7% for 1993 and 1994 (N$651.4 and N$675.6 million respectively). 12. This policy and the response capacity showed by the state and municipal governments and other agencies involved in the development of the project, but most of all the responsible participation of the communities, permitted the compliance with the execution schedule and the goals proposed by DRD. 13. In relation to the goals achieved in global terms, the investment made during the execution of the project was N$2,537.3 million of which 94.0% was to component A (infrastructure), 0.5% in component B (cultural sites), 3.3% in component C (environmental protection), and 2.2% in component D (strengthening state and municipal finances), achieving the completion of 85,861 public works and activities of different sorts, including those in programs of drinking water, educational infrastructure, schools, rural roads, electrification, and sewage. Financial Issues 14. The project was financed partially by the World Bank loan of US$350 million. Originally, the project had six categories of eligible expenses and financing. During the execution of the project, several concepts not included in the original loan were covered. Therefore, the Bank was asked to modify the original loan contract on November 9, 1992. The contractual modification approved by the Bank added three extra categories of expenditure. 15. The importance of this modification was that it allowed the execution of public works by the communities. This new modality allowed the Bank to apply new ways of community participation and a new altemative of implementation of social-oriented projects. These changes in implementation in the DRD were included as part of the Bank's new project's guidelines established in 1995. 16. Additionally, during the operation of the project two other categories were suggested (November 16, 1994 and December 5, 1994) for the Solidarity Municipal Funds and the Worthy School (Escuela Digna) program. Both of these programs were very successful since they allowed the decentralization of the resources to the municipalities, in addition to strengthening community participation in the activities proposed. 17. After the analysis made by the Bank of these programs, additional funds were approved, increasing an original amount for them of US$50 million to US$146 million. The Appraisal Report made by the Bank established a project cycle of six years, starting in 1991  - 33 - and finishing in 1996. However, it was reduced by almost two years since the funds were spent in about four years (1991-1994) with the last disbursement made in August of 1995. 18. The schedule of disbursements was the following: . 1990-1991 US$1,73.4 million, 4.9% more than estimated in the Appraisal Report; * 1992, US$105.8 million, 5.8% more than estimated; * 1993, US$91.8 million, 8.2% more than estimated; * 1994, US$79 million, 75.5% more than estimated. 19. The reduction in time of disbursement reflects the success of the project, and the convenience of establishing Resident Missions in different countries with decision making and payment capacity. 20. The disbursements by components were made in the following way: US$317.1 million to the Infrastructure Component; US$3.1 million to the Restoration of Cultural and HIistorical Sites Component; US$16.7 million to the Environmental Protection Component (which included US$15.5 million of the Lacandona Jungle Project; and US$13.1 million for the Institutional Development Program (which included the Strengthening of Municipal and State Finances). 21. In national currency, the US$350 million are equivalent to N$2161.7 million of which 52.5% were financed by the World Bank and 47.5% by the Mexican Government. 22. These figures do not include other kind of expenses related to the Project (such as salaries for Project Coordination Unit employees, airfare, etc.) and works that were rejected because they did not followed the Bank's design or procurement criteria. Therefore, the Mexican Government participation was about 60% and the Bank's participation, 40%. 23. The most important project activity, with 47% of disbursements, included drinking water, electricity, rural roads, schools, drainage and production projects, equivalent to N$1058.4 million. 24. The other categories in importance were those including education and Municipality Financing, which had 41.7% (N$949.8 million). The remaining 11.3% of disbursements were used to finance Restoration of Historical and Cultural Sites and Institutional Development. 25. On a state basis, most of the investments were in Chiapas, which had 30% (equivalent to N$616.8 million), followed by Oaxaca with 27.7% (N$612.9 million), Guerrero (22.7%, N$499 million), and Hidalgo (18.8%, N$423.4 million). The Central Authority accounted for the balance expenditure. 26. The distribution of the funds reflects a high level of decentralization of resources to each state where each activity was proposed, and where the payments were made. In addition, 99.9% of the contracts were awarded to regional and/or national companies. Only  - 34 - one international public bid took place (for consulting services of the institutional development component, which was awarded to a Chilean company). 27. Most of the works were contracted thorough a process of public bids which represented 42.2% of the individual works and 33.7% of the resources, and by an invitation process of at least three bidders, which represented 27.6% of the works and 14.8 of the total cost. 28. The mechanisms and procedures applied for control and management of the project allowed a better tracking of the works on the quality control, financing and bids in more than 100,000 activities and/or projects implemented during the project. For this purpose, the data base used can access expenditure information according to exchange rate, date of payment, and other important information of any of the projects or activities implemented. 29. An operational manual was created specifically for this project which contained rules of procedures with the Bank and national rules for the budget, follow-up, bidding procedures, payments and auditing. In effect, it established the terms of reference of the project. Conclusions 30. Considering the level of execution of the program and financial goals (116% compliance in relation to what was specified in the Loan Agreement), and on the other hand the conclusion of the project including full disbursement in 1994, and the little experience of the municipalities and communities in making public works, the objective of fighting poverty in the selected states were accomplished satisfactorily. 31. It would not have been possible to accomplish the objectives without proper coordination between and within government agencies at various levels, and with the monitoring for the control and progress of the project. The direction of the expenditure allowed priority care of the poorest and mostly rural population. Obviously, the activities of DRD must be strengthened, as agreed in its second phase, which must constitute the development base of social development in the states included. 32. The project has been very productive in achievements and experiences that allowed substantive advance in strategies and methodologies for the development of the following social policy driven by the Mexican Government: a) Encourage at the national level the necessity of advancing the strengthening of federalism through a policy of deeper decentralization of resources to the municipalities and the communities in the country. b) The establishment of institutional development programs for the improvement of technical and administrative capacity of the municipalities and government agencies, as a policy of the Social Development Secretariat.  - 35 - c) Development of methodologies for the support of the transfer of resources and responsibilities to the local authorities through equity and transparency schemes. 33. The following are the recommendations for the second phase of the project: a) Concentrate efforts in promoting the municipality as the main executive agency, using this strength to accomplish broad goals, unlike partial approaches which channels resources in a sectoral fashion. b) Substantial strengthening of institutional development programs of the municipalities, especially in those areas concerning training and equipment, that allow them to have more efficient management of their expenses. c) Extend the coverage of the project to other states with poverty conditions, in order to cover with similar policies, methodology and processes, most municipalities of the country.      I4. - I At j >" e;p, -irt N_1- L) -, ''''- T vs 1 I -. I. C( 'k