Document of The World Bank FOR OFFICIAL USE ONLY Report No. 21708-ME IMPLEMENTATION COMPLETION REPORT MEXICO SECOND DECENTRALIZATION AND REGIONAI, DEVELOPMENT PROJECT (Loan 3790-ME) December 29, 2000 Environmental and Socially Sustainable Development Sector Managemen t Unit Colombia, Mexico and Venezuela, Country Management Unit 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 Exchange rate as of December 2000 Currency Unit: Mexican Peso 1US$ = 9. 40 Pesos BORROWER FISCAL YEAR January 1 - December 31 ABBREVIATIONS AND ACRONYMS CDS Social Development Agreement between Federal and State Governments COPLADE State Planning Council DRD I First Decentralization and Regional Development Project DRD II Second Decentralization and Regional Development Project ITG Inter-State Technical Group MSIF Municipal Social Infrastructure Funds MPWG Municipal Presidents Working Group MTR Medium Term Review NAFIN Nacional Financiera PCU Project Coordination Units SAR Staff Appraisal Report SECODAM Ministry of the Comptroller and Administrative Development SEDESOL Ministry of Social Development SHCP Ministry of Finance SOE Statement of Expenditures SRA State Roads Agency SWA State Water Agency Vice President: Mr. David de Ferranti Country Director: Mr. Olivier Lafourcade Sector Director: Mr. John Redwood Sector Leader: Mr. Adolfo Brizzi Task Team Leader: Mr. Julio Cordoba FOR OFFICIAL USE ONLY IMPLEMENTATION COMPLETION REPORT MEXICO SECOND DECENTRALIZATION AND REGIONAL DEVELOPMENT PROJECT (Loan 3790-ME) TABLE OF CONTENTS PREFACE....................................................................................................................................... I EVALUATION SUMMARY...........................................................................................................I A. PROJECT DATA..1 B. PRINCIPAL PERFORMANCE RATINGS ........................................................................2 C. ASSESSMENT OF DEVELOPMENT OBJECTIVES AND DESIGN ...............................................2 D. ACHIEVEMENT OF OBJECTIVES AND OUTPUTS ...................................................... . .6 E. MAJOR FACTORS AFFECTING IMPLEMENTATION AND OUTCOME .................................... 14 F. SUSTAINABILITY ..................................................... s15 G. BANK AND BORROWER'S PERFORMANCE ................................. 15 H. LESSONS LEARNED ................................. 17 I. PARTNER COMMENTS ................................. 18 K. ANNEXES ................................. 21 This document has a restricted distribution and may be used by recipients only in the performance of their official duties. Its contents may not be otherwise disclosed without World Bank authorization. MEXICO SECOND DECENTRALIZATION AND REGIONAL DEVELOPMENT PROJECT (Loan 3790-ME) IMPLEMENTATION COMPLETION REPORT Preface 1. This is the Implementation Completion Report (ICR) for the Second Decentralization and Regional Development Project in Mexico, for which Loan 3790-ME in the amount of $500 million was approved on September 13, 1994, signed on October 2, 1994 and made effective on February 2, 1995. 2. The Project closed on June 30, 2000, one year after the original estimate of June 30, 1999. Moreover, in order to permit pending loan disbursements, the Bank conceded an administrative extension to October 12, 2000. The loan was almost fully disbursed (94 percent) and the remainin.g resources were cancelled ($29.3 million). 3. The ICR was prepared by Julio Cordoba (Task Manager, ESSD) and Manuel Lasaga (Consultant). Adolfo Brizzi (Sector Leader) and Anna Sant'Anna (past Task Manager, HD) reviewed the ICR. Preparation of this ICR was begun during the final Supervision/ICR Mission, which took place in September 2000. The ICR is based on materials in the project file, interviews with Bank staff involved in project preparation and implementation, interviews with the Mexican Government's staff in Mexico, and site visits to participating states and municipalities. The Borrower (NAFIN) prepared the Project Review from the Government of Mexico's perspective. Comments on the draft ICR received from NAFIN and SEDESOL have been largely incorporated into this final version. IMPLEMENTATION COMPLETION REPORT MEXICO SECOND DECENTRALIZATION AND REGIONAL DEVELOPMENT PROJECT (Loan 3790-ME) Evaluation Summary Introduction 1. The Second Decentralization and Regional Development Project (DRD II) was designed as a continuation of the predecessor DRD I. Increasing decentralization and strengthening the authority and capacity of state and local governments to manage public finance, to establish investment priorities and to provide basic infrastructure services has been a central goal of the Federal government during the past decade. During the implementation of the l)RD II project the government further defined strategies for accelerating the decentralization process, improving the targeting and quality of investment programs and strengthening state and local institutions involved in rural development and rural poverty alleviation. Project Objectives and Design 2. The objectives of this project were reasonable and consistent with the Bank's country assistance strategy. The project's decentralization objective dealt with the reinforcement of state and local institutions' capacity to manage infrastructure projects, while the regional development goal was defined in terms of investments in social and economic infrastructure in the rural areas of the eight participating states (Chiapas, Oaxaca, Guerrero, Puebla, Michoacan, Hidalgo, Veracruz, and Zacatecas). Compared to its predecessor, DRD II covered a broader geographical area while focusing on basic infrastructure and institutional development. Whereas DRD I dealt with investment programs implemented by both the states and the municipalities, DRD II's targets and focus shifted more towards municipalities in rural areas. Project Implementation and Outcome 3. The DRD II started as an investment program for rural communities with institutional strengthening of municipalities, states and the Federal government. The implernentation of the project had two distinct phases: a) from effectiveness to the end of 1997, the project supported a federal program led by SEDESOL under the umbrella of an array of programs that were intended to fight poverty, known as Ramo 26 1; b) from January 1998 to-date, the project supported the decentralized Municipal Social Infrastructure Fund (MSIF), under the umbrella of a new line- item of the budget known as Ramo 33 (Contributions to States and Municipalities). 1 A "Ramo" is both a line item in the National Budget and the set of purposes, norms and institutiDnal arrangements under which the programs financed through this line item are to be executed. Frequently, Ramo refers also to the programs themselves. ii 4. The project was able to achieve its objectives of increasing access of rural communities to basic infrastructure and supporting the government's decentralization strategy. A modification of the decentralization framework in January 1998 implied major policy changes by unexpectedly accelerating the decentralization process of the social investment program towards the municipalities, and requiring a major restructuring of the project implementation arrangements. 5. The SEDESOL post-implementation report for the first phase (1995-97) provided abundant information on the number of sub-projects, their characteristics and a basic demographic profile of the beneficiaries. It may be still too early to assess the impact on poverty reduction. However, empirical evidence shows the remarkable contribution of the FMIS to address the needs of the rural populations and facilitating access to basic services and infrastructure, thus improving their standards of living. The state governments that participated in the second phase of the project (1998-to-date) have not been able to provide as yet, a consistent and harmonized set of information. 6. The 1998 change in the federal decentralization policies altered the nature and the scope of the program being supported by the Bank, including the elimination of the involvement of SEDESOL as executing agency, and of SECODAM as administrative controller. While there were sufficient reasons to possibly proceed with the cancellation of the loan, the Governnent and the Bank decided to embark in a major restructuring effort so as to adjust the legal documents of the project to the new framework. These made it possible for the Bank to play a role in solving practical issues regarding monitoring and auditing processes during the transition process from federally mandated to state and municipal social investment programs and establish, for the first time, a direct operational dialogue at the state level. 7. Some of the merits in the outcome of the project are not quantifiable. The project helped to further solidify the success of the decentralized infrastructure investment approach through the promotion of community participation. The reliance on standardized policies and procedures reinforced fiscal discipline at the state and local levels. 8. The contribution to the institutional capacity at the state and local levels to assume greater responsibility in supporting rural development was a major achievement, although it was not as envisioned in the SAR. Sustainability 9. The sustainability of the project is judged to be likely since there is evidence that the government will continue to support the decentralization process. The decentralized approach has become an important vehicle for community involvement in infrastructure investments. A critical area for improvement is the strengthening of institutional capacity of the states and municipalities, which is key to sustainability. 10. A most important result of the project was the consolidation of its main operational arrangements into the Fiscal Coordination Law, during the 1998 reform. This major achievement ensures the sustainability of the conceptual and methodological approach of the iii DRD I and DRD II projects. It also puts a greater responsibility on state and municipal implementation as a key element to sustainability and to the delivery of quality services. Bank and Borrower Performance 11. Bank performance was up to the task except for some weaknesses in project design and supervision. In general, the project was well designed and focused on few important sub-sectors. The Bank's decision in 1998 to continue with the project despite the operational obstacles and its persistence in searching viable restructuring options proved to be an important factor of project success. 12. The commitment of the government to support regional development and decentralization initiatives goes back to the late 1980s with the creation of the Solidarity program. After a positive outcome for DRD I, the government was very supportive of this follow-up operation and committed considerable resources to it. However, the drastic shift in decentralization policies in 1998 caused a period of uncertainty during which SEDESOL lost most of its jurisdiction over the matters covered by the project, not only in project implementation but also in monitoring, a function that was not really taken over at the federal level. After 1997, SEDESOL stopped providing data about municipal spending although the law allowed it to maintain this function. However, SEDESOL produced a very valuable report on the implementation experience of the first phase (1995-1997). 13. During the transition from Ramo 26 to Ramo 33, NAFIN was heavily involved in coordinating some aspects of project implementation and monitoring, and considerably helped the states in organizing disbursement arrangements. Its participation was instrumental in the successful outcome of the project. SECODAM, while it had to relinquish its authority to supervise the use of public funds at the state and municipal level under Ramo 33, was nevertheless helpful in providing its assistance to the definition of a new audit compliance scheme. Lessons Learned 14. Several lessons are to be learned: * The follow up to an earlier successful project is likely to also succeed. In the case of DRD II the very successful Municipal Funds Program introduced during DRD I provided an important vehicle for continuation and expansion of the government's decentralization strategy. * Unforeseen changes in the external environment can lead to important modifications of the operational arrangements of a project without necessarily compromising the final outcome. Persistence, combined with a flexible approach to project implementation and restructuring, backed by strong Bank's management commitment, can result in a successful outcome. * Decentralization of resources for rural development, in the absence of careful supervision and strengthening of those entities responsible for project planning and implementation, may iv compromise the returns on community investments and result in uneven coverage and quality of the infrastructure services being delivered. * The usefulness of simple project design cannot be over-emphasized. A simple and well- focused project (though very large) has a much greater likelihood of success and of contributing to national development objectives. * The promotion of a participatory approach to the delivery of basic infrastructure will contribute to a greater understanding of the project benefits, enhanced maintenance and sustainability of the investments. . The 1998 reform demonstrated that municipalities and state governments were able to proceed with project implementation in the absence of the federal agencies previously responsible for it. However, the monitoring and evaluation function, while still carried out at the state level, was largely lost from the federal perspective due to the lack of central coordination. • As seen from the Borrower's perspective, the project contributed to gathering the following lessons: (i) the good experience of municipal authorities working together with communities; (ii) the use of formulas as a non-discretionary way of allocating funds to municipalities; (iii) the importance to adapt Bank's procedures to prevailing circumstances and to changing strategic directions; and (iv) the importance to put in place monitoring requirement in the context of complex decentralized projects with a large number of local actors. - A major lesson can be derived from the practical problems faced by the project after 1998. In this new scenario, there is no clear way to implement Bank-financed projects or programs that address sectoral issues that are being decentralized to local governments. A joint effort of the federal government and the Bank to study local legislation and institutional capacities could lead to establish new schemes and procedures to make possible international financing of relevant decentralized programs. MEXICO SECOND DECENTRALIZATION AND REGIONAL DEVELOPMENT PROJECT Loan 3790-ME IMPLEMENTATION COMPLETION REPORT A. Project Data Report Date December 29, 2000 Name Second Decentralization and L/C Number: :3790 ME Regional Development Project Country/Department Colombia, Mexico and Venezuela Region: LAC Sector/Subsector ESSD Original Revised/Actual Key Dates PCD 04/07/1993 04/07/1993 Appraisal 01/31/1994 01/31/1994 Approval 09/13/1994 09/13/1994 Effectiveness 01/03/1995 02/02/1995 MTR 03/31/1997 05/23/1997 Closing 06/30/1999 06/30/2000 Borrower NAFIN Implementing Agency SEDESOL Other Partners SHCP, States and Municipalities 2 Staff Current At Appraisal Vice President David de Ferranti Shahid Javed Burki Country Manager Olivier Lafourcade Carl Dahlman Sector Manager John Redwood Michael Baxter Team Leader of ICR Julio Cordoba -- ICR Primary Author Manuel Lasaga -- Team Leader of Appraisal -- Andrea Silverman B. Principal Performance Ratings 1. The principal performance ratings are the following: Outcome Satisfactory Sustainability Likely Institutional Development Impact Satisfactory Bank Performance Satisfactory Borrower Performance Satisfactory QAG ICR Quality at entry N/a Satisfactory Project at Risk at any time N/a Yes C. Assessment of Development Objectives and Design Background 2. The Second Decentralization and Regional Development Project (DRD II) was designed as a continuation of the predecessor DRD I. The government expressed strong interest in DRD II, which would continue and further develop the successful experience of the first project, while expanding its geographical scope. Increasing decentralization and strengthening the authority and capacity of state and local governments to manage public finance, to establish investment priorities and to provide basic infrastructure services has been a central goal of the Federal government during the past decade. During the implementation of the DRD II project, the government further defined strategies for accelerating the decentralization process, improving the targeting and quality of investment programs and strengthening state and local institutions involved in rural development and rural poverty alleviation. 2 See ICR for DRD I, Loan 33 1O-ME, Report Number 16229-ME. 3 3. DRD I underlined the importance of decentralization in order to improve the effectiveness of municipal investments in basic infrastructure. The key lessons from the earlier project pointed to some specific problems, like: (i) spread of investments over too many sectors with a large number of implementing agencies; (ii) insufficient development of the techrical capacity of SEDESOL; and (iii) lack of strategy for the implementation of the productive investment component. General Objectives 4. The DRD II started as a federal investment program for rural communities with components that also addressed institutional strengthening of municipalities, states and the federal government. It was then caught up in a major institutional change -in 1998-, which unexpectedly accelerated the process of decentralization of the social infrastructure program towards the municipalities. Although it was not anticipated at appraisal, the pro ject had two well differentiated phases: (i) from effectiveness to the end of 1997, it supported a federal program for municipal infrastructure development, closely following the original institutional design, and (ii) from 1998 to its closing in June 2000, it supported a program with the same content, but managed by local governments. These changes did not mean modifications of the project. 5. As per the Staff Appraisal Report and the Loan Agreement, the project development objective was designed as follows: To increase access of rural poor and indigenous communities to basic social and economic infrastructure and to income-generating activities, and thus alleviate poverty; and to strengthen the institutional capacities at the state and local levels to assume greater responsibility in supporting rural development. Project Components 6. In order to achieve these general objectives, the project was designed with the following five components, as defined in the SAR: (a) Municipal Investment and Institutional Development (US$737 million). This component would finance small-scale community selected and managed works, for water supply development, rural roads, school rehabilitation, productive activities, and other small infrastructure works of priority to rural communities, through SEDESOL's Muiicipal Funds and school rehabilitation (Escuela Digna) programs. The component would also develop the fiscal, administrative and technical capacity of rural and semi-rural municipalities, through studies, technical assistance, training, and support for legislative and regulatory reforms at the state level. Individual infrastructure investments in this component would not exceed US$50,000. Communities would be required to provide 20% of total project cost in cash or kind. (b) Rural Water Supplv (TJS$178 million). This component would provide potable water to small rural localities (between 500 and 5,000 inhabitants), by financing investments whose technical complexity and cost make them inappropriate or ineligible for the Municipal Funds program. The component would also improve the capacity of state water agencies and municipalities to 4 manage water supply investment and operations and maintenance efficiently and effectively. Sub-projects would be selected and contracted by state water agencies, using established investment selection criteria and standard parameters for engineering design. (c) Rural Roads Rehabilitation and Maintenance (US$133 million). This component would finance the rehabilitation and maintenance of priority sections of the rural road network, while developing the capacity of state road agencies and municipalities to do so efficiently and effectively. Rehabilitation investments would be managed by state-road agencies, selected based on an established set of criteria, and designed according to standard engineering parameters and costing on average US$7,000 per kilometer. Municipalities and communities would have an active role in implementation, with labor intensive works contracted to municipalities through inter-governmental agreements. A municipal maintenance program to cover roads rehabilitated would also be financed. (d) Income-generating Component (US$12 million). This component would provide technical assistance, training, and limited investment to develop and pilot new strategies to support income-generating projects in rural areas, by strengthening income-generating activities in the Municipal Funds program and by supporting the development of income-generating programs by state agriculture agencies in one or two pilot states. (e) SEDESOL Institutional Development and Project Coordination (US$35 million). SEDESOL's capacity to manage the project-financed programs would be strengthened. Financing would be provided for a program of institutional development technical assistance and training, contracting of technical expertise required for the supervision and review of each component, and continued development of the project information and monitoring system. Annual project audits would also be financed. 7. The objectives of the project were reasonable and consistent with the Bank's country assistance strategy. The project's decentralization objective dealt with the reinforcement of state and local institutions' capacity to manage infrastructure projects, while the regional development goal was defined in terms of investments in social and economic infrastructure in the rural areas of the eight participating states of Chiapas, Oaxaca, Puebla, Veracruz, Hidalgo, Guerrero, Michoacan and Zacatecas. 8. The design of DRD II addressed the main problems raised by the evaluation of DRD I. First, the majority of the project resources were directed to the Municipal Social Infrastructure Fund, based on its successful track record during the earlier project. The program would be reinforced through enhanced monitoring and technical assistance. Second, other investments at the state level were limited to three sub-sectors: water supply, rural roads, and income generating activities. Third, as a consequence of the reduction of the scope and nature of the investments, the number of implementing agencies was reduced in each state. 9. While DRD I dealt with investment programs implemented by both the states and municipalities, the focus of DRD II showed a clear shift towards municipalities for the implementation of basic infrastructure, especially in rural areas. Also, DRD II expanded the coverage from four to eight states, covering 1,437 municipalities. While there were fewer 5 sectors, the vast number of communities and sub-projects created a major challenge for project implementation and supervision. The Bank's financing was aimed at an investnment program of $1.1 billion, but most of the sub-projects were limited to $50,000. 10. The Municipal Social Infrastructure Fund was driven by community participation both in terms of the selection of projects and in their implementation. Community participation in investment selection effectively complemented the use of eligibility criteria and improved the quality and the relevance of the selected projects. The MSIF had been initially conceived in the DRD I as a small sub-component of that project. However, due to its recognized success, its size was more than tripled through successive increases of the budget allocations, so that DRD II effectively captured the importance of this vehicle in its design. 11. The rural water supply and road rehabilitation components were introduced in order to handle larger projects that exceeded the Municipal Fund Program criteria of $50,000. As per the SAR for DRD II, these sub-sectors were expected to maintain a similar profile as in DRD I, i.e. about 18 percent and 15 percent respectively of project's total investments. One of the innovative features of this project was the introduction of a maintenance and repair crew for rural roads (peones camineros) to be contracted by the municipalities. 12. The income-generating component was perhaps the more controversial component of the project design. It did not fit well within SEDESOL's operational structure. This component also detracted from the basic structure of the project, which was the development of social infrastructure by municipalities in rural areas. The management of this component called for a new set of policies and procedures regarding sub-project appraisals and evaluations thus adding to the already burdensome policies of the Operational Manual. This component could have been developed as a separate project. 13. The municipal institutional strengthening component was comprised mostly of technical assistance to municipalities in: i) works planning and programming; ii) the development of municipal budgeting and accounting practices; and iii) the promotion of community participation. 14. The backbone of the project's operations was the Project Operational Manual, which was developed by SEDESOL and distributed to the participating states. This was an extensive document with numerous procedures required by the Bank that went beyond the technical capacity of many municipalities. Perhaps a more simplified document may have been more effective for the implementation of the project. The Bank also included specific selection criteria for the sub-projects, which added to the administrative burden during implementation. For example, some municipalities were excluded based on economic status, and the amount of funds that could be invested in the seat of the municipality (cabecera) was also limited. While these conditions seemed logical in terms of ensuring the coverage to the neediest areas, they may have proved contradictory with an increasing focus towards decentralization and devolution of decision making at the local level. 6 Quality at entry 15. Quality at entry of the project is rated as Satisfactory. Project objectives were consistent with CAS and Government priorities. Project design complied with safeguard policies, although no specific studies were made on its likely impact over indigenous peoples, since DRD II was considered as a continuation of a former project. Quality of design was satisfactory at entry: most of the changes the project suffered afterwards were attributable to non-foreseeable circumstances. One of the few weaknesses of the design was the supervision burden of working with over 100,000 sub-projects spreading over 8 states, meaning over 1,400 municipalities and near 100,000 communities, and monitor detailed data for each of them. This proved to be excessive. Risks 16. During implementation, the project was at risk on two different occasions. In 1997, disbursements were suspended for a few months due to considerable delays in the provision of audits; this occurred at a time when monitoring reports had also a considerable lag. Things straighten up and disbursements resumed promptly. In 1998, the project suffered a hard blow when changes in the Mexican Law made impossible the operation of the existing institutional scheme, on which the project operation was based. It took two years to the Borrower and the Bank to re-design the institutional framework, restart monitoring and production of SOEs and resume disbursements. During this lapse, both the Borrower and the Bank considered the option of canceling the project, but decided to give it an opportunity to re-organize under an unprecedented scheme that could be the base for future operations. D. Achievement of Objectives and Outputs 17. This section separately discusses the implementation experience during two main periods: first under the original federal institutional environment (known and referred to as Ramo 26); and second, under the decentralization scenario established in 1998, under local governments responsibility. This last scenario is known and referred to as Ramo 33. The section concludes with an assessment of the final outcome. IMPLEMENTATION UNDER RAMO 26 The municipal investment program 18. The project became operational in February 1995, with the adoption of the Operational Manual by SEDESOL. Physical implementation of sub-projects under the Municipal Fund Program proceeded in a satisfactory manner. However, project management by SEDESOL was slow in the areas of monitoring and evaluation, state level project management, and institutional development. While these delays could be attributed to project start-up, their full justification on this ground may be questionable in view of the fact that DRD II was a follow-up project on four 7 out of the eight states covered. On the other hand, although the conceptual design of DRD II was less ambitious than that of its predecessor, the addition of four states and nearly 500 municipalities, coupled with the enhanced role given to communities as executors, added complexity to its operational design. Also, the unsuccessful efforts of SEDESOL to secure budgetary allocation to carry out the annual audits, caused delays to the ability of SEDESOL to comply with this requirement. As a consequence, in February 1997 the Bank rated the project as Unsatisfactory and suspended disbursements. The audit issue was eventually resolved and disbursements were resumed in October 1997. 19. The implementation of the Municipal Fund Program was largely successful, repeating the performance of the earlier DRD I. The quality of the information improved gradually, although there were significant lags with some of the indicators such as the proportion of investments in the seat of the municipal government (cabecera) and the rest of the municipality, where the poorer communities were located. 20. The quality of the output under the Municipal Fund Program was satisfactory as indicated by the field audits. Based on a random sample of sub-projects for the eight states, the field review found the community-based projects of the Municipal Fund Program to be more efficient than those managed by outside contractors. The costs of community-managed sub-projects were found to be 30 percent below the comparable cost of a contractor. At the same time, the quality of the sub-projects that were managed by community groups (Comites de Obras) was found to be significantly better than the ones managed by the municipal government directly. This was also due to the fact that works decided by communities were generally simple and small. Comites de Obras would have also benefited from additional guidance and training in exploring alternative sub-projects. 21. Community participation in the identification, design and implementation of Municipal Fund Program projects was mixed across the participating states. According to supervision reports, participation in the execution of works was considered good, in selection and maintenance uneven, and in the design and control aspects only marginal. SEDESOL and some states could have put more emphasis on the promotion of community participation and the fimctioning of the Comites de Obras. However, the lack of resources prevented a more ambitious approach to strengthening communities' capacities. One of the independent evaluations carried out under the project concluded that the success of the works under the Municipal Fund Program was dependent on the degree of participation by the beneficiary population. Rural water supply and rural roads rehabilitation and maintenance 22. This component was originally included to finance investments whose technical complexity and cost made them ineligible for the Municipal Fund Program. Disbursements for these components were exceedingly slow. One of the reasons for the delay was the insufficient budget allocation for this component due to the Government's strategy to put an always increasing focus on budgetary allocation to the municipal level, therefore reducing the scope and resources for the 3 See PROINFRA, Estudio de Evaluaci6n Sobre la Sostenibilidad de las Inversiones del Fondo de Desarrollo Social Municipal, February 1998. 8 state components. In 1997, 65 percent of Ramo 26 funds were destined to the Municipal Fund Program, while the remaining 35 percent were for regional development, which included the state's rural roads and water components. 23. The characteristic of the rural road component was not clearly understood by the states and municipalities. The focus was on low cost rehabilitation and maintenance of the most important road links. The concept of the pe6n caminero was considered mostly a job creation instrument rather than as a mean to carry out low cost routine maintenance. Fortunately, the spot improvement approach was also well adapted to the Municipal Fund Program, particularly the small drainage works that could be implemented by communities on rural roads. 24. In general, states did not provide any specific funding to the state water agencies (SWAs) for implementing the rural water component. In addition to the reasons mentioned above, the project's rules and strategies were not sufficiently disseminated to the participating agencies. Field reports indicated that although systems appeared to be well constructed, little assistance was provided to the communities and municipal governments to help them form water committees, set tariffs, and train system operators. 25. As a result of the limited institutional and budgetary capacity, the rural roads and water components were eventually modified and reconfigured by the municipalities to fit the organizational structure of the Municipal Fund Programs. The size of the sub-projects was reduced in line with the $50,000 Municipal Fund Program ceiling. The spot improvement concept for rural roads was compatible with the Municipal Fund Program strategy. Thus, when the decision was made to cancel these two stand-alone components, the communities were still able to benefit from these works through their incorporation into the Municipal Fund Program. Income-generating 26. Disbursements under the component were quite low, less than one percent of allocated project funds at the time of the Mid-term review. As explained in the earlier section, this component did not fit well within the structure of this project. While municipalities are eager to develop economic development programs and productive investments, they consider basic social infrastructure as a priority for the well being of their communities. Institutional Development 27. Delays in initiating a number of institutional development activities caused disappointing performance early on. Little progress was registered in the technical assistance to municipalities. One of the limitations of this component was the lack of resources allocated within this program to the implementing agency -SEDESOL- to promote, identify, and design technical support programs for hundreds of municipalities. Nevertheless, SEDESOL widely addressed municipal development through other programs. The existing complexity of budget mechanisms made not possible for SEDESOL to include these programs (belonging to the budget Ramo 20) as part of the component financed by the loan (which was classified as Ramo 26 in the budget). Delays in institutional strengthening of state water agencies, local water organizations, and state road agencies impaired their capacity to manage operations effectively. On the other hand, the 9 institutional strengthening of SEDESOL exhibited satisfactory progress. The main thrust of SEDESOL's strategy to improve its own management capacity was the hiring of long-term consultants to provide technical expertise to the central and state-level PCUs. However, at the time of mid-term review less than 2 percent of allocated funds for this component were disbursed. 28. The project successfully instituted an equitable formula for distribution of SEDESOL's resources among the states, which took into account the incidence of poverty as well as population. It was applied nationally and published in the Official Gazette every year. Loan Amendment 1 (1998) 29. After the first two years of project implementation and coinciding with the mid-term review, the Bank and the Borrower proposed several amendments to enhance the effectiveness of the project. These changes dealt basically with the operational aspects of the project. The modifications to the Loan Agreement were as follows: (i) the elimination of the rural roads and water components with the re-allocation of the funds to the Municipal Fund Program; (ii) an increase in the disbursement percentage in certain categories, from 50 percent to 65 percent, given the devaluation of the peso in December 1994, and the poverty focus of the project; (iii) an increase in the aggregate amount for civil works under direct contracting for the Municipal Fund Program; and (iv) minor adjustments to definitions of various components of the project. IMPLEMENTATION UNDER RAMO 33 Municipal Social Infrastructure Fund 30. Without prior waming, in December 1997, the Federal Congress included the Municipal Funds among the seven national programs for which the law of Fiscal Coordination was amended. This changed the rules of the game regarding the transfer and management of funds to the states and municipalities, creating a new concept -that of Aportaciones- and a mechanism known as Ramo 33. According to this mechanism, budget funds alloc;ated to cover the program's actions would be transferred directly to the states and municipalities, with no direct intervention by federal agencies4. Actually, transferred funds cease to be federal and are to be managed under local jurisdictions. 31. With respect to the operational structure of the DRD II project, the new system of funds transfers under Ramo 33 created several problems. First, SEDESOL could no longer act as implementing agency. Second, the national Operational Manual, which was used by this project, was no longer valid, since it was up to the states to develop their own policies and procedures. And third, sub-project supervision and audits were delegated to the states, without the need to 4In Mexico, all expenses of Bank supported projects are paid out of already budgeted funds. Disbursements from the loans are received afterwards by the Treasury. The budget law allocates resources regard less of the sources of finance. Executing agencies never can spend against the loan proceeds if not previously budgeted by law. This is the "non-additionality" principle. Additionally, under the Mexican Constitution, local governments cannot borrow from international sources, which prevents the Bank from directly financing prograns that axe not federally supported. 10 report back to the Federal government, thus posing a major problem in terms of information gathering and compliance with the Bank's auditing policies. 32. It took about a year to understand the implications of Ramo 33. In the meantime, it created a vacuum in the project, and the generation of SOEs was discontinued during 1998-1999, stopping new disbursements. During all that period, however, the program continued entirely under domestic financing. The new law had also created some ambiguities regarding the assignment of responsibilities between the states and municipalities, as well as the role of SEDESOL, which was mostly relegated to some institutional development of these entities. The Bank held workshops during 1998 in the eight states in order to discuss the new fiscal structure and its implementation. At that time the Bank faced two options regarding the implementation of the project: either find a way to adapt to the new structure, or cancel the project. In October 1998. the Borrower requested the Bank to consider the possibility of restructuring the project. This coincided with a new reform to the Fiscal Coordination Law that cleared up some of the ambiguities from the original version of the legislation. On this basis the Bank made the decision to pursue an acceptable mechanism that would allow the project to continue. 33. These changes did not affect the objectives of the project, but rather the strategy to achieve those goals. It "de facto" established new clients at the state level whose views of the project varied. The responsibility shifted to the states as regulators and supervisors and to the municipalities as independent entities with the authority to make their own decisions regarding the application of funds to individual social infrastructure projects. The relevance of the Bank's role was now in strengthening states' and municipalities' capacities needed for implementation of monitoring and auditing under conditions of Ramo 33, as well as in providing support to discuss options and national and external experiences that may contribute to a better understanding and management of decentralization processes. 34. SEDESOL's loss of status as the implementing agency did not necessarily eliminate its role within the project. It was still responsible for the implementation of institutional development programs in the states and municipalities. At the same time, the new law assigned SEDESOL the responsibility to compile information on the use of the Social Municipal Infrastructure Fund under Ramo 33. Nevertheless, SEDESOL stopped providing the detailed information from the states and municipalities to the Bank. The Federal government as well did not insist on maintaining a flow of information on social infrastructure expenditures. The Bank and NAFIN thus had to negotiate with state governments the development of an appropriate mechanism for each individual state in order to gather the necessary information. 35. Continuation of the project hinged on the adoption by the states of adequate policies and procedures regarding investment expenditures in social infrastructure. In January 1999, the Bank and NAFIN, which by then had assumed some of the functions of the implementing agency, began to meet with each of the eight participating states to determine their interest as well as the compliance of their Operational Manual with Bank's requirements. The new scheme was first tested in the state of Zacatecas, which had established satisfactory procedures and was able to prepare the SOEs in conformance with Bank guidelines. By then the Bank and the Government were able to reach an agreement with three other states: Guerrero, Puebla, and Veracruz. 11 36. The restructuring of the Loan Agreement was finalized on November 29, 1999. During the first phase of the project under Ramo 26, the Bank had disbursed $355.9 million, the remaining $144 million would be disbursed under this new arrangement. Basic Infrastructure 37. During this phase of implementation, the project focused exclusively on the MSIF, since the other components had been cancelled by the April 1998 amendment. Some of the states had signed an agreement with the respective municipalities, which established priorities in the investment program such as water, electricity and roads. Disbursements were made in 2000, based on actual expenditures accumulated during 1998-1999. Each of the participating states presented the corresponding SOEs to NAFIN, which then processed them for final presentation to the Bank. 38. Community participation continued to be a critical factor in the success of the MSIF although performance varied. In most of the municipalities the use of Comites de Obras for the contracting and supervision of works was continued and strengthened while in others communities were basically providing the unskilled labor. In general, despite initial fears that the exit of SEDESOL from the program would have affected implementatiorn performance, the MSIF continued to perform strongly with enhanced accountability at the local level and good participation. However, the weak management capacity of many small municipalities appeared to be a main source of concern. Institutional Development 39. The restructuring of the project involved, in effect, the redefiniticn of the project implementation arrangements with four, instead of one implementing agency. Since many municipalities exhibited significant weaknesses in project management, the focus of this second phase of the project was on assisting the states and municipalities with the management of the MSIF. The continuation of this project allowed Bank staff to initiate a productive dialogue with the states regarding the new institutional requirements of Ramo 33. 40. The Ramo 33 included a Municipal Institutional Development program with an optional amount for each municipality of up to 2 percent of its MSIF, which led to an excessive atomization of resources. During both the pre-1998 period and the Ramo '33 period, resources for institutional development were clearly insufficient. This is even more striking if compared with the considerable resources transferred to municipalities for infrastructure development. 41. The Bank's diagnosis of the institutional weaknesses of the states in administering the new Ramo 33 program led to the negotiation of an IDF grant for $500,000, that would assist in the strengthening of the states' oversight of public administration. The purpose of the grant was to strengthen the states' auditing and monitoring capacity through the offices of the General Accounting Office (Contadurias Mayores de Hacienda) and the States' C'omptroller's Office 12 (Contraloria del Estado). 5 The goal was to provide those agencies with the knowledge and skills needed for upgrading their auditing standards to levels compatible with international practices and to improve their control and monitoring systems in line with the new responsibilities of the States within the context of Mexico's decentralization framework. The IDF grant was approved in August 1999 and became effective in December 1999. 42. According to the new Fiscal Coordination Law, the distribution of funds for social infrastructure to the states and the municipalities was determined by a formula that weighed more favorably those areas with greater incidence of poverty. The distribution arnong states was based on a poverty index comprised of the following variables: household income per capita, average educational level per household, adequacy of housing, availability of adequate sewer infrastructure, and availability of electricity and fuel for cooking. The states were in turn to use a similar formula in determining the allocation across municipalities. Loan Amendment 11 (1999) 43. The loan amendment basically restructured the operational characteristics of the project. The objectives, components and description of the project remained essentially the same, except that each of the states would now assume the role of the implementing agency. The institutional development component focused on municipal strengthening and community participation; therefore the sub-component dealing with the institutional strengthening of SEDESOL was eliminated. PROCUREMENT POLICIES, PROJECT INFORMATION, AND QUALITY CONTROL 44. Procurement procedures were maintained wisely simple by keeping a ceiling of $50,000 on all works to be financed by the Bank. For works contracted through the Comites de Obras, this permitted to carry out direct contracting through community participation, which permitted the project to be executed relatively smoothly. The ceiling was slightly relaxed to $100,000 in 1997 for some water supply contracts, which required a larger economy of scale and, an inter- municipal approach. In general, community participation in each sub-project ensured a higher quality of the output and better sustainability. For this reason no ex-ante feasibility studies needed to be reviewed by the Bank, the identification and appraisal process was mostly delegated to the municipalities. The relevant sub-project documentation was contained in the so- called "technical file" (expediente tecnico) which field supervision generally found to be satisfactory. However, cases of weaknesses in the technical design of the project such as the engineering drawings or the lack thereof, implementation problems or unfinished works were also detected by the audits. 45. The project's complexity in terms of scope and large volume of sub-projects required a highly efficient implementation system comprising of: (i) a computerized data base that could provide timely information; (ii) intensive supervision; (iii) frequent reports; (iv) independent evaluations by consultants; and (v) periodic meetings of participating agencies. The system was 5 The General Accounting Office is in charge of auditing the use of all public funds at any level of govermment, and reports directly to the corresponding State Congress. The State Comptroller's office resides within the executive branch of each government. 13 to a large extent put in place by SEDESOL; however, its complexity caused significant delays in submitting the progress reports and the audits. SEDESOL, eventually produced an extensive and very detailed report on pro ect implementation during 1995-1997, after the completion of its participation in the project. 46. The basic instrument for disbursements was the SOE. The $50,000 ceiling permitted to filter the most complex or problematic works. The Bank assigned a staff member full-time to inspect and review these documents. It was through his analysis that many discrepancies were identified and passed on to the Borrower for further action. The information system that was established was not very user-friendly with respect to disbursement processes, whic;h required constant screening to ensure compliance with the Operational Manual. 47. From the perspective of the sub-projects, the restructured operation after 1998 did not allow the type of detailed monitoring that had been achieved under SEDESOL, mostly because the very high cost of monitoring had no visible rewards for the local governmenlts. 48. Under Ramo 26, SECODAM had been responsible for auditing the project. By the end of 1998, the decision was made that the responsibility for auditing the project was with the state legislatures. For the purposes of this project, the Bank accepted that the states would contract the audits with the assistance of SECODAM. The cost of these audits became an issue, since the Ramo 33 did not earmark money for additional audits as requested by the -Bank, and the audits normally carried out by the states would not be considered satisfactory to the Bank. This was eventually resolved with the federal government taking over the cost of carrying out the Bank- required audits. ACHIEVEMENT OF PROJECT OUTCOME 49. The project was able to achieve its objectives of increasing access by irural communities to basic infrastructure and assisting with the government's decentralization strategies. The SEDESOL's post-implementation report provided abundant information on the number of sub- projects, their characteristics and a basic demographic profile of the beneficiaries. The available information is not sufficient at this time to establish the impact of the project on poverty alleviation, and the benefits of a number of the infrastructure improvements as well as institutional development initiatives may still take some time to be realized. However, empirical evidence shows the remarkable contribution of the FMIS to address the needs of the rural populations and facilitating access to basic services and infrastructure, thus improving their standards of living. The decision to continue with the project, when there were reasonable grounds for its cancellation, made it possible for the Bank to play an influential role during the transition from federally-mandated programs to state and municipally-led social investment programs. 50. Some of the merits in the outcome of the project are not quantifiable. T]he project helped to further solidify the success of the Municipal Fund Program through the promotion of community participation. The reliance on standardized policies and procedures reinforced fiscal discipline at 6 SEDESOL, DRD II, Reporte de Conclusi6n del Proyecto 1995 - 1997. 14 the state and local levels. There was good targeting of sub-projects. Field supervision missions reported an improvement in the quality of basic services. 51. With respect to poverty targeting, this project helped to institutionalize the use of formulas in the allocation of investment resources, thus eliminating discretional and ad-hoc approaches. The new Fiscal Coordination Law established explicit formulas and required that all states and municipalities publish the information related to the distribution of funds as well as their destination in the Official Register. The operational infrastructure supported by this project under the Ramo 26, subsequently served as a valuable input for the states and municipalities as they began to build their own institutional framework under Ramo 33. On the other hand, the development of institutional capacity at the municipal level could have been enhanced. The relatively short duration of municipal governments (a mayor is elected for a non-renewable term of three years) combined with the limited resources earmarked for this purpose, hindered the possibilities to make a major contribution to institutional development issues at the local level, while acknowledging that the capacity of municipalities has improved considerably compared to where it was ten years ago. E. Major Factors Affecting Implementation and Outcome 52. Problems of Incentives and Non-Additionality in Bank Lending to Mexico. The public sector in Mexico operates under the policy that lending from the Bank -as well as other multilateral agencies- does not provide additional resources to the government's expenditure program. Based on the annual target of budget deficit, the government determines the expenditure budget and the Bank's financing are considered part of the foreign currency financing of the overall deficit. In effect, the direct impact of the Bank's financing is on the government's macro-economic management of the fiscal deficit in terms of the domestic and foreign currency financing components. Therefore, the Bank's loan did not provide additional resources to states and municipalities. Because of non-additionality of financial resources, Bank projects lack incentives. Some states expected that the Bank project would bring them additional resources over and above what they had been assigned by the Federal budget. One of the problems with the rural road and water components, that eventually led to their closure, was that the Federal government had not assigned sufficient resources to the state budgets to cover the cost of these sub-projects. 53. Factors outside the control of government. During 1999, heavy flooding in some of the participating states delayed implementation and resulted in a reallocation of MSIF resources towards the affected communities; however, this did not have a negative impact on implementation. 54. Factors generally subject to government control. Perhaps the key factor from the government's perspective was a strong commitment to decentralization. Beginning with DRD I, the government was convinced that the process of regional development was most effective when managed by the communities themselves. There was some resistance early on in the project, which mostly reflected the traditional centralized nature of public administration. 15 However, the policy shift in December 1997, which was triggered by the legislature, clearly indicated that the decentralization trend was going to strengthen. 55. Factors generally subject to implementing agency control. There vwere two kinds of implementing agencies during this project. SEDESOL served as implementing agency during the first part of the project, and when Ramo 33 was approved, this responsibility was assigned to the participating States with the help of NAFIN. SEDESOL's staff was highly qualified, but there were significant delays in providing basic project information. However, it has to be recognized that without a regular flow of information from the municipalities, SEDESOL's tasks were made more difficult to execute and it was not possible to disburse on schedule. With the change to Ramo 33, the states became the implementing agencies, although institutional weakness in terms of management effectiveness and lack of experience in monitoring investment programs was a cause of concern. However, although these entities had not been considered for this role at the time of project preparation, the new policy called for a much stronger involvement of sub-national governments, and therefore highlighted the importance to assist them in implementing this new role. F. Sustainability 56. The sustainability of the project is judged to be likely, as there is evidence that points to the continuation of the decentralization process. The Municipal Fund Program zand its successor, the MSIF, have both been successful and they have become an important vehicle for community involvement in infrastructure investments, which is key to sustainability. 57. The consolidation of part of the features contained in the Operational MAanual as articles of the law of Fiscal Coordination is a major achievement regarding the sustainability of the conceptual approach of the project. 58. A critical area for improvement is the institutional capacity of the states and municipalities. The states welcomed the Bank's support in the implementation of their regional development strategies and the opportunity to establish a direct dialogue with the institution. The Bank's initiatives resulting from a continuous dialogue with the States are expected. to expand, building on the achievements of this project. A Bank follow-up operation in this area will help to sustain the benefits accumulated so far. The IDF grant to strengthen the monitoring and auditing capabilities of the states is a step in the right direction. G. Bank and Borrower's Performance Bank 59. Bank performance was up to the task except for some weaknesses iri project design and supervision. In general, the project was well designed in focusing on fewer sectors. The Bank's decision to continue with the project despite the operational obstacles and its persistence in 16 searching for viable restructuring options proved to be an important factor in its success. The Bank was thus able to support institutional development and, to some extent, support the re- organization of the flow of information from the states under the Social Municipal Infrastructure Fund of Ramo 33, a function which was not sustained by the federal government. Lending 60. The identification and preparation phase was facilitated by the success of the earlier project. Nevertheless, Bank staff worked diligently in terms of the sector work and extensive project analysis contained in the SAR. On the other hand, the appraisal was not as accurate in its evaluation of the commitment by SEDESOL to the provision of project information and to the institutional development program. Part of the problems were attributed to differences of opinion between the Bank and SEDESOL regarding some of the implementation criteria such as the ranking of municipalities according to their so-called "development needs," which is not normally used by the government in targeting the rural poor. Supervision 61. The complexity of the project made on-going supervision a critical factor. The decentralization of the Mexico Department made it possible to keep abreast of project developments and to meet frequently with the implementing agency. A possible setback was the turnover in project task managers. From appraisal to closing, the project had four task managers. This created some bumps along the way, particularly for the Borrower, as each change in task manager led to a temporary loss of momentum in terms of the implementation. The frequency of supervision missions was not an issue in the sense that the project was managed out of the regional office. Nevertheless, more intensive and frequent field supervisions by the Bank may have been advisable. Overall Bank Performance 62. After consideration of the above-mentioned factors and of the outcome of the project, the overall performnance of the Bank is rated satisfactory. Borrower 63. The commitment of the government to support regional development and decentralization initiatives went back to the late 1980s with the creation of the Solidarity program. After a positive outcome for DRD I, the government was very supportive of this follow-up operation and supported its preparation fully. However, it was the notable shift in decentralization policies that presented significant challenges to the project. The implementing agency was legally impeded to respond to the 1998 changes in operational aspects of the project, although it eventually produced a valuable report on the implementation experience. Government and Implementing Agency Performance 64. Overall, government performance was satisfactory. However, performance of the implementing agency was not satisfactory during the initial years of implementation, because of 17 extensive delays in reporting and auditing compliance. This was largely due to the difficulty of organizing an harmonized and regular flow of information from thousands of communities and municipalities to the central level. As stated above, the government's strong commitment to decentralization was instrumental in maintaining the project's momentum and contributing to its positive outcome. During the transition to Ramo 33, NAFIN was heavily involved in monitoring and supervision, in view of the absence of SEDESOL as implementing agency. Its participation was instrumental in the successful restructuring of the project. However, collection and coordination of the flow of information from the States remained an issue. SECODAM, while not empowered under Ramo 33 to supervise the use of public funds by states and municipalities, was nevertheless helpful in providing technical assistance with field audits. Overall Borrower Performance 65. After consideration of the above-mentioned factors and of the outcome of the project, the overall performance of the Borrower is rated satisfactory. H. Lessons Learned 66. The principal lessons derived from the outcome of this project include the following: v lThe follow up to an earlier successful project is likely to also succeed. In the case of DRD II the very successful Municipal Funds Program introduced during D)RD I provided an important vehicle for continuation and expansion of the government's decentralization strategy. * Unforeseen changes in the external environment can lead to important modifications of the operational arrangements of a project without necessarily compromising the final outcome. Persistence, combined with a flexible approach to project implementation and restructuring, backed by strong Bank's management commitment, can result in a successful outcome. * Decentralization of resources for rural development, in the absence of careful supervision and strengthening of those entities responsible for project planning and implementation, may compromise the returns on community investments and result in uneven coverage and quality of the infrastructure services being delivered. * The usefulness of simple project design cannot be over-emphasized. A simple and well- focused project (though very large) has a much greater likelihood of success and of contributing to national development objectives. * The promotion of a participatory approach to the delivery of basic infrastructure will contribute to a greater understanding of the project benefits, enhancecl maintenance and sustainability of the investments. The 1998 reform demonstrated that municipalities and state governments were able to proceed with project implementation in the absence of the federal agencies previously responsible for it. However, the monitoring and evaluation function, while 18 still carried out at the state level, was largely lost from the federal perspective due to the lack of central coordination. * The possibility for the Bank to continue supporting and financing decentralized programs under Ramo 33 at the State and Municipal level will require to consider and adjust to the legal, economic and financial framework that govern the relationship between the three levels of government. * As seen from the Borrower perspective, the project contributed to gathering the following lessons: (i) the good experience of municipal authorities working together with communities; (ii) the use of formulas as a non-discretionary way of allocating funds to municipalities; (iii) the importance to adapt Bank's procedures to prevailing circumstances and to changing strategic directions; and (iv) the importance to consider monitoring requirement in the context of complex decentralized projects with a large number of local actors. * A major lesson can be derived from the practical problems faced by the project after 1998. In this new scenario, there is no clear way to implement Bank-financed projects or programs that address sectoral issues that are being decentralized to local governments. A joint effort of the federal government and the Bank to study local legislation and institutional capacities could lead to establish new schemes and procedures to make possible international financing of relevant decentralized programs. I. Partner Comments7 SEDESOL's comments 67. The Decentralization and Regional Development Project (DRD II) addressed social development issues defined in the National Development Plan 1995-2000, emphasizing the strengthening of the new federalism process promoted by the Federal Executive Branch. 68. In this context, the project accentuated the participation of municipalities and communities, leading to an increased demand for transfers of Ramo 26 resources to be invested directly by Municipalities according to the priorities defined by the Municipal Development Councils. 69. This strategy developed a symbiosis: on one hand, with the DRD II the government at its federal, state and municipal levels promoted the social and microregional development and, on the other hand, the population obtained tangible benefits by diminishing infrastructure deficits to improve standards of life, while fostering local economy by using regional materials and local labor force in the execution of the works, thus mitigating the ancestral poverty that characterizes the social environment of the project areas. 70. Permanent efforts of coordination by SEDESOL and state governments were required to achieve these objectives, mainly in the situation created by the change of mayors every three years, which implies to carry out training actions practically on a permanent basis. 7 The full Completion Report prepared by Sedesol is available upon request. Here reported is a translation of the Executive Summary. 19 71. The great motivation and participation of rural communities in the implementation of the project is worth noting. This was made possible thanks to the credibility on the three levels of government. It resulted in responding to the requests for social and productive works in thousands of small and disperse localities. 72. It is equally important to mention that the involvement of the municipalities -and mainly of the beneficiaries-, in the execution of the project, has resulted in lowering time and costs of execution of the works. For this reason, the DRD II through the MSIF presents an innovative aspect in the consolidation of the "federalism," which the Mexican goverrinent and the World Bank can implement in other development projects. 73. For all the above-mentioned, it follows that the success of the DRD II was based on targeting investments for the population with the highest degree of poverty, who generally inhabits the rural and indigenous areas. In fact, 61 % of the investment was targeted on. the municipalities considered of high and very high marginality, and a strategy was implemented to give preferential attention to those localities outside municipal government seats ("cabeceras"). 74. Also, it is worth to point out that the monitoring of investments and follow-up processes allowed to successfully achieve the disbursement goals, in spite of the devaluation of the Mexican peso during 1995. 75. In this same order of ideas, it should be noted that although only 295.2 million dollars were disbursed [under SEDESOL's execution], equivalent to 590%/0 of the 500 million dollars loan, this was due to the ending of the contract 3790-ME with the World Bank, caused by the transfer, in 1998, of the resources formerly allocated to Ramo 26 to the Municipal Social Infrastructure Fund of to the newly created Ramo 338. 76. The success of the execution of the DRD II according to the results presented in the report, is corroborated with the conclusions and recommendations mentioned in the Aide Memoire of the mid term evaluation mission of the World Bank. Among those conclusions it is worth to menlion the following ones: * "a) The organizational arrangement implemented at municipal level, centered in the Council of Municipal Development and the local Community Committees; b) the opportunity given to poor peasants to manage investments that improve their conditions of life; c) the social pressure felt by the local committees to deliver complete works at the smallest cost to their communities; d) the demonstration effect arnong localities in the same Municipality and among municipalities, that leads to seek better performance levels; e) the honor that participants of local committees experience when feeling that they are contributing to improve the opportunities of life of the future generations in their own communities; and f) the fact that the payments made on the basis of actual concrete progress in the execution of the works." * "The diffusion of the MSIF should be promoted with special emphasis in the most isolated localities. For doing that, among other actions, it is advisable to reinforce the training of the promoters of COPLADES and SEDESOL." x Actually, under local governments execution (referred to as Ramo 33), the loan kept on disbursing, up to a total of 470 millions, before closing on 2000. 20 *"The mission considers as positive the performance of the project in general, although problems still exist. Note that due to the high number and the great dispersion of works, it becomes difficult to register accurately specific indicators of execution and performance of the investments." * "It should also be noted that although the project is definitively contributing to the reduction of poverty, it is difficult to exactly specify the measure of this impact. However, it could be enough to say that when the project provides basic infrastructure, it is satisfying high- priority necessities of the poor with regard to water, electric power and schools, among others." * "The prioritized goals for the future of the project, in its current consolidation phase, could be summarized in six main actions: a) to consolidate the social participation of the communities and of the beneficiaries; b) to promote and support the execution inside the MSIF of inter- municipal works such as rural roads, to facilitate the development of the rural production through better connections with markets and better conditions of rural transport; c) to improve the monitoring system and to develop easier indicators for both the project and Ramo 26; d) to consider the inclusion in the project of urban municipalities that have necessities of basic services, and to foster social participation in the development of the MSIF in the cities; e) to look for mechanisms to guarantee an efficient maintenance and operation of the infrastructure; and f) to develop and implement a systematic program to support the institutional development of municipalities and communities." 21 K. Annexes Annex 1 Key Performance Indicators a) Output Indicators: number of sub-projects Indicator Projected Ln SAR/PAD Aciua/Latest Estimate* MUNICIPAL FUNDS (Total) 40,000 106,427 * Escuela Digna 20,000 10,991 * Potable water 1,300 12,344 * Rural roads 1,560 10,203 * Support to production 9,056 * Schools 17,431 * Electrification 7,027 * Sewage systems 7,457 * Pavement 7,322 * Urbanization 8,661 * Housing 5,219 * Health centers 1,768 * Sports infrastructure 2,371 * Community services 3,347 * Others 3,230 POTABLE WATER _ * Construction 35 * Enhancement 13 * Rehabilitation 14 RURAL ROADS * Rehabilitation 1,388 164 * Maintenance 36 * Data for 1995-1997 only. From 1998 on, there are no statistics available 22 b) Disbursement: geo2raphical indicators DISBURSEMENTS Millions of Dollars STATE 1994-97 1998-2000 Total CHIAPAS 63.8 63.8 GUERRERO 36.2 13.2 49.4 HIDALGO 43.6 43.6 MICHOACAN 32.2 32.2 OAXACA 34.9 34.9 PUEBLA 49.9 41.6 91.5 VERACRUZ 78.7 29.7 108.4 ZACATECAS 13.7 30.2 43.9 SEDESOL 2.8 2.8 TOTAL 355.9 114.7 470.6 c) Disbursement: component indicators DISBURSEMENTS Million dollars Original Actual Municipal Funds 308.5 465.5 Roads and water 146.5 2.4 Income generation 9.0 0.0 SEDESOL 23.0 2.8 Unallocated 13.0 29.3 .__ __ __ __ __ __ __ _ 500.0 500.0 23 d) Disbursements: by categories (thousands of dollars) Category Original Real Municipal Funds * Works under direct contracting 236,500 463,589.7 * Construction materials 10,000 0 * Goods (different from 8,000 577.3 construction materials), * Technical assistance and training 27,000 1,318.8 Rural Roads and Water * Works under direct contracting 40,000 2,409.1 * Works other than direct 90,000 contracting ___ * Goods 2,000 * Engineering services 6,500 * Technical assistance and training 8,000 Income zeneration + Works 3,000 0 * Goods 1,000 0 * Technical assistance and training 5,000 1.1 Institutional Development SEDESOL * Goods 2,000 75.3 * Consultants (long term) 13,000 1,699.5 * Technical assistance and training 5,000 635.1 * Audits 3,000 357.2 Non allocated/Unused 13,000 29.3 Total 500,000 500,000 24 Annex 2 Proiect Costs and Financing a) Project Costs by Components (in US$ million equivalent) Project Component Appraisal Estimate Actual/Latest Percentage of Estimate Aporaisal Municipal Funds 336.5 465.5 138.3% Rural Roads and Water 142.0 2.4 1.7% Income generation 5.5 0.0 0.0% Institutional Development 3.0 2.8 93.3% SEDESOL Unallocated/Unused 13.0 29.3 225.4% b) Proiect Costs by Procurement Arrangements (In US$ million equivalent) Expenditure Procurement Method Procurement Method Cateories ADraisa Estimate ActlLatest Estimate ICB NCB Direct NBF Total ICB NCB Direct NBF Total I Works 80 882 962 466.0 2 Goods 16 20 36 0.6 3 Services 97 97 4.1 4 Misc. _ Total 96 999 1095 470.7 c) Disbursements: timing Year (calendar) Programmed (millions) Actual (millions) 1995 80.0 112.9 1996 91.0 58.8 1997 116.0 85.3 1998 147.0 98.9 1999 66.0 0.0 2000 0.0 114.7 Total 500.0 470.7 25 d) Proiect Financing by Component (in US$ million equivalent) Component Appraisal estimate Actual/Latest Percenitage of . _ _ aiEt r nateasal Bak 1=lvm C Ok GX_ w 0V- mnt mi n S m.en Municipal Funds 281.5 152.8 465.5 231.6 165.37% 151.6% Rural Roads and Water 138.5 136.7 2.4 2.4 1.74% 1.8% Income Generation 4.0 3.1 0.0 0.0 0.00% 0.0% Institutional 63.0 1.7 2.8 0.2 4.39% 9.3% Development . Miscellaneous/Unused 13.0 29.3 225.38% TOTAL 500.0 294.3 500.0 234.2 . 79.6% 26 Annex 3. Cost Benefit Analysis NOT APPLICABLE 27 Annex 4. Bank Inputs a) Site Visits '. 5=e of ProjM t/erycles a r - R irnpkpientation Development _______________________________ ProR ess O jec ive Identification/Preparation _ _ _ Appraisal/Negotiations Supervision Feb-95 4 (Institutional Development, JHS HS Participation, 2 Engineers) Dec-95 3 (Portfolio Manager, 2 Engineers) S S May-97 6 (Mission Leader, 2 Engineers, lJ U Financial, Environmental, Social) Feb-98 6 (Mission Leader, Sector Leader, U S Sector Manager, 2 Engineers, Financial) Aug-98 6 (Mission Leader, Sector Manager, UI S Engineer, Financial, Procurement, Monitoring) Dec-99 7 (Mission Leader, Sector Leader, S S Sector Manager, Monitoring, Procurement, Financial., Institutional) Jun-00 5 (Mission Leader, Sector Leader, S S Monitoring, Financial, Institutional) ICR Jun-00 5 (Mission Leader, Sector Leader, Monitoring, Financial, Institutional) Sep-00 3 (Mission Leader, Writer of ICR, Institutional) b) Staff StgAt of Project Cycle /Ac/ t Estimage Staff weeks . Identification/Pre aration 87.7 196.4 ApRsae a 54.6 131.5 Supervision 295.2 680.9 ICR 13.0 29.8 Total 450.5 1,038.6 28 Annex 5. Ratings for Achievement of Objectives /Outputs by Components Objectivesotuts Hi Hatfac Not Unsatsfactory : h Not :________________ Satsar0X:yL :___ _:__ Unsatisfactory uAlicable Macro Poiis ______X Sector Policies X Physical X Financial _____ X Institutonal X Development Envimronmental _____X Perty Reduction X Gender _ X QLhLr LSR__ _ _i_f) x Private Sector X Development _ Public Setor X Manement Other (spcif_) 29 Annex 6. Ratings of Bank and Borrower Performance i U aSatisfactor Unsatisfactorv Bank Pe_fornance _ _ Lending X _ Supervision _ X Overal X Borrower Perfomance: _ Praration X _ Govern tx IMplementation Performance Oyerall X _- 30 Annex 7. List of Supporting Documents 1. Staff Appraisal Report: Report 13032-ME August 1994 2. Loan and Guarantee Agreements: October 1994 Amendment I April 1998 Amendment II November 1999 3. Supervision Missions Aide Memoires 4. Project Status Reports # 1 to 14 5. SEDESOL: Evaluaci6n sobre la sostenibilidad de las Inversiones del Fondo de Desarrollo Social Municipal. April 1998 6. SEDESOL: Reporte de Conclusi6n del Proyecto 1995-1997