Participation in the South Asia region's - project portfolio : towards defining conditions for success

THE WORLD BANK Internal Discussion Paper SOUTH ASIA REGION Report No. IDP-146 Participation in the South Asia Region's Project Portfolio: Towards Defining Conditions for Success Barbara Parker August 1994 Office of the Chief Economist SASVP The views presented here are those of the author, and they should not be interpreted as reflecting those of the World Bank. SOUTH ASIA REGIONAL SERIES Title Author Date Oriqinator IDP1 11 How Composition of Public Expenditure Affects Competitiveness: The Case of Bangladesh K. Matin March 1992 P. Mitra (80419) IDP1 21 Labor Retrenchment and Redundancy Compensation in State Owned Enterprises: The Case of Sri Lanka A. Fiszbein December 1992 G. Nankani (84641) IDP126 Some Guidelines for the Appraisal of Large Projects W. Jack February 1993 A. Estache (81442) IDP109 Reforming Higher Secondary Education in in South Asia: The Case of Nepal H. Abadzi May 1993 H. Abadzi (80375) IDP1 27 Some Lessons for South Asia from Developing Country Experience with Foreign Direct Investment M. Fry June 1993 A. Estache (81442) IDP129 Quasi-Fiscal Deficits: Latin American Lessons for South Asia C. A. Rodriguez August 1993 A. Estache (81442) IDP131 The Impact of Rural Infrastructure on Rural Poverty: Lessons for South Asia E. Goldstein June 1993 G. Nankani (84641) IDP1 34 Infrastructure and Industrial Policy in South Asia: Achieving the Transition to a New Regulatory Environment P. Seabright June 1993 A. Estache (81442) IDP142 Regional Trading Arrangements and Beyond Exploring Some Options for South Asia Theory, Empirics and Policy T.N. Srinivasan July 1994 G. Nankani (84641) IDP146 Participation in the South Asia Region's Project Portfolio: Towards Defining Conditions for Success B. Parker August 1994 G. Nankani (84641) PARTICIPATION IN THE SOUTH ASIA REGION'S PROJECT PORTFOLIO: TOWARDS DEFINING CONDITIONS FOR SUCCESS Barbara Parker August 1994 TABLE OF CONTENTS Page FOREWORD ACKNOWLEDGMENTS Executive Summary ....................................... i Introduction: W hy This Review? ............................. 1 Goals and Methods of theReview ............................. 2 D efinitions .. . .. .. .. .. . . .. .. .. .. . . .. .. ... . . . . . . . .. 3 W hy Participation ....................................... 6 Stylized Propositions ..................................... 8 Incentives to Participation .............................. 9 Organizing Effectively ............................... 14 Optimum Scale and Replicability ......................... 24 Achieving Local Control .............................. 28 The Role of the WorldBank ............................ 31 Suggestions for Further Implementation .......................... 35 Annex 1 - List of Projects Discussed ........................... 37 List of Text Boxes Box 1 Some LessonsLearned ............................ ii Box 2 Pakistan: Integrated Hill Farming Project Adds a Continuing Stream of Benefits ............................... 10 Box 3 Landless Contracting Societies in the Bangladesh Coastal Embankment Rehabilitation Project Pilot .................. 11 -2- Box 4 India's West Bengal Forestry Project: When Local Government Lacks Accountability .............................. 13 Box 5 Women Volunteers Support India's Fifth Population Project ...... 15 Box 6 Sri Lanka's Community Water Supply and Sanitation Project Finding NGOs Appropriate Partners .................... 15 Box 7 Farmers Identify Assistance Needs in Nepal's Agricultural Extension I Project .......................... ... 17 Box 8 Participant Observation Strengthens Pakistan's Second Karachi Water Supply and Sanitation Project ............... 18 Box 9 Waranbandi Water Distribution System .................. 19 Box 10 The Problem of the Landless in Pakistan's On-Farm Water Management Project .............................. 20 Box 11 India's National Social Forestry Project: Does Participation Always Help the Poorest?........................... 21 Box 12 India's Tamil Nadu Integrated Nutrition Project Achieves Broad-Based Community Support ...................... 22 Box 13 India's Uttar Pradesh Basic Education Project Will Involve LocalW omen ............................. 23 Box 14 Tiered NGO Training in India's Uttar Pradesh Sodic Land Reclamation Project .............................. 25 Box 15 Users Groups and Subgroups in Nepal's Bhairawa Lumbini GroundwaterilProject ............................ 26 Box 16 Bangladesh General Education Project is Urged to Scale Un . . . . . . 27 Box 17 Decentralization Suceeds in Sri Lanka's Second Rural DevelopmentProject.............................. 29 Box 18 Partner Organizations in Sri Lanka's Janasaviya Trust Fund . . . . . . 30 Box 19 Participation Requires Up-Front Investment ................ 34 FOREWORD I am pleased to introduce this review of issues and experience from the South Asia Region's project portfolio in the areas of participation and poverty reduction. This review follows from a call issued at the October 1993 Workshop On Poverty Reduction in South Asia: Promoting Participation by the Poor for experimentation, research and further analytic work on the subject of participation. A stocktaking of the Region's experience to date was deemed to be an essential prerequisite. Its role has been to document task managers' efforts to utilize and improve participatory approaches in their work, to identify the mechanisms that have worked well under various conditions, to call. attention to the problems and constraints that are commonly encountered in the course of implementing these approaches, and to reveal the gaps in our current understanding of particiration. The review concludes with some suggested directions for further experimentation that could fill many of thes- gaps. I hope this review -- particularly its specific illustrations drawn from actual practice -- will serve as a source of information to task managers and will also provide food for thought to a larger audience within the Bank. It provides many possible practical and operational implications for the design and implementation of projects in tne South Asia Region. D. Joseph Wood Vice President South Asia Region The World Bank ACKNOWLEDGMENTS Gratitude is extended to Gobind Nankani, Robert Drysdale, Valerie Kozel, Ann Duncan, James Socknat, Renee Dankerlin, John Clark, Bhuvan Bhatnagar, Paul Francis, Veena Siddharth, Robert Wade and Paul Seabright for comments on earlier drafts of this paper or for contributions to its 13 "stylized propositions." Many task managers in the South Asia Region contributed their time and insights. In particular, I would like to thank Neil Boyle, David Klaus, A. Graeme Lee and Mae Chu Chang. Research assistance was provided by Salahuddin Admad, and Elizabeth Miller and Mary Schmidt generously shared notes from their own research. PARTICIPATION IN THE SOUTH ASIA REGION'S PROJECT PORTFOLIO: TOWARDS DEFINING CONDITIONS FOR SUCCESS EXECUTIVE SUMMMAY In the South Asia Region, there is a growing interest in the potential of broader participation in project design and execution to enhance the performance of poverty reduction programs supported by the World Bank. Recently, the Region has conducted a review of lessons learned during the implementation of its own participatory interventions. WHY THIS REVIEW? In November, 1992, the Independent South Asian Commission on Poverty Alleviation (ISACPA), which had been commissioned by the South Asian Association for Regional Cooperation (SAARC) six months before, released its report entitled Meeting the Challenge. The report sparked an interest in initiating a dialogue on poverty between the Bank and development professionals in South Asia. As an expression of this interest, the Bank's South Asia Region and SAARC collaborated on the organization of a workshop entitled Poverty Reduction in South Asia: Promoting Participation by the Poor. The workshop was held in Annapolis, MD on October 12-15, 1993, and its goal was to review current poverty reduction strategies in the region with a special emphasis on the role of participation by the poor in those strategies. The participants included the SAARC Secretary General, former members of ISACPA, representatives of several multilateral organizations and NGOs active in the region, and World Bank staff members. Prominent among the workshop's findings and conclusions was the conviction that bolder experimentation with community participation and other innovative approaches was needed in order to close existing gaps in understanding and to point the way toward effective use of promising models in future poverty reduction initiatives. If these experiments are to be useful, however, they should be based on current knowledge and experience. It was decided, therefore, to undertake a stocktaking of the Region's experience with participatory development. The purpose of this review was to derive and disseminate key lessons learned from the Region's past, ongoing and planned participatory work, to identify gaps in information or understanding, and to propose further experimentation that would fill these gaps wherever possible. INSIGHTS AND QUESTIONS Although most of the Region's projects that include participation are now in an early stage of implementation, a fund of experience and knowledge has begun to accumulate. A series of 13 "stylized propositions" was collected from many sources in the Bank as a means of encapsulating lessons learned to date; and task managers and other staff members were asked to respond to them. Although many of the propositions were readily accepted as verified by experience, others could not be definitively analyzed and assessed using the evidence currently available. The debate surrounding these propositions underlines the need for experimentation that can definitively answer the questions they raise. - ii - BOX 1: SOME OP'IE LESSONS LEARNED * LocaL governmentO pan worX effectively at the community level. When .NGOs -assist development agenc.ies to reach down to small communities, they should: be. encouriagpd to collaborate eith local governments insteaa of oreating parallel and redundant programs and services. NGOs can, however, -be higly effegtive in linking tne poor with. existing ser-ices, monitoring services to ensure that they are responsive to the poor, and calling attention to the needs and preferences of poor communities. * Local elites and the nonpoor can have a positive role in promoting -community participation in initiatives that benefit all classes equally, but when-competing with-the poor for scarce public benefits, they may use their power to capture more than their share. The wealthy can be omitted from many types of beieftts without excU%ionary rules, however, if goods useful only to the poor'are provided, or if users' associations have control over distribution. * Participatory projects are expensive in time.and funding, and there is a dgAger of dissipating resources by spreading them -too thin. To conduct a .:valid test of. the value 'of. community participation, available resources shbuld be.cpncentrated in a small number of "flagship" projects that are .closely monitored to establish the conditions necessary for success. * .The World Bank can and abould work.at the "wholesale" and uretail" (or .local) levels"simultaneously. At the wholesale level, the Bank can develop coherent programs that are-based on what it has learned at the retail level aoplt the requirements for effective use of innovative and participatory .approaches. The review concludes with a series of suggestions for further experimentation that could begin to resolve some of these unanswered questions. Among them are the following: * Accountability: Innovative project designs wherein community-level service users could monitor project progress, track the movement of funds, register complaints and initiate corrective measures should be developed and tested. Users' groups such as parent-teacher associations are one possible model. * Indigenous Organizations: The usefulness of indigenous community groups as an organizational tool could be tested through a pilot study in which the outcome of community-based projects that function through pre-existing groups is compared with the results of an identical project using community groups organized by outsiders. In addition, a categorization and mapping of traditional organizational forms in South Asia could offer a useful resource for task managers. * Single-Purpose/Multi-Purpose Groups: The impact of adding functions to the workload of successful single-purpose groups could be tracked and evaluated - Lii - in order to examine the question of whether community groups can profitably undertake several functions simultaneously. * Scaling Up/Replication: Controlled experiments should be initiated to test the results of planned replications of promising programs in their new settings. Projects attempting to transfer methods from effective NGO programs to public sector systems should also be evaluated carefully. * Evaluation of Participation: Careful evaluation of the value added by participation to the achievement of objectives of selected pilot projects should be undertaken by comparing project performance in participating communities with performance in control communities without participation. The go. 1 of these studies would be 1) to establish whether community participation is or is not associated with improved outcomes, and 2) to establish protocols for best practice in monitoring and impact evaluation of participatory projects. PARTICIPATION IN THE SOUTH ASIA REGION'S PROJECT PORTFOLIO: TOWARDS DEFINING CONDITIONS FOR SUCCESS INTRODUCTION: WHY THIS RNVIEW? Participation by the poor in poverty reduction initiatives commands a great deal of attention in the World Bank of the 1990s. Voices both within and outside the Bank have called on Bank staff to find an active role for those affected by its programs in the planning, implementation and maintenance of those programs. Participation has been proposed as a remedy for weaknesses of past projects and initiatives, particularly for those sectors in which local opposition or disinterest has been a problem. Even in sectors in which no specific problem with local cooperation has been documented, many proponents of the participatory approach suggest that the best results will not be achieved unless project beneficiaries are involved in planning and decision-making, regardless of the type of project or its specific goals. Development practitioners of the South Asian countries are particularly well-disposed toward community participation. The region is host to some of the world's best known grassroots initiatives, such as Bangladesh's Grameen Bank, the Bangladesh Rural Advancement Committee (BRAC), India's Self-Employed Women's Association (SEWA), Pakistan's Aga Khan Rural Support Project (AKRSP) and Sri Lanka's Sarvodaya. There is no doubt that each of them has been successful in generating interest and active involvement on the part of the poor. The replicability and independent sustainability of these initiatives have not been clearly established. Nevertheless, given this promising beginning and the positive climate of opinion in the region, it is not surprising that a significant number of projects currently in the planning stage in the World Bank's South Asia Region include a community participation component. From the World Bank perspective, however, participation is still a relatively new approach to poverty reduction, and quantitative evidence of its value and effectiveness in various contexts is just beginning to accumulate The best ways to employ participation in either the preparation or implementation of projects in the South Asia Region are not yet definitively identified, and although it has already become evident to the World Bank's task managers and their supervisors that obstacles are commonly encountered in the process of introducing community-level participation, the exact nature of these obstacles and the most effective means of surmounting them are far from clear. A body of experience has begun to accumulate nevertheless. A few pioneer projects that enlisted community members in carrying out project interventions are now complete, and a great many more are underway. Although a majority of the latter are too new to have provided any "lessons learned" for the record, there is a growing fund of experience on which to draw in framing future initiatives. Most of the current wisdom on participation exists in anecdotal form, in informal ,1\ The final .report of the World Bank's Learning Group on Participatory Development will soon be available for distribution. Within the South Asia Region in particular, the SA2 (India) and SA3 (Pakistan/Sri Lanka/Afghanistan/Maldives) Departments are reviewing their own experiences with participatory approaches. - 2 - memos, in scattqred supervision reports or in the minds and memories of task managers. Task managers do not normally share insights and information on this topic unless it is done informally and on an ad hoc basis. As a result, opportunities for productive cross-fertilization are often lost, and the Region may fail to build on what has already been learned about participation by individual departments and divisions when designing innovative projects. This paper will examine the experience of the World Bank's South Asia Region with participatory and community-based approaches to poverty reduction. Its purpose is to derive and disseminate key lessons learned from the region's past, ongoing and planned participatory work, to identify gaps in information or understanding, and to propose further experimentation that would fill these gaps wherever possible. Goals and Methods of the Review Several efforts are now underway to systematize the collection of information on participation and community mobilization within the World Bank. Among them is the Participation Sourcebook, which will help Bank staff to understand participation and its methods, to identify partner organizations who can assist in carrying out these methods, and to interact more confidently with borrowers who are undertaking participatory programs. The Sourcebook is still in preparation at this time. Although the South Asia Region is coordi-.iting with contributors to the Sourcebook, it was felt that, because of the unique contributions that the South Asian countries have made to the practice of community participation and the consequent importance of this approach to the region, a separate region-specific assessment was necessary and would be a useful complement. The high level of interest in and commitment to community participation displayed by many South Asian development practitioners, together with the region's broad tapestry of innovative NGO-led programs using participation by the poor, provide the backdrop for this review of participatory projects in the South Asia region supported by the World Bank. The exercise can b^ said to have had its genesis on October 15, 1993, during the closing session of an informal workshop entitled "Poverty Reduction in South Asia: Promoting Participation of the Poor" which was held by the World Bank in cooperation with the South Asian Association for Regional Cooperation (SAARC) in Annapolis, MD. The workshop had been inspired by a report entitled Meeting the Challenge that was produced by the SAARC-sponsored Independent South Asian Commission on Poverty Alleviation (ISACPA). The report described the weaknesses of past South Asian strategies for poverty alleviation, documented successful grassroots programs in the region, and offered a series of recommendations for improving the results of poverty alleviation efforts. Prominent among these was the proposition that the productive capabilities of the poor could be dramatically enhanced if they were encouraged to participate actively in the development process. In order to open such a dialogue, to contribute to the search for an effective regional strategy for poverty reduction, and to develop an agenda for future actions, policies and research, the workshop was developed. A series of recommendations was produced. Among them was the call for a bolder and more experimental approach to development that includes experimentation with projects involving community action and the participation of those affected by interventions in the planning and implementation of poverty programs. Because the World Bank is the largest of the development agencies, its role in this - 3 - process is expected to be a significant one. Before it can determine the magnitude of that role and the specific actions it entails, however, the Bank's South Asia Region must first take stock of experience to date. Since participatory approaches are new to the Bank, most of the projects discussed are in planning or the early stages of implementation. Very few task managers were able to offer concrete results. Nevertheless, preliminary efforts to organize local groups, to motivate government counterparts to cooperate with these groups, or to interest communities in undertaking responsibility for planning and implementing project components have been initiated in a significant numher of projects. These efforts are showing positive early results in many cases, but sometimes only after clearing unexpected hurdles and meeting unexpected challenges. The information included in this stocktaking has been gathered through document review and informal discussions over a period of several weeks with task managers, division chiefs and others working in the South Asia region. Staff members who have been active contributors to the Bank's ongoing discussion on the role of participation in poverty reduction programs were asked to contribute their ideas. The final document, therefore, represents the thoughts and insights of a wide variety of persons both within and outside the World Bank. Definitions A rich and varied dialogue on participation has developed within the World Bank during the 1990s. A wide range of opinion has been aired during that period. It is becoming increasingly clear, however, that "participation" does not necessarily carry the same meaning for all who use the term. Before proceeding further, it would be worthwhile to clarify the meaning of terms that are commonly used when discussing participation, and to establish common areas of reference. In recognition of the problem of floating definitions, the Participation Sourcebook will devote a section to establishing a common vocabulary of participation. In the meantime, the most useful definition of participation may be one proposed by a recent review of the World Bank's participatory development experience. According to this document, participation is "a process through which stakeholders influence and share control over development initiatives, decisions and resources which affect them.2 Although this definition of participation is widely accepted within the Bank, there is no generally accepted definition of successful participation. The success of a development initiative is judged on the basis of the extent to which it has achieved its stated goals and objectives. The question of whether "participation" has been a critical component of this success is more complicated. If participation is not central to its planned implementation strategy, a project may meet its objectives even if attempts at generating community involvement have failed. On the other hand, participation may be active and vigorous, yet the project may fail to meet its objectives due to unexpected external constraints. In such a case, the community might still have gained an important benefit if the organizational impetus catalyzed by the project carries over into other constructive activities. While acknowledging 2\ The World Bank and Participation: Report of the Learning Group on Participatory Development. - 4 - that it is difficult to calculate the exact value added by participation, this review will assume that successful participation has been achieved by a Bank- supported project when the project itself is judged successful and active involvement of beneficiaries has been a factor in this achievement. Providing a precise definition of a community is also problematic. Although the most appropridte definition may be among the alternatives offered by Webster, "a group of people with a < >mmon characteristic or interest living together within a larger society," a community in everyday parlance commonly refers to a group of people who are co-resident in a specific (usually small) area. What is unclear is who should establish the boundaries around a given community and define who is to.be included or excluded. These boundaries may be established in two ways: 1. Self-Identification: There is a pre-existing sense of community among many groups of people in South Asia that can, at least in theory, be channelled into productive development-related activities. Many of these groups have emerged from the need to manage a commonly-held public good. However, in South Asia as in other regions self-identified communities may be organized on the bases of caste or language group and may exclude those persons (women or scheduled castes) ir. greatest need of organized assistance. This point is discussed in greater depth below. 2. External Identification: A government or development agency may define a community as the population of interest with respect to a specific purpose. A development organization can, as suggested by the formal definition above, identify those individuals within a defined geographical area who share a common interest or problem as a community. An urban slum's or village's poor households or women could then be treated as a community, bypassing local elites or landlords, when mobilizing groups for income-generating activities. An obvious difficulty with this approach is in convincing the members of an externally defined group of the fact that they do indeed share a commonality of interests and that they should therefore work cooperatively on joint endeavors. Those persons who have been bypassed or excluded may also constitute an obstacle if they are in any way opposed to the group or its objectives. The sustainability over time of such externally defined groups, and the conditions under which they might take root and persist, are questions that might be resolved by the results of ongoing experiments. It may also be useful to establish clear distinctions between the types of organizations that are active in the mobilization of communities for development- related tasks. As with communities, boundary lines between sectors and organizational types are not simple to draw in many cases. The following may illustrate the complexities involved in identifying and distinguishing the various organizational players in community participation: 1) The Private Sector/Community Group Intersection: Some World Bank staff members have pointed out that, in many important ways, a community-based organization (CBO) with development goals is very similar to a private sector organization such as a professional association or a corporation. The 3\ Webster's Ninth New Collegiate Dictionary. 1991. Springfield, Massachusetts, U.S.A.: Merriam-Webster, Inc. - 5 - resemblance is clearest in the case of a community group that establishes a common resource or utility such as an irrigation system, and then maintains and manages it for the benefit or profit of group members. For example, under the On-Farm Water Management Prolect in Pakistan, legal recognition was achieved for irrigation water users groups. Although they have not yet been granted the right to sell canal water except in tribal areas, they distribute the benefits of their common resource (water) among themselves and are able to undertake some types of contracts and agreements. Under a project now in planning, the Baluchistan Community Irrigation and Agriculture Prolect, farmers' water users' groups will be shown how to write contracts and trained in marketing so that they may eventually hire their own marketing consultants and contractors. 2) The NGO/Community Grout Intersection- NGOs are often the moving force in organizing community groups to undertake income-generating or other development activities. Not all community groups have been formed in response to an external impetus, however. Some have arisen spontaneously over time or in response to a particular problem, such as the need to manage a local resource, to mediate property disputes between neighbors or to apportion routine public maintenance tasks. When these indigenous groups are induced to take on broader development- related functions, they may begin to resemble the small, localized "grassroots" nongovernmental organizations. NGOs are expected to expend voluntary and "altruistic" labor or resources for the benefit of others. Community groups, by contrast, exist for the purpose of benefiting their members, and in the view of some economists, are more likely to be sustainable if they can limit the beneficial impact of their activities to those who actually participate . Nevertheless, a community group occasionally extends the scope of its activities outside the home community. When it does so, or when members systematically begin to expend time, effort and funds for the benefit of non-members, then the contrast between the community group and an NGO becomes less clear. 3) The NGO/Private Sector Intersection: There are many ways of categorizing NGOs -- for example, current discussions within the Bank draw a distinction between NGOs that are advocacy oriented and those that are service oriented. Nevertheless, all types of NGOs are thought to have at least one feature in common, which is that they are non-profit organizations. The distinction between an NGO and a private sector organization is normally sought in the motives of staff members. Private sector organizations are profit-oriented, while NGOs are usually humanitarian and "voluntary." As a successful NGO matures, however, its early reliance on the idealism of untrained volunteers may prove to be unsustainable, or these volunteers may be inadequate to the task of providing high-quality technical assistance to poor communities or local governments. The organization is then faced with the necessity of offering higher salaries to attract experienced and qualified staff members. Funding requirements rise accordingly. In the end, complaints are sometimes heard to the effect that the former NGO has come to resemble a consulting firm or a government contractor in the private sector. 4\ Philip R. Gerson. Popular Participation in Economic Theory and Practice. Human Resources Development and Operations Policy Working Paper #18. The World Bank. December, 1993. In summary, the dialogue on methods and models for participatory development includes terms fo. organizational types that may not be as distinct as they initially appear. The category "community groups" can blend seamlessly into other types of organizations, and it should not be assumed that community groups organized or utilized by projects of different types are alike in their aims, possibilities or limitations. By the same token, NGOs may be similar to community groups and suffer from their limitations when they are very small, highly localized and lacking in skills or training. Alternatively, NGOs may be large, well-funded, and powerful enough to be perceived by government as a political or economic force requiring regulatory control. NGOs of this type may be difficult to distinguish from political parties. Because of these ambiguities, task managers would be well advised to avoid simplistic standardization and to design their approaches to NGOs and community organizations on a case-by-case basis depending on the requirements of their sector and the existing organizational potential of the borrower countrys Why Participation? The ultimate goal of the World Bank's IDA is to assist member countries in the reduction of poverty. The lending program, as well as the program of economic and sector work, are the principal means of assisting with the attainment of this goal. One way that these programs achieve their results is through a series of poverty-reducing interventions, such as the projects discussed below, that can be judged successful in so far as they measurably improve living standards and life quality indicators in the borrowing countries. It is proposed that participation by the poor in the choice and management of these interventions is a valuable mechanism for achieving project success, for reasons which the following will make clear. In order to define the role of participation in achieving project success, it may be worthwhile to consider the contrasting or counterfactual condition -- nonparticipation. From the point of view of the poor, a nonparticipatory intervention could be defined as one that is selected, planned, implemented and maintained by hired organizational functionaries (in government, development agencies, NGOs or the private sector) who will never use the good or service in question and whose personal well-being will not be affected by it. Several implications may be drawn from this definition. First of all, an intervention developed and executed by a hired functionary may be handicapped by an array of conflicting agendas that are similar in type to what economic theory has termed the "principal-agent problem." The essence of the principal agent or agency problem is in the fact that the owner or employer (principal) has no guarantee that the employee (agent) will perform the mandated task in the manner specified. With respect to the intended beneficiaries of the intervention, there are additional implications: 1. The objectives of these functionaries, or agents, in establishing the intervention may be incongruent with the objectives of its users. For example, although both the agent and the user may have a common interest in the successful .\ As experience accumulates, however, it should become possible to identify prototypical design elements suitable to various types of projects and categories of players. - 7 - installation of the intervention, the agent's interest is usually based on the need to satisfy a scope of work, please supervisors, and achieve success in a career. The agent, therefore, may be inclined to ignore negative externalities such as environmental damage or social deterioration that are outside his or her formal area of responsibility. Of course, the reward structure of the implementing organization will influence the degree to which the agents' and users' interests coincide. 2. The users may possess a fund of information about local factors affecting the outcome of the project that the agent lacks. For example, an agent who plans and builds a school or clinic during the dry season may not be aware that, during the rainy season, the chosen site is a point of concentration for malarial mosquitos. Participation by local residents in planning and decision-making makes it possible to tap into their understanding of local environmental, health and cultural constraints and opportunities. 3. The agent's approach to the problem is always defined by pre-established organizational policies and procedures to a greater or lesser extent. That approach may, therefore, be too rigid to be adaptable to external changes that may occur during implementation. Localized control by users permits a quick response to changes in local conditions. Centralized, top-down decision-making systems react more slowly, because choices must be referred to higher-order organizational units at the state or national level. Particularly for a country such as Bangladesh that experiences frequent natural disasters, rapid response capability is a significant advantage in poverty reduction initiatives. To summarize, two basic assumptions underlie this review's approach to community-level participation. They are: 1. Individuals act to further their interests, and those individuals with the greatest interest in achieving poverty reducing outcomes are those whose well-being is most directly affected by those outcomes -- in other words, the poor. 2. When decision-making is centralized and top down, the interests of individuals most directly affected by the project's outcome may become secondary to other bureaucratic, organizational and, sometimes, rent- seeking considerations. Decentralized development that involves users at the grassroots level in planning and management is important to maintaining a focus on the ultimate objectives of reducing poverty and improving well-being in the project area. Without such a focus, peripheral concerns can become central and efforts can easily be misdirected. Once we have accepted participation by the poor as one of the ways in which this focus can be maintained, however, many questions remain. Experience to date has revealed that participation is not simple to implement and may not succeed without careful handling of key problematic issues that are commonly encountered in participatory interventions. These challenges, and the techniques which World 6\ See also Paul Seabright's "Managing Local Commons: Theoretical Issues in Incentive Design." Journal of Economig Perspectives. 7, #1, 1993. p.132. - 8 - Bank staff members in the region are using to meet them, are condensed in the 13 Stylized Propositions that follow. Stylized Propositions A series of "stylized propositions" are suggested as a useful device for organizing current thought and experience in the South Asia Region on the issues surrounding participation in poverty reduction. The propositions below were culled from various sources within the World Bank, and reflect the thought and experience of a number of staff members in the South Asia Region and in other departments. They are not formal hypotheses, since there are no specific plans at present to put any of them to the test. They are not offered as proven propositions; rather, they are presented as a starting point from which to begin examining the experience of task managers and others who have operational experience with participatory approaches. The opening set of hypotheses is as follows: Incentives to Participation: 1. Participation will not be active, effective or sustainable without strong individual incentives. These incentives must outweigh the expected private transaction and opportunity costs of sustained participation in project or program activities. 2. By taking advantage of economies of scale, individuals working together can achieve more than they can working alone. Common interests must be made to outweigh diverse individual interests, however, if collective action is to be sustained. 3. Community participation will not be sustainable unless accountability to beneficiaries is built into the project process. Organizing Effectively: 4. Community participation cannot be employed in uniform fashion across all sectors and in all countries. It is necessary to plan participatory components with the specific sectoral, economic and socio-cultural conditions of the project and target community in mind. 5. Local knowledge of local constraints and conditions should be taken into account before introducing new technologies or organizational forms. 6. Community participation in project implementation and maintenance will be more sustainable if culturally appropriate and familiar models of organization are employed. If possible, indigenous organizations and/or locally recruited staff should be used as a starting point. 7. Homogeneous populations are more likely to form cohesive groups than are sociologically or economically heterogeneous populations. 8. Single-function community groups are easier to organize and sustain, and are more effective in reaching their goals, than are multi-function groups in most - 9 - circumstances. In other words, a project that undertakes only irrigation-related activities will be more successful than one which combines irrigation, reforestation and income-generation initiatives. Optimal Scale and Replication: 9. Community development groups should be kept small in order to maximize effectiveness and to control free riders; but to achieve maximum impact in terms of national development goals, they can be coordinated or federated at higher levels of organization. 10. In order to replicate or scale up successful programs, it is important to identify 1) the internal factors and external conditions that determine the success of the program, and 2) which of the projects' "elements of success" are found in many settings and which are unique characteristics of the original project site. Achieving Local Control: 11. Local authority over resources that are to be used at the local level is important to the success of community-level organizations and to government decentralization/devolution initiatives. Local authority to collect revenues or fees is a useful approach to generating local resources and building commitment. 12. Clearly specified rules and allocation of rights and responsibilities are critical to the smooth functioning of levels of government, to government/NGO collaborative efforts, and to the success of community groups. The Role of the World Bank: 13. Although it has been suggested that the World Bank's initiatives are too large to be tailored to each local community in which it operates, and thus the Bank is suited only to working at the "wholesale" level -- to providing resources and guidance for putting participatory initiatives into motion -- it may also be necessary for the Bank to work at the "retail," or local, level (in collaboration with NGO/CBOs) in order to acquire the necessary knowledge for informed decision- making. Incentives to Participation 1. Participation will not be active, effective or sustainable without strong individual incentives. These incentives nust outweigh the private transaction and opportunity costs of sustained participation in project and program activities. As economic theory predicts and experience demonstrates, projects that are designed with built-in incentives, particularly when these incentives are monetary, are more likely to generate participation than those without. This is particularly true of projects that call for participation of the poor in organized activity. Participation in organized activity implies a time - 10 - commitment that is sometimes substantial, and the poor are likely to have less leisure time than the more affluent. Community members are not always willing to continue expending time in altruistic work indefinitely, unless they receive a clearly identified personal benefit in return for their participation. Closely related to this point is the discovery, voiced by some Bank staff members, that it is easier to form lasting associations around mutually advantageous public assets than around nebulous concepts of public benefit. For this reason, a community forestry group may continue to operate indefinitely if their efforts are clearly leading to improvements in the condition of the local forest and the resources that all members can rece' .3 from it. A voluntary health education group may be more difficult to sustain unless there is a focal asset, such as a clinic, that might be gained or maintained by continued activity on the part of the group. Task managers have noted, however, that an introductory incentive at the time of project initiation, or a long-term return expected in the distant future, are not enough to guarantee sustainability. Instead, it is important to build a continuing stream of benefits into the design of a project. For example: During imnplementat ion of this resource cosvtion and envretal. pesevationx pject, it was leard that com ite tha were oranized~ to *undtake reforestation of commuunal lands were nt disposed to mitain and protect the saplings during the five year matuiatio>n pe4iod. Th~1erer expected from the mature treesa were too ditant to.serve a.s a force for' was ecesary to add underrwth plants to the. platations that could be harvested by community m~embiers on ~a otating basis whil.e the nxew 1ress w'ere* maturing. Us~e of these tod.der planta was available only to those who pariciated in~ pro tecting and maita~.inn the grwing forest. Rigid definitions of sustainable participation should be avoided, however. A community or resource users' group may decide that joint or rotating labor responsibilities are an unnecessary burden on its members over the long run. Particularly if many of the group members have been able to escape poverty, they may decide that it is in their comparative advantage to hire a less affluent group member or outsider to perform routine maintenance on a continuing basis. Although the form of participation and the functions group members perform may evolve in this scenario, the local management goal is realized nevertheless. As long as the community of users continue to manage and provide for the execution 7\ Robert Wade, in Villagie Republics: Economic Conditions for Collective Action in South India, has suggested that a fear of losing assets or benefits may be a stronger motive for participation and mutual cooperation than a desire to acquire additional benefits in the future. - 11 - of this responsibility, sustainable participation can be said to have been achieved. 2. By taking advantage of economies of scale, individuals working together can achieve more than they can working alone. Common interests must be made to outweigh diverse individual interests, however, if collective action is to be sustained. In many discussions of the subject, it is assumed that participation is coterminous with local group formation. "Strength in numbers" is one significant argument for group formation. The poor, who may be helpless individually by dint of their lack of resources and influence, are better able to defend their interests against wealthier and more powerful neighbors if they marshall their combined forces (see proposition #7 for a discussion of "horizontal" vs. "vertical" organization). What is more, skills, information or other nontangible goods can be shared among group members to the mutual benefit of all. There are also cases in which individuals without skills, contacts or formal labor market experience have been able to gain access to opportunities through group membership: Qfw w a . ... ...... ............... I . n a pilo,t project. landless a tters sure. assisted'by :&$0s topreeve tis initiatie,. "landless contracting societiesa were formed toprfr Widows and other dest ewmn wer g n prioityW twhen thes. .tie value pliaced on female secusion im tangladesh, most of them had~ nevre been employed and kad little exp ein ineatn with' stranges in the p~ul aphere, $4ithbout the contracting societi.es,: they would have had little c,hance of securing paid emplynt ii &. matu.rated job. market. Begn or postitut iighbt be. th destitute woman's only options in such a case. Despite these and many other advantages of group membership, however, it is not without costs. In assessing the value of this approach, some of the problems that must be considered include the opportunity and transaction costs of group membership. The importance to members of the amount of time absorbed by group activities should not be overlooked. Members of credit associations, for example, must spend time attending meetings, setting ground rules, keeping abreast of one anothers' repayment schedules and monitoring compliance, etc. Described in a recent study as the transaction costs of group membership, these time requirements must be offset by observable advantages if group membership is to appear worthwhile. For this reason, it is conceivable that the poor might prefer to interact with banks as individual savers and borrowers if that option were available to them. - 12 - On the other hand, transaction costs to banks and rural development agencies may be lower in dealing with groups than with individuals. Group members save financial service agencies considerdble time and effort by negotiating collectively, by policing each other's repayment schedules, and by making good fellow group members' defaults. Since small non-collateral loans would often be unfeasible if these functions had to be performed by lending or service institutions, borrowers' groups with joint liability may offer the poor their only access to credit beyond the informal sector. There are also advantages to using the group process when agreements must be negotiated or an understanding established with communities, government agencies or other "stakeholders." A task manager working to improve the performance of Pakistan's public sector in urban water management commented that, in that country, working with groups has been found to be more efficient than working with individuals because it generates a higher level of trust. The transparency of the group process deflects suspicion and defuses fears of hidden transactions and special privilege. Participants in the group can assure themselves that they are fully aware of pertinent transactions and agreements. Commitment to participation in community groups is often fragile, however, and it can be undermined when nonmembers are able to appropriate benefits that the group has labored to create. These "free riders" are particularly likely to subvert group commitment when the benefits they capture are finite and subtractable (rival goods) , so that less is left for those who participated by contributing labor or materials. For example, a forest users' group that has worked together to reforest a denuded slope might discover that villagers who did not contribute labor are harvesting wood. If so, then the motivation of group members to labor further may begin to wane; and the collective interest of the group will be damaged by the diverse individual interests of those choosing to act as free riders. If continual maintenance and preservation of a forest or other common good is a project goal, then a means must be found to control the free rider problem - - either by compelling all potential users to contribute (through peer pressure or litigation) or by making the benefits excludable (limiting noncontributors' access to the forest). Either of these solutions will be easier to implement if the universe of potential users is small, as is the case in a rural area. 3. Comunity participation will not be sustainable unless accountability to beneficiaries is built into the project process. This review did not uncover an array of projects in which accountability to beneficiaries is a built-in feature of project design. The usefulness of such a feature, however, was clearly illustrated by instances in which participation not only failed to prevent project benefits from bypassing the poor, but actually cheated them of time and resources in the worst cases: 8\ "Non-excludable rival goods" of this kind are sometimes contrasted with "toll goods." Access to toll goods can be limited, although use by one individual does not diminish the total available to others. - 13 - K)X 4: INDM'S WEST BENGAL FORESTIY PROECT: WHEN LOCAL GOVERNMENT LACKS ACCOUNTABLTY In a district of West Bengal, villagers using local sbsistence forests, final yiel~d., Instead, a goermen oaizto used bueautic Prc to sty retaenig mos o tst. r ht what the tsers' grouips actually recei.ved was far les than the values o~f the -- - - - . ~. . ..~ Task managers pointed to cases in which community members were not motivated to participate in the implementation of a Bank-supported project because their experiences with empty promises of free services from government had left them suspicious of all development initiatives. Some task managers stressed the importance of World Bank supervision as a means of controlling abuses, forestalling political interference and otherwise keeping the project fixed on its original goals. Bank presence at the field level, however, is usually too sporadic for consistent surveillance to be maintained or for task managers to be dependably aware of political lines of cleavage, particularly at the local level. It is a challenge, therefore, to avoid being drawn into political disputes while executing necessary project supervision. A more sustainable and dependable solution may be to allow community members themselves to become involved in the monitoring of project progress and personnel. A World Bank publication on participatory evaluation in the water and sanitation sector offers insights on how the priorities of participating communities can be included in the evaluation process. A key step in fostering accountability is to identify an appropriate mechanism whereby community members can be directly involved in monitoring performance and in assuring that good performance is rewarded and bad performance penalized. One such mechanism could be inferred from a discussion of the Indian and Korean irrigation systems by Robert Wadeo. In the Korean system, the field-level worker (irrigation patroller) resides in the community in which he controls the release of water. He is selected by the village chiefs, and he must be re-nominated by them every year to retain his position. By implication, the patroller could be recalled at the end of any year in which he did not perform his work to the satisfaction of the community (although the chief is appointed by the government, he is normally responsive to the concerns of his community). 9\ Deepa Narayan, Participatory Evaluation: Tools for Managing Change in Water and Sanitation. World Bank Technical Paper # 207. 1993. !0\ Robert Wade, "Organizational Determinants of a 'High-Quality Civil Service': Bureaucratic Incentives in Canal Irrigation in India and Korea." Background Paper for World Development Report 1994. Institute of Development Studies, Sussex University. - 14 - It is not clear that this system would be equally effective in South Asia without modifications to accommodate a very different social and cultural system; however, it suggests a mechanism for supervision by communities that could be adapted to some South Asian settings. On the other hand, alternative mechanisms such as users' groups might serve this purpose as well or better. In general, it would appear that this is an area in which further experimentation and analysis are warranted. Organizing Effectively 4. Community participation cannot be employed in uniform fashion across all sectors and in all countries. It is necessary to plan participatory components with the specific sectoral, economic and socio-cultural conditions of the project and target community in mind. By the very localized nature of participation, a cookbook or blueprint approach to it is not likely to yield best results. Requirements of successful participation can be expected to vary by sector and country.'1 A notable body of analytical work has emerged on the conditions necessary for successful participation, but there is still a great deal to be learned, and this proposition represents a particularly fertile area for additional research, analysis and experimentation. Sectoral Requirements: In sectors concerned with infrastructure, agriculture and natural resources, organizing users' groups that will receive a direct material return from their investment of labor or funds may be the most effective and sustainable approach, as suggested above. A different approach may be called for when community participation is used to implement projects in sectors such as health, population and education, where there is no automatic material return to the participants themselves, and where the free rider problem cannot be controlled. For example, some projects rely on semi-volunteer labor to carry out health promotion or nutritional programs in their communities. An example is the following: ;1\ in addition, -:he requirements of successful participation will differ depending on the type of participation, and on whether it is participation in the choice or management of an intervention, that is being sought. The draft report of the Learning Group on Participatory Development includes a list of the types of participatory mechanisms that are used at various points in the project process. 12\ See, for example: Deepak Lal, Participation, Markets and Democracy. Human Resources Development and Operations Policy Working Paper #20, The World Bank, January 1994; and Jean Dreze and Mrinalini Saran, Primary Education and Economic Development in China and India: Overview and Two Case Studies. The London School of Economics Development Economics Research Programme, #47, Sept. 1993. See also Philip R. Gerson, and Robert Wade, op. cit. - 15 - BOX 5: WOMEN VOLVNTEERS SIlPPORT INDIAYS FWEr POPU ATION PRJC ifoaton~ and ote services to thi negbos Toofrte atre patcpt .in the form. ofpie or 4 salhnrttdpi oe volnter, Muneer rport that hey £1s adi na nenie tn tah stu aordd them. by the. coiminniy, and in the epi they gai Many of he bope this experencewl lwte to imake the ~t ton to a foral and muore reihttv posiixt pitMg in the priaewtr Even the provision of minimal honoraria may separate the work of these women from true voluntarism, but the non-salaried character of their positions nevertheless encourages a part-time and temporary orientation. While it is reasonable to expect that irrigation system users' groups could be motivated to continue participating in the maintenance of the system perpetually by virtue of the fact that extra water improves crop yields, a lifetime of largely altruistic voluntarism or semi-voluntarism on the part of any individual may be an unreasonable expectation. Incentives such as honoraria may be needed to establish a network of "volunteers." A self-sustaining voluntary system without honoraria may be impossible unless it has built-in rewards such as allowing the volunteers to charge small amounts for minor medicines or other government- provided materials and supplies. Projects that rely on voluntarism or semi- voluntarism should also be prepared for rapid turn-over by providing periodic training of new volunteers throughout the life of the project and, if possible, beyond. Country or Project site Requirements: Pre-existing conditions at the project site must also influence choice of approaches. In addition to an understanding of requirements of the specific sectors in which they are working, task managers must also examine the basic social conditions that characterize the target area in terms of the ways in which they could affect project outcomes. For example, the most appropriate entry point into communities will vary by country and ethnic group: that could act asprne raiztos itroducig the prjc t*o comuniie and inviting them to patcpt in the installa manenance adoperation of water sytm i.Whe a similar project s thrugoutth conty,and different aprahswre requrt thrug the.tribal chie . . .§ ..................A . ............JA M .............**. . - 16 - It should be kept in mind that most of the countries of South Asia, particularly Nepal and India, are characterized by extensive ethnic, linguistic, and ecozone diversity. In response to Pakistan's diversity challenge, the Environmental Protection andl Resources Conservation Project in that country will employ a research team, comprised of both consultants and local people, to visit each project community and identify an existing community unit - - a local NGO, watershed organization, or village council - - with which to discuss ideas and alternatives. Tailoring Participatory Projects Correctl: Consulting the sociological or anthropological literature on the project area can sometimes provide an overview of the ethnic, political and language groups that will be affected by the project. At present, the Asia Technical Department assists the South Asia Country Departments by reviewing all initial Executive Project Summaries to determine whether indigenous peoples or resettlement questions need to be resolved, and by identifying relevant social issues and recommending ways to enhance the project's handling of these issues. For many investments, however, such an overview may prove to be too brief to provide critical information about the most suitable techniques and opportune pacing for the introduction of project activities in specific sectors. More intensive examination of the social and political context may be required. This examination can be of varying levels of depth, depending on the project's degree of commitment to the participatory approach, its resources and its practical requirements. A key component of such an examination is consultation with members of communities that will be affected by the project, with local, state and national government officials who will be involved in implementing it, and when the project plans call for cooperation with NGOs, with local NGO staff members at the management and field levels. These discussions can reveal areas of opportunity and points of community concern that will affect the progress of the project. A number of projects have gone beyond consultation, however, to offer project beneficiaries the opportunity to make informed decisions about local project implementation. When communities are being asked to undertake responsibility for aspects of project implementation, they are more likely to do so with enthusiasm and commitment if decision-making authority is shared with them: - 17 - Fa .1...... .G F& notceb, XimprovK .....me..... ..o....upt,anw prah a evlp A iltprojet has beg9un ttilg a component called #xrble Census prolems that will be h focus of the technial astance provided ini thei are . £griultralextension~ verker ("faciietaore" ratne i ..... . 5.e Lona knwdeanof lscalcdiions whh ahyndni h constraints lde tknto accoenttheforoe introeving neatecrnoltputs or oranicutonal forthe Alsonf the ldaoilustraes, por farmers iantdz othe affed by Bank- anhe woulelere pacdt pervioroeize theeno nutske suc ahs teMcnc assistavee Ont lothe hand,rte poor can te d toe in fupesdonaol teeinormaionneedaed oitive deitudiohgs aotloa deelpment poites. onThey ay be nawaricopateflsageo.pin availal kowlhedge ofhe moayb otoly ansrinformedoudtbe poetakenefints orcsso thosxape oabonse illustrampe, genr iaes and onhesafeected bycankel auptedc proectsareciaeneenly are of ocalsonstratiand Fporntes, ran,r teefouth Asian pcmuitio pilotaeoheir choed educint ofc gias aseahniglpsistityncea orh impoeerhnt. neo staffo mebersxpepte that i recreatonaleaiitiesfsuhe ifratioys basetal coutskev dcsioeabout local slceinpeeectomrprsigdevelopment priorities. Educationaar o hefulrag of then covaibt hemr atre ma be suchasly informdnbou the oteniportanceft orcsof leosel fpinFreamileplmae, edessr befor the ommreenity can makeldn athenc pomaalb unawcaeo thevle largrlsonmi ndatiticalPontext ienic thew oc sal coditinis will siuatnTeold Bankose eductter abl girl taca ening-pr i tinygre ondiovemet. thOnae sad national leepors,n to foeratonaifiositiendl tc loal decsionbaskbal thoutshav isoktimes pblic seectr inpfeencen intitutons. prsing oderveloment wrorktiest effecatively ofnth feman edrezeind rilo ini Saran. Pmry Eduats and Epovmic track enblin anCltinan onditOvis antwo stae Sudaional leveldoan Scoo of Economics, The Development Economics Research Programme. #47. Sept. 1993. - 18 - depth knowledge of local conditions is again an asset, only in this case it is public sector institutions and their functioning that must be known and understood. A period of residence in the borrowing country can yield a familiarity with habitual patterns of work and 4 teraction which could be critical to avoiding potential pitfalls that might alienate constituents or otherwise undercut project achievements: BOX 8: PARTICIPANT OBSERVATION STRENGT MENS PARIStA'SSECOND To strengthen the performance of an urban water authority in Pakistan, procedural chinges suh s :trainifng a4 capacity ildih' were .,initially. assuited to be the project' s firat priority interventidh The ptoject had plaed& a. long-tdrm consiltant-..in. the d6verment *a participaOnt observ6t.- -Feedb1xck from- the consultant revealed thAt the. water lauthrity' . mpst serious broblems were not in fact. proceduta:.but suctural. F instancei 124 clerical steps were required in order for the organization to issue: a. sing1e payment. to a contractor. on .the advice of the congultant, the..task .manager- decided to..delay the: initiation, of training programs until a coplhte resiganizationdf..the wate ahotysahaLe stue.had,bben undertaker. The reorganization was succegsfdal,, ad .the governioit -has asked :or the-project to be scaled up.to the provincial level.. In the view of the task-manager, long-term participant observaton w;s a kiy...factor in the 6fue he. project. Without. it, the.iankwould,hot.4ave identified the :problem-correctly,and its efforts would have been misdirect4d. Face-to-face involvement with counterparts and stakeholders in the borrowing country over the long term can provide information and feedback of a depth and magnitude that is impossible to achieve through fax and telephone communication supplemented by periodic visits. For that reason, some task managers are convinced that their work could be carried out more consistently and effectively if the task management function itself were based in the field. The task manager could keep a finger on the pulse of local conditions, concerns and issues, and s/he could conduct more frequent and informal visits to project sites. Bank staff members have observed that an insider's understanding of the "informal organization below the surface of client institutions"14 is essential if the goal of the project includes an effort to change or improve the institution's functioning. What is more, a relationship of confidence between Bank and borrower could be built more easily if interaction were frequent and regular, since establishing interpersonal trust is an expected and crucial component of professional relationships in most South Asian countries (and indeed, in most countries worldwide). The cost-effectiveness of a significantly 14\ Neil Boyle and Albert Wight. "Policy Reform: The Role of Informal Organizations." World Bank. Infrastructure and Urban Development Department. Urban No. OU-5. Sept. 1992. - 19 - increased field presence, however, is under debate and an assessment of this issue is ongoing within the Bank (see proposition #13, "The Role of the World Bank"). 6. Community participation in project implementation and maintenance will be more sustainable if culturally appropriate and familiar models of organization are employed. If possible, indigenous organizations and/or locally recruited staff should be used as a starting point. Locally-Recruited Staff: Task managers across a broad spectrum of sectors and countries recognized the advantages of using locally recruited staff to implement interventions at the field level. India's Tamil Nadu Integrated Nutrition Project achieved greater efficiency when recruiting workers from the participating villages than when attempting to introduce an outsider, particularly in tribal areas. In Pakistan's education sector, dramatically improved enrollment rates for girl students were seen in a small number of pilot villages where the Baluchistan Primary Education Project has supported the training of female teachers selected by villagers from among literate women in the project village. Even elderly male teachers appear to be acceptable to parents of many girls if these teachers are native to the village in question. A similar design was developed for a primary education project in Bangladesh (General Education Prolect), and community response to this project has also been positive. Indigenous Orcanizations and Models: Use of indigenous local organizations or local arrangements for development purposes is a more controversial proposition. Some Bank-supported projects have made effective use of traditional agreements or organizational models that are familiar to the local population and have proven to be compatible with the local culture and environment. An example is: 110 M9ii 110~NI 011,~Sl~ yage 94 ystem, turnrs receive water i~n tur~n according to the ameoat o land they have adder c vain Since this~ syatem is at least aceuy .eld.,has officia.1 sanctin..a4. appeare -satisfactory to users, it has been. 4oemed a useful ajkotw6 the* fre' irrgtinassociation approach. --... ..-.. ... -.....: A drawback of traditional systems is that the most disadvantaged members of a community - - the poorest of the poor and women - - may lack representation in spontaneously occurring local organizations. Customary agreements and systems of distribution may also exclude the poorest and least powerful. For example, because traditional warabandi rules distribute irrigation water on the basis of land under cultivation, those who have no rights in land may not be eligible for membership in a water-user's group that is based on warabandi arrangements. Some - 20 - water users associations in Pakistan, which were organized by project staff from existing groups, have displayed this weakness: Although acievin A water users asociation (UA) ordinance in all fou- o k a povincesas beakt gh or atea nagement project it was learned that the indh ordinance euded the landes roM jOinn tbeir' landlo'a' so"refrin hi ,A consequnwce, they were inadvertently excluded from acc to gop creit whih asavalaleto their landed as mebrso the WT3As. In uh therfore thepoorstCarmers are denied 'participattin in diect bnft *..-:..:..:............ As long as the WUAs are defined as single-purpose groups (see proposition #8 below) whose sole responsibilities are operation and maintenance of an irrigation system, there is no clear and obvious way to include landless laborers. Solutions to this problem include developing compensatory interventions for the landless, such as inviting them to participate in income- generating and small enterprise projects adapted to their particular needs. The warabandi system of water distribution also provides an illustration of a dilemma that maybe encountered when traditional arrangements are formalized for development purposes. Water users in Pakistan have commented that the traditional distribution system (kaSha warabandi) is informal and flexible, and that it allows for ad hoc modifications as warranted by circumstances. The state-sponsored system (Pakka warabandi), on the other hand, is too formal to allow for much flexibility. Procedures are established and distribution sequences are fixed. As a result, a farmer has no opportunity to request changes or concessions, even if he repeatedly finds the canal dry on his allotted watering day. The lessons here are that, 1) when traditional institutions are formalized, an effort should be made to preserve the adaptability to local conditions that is a primary virtue of indigenous management systems, and 2) traditional Listitutions are most likely to be useful when program design builds on their strengths while sidestepping or mitigating their weaknesses. 7. Homogeneous populations are more likely to form cohesive groups than are sociologically or economically heterogeneous groups. This proposition must be considered separately for the two types of homogeneity -- sociological and economic -- since different principles may be involved. Sociologically Homogeneous Groups: The proposition that groups are more likely to be cohesive and to work together effectively if they are sociologically similar appears to be well-supported by experience. This is particularly the - 21 - case in settings in which minority ethnic (or "tribal") groups have maintained a separate and distinct identity. In India, for example, tribal groups with a separate language often feel that they are at a disadvantage working with non- tribals in the majority language of the area. In a country that has experienced recent ethnic strife, such as Sri Lanka, attempting to secure cooperation between the poor members of two hostile ethnic or religious groups may be impossible. An awkward problem then arises, wherein separate components may need to be designed for ethnically dissimilar residents of a single local area. Economically Homogenous Groups: The question of economically homogeneous groups is more complex. The international development community, and to some extent the staff of the World Bank, are divided on the issue of whether programs for the poor must be protected from interference by members of other economic strata. There is a tendency in some quarters to view local elites and the urban middle class as the enemies of the poor. This view is particularly common in South Asia, as a result of the semi-feudal relationship that pertained historically between landlords and tenant farmers in many parts of the region. In many South Asia rural settings, landless and smallholding villagers are informally organized into two or more factions that are led by wealthy landowning families. These privileged faction leaders may try to obstruct project efforts to institute an alternative system of organization in which a homogeneous group of the poor works together to overcome shared obstacless. Seizing control of these groups is the most effective way for elites to resist their potential threat to the status quo. Local elites, therefore, may seek to usurp leadership of community groups, particularly if these groups come to control locally important resources or decisions. In such a setting, barring the prosperous from membership in groups organized to undertake poverty reduction initiatives may be an option. Exclusionary options such as means testing are controversial and often politically unpopular, however. Alternatively, mechanisms may be found which automatically target benefits to the poorest members. For example: BOX 11: INDIA NA.IONAL SO.L TRY. OJECT:; .O .ARTICIAION ALWAYS HELP THEI POOREST? Although a forestry project in India was sticessful in establishing village woodlots on publicly owned wastelands,"it.was.later learned that the poorest of the poor were not receiving the-expected quantities of fuelwood and fodder benefits. Since the village-organization -ipanchayat) placed its highest priority on generating cash ipcome, a large number of woodlots were diverted to commercial use. Subsequently,.,project documents offered the recommendation that future emphasis should be placed on fuelwood rather than commercial use of these lands, as a means of targeting benefits to the poor. 15\ It should be noted that poor households may also be wary of joining horizontal organizations because they fear the loss of vertical patronage ties that they have relied on for access to credit, employment or emergency assistance in the past. - 22 - Creative intervention design is sometimes required if coaptation of benefits by the wealthy is to be averted without constant time-intensive policing. Normally, government-financed services such as clinics or schools are not excludable goods. Nevertheless, elites will usually opt out of programs that provide benefits useful only the disadvantaged. For example, supplementary food distribution can be targeted automatically to the poor if the program offers only unrefined, locally produced foodstuffs that are not normally consumed by the middle class or the wealthy. Forming public school users' groups, along the lines of a parent-teacher association, is one way to omit those who are wealthy enough to use private schools. For services developed specifically to attack poverty-related problems, then, a form of self-targeting can be built into the design of the service. On the other hand, the interests of the poor are not always opposed to those of the middle class or local elites. For example, programs aimed at controlling malaria, constructing a feeder road or improving sanitation benefit community members of all socio-economic strata. For projects of this type, elites may prove to be valuable as leaders of community participation in development initiatives that have a poverty-reducing impact. What is more, interventions can achieve wider support if membership or benefits are extended to all, as the following illustrates: BROAD-pSED) OMMUNT1SUPPORT Because aippemental feeding was made available to any manourished chi.d regardless of the incom of its parents, a Bank-aupported nutrition project in~ rural Tamil Nadu~ was popular with all social. and economic classes, The middle and upper-classes did not try to capture benefits aimed at the poor since they did not view themselves a having been excluded. OA* epression of commity backin was the spontaneous formation of women's support group around the project. Although some of these volunteers were poor, women from better-off families were represented as well The women's working groups were so successful that they have been. made an imora:- comppent of. the follow-on project. Gender-Exclusive Groups: It has been observed that no community is truly homogeneous, in so far as development initiatives are likely to affect men and women differently. Although gender is not a class or ethnic line of cleavage, community groups in South Asia are inclined to be gender-exclusive. To some extent, gender-exclusive community action groups are a reflection of traditional culture and the associated division of labor. Water users' groups based on traditional irrigation arrangements, for instance, are composed entirely of men because women have traditionally had little or no role in relation to irrigation systems. Because men are not responsible for the routine care of children, their inclusion in community nutrition or child health initiatives may not be an efficient use of resources. In addition, the cultural expectation of physical and social separation between unrelated men and women often results in a degree - 23 - of awkwardness between the sexes that could damage group cohesion and undermine the capacity for effective joint action. Traditional roles and expectations are not the only justification for organizing female-exclusive community groups. Many development practitioners, both in the Bank and in South Asian countries, are of the view that women are more powerless and isolated than their male counterparts, particularly in those South Asian societies in which Purda (female seclusion in the home) is practiced. Women who have spent most of their lives confined to the home may have few acquaintances outside the family, even among women. In addition, a majority of these women are illiterate or semi-literate and their access to information about services, markets or other resources is very limited. Individually, they are usually powerless to influence events outside the home. Their capacity to participate in development-related activities will be low unless special efforts are made to provide them with information and to establish linkages between individual women and networks or resources that would be useful to them. Once they are informed and linked to groups and networks in the public sphere, women can be effective agents of positive development change: & K L.O!ML....MF availabl puli services. It als providea a forum in which villa wme the Bankzppoted project the M.S grou will rtain their autonomy and their~ rght t- -e-lt appo-ae at$vtie f t-r v g--. Since howver they ar cosiere to be a sitable sad4 prmsn eil o h implemntation and maaemn of gil' dcain intrvetons in their 8. Single-function community groups are easier to organize and sustain, and are more effective in reaching their goals, than are multi-function groups in most circumstances. It is generally assumed in the Bank that multi-purpose rural or integrated development projects are inefficient and less effective on the whole than more focused single-sector projects. Some task mangers felt that many integrated development projects have dissipated resources by spreading them too thinly to achieve measurable results in any one sector. In their view, simplicity and standardization was the best way to avoid becoming overextended and to operate within limited means. Efficient management of multiple sectors is also difficult to achieve, particularly in view of the fact that it is a rare manager who has expertise in a broad array of disciplines. - 24 - By extension, it would be easy to conclude that community groups, too, can be managed more easily and can achieve their objectives more surely if those objectives are kept simple and unidimensional. If so, then single-purpose community groups such as rotating credit associations or forest users' groups can be expected to outperform multi-purpose groups that embark on a panoply of activities such as infrastructure, credit, and health promotion, simultaneously. In the view of Robert Wade , however, secondary functions might be sustainably added to a group's core function, depending on whether the secondary functions are complementary or isomorphic (organized according to the same principles) to the core. Common sense suggests that a community group organized to upgrade an irrigation system might easily extend their activities to include the construction of a drinking water system, even when other aspects of disease prevention (such as promotion of oral rehydration therapy) are thought to be incongruous and beyond the scope of the group's interest. Multiple functions way actually increase the positive potential of a project in cases wherein the effect of these functions is additive. The Nepal Rural Water Supply and Sanitation Project, now in planning, is an example. Although community-implemented water schemes and latrines are the core of the project, it will include a sanitation/hygiene education component as well. The two project components serve complementary and mutually reinforcing purposes, and together they will lead to health improvements. The provision of infrastructure, even with community "ownership," might not have the desired public health impact if the attitudinal and behavioral components of sanitation were ignored. Task managers were not able to offer definitive evidence on this proposition, since none of them had managed projects that generated comparable data on the single- and multi-function types of local organizations. This issue could Ye a suitable question for structured experimentation on participatory development. In the meantime, some projects are planning to add functions to community groups that have proven to be successful in executing single-purpose interventions, and it may be possible to document whether the cohesion and effectiveness of these groups is weakened or strengthened by the addition of new functions. Optimum Scale and Replicability 9. Community development groups should be kept small in order to maximize effectiveness and to control free riders; but to achieve maximum impact in terms of national development goals, they can be coordinated or federated at higher levels of organization. It is widely recognized that organizations of the poor that have income- generating and credit components are able to utilize peer pressure to insure that individual members will conform to repayment schedules; and functional organizations such as forestry associations are able to enforce compliance with agreed work obligations. This advantage may be lost, however, when the size of the group exceeds that which allows daily or frequent "face-to-face" contact between members, or when members are drawn from distant locations. 16\ Personal communication - 25 - A 1992 World Bank working paper noted that "In the absence of countervailing arrangements, the incentive for group action diminishes as the group expands." 17 This argument suggests that, particularly in large groups, penalties may be required to control problems such as the "free rider." In small, face-to-face groups, communication is frequent and direct, consensus is possible, and monitoring is simple. As a result, responsibility can be clearly allocated and easily shared among all members, and any deviation from group norms is quickly noticed. It is a common contention that the celebrated success of many NGO-organized community development programs is contingent on the small size of the groups with whom they work and of their programs overall. Small programs aimed at small groups can be tailored to the specific setting in which they are implemented, and they often absorb a very high level of funding and technical assistance in 18 relation to the number of persons helped . As with community groups, these advantages may be lost in attempts to expand or replicate these initiatives on a larger scale -- particularly when those attempts are undertaken by a financially stressed public sector. Government programs, by virtue of their more limited resources and the requirements of broad coverage, may never be able to achieve the same level of success in their effcrts to include grassroots participation. Even NGOs operating at the national level may be unable to provide the intensive assistance that is needed at the local level for effective organization, training and technical support of beneficiaries who are expected to take control of their own irrigation systems, forests, savings and loan groups, or other poverty-alleviation activities. In the meantime, some Bank- supported initiatives that work through NGO partner organizations are testing a method of implementing participatory programs that might be called "nested" management: BIOX 14: TIERED~ NGO TRADHING IN IN1D1A'M UTTAR RtDESH 50Dq LAN4D could ervea a m l b ld broadly .i the fu.tur Farers0 wae maaeetgop will be oraiedadco .ty Voata wilbe trained :b technooy tra .fe bymall. letal rNGO. the stf ,>f ....... ... .. . * ** i7\ Robert Picciotto. Participatory Development: Myths and Dilemmas. Policy Research Working Paper, WPS 930. July 1992. 18\ Mark A. Robinson. Evaluating the Impact of NGOs in Rural Poverty Alleviation: India Country Study. ODI Working Paper #49. London: Overseas Development Institute. 1991. - 26 - Other projects, such as the Participatory Forest Development Project planned for Bangladesh, will use a similar structure in which advisory NGOs would coordinate the implementation activities of small locally-based NGOs. These projects could be coordinated at the national level by a single organization that works directly with the government to insure compatibility with national goals and policies. Community groups can heighten their visibility and influence through similar mechanisms. Specifically, a number of small local groups can be tied together on a regional or common-interest basis in what could be called the "federation" model. This may be coordinated by government, as in the two-tiered Comilla model pioneered in Bangladesh. Under the Comilla approach, representatives of village- level farmers' cooperatives participate in federated councils. Representatives also receive training and technical inputs, which they pass on to members of their respective cooperatives. Coordination of village-level development groups can also be performed by NGOs, as described above. A related strategy is to organize a large assembly that subsequently subdivides into local decision-making and action groups: SOX Is: USERS GRMPS.An'SVBGROUPS IN NEPAL'S BRAIRAWA LUMBJNI Deep tube.wells serve large areas with irrigation water under the Bhairawa-Ltifnbin irrigation project. Because the approximately 150 farm families . served - by --ea.ch tubewell :constitute too large a group to work %'together effectively, the tubewell users' groups are now subdivided into -sub-units of about 0' families each. About seven sub-units are conftderated in a water users,group (WUG), and 4-5 WtJGs form a water users association (WUA) with an elected chairman. The sub-units and WUGs control distribution of water at the local level, while decisions affecting the system as a whole are undertaken at the level of the WUA. 10. In order to replicate or scale up successful programs, it is important to identify 1) the internal factors and external conditions that determine the success of the program, and 2) which of the projects *elements of success" are found in many settings and which are unique characteristics of the original project site. Assessment of this proposition would be premature at this time, since pertinent evidence from Bank-supported work in South Asia is scarce as of 1994. Many of the community-based projects receiving World Bank support can be described as pilot projects that will later be implemented more broadly. It should be possible to say more about the feasibility of "scaling up" small projects when results of those efforts are in. One early lesson on the upscaling issue is that projects that show positive preliminary results may come under pressure to scale up before effectiveness has been clearly demonstrated. For example: - 27 - JOC 16: BANGLADESH GENERAL EDUCATION PROJECT JGID'TO SCA[t TP Aprimary education program in Bangladesh that involves -communitiq. in. the creation of satellite schools ib -their villages ha: p:.gae:rated. enhtuiastic public response. .Communities vw4tee) or s , in the. project and a waiting list has been established, Az gheventual expa'sion was planned, the govexnment is, auxi.d : %t.move EwmimA: a anticipated-in the project design. The Bank holds the :view that; sigiicant quetions about austainability should be reso] Ad e de ' prcsject. S substantially expanded, but government official. d ot ar thisviewk. Premature attempts to replicate or expand small-scale pilot projects defeat the objective of testing a promising approach. A thorough quantitative and qualitative assessment of the effectiveness of pilot projects is an important requirement for expansion. What is more, it has been proposed within the Bank that full national-level implementation may not be warranted without additional evidence of success from an intermediate-level demonstration project. According to this view, an orderly progression through the pilot, demonstration and full- scale stages is the best way to insure that significant investments in a particular approach are justified. Monitoring and Evaluation: Another issue that this proposition raises is the question of when a program is deemed to be a success that is worthy of expansion or replication. Careful monitoring and evaluation of the participatory project's effectiveness in reaching its goals is the route to separating successful from less successful strategies. Bank-supported projects normally report the project's completion rate as a means of measuring its progress toward its targets and objectives. Community participation may be assumed to have taken place effectively if a project relying on that approach achieves its overall goals. A few projects, however, are being designed so as to include process indicators of successful participation. For example, the monitoring plan of Sri Lanka's ongoing Community Water Supply and Sanitation Project includes indicators such as the number of women or the percentage of villagers who participated in the project, the number of users groups formed or reached, the number of meetings held by community groups, etc. Pakistan's proposed Northwest Frontier Province Community Infrastructure Project will not only use indicators similar to these, but will also use a phased implementation schedule to monitor the sustainability of community involvement in the project. After each phase of community co- financed infrastructure is completed, a period of time will elapse in which the community will accumulate funds for the next phase and project staff will observe the degree to which the new infrastructure is maintained. Only those communities that maintain their new systems adequately during the hiatus period will be offered the next phase of construction. Although these indicators can be expected to yield information on whether or not participation was actually achieved, they do not conclusively prove that the pivotal factor in achieving project success was effective participation. For - 28 - that purpose, compar.sons between the outcomes of participatory and conventional projects, such as a recent exercise performed by the Bank's Participation Learning Group, are essential. Oveiall, the most appropriate indicators for assessing the effectiveness of various models for community participation and the value added by these models to overall project success are still evolving. The more difficult question of whether models judged successful in a particular setting have a broader replicability under diverse internal and external conditions is yet to be resolved in systematic fashion, and so it is an area in which further experimentation is in order. Achieving Local Control: 11. Local authority over resources that are to be used at the local level is important to the sucmess of community-level organizations and to government decentralization/devolution initiatives. Local authority to collect revenues or fees is a useful approach to generating local resources and building commitment. 12. Clearly specified rules and allocation of rights and responsibilities are critical to the smooth functioning of levels of government, to government/NGO collaborative efforts, and to the success of community groups. Even when beneficiary opinions on planned interventions are sought, government-administered programs that are intended for uniform implementation over large regional, national or state levels may be unable to respond to needs and preferences that are unique to specific local areas. Within a single state, local conditions may vary to such an extent that beneficiaries in different areas face different concerns and constraints. For that reason, community participation is compatible with planning that allows for local variation in choosing and carrying out activities, and with management plans that make expertise available to personnel operating at the community level. Without an allocation of development funds to local governments, however, the resources needed to carry out local decisions with respect to development priorities are usually inadequate and the devolution of authority is little more than a slogan. Along with the World Bank more generally, the South Asia Region has shown strong interest during the 1990s in the topic of decentralization or devolution of go- ernment functions and services. Little quantitative data exists, however, that is adequate to the task of assessing the value of decentralization schemes. Presently, the Bank is considering the feasibility of undertaking a major research project that would use country case studies (including India) to examine questions such as the critical components of decentralization, how measures of decentralization can be used to quantify its impact, and what determines the capacity of local institutions to deliver rural public services. The role of appropriate fiscal systems would also be assessed as part of this initiative. There is evidence in support of the suggestion that local governments can be effective agents of participatory poverty reduction programs: - 29 - ..,. . ..,.. . ;,:..... . .... :::<.. :г>: : ::;� :: . . .. :. .. ..... ... ... ... . 4 ... .. .....:.: ..< :::::;'_. . . . : .. .............. ............... ...: й.:. ..... .:::::.:::. . . .,::: :.:.:.:..... .. ... ;::{.;:.; {`;: ::4w:.+.: ..... . . .: .. .:i::4гг:йУ. . ..::. г'�: ' ' .:: л..:....н:i}.{iгл'f..:4:гчг{г':•i:.г:.:':iгг:у.i{г:::iiг:.i•'гц : :: ч..: ::: :::... . .:. :::ч."::г: ...... •{:л:' i:С:: г:{i "г': гi:i::v:�г: !iц: гiг.: :г:. �:`:: ::::::.жл.......д .................:г.и:..ч•нн. мv.::•.:::::::::и::..:: .....: •г:{й::г:гг'г:йг:i^ггi:.?:{г:iггг�:i:`^:..:.�:: н<. �:.•:�:..,::.: :................и:ч....:v•:::..::::...л.....г....л.:.{..и.иг..л:.чf.... ч..:... � ...г....... ............ . ::::.... .г.иг�:::,.нг. ...л...... i:г: ,:и.. .г.....:.::::�::.:::.::.:.:•:::.:г'.г::.:.л :..: ::. ...:.:{:::.{.:....:..... ::i}i•:i . .:. ..:: . ....:ил v..::.:: .::.v.:{:•.::'г :г:":г{:г: :.{: .:гК.•:.{г:г{•ггг'f.'г;?г:л';:::}5:v'{{:::i::{:i::<г:{:':{t:::::{::ii.::::l:г';:::ггг':::,i'гггi.ггi:гл:�:гi:':.- �_ ..::.: чх::..:..:...:л:::: .... .....л..:.::л. :w...... ..:. . ,. _ ::-л ::.: .. :: :: ..:. .i�г:: .: .:: :.; . v. ..: .: . .4 : :. •:::.... , .: .: •,..:. . . : :;.: л ;г:. : . : .. . . , :: .: .i: �: {. � . vг. •. {:г'� : ♦чг.- . . .}: .. : •. г •: ;: , .: .,. .. •: .и . . . . .v .. . ... . . .... . . . .. . : i. �. . . , ::i .г .. . �уΡ +:��г:��4.:л:•: . :. :.; �:. ..... ... :.: ......... ..:..., ... .. . ..У............ •:йн. . .: и: � .: . :.... л'} . ... ..... .. ... .::. . . .: .:: • : • ::� . . . . . , . ; ::: .: •: ••::::: г:. �.:: . '. {. . ° : : л'. , . :.и...;: { •: �,:..; ,'': ::..: +.,...::.::и.. ; :::.:: ::: �+: л::.;. ... .... ::.сг:.:':. ::: •. � ..;.:� . . :::.. . :.:.......:...::.::. :•:::..: :•.::.и .. .:. ... .:.:..........:<: ..::,. ifr%: ;';::.. : ,У,.. : . . •.: •. ...:•>:г,{и::., : .:. . .. ..... ...:.:: .:.. �. r, ч.r �: �:.:. :, ..... .:. .:' , . .:: :. . . , .. .. . • :.:...:.. : ... :::: •. .: �: � . .. .. :.. .....: ,:.:::: •. :>:.х : .: ::::, и.::: ..: ••>::;;�:`•� ... . . .:.'.гo-:: • :: : :. . .. . ... ........ . .. ::.... . . . ... ;,,: ..:. � . . . .. .. . ... .. . .. . ..: .. : ..... . .... :... .. :.: . ...: .::, •::: .::: .:: :.::: . : .. . • •: •. . . ..: :. . г . , .:: . .. : •г::. . :.... : ...;,.. н':%::Е .5:;г:". ,iiiiifi: �ii;::;г;:: iiiSгiг:.2:i : ... .;..ч:.: .>:::. :. ... �::,., .: , .. . .���;.;::�.�`�sz�<���5. ::.����� �•���X:�ЁF��:: �'�?��;::'г;::.:�>г>::;<:;:.г:г;,:>:;:::;:;>:;>:::::.:::<:��'`::'�'._.:, :.<..�`{й::•::.::,:,.:'.::':.;.:;;>:;::':`;::?.;:>г::::�`.:г� �� �`.i:г :;: �� :чгi .� �a:��fii;:% `f.1.� • ��'•t� :г:; "'�`: �� `�� : й � �'r': . i7F� . ... •��Q:Clа. .. . .� �'..�.... $�' �. �... ��;A.I1... ;.;:� .й!. ......�'r.. .. .. ���...... .. .. �ч.. . . �...... �:. . .� ... � .::.::::. ... ..�и:.и ::.::......... �и.::.ч::и:::::::.::и::::.•�:.............:.::::.и:::и:.::и::и.:;::�.н...:.;�..:�...... .::.::: ..� .:........ ...........�?:::.. ........ ' ' ..�?.. л.. .........}г.•г:лчи\vw: и...::.г:::::::л:::..:ц.....и:чг:и..v:::•.., м: ............................ ............л..................л.ч....:г:л............ ...::::::.г.:'•гг:.i}'{• :::::::::...... w...:: . :•:: ..., ц4.:'::: {.: :::г�:мг}'<S{: ::.гv: г.. :л.-.: ил...ч..:::...... :::......... �:::Уг:'г'::::4:f:i::г'� :::{:<:..::и ::ггл :i'::::i.�.:::! i:'г,:гг{•:::''гi::f �'::г ::::г•:{::гi::.i:i::i::г:г::}::{iг`:iv:::: i:..y' .:<:г::::г::г:::::.:i:::::t::С:{; :::?i:::::::5:::;::г:�::.: :и:i �'г:�:•:}.•:••+•:{{.. й::•:'.i:i:�<и.,, .... ..й.i'г:i,-:.i'::::;y:i::: ':::С•::. :.i:r,i:i'.:ч.::}:г:t�::г.. ;:4. :. .. �' �?::: :: . .��г�.�. d::::��:�;::::;:�t�;d����::>::�ж�>;f�t�:;.:: . .:�.r�i�:.: ��}:с��::�г�>��.����•<�� ��; • .;..: : �;.::: � � <,:.: : . ..: �: �.� �:::::::>:::: . .ч ... �..;4.:..:>.и�:::.и .:.::.::::::.,�°':,�.:.:.::.и:...::.:::.::..::::::.:::...::..:::нл:.... �а.......:�� ..........г..�::.4�.�:..::: ех����,.:�:��с�':::� : .�����:... ..:�h�;..: : :�:и . : �.:; . . �.. . .. {:.и:::... . ...�,�..и>� и. ...и и.{..{.; <:::< .::..: .. :.: :.:.: л • •:4:: •:: •:.и:.и:::•:::::.и •:: �::•::::: :.:::::::::: ч:: й:: r>. ...::::: ....:: л:::.::::: ч:ч......:;: ч:;{:'.:::•г.лr::г<•:й•:.: r:•>: г:{•r•:.: .:.и �:. ::: н..:..: . :•:.и:..::: г • .... . .. . . :У{;.(;�•�:{:�{::q:г:r.:: •::} •�.ч�.ч„:.чч•..:.:.:::.:.tи иуΡ,,...,уΡ•. :•::: :•:.и:..; ...::.:•. .::,: -r::::::.f�:::::.: .•.::::.ч•:�:. уΡ•::уΡч.: л.:s ,::..'.b -..и: •:::: :•.:уΡ:� �� :3`с :::уΡ •�:�:;?:.:х...::г:>;.:..; ..г; :...<::.:{:: ::а:.::ч .г.•::::::.ч::: у: :.: :•: :• :• eta'•:r..;•: • ...�. W. :�F'::�Yi"{�и44. �}�п:):�{ �.`.'•�'i(�рΡ '..:},�.� .:4v,:1�лИ'д :�..(�:;}���:?:г('{.,R. чуΡ (� :;л:д�Х��.:.уΡ7: у.�е.}гд Ч' �} •.{ .:ti . q �� .N... .::г:� . ...... ... Su`К�'hCi,й,t�vл�f.A..�S}.,i,И ... ......... .. :. :RSч{:'�. ...:...........�?Т::?.. ... 'VR�i7�::'г.ч?:7�"�:�•::ti��.�. •.�.л.;/�.••/•q�гл:��::�: •?R.:.цж:�.i:й:^:{:�.•г.:• :�С:1\:`.;.?;;И;,�:{4• v:F...wлw.чг.vv: .::;;.:,.М',.лч:. . vi,'v,;.;�FS . .v. ч�,..,•:.::::•::::.;...v ..ч�., л.�;.F.й4!S!;.;б.......л ..:......:.:k.. г.:г..:л.ч..•......л. �•У:.ч.х :.;;.чй... ч •.:......й:::и{.ч;,ггхi{йчг'vп;., г }..::::::.... ...г.....�..гС?ii:•.ии.ил... ?:•:.•:4 ::{:й...и::.:гцг':йгF,.:v:::::..... г, �, йч,г..г.....4::,:::i:v::;гг.i: :чгi?г,.цУй , t....i . . �.fFF..'н.й.и:i'г:ц:У'•: .... .::::: ::::.....':�..:....: ч ч{.f. •v• 4�гг ич .л{ .:..........,..'�]! . .....fi�......;..л.и..:...л:..:уΡ:!...:��....:уΡ �. :ч; .:� ......{♦,+:�}...(.� дуΡ�рΡ !{, �}� '�:i{й . ч,л.}::{}...:•.•::{•:{•:н':гг}:.::{: • � �уΡ/{ { (�уΡ!�/ (к�и...:.:.4':...,.лf г..;.лц.:.4.:.г:.::....:г.г..;...:...:�.'•..г'?..ч:fуΡ�1���i::гг:'(ц•;•;•,.::....'г' :ц. .::' .'.i•::г:ч':: ?�� :У� и�.�'•.У-'�� :::^Я... � �i%� V:1'-АдгиW ':Т�.FТИ,:Э�:• N. � j..�.���•:.чуТF�.�lУ.��•. '.-ч'�•}.ftY3:��� :и:�Ч�.•.•.г.�t�'}. �У}:�� }� �:.ч�.ч ....м�......:�...;.�::�? :?�.,-л.....:.::г...л.'�s:..:: ......:. .....�'.� .. �.... ...��...',�.,''`г'•,�,.,.'У".г4.:.�:`,�'>:':''п�',�,,:.,�.�',м.�,F.:..Г'.,��:л;{'л: {и . .. . \•.••..v.i'л'лv . ...... . ......: � г. . ... ..... ...:..... .......... �.:..,К?.vv. v.vж:::::::н::::н:: �i..:.г.г... .::.::::: :ц: ::и�:::::: и::::::мvw.чл:: v .г. .. .. ч ..f.... ..F•.л .fг л:.:4•..:::.л: ..г::::: v..и:..л...л..�}.:.w:::::л.:::л:w..:,::.:::.н:.чл::гн:::�:� ..:::::::::::: .. ......ч .........г..,..........:: ............. : , . л::::: ч:•:г.и}гг.^гM1.и.4•:•::•г+:G.iггг'г':{,?.,:i,.:;.,г..iнг:. •: :.: :•::� ::..:.........:... ....ч..... ......-..ч.\ ........ч.::. чл.ч::w.;........... ,У . ..::::. ..... ..... .. . ......: •: •: ...;.: • ... .. .. , . •. ::.: �::•::::,.:.; ...:.ч;�й:.::: :{.::.::й�:: :'tг.:. ; { ..- .;. � :;:::;•',f.:;;:•.,•::o-::.:;: {{:{:. ...:.:. :.,...: � .. 5:i:':г...:::: :.:: :л..,.:.и .:.:.:::...:::г.:, •.; ;.,..ч,. :: :м: .::��. :�:��.^��:.•:s ' ч :�1�;� �� " :<�`�. •� ���:' �:?:��i�к�`:е[;�{:>. • � • 'е »s:a�GиEr::r; • @ � �;:;:,:: ' �J1Yf�.exi: •::"<.�:. . С�:::: :�:: ..�.. .:t:: :>�.. ч..�.,�.�.:.. .:А� �<�..�;:, и�€�t?t. .:�:r.:<::...:.... ..�.:.Х����...... .:� .................�',У...: ...1.Е1Х3-.. .::.:. ...... �:.. '>:.и....:.�.::.;.:. � :..н..� ................ ......... . ... , .. . ... . .. .::{:{{:: :,.х'�..�!�..:.: .... : ...::::.<.:{•:•..... ...�...:,... : .:... ......:,•..:;.4ч{.,:.::::::г:::::::::.. ,.:::::::.:::. ... ... : .:•�..;:{.. :�.У.:::::>'{: �:s: й::.,::::<::.::. :.: ..:: •::.и::::........ч :. : ч ...:..::,:'.:::s<::::.:.>::....::..... . :.:.ч�::. .:::: :.::::::. :.::: г.,{.:: . ч......... . ..й .:.::.....ч:: .............. . чл............\ .. . ........... ..:.f. .: :\л. ь. . .....У..... .... .... .г.:..й:::: .............. ..... ��.........�. ,.... ...: :� �::'::{ч4г}:::гiУ:{::i:г..:.F............л{�}ьг:чггi:iгi:г::::}::л:�:::f:":Sгi:::::?.:..г..,:.,,,.:�д , А . 1. . .. ; /. ; .г.: гL .. . ч :уΡ�. .г.г.г.г! .{ . уΡ �t ,�в : {' :+г: г�.. .�. ., .'лг.;k�г: -:: •г . ' . '. :::i:�г. • : <{;.�'if. � :?< .. :::: : �.�гуΡ �p �{ >:� {.. .•;d.i�. . , {; :; .',�..�; :. .`,У . . . . }� .•.�..�,sF.� . �й�%��.г�.�С.�.. .�:ч .....,l,ч4. . .. Ъг.л�...:7�� . ��. . ::::{ч4��к��� ... . ��.�v'h��...�r..�r.•�,�н+.y���iii:':::t::;�`�с ��:'�d::r�:ti:3i.�ь.Sч л•.:,:..ч•.г.::::гг:.:::нн••чч::. •:.ч•:::...:л :.:: :...:::::......и:: �:.�: �:::.. :.:::..д'г.дг:::::::.:..:'!.qл{!3';. ..;... ...:: ::•:.�::.:..:::. :::::......:.........::::: :.. ... , ................:.ч•::::: ..;; .. ..4.. чл4 ......................... . . ..г..г:: :ли'{й:vйi/.w :ww::::::r;:::':::: гнгггг':.гг�.г �:..;..:г..: лггг ............г глv:.v. v::г.:.:г.� vv:::.v::.иv... :w:: л..:: г.ч4ч}г.. л...чv:: .::: ::wли:: :м. ч., v. :..:. v. . . ...........: ...: : .л. .:..::.::л... г... ;г . ...v : ..: .:...::..;:' ц :лл:лf:::w• ::.:: : ..:....и ...:.:.и.:и:г. ,ц . . .. г , .:. . .. :.:: :у}v:. � j� уΡ : v . ... :. .. .. ...г ч . . .. �:{`й: . . .. , .ч .уΡ�.уΡ jуΡ :.,.{,,рΡуΡ� .._ :. . уΡ-� .с� ч4уΡ�.уΡ� . �уΡ.�i� /� �q �$ /уΡ �уΡ�{ цi�в� уΡ�'гг'{{ :i: .:' [.}.�&.н.?�'.��{ tl �r.�уΡ�уΡуΡ} уΡ� /�f�рΡ.�i� ���!(�•уΡ%., .: г�: }:fi%�....:' гг,:�: ':, �•:::г':::': .`'}:ч: :: �3i(АУ)А;•::ггг' ' ���i:ivv' � . .ч :bЛ� ' ц{�{G' �SК�:гi: .•1i.•JG!'i' :. 'М'.�: ггг:"...�: ' °вi:iг.. .Т'У.' ':Fi��.iC.;Yв.'iI•{itfi г!v? Ч�Р � . � ��,' �, :`.•� �А. 'LУ:гг ь;.; ., ..: йчч,., . �.'гТl.•R;(. ...�.�. �iC�•fi}:..; . �. .�.. . � � ...АЛ.. ..R ...... .�.. ...i.....г.`�М'.. . г'.4С�.� . �'��i::.�F... R......... ��W.,4i,7.г.....��:i,v й:\г.:.�ч. v;Nл{�;.,.., ..�...х:f.w:..:�:::.ч :n}г,У{,.}i:• . ..г, .: x.w:.x:M{.iгг'г::........ . н. С.. : ч.i:г$:• .. ..::: ::ц: . .. v::i• • {• й v:: 4: ..:: ..; . .:.. .. r:::v .......: v:•. чг'ri:�'i{:$гiiг$::i:ii::i:г`:v:г::::::hг}..гг : .гh.:hчйiiгчг:{:.й:й:г{4г}г:�г:г'гг•глS:{{•. й.:. ..Уч д�}ч•глин: .i ц..л .: : гi: гг}г:: i:{ц:{v:v.:v... ...: hvv.: .. . и:....: мii ли:... г:........чг. w . •. v .. :ч:.........й. .::<w::::: � ... •.. :.:.и...: . л. .: : .::.v: . :...::::: . :...::. ..; ...л.;ч ...:.::::... . .. .: ............ч.: ч } "�♦''. . .... :........ . .{• �4 .......:...... ..... и:::� :г:i',л,f:.5 . : •Ь;: �`i.. .. �..::,j�' {.F::.�уΡуΡМΡ .ггн .,.i�.. г::г. �. },� иМ уΡ� ,л �Г.{уΡ�i. }/-� ...и .. . . ' .д '..'г.г� • :: и:::: , ч.}F:Y. : . �.:С . .. •!16'.!/..:�:!{Т :и.i::И ...'3ft .W.d!f:Y:: .. . : �' . :'.}::: . . .......�Э!!iY{i'ц . • .�I:jг:::: '.. л.ч � .�'г::::i}::�:�.ji• ?::.:;�i{'':. t?,.{, . �.Й..ч л ..t..... �# .................й �:��:..:.... :.. ' �• ••� ��... *и.::::.. :�а�,�.?:�:�:.;. :��.��'.�.а"'�. ;�л�:�::..:::.::..г`.::�й�: ���..{ .> :�::й::::::::�{.::.н:�:;::..:?�::и:::.и:.:::.::::��:..:::�`• ...............................й...::.....::.......�:.�?.:........ .......................:.:.:........ч.....?�.#�..:.:...и::::.�:.н�. .: �:::::..:::,.::::::::: т.:..::>::.,» ::::..::: :.::.и:: •:....::....::. r.�v:.:.::::::::::::::.:::::: :.:::::.:.:г. .. :...,::::.. ....ч.......... .....::г:.........::.: ::й...`��.:<>.и..ч....:.:....::..:..::... 4.:...:: .....ч...и,. .......�.�::,::::... . �:ий,�:{::�',и.... . ;.ч . . � .:: �.: :>'.;:. и ..: ..{....: ..:::.и•:.::.,>,.;...�::.:.и::.::..::.:::.: ':: и: :..::.> ::.>:<й:::: :::;:;:>.>;>;. , . ..:;.>:<::..:;::>�:::.::..<::: :,. : ..... { 'й:.::й:::::: ::и.:.:::{.,::,:.:;>.::ч.:_ :' {::ч;и:::;. .: с�.г�:� �:�:. � «::>:� =::: ;: �. .:»>::а�. �в���д�,.:г.:: ::.::.::� е:,г>:;�i�����: . i�.:::;::o�'���о�:�а: ���::;;::::.>..:� . sя :.v:.�°::>::;:::: �' �<. •��::. .; �?:.:.::г�.:::. ч...::.::::��.::::.� ,.��..::ч.: ,:.:.:�� ::::...::::::.::::::: �у':,.::::.:�.и::.:::::::::: ..... .... �., ...:.. ::.: .:., . ..`�.. .�. .,:..�:с�.:: �t.... :.::: :�.::: ::к�:�:�: ::.::.. .. ... . ..........:.......�....................:.. ..., :...:,......г..... ............................... ..... :::::::::' . .::::. : ::......: . ..: ... . . � : ..; ... и:; {�:: .:: .....:. ::: .: ::......:..., :. ... . ... . :: :.. .. ,.:.. .:..; ...,::. . . ...: : .: ..:: .. .:: :.. .::::.:...: .. ........ . ......... ............:.:':<:{.::::: .:....... .. ::..и::...: . .... .. <'<::�Р �>::::��<:,>...,:{:;.::::<. ::�>: ч': j� . . :�з�`� ::.3:'�> �� ":<:��:: •: `#:.€<<�i`:�:��?.�:s .. �.... . .. �:��..... �::::<:#�'��::����з:::::��=г`��3;�����:��:�::,�:::: :;:::'�':��::�::;;�'�".:.:.:.:'`:�:< �►::�::;:�: �ж ��чи . •.� •. vг.�:: • и4:г:::лили:......:..:......г..:.ч••�.�.•�•:.:ж:., ..�. . ....,.�.{.,,..и:лч•г :: �;5: гР.ч :: ���.,-:. ч л.?г'л`�..4лtiй4}.. ..... :..:..г.л..... л4•...:vггi:v.i :::........::x.и.\•гчгг:гг'г':гf.г?г:ц'гчг:�:::[...... ..... мл. •4 л...v.ц:н:::{.г:•:v:Слг:• ::.и.:ч.чv.и: '.ггг:vk:гвiiК•::{и:г:...... ..г4..л .. • •.>г .:й}л{ .... .:{•:Fг:: •. .л ........... .......F.... ........ ...... .. ....::;....и • :.: �нчх�{i:. °:4 .:::л:w:v:.иv:.;vi:rc•г::?�::{гг{•iiiгk;iYi:i:г:::}:vч{?{::::i::vггyг::iw..::.v. .. }.... .. ....: :. г.... лv {,. . . : .......г:•г. : .. .: .... г : чч.; . "чь ... . ... :: ::, r :и :и: . :,.: , . �4; г•. л. ...�.й. ...Ч . ,. У... . :: 4• г :v.G.• г: г. :.: . ... л . •г ч , г, ч .\.;, . . : ;�г.. ;С. . . . . : { : . � :'.�t+�Х3��.�;г:.:1' � ''� :��::::._;...:.. •.�. ..:; : � ••;::. ::: •:•� . :.. �3.:�''.�..:..' ' .. �;:.. : : ��::.:::�'?.�:::;{: � �.:' :.. ;�г::: ; �`L?::• ;�й��:r..:?'> ��: � ::................ .... �>:::>:' :��.�'........�r . . 3э'�... и: ��<:>' {.�'г::<: и.;��..чл...�е . ..,��.. ч.:�ч. .. �т. .ч . {�. . . ��. :.:;::�:��... :.�, �:� .... ...................:. .:..... .......:4.,:..г:.::::::::::,ч.:�...:�?`.::>.......:.и.и :. г :..,<:.:::..ч •• ..�::� � г. г.ч г.......... и.... f{.� 4и:`:iч:и.л':::.г:.:::.г::ц: л..... ....г•: л:'•гi:.; ::и:.г}гг:{{и::::::i`:л::г:':..v.: .:\..й�$:. :С.vи:гi:iи: :.:•;ч. ?ц..и::ч::.::::::.и::::•:г.v :.гг::лч..\ч•.и::::vди:::х:v::::м:.и:ц..f {...... .ч . .:.. ................ . ..... ........ ч.........:.. _.. .. . ... :.::: ..; ..:..::{г{й:: .. {г:':4и.ч.:. .. :: {. •::::: гг.:f•:: ::У ::}\v:: :^ :•� {{:. и:. . : и::: .. .. ..: :: и и:: .. /..-...... :... :.:� .:}...г::г:.г.•.{•:•.::..;:.{:.ч .. д уΡ� гf::в... ��.�{:{ ' У. �рΡ . �. �: �:::�:•г:::Y..v.'•i -���:д:.»:� �.� :�:�...с�::г>:� .. . . .е�д�:>::::: :�:::::, •. : � ::���_:�:�:г:�'�г:.::���:... .. �<�::��:..... � :.��:.�::: ��<��- ; .:...........::..ч ..::. .::.......�:.............:..�....::::::.-..........'�.........ч ::.: .:..::.......................�..�.. ......: ...�� ..���'����>:::::ч�г.....:.........:...л..: v4•.w:::л . .. .. 4. хw :::.:::::::::::::.иn.и: ии•.:ич.. ��v.�.:............ гг. ...г..:....г л .... .... :... . ..........ч.{..{.г.ч . ч ...........................г...г... ч..ли::i::.:fiгг::.гг:iйk{vг::г:{•::::.:: :ч.г. . .......... . .. :.: :..: .......... .... . . . .. ......".:'.:::... . ...ч:.::4iiчи'{.г}:гУ.{{и'F:гi}:й:i:г::.:::. i:. .�.... .... й�чч.ч. .......................л........ ............,...........�гhvw:::v::.1::г::}:л}:г:г`г:й.г�л:л.::л: '.�f.,::::. .::F.;,f . . .. . .. . :w: v: и. . ... и..::г.. .....: :.... .....г. :..-........л?:: :.:г ::........ ..................f..... ............................ ;........ ., ..,.. . :.:.::.::r :цг:�:: :.•л:..vлч..i'л.: ч\.:гw: w::::лли: :{{ti•iiiгil:.• ': •. :.ч. . . ....: .. . .. . ........... ::. г.::: v•н:?.:.ич: ч: ::w:: •.v v w::: л4. : �{ . } ч$:}'г.iг i:.гг:: ::гiгг: i:i4i::::{:iii г:�гi'fг:}i'гi:iл: :гл::г:i гi};::{:.:: iv<: г::::: л:. {':.:'i: �.'ггг:.:i{} ::::; ::i �::::�:iг ::iW::ii.v: г::: ..... ;л....; л, л.-ч. .. ..::: . . � :..::::л':'::.:{:: v:.:....vn:.. .............г...:.\. �. . �.и. .�1.�'�. 8 : .:«::.>;•{•::,:<{��:::�:.::: :::.:::::.:.::{•::>::::-:::�..:.::.:::....................ч•:.:::.::::.:::.:::.:::.. :.::.:.::,;и.:.::::..�:•;:::::::::::::.и::«•:т::��.::>::•:.,;. <::.zs>s»»s•::> ... .... . .. ..................: ........................: .: ...: .: ... � • .��::::и�..:::::�.::�.�.:.:4•::�.и :::.::::::::::::::::.и:�:.�:.:.,и:::.ч•.::..:.:::::,:::::•.::::::::.•-,.;.....:::::::..:... �ц:•:w:..::::::v:: � f•�.ч ,. :.::::: :{•::: •::::�..:.и:: {.::.: •::ч...: :::: :•:,:.::::�::::::>:.и.г:.:. ::и:.и. :ч• :.:..:::.::::::::::::: •:: :,.:?:{.::::.ч• :•;:•::�:::.::::г;,:.:.:::::;.;.:;.:::;'•::•У:•::.,;>..:::::•:>:::: {:::•;•:..{:.;::.::.:.:{.;г:.:.:.. ;�.....:- •: ..�н::�:.�4.чи•::и:::::::н.и:::::::::н..:..,.: .и:: � ::::::::::::::::: и:::..::..::.•r::..,ч�и:::н:: :•.:::::::::::::::л..:::н:.и. �.и::: и: :.::::::.::..:.::.::::::::::. и: :й::::::: йи::..,- {{.:л:{.:н:йw. ........... ....... ч............................г.лR...л .................................. .. .. . .. ...................... .....4..... .. .:.. . ........ . г: :и::: �:::: :и:::: .-:: н::::: �.: н: н::.ьгнч ч 4•{;:vr:•кйг:Lйгг::й:::{ {?::S:a:::: �. .:; .. ...........� й.. ..: ..:::::::::..::::::: :•::: :•::::.ччч.: �:::::::.::::::::::.:::::.и::г:l:::•.чйг..:. •:::::::: •::::: �:: ::.::::::::ч.и::�,. ...4w,.v.... .4..чи.\ йг\ : \44 ч.• ................ . 4.. .. �.. . ...� ;:::.У.:'. :.г•iг:{•:, ..... У ..4.... .. . м..мv:н::::::::г ::::•::и::::::n: :ч. :.г....... ч....... ..л ............ ..............:и:::::г:::г....ч... л.. чv.и:.•г:::::::г.и:: ::w::г ...лн.. ч .::::::..:::.i:'::•:::йгг:•::fv.... ::..нлл.•i:::::::::::�::::•:..:::У••:.г'.i?г::г..::р:::ч•}гг::гii}г:ли::w:ж::::г.и.{г.44ww:л:и::w:::::w:::iг:гч .:::.л::v.........г ч и:г. ..v:.....:..n.. . .. . . . . . .....чччи:.�:. ........ ......... и::гл\С. ..k.�:::::лvи:: .. .... . ..fг: f:лv .::fi:h£'.:гггг: •г:и:.гh{::v'�.г:•}iS . ... . v: •:::::::::. v� :.:. и:::.Ч{::.: v{гг: v::: �.. ...........г:::.и...:.г::нлчч{ч{чл..ч4ь.лч..-::::::.й::л::и::: •.�:ч{'i:.+f.. Bank staff inembers asaociated with the project have suggested that 1оса1 autonomy апд resources, combined with high-quality 1оса1 adminiяtration, were among the key factors that facilitated both the active participation of farmers апд the overall effectiveness of the project. In this саве, devolution of authority had а positive impact partly because the project was implemented and managed entirely through 1оса1 government structures with а clear mandate. Project documen�s апд task managers alike attribute а large measure of the project�s effectiveness to the fact that, цnlike тапу Bank-suppo�ted rural deveiopmen� projects, The Sri Lanka Second Rura1 Development Project did not create а para11e1 structuкe through NGOs or other intermediary organizations. This pкoject highlights the question of whether alternative institutions and organi2ations should be created to implement community development programs, or whe�her the goal should be to link the poor to existing governmental institutions and services. Many of the proponents of participation by the poor, both within the Bank and external to it, assume that the most appropriate vehicles for organizing and involving communities are 1оса1, national ок international NGOs. NGOs are we11-positioned to undertake 1оса1 participation initiatives, since they can work intensively in а sma11 area апд as а result they often have extensive knowledge of the specific proЫems faced by the poor in these areas. Some of the most celebrated community mobilization projects in South Asia are the products of NGO activity. Nevertheless, 1оса1 government сап also work оп а sma11 scale, in some instances successfully. Under the Bank-aupported Mexico Decentralization апд Regional Development Pкoject, funding has been таде availaЫe to 1оса1 municipalities for саккуinд out sma11-scale community-based infrastructure and productivity projec�s that have been identified by 1оса1 peasants as high priority inteкventions. This project соиlд be viewed as а test of а тодеl that, if successful, can be кeplicated in other regions, including South Asia. - 30 - The advantage of utilizing local governments, parastatals and rural development banks as intermediaries is that, because they are subsystems within larger national systems, expertise and experience that is acquired in the course of implementing small-scale pilot projects can later be utilized ac the national level for scaling up and replication of successful programs. An experimental Bank-supported multisectoral program that is now ongoing will test the use of a variety of non- governmental organizations, local government organizations, rural development banks and other institutions as intermediaries to implement a package of poverty reduction activities, each in its own location: :.SOX IS* R&INE _X QRGAMA= 5 R( M1 "NEA'91AN RUST &JMIYA X__ The.:, Janasaviya...Trus!t Fund was establishe& to provide services to the poor --are P -.of - p lie works,, i4ioved.. " trit on, human resource development `d::Adc6s s,.: t,6.. ctedit. At the- oc'al-.ieve1,, dommun ti;es ate being Mo e_ amomr., %460 0 'CoinpooeAts - - an array of .part.er .: -orqanizatipns .(RO) These Pos.include local gove"rnment, XG0s, cooperatives, : 4hd % f lnanoii l -instituti 45i org"anizatio -can become a P0 if it has.1). -1 oz wes 0 tb t, 2) exp rienc6:ih prSantzing. cons steM.With-those f- 0:-p.:rq ec 'cial management. . oa a y. dmmn ty. it An"Advisory 36a'id., 'Of. 9006 -and 009plefi or ations.:--tnf6rM the D :r9p 0",r4b ui- and ob and, ru. tee Pr. Xems., 'Offtativenesa, and, "$at oilitY, of' ongoing programs at the, Although this project experienced early difficulties due to its political environment, and results are as yet unknown, its final evaluation may shed some light on the question of which type of intermediary is most effective and under what kinds of local conditions the various types of intermediaries function best. Governmental structures are not likely to be as effective as NG's at stimulating local participation, however, unless local autonomy and the funds to support it are made features of governmental programs aimed at involving the poor. This is not to imply that local governments and NGOs should be rivals. In many cases, task managers have found local governmental structures to be hierarchical, dominated by wealthy elites and uninterested in working with the poor. NGOs can perform an essential role in these settings, in so far as they alone may possess the motivation and skills required for fostering grassroots involvement. NGOs can also work with and advise local governments in order to make public sector plans and programs more responsive and accessible to the poor. Although Pakistan's Orangi Pilot Project (OPP) is not a Bank-supported project, it is utilizing a model of NGO-local government cooperation that offers -useful lessons for the Bank. OPP facilitates the installation and maintenance by users of water and sanitation infrastr.cture in urban slum areas. This user-created infrastructure would not solve the solid waste disposal problem if these local - 31 - systems could not unload into larger urban sewer systems. OPP collaborates with local government in establishing these connections. The same model is operative in OPP's approach to improving access to human resource services, in that OPP has established new and stronger links between the urban poor and the government facilities -- such as immunization programs and schools -- that are supposed to serve them. Instead of creating alternative service systems, OPP has assumed a complementary role in relation to local government. As a result, conflicts of interest and duplication of effort are minimized, while participation of and accountability to the poor are enhanced. NGOs such as OPP are well positioned to monitor government service providers' performance and to insure that users are actually receiving the services that they are legally entitled to receive. The Role of the World Bank 13. Although it has been suggested that the World Bank's initiatives are too large to be tailored to each local community in which it operates, and thus the Bank is best suited to working at the "wholesalen level - - to providing resources and guidance for putting participatory initiatives into motion -- it may also be necessary for the Bank to work at the "retail," or local, level (In collaboration with NGO/CB0s) in order to acquire the necessary knowledge for informed decision- making. The community of opinion in the World Bank is still in the process of defining the appropriate role of the Bank in fostering community-based and participatory interventions. According to its mandate, the Bank must work in collaboration with member governments. What is more, Bank projects must satisfy a minimum size constraint in order to absorb the overhead costs of an international organization based in Washington. The usual scale of Bank operations is, therefore, large. Participation of the poor must take place at the actual site of poverty-reduction activities -- in villages and urban slum communities -- and so interventions using this approach are necessarily small- scale and localized. How, then, can the Bank reach down to the local level? This discrepancy in scale raises a number of questions. Many of the important issues have been discussed in published World Bank documents, while others are being identified as questions arise in the course of project implementation. Both informally and in forums such as the October 1993 Workshop on Poverty Reduction in South Asia, staff members involved in the South Asia Region's operations at all levels have begun to ponder the question of the Bank's comparative advantage in relation to other organizations active in promoting community participation. Comments by staff members on the Bank's relationship to other organizations emphasize the fact that the World Bank is the single largest development organization in terms of the amount of funds transferred. It was suggested that 19\ Bhuvan Bhatnagar and Aubrey C. Williams, eds. Participatory Development and the World Bank: Potential Directions for Change. World Bank Discussion Paper #183. 1992. - 32 - bilateral donors can and sometimes do select a specific sector or subsector in which to invest intensively, at times ignoring the overall development requirements of the country. The World Bank, by contrast, is responsible for assisting countries to achieve their overall development goals. Investment in specific sectors, no matter how critical from the humanitarian point of view, must always be viewed from the perspective of how they fit within an overarching plan. Given these constraints, one task manager suggested that a suitable division of labor might be for bilaterals to fund small pilot programs that test various models for community involvement and participation in the sectors that interest them. The World Bank, with its greater resources, could use the results of these pilot programs to identify models that merit expansion across sectors or scaling up to the national level. Through its economic and sector work, the Bank would be well positioned to propose effective models to governments and to help them to recognize the appropriate contexts and mechanisms for involving local communities in poverty reduction activities. The World Bank, acting at the meta-level but working closely with governments, NGOs and other donors, can act as national or regional coordinator and clearinghouse for these types of initiatives. It can collect information and examine results from ongoing experiments using local participation, deploy these insights in the context of larger poverty reduction plans, and pinpoint gaps in current knowledge. Not every important issue has been addressed by research or experimentation. From its broad overview position, the World Bank can identify and, in some cases, begin to marshall support for, experimentation on problems and issues that have been overlooked (see "Suggestions for Further Experimentation," below). When the World Bank is acting as a "wholesaler" of development initiatives, it may find the small community-based organization or local NGO to be an inappropriate partner due to coverage limitations. Instead, the Bank may find that it is able to achieve wider poverty-reduction effects by providing support to apex or seed organizations (see the discussion of "nested" management under Proposition #9, above). These organizations, in turn, can offer a prototypical model for local interventions and they can select, monitor and provide training and technical or funding support to these small-scale initiatives. An example is the Grameen Trust, which provides support and guidance worldwide to projects based on the example of Bangladesh's well-known Grameen Bank. Ne*,ertheless, it should be noted that opinion within the Bank is not uniformly supportive of the "wholesaler" proposition. Information acquired through innovation and monitoring at the "retail" level can be a critical input into the Bank's higher-order decision-making. What is more, some task managers maintain that in order to execute project management effectively, they may in fact need closer familiarity with conditions at the sites of project implementation than they are now able to have. Those who hold this position have raised the issue of whether a different and more decentralized deployment of staff would not be better for the management of programs tailored to the expressed needs of specific local or even national constituencies. If project management functions were based in-country, the task managers might gain a better underst-.nding of the conditionr and stakeholders that will determine the success or failure of the project, and they could also build capacity among counterparts - 33 - and transfer the skills that would improve the project's potential sustainability. This implies an expansion of the size and mandate of the resident missions. Under this scenario, task managers would be based in the resident missions and a greater number of borrower-country professionals would also be attached to them. Alternatively, task managers could be based in regional offices that would place them within a few hours travel time of the project country. Regional offices would facilitate regional-level planning and could be the base for technical staff who would provide short-term assistance throughout the region as needed. It is important to ncte, however, that the cost-effectiveness of a dramatically increased field presence is still an unresolved issue. At present, a committee within the Bank is examining the advantages and disadvantages of expanding the Bank's resident presence in borrowing countries. An in-depth study of the experiences of the bilateral donors and other multilateral organizations with these and other staff-deployment alternatives could also be a useful adjunct to this debate. The "wholesale" and "retail" points of view can be reconciled, then, if we concede that the World Bank possesses the capacity to work at both levels simultaneously. The Bank can work at the wholesale level, developing coherent programs that are based on a clear conception of what is required for effective use of innovative and participatory approaches, while cooperating with partner and intermediary (including local government and NGO) organizations and consultants to carry out the on-the-ground work of community mobilization. At the same time, management of ongoing projects is the clear responsibility of the task manager, who must be given the institutional support necessary to enable them to function in an innovative and flexible manner. World Bank work at both of these levels will be stronger if information and feedback loops between policy and operational staff within the Bank are established or, where they exist, tightened. The results of this review suggest that, while task managers are aware of the priorities of senior management, there may be fewer conduits for feedback whereby the experiences of task managers can be used to sharpen the understanding of senior management policymakers. For example: - 34 - BJOX 19: PARTICIPAVTON~ REQUIRES UJP-V1RONT INVESTMENT Taskc managers are aware that Bazk policy fao the~ inciein of community participation components i project design where it is apropriate, and some of them have aepted this callnge wit entusam. At the sam~e tiwe, many have observed that this requiires that adeqate time and resources be tansferr~ed if participation is to be proerty iutiiked 4uing the project preparation process to shape the proposed. intervention to~ the needs and expcttinsof clien~t communities. Thr is genexal agreement to the effect that participation cannot be fully.effective if it is added on as an afterthought with noc additionial staff time, prjet prpartion~ time or funding~ to support it . Guggestions for. copi$ng with this problem have the additional costs and. time~ rquired, lenthening the project cycle, m iarsallixg bilateral trust funds where avalabe- for this pturpose, and satregthning the incentive struncturs w3iin the Bank to reward task a-el% .. .be .... _X; Resources of this density are not likely to be available across the board. While resources are limited, it may impossible and inadvisable to suggest that community participation must be cbligatory for each and ever-y project. To do so would be to encourage a superficial and cosmetic use of the concept of participation. Instead, it is important to carefully identify, on a case-by-case basis, the projects for which participation by the poor is most strongly indicated, and to allocate sufficient staff time and funding resources to implement the approach properly and thoroughly in these projects. If, as the results of the Learning Group on Participatory Development would predict, these projects are demonstrated to be more successful than conventional projects, then they may serve as models for a wider array of future projects. The recently announced Fund for Innovative Approaches in Human and Social Development will provide some resources for operational support and capacity building in World Bank operations that promote participation. The Fund will also be used 'for project-related social assessments. Within the Asia Technical Department, the Human Resources and Social Development division is conducting discussions with country departments in the region to identify "flagship" participatory projects in which to concentrate Fund resources. It should be noted, however, that without changes in the project cycle that leave space for more intensive involvement of stakeholders at all levels, additional funding may not be effective in providing an opportunity for a demonstration of the impact of participation on investment performance. - 35 - Sugqestions for Further Experimentation This discussion of the South Asia Region's experience with community participation has centered around a series of stylized propositions. Although many of these propositions would appear to be supported by the preliminary results of Bank experience with community participation in South Asia, evidence on others is absent or inconclusive. Although the range of topics for future experimentation should be broad, and the present exercise is in no way definitive of this range, the most controversial and unresolved of the stylized propositions could be used as a starting point in the effort to construct a research agenda on participation. Some of the following might be included in this agenda: Accountability to Beneficiaries: Methods of insuring the accountability of public services could be made more sustainable if service users were enabled to take responsibility for monitoring. Innovative project designs wherein community-level service users are able to monitor progress toward stated development goals, track the movement of funds, register effective complaints in cases of abuse and initiate corrective measures where needed should be tested. The use of group process and other means of achieving greater transparency in planning and negotiation of agreements could be tested in the context of Bank projects. Economically Homogenous Groups: The decision to include or exclude the middle class and local elites when developing interventions that employ community participation must be made on a project-by-project basis, since the above discussion suggests that in the context of some types of projects, their input can be useful to the poor and in others it may be detrimental. A typology of such projects could be developed as a guide for task managers on this issue. A useful study might compare the advantages and disadvantages of homogeneity, particularly in the context of interventions that would be of benefit to all community members irrespective of economic standing. For example, a water and sanitation project could compare the effectiveness of a community mobilization intervention that restricts participation to the lowest income decile with that of an equivalent intervention in an adjacent community wherein all residents are invited to participate. Indigenous Organizations: Clearly, some culture areas within South Asia display greater propensity toward spontaneous formation of local organizations than do others. Do these areas display an "organizational potential" that predisposes them to successful participation? Do areas with little history of indigenous local organizations require a different or more time-intensive approach? Relevant experiments could include pilot studies in which the cohesion and sustainability of community mobilization efforts that have worked through pre- existing groups are compared with results from communities that are similar in other ways but where community groups have been organized by "outsider" NGOs. Comparisons of local-level organizational forms in South Asian countries with those of countries in other regions (as Robert Wade has done for the irrigation sector in India and Korea) could also shed light on the issue of organizational propensities and potential. In addition, because there is little data on indigenous groups in the development literature, a systematic inventory, categorization and mapping of traditional organizational forms in South Asia could provide a resource for task managers seeking to involve existing local associations in development initiatives. - 36 - Single-Purpose vs. Multi-Purpose Groups: The question of whether single-purpose or multi-purpose community groups are most likely to be effective in lifting participants from poverty could be explored in a variety of ways. Projects that have organized effective single-purpose groups could test whether, by expanding the goals and activities of the groups, the poverty-reducing impact accelerates. In an effort of this kind, the effect of adding extra functions on group cohesiveness and cooperation should also be tracked. Alternatively, existing NGO and public sectors of the two types could be compared in terms of cost effectiveness and impact. Scaling tip and Replicability: Ongoing or planned efforts to test the replicability of established small programs (such as the Grameen Trust proposes to undertake) should be followed and controlled experiments initiated to test the impact of the transplanted programs in their new settings. Projects attempting to transfer insights and methods from successful NGO programs to national-level public sector programs (such as Pakistan's NRSP) should also be evaluated using a rigorous comparative methodology. The basic question to be resolved is the following: is density of resources the single key to success or are the essential elements of success in resource-rich NGO programs transferrable to relatively resource-poor public sector ones? Evaluation of Effectiveness: Identifying the most successful projects for possible replication is dependent upon our ability to monitor and evaluate, but the most appropriate indicators for measuring the effectiveness of participatory projects and programs are still to be determined. To begin identifying a set of tools appropriate to the purpose, selected pilot projects that involve participation could be designed so as to include controlled studies of the impact of those projects on beneficiaries. These studies would establish baseline measures of the economic well-being (or health status, educational attainments, etc., depending on the project's goals) of the target communities and of participating households; and also of a control population who will not participate in the project. During the life of the project, mid-course corrections should be based on assessments of progress that are monitored in relation to pre-established benchmarks. Determination of overall effectiveness should be based on a comprehensive final evaluation in which gains made by project participants are compared both with their baselines and with gains made by the controls. The goal of these studies would be to establish protocols for best practice in monitoring and impact evaluation of participatory projects. Data Bank and Information Exchange on Participation: In its final report, the Participatory Learning Group has suggested that electronic data banks and networks on participatory development would support an exchange of information and experience on a continuing basis both within the Bank and with NGOs, governments and local groups. Within the Bank, a participation database could collect evidence on the usefulness of various models and approaches, and could be a resource for identifying the factors contributing to the outcome of participatory interventions under various conditions. It would also serve as a clearinghouse for inormation on existing community groups, NGOs, local consultants with community development expertise, and other resources task managers might wish to tap when designing and implementing new initiatives. - 37 - LIST OF PROJECTS DISCUSSED ANNEX 1 Bangladesh General Education Project Coastal Embankment Rehabilitation Project Participatory Forest Development Project India Fifth Population Project Tamil Nadu Integrated Nutrition Project National Social Forestry Project Uttar Pradesh Basic Education Project Uttar Pradesh Sodic Land Reclamation Project West Bengal Forestry Project Nen 1 Agricultural Extension II Project Bhairawa Lumbini Groundwater II Project Rural Water Supply and Sanitation Pakistan Baluchistan Community Irrigati.on and Agriculture Project Baluchistan Primary Education Project Environmental Protection and Resources Conservation Project Integrated Hill Farming Project Northwest Frontier Province Community Infrastructure Project On Farm Water Management Project Second Karachi Water Supply and Sanitation Project Sri Lanka Community Water Supply and Sanitation Project Janasaviya Trust Fund Second Rural Development Project