51410 Paper No. 114 /April 2009 Mobilizing Rural Institutions: A Comparative Study of Rural Institutions for Improving Governance and Development: Afghanistan, Ethiopia, India, Vietnam, and Yemen Arun Agrawal & Nicolas Perrin SOCIAL DEVELOPMENT WORKING PAPERS Paper No. 114/April 2009 Mobilizing Rural Institutions: A Comparative Study on the Role of Rural Institutions for Improving Governance and Development in Afghanistan, Ethiopia, India, Vietnam, and Yemen Arun Agrawal and Nicolas Perrin This Working Papers Series disseminates the findings of work in progress to encourage discussion and exchange of ideas social development issues. The papers carry the names of the authors and should be cited accordingly. The series is edited by the Community Driven Development and Local Governance team in the Social Development Department of the Sustainable Development Network of the World Bank. This paper has not undergone the review accorded to official World Bank publications. The findings, interpretations and conclusions herein are those of the author(s) and do not necessarily reflect the views of the International Bank for Reconstruction and Development/ World Bank and its affiliated organizations, or its Executive Directors, or the governments they represent. To request copies of the paper or for more information on the series, please contact the Social Development Department Social Development The World Bank 1818 H Street, NW Washington, DC 20433 Fax: 202-522-3247 E-mail: socialdevelopment@worldbank.org Printed on Recycled Paper Table of Contents Executive Summary.................................................................................................................................................. iii Foreword ................................................................................................................................................................ vii Acknowledgments..................................................................................................................................................viii 1. Introduction.............................................................................................................................................................1 2. Research Questions and Approach ......................................................................................................................3 2.1 Four Key Questions......................................................................................................................................3 2.2 Two Sets of Factors Determine Institutional Effectiveness.....................................................................3 2.3 Research Approach .....................................................................................................................................4 3. The Research Framework ......................................................................................................................................6 4. Study Context and Methods..................................................................................................................................8 4.1 Study Context ...............................................................................................................................................8 4.2 Research Design and Methods ...................................................................................................................9 5. Membership-Based Rural Producer Organizations .........................................................................................11 5.1 Field Research Findings for RPOs............................................................................................................11 5.2 Lessons from the Comparison of Membership-Based Producer Organizations ...............................14 6. Natural Resource Management Institutions .....................................................................................................17 6.1 Field Research Findings for NRM Institutions.......................................................................................17 6.2 Lessons from the Comparison of NRM Institutions..............................................................................18 7. Local Governance Institutions ............................................................................................................................21 7.1 Field Research Findings for Governance Institutions ...........................................................................21 7.2 Lessons from the Comparison of Local Governance Institutions........................................................22 8. Major Research Findings .....................................................................................................................................25 8.1 Inclusion, Accountability, Participation, and Adaptive Capacity: Relationships Among Different Aspects of Governance ..............................................................................................................................25 8.2 Governance, Livelihoods, and Equitable Development ...........................................................................27 8.3 Context, Institutional Linkages, and Outcomes.........................................................................................30 8.4. External Interventions and Institutional Performance ............................................................................31 8.5 Research Findings and Territorial Institutional Development ................................................................34 9. Implications of Research Findings for World Bank Projects...........................................................................36 9.1 Strengthening Local Governance will Require Greater and Better-Directed Efforts .......................36 9.2 Promoting Better Institutional Linkages is a Key New Area for External Interventions ................36 9.3 Partnering with Existing Institutions versus Creating New Institutions ..........................................37 9.4 Assessing the Relationship Between Governance and Livelihoods ....................................................37 9.5 Assessing the Institutional Context .........................................................................................................38 9.6 Promoting Local Institutional Autonomy and Adaptability................................................................38 10. Conclusion ...........................................................................................................................................................40 Glossary of Key Terms.............................................................................................................................................42 i References .................................................................................................................................................................44 Appendix 1. The IAPA Governance Index...........................................................................................................51 Appendix 2. Institutional Network Analysis and Mapping..............................................................................55 Appendix 3. Membership-Based Rural Producer Organizations in Ethiopia, Vietnam, Yemen, and Orissa (India)..................................................................................................59 Appendix 4: Natural Resource Management Institutions in Ethiopia, Yemen, Orissa, and Vietnam ..........65 Appendix 5 Local Governance Institutions in Afghanistan, Ethiopia, Vietnam, and Yemen.......................70 Appendix 6. Data for Assessing the Overall Relationship Among Selected Factors (Based on Institutional Case Studies for Study Countries) .....................................................................................................75 Boxes Box 1. Rural Organizations at the Intersection of the Market, State, and Civil Society ....................................5 Box 2. Taramessa Multipurpose Cooperative in Ethiopia...................................................................................14 Box 3. Agricultural Cooperatives in Yemen..........................................................................................................16 Box 4. Gender and CDCs in Afghanistan ..............................................................................................................23 Box 5. Contributing to Livelihoods Despite Low Accountability: Tribal Structures for Water Use in Yemen ................................................................................................27 Box 6. A Tale of Two Water Users' Councils in Orissa (India)...........................................................................29 Box 7. Local Forestry Institutions in Vietnam.......................................................................................................32 Figures Figure 1 Institutional Analysis, Governance, and Livelihoods Framework......................................................6 Figure 2: Governance Outcomes.............................................................................................................................25 Figure 3: Governance and Livelihoods Outcomes ...............................................................................................27 Figure 4: External Interventions, Institutions, and Outcomes............................................................................31 Figure 5. Territorial Institutional Development ...................................................................................................35 Figure A2.1 ................................................................................................................................................................57 Tables Table 1. A Typology of Local Rural Institutions.....................................................................................................4 Table 2: Basic Information on Study Locations ......................................................................................................8 Table 3: Case Study Institutions in the Selected Countries...................................................................................9 Table 4. Membership-Based RPOs, Their Context, and Their Outcomes ........................................................12 Table 5. NRM Institutions, Their Context, and Their Outcomes.......................................................................18 Table 6. Local Governance Institutions, Their Context, and Their Outcomes..................................................21 Table 7. Relationships between Different Aspects of Governance ....................................................................26 Table 8. Associations between Governance and Livelihoods Outcomes ..........................................................28 Table 9. Associations Between Institutional Linkages and Livelihoods Outcomes .......................................31 Table A1.1 Calculating the IAPA Governance Index Score ...............................................................................54 Table A2.1 Checklist of Institutions According to Formal/Informal and Public/Market/Collective Dimensions......................................................55 Table A2.2 Hypothetical Data for Institutions in a Given Context ...................................................................56 Table A2.3 Hypothetical Governance and Livelihoods Scores for Institutions ...............................................56 ii Executive Summary Rural local institutions and organizations are central agents of development. They support local livelihoods, contribute to the quality of everyday democratic participation, and are the means through which central governments and donors channel resources for local development. A central question concerning their function and performance is whether there are ways to help them improve rural livelihoods and local governance at the same time. This report presents the findings of a multi-country project in five countries--Afghanistan, Ethiopia, India (Orissa), Vietnam, and Yemen--that addresses this question. The TFESSD-supported project partnered with in-country research organizations to study the role of local institutions in improving governance and livelihoods outcomes, particularly for more marginal and disadvantaged populations. Local institutions in three different domains of governance and livelihoods were the focus of the research: · Agricultural services; · Natural resource management; and · Local governance. The major findings of the report concern four critical areas of performance of rural institutions: · The relationships among different measures of governance, specifically inclusion, accountability, participation, and adaptability of local institutions; · The relationship between governance and livelihoods outcomes in given socioeconomic and political contexts, including equity outcomes; · The role of contextual factors in the performance of local institutions; and · The nature of external interventions necessary to improve the performance of local institutions, particularly with respect to improvements in the ability of poorer and marginal social groups (such as women) to improve their ability to participate, hold institutions accountable, and improve livelihoods. I. Relationships among Major Dimensions of Local Governance: Inclusion, Accountability, Participation, and Adaptability 1. The study finds that institutional performance related to local governance should be assessed along multiple dimensions. The studied local institutions have varying levels of performance along different dimensions of governance, such as inclusion, accountability, participation, and adaptability. This is because a) different dimensions of governance differ in what they require of institutional capacity; b) there are tensions among the four dimensions of governance; and c) different dimensions of governance vary in the benefits they provide to those served by institutions. 2. Specifically, higher levels of inclusion require significant outreach and openness on the part of institutions to meet the variable needs of different social groups. Accountability requires that institutions have appropriate procedures in place that help participants gain information and demand better performance from decision makers. Participation is driven by a variety of factors, but unless participants get material benefits from participation, they do not get involved in activities that demand their time and resources. Finally, adaptability requires expertise in long- range planning and understanding of changing conditions. 3. Empirically, the study finds only a very small association (positive or negative) between inclusion and participation, and between inclusion and accountability. Participation and iii accountability are strongly related and it is likely that promoting higher levels of participation will in time lead to demands for greater accountability. 4. Finally, studied institutions had low adaptive capacity in general. Their capacity to address new environmental challenges was very small, but they had somewhat greater ability to address expressed needs of their constituents, even if such needs were not identified when institutions were created. Recommendations for World Bank Projects These findings suggest that Bank projects aiming to improve governance and livelihood outcomes by attending to local institutions need to pay attention to multiple institutional requirements. Provision of more financial resources by itself can serve to improve inclusion, but the following are also crucial to address the different dimensions of governance performance: incentives to promote information dissemination and uptake about the services institutions provide; training of institutional decision makers in creating accountability mechanisms and linking their remuneration to greater accountability; and finally, help in building planning capacities. II. Governance, Livelihoods, and Equity The study has three major findings on the relationship between governance and livelihoods outcomes. 1. Stronger governance is not necessarily associated with increases in livelihoods. Institutions with high performance on governance indicators tend to improve the quantum of livelihoods benefits in some cases, but in other cases effective governance does not translate into higher levels of livelihoods. 2. At the same time, we find strong evidence that better performance on different aspects of governance (especially accountability) is positively associated with greater equity in benefit distribution. The positive relationship between governance and more equitable distribution of livelihoods benefits is particularly evident in relation to downward accountability. High level of downward accountability to members is directly related to more equitable allocation of benefits. Explicit incorporation of downward accountability mechanisms in institutional design has, however, been rare because governments tend instead to institute upward accountability mechanisms to maintain control over local institutions. 3. The lack of equal benefits for women is a particularly unfortunate consequence of the unequal nature of involvement in governance. Most of the institutions on which the study focused show low levels of inclusion of women. Indeed, this observation is also true for other dimensions of governance and livelihoods. Women's participation in institutional decision making is consistently low and institutional accountability to women members is almost nonexistent. The absence of women from most governance functions is also reflected in the low levels of benefits distributed to women from local institutions. Part of the reason for the invisibility of women in institutional relationships is structural--women suffer from various forms of marginalization in the areas in which the study was carried out. But another reason is the design of the studied institutions. Recommendations for Bank Operations Together, the above findings have a major implication for Bank projects. More equitable provision of services and benefits through local institutions requires institutional design that incorporates accountability procedures. Bank projects aiming for more equitable benefit allocation need to ensure that local institutions include accountability mechanisms through which institutional members and constituents can gain information about institutional decision making, and can sanction decision makers when institutional performance is below par. Improved accountability can be secured both through iv institutional means and through better communication of what constituents should expect of institutions. Competitive local elections are a common and practical means of holding decision makers accountable, but they need to be accompanied by better information sharing between institutions and their constituents. III. Social and Institutional Context and Institutional Performance 1. The chief finding of the MRI study in relation to institutional context concerns the nature of institutional linkages. We find a clear and substantial relationship between stronger institutional linkages and improvements in livelihoods outcomes. But effective local institutions and institutional partnerships within and across levels are relatively uncommon. They are uncommon because such partnerships require overcoming serious challenges related to a) the tensions between poverty-stricken, hierarchical social contexts, equitable service delivery, and short-term and sporadic external support for local institutions; b) the tradeoff between inclusion and representation (since higher levels of inclusion can make it more difficult to coordinate stronger participation); and c) the conflicting demands of greater inclusion and more effective delivery of services (including enhancement of livelihoods). Overcoming these tensions is necessary because stronger institutional linkages play a positive role both in enhancing livelihoods and in improving the distribution of livelihoods benefits. Based on the findings of the research, we suggest that institutional linkages can compensate for missing competence or resources on the part of a given institution. Recommendations for Bank Projects The implications of these findings for Bank projects are twofold. · First, local institution-focused Bank projects need to incorporate analytical methods that can systematically assess the nature of institutional linkages in project areas. One promising approach is Institutional Network Analysis, a technical tool developed by the study based on social network analytical tools. · Second, Bank projects need to help incentivize better connections among local institutions to improve the links not only between localities and higher levels of administration but also within the locality. In building such connections, it is important to identify which institutions are more inclusive of poorer, more marginialized groups in a given site. IV. External Interventions and Local Institutions 1. The major finding of the study in this regard is that there are major gaps in the existing capacity of local institutions and the extent of support available to them. Effective external support to local institutions can come in the form of financial and technical contributions, but can also occur through the facilitation of stronger links between relevant local institutions, government agencies, and other local institutions. The study finds that consistent and long-term support to local institutions--both when new institutions are introduced and when development policies aim to work through existing institutions--is of critical importance. This is in part the result of the low existing capacity of most local institutions in the developing countries where the studies were carried out, but also because existing institutional arrangements often reflect existing power relationships and changing them requires a long-term commitment on the part of external actors. Concerns about external support creating dependency are misplaced because the existing capacity of local institutions is extremely low. v 2. Further, institutional linkages with different social groups at the local level determine how institutions channel external support. Local institutions have different types of linkages to various social groups. External interventions that channel greater support to more inclusive institutions or to institutions that are connected to disadvantaged social groups are more likely to generate equitable outcomes. Recommendations for Bank Projects We conclude this summary of the study by reemphasizing the overwhelming evidence of: · The fundamental role of local institutions in governing access to resources, shaping delivery of services, and determining the experience of governance; and · The need for much higher levels of investments in building the capacity of local institutions to achieve the promise embodied in them. vi Foreword This paper is a contribution to the TFESSD-supported Mobilizing Rural Institutions for Sustainable Livelihoods and Equitable Development Project conducted by the Social Development Department of the World Bank for the World Bank, Washington, DC. Arun Agrawal is a professor at the School of Natural Resources and Environment, University of Michigan, Ann Arbor. Nicolas Perrin is Senior Social Development Specialist at the Social Development Department (SDV) of the World Bank and the task team leader for this project. vii Acknowledgments This study is a component of a project implemented jointly by the Social Development Department (SDV) and the Agriculture and Rural Development Department (ARD) of the World Bank titled "Mobilizing Rural Institutions for Sustainable Livelihoods and Equitable Development". The task manage of the overall project is Nicolas Perrin (SDV). The supervisors of the ARD specific component "Development of a Benchmarking tool for Agricultural Innovation systems: Pilot Study in Vietnam and Ethiopia", Eija Pehu and Riikka Rajalahti (ARD) are acknowledged for their guidance and support. This study elaborates on the findings of 5 country studies and benefited from the comments and suggestions of their authors: Ethiopia (International Food Policy Research Institute, Addis Ababa. David J. Spielman, Marc J. Cohen,Tewodaj Mogues, Tanguy Bernard and Dawit Kelemework), Vietnam (Center for Agricultural Policy: Anh Tuan Nguyen Do, Dinh Tuan Minh, Tran, T. Ha, Hue Le), Yemen (Ismail Mohamed Sharief); India (Sutra Consulting: Sanjeev Sasmal) and Afghanistan (Afghanistan Research and Evaluation Unit: Nixon, Hamish). The authors would like to to thank and acknowledge the following for their guidance and support : Jeeva A. Perumalpillai-Essex (AFTAR), Achim Fock (AFTAR), Pierre Rondot (MNSSD), Marie Helene Collion (LCSAR) , Robin Mearns (SDV), Asta Olesen (SASDI), Maitreyi Das (SASHD), Catherine McSweeney (WWF) and Robert Chase (AFTH3) . Review comments from Jesse Ribot (University of Illinois), Regina Birner (IFPRI), Eija Pehu (ARD), and Stephen Mink(AFTSN) have greatly improved the clarity of the arguments. viii 1. Introduction Mobilizing local institutions for better governance, improved livelihoods, and sustainable development is a major challenge of the twenty-first century. Meeting this challenge is necessary because the quality and effectiveness of local institutions are central to development progress for hundreds of millions of disadvantaged and poor households (Grootaert 2001). Local institutions shape political action, social interactions, and economic choices and outcomes in the daily lives of millions of men and women. They are also the principal mechanism to channel external interventions, resources, and assistance to poor households, women, and disadvantaged groups in general. Local institutions--in the form of organizations (Uphoff 1997)--are particularly important for project-based development initiatives. They shape project implementation, access to resources of different kinds, and local distribution of external benefits. The findings presented in this report are based on a five-country research project on local institutions supported by the Trust Fund for Environmentally and Socially Sustainable Development (TFESSD). The objective of the TFESSD is to help the World Bank and its client countries better understand how rural organizations can be strengthened to contribute to sustainable rural development, overcome spatially induced exclusion in livelihoods through adaptation and partnerships with other rural institutions, and enhance technological and institutional change. This TFESSD-financed work builds on earlier work on local-level institutions (Helling, Serrano, and Warren 2005; World Bank 1996), particularly on the findings of the rural institutions stocktaking initiated by the Bank in FY06 (Perrin, Van Der Riet, and Sjorslev 2006; Uphoff and Buck 2006). This report analyzes local institutions and the role of institutional networks in shaping governance and livelihoods outcomes in rural locations in five areas that face significant governance challenges: Afghanistan, Ethiopia, Orissa (in India), Vietnam, and Yemen. In each location, the focus of the study is on institutions that take the form of organizations (see section 3) in part because a significant proportion of the Bank's development projects rely on local rural organizations as entry points for the intervention. According to one study, 61 percent of community-based rural development (CBRD) and community- driven development (CDD) projects use public or civic rural organizations as partners to deliver benefits, and another 23 percent rely on a combination of public/civic/private organizations (Perrin, Van Der Riet, and Sjorslev 2006, 21). The report focuses in particular on the impact of organizations on marginalized and disadvantaged groups such as the rural poor and women. Empirically, the report examines three domains of institutional functioning and impact: agricultural production, natural resource management, and local governance. Although its major focus is the role of local rural institutions and institutional networks in development, the report recognizes that institutions alone do not determine development outcomes (Crouch and Farrell 2004; Mahanty 2002; Roland 2004). Local institutions of different kinds always exist within sociopolitical and ecological contexts (Agrawal 2001, Agrawal and Chhatre 2006; Day 1994). The context in which they exist deeply influences their functioning and outcomes associated with their functioning (Sick 2008). The report focuses on how the capacities of rural institutions to improve livelihoods and reduce livelihoods risks can be strengthened. It also examines how disadvantaged rural households can adapt to livelihoods challenges with the help of rural institutions. Case study evidence is presented in the context of findings from other work supported by the Bank and the existing literature on local institutions. The purpose of this report is threefold: to contribute to a better understanding of how external interventions can promote inclusive, socially accountable, and resilient development; to examine the relevance of its findings to the Bank's development projects that focus on local institutions--in particular, to Social Development Department (SDV), Agriculture and Rural Development Department (ARD), and 1 Environment Department projects; and to apply its findings to a concept of territorial development in which local institutions and their interactions are critical for greater inclusion, accountability, and resilience. To examine the role of rural institutions in addressing development failures, our research also developed a set of preliminary methods that can help assess governance arrangements and livelihoods outcomes in the context of such arrangements. One of these methodological tools-- the IAPA governance index--is useful in the rapid initial assessment of the governance contributions of institutions. The second tool-- Institutional Network Analysis--is based on extensive existing work on social networks and network analysis, and focuses on the network relationships of local institutions to other institutions and major social groups (Entwistle et al. 2007; Scott 2000; Wasserman and Faust 1994). Descriptions of these tools are provided in appendix A1, The IAPA Governance Index, and appendix A2, Institutional Network Analysis and Mapping. 2 2. Research Questions and Approach 2.1 FOUR KEY QUESTIONS The report addresses four key questions related to local governance and livelihoods. 1. What are the relationships between different measures of governance such as inclusion, accountability, participation, and adaptability of local institutions? (Operational definitions of these terms are provided in appendix A1). 2. What is the relationship between governance and livelihoods outcomes in given socioeconomic and political contexts, including equity outcomes? 3. How does the social and institutional context of the studied local institutions affect their performance? 4. How do different types of external interventions affect the performance of local institutions? These four research questions aim to uncover how institutions affect governance and livelihoods outcomes given the micro- and macrolevel social, political, and economic context in which institutions exist, and how external interventions can improve the desired efficiency, equity, and sustainability outcomes associated with local institutions and governance. 2.2 TWO SETS OF FACTORS DETERMINE INSTITUTIONAL EFFECTIVENESS · INSTITUTIONAL DESIGN: One set of factors that affects institutional effectiveness comprises features of institutional design. Institutional design features refer to the different rules that define institutions. They shape how institutional actors make and implement decisions as well as social outcomes. · FEATURES OF THE MICRO- AND MACRO-CONTEXT: The second set of factors concerns the macro- and micro-context within which institutions exist and the problems they seek to solve (Mitchell 2003). Micro- and macro-context refer to the many cultural, economic, political, and social factors characterizing the location of institutions. The recognition of the role and influence of both institutional design and institutional context is apparent in a broad array of research on local institutions. A significant amount of work in the Bank's own research portfolio addresses similar issues (Lele 1981; Grootaert 2001). For example, in a review of the role of the Bank in improving agricultural outcomes in Sub-Saharan Africa, Cleaver argues for a consideration of both socioeconomic context and institutional design issues in promoting greater inclusion of farmers in decision making, development and distribution of new technologies, social programs, infrastructure provision, and improved management of natural resources (1993). Conclusions of a number of Bank- sponsored studies on social capital and local institutions show the role of both institutions and social context in shaping development outcomes (Bebbington and Carroll 2000; Grootaert and Narayan 2000). Similarly, CDD and SDV research recognizes the importance of community-level institutional arrangements, their linkages with higher-level organizations, and the need to examine the social and political context of communities.1 The case evidence from field research also shows that when institutions function well, are participatory and accountable, and are well connected to higher-level institutions and local residents, they generally help households improve their quality of life. But the effective functioning of institutions is in no small 1See, for example, SDV Sector Strategy (2004). Empowering People by Transforming Institutions: Social Development in World Bank Operations. Washington, DC: World Bank. 3 measure related to their social and institutional context. When this context is unfavorable, there is a strong need for effective external intervention to redress weak institutional functioning and performance. The research finds that an immense amount of work still remains to be done if external interventions are to help improve the functioning and capacity of local institutions. 2.3 RESEARCH APPROACH The study relies on the Institutional Analysis and Development (IAD) approach to guide the research. Institutions are at the center of this framework of analysis. Since the term institutions is used in two major ways, it is important to clarify the approach adopted in this study. Institutions often refers to roles or organizations--this is especially true of approaches inspired by organizational theory, public administration, and management. Another large body of scholarship, particularly in economics and in rational-choice approaches in political science and sociology, refer to institutions as rules. This latter view of institutions includes organizations, but also includes more diffuse and loosely connected groups. In this report, institutions are broadly conceived as rules that structure interactions and expectations by providing the incentives for human actions. At the same time, in examining the specific institutions engaged with rural development processes in the studied sites, the report focuses particularly on organizational arrangements. Accordingly, the report pays special attention to three types of organizational arrangements: rural producer organizations such as cooperatives in Ethiopia, Vietnam, and Yemen; natural resource governance institutions in India, Vietnam, and Yemen; and local governance institutions in Afghanistan, India, Vietnam, and Yemen. The rural institutional typology for this study is based on that developed in Uphoff and Buck (2006) and used in Perrin, Van Der Riet, and Sjorslev (2006). According to this typology, three major types of organizations can be found in rural contexts: civic, public, and private. Rural institutions have also been categorized in other ways--differences among attempts to classify institutions hinge on institutional design and performance aspects. Analytical approaches thus have focused, among other dimensions, on whether institutions are sector-specific or multisectoral/general purpose, and on their levels of hierarchy (IFAD 2003; Shotton and Winter 2005). Within the three broad domains of institutions identified by Uphoff and Buck, it is possible to recognize additional relevant distinctions. Focusing mainly on organizations because of their formal and concrete nature, Uphoff and Buck (2006) highlight two important types of local public institutions: local governments (organizations accountable to a local constituency through elections or some other mechanisms) and local agencies (branches or arms of higher levels of government operating at local levels). They similarly identify two types of civic institutions (membership organizations that function in a manner analogous to companies and advance some common interest of their members; and cooperatives, which function more like partnerships and help members pool resources for improved economic outcomes) and two types of private institutions (service organizations such as NGOs and charities; and private businesses). This typology is summarized in table 1 (adapted from Uphoff and Buck 2006). Table 1. A Typology of Local Rural Institutions Public (state) Private (market) Civic (civil society) Local agencies Service organizations Membership (bureaucratic) (charitable) organizations (common interests) Types of institutions Local governments Private businesses Cooperatives (political) (for profit) (resource pooling) 4 The fieldwork for the study covered all these different types of institutions, focusing in particular on public and civil society institutions, and examining their linkages where relevant to private institutions including nongovernmental organizations (NGOs) and private businesses. It should also be noted that rural institutions do not always fall neatly in one of the above categories. Specific organizations can straddle the boundary between civic and public, or public and private (Lemos and Agrawal 2006). Consider the case of cooperatives in Ethiopia (box 1). Box 1. Rural Organizations at the Intersection of the Market, State, and Civil Society Cooperatives in Ethiopia can be considered an example of an organization form that lies at the intersection of the market, state, and civil society, and of the interaction between formal and informal governance. They are established and organized for the explicit purpose of increasing smallholder production and commercialization--activities that rely largely on private entrepreneurship--but they also seek to promote and manage community-based collective action. Although they are often managed as private enterprises for the goal of generating a profit for members, they are also designed and managed on clearly democratic principles of governance appropriate to collective action-based organizations. They thus share many characteristics with nongovernmental, membership-based civil society organizations and with market-based private enterprises, and they have been used in the past as a means of extending state control and authority during the Derg regime (1974­91). 5 3. The Research Framework The basic IAD framework is outlined carefully in Ostrom (2005) and has been crafted over the past 20 years by a number of scholars (McGinnis 1999; Ostrom 1998; Ostrom, Feeny, and Picht 1988; Ostrom and Ostrom 2004). The framework that guides the ensuing case studies and analyses of local institutional arrangements, their functions, and outcomes is presented in figure 1 as the Institutional Analysis, Governance, and Livelihoods Framework. Figure 1 Institutional Analysis, Governance, and Livelihoods Framework Figure 1 combines the different elements of the social, economic, and institutional context, institutional design features, and governance and livelihoods outcomes. It suggests that the kinds of problems local institutions address and the context in which they attempt to do so have a significant effect on the ability of institutions to solve development problems. In turn, institutions are characterized by their design features, processes, and endowments/capacities. Their capacities, coupled with their context, affect their ability to produce governance and livelihoods results. The thicker and thinner arrows in the figure indicate that the effects of a range of different factors occur both directly (thin lines), and are also mediated by institutions (thick lines). One of the most important sets of factors that affect outcomes, particularly for policy action, is external interventions toward equitable and sustainable development. The effects of external development interventions--in the form of funds, technical inputs, new information (including training and skills enhancement), and new institutions--are mediated by existing institutions. New institutions are often created because of dissatisfaction with the performance of existing or traditional institutions. This is confirmed in our study in the case of governance institutions in Afghanistan, irrigation institutions in Yemen, and self-help groups (SHGs) in India. In these and other cases, the impact of new inputs and the creation of new institutions alike are shaped by existing institutions and their relationships to each other. 6 The four research questions for the study are encapsulated in figure 1. The first research question concerns the relationships among different governance outcomes, the second between governance and livelihoods outcomes, the third between context and outcomes, and the fourth on the role of external interventions in promoting better outcomes. In this sense, the framework provides a useful way to present the different areas of emphasis in the research as well as the findings for each of these areas (see sections 7 and 8). The report is organized through an analysis and synthesis of the findings in different institutional domains across the studied countries. Recall that the three institutional domains cover local institutions focusing on agricultural production (typically cooperatives, although in the Indian case, SHGs were studied), natural resources (irrigation and forestry), and local governance (both elected and appointed bodies). There is significant diversity in the types of institutions across these domains and, within each domain, among institutions in different countries. This diversity pertains to the features of institutions, not just to the many different features of their context. Efforts to analyze the differences in the functioning and outcomes of institutions clearly would be related to the variations in the contexts of these institutions; but given the relatively short periods of field research and quick data collection for this study, attempting to identify the specific contextual variables that introduce specific differences in institutional functioning would be very difficult. It is precisely this diversity, however, that forces generalizations related to the questions that motivate this study, which in turn allows broader similarities to emerge. Given the differences across the institutions and contexts, it is the similarities in observed patterns that are of the greatest interest for the study. 7 4. Study Context and Methods 4.1 STUDY CONTEXT Two different aspects of the context are important for the study: a) the macropolicy context of the countries where the study was carried out and b) the microsocial and economic context of the institutions whose functioning is being analyzed. 4.1.1 Macrogovernance and social context To examine the relevance of macrolevel context on outcomes, the study was carried out in five different countries representing significant variation in macropolicy and political environment, political system and organization, and social groups and their relationships. Some basic economic, demographic, and social information about the study countries (and the state of Orissa in India) is presented in table 2. The table includes macrocontextual information on the five countries initially included in the study. For a variety of reasons having to do with security problems for fieldwork in Afghanistan, research could be completed only for governance institutions in that country. The approach adopted in the research simplifies the research problem considerably by selecting different kinds of institutions in different contexts in different countries, and by focusing on common patterns between institutional design features and governance and livelihoods outcomes across multiple contexts. Table 2: Basic Information on Study Locationsa Afghanistan Ethiopia Orissa (India) Vietnam Yemen 24.8 M 79.1 M 37.5 M 85 M 22.3 Population (approx) Population 3.9% 2.5% 1.6% 1.2% 3% growth rate Rural 76% 85% 85% 74% N/A population as percent of total Per capita $377 $780 $710 $2550 $2200 GDP (PPP) GDP growth 7.1% 11.1% 4.9 8.5% 3.6% rate * Poverty N/A 44% (2000) 47% 15 45% Literacy 28% 42% 63% 92% 60% Life 43.1 years 52.5 years 61 years 71 years 62 years expectancy (2006) (2006) (2006) (at birth) *for rural population 4.1.2 Microinstitutional context The goal of mobilizing rural institutions to address the competitive disadvantages of rural areas also means that the study focused on a variety of different institutional types and domains. This is because of two reasons. The first is simple: many different types of rural organizations and institutions affect the a Most of the data in the table is for the latest available year (typically for 2007, unless mentioned otherwise) and is drawn from , the World Bank's website for country level economic and demographic data. The data for Orissa is based on the state government's information website: < http://orissa.gov.in/portal/default.asp> (Both sites accessed December 20, 2008). 8 lives and livelihoods of the poor. Focusing on just one type of organization therefore could provide a very one-sided, limited picture of how rural organizations can help improve livelihoods and governance. Second, the relevant features of natural resource governance organizations that affect performance can be quite different from those of cooperatives or local governments or traditional institutions. Thus, higher levels of inclusion may be detrimental to effective contributions to livelihoods for organizations such as cooperatives that aim to contribute to economic improvement. But higher levels of inclusion are critical in assessing the performance of institutions that provide governance-related benefits, such as local governments. To understand these differences, one must understand commonalities and differences across multiple institutional domains and their contexts. 4.2 RESEARCH DESIGN AND METHODS Table 3 lists the organizations and institutions on which the case studies concentrated in each country even as they examined the links of these institutions to all the other relevant local and external institutions. These institutions were selected on the basis of their relevance to local governance and livelihoods. For each institutional case study in a country, four different examples of that institution were selected. Overall, the study examined a total of 54 distinct institutions. The choice of different types of institutions ensured that the study would gain insights from the different processes at play in the different institutional domains. According to the classification of institutions in table 2, the studied cases are examples of either public (governance cases) or civil society (cooperatives and natural resource management [NRM] cases) institutions. Table 3: Case Study Institutions in the Selected Countriesb Country Name Institutional Case Studies Afghanistan 1. Community development councils (LG) Ethiopia 1. Smallholder agricultural cooperatives (RMBO) 2. Irrigation cooperatives (NRM) 3. Cooperative governance arrangements (LG) India (Orissa) 1. Self help groups (RMBO) 2. Pani panchayats (NRM) 3. Gram panchayats (LG) Vietnam 1. Agriculture cooperatives (RMBO) 2. Forest protection committees (NRM) 3. Democracy steering committees (LG) Yemen 1. Agriculture cooperatives (RMBO) 2. Tribal structures for water allocation (NRM) 3. District local councils (LG) 4.2.1 Research methods2 The research relied on a case-based approach to gather data on the above institutions. Semi-structured household interviews, discussions with institutional stakeholders, focus group discussions, and, where available, existing case materials and data sets were used to inform the description of the cases and an analysis of the causal processes represented in figure 1. b LG= Local governance RMBD= rural membership-based organization NRM= natural resource management 2 The different country teams for the study used the same analytical framework (figure 1) to guide their fieldwork and followed a common approach for data collection. But differences owing to contextual variation, resource constraints, and institutional capacity are inevitable in cross-country comparative studies. The concrete implication of such differences is some variation in the nature and amount of data available across the cases. 9 For the most part, the case studies in each country relied on a descriptive and comparative approach to examine ongoing institutional and political processes. In-country research partners were responsible for data collection, case study preparation, and a country report. This report is based on the case study and country descriptions prepared by the country research teams. Each country research team selected four locations for empirical work, based on representation of ecological and social variation in the selected study sites. Research sites were also selected to ensure that within each there were representative examples of institutions belonging to three different domains-- rural producer organizations, typically cooperatives; natural resource management institutions; and local governance institutions. Thus each country team collected data to produce three institutional case studies where in each case study at least four examples of the relevant institution were compared. The findings from each study were encapsulated into a country report that the research teams also prepared. The ensuing discussion is based on these studies. After providing a brief introduction to the social and policy context, the sections on each of the three institutional domains below compare findings related to institutional performance and outcomes. The objective is to identify the chief common features relevant to institutional performance, and highlight the most prominent tradeoffs and synergies in such performance. Sections 5­7 in the report examine findings from the study of membership-based rural producer organizations (RPOs), NRM organizations, and local governance (LG) organizations. Section 7 examines the relevance of the findings for structuring external support to local institutions so as to improve their capacity to contribute to more effective governance and improved livelihoods. The report ends with both general and specific implications of the findings for how external interventions can promote better governance and livelihoods outcomes through rural producer organizations, NRM organizations, and governance institutions. 10 5. Membership-Based Rural Producer Organizations Rural producer organizations can help smallholders overcome market, state, and institutional failures related to development by providing economies of scale and thereby reducing costs of production and marketing, increasing the bargaining power of smallholders, and helping improve productivity and governance (World Bank PRMPR and SDV 2003; World Bank 2004; Collion and Rondot 1998; Chen et al. 2006). There is widespread agreement among observers of membership-based organizations that if MBOs can overcome barriers to effective integration of rural producers' economic activities with credit, input, and output markets, they can enable improved livelihoods (Krishna, Uphoff, and Esman 1997; Rice 2001). This is one reason why a number of governments in different parts of Asia and Sub-Saharan Africa have given renewed attention to cooperatives and various SHGs. Although some of these programs have been successfully implemented in a number of regions in that they have increased agricultural production and contributed to economic growth and development of social capital, in other areas rural producer organizations have been relatively less effective. Indeed, this experience is similar to what Uma Lele noted in an early review of the experience of cooperatives: "Co-operatives have a universal appeal as an instrument of poverty alleviation and yet their record has been less than exemplary." (1981, 55) The findings and lessons of the empirical research conducted in the selected countries assume importance in this context of the high expectations of membership-based rural producer organizations, more specifically cooperatives. The case evidence from the examination of cooperatives and SHGs suggests that such organizations may have a long way to go to overcome the obstacles in their environments, but that enabling policies can provide a strong impetus to their accomplishments. 5.1 FIELD RESEARCH FINDINGS FOR RPOS The findings for the rural membership-based producer organization cases are based on an examination of cooperatives in Ethiopia, Vietnam, and Yemen, and SHGs in Orissa. The performance of these organizations depends in part on the macro-level socioeconomic environment as well as the macro-policy environment. But it also depends significantly on the social and economic forces in the rural contexts in which the organizations are located. Our case evidence suggests that governance plays a very important role in improving performance for better livelihoods, but is not conclusive. Further, different aspects of governance--inclusion, accountability, participation, and adaptability of the selected organizations themselves--do not necessarily register improvements simultaneously and can often be in tension. The macrosocial and economic context of all the four studied cases was characterized by significant constraints. Nonetheless, there are important variations in their performance. Understanding these variations requires a better understanding of the role of local institutional design and the critically important role of government policy in shaping the effectiveness of membership-based producer organizations. Table 4 provides a summary statement of the main characteristics of the context in which the studied rural producer organizations in Ethiopia, Vietnam, Yemen, and Orissa are located, and their institutional features. These summary indicators (low/medium/high) for the different aspects of the context and outcomes are based on the institutional case studies prepared for each country (for basic details on each country, see appendix 3; detailed case studies of the different institutional domains for each country and the country reports are available on the World Bank website). The low/medium/high values in tables 4, 11 5, and 6 are based on the judgment of the field investigators, and therefore subject to criticisms of subjective bias. We attempted to reduce this bias in part through discussions with each country team. For the different aspects of governance, we also used a simple conversion of the IAPA scores such that a low value was assigned to numerical scores between 0 and 1/3; a medium value to scores higher than 1/3 up to and including 2/3, and a high value to scores greater than 2/3. For livelihoods the assigned values are more subjective, and are based primarily on the discussions in the case studies. Note also that livelihoods contributions of governance institutions tended to be low to medium in most cases (except for Ethiopia, where the governance contributions of cooperatives were used in the case study). A quick review of the table shows some evident parallels across the studied locations as well as clear patterns in the relationships between institutional capacity, linkages, and outcomes related to livelihoods and governance indicators. Table 4. Membership-Based RPOs, Their Context, and Their Outcomes Rural producer Ethiopia: Yemen; Orissa: Self- organizations Vietnam: Rural Agricultural Agricultural help credit cooperatives cooperatives cooperatives groups Aspects of context and outcomes Extent of integration of Low Low Low Low smallholders with factor and output markets Extent to which an enabling policy Medium Low/Medium Low Medium/High framework is in place Extent of interventions in Low Medium/High Low Low functioning by external higher- level officials Links with other rural institutions Low Low/Medium Low Medium/High Local institutional capacity Low/Medium Low/Medium Low Medium Inclusion Low Low/Medium Low Medium/High Accountability Medium Low/Medium Low/Medium Medium Participation Low/Medium Medium Low/Medium High Adaptability Low Low/Medium Low Low/Medium Contributions to livelihoods Low/Medium Low Low Medium Equity of outcomes Low Low/Medium Low Medium The context of the studied rural producer organizations (RPOs), as is usually the case for most such organizations in the developing world, is characterized by production, distribution, and marketing constraints. These constraints work against the poor (Cleaver 1993; Huppi and Feder 1989). For example, there are few well-established channels for providing rural households with information, transportation, credit, insurance, marketing, and services at a cost they can afford. In all the studied locations in Ethiopia, Vietnam, Yemen, and Orissa, marketing chains for agricultural products are long, and the rural poor are in a highly unequal bargaining position with traders, credit providers, and government officials. Although the rural social and economic context across the different cases is quite similar, the policy landscape is not. In Vietnam and Ethiopia, there is significant interest in the further development of cooperatives and the policy framework is favorable to such development. However, the status of informal cooperatives in Vietnam, despite their large numbers, is ambiguous. There have been some efforts to use cooperatives to assist farmers in Yemen as well, but the extent of support available to cooperatives is very low. Most cooperatives cannot count either on the government or on donors for significant material assistance. In Orissa, in contrast, the government has been quite active in promoting SHGs, particularly through the Mission Shakti campaign. Its interest has helped build on the efforts of many NGOs that took the lead in promoting SHGs in the state. Higher-level officials charged with overseeing registered cooperatives have made different choices about influencing the functioning of RPOs. Government officials in all the countries have a keen interest in the 12 success of RPOs. But the degree of external support to RPO functioning is particularly strong in Vietnam and to a lower extent in Ethiopia as well. This is likely because the Vietnamese government is particularly invested in the functioning of local organizations, and in Ethiopia the government has placed a major emphasis on cooperatives as a means of channeling support to communities and mobilizing agricultural development. The effectiveness of external support, however, is higher in Vietnam, in part owing to the stronger local linkages of rural organizations. In addition, the presence of strong leaders in RPOs is often crucial in making external support count. Interestingly, the local institutional context itself is quite variable across the cases. Variations in the nature of linkages among institutions (at the same level, and with higher-order institutions) and institutional linkages with different social groups are particularly important to note. In both Ethiopia and Vietnam, the number of different kinds of local institutions is low in the studied locations. More importantly, the cooperatives in these two countries are not very well connected to the local institutions that do exist. The local institutional context seems usually to be dominated by one institution. This means that institutional linkages tend to be weak for other institutions. The weakness of institutional linkages is also evident in Yemen despite the presence of a significant variety and number of institutions in the studied areas. In contrast, the SHGs in Orissa are well connected to other institutions within the village and also to institutions outside the village. They have also had sufficient experience and history to recognize the importance of linkages to higher-level institutions. As a result, they have organized as federations in several cases and combined their strength. The suite of governance and livelihoods outcomes in the bottom half of the table needs to be understood partly in terms of the objectives of the cooperatives themselves and partly as an indicator of their performance. The low levels of inclusion for agricultural cooperatives in Ethiopia, Vietnam, and Yemen is a consequence of the way institutional goals of cooperatives are defined to support smallholders and landowning farmers. Agricultural marketing cooperatives are seldom interested in providing services to households that do not have products to market! In contrast, membership in the SHGs is more open because all rural households can participate in savings and credit efforts. All the case studies also suggest that the relatively better-off households tend to participate more in the decision making of the cooperatives. The variation in levels of accountability in the table is also a result of the fact that the cooperatives tend to be more accountable to higher-level officials and better-off members. (It should be noted that although the cooperatives have tended to serve "better-off" rural residents, the overall levels of income and wealth are sufficiently low in the studied locations that even the better-off households do not enjoy high standards of living). The degree of adaptability of cooperatives turns out to be low, particularly when it comes to addressing challenges faced by rural households that do not concern the relatively narrow mandates of the cooperatives. Indeed, the low level of institutional adaptability is primarily a result of the fact that cooperatives focus on a relatively narrow set of goals. An examination of the livelihoods contributions of the studied cooperatives and SHGs finally indicates that they do make a difference in the ability of their members to earn more. The Taramessa Multipurpose Cooperative in Ethiopia provides a good illustration of the low levels of inclusiveness and accountability found in many cooperatives that nevertheless help improve the livelihoods of their members. 13 Box 2. Taramessa Multipurpose Cooperative in Ethiopia The Taramessa Cooperative is viewed as a model coffee producers organization in the Sidama Zone of Ethiopia's renowned high-quality export coffee. The cooperative has more than 2,000 members, is integrated into the coffee-marketing chain, and has close ties with local financial and training organizations. The Taramessa Cooperative membership has been concerned about the accountability of its leadership, and many members believe that the cooperative's leaders have sought personal gain--employment for relatives, access to the cooperative's resources, and social recognition and stature. And although the local office of the Bureau of Agriculture and Rural Development has audited the cooperative, there is little capacity among the members to hold their leadership accountable by investigating suspected wrongdoing. The actions of the leadership have also created some financial difficulties for the cooperative. Despite these problems, the cooperative has performed well over the last several years to weather environmental shocks, expand the scope of its activities in response to members' interests and market signals, and improve members' livelihoods. Part of the explanation lies in the ability of the cooperative to benefit from the institutional linkages and the fact that the cooperative is involved in the marketing of a high-value cash crop. 5.2 LESSONS FROM THE COMPARISON OF MEMBERSHIP-BASED PRODUCER ORGANIZATIONS The four country case studies summarized in table 4 share patterns that reveal common puzzles and challenges related to improvements in performance. Patterns related to two areas are particularly important: 1) the relationship between performance on three dimensions of governance--inclusion, accountability, and participation--and the determinants of such performance and 2) the extent to which better governance and livelihoods outcomes are institutionally related. 5.2.1 Relationships among governance dimensions · NEGATIVE RELATIONSHIP BETWEEN INCLUSION AND PARTICIPATION. Among the examined cooperatives, there is a negative relationship between levels of inclusion and other governance indicators, particularly participation but also accountability. Higher levels of inclusion in the studied cases appear to correspond with lower participation and accountability. This inverse, relationship in the case of cooperatives seems to result from the fact that cooperatives are often created in a top-down manner to serve specific social groups (landholders, medium and large farmers) or type of production (crop specific, irrigation related). Servicing a smaller, higher- income constituency increases the efficiency of cooperatives in performing the tasks with which they are charged, but hinders inclusion. · NARROW MANDATES OF COOPERATIVES HINDER THEIR ADAPTABILITY. Although cooperatives are efficient in terms of performing the tasks for the target groups for which they are created, the narrowness of their mandate also hinders their adaptability. Adaptability of local institutions tends to emerge over time as they find the resources and the flexibility to address new challenges that emerge either because of changes in the environment or through new felt needs of members. When such flexibility is low--either owing to the way their objectives are defined or because of low resource availability--then adaptability remains low. · HIGHER LEVELS OF INCLUSION ARE NEGATIVELY RELATED TO EFFICIENT INSTITUTIONAL PERFORMANCE. The main reason for this seems to be that the inclusion of more people in a local organization that lacks sufficient resources stretches organizational capacities too thin. The negative relationship can in part be ameliorated by appropriate external support to institutions. Greater institutional 14 capacity, either as a result of contributions from members as payment for institutional services or through carefully designed external support, can increase livelihoods benefits from membership- based rural producer organizations. Other dimensions of governance performance--participation and accountability--are positively associated with livelihoods enhancement. The causal relationship here concerns the ability of participating members to ensure that the organizations will function in ways that better serve the interests of those who are more closely involved in institutional governance. 5.2.2 Governance and livelihoods The factors that have a positive impact on the ability of cooperatives to improve livelihoods have most often to do with their linkages with other local institutions, support from relevant government agencies, and an enabling policy environment. Further, cooperatives tend to benefit better-off rural families and households in part because costs of membership restrict the ability of cooperatives to draw members from among poorer households. Even in Orissa, where the SHGs provided benefits to poor households, they performed better in locations where overall levels of poverty were lower and their members were better-off. · LIVELIHOODS AND GOVERNANCE ARE POSITIVELY ASSOCIATED FOR COOPERATIVES. The positive association between participation/accountability and livelihoods enhancement is partly a result of the fact that in the studied cooperatives, better-off households have higher participation rates and the cooperatives are also more accountable to the better-off member households. As a result, the enhancement of livelihoods is greater for such households. · COOPERATIVES ARE DOMINATED BY MALE DECISION MAKERS. Benefits from cooperatives are at best available to women only in a limited fashion. Cooperatives are dominated by male heads of member households, and women are for the most part absent from participation and a direct share in available benefits. · WOMEN ARE ACTIVE PARTICIPANTS IN SELF-HELP GROUPS IN ORISSA. The SHGs in Orissa are distinct from cooperatives with regard to gender. Many of the SHGs focus explicitly on women's participation, have no entry barriers in terms of membership fees, and support activities that are either gender neutral or have a bias in favor of supporting women. 5.2.3 Local institutional context · THE INSTITUTIONAL LINKS OF SHGS HELP EXPLAIN THEIR BETTER GOVERNANCE AND LIVELIHOODS OUTCOMES. An important element in the better livelihoods and governance outcomes of the SHGs is their link to other local institutions. Unlike the agricultural cooperatives of different kinds in Ethiopia, Vietnam, and Yemen, the SHGs have more extensive linkages to important local institutions such as banks and rural governance organizations. In addition, they also often have links to relevant institutions outside the local context. Cooperatives' linkages with institutions outside the local context tend to be limited to mandated links. Further, the SHGs operate in a context where there are a number of local institutions that have multiple links with other institutions. In contrast, both in the Vietnam and Ethiopia case, the number of institutions is small, and the links show the importance of a single institution--essentially this means that if the dominant local institution is not functioning well, other institutions would be unable to take up the slack. 5.2.4 Role of external support In terms of external assistance, all of the studied local institutions were created through specific external interventions. The nature and continuity of external support, however, has been quite variable. These 15 differences raise two basic issues: what should be the sequencing and length of period for which external support is provided, and should such support be for specific project-related activities or for general institutional strengthening? In most of the case studies, support was typically available for project- specific activities rather than for general institutional strengthening. Such support can help improve the ability of local institutions to deliver specific services, but may not be helpful in improving their adaptive abilities or their capacity to address emerging needs in a changing environment. · LONG-TERM SUPPORT TO EXISTING INSTITUTIONS IS NECESSARY TO ENABLE MANDATED FUNCTIONS. A final set of findings relates to the way and length of time for which external support was made available. The cases bring out two major points on this score. The first is that long-term continued support is necessary for new institutions if they are to fulfill their mandated functions. Cooperatives in Yemen have suffered owing to the lack of sustained support, and a simultaneous absence of links with other local institutions (see box 3). In contrast, the SHGs received support from many NGOs initially, and in a number of cases the state government has begun to provide institutional and financial support to maintain their ability to provide services to their members. Box 3. Agricultural Cooperatives in Yemen Like many countries, Yemen has a long history of experience with cooperatives. Over the past several decades, cooperatives have been viewed as potentially crucial social and organizational mechanisms to improve livelihoods, undertake infrastructure construction, overcome barriers to marketing and input provision, and facilitate collective action. A number of legislative measures have recognized different kinds of cooperatives. Among the 1,410 cooperatives registered with the government, more than 60 percent are agricultural cooperatives (Sharief 2008a). Improvements in the functioning of cooperatives in the past decade have not kept pace with even the slow increase in the number of cooperatives. Lack of support from the government has not been offset by an enthusiastic embrace of cooperatives by farmers. The support from the state that is available has often been misused or used inefficiently. The existing agricultural cooperatives are not well connected with government agencies and have limited linkages at best with the marketing or veterinary services farmers need. Their contributions to farmer livelihoods and improvements in local incomes are, consequently, also quite low. · CONSISTENT SUPPORT FOR LOCAL INSTITUTIONS IS NECESSARY IN LIGHT OF THEIR LOW CAPACITY. The basic tension in providing external support to local organizations is about how to ensure that external support builds local capacity but does not create dependency relations. Although the case studies do not provide a clear message on this score, what they do make evident is the need for longer-term support than is typically available to most local organizations. One-shot, project- based assistance to new local institutions means that they may be overwhelmed by existing social relationships once the project ends. It is unlikely to yield sustained results. 16 6. Natural Resource Management Institutions Local water users and irrigation organizations, together with forest management bodies, are the focus of the case studies in the domain of natural resource governance. Such institutions number in the hundreds of thousands around the world (Pretty 2003). They play a fundamental role in shaping the access of poor households to basic resources such as water, fodder, and firewood (Ostrom 1990; Agrawal et al. 2008). Where they function well, they can generate equitable livelihoods-enhancing outcomes that take into account resource availability and improve sustainability. This study's focus on irrigation institutions derives from the important role they play in enhancing agricultural production. But such institutions are also important for managing water more sustainably, particularly in water-scarce areas. The role of community institutions in improving irrigation and agricultural production has been recognized in many World Bank studies (Gopal and Marc 1994: Jones 1995; Svendsen and Groenfeldt 2000). The study's concern with forestry institutions is particularly timely in light of the important role of forests for both climate change mitigation and contributions to local adaptation. Over the past two decades, water users and forestry institutions have increased significantly in number and importance as a result of more widespread decentralization of resource governance policies. Simultaneously, their relationships with external agencies have become more important in influencing their performance because in many contexts such relationships with higher-level authorities have become more formalized. 6.1 FIELD RESEARCH FINDINGS FOR NRM INSTITUTIONS The NRM institutions examined during the field research include irrigation cooperatives in Ethiopia, water users' associations in Yemen, irrigation councils (pani panchayats) in Orissa (India), and forest protection committees in Vietnam. In each country, these institutions can play a potentially fundamental role in enhancing environmental sustainability and improving rural livelihoods. But this potential is realized only infrequently. Our empirical work suggests that the key factors that influence outcomes include the nature and quantum of external support and the capacity of local institutions together with their links to other local and external institutions. There are important differences in the sets of variables represented in table 5. Thus, in the four countries there are significant variations in the resource endowments that NRM institutions aim to manage-- irrigation water and forests. Water is available in relatively high amounts in Orissa (despite some important spatial variations), particularly in comparison to the availability of resources in Ethiopia and Yemen (water) and Vietnam (forests). The other important feature of the context--an enabling macro- policy framework--is also variable. Although the governments in all the studied countries want to promote irrigation and improve rural incomes from forests, the extent to which they are willing to allow a critical role to local institutions varies. It is highest in Orissa and perhaps lowest in Vietnam.3 3 As in table 4, the "low," "medium," and "high" values in tables 5 and 6 are also based on the case descriptions provided by authors of the country case studies. More elaborate descriptions of the cases are available in appendix 4. 17 Table 5. NRM Institutions, Their Context, and Their Outcomes NRM Ethiopia: Vietnam: Yemen; Tribal Orissa: Pani Institutions Smallholder Forest water panchayats cooperatives' management management Aspects of context and outcomes irrigation institutions bodies management Extent of resource endowment Low/Medium Low/Medium Low Medium/High Extent to which an enabling Medium Low Low/Medium Medium/High policy framework is in place Extent of interventions in Low Medium Low Medium functioning by external higher- level officials Links with other rural institutions Low Low High Medium Local institutional capacity Low/Medium Low/Medium Medium Medium Inclusion Low Medium Low Low Accountability Medium/High Low Medium Medium/High Participation Low Medium Medium Medium/High Adaptability Low Low Low Low/Medium Contributions to livelihoods Low/Medium Medium Medium Medium Equity of outcomes Low Medium Low Low The studied water institutions in these countries are linked to other local and higher-level institutions. However, these links are quite low in Ethiopia, and medium to high in Yemen and Orissa. In Vietnam, the NRM institutions for forestry were established typically with the help of donors, and tend not to be well connected with other local institutions. 6.2 LESSONS FROM THE COMPARISON OF NRM INSTITUTIONS 6.2.1 Relationships among governance dimensions · EQUITABLE PARTICIPATION FOR THE SELECT FEW. In both Yemen and Orissa, the study examined two different kinds of water users' associations. In Yemen, the two types of irrigation institutions were formal and informal. The formal institutions tend to seek equitable levels of participation from different landholders, but they have not been very well linked to existing institutions. In contrast, the traditional institutions, typically headed by tribal elders, are well linked to other institutions, but tend to favor the better- off, upstream landowners. In Orissa, the study looked at lift and flow irrigation-based systems. The flow irrigation associations receive higher levels of support from the government, but are less accountable to their constituents and are less effective than the lift systems that rely on higher levels of contributions from members and are also more accountable and effective. · NEGATIVE ASSOCIATION BETWEEN INCLUSION AND PARTICIPATION. The governance outcomes of NRM institutions in Ethiopia are low for the most part except where their accountability is concerned. The irrigation cooperatives, like the agricultural cooperatives, are characterized by low levels of inclusion. They are, however, accountable to their members. In Vietnam, in contrast, the levels of inclusion are medium but accountability is low. The pattern observed in the case of cooperatives and membership-based organizations--a negative relationship between inclusion and participation/accountability--is also observed in the case of the studied NRM institutions. · WHEN INCLUSION IMPAIRS ACCOUNTABILITY AND EFFICIENCY. The studied NRM institutions also evidence tension between inclusion and effectiveness, as was the case for the RPOs. Greater levels of inclusion are associated with lower levels of accountability; the reverse also holds true. The attempt to increase inclusion and provide benefits to a larger number of households reduces 18 the ability of institutions to remain accountable to those they seek to serve, and also reduces their ability to provide benefits effectively. · FOCUSED MANDATE LIMITS ADAPTABILITY. None of the investigated NRM institutions demonstrated any significant adaptive capacity, or the ability to help their constituents adapt to significant changes in the environment. The narrowness of their mandates means that they are not designed to cope with major changes in their environments. This is a particularly important gap in local institutional capacities in the context of ongoing climate change, the likelihood that rural livelihoods are going to be adversely affected by increasing climate variability, and the closeness of many rural households to the poverty line. 6.2.2 Governance and livelihoods · CLEAR AND CONCRETE LIVELIHOODS BENEFITS FROM NRM INSTITUTIONS. The NRM institutions provide clear and concrete livelihoods benefits to their members. Even in the Ethiopian case, the low/medium ranking of the irrigation cooperatives is because of the low coverage of farmers-- the cooperatives generate clear livelihood benefits to member households. In Orissa, respondents attributed a close to 10 percent improvement in their livelihoods to the provision of irrigation water. In the Vietnam case of forest institutions, the outcomes are mixed as far as livelihoods contributions are concerned. In two of the four locations, local forests have helped improve livelihoods; in the other two cases, the forests themselves are threatened and their livelihoods contributions are limited. · LOW INCLUSION, HIGH INEQUALITY IN BENEFIT DISTRIBUTION. The important issue around livelihoods performance of institutions is the degree of inequality in the distribution of benefits from irrigation and forests. Benefits from irrigation institutions are clearly tilted toward the relatively well-off households. This is not surprising given the low levels of inclusion of many irrigation organizations. For the forestry institutions, the evidence is again mixed: at least in some of the cases, benefits from forests are distributed more equitably (see appendix 4). 6.2.3 Local institutional context Two important factors that work against local institutions' being more effective are the relatively unequal social environments in which they function and the limited resources they command. · INSTITUTIONAL CHARACTERISTICS COMBINE WITH CONTEXT TO AFFECT EFFICIENCY AND EQUITY. For example, in the Yemen case, new irrigation institutions have found it difficult to change existing rules that allocate water inequitably because the new institutions depend for their effectiveness on links to established leaders and authorities. The new irrigation institutions do not have sufficient resources to overcome traditional power relationships. In Orissa, on the other hand, the differences in the performance of lift irrigation and flow irrigation institutions suggests that where institutions are more accountable to their constituents, they are also more likely to allocate livelihoods benefits more equitably. · LOW CAPACITY OF NRM INSTITUTIONS. The studied NRM institutions struggle to achieve financial and economic sustainability. For example, the irrigation cooperatives in Ethiopia, the forest organizations in Vietnam, and the water users' councils in Orissa each face resource constraints that prevent them from hiring staff, enforcing institutional rules, and sanctioning those who break rules. These organizations run into monetary problems after the initial infusion of funds, and have only limited leeway to raise funds from the renewable resources they control. When they receive funds from central governments, these are seldom sufficient to facilitate day-to-day operations. Nor do they have the right to levy adequate membership or use charges on households that depend on the forest or water resources. The absence of resources and the 19 dependence on higher- level bodies for financial stability compromise local institutional capacity to manage resources. All the cases demonstrate the need for local institutions to have independent sources of finances instead of relying only on government handouts. 6.2.4 Role of external support · GOVERNMENT POLICIES ARE CRUCIAL IN SHAPING NRM INSTITUTIONAL PERFORMANCE. In all four cases, government actions and policy measures, rather than the amount of available resources, appear to be the more important influences shaping successfully functioning NRM institutions. The studied institutions came into being as a result of government interventions, many of their rules and design features are shaped by the government departments that oversee their functioning, and both governance and livelihoods outcomes are strongly affected by the way policies regarding resource governance are implemented. Particularly important aspects of NRM policy implementation are the extent to which devolution of decision-making powers is actually practiced and the resources that become available to the NRM institutions. · CAPACITY BUILDING WITH A LONG-TERM PERSPECTIVE IS KEY TO SUCCESS. One important lesson from the comparison of the NRM institutions concerns the importance of continued external support for improvements in the effectiveness of local institutions. When external actors (such as governments, but also including NGOs and donors) introduce new NRM institutions into unequal and hierarchical social contexts, long-term support is a prerequisite to change resource management outcomes. In the Vietnam case, the newly introduced forestry institutions need better links with other institutions as well as to government agencies to convert their initial achievements into longer-term livelihoods improvements. The new irrigation institutions in Yemen, the pani panchayats in Orissa, and the irrigation cooperatives in Ethiopia will continue to benefit primarily better-off households if longer-term support from the respective government to promote more equitable outcomes remains unavailable. 20 7. Local Governance Institutions The potential contributions of decentralized local governance to sustainable and more effective development outcomes have been described by many scholars (Rondinelli 1981; Smith 1985). The quality and performance of local governance institutions shapes how local populations relate to decision makers. Benefits of local governance include lower transactions costs associated with planning and coordination of projects at the local level, closer and more direct interactions between decision makers and local residents, and better representation of different local interests, greater responsiveness of institutions and decision makers that are closer to their constituents, and higher likelihood of public accountability (Davis, Hume, and Woodhouse 1994, 253). Whether these benefits are realized, however, depends on whether local governance arrangements are effective. All the countries included in the study are undertaking some efforts to reform local governance. These reforms are prompted by earlier development experiences that have often been disappointing both in the way local institutions have related to households and in the improvement of incomes and livelihoods. Reforms hinge on decentralizing control over decision making and resources, creating better channels for communication and service delivery, ensuring inclusion of the poorest, and creating mechanisms for greater participation and accountability. 7.1 FIELD RESEARCH FINDINGS FOR GOVERNANCE INSTITUTIONS The governance institutions studied in the research sites include community development councils in Afghanistan; village councils, or gram panchayats, in India; grassroots democracy committees in Vietnam; and district local councils in Yemen. The concluding discussion at the end of this section is also informed by a review of governance capacities and performance of agricultural cooperatives in Ethiopia. The focus on local governance institutions reveals a concern in each country about improvements in governance through local institutions, either by reconfiguring their structure and relationships with other institutions or by launching entirely new institutions. Clearly, policy makers in the relevant countries believe in the need for making local governance more effective. Table 6 summarizes the study's findings for local governance institutions in the four countries. Table 6. Local Governance Institutions, Their Context, and Their Outcomes Local governance Afghanistan: Orissa: Vietnam: Yemen: Ethiopia: institutions Community Village Democracy District local Cooper- development councils committees councils atives Aspects of context and councils (gram outcomes panchayats) Extent of social obstacles to High High Medium/ High Medium/ effective governance High High Extent to which an enabling Medium Medium Medium/ Medium Medium/ policy framework is in place High High Extent of interventions in Medium Medium Medium/ Low/ High functioning by external higher- High Medium level officials Links with other rural Low/ Medium/ High Low/ Medium/ institutions Medium High Medium High Local institutional capacity Low/ Low/ Medium Medium Medium Medium Medium Inclusion High High High Low/ Low Medium Accountability Low/ Low/ Low Low Medium Medium Medium Participation Low/ Low/ Low/ Low Medium Medium Medium Medium 21 Local governance Afghanistan: Orissa: Vietnam: Yemen: Ethiopia: institutions Community Village Democracy District local Cooper- development councils committees councils atives Aspects of context and councils (gram outcomes panchayats) Adaptability Low Low Low Low Low/ Medium Contributions to livelihoods Medium Medium Low/Medium Low Medium/ High Equity of outcomes Low/ Medium Low Low Low Medium In three of the four study locations--Afghanistan, Vietnam, and Yemen--local governance institutions have been created or introduced within the last decade. In each case, local governance reforms were an expression of national concern to overcome existing social obstacles to effective local governance. Orissa, where gram panchayats have existed for a long time, also saw significant local governance reforms through the 73rd amendment of the Indian Constitution (1992) with regular elections, reservation of leadership positions for women and minorities, and devolution of decision-making powers. In all cases, however, policy interventions to promote more democratic local governance occurred against a backdrop of significant social and political obstacles. These include highly unequal rural social relationships; political dominance of particular ethnic, caste, or propertied groups; and elite attempts to control organizations through which policy benefits are allocated. The experience of local governance varies across the selected cases. The availability of resources to implement development projects has led to the formation of community development councils (CDCs) in a majority of districts in Afghanistan. In the uncertain security context of Afghanistan, even the formation of the CDCs may be viewed as an accomplishment. Where they have succeeded in implementing development projects, the success is in no small measure the result of their ability to work through connections with traditional leaders and institutions. The story is similar for the district local councils (DLCs) in Yemen. The story of the gram panchayats in Orissa also bears rough parallels: they are the central rural governance institution, but their effectiveness has been limited as a result of low levels of financial devolution and limited ability to influence government agencies. Thus, although most rural governance institutions have gained additional resources for implementing a variety of infrastructure and development projects, they operate in a context of high inequality, government agencies that guard their powers, and strong patron-client relationships. In Vietnam again, the democracy committees aim to invigorate local democracy but have had only limited success for the same reasons. 7.2 LESSONS FROM THE COMPARISON OF LOCAL GOVERNANCE INSTITUTIONS 7.2.1 Relationships among governance dimensions · INCLUSIVE LOCAL GOVERNANCE INSTITUTIONS TRADITIONALLY LEAVE SOME GROUPS (PARTICULARLY WOMEN) ON THE MARGINS. Governance organizations in all the four cases are relatively inclusive on income and wealth criteria. Their high level of inclusion along these criteria is a result of legislatively mandated requirements. But even so, they are neither inclusive nor participatory where gender and ethnicity are concerned. The lack of inclusion of women is particularly noticeable in Afghanistan and Yemen, where women are not only absent from most local governance institutions but also gain little material benefits from the development projects implemented by the local councils (see box 4). Further, in nearly all the cases, the main forms of accountability exist in an upward direction. There is some downward accountability in the case of the gram panchayats in Orissa owing to electoral competition, but even here the party-based election process means that local officeholders are only rarely sanctioned for poor performance. 22 Box 4. Gender and CDCs in Afghanistan The intricacies of gendered participation in community-level local institutions are illustrated well by the example of community development councils (CDCs) in Afghanistan. The CDCs are a new institution, launched through the National Solidarity Program (NSP). CDCs are elected local bodies tasked with channeling resources into different localities, creating infrastructure, and improving local governance. Between 2002 and 2007, more than 16,000 new CDCs were created in 34 different provinces in Afghanistan and nearly 25,000 local development projects were approved through them. Among the debates about the functioning and achievements of the CDCs, the highly gendered nature of participation given the social realities in Afghanistan is unquestioned. Although the ideal of women's participation according the NSP operations manual is a single CDC in each community with mixed membership, the reality is that CDCs have either two bodies (one for men and the other for women) that meet separately, an all-male elected body that is then strengthened by the appointment of a woman as a member, or an all-male body without any women members. Without participation, women benefit only to the limited extent that male members of the CDCs think about allocating available resources for women's needs. 7.2.2 Governance and livelihoods · THE DIRECT IMPACT OF LOCAL GOVERNANCE INSTITUTIONS ON LIVELIHOODS IS LIMITED. The contributions of local governance institutions to livelihoods are limited, and occur mainly as local public goods such as infrastructure, schools, better water supply, and so forth. The private livelihoods benefits they provide are in the form of employment related to different kinds of construction projects. For the most part, collectively oriented as well as private benefits are skewed in favor of relatively better-off households, and indeed, often favor the families of the leaders of the governance institutions. 7.2.3 Local institutional context · TIES TO OTHER LOCAL INSTITUTIONS ARE WEAK. In all four cases, local governance institutions have multiple formal ties to existing local institutions. But in reality, these ties are weak and have failed to mobilize support for improved governance. Weak monitoring and management capacity have also meant that in most cases the implementation of development projects has typically been inefficient. 7.2.4 Role of external support · LINK NEW INSTITUTIONS TO TRADITIONAL INSTITUTIONS FOR LEGITIMACY AND OPERATIONAL EFFICIENCY--BUT ENSURE ACCOUNTABILITY OF DECISION MAKERS. The effectiveness of newly introduced institutions in achieving better governance outcomes (compared with that of traditional governance institutions) depends in significant measure on their capacity to provide needed services and elicit participation from groups that traditional governance institutions exclude. The case studies reveal that these two aspects of what makes local governance institutions effective are often in tension. Newly created institutions often rely on connections with those who are powerful in the existing hierarchy to provide services effectively. But these very relationships militate against meaningful involvement of excluded groups in the process of governance. This pattern is visible in both Afghanistan and Yemen. One of the central means of ensuring accountability--regular elections--is missing in most of the studied organizations, even when elections are required by law or by the bylaws of an organization. The difficulties confronting effective service provision differ between Orissa and Vietnam. In these two locations, existing government line agencies control needed resources to provide local services, and newly empowered local governance institutions must work in collaboration with these agencies. 23 However, such agencies guard their resources and prerogatives, in the process undermining the governance capacity of local bodies. The examples illustrate that better coordination and articulation between local institutions and social groups thus remains a major issue in all the studied locations. The relevant questions concern the access of disadvantaged groups to institutions, the availability of resources through different institutions, and the allocation of decision-making authority within them. · COHERENT DEVELOPMENT INTERVENTIONS REQUIRE LOCAL INSTITUTIONAL COORDINATION. In any given territorial location, there are always multiple institutions that vary in their capacities and responsibilities to provide services and benefits. Multiple avenues of service provision introduce questions about coordination of services. At stake is whether a single institution can be assigned more tasks so that users do not have to gain access to multiple institutions and decision makers, or if focusing all activities and responsibilities creates problems related to elite capture, and local institutional capacity more intractable. The dilemma of a single multipurpose institution serving multiple mandates versus multiple single-purpose institutions serving more narrow mandates is particularly evident in the case of gram panchayats in Orissa, but is also relevant to local governance bodies more generally. 24 8. Major Research Findings4 The domain-specific findings above are summarized in a small set of general lessons for mobilizing local institutions and governance. These lessons pertain to the four basic questions stated at the beginning of the report. Recall that these questions concern: a) Relationships among different dimensions of governance (inclusion, accountability, participation, and adaptability); b) Links between governance and livelihoods (in terms of enhancement as well as equity of livelihoods outcomes); c) Links between social/institutional context and institutional outcomes; and d) The role of external support in reinvigorating local institutional performance. 8.1 INCLUSION, ACCOUNTABILITY, PARTICIPATION, AND ADAPTIVE CAPACITY: RELATIONSHIPS AMONG DIFFERENT ASPECTS OF GOVERNANCE Questions about the relationship between different governance outcomes pertain to the right half of the research framework (figure 1). The fieldwork undertaken in different countries during the research clearly shows the multifaceted nature of local governance and the absence of straightforward relationships among different aspects of governance. Indeed, the first thing the research shows is that although it may be customary to talk about governance as if it is a single outcome, different aspects of governance are not necessarily related, different rural institutions promote different aspects of governance, and tradeoffs can occur among the different aspects of governance. Figure 2: Governance Outcomes Governance Outcomes Inclusion Participation Accountability Adaptability · MOVING BEYOND TRADEOFFS BETWEEN INCLUSION, PARTICIPATION, AND ACCOUNTABILITY. Higher levels of inclusion, we have seen, are often negatively associated with participation as the costs of coordination rise as inclusion increases. Higher inclusion can also increase the demand for greater participation. But participatory or inclusive institutions can remain unaccountable. Changes in institutional design can enhance accountability, but a large number of contextual factors can also promote accountability of decision makers. Tables 7a, 7b, and 7c present summary scores for inclusion, participation, and accountability across all the cases. The values of low, medium, and high are used to build the tables.5 4 It is worth mentioning that this study of rural institutions has two significant limitations. The first concerns the geographic scope and qualitative approach of the study. The findings are drawn from a relatively small sample of countries, and within each country, the analysis rests on case studies of rural institutions in three domains drawn from four different locations. As a result, the findings of the study are not representative; they indicate empirical regularities across a variety of contexts, and therefore an occasion for further investigation and testing. A second limitation stems from the fact that of the three major types of rural institutions in rural contexts--public, private, and civic--the study focuses primarily on two forms. Private rural institutions are not represented directly in the study, although they are included in the examination of institutional linkages of civic and public institutions. This approach can be defended on the ground that it represents a significant proportion of the institutional entry points used by Bank projects in rural areas. 5 Appendix 6 contains the values on which the correlation scores are based (low = 1, medium = 2, and high = 3). 25 Table 7. Relationships between Different Aspects of Governance Table 7a. Inclusion and Table 7b. Inclusion and Participation Accountability Inclusion Grand Inclusion Grand L M H L M H Participation Total Accountability Total L 15 5 3 23 L 16 7 8 31 M 8 3 7 18 M 2 1 9 12 H 8 4 1 13 H 8 3 0 11 Grand Total 31 12 11 54 Grand Total 26 11 17 54 Table 7c. Participation and Accountability Grand Participation L M H Total Accountability L 16 6 4 26 M 3 4 4 11 H 4 8 5 17 Grand Total 23 18 13 54 Tables 7a, 7b, and 7c suggest that the relationship between inclusion and participation and that between inclusion and accountability is relatively unclear. But the high (16), medium (4), and low (5) values for participation and accountability tend to occur together in 25 cases. At the same time, there are only four cases in which high values of participation occur with low values of accountability, and vice versa. This information suggests that high, medium, and low levels of participation tend to occur with high, medium, and low values of accountability The causal relationship between these aspects of governance is difficult to disentangle given the available data, but one explanation for this close association between inclusion and participation may be that stronger participation by members of local institutions can promote demands for greater accountability. · LOW INCLUSION OF WOMEN. Most of the institutions on which the study focused show low levels of inclusion of women. Indeed, this observation is also true for other dimensions of governance and livelihoods. Women's participation in institutional decision making is consistently low and institutional accountability to women members is almost nonexistent. The absence of women from most governance functions is also reflected in the low levels of benefits distributed directly to women from local institutions. A major reason for the invisibility of women in institutional relationships is structural: women suffer from many forms of marginalization in the areas in which the study was carried out. But the other part of the explanation lies in the design of the studied institutions. Few institutions require explicit attention to the interests of women. Where institutional design features make the involvement of women part of the institutional design, as is the case for the some of the gram panchayats in Orissa and many of the SHGs, women receive at least some formal recognition of their needs and at times benefits are also channeled toward them. There is considerable scope for greater involvement of women in local institutional processes through institutional redesign. 26 8.2 GOVERNANCE, LIVELIHOODS, AND EQUITABLE DEVELOPMENT The second general question focuses on the relationship between governance and livelihoods. Ideally, effective governance should lead to improvements in livelihoods outcomes. But it does not always do so. The examination of the three different institutional domains shows that the reason for variation in the nature of governance and livelihoods relationships can be traced again to institutions as well as to their context. Figure 3: Governance and Livelihoods Outcomes The basic two points can be stated simply: · Local institutions can often help align livelihoods and governance; · But when resources are inequitably distributed, policy support is inconsistent, and local institutions are not well connected, the negative force of contextual conditions is too strong to be overcome by local institutions. The research finds in general that governance does not have a clear relationship with livelihoods outcomes, in terms of both livelihoods enhancement and equity of benefit distribution. The lack of a clear relationship works in both directions--some local organizations, despite being well governed, do not contribute much to livelihoods improvements. In other cases, local organizations that are not very well governed still contribute to improvements in livelihoods. An example of the latter is presented by the example of water use in Yemen. Box 5. Contributing to Livelihoods Despite Low Accountability: Tribal Structures for Water Use in Yemen Tribal leaders and elders--sheikhs--play an important role in irrigation water management in Yemen. Application of traditional laws related to agriculture, grazing, common lands utilization, and irrigation is widespread. Tribal structures are also critical in the mobilization of communities for flood control, protecting riverbanks, and constructing and repairing terraces. Within the selected sites, spate and masque irrigation are the two main forms of irrigation. Under spate irrigation, elementary stone or earthen bunds together with diversion canals are used to divert floodwaters from riverbeds or valleys (wadis) into fields. Runoff can also be harvested from small surfaces and diverted by gravity directly onto fields using ditches (masques). The basic principle of water allocation is that landowners located upstream in a wadi irrigate their lands before those located downstream. Tribal structures and decision makers in the four selected locations are well connected to most of the other local institutions, but less so with higher-level or central institutional arrangements or decision makers. In contrast, the newer water users' associations tend to have fewer links with local institutions, as is the case for the Bank- assisted projects to improve spate irrigation and soil and groundwater conservation. Local residents often view tribal structures as relatively inclusive and participatory but have little capacity to hold tribal leaders accountable or to change their leaders. Despite the low accountability of tribal leaders, their role in applying water allocation principles means that many of the upstream landholders benefit significantly from the existing institutions. These benefits are distributed inequitably, based on who owns upstream land. The information across the cases does not show any clear patterns in the association between medium, high, and low values of governance indicators and livelihoods improvement. On the other hand, the relationship between governance and livelihoods is stronger in relation to equity of benefit distribution. It is the strongest for accountability and equity, but also high for participation and equity. Tables 8a and 8b summarize the 27 association between low, medium, and high values of equity on the one hand, and accountability and participation on the other. Table 8. Associations between Governance and Livelihoods Outcomes Table 8a. Participation and Equity Table 8b. Accountability and Equity Participation Accountability Grand Grand Equity in L M H Equity in 1 2 3 Total Total Benefits benefits L 15 5 3 23 1 17 3 3 23 M 6 7 7 20 2 8 6 6 20 H 2 6 3 11 3 1 2 8 11 Grand Total 23 18 13 54 Grand Total 26 11 17 54 In tables 8a and 8b, the low, medium, and high values of equity occur together with the similar values of participation (8a) and accountability (8b). In contrast, there are only a few cases where high or low values of equity occur with the reverse values of participation or accountability (only 5 such cases are present in table 8a and 6 in table 8b). 8.2.1 Explaining the lack of a clear association between better governance and increases in livelihoods Two important reasons, apart from sporadic, inconsistent, and low levels of external support and inhospitable local context, help explain why the association between different aspects of governance and improvements in livelihoods is low for the studied local organizations. a) Local organizations do not have the authority to raise funds that would allow them to function effectively (certainly not adequate funds); and b) most newly created local organizations are not well connected to other institutions in the local context, nor to higher level government agencies. As one might expect, the lack of resources and opportunities to raise resources adversely affects their ability to support the livelihoods of their constituents. The example of water councils in Orissa is representative of local organizations more broadly. 28 Box 6. A Tale of Two Water Users' Councils in Orissa (India) The two different kinds of water users' councils in Orissa (pani panchayats) are based on flow irrigation and lift irrigation. These councils are statutorily empowered to raise the resources they need to function effectively from five different sources: · grants by the state and central government; · resources raised from any financing agency; · income from the properties and assets managed by the councils; · fees from water users for the irrigation services that the councils provide; and · other sources including funds from legislators. However, the councils based on flow irrigation have received funds only as annual grant-in-aid from the state government and small amounts as membership fees. In case of lift irrigation arrangements, the relevant council has received funds from more diverse sources, including water charges collected from water users and membership fees. Lift irrigation­based systems do not receive grant-in-aid from the state because the council collects and retains the water and service charges from its members. In contrast, the councils in the flow system are mobilizing at best very small funds from members. The consequent difference in the functioning of the two types of water users' councils is striking. The council based on flow irrigation, which receives most of its resources as a transfer from the state, is not very accountable to its members and has tended to improve the livelihoods primarily of better-off individuals. The lift irrigation­based water users' council is highly accountable--members want to be sure that their fees and charges are being put to good use--and the livelihoods impacts of the council's activities are also superior. In the case of SHGs in Orissa, there is a stronger relationship between better governance and improved livelihoods. Three pressures work to make this possible. First, the SHGs are designed to facilitate the involvement of low-income households as well as marginalized groups such as women. As a result, their rules for membership and focus on types of services to be provided cater to the needs of poor households, and costs of entry are low. Second, the activities they seek to promote do not require significant investment. Therefore, it is easier for external actors to support their work. Finally, they have received consistent support from external organizations such as NGOs and later at a larger scale from government agencies. This has also meant that they are well connected to these higher-level organizations and also to other local institutions. They are, thus, distinguishable from other local institutions we studied in three key areas: institutional design and goals, resource availability and external support, and local and higher-level institutional networks. 8.2.2 Accountability and equity From among these findings, it is particularly important to focus on the observed relationship between accountability and equity in distribution of livelihoods benefits. Our research provides two types of evidence to support the proposition that higher levels of downward accountability lead to more equitable allocation of benefits. The first set of examples concerns the cases where downward accountability is low or nonexistent. The examples of most of the governance institutions are a case in point. In these institutions, the flow of resources occurs from higher levels to lower levels and the institutional decision makers feel accountable to higher-level organizations. The allocation of benefits in these institutions is not very equitable. Several of the NRM institutions and the RPOs--the lift irrigation institutions in Orissa and the irrigation cooperatives in Ethiopia, for example--are accountable to relatively well-off households. Benefit flows from these institutions are directed more toward better-off households. In the case of the SHGs (and to some extent also for the lift irrigation institutions), there is effective participation and accountability to members. Benefits flows in this case are more equitable. · INSTITUTIONS CAN BE DESIGNED TO PROMOTE DOWNWARD ACCOUNTABILITY. The study shows that it is possible to institute downward accountability mechanisms through careful institutional design. But this is rare because governments tend instead to institute upward accountability mechanisms to maintain control over local institutions. Longer-term and greater support to local institutions (whether they are newly created or whether government policy seeks to invigorate 29 existing institutions) is necessary if local constraints to effective institutional performance are to be overcome. A number of institutions in our study have helped improve local livelihoods even when their governance arrangements are neither very inclusive nor very participatory. The marketing and agricultural cooperatives in Ethiopia, but also in Vietnam, are a case in point. The study also shows that better governance is strongly associated with more equitable livelihoods outcomes. · REDUCING BARRIERS TO PARTICIPATION. The lack of clear relationships between effective governance and livelihoods improvements, coupled with the stronger relationship between better governance and higher equity, suggests that local institutions charged with improving access to resources should also be provided with the resources to do so, and that the allocation rules for available resources as well as decision-making procedures should discourage wealthier and better-off households from attempting to maintain exclusive control over local institutions. This can be accomplished by o reducing or eliminating membership barriers; o requiring local institutions to disseminate information widely about the benefits they provide; o requiring performance reviews of decision makers by constituents; and o instituting mechanisms for rotation of leadership positions. 8.3 CONTEXT, INSTITUTIONAL LINKAGES, AND OUTCOMES The research throws some light on the extent to which contextual factors, both social and institutional, shape the tensions between different dimensions of governance and the relationship between governance and livelihoods. Greater accountability is a central issue for improved governance. In nearly all the studied cases, the local context is characterized by different kinds of social and economic inequalities, weak integration with markets and higher-level governance arrangements, and low availability of different services. Effective institutional performance in such unfavorable contexts requires visible and consistent provision of resources that all organizations need to function. Because such resources are typically not available, three kinds of tensions in institutional functioning are manifest: between inclusion and representation/ accountability, between inclusion and effective service provision/livelihoods enhancement, and between indifferent resource availability and the demand for effective service delivery. These tensions are visible in all the studied domains. Most of the studied cooperatives in Ethiopia, Vietnam, and Yemen have been inclusive only to a limited extent. Their low levels of inclusion are a result of entry costs that many rural households are unable to meet, or because they provide services that can benefit only a small proportion of rural households that they aim to serve. Low inclusion has been consistent with higher involvement of members in decision making and greater accountability to members. Attempts to improve inclusion have met much success, but where they have been successful, higher levels of inclusion have adversely affected downward accountability and member involvement in decision making. A similar pattern also holds for nearly all the irrigation institutions studied for the NRM domain. Claims over water in water-stressed contexts are defended vigorously, and without major infrastructure investments, improvements in water supply are difficult to accomplish. In the case of local governance institutions, the reverse pattern holds. Broadly speaking, they are more inclusive than agricultural cooperatives and irrigation institutions; nevertheless, decision making in these institutions is not shared broadly. Formal inclusion is not matched by effective representation and accountability. 30 · INSTITUTIONAL LINKAGES IMPROVE LIVELIHOODS AND EQUITY. Apart from the social, political, and economic aspects of the context, it is also important to attend to the institutional linkages among the studied local institutions. Stronger links among different local institutions can help address weaknesses in any single institution by permitting it to draw on resources available from other institutions to which it is linked. Weak linkages with higher-level agencies and institutions prevent the flow of information and resources to constituents, as has been the case with the water users' councils in Orissa and the studied NRM institutions in Yemen. Tables 9a and 9b provides summary associations for the strength of institutional linkages and livelihoods outcomes, showing that stronger institutional linkages are associated in the studied cases with better institutional performance related to equity in distribution of livelihoods benefits, but a less clear pattern related to linkages and livelihoods. Table 9. Associations Between Institutional Linkages and Livelihoods Outcomes Table 9a: Institutional Linkages and Livelihoods Table 9b: Institutoinal Linkages and Equity Institutional Institutional Linkages Linkages Grand Equity of Grand Livelihoods L M H Total Benefits L M H Total L 12 12 2 26 L 13 10 -- 23 M 8 7 4 19 M 6 10 4 20 H 1 8 -- 9 H 2 7 2 11 Grand Total 21 27 6 54 Grand Total 21 27 6 54 The information in the tables above should be treated with some caution, because it is based on summary measures but also because it is based on subjective assessments of linkages as perceived by field investigators. But high, medium, and low levels of linkages as indicated in table 9b seem to occur together with high, medium, and low levels of equity in distribution of benefits from the institutions. 8.4. EXTERNAL INTERVENTIONS AND INSTITUTIONAL PERFORMANCE The final research issue--the relationship between external interventions and local institutional outcomes--is perhaps the most important from a policy and programmatic perspective. When and what kinds of external interventions are more likely to improve the governance and livelihoods performance of local institutions? Figure 4: External Interventions, Institutions, and Outcomes 31 Local improvements through external interventions are necessarily channeled through existing local institutions, which are linked with one another and with different rural social groups. Particularly important for our research on this theme are questions about the size, sequencing, distribution, and continuity of external interventions and their relationship to the autonomy and adaptability of local institutions. · FINANCIAL AND TECHNICAL SUPPORT FOR LOCAL INSTITUTIONS CONTINUES TO BE OF BASIC IMPORTANCE. The study documents the importance of external financial and technical contributions to local institutions as key elements in efforts to develop vigorous local governance. The absence of such contributions is in large measure responsible for the lack of performance by LDCs in Yemen, democracy committees in Vietnam, and the irrigation cooperatives in Ethiopia. Similarly, availability of external support has helped the SHGs in Orissas and to some extent the CDCs in Afghanistan and the gram panchayats in Orissa. The example of the forestry management bodies in Vietnam shows clearly the importance of external support in strengthening local institutions (box 7). Box 7. Local Forestry Institutions in Vietnam During the study, two different types of forests were examined in Vietnam: mangroves in coastal areas and upland forests in the interior. In general, the coastal forests seem to have been managed more carefully and sustainably and with better returns to local participants. Better forest governance may be attributed in significant measure to greater available resources. Coastal communities have effectively combined forest protection with economic development as they have used external resources for training and improvements in management. In some cases, the pilot effort to protect forests was not continued and once external resources for the pilot project ceased, local participants did not know how to continue to raise additional resources. Improved management in the coastal areas also has to do with better education, stronger social capital, and stronger infrastructure in the coastal forest management sites. A major lesson of the example of forestry institutions in Vietnam is to underline the importance of resources for improved local institutional functioning. External support can be made available in several ways. One of the most important types of external support-- one that can make a real difference in local institutional performance--is clear assignment of rights and closer linkages to existing institutions. These factors have made a significant different in the performance of SHGs in Orissa. They are also at work in improving the performance of tribal irrigation arrangements in Yemen, the CDCs in Afghanistan, and the informal cooperatives in Vietnam. As we noted in the previous section, the nature of the institutional context, particularly in terms of the linkages among different local institutions, makes a clear difference to livelihoods outcomes in the studied locations. Higher levels of external support to local institutions also create the risk that local elite may appropriate most such resources and little of the benefit of improved support will flow to the poorer and more marginal groups of rural residents. (This risk is in addition to the risk of developing dependency relationships between local and external agencies). To counter this risk, better accountability of institutional decision makers to their constituents is necessary. · NATURE OF INSTITUTIONAL LINKAGES. Not only the extent but also the nature of institutional linkages makes a difference to the distribution of benefits. Where such linkages are concentrated in a single dominant institution, the prospects for equitable distribution of benefits depend mostly on the performance characteristics of the dominant institution itself. But when such linkages are distributed across multiple institutions, and when there are a number of different important institutions in the local context, there is increased potential for more equitable institutional performance through the influence of the different institutions on the focus institution. Linkages between local and higher-level institutional arrangements are especially important. Even without a flow of finances from external sources, the ability of local institutions to access higher authorities can help improve local effectiveness. 32 · INSTITUTIONAL AUTONOMY IS LINKED TO ADAPTIVE CAPACITY. Effective external support is needed to increase local adaptive capacity. Indeed, adaptation can be viewed as the autonomous ability to change in response to environmental shifts, both sociopolitical and natural. The low adaptive capacity of most of the studied organizations is particularly apparent when they are faced with environmental changes and new environmental challenges. But the adaptive capacity of local institutions is also not very high in relation to new expressed needs of their constituents. Three factors that lead to lower adaptive capacity are related to institutional autonomy. Absence of strong leaders, lack of resources, and inadequate levels of participation all contribute to low levels of adaptive capacity of local institutions. Institutions such as agricultural and marketing cooperatives, forest management bodies, and irrigation groups all aim to improve performance in specific areas. They cannot be expected to address broader challenges thrown up by shifts in the sociopolitical or natural environment. · LOCAL INSTITUTIONS HAVE DIFFICULTY ADAPTING TO NEW CHALLENGES. The governance institutions we examined are different from sector-specific institutions; they should be expected to have high adaptive capacity. However, both sector-specific and general-purpose local institutions find it difficult to adapt to environmental shifts because of a lack of autonomous capacity to raise resources and deploy them for locally agreed upon purposes. We saw that the major sources of revenue for local governance, where such sources exist, are external transfers. The CDCs in Afghanistan, the gram panchayats in Orissa, the democracy committees in Vietnam, and the LDCs in Yemen all suffer from similar difficulties regarding resource access. Either they do not have resources to speak of (democracy committees in Vietnam that are essentially expected to coordinate local development activity and the actions of other organizations) or their resources are monetary transfers from higher-level authorities (the CDCs, the gram panchayats, and the LDCs). These organizations have little discretion over how they can spend resources because of the conditions attached to available resources. · CONCERNS ABOUT DEPENDENCY OF LOCAL INSTITUTIONS ARE MISPLACED. An additional tension between continued external support and development of relationships on the part of local institutions is important to discuss. Concerns about dependency relations are conceptually well founded. But the local organizations that were studied during the research receive nowhere near a high-enough level of prolonged external support from governments and NGOs to convert them into dependent institutions. For the most part, the studied organizations suffered from a lack of external support and statutory constraints on their resource-raising capacities such that they seldom enjoyed adequate resources for the tasks entrusted to them. Further, concerns about dependency relations are better expressed against the limited autonomy and resources that local institutions are permitted. Sporadic and limited support, coupled with limited devolution of decision- making capacity, is what makes local institutions dependent on external support. Sufficient external support in terms of material resources and autonomy in decision making is more likely to lead to greater adaptability and autonomy. · INSTITUTIONAL EFFECTIVENESS REQUIRES INSTITUTIONAL ADAPTIVE CAPACITY. To be effective in promoting sustainable rural development, institutions must be adaptable. The nature of challenges to development is constantly changing. Social and institutional learning, therefore, is critically important for long-term institutional adaptation. Such learning requires the incorporation of flexibility in institutional design. Flexibility in institutional functioning, however, also opens a window for institutional takeover by local elite. Explicit recognition of the tradeoffs between flexibility and elite capture is a necessary step in trying to negotiate social divisions and include excluded groups (Johnson and Wilson 2000, 1891). Collectively, these findings are highly relevant to concerns about institutional sustainability. Nearly all the institutions examined during this project are newly introduced institutions, although in several 33 cases--Afghanistan, Yemen, and particularly in Orissa--one of the issues that the study teams examined was about the relationship of new institutions to existing institutions. Where there is a lack of fit between new and existing institutions, as for example was the case in some water councils in Orissa and some cooperatives in Yemen, the prospects for institutional sustainability are low at best. This is one of the reasons for thinking more carefully about how external support for local institutions can be structured in ways that promote longer-term sustainability and effective functioning. 8.5 RESEARCH FINDINGS AND TERRITORIAL INSTITUTIONAL DEVELOPMENT The study examined nearly 50 distinct local institutions in three different domains in five countries. (See appendix 6 for a summary table of the specific institutions included in the study.) The common elements and patterns in this set of empirical cases should be emphasized. The reason is simple. The cases are spread across a multitude of problem domains, microlevel social and institutional contexts, and macro-level policy environments. We expect therefore to find many differences across the studied institutions and contexts. But where we find commonalities among different aspects of governance, between governance and livelihoods outcomes, or between types of external interventions and institutional performance outcomes, the patterns deserve greater attention. These commonalities in institutional outcomes are more likely to be broadly reflective of local institutions in rural settings in developing countries. 8.5.1 Local institutions necessary complements to national and regional institutions It is important to highlight three fundamental points on the basis of the case studies. One has already been mentioned: local institutions are necessary complements to national-level development project and policy interventions. They are necessary because they structure local social, political, and economic interactions and exchanges; they are also necessary because they channel flows of external interventions and resources into local contexts. The CDCs in Afghanistan, SHGs in Orissa, marketing cooperatives in Ethiopia, and the DLCs in Yemen are important to the local success of different kinds of development projects and policies. They are the means through which national governments and their agencies connect to rural social groups and households. The success of development projects, particularly in terms of benefits to more marginal, poorer, and disadvantaged groups, depends on the effectiveness with which institutions can connect to and serve such groups. 8.5.2 Government policies do not provide adequate support to local institutions The fact that local institutions are necessary to improved development performance, in terms of both livelihoods and governance, is not reflected, however, in the support available to these institutions. A quarter century after the current wave of interest in local institutions and decentralization began and despite the thousands of reports and papers exhorting the importance of local institutions, little seems to have changed as far as local rural institutions are concerned. In all our study sites, local institutions are hobbled by lack of effective decision- making authority, paucity of funds, low technical capacity, weak links to higher-level governance arrangements, and incentive and accountability relationships that benefit the better-off in the cases where local benefits are generated. Despite some significant exceptions that provide important lessons for future interventions, case after case in the study--in the gram panchayats in Orissa, the agricultural cooperatives in Yemen, the irrigation cooperatives in Ethiopia, or the forestry institutions in Vietnam--contains evidence that local institutional and governance arrangements have not met the development hopes invested in them. 8.5.3 Institutional connections, in addition to resource transfers, important to improve outcomes Finally, the research advances an argument about the importance of institutional connections. Our findings suggest that local institutions depend not just on the internal capacity of the institution or resource transfers from external agents, but also on the nature of links with other institutions. In all the cases we examined, three 34 different types of local institutions were present. These include sector-specific institutions (agricultural, marketing, and irrigation cooperatives in Ethiopia; lift and flow irrigation institutions in Orissa; and forestry institutions in Vietnam), multipurpose institutions (e.g., CDCs in Afghanistan, DLCs in Yemen, and governance organizations), and community-based institutions (SHGs in Orissa, informal cooperatives in Vietnam, and tribal irrigation arrangements in Yemen). Interactions among these different institutions through network relationships turned out to be a key element in promoting governance and livelihoods. We conclude that where such interactions exist or are promoted through external interventions, they are likely to lead to more balanced local institutional development in a given territory, and in turn to better prospects for improved institutional performance toward development. Figure 4 presents the argument graphically. Figure 5. Territorial Institutional Development The basic point of the figure is straightforward. In any given territorial location, there are many different types of institutions. Interactions and linkages among these institutions are critical to development outcomes within that territory. To the extent territorial development involves institutional changes among multiple parts of the territory, it also depends on the interactions among different types of institutions within the territory and the sectors in which these institutions are located. Improved territorial development is likely to be facilitated by positive interactions among different institutions within the territory. 35 9. Implications of Research Findings for World Bank Projects Ineffective institutions diminish the lives and life chances of people (IFAD 2003). In this context, the findings of this report on rural institutions are relevant to a large portfolio of Bank development projects, particularly those focusing on community-driven development (CDD), natural resource management (NRM), and local governance (LG). According to a recent review, the vast majority of such projects rely on a combination of civic and public organizations as entry points to achieve their objectives (26 out of 31; another 4 were hybrids, and only 1 relied entirely on private-sector organizations (Perrin, Van Der Riet, and Sjorslev 2006). Indeed, this is one reason why the focus institutions in the study are primarily in the public and civic domain. 9.1 STRENGTHENING LOCAL GOVERNANCE WILL REQUIRE GREATER AND BETTER-DIRECTED EFFORTS The most consistent implications of the study's findings stem from the low capacity of the studied organizations. An hierarchical social context and limited cross-scale institutional interactions mean that local institutions have to go a long way to be effective partners in inclusive, coherent, and accountable development. The study was not designed to focus on local institutions that are ineffective. But despite the selection bias in favor of better-known organizations in all the selected domains, the research mostly encountered institutions that enjoy limited success at best. Whether trying to promote sustainable and equitable livelihoods, or attempting to enhance inclusive, participatory, and adaptable governance, the record of local institutions can be improved substantially. Strengthening local governance and local institutions to improve livelihoods will require greater and better-directed efforts than has hitherto been the case. These efforts should include channeling of higher levels of resources to support local institutions, clear definition of the tasks rural institutions are expected to perform, and support to decision makers in rural organizations to improve their ability to undertake organizational governance. 9.2 PROMOTING BETTER INSTITUTIONAL LINKAGES IS A KEY NEW AREA FOR EXTERNAL INTERVENTIONS A second major implication of the study concerns the type of external supports that can improve local institutional performance. Existing studies of institutional performance have focused on two major issues in this regard: i) the quantum of external support and whether it creates institutional dependencies and ii) whether the choice is between partnering with existing institutions or creating new ones. Our review of evidence from nearly 50 cases suggests that an additional important form of support to local institutions can occur by facilitating greater linkages across institutions--within the locality, but equally and perhaps more importantly across levels so that local institutions are better articulated with higher-level organizations. Such support may cost relatively little in monetary terms, but can greatly improve the effectiveness of local institutions. Nonetheless, the costs of improving linkages need to be taken into account. In particular cases, building linkages may place greater strain on the limited resources available to local organizations. Where necessary, better linkages can be facilitated, for example, by better information sharing and making information publicly available and creating monitoring relationships between specific government agencies and local organizations. 36 9.3 PARTNERING WITH EXISTING INSTITUTIONS VERSUS CREATING NEW INSTITUTIONS The study found only a small association between the effectiveness of an institution in promoting livelihoods and whether it was new or whether it predated an external development intervention. Thus traditional irrigation institutions dominated by chiefs in Yemen as well as newly introduced irrigation cooperatives in Ethiopia effectively allocated available water, but did so in a way that benefited a small group of well-off landowning households. Newly introduced forestry institutions in Vietnam turned out to be less effective in enhancing livelihoods once external support was no longer available, and informal cooperatives were quite effective in promoting livelihoods. The choice between traditional and new institutions seems more to be a function of whether institutions with desired characteristics and capacities are present in the location where a development project is implemented. Existing or traditional institutions are more likely to be dominated by a group of local elite, as was true for all the cases covered in the study (except for the informal cooperatives in Vietnam), but they also have tend to have strong local linkages and knowledge about what works. The creation of new institutions allows greater leeway in defining more equitable rules of membership, participation, accountability, and benefit allocation. But it is also a far more costly choice. Therefore, creation of new institution should be undertaken only when external interventions can assure long-term support to such institutions. Inconsistent or single-time-period-oriented, package-approach-based support for new institutions is likely to strengthen the less preferred traditional institutions in the long run. Thus the CDCs in Afghanistan will require long-term and consistent support to garner greater legitimacy; the role of DLCs in Yemen without continued and higher levels of support is likely to be even less positive than it has been till now. 9.4 ASSESSING THE RELATIONSHIP BETWEEN GOVERNANCE AND LIVELIHOODS The lack of a strong and systematic relationship between more effective governance and improved livelihoods is less important from the point of view of Bank operations than the identified relationship between greater downward accountability and more equitable distribution of benefits. This is because the focus of much of the Bank's development work is on improving the livelihoods of the poor and the marginalized. The findings of the study suggest that Bank-supported local development projects should aim to institutionalize clear accountability relations where decision makers are answerable to their poorer and socially marginalized constituents. Although periodic elections may be viewed as one means of achieving accountability of decision makers, they are by no means the only such mechanism. Indeed, they may not necessarily be even the best such mechanism, depending on the extent to which they are fair, regular, and competitive. A significant number of other social accountability mechanisms can also be promoted, including those that emphasize broad information sharing about projects, sharing of financial information, citizen report cards linked to sanctions on decision makers, periodic audits of the functioning of institutions, and improving the ability of constituents to voice dissatisfaction (Ackerman 2005). Many accountability mechanisms are already used in various efforts to improve rural livelihoods. Broadly speaking, they can be divided into three major categories, based on information, on financial requirements, and on institutional changes. Information mechanisms include various kinds of audits, mechanisms for reporting corruption, and mechanisms for performance monitoring. But different forms of financial mechanisms--through accounting, judicious use of taxes, and control over some proportion of salaries of decision makers--have also been used to improve downward accountability. Finally, 37 creation of new rules, dismissal of decision makers upon evidence of wrongdoing, and in some cases even training and advice can serve to improve accountability of institutional decision makers. In this last set of accountability mechanisms might also be included more traditional models, in which social norms and custom act as a check on the arbitrary exercise of power. What should be clear, however, is that improved accountability is a fundamental requirement if there is to be a better match between improvements in livelihoods and improved governance, and that elections are not the only means to create better accountability. 9.5 ASSESSING THE INSTITUTIONAL CONTEXT A key contribution of this study concerns the importance of the social and institutional context in which local institutions function. The difficult social context of rural institutions in the developing world has been highlighted in many studies. But the importance of institutional linkages, across different types of institutions and across levels, is also fundamental. Local institutions not only are constrained by their limited financial and technical capacities, but too often are also isolated and without the necessary linkages even to the right sources of information and influence. Implementation of development projects that rely on local organizations should follow an assessment of existing institutional linkages, both among different types of institutions in the relevant sectors and across scales, and of links of the relevant institutions to different social groups. Such an assessment can be undertaken through widely available institutional mapping and assessment tools. This study has taken the first steps toward generating two such tools: Institutional Network Analysis, drawing on the literature on social networks and their analysis, and the Inclusion, Accountability, Participation, and Adaptive Capacity (IAPA) Index of institutions. Both tools can and should be refined further. But the use of the right tools to understand the nature and depth of institutional linkages will provide needed information to select appropriate local institutional partners. An assessment of the capacities of potential institutional partners will further improve the choice of partners. Appendixes A1 and A2 outline the methods used in this study to undertake a rapid and preliminary assessment of the governance capacities of local institutions and their network relationships and scores. 9.6 PROMOTING LOCAL INSTITUTIONAL AUTONOMY AND ADAPTABILITY For long-term sustainability, it is clear that local institutions need to be both more autonomous and more adaptable (Johnson and Wilson 2000). It is also clear from the case studies that self-reliant and adaptable local institutions are rare. But the reason for the lack of self-reliance and adaptive capacity among local institutions is not that they have received too much external support for too long a time. Rather, the absence of autonomy and adaptive capacity among local institutions seems to be a result of inadequate external support coupled with almost nonexistent capacities to generate resources or revenues locally, and substantial limitations on the ability of local institutions to manage their own affairs--both in terms of crafting rules and in terms of spending available resources. Greater autonomy and adaptive capacity would require stronger support to local institutions and a stronger defense of their ability to make decisions to organize their own affairs. Greater autonomy to local institutions can be enabled in a number of ways. Drawing on researchers' study experiences, autonomy and adaptive capacity can be increased by · Broadening institutional mandates; · Initially providing financial and technical support to institutions but in the longer run creating opportunities for local institutions to generate resources locally even if the amounts are small; · Supporting planning capacities within institutions; 38 · Developing institutional linkages to diverse social groups as well as to higher-level institutions; and · Creating greater transparency regarding institutional decision making and making decision makers accountable to constituents. 39 10. Conclusion This report has sought to identify the major factors that shape institutional impact on development outcomes. The study on which the report is based focuses on two types of outcomes--those related to governance itself, and those related to livelihoods. The findings of the study suggest that local institutional change is necessary to reshape development outcomes because institutions are ubiquitous and tend politically not to support existing social and economic asymmetries and inequalities. More than three decades ago, Schwartz observed that "externally introduced resources--technical, economic, political, social service--are distributed along lines of existing inequality and thus reinforce it unless accompanied by radical, massive institutional changes . . ." (1978, 249). His words are still true. The nature of existing inequalities and patterns of access to assets and resources of different kinds are key factors affecting developmental outcomes (Bebbington and Perrault 1999). In examining the role of local institutions, the report focuses particularly on four different areas: the nature of governance, the relationship between governance and livelihoods, the interactions between an institution and its context, and finally, the role of external interventions in improving development outcomes. The findings of the report can be summarized in four major statements. One, different aspects of institutional governance--inclusion, accountability, participation, and adaptive capacity--represent different areas of institutional capacity and performance. Tensions often exist between high levels of performance on different governance indicators. In the MRI study, inclusion appeared most often to have a negative association with participation and accountability. On the other hand, participation and accountability had a positive relationship. Finally, the studied institutions mostly had low adaptive capacity. Two, higher levels of governance seem to have little positive correlation with improvements in livelihoods. In specific countries, and for specific types of institutions--among them SHGs in Orissa, India, and NRM institutions in Yemen--there is some positive association between differences in levels of livelihoods and those in governance. But the research identifies a more interesting relationship between improved governance (in terms of greater downward accountability) and improvements in the equity of livelihoods outcomes. Three, the report finds that most rural institutions are located in social and political contexts that hamper institutional performance. At the same time, institutional networks--the links between different institutions--have a positive impact on outcomes. The report uses social network analysis to examine institutional network arrangements and discovers a clear relationship between institutional linkages and livelihoods improvement as well as equitable allocation of benefits through institutions. Finally, the study finds that substantial external support is necessary so that local institutions can fulfill the promise embodied in them. External support can occur in many different forms. One arena in which institutional support of local institutions has been particularly lacking concerns efforts to build relationships between different institutions at a local level, across levels. The above findings have several important implications for development projects implemented by the Bank. Given the low capacity of most local institutions, it is necessary to channel substantial resources to them. Improving the ability of local institutions to play a useful role in the implementation of development projects will require such investments. Fears that external support will lead to dependency relationships between external support institutions and local institutions are likely exaggerated in light of the generally low resources local institutions possess and receive. Together with promotion of local 40 institutions, strengthening links between different institutions both at a given level and across levels is likely to yield positive livelihoods outcomes. Moreover, strengthened linkages make it possible for local institutions to draw on the resources of other institutions so as to complement or fill gaps in their own resources. Improvements in governance outcomes are more likely to yield dividends in terms of improving equity in distribution of institutional benefits. But greater equity in livelihoods benefits would be no bad thing as far as the Bank's development projects are concerned. 41 Glossary of Key Terms Accountability: the proactive process through which public officials provide information about their plans, their actions, and results; justify and explain them; and can be sanctioned by those on behalf of whom they act (Ackerman 2005, 5­6). Accountability can be upward, downward, or horizontal. It can also be internal or external. The direction and orientation of accountability depend on the direction of flow of influence among actors who are in an accountability relationship. Accountability is viewed as necessary to combat three kinds of threats to good governance: corruption, clientelism, and capture. Adaptability (of institutions): the ability of institutions to address new environmental challenges or take advantage of new opportunities in different functional areas. Community: a group of individuals or households often viewed as being small in size, having multistranded interactions, and sharing a place. Community members are expected to be united by commonality of purpose, characteristics, beliefs, and/or actions. Nonetheless, most communities are also internally differentiated (Agrawal and Gibson 1999). Community-based organizations: membership organizations comprising a group of individuals/households in a self-defined community who have joined together to further common interests (Perrin, Van Der Riet, and Sjorslev 2006, 17) Decentralization: downward redistribution of resources, responsibilities, and decision-making powers in a territorial and administrative hierarchy (Agrawal and Ribot 1999). Inclusion (by institutions): the extent to which an institution effectively provides services to intended constituents. Coverage of poorer and more marginal social groups is particularly important for effective institutional inclusion. Institutions: human-created formal and informal mechanisms that shape social and individual expectations, interactions, and behavior (Ostrom 1990; North 1990; Bates 1980). Institutions structure and shape outcomes through the actions of individuals and decision makers associated with them. To understand their impacts it is necessary to examine their internal processes, external relationships, and linkages with different social groups and households. Livelihoods: the capabilities and material and social assets necessary for a means of living; includes the idea of coping with and recovery from external stresses (Chambers and Conway 1992; Scoones 1998), and the sustainability of the resource base on which livelihoods depend (Ashley and Carney 1999; Norton and Foster 2001). local governance: the political and administrative effects of a set of local institutions in a given context. For the purposes of this report, local governance hinges on performance related to inclusion, accountability, participation, and adaptive capacity. Local governments: constitutional or statutory legal entities mandated to govern within a local jurisdiction with some degree of fiscal, administrative, and political discretion. Organizations: concrete manifestation of institutions with an identifiable location, personnel, and rule structure. For this report, classifiable into three broad, sometimes overlapping sectors: public (bureaucratic administrative units and elected local governments), civic (membership and cooperative organizations), and market sectors (service and business organizations) (Uphoff and Buck 2006, 47). 42 Participation: the process through which stakeholders influence and share control over development initiatives and the decisions and resources that affect them (World Bank, 1996). Participation is operationalized in this research as the extent to and ways in which those served by an institution are involved in meetings and decision-making processes of the institution. Resilience: the amount of change a system can undergo and still retain the same function and structure while retaining options to develop in desired directions (Berkes et al. 2003; Holling 1973; Nelson et al. 2007). Territorial development: a development approach focusing on productive transformation of agricultural and nonagricultural activities in a well-defined distinct territory through institutional development that links people, local governments, economic and civil society organizations, and higher-level government institutions (de Janvry and Sadoulet 2004). The heterogeneity of territories necessitates careful modulation of public policies to meet the needs of affected parties in a given territory. 43 References Abdolmaola, H. A., and A. S. Ali. 1987. "The Role of Agricultural Cooperatives in the Development of Yemen Society." Arab Agricultural Engineer 21: 49­56. Ackerman, John. 2005. "Social Accountability in the Public Sector: A Conceptual Discussion." Social Development Department (SDV) Paper 82, World Bank, Washington, DC. Agrawal, Arun. 2001. "Common Property Institutions and Sustainable Governance of Resources." World Development 29 (10): 1649­72. Agrawal, Arun, and Gautam Yadama. 1997. "How Do Local Institutions Mediate Market and Population Pressures upon Resources?" Development and Change 28: 435­65. Agrawal, Arun and Clark Gibson. 1999. Enchantment and disenchantment: The role of community in natural resource conservation. World Development. 27(4): 629-49. Agrawal, Arun and Jesse Ribot. 1999. Accountability in decentralization: A framework with South Asian and West African cases. Journal of Developing Areas 33: 473-502. Agrawal, Arun and Ashwini Chhatre. 2006. Explaining Success on the Commons: Community Forest Governance in the Indian Himalaya. World Development 34(1): 149-66. Agrawal, Arun, Ashwini Chhatre, and Rebecca Hardin. 2008. Changing governance of the world's forests. Science 320(5882): 1460-62. Alemu, D., E. Gabre-Madhin, and S. Dejene. 2006. "From Farmer to Market and Market to Farmer: Characterizing Smallholder Commercialization in Ethiopia." Ethiopia Strategy Support Program (ESSP) Working Paper, IFPRI, Washington, DC. Alemu, D., H. de Groote, and D. Bacha. 2006. "The Role of Market Information System in Improving Rural Livelihood and the Status of the Service in Ethiopia." Ethiopian Journal of Agricultural Economics 6 (1): 87­108. Ashley, C. & Carney, D. 1999. Sustainable livelihoods: Lessons from early experience. London: Department for International Development (DFID). http://www.livelihoods.org/info/docs/nrcadc.pdf) Barakat, Sultan. 2006. "Mid-Term Evaluation Report of the National Solidarity Programme (NSP), Afghanistan." Post-War Reconstruction and Development Unit, University of York/Ministry of Rural Rehabilitation and Development, York, U.K. Bates, Robert. 1980. Markets and States in Tropical Africa. Berkeley and Los Angeles: University of California Press. Bebbington, A. J., and T. F. Carroll. 2000. "Induced Social Capital and Federations of the Rural Poor." Social Capital Initiative Working Paper 19, World Bank, Washington, DC. 44 Bebbington, A. J., and T. Perreault. 1999. "Social Capital, Development, and Access to Resources in Highland Ecuador." Economic Geography 75 (4): 395­418. Berkes, F., Colding, J., Folke, C. (Eds.), 2003. Navigating Socialecological Systems: Building Resilience for Complexity and Change. Cambridge University Press, Cambridge, UK. Brown, R. 2008. "Local Institutional Development and Organizational Change for Advancing Sustainable Urban Water Futures." Environmental Management 41 (2): 221­33. Carapico, Sheila. 1998. Civil Society in Yemen: The Political Economy of Activism in Modern Arabia. New York: Cambridge University Press. Chambers, Robert and Gordon R. Conway. 1992. Sustainable rural livelihoods: Practical concepts for the 21st century. Discussion Paper # 296. Institute of Development Studies, University of Sussex: Brighton. Chen, M., R. Jhabvala, R. Kanbur, and C. Richards. 2006. "Membership Based Organization of the Poor: Concepts, Experience and Policy." In MBOP: Membership Based Organizations of the Poor, ed. M. Chen, R. Jhabvala, R. Kanbur, and C. Richards. Forthcoming. New York: Routledge. Cornell University College of Arts and Sciences. http://www.arts.cornell.edu/poverty/kanbur/MBOPVolumeOverview.pdf (accessed April 9, 2008). Cleaver, Kevin. 1993. A Strategy to Develop Agriculture in Sub-Saharan Africa and a Focus for the World Bank. Africa Technical Department Series. Washington, DC: World Bank. Colburn, M. 2002. The Republic of Yemen: Development Challenges in the 21st Century. London: Catholic Institute of International Relations. Collion, M-H., and P. Rondot. 1998. "Background, Discussions, and Recommendations." In Agricultural Producer Organizations: Their Contribution to Rural Capacity Building and Poverty Reduction, ed. P. Rondot and M-H. Collion. Washington, DC: World Bank. Crouch, C., and H. Farrell. 2004. "Breaking the Path of Institutional Development: Alternatives to the New Determinism." Rationality and Society 16 (1): 5­43. Davis, D., D. Hume, and P. Woodhouse. 1994. "Decentralization by Default: Local Governance and the View from the Village in the Gambia." Public Administration and Development 14 (3): 253­69. Day, G. 1994. "Working with the Grain? Toward Sustainable Rural and Community Development." Journal of Rural Studies 14 (1): 89­105. de Janvry, A. and E. Sadoulet. 2004. Toward a territorial approach to rural development: International experiences and implications for Mexico's micro-regions strategy. Mimeo. University of California, Berkeley Dinh Tuan Minh. 2008. "TFESSD Mobilizing Rural Institutions for Sustainable Livelihoods and Equitable Development: The Case of Rural Producer Organizations in Vietnam." Country case study, Social Development Department and Agriculture and Rural Development Department, World Bank, Washington, DC. 45 Dorward, A., J. Kydd, J. Morrison, and I. Urey. 2004. "A Policy Agenda for Pro-Poor Agricultural Growth." World Development 32 (1): 73­89. Entwistle, Barbara, Katherine Faust, Ronald R. Rindfuss, and Toshiko Kaneda. 2007. "Networks and Contexts: Variation in the Structure of Social Ties." American Journal of Sociology 112 (5): 1495­ 1533. Gabre-Madhin, E.Z. 2001. Market Institutions, Transaction Costs, and Social Capital in the Ethiopian Grain Market. IFPRI Research Report 124. Washington, DC: IFPRI. Gabre-Madhin, E.Z., W. Amha, E. Tafara, J. Schluter, T. Teshome, and G. Kilkile. 2003. Getting Markets Right in Ethiopia: An Institutional and Legal Analysis of Grain and Coffee Marketing. IFAD (International Fund for Agricultural Development) project document. Washington, DC: IFPRI. Goldsmith, A. A. 1992. "Institutions and Planned Socio-economic Change: Four Approaches." Public Administration Review 52 (6): 582­7. Gopal, Gita, and Alexandre Marc. 1994. "World Bank­Financed Projects with Community Participation." Discussion Paper 265, Technical Department, Africa Region, World Bank, Washington, DC. Grootaert, Christiaan. 2001. "Does Social Capital Help the Poor? A Synthesis of Findings from the Local Level Institutions Studies in Bolivia, Burkina Faso and Indonesia." Local Level Institutions Working Paper 10, Social Development Department, World Bank, Washington, DC. Grootaert, Christiaan, and Deepa Narayan. 2000. "Local Institutions, Poverty and Household Welfare in Bolivia." Local Level Institutions Working Paper 9, Social Development Department, World Bank, Washington, DC. Helling, Louis, Rodrigo Serrano, and David Warren. 2005. "Linking Community Empowerment, Decentralized Governance, and Public Service Provision Through a Local Development Framework." Social Protection Discussion Paper 535, World Bank, Washington, DC. Holling, C.S., 1973. Resilience and stability of ecological systems. Annual Review of Ecology and Systematics 4: 1­23. Hue Le. 2008. "Mobilizing Rural Institutions for Sustainable Livelihoods and Equitable Development: Natural Resource Organizations, Vietnam." Country case study, Social Development Department and Agriculture and Rural Development Department, World Bank, Washington, DC. Huppi, Monica, and Gershon Feder. 1989. "The Role of Groups and Credit Cooperatives in Rural Lending." Policy, Planning, and Research Working Paper 284, Agriculture and Rural Development Department, World Bank, Washington, DC. IFAD (International Fund for Agricultural Development). 2003. "Transforming Rural Institutions in Order to Reach the Millennium Development Goals." Discussion paper, IFAD, Rome. IFRI (International Forest Resources and Institutions). 2005. "Data Collection Instruments." School of Natural Resources and Environment, University of Michigan. http://www.sitemaker.umich.edu/ifri/files/ifri_manual.pdf (accessed March 16, 2007). Johnson, H., and G. Wilson. 2000. "Biting the Bullet: Civil Society, Social Learning, and the Transformation of Local Governance." World Development 28 (11): 1891­1906. 46 Jones, William. 1995. The World Bank and Irrigation. Washington, DC: World Bank. Kloos, H. 1991. "Peasant Irrigation Development and Food Production in Ethiopia." Geographical Journal 157: 295­306. Krishna, A., N. Uphoff, and M. J. Esman, eds. 1997. Reasons for Hope: Instructive Experiences in Rural Development. West Hartford, CT: Kumarian Press. Lele, U. 1981. "Cooperatives and the Poor: A Comparative Perspective." World Development 9 (1): 55­72. Lemos, Maria, and Arun Agrawal. 2006. "Environmental Governance." Annual Review of Environment and Resources 31: 297­325. McGinnis, M., ed. 1999. Polycentric Governance and Development. Ann Arbor: University of Michigan Press. MacLeod, G. 2001. "Beyond Soft Institutionalism: Accumulation, Regulation, and Their Geographical Fixes." Environment and Planning A 33 (7): 1145­67. Mahanty, S. 2002. "Conservation and Development Interventions as Networks: The Case of the India Ecodevelopment Project, Karnataka." World Development 30 (8): 1369­86. Messer, Norman, and Philip Townsley. 2003. Local Institutions and Livelihoods: Guidelines for Analysis. Rome: Rural Development Division, Food and Agriculture Organization of the United Nations. Mitchell, Ronald B. 2003. "International Environmental Agreements: A Survey of Their Features, Formation, Effects." Annual Review of Environment and Resources 28: 429­61. Nelson, Donald R., W. Neil Adger, and Katrina Brown. 2007. Adaptation to environmental change: Contributions of a resilience framework. Annual Review of Anthropology 32: 395-419. Nixon, Hamish. 2008a. "The Changing Face of Local Governance? Community Development Councils in Afghanistan." Working Paper Series, Afghanistan Research and Evaluation Unit, Kabul. ------. 2008b. "Afghanistan Institutional Case Study: Community Development Councils." Country case study, Afghanistan Research and Evaluation Unit, Kabul. North, Douglass C. 1990. Institutions, Institutional Change, and Economic Performance. New York: Cambridge University Press. Norton, Andy and Mick Foster. 2001. The potential of using sustainable livelihoods approaches in poverty reduction strategy papers. Working Paper 148. London: Overseas Development Institute (ODI). Pp. 44. Ostrom, Elinor. 1990. Governing the Commons. New York: Cambridge University Press. ------. 1998. "The Institutional Analysis and Development Approach." In Designing Institutions for Environmental and Resource Management, ed. Edna Tusak Loehman and D. Marc Kilgour, 68­90. Cheltenham, UK: Edward Elgar. ------. 2005. Understanding Institutional Diversity. Princeton: Princeton University Press. 47 Ostrom, Elinor and Vincent Ostrom. 2004. "The Quest for Meaning in Public Choice." American Journal of Economics and Sociology 63 (1): 105­47. Ostrom, V., D. Feeny, and H. Picht, eds. 1988. Rethinking Institutional Analysis and Development. San Francisco: Institute for Contemporary Studies. Perrin, Nicolas, Rex Van Der Riet, and Jens Sjorslev. 2006. "Stocktaking of Selected World Bank Projects Working with Rural Institutions for Sustainable Livelihoods and Equitable Development." Draft, World Bank, Washington, DC. Pretty, Jules. 2003. Social capital and the collective management of resources. Science 302: 1912-15. Rice, R. A. 2001. "Noble Goals and Challenging Terrain: Organic and Fair Trade Coffee Movements in the Global Marketplace." Journal of Agricultural and Environmental Ethics 14 (1): 39­66. Roland, G. 2004. "Understanding Institutional Change: Fast Moving and Slow Moving Institutions." Studies in Comparative International Development 38 (4): 109­31. Rondinelli, D. A. 1981. "Government Decentralization in Comparative Perspective: Theory and Practice in Developing Countries." International Review of Administrative Sciences 48 (2): 133­45. Sasmal, S. 2008a. "Mobilizing Rural Institutions for Sustainable Livelihoods and Equitable Development: Case Study--Rural Institutions in Orissa (India)." Draft report prepared for the TFESSD Mobilizing Rural Institutions for Sustainable Livelihoods project of the World Bank, Sutra Consulting, Bhubaneshwar. ------. 2008b. "Mobilizing Rural Institutions for Sustainable Livelihoods and Equitable Development: Institutional Case Study--Self Help Groups in Orissa (India)." Draft report prepared for the TFESSD Mobilizing Rural Institutions for Sustainable Livelihoods project of the World Bank, Sutra Consulting, Bhubaneshwar. ------. 2008c. "Mobilizing Rural Institutions for Sustainable Livelihoods and Equitable Development: Institutional Case Study--Pani Panchayats in Orissa (India)." Draft report prepared for the TFESSD Mobilizing Rural Institutions for Sustainable Livelihoods project of the World Bank, Sutra Consulting, Bhubaneshwar. ------. 2008d. "Mobilizing Rural Institutions for Sustainable Livelihoods and Equitable Development: Institutional Case Study--Gram Panchayats in Orissa (India)." Draft report prepared for the TFESSD Mobilizing Rural Institutions for Sustainable Livelihoods project of the World Bank, Sutra Consulting, Bhubaneshwar. Schwartz, Norman B. 1978. Community development and cultural change in Latin America. Annual Review of Anthropology 7: 235-61. Scoones, Ian. 1998. Sustainable Rural Livelihoods: a framework for analysis. IDS Working Paper no. 72, Brighton: Institute for Development Studies. Scott, J. 2000. Social Network Analysis: A Handbook. 2nd ed. London: Sage. Scott, James C. 1998. Seeing Like a State: How Certain Schemes to Improve the Human Condition Have Failed. New Haven, CT: Yale University Press. 48 Sharief, Ismail. 2008a. "A Report on: Institutional Case Studies: Agricultural Commodities Domain, Agricultural Cooperative Societies (ACSs)." Draft report prepared for the TFESSD Mobilizing Rural Institutions for Sustainable Livelihoods project, World Bank, Washington, DC. ------. 2008b. "Report on: Institutional Case Studies: Republic of Yemen." Draft report prepared for the TFESSD Mobilizing Rural Institutions for Sustainable Livelihoods project, World Bank, Washington, DC. ------. 2008c. "A Report on: Institutional Case Studies: Natural Resources Management Domain, Tribal Structures/Systems and Water Users Associations." Draft report prepared for the TFESSD Mobilizing Rural Institutions for Sustainable Livelihoods project, World Bank, Washington, DC. ------. 2008d. "A Report on: Institutional Case Studies: Governance Domain, District Local Councils (DLCs)." Draft report prepared for the TFESSD Mobilizing Rural Institutions for Sustainable Livelihoods project, World Bank, Washington, DC. Shotton, Roger, and Mike Winter. 2005. Delivering the Goods: Building Local Government Capacity to Achieve the Millennium Development Goals. New York: United Nations Capital Development Fund. Sick, D. 2008. "Social Contexts and Consequences of Institutional Change in Common Pool Resource Management." Society and Natural Resources 21 (2): 94­105. Skocpol, Theda, and Margaret Somers. 1980. "The Uses of Comparative History in Macrosocial Inquiry." Comparative Studies in Society and History 22: 174­97. Smith, B. C. 1985. Decentralization: The Territorial Dimension of the State. London: George Allen and Unwin. Spielman, D. 2008. "Mobilizing Rural Institutions for Sustainable Livelihoods and Equitable Development: A Case Study of Farmer Cooperatives in Ethiopia: An Overview." Report, IFPRI, Addis Ababa. Spielman, D., and D. Kelemework. 2008. "Mobilizing Rural Institutions for Sustainable Livelihoods and Equitable Development: A Case Study of Irrigation, Water Management, and Smallholder Cooperatives in Ethiopia." Draft report, IFPRI, Addis Ababa. Spielman, D., M. J. Cohen, and T. Mogues. 2008. "Mobilizing Rural Institutions for Sustainable Livelihoods and Equitable Development: A Case Study of Local Governance and Smallholder Cooperatives in Ethiopia." Draft report, IFPRI, Addis Ababa. Sunderlin, William, Arild Angelsen, Brian Belcher, Paul Bergers, Robert Nasi, Levania Santoso, and Sven Wunder. 2005. Livelihoods, forests, and conservation in developing countries: An overview. World Development 33(9): 1383-1402. Svendsen, Mark, and David Groenfeldt. 2000. Case Studies in Participatory Irrigation Management. Washington, DC: World Bank. Tanguy, Bernard, and David J. Spielman. 2008. "Mobilizing Rural Institutions for Sustainable Livelihoods and Equitable Development: A Case Study of Agricultural Marketing and Smallholder Cooperatives in Ethiopia." Country case study, IFPRI, Washington, DC. 49 Thomas-Slayter, B. 1994. "Structural Change, Power Politics, and Community Organizations in Africa: Challenging the Patterns, Puzzles, and Paradoxes." World Development 22 (10): 1479­90. Tran, T. Ha. 2008. Governance institution: The case of the Grassroots Steering Committee. Mimeo. Center for Agricultural Policy, Institute for Policy and Strategy for Agriculture and Rural Development, Hanoi. Uphoff, Norman. 1997. Local Institutional Development: An Analytical Sourcebook with Cases. Ithaca: Cornell University Press. Uphoff, Norman, and Louise Buck. 2006. "Strengthening Rural Local Institutional Capacities for Sustainable Livelihoods and Equitable Development." Paper prepared for the Social Development Department, World Bank, Washington, DC. Van den Berg, M., and R. Ruben. 2006. "Small Scale Irrigation and Income Distribution in Ethiopia." Journal of Development Studies 42 (5): 868­80. Varisco, Daniel Martin. 1991. "The Future of Terrace Farming in Yemen: A Development Dilemma." Agriculture and Human Values 8 (1-2): 166­72. Wasserman, S., and K. Faust. 1994. Social Network Analysis. New York: Cambridge University Press. Weinberger, Katinka and Johannes Paul Jutting. 2001. Women's participation in local organizations: Conditions and constraints. World Development 29(8): 1391-1404. World Bank. 1996. The World Bank Participation Sourcebook. Washington, DC: World Bank. ------. 2001. "The Role of Rural Producer Organizations in the World Bank Rural Development Strategy." Rural Development Strategy Background Paper 8, World Bank, Washington, DC. ------. 2003. Reaching the Rural Poor: A Renewed Strategy for Rural Development. Washington, DC: World Bank. ------. 2004. "Community Driven Development in Local Government Capacity Building Projects: Emerging Approaches in Africa." Social Development Notes 86, World Bank, Washington, DC. ------. 2005. Linking Community Empowerment, Decentralized Governance, and Public Service Provision Through a Local Development Framework. Washington, DC: World Bank. World Bank Poverty Reduction Group (PRMPR) and Social Development Department (SDV). 2003. A User's Guide to Poverty and Social Impact Analysis. Washington, DC: World Bank. Yin, Robert K. 2003. Case Study Research: Design and Methods. London: Sage. Zia, M. Ehsan. 2008. "Afghan Aid That Works." Christian Science Monitor, May 16. 50 Appendix 1. The IAPA Governance Index Objective The Mobilizing Rural Institutions Project seeks to identify organizational characteristics that enhance rural livelihoods. It does so by focusing in part on institutional governance. It seeks to understand how institutions are related to each other so as to investigate whether and how better governance through institutions can improve livelihoods outcomes. To develop a concise and comprehensive measure of institutional governance, the project focuses on four major aspects of governance: Inclusion, Accountability, Participation, and Adaptability. A given institution can be scored on these four dimensions; the combination of the component scores creates the IAPA Governance Index. Content While the numerical value of the index is itself a measure of the governance capacity of the given institution, it can also be used to understand the relationship of governance to livelihoods outcomes. To the extent the index can be used to assess the governance capacity of institutions in any local context, it can be a useful tool for policy and implementation projects that view local institutions as an important actor/partner in development and conservation efforts. The following dimensions are covered by the index: · Inclusion by the institution refers to the extent to which members of the group supposed to be covered by the group are actually included, and the extent to which the institution focuses on marginalized members in the area of its operation. · Accountability of institutional decision makers relates to two aspects of accountability: information sharing and power. · Participation in the institution concerns involvement in day-to-day, regular activities of the institution. · Adaptability is the capacity of the institution to address challenges and take advantage of opportunities arising in its environment. The construction of the IAPA index relies on information that can be collected in a short focus group discussion with institutional stakeholders. Operationalizing the IAPA Governance Index Inclusion The first aspect of Inclusion is the extent to which households that are supposed to be covered by the institution are actually covered by it, either as members or as beneficiaries depending on the kind of activities performed by the institution. The ratio of the number of households that are actually covered by the institution to the number of households that are eligible to be covered by the institution yields the percentage of households included by the institution. The second aspect of Inclusion is the extent to which it covers marginalized households. (Depending on the context, marginality can be defined in relation to indigenous groups, caste, women-headed households, or the poor). This aspect is calculated as the ratio of the number of marginal households actually covered by the institution to the number of marginal households in the village (or settlement). The Inclusion score of the IAPA index is the simple average of these two percentages. Example: Suppose the total number of eligible households in a settlement is 100, of which 60 are covered by a microfinance institution. Fifty of the eligible households belong to marginal groups. If the number of 51 marginal households covered by the institution is 40, then its scores are 0.60 on the first aspect of inclusion and 0.80 on the second aspect. The average of these two scores, i.e., the microfinance society's Inclusion score, is 0.70. Accountability Accountability has been viewed as a critically important aspect of governance. Accountability is the ability to sanction decision makers for inappropriate decisions or for nonperformance. For the IAPA index, we quantify two elements that are critical to effective institutional accountability--information and power. In examining information-related aspects of accountability, we seek to identify the extent to which the leaders of the institution share information about institutional activities and different aspects of its performance with rank-and-file members. Information sharing and transparency is critically important for members to hold decision makers and leaders accountable in relation to the institution's strategic direction, finances, personnel changes, and decision making. Sharing of information in a given time period about any of these areas in a given time period earns a score of 0.25 in relation to information sharing, and if decision makers share information in all the four areas with members and staff, the measure should be scored as 1.0. In terms of accountability, we are primarily interested in downward accountability of the institution to its members, but also recognize that upward accountability can be a means to check arbitrary decision making. Decision makers can be accountable to superiors or to constituents--if they are accountable to both, the score for the Index is 1.0; accountability to one or the other group earns a score of 0.5. Example: Suppose the same microfinance society's decision makers are willing to share information about their activities and decision making in three of the four domains (strategic direction, personnel changes, and decision making), but not about the society's finances. In this case, the information accountability score of the society is 0.75. If the decision makers can be sanctioned by superiors but not by the members of the institution, then the power accountability score of the institution is 0.5. The averaged accountability score of the microfinance society is 0.625. Participation The complexity of the term and its many different meanings suggest that any simple measure of participation will necessarily be difficult to defend. For the purposes of the IAPA index, we focus on the most common and intuitively obvious meaning of the term: to what extent are the institution's members involved in the meetings and discussions of the institution. The index takes into account both the frequency of meetings and involvement of members in meetings. The main difficulty in representing the frequency of meetings is the need to establish how members feel about what the appropriate number of meetings in a given time period should be. For many formal organizations, this number may be indicated in the organization's bylaws or constitution. But for both formal organizations and informal institutions, it is also necessary to identify what the members believe is an appropriate or desired number of meetings. The proportion of actual to the desired number of meetings yields one score on the participation dimension. The second aspect of the participation dimension concerns the attendance of members at the organizational meetings. The average attendance at meetings provides the second score for the participation dimension. Example: If the constitution of our microfinance society provides for a monthly meeting and the number of meetings that were actually held in the year is 10, the first score of the participation dimension would be 0.83. If the average attendance at these meetings (members present as a proportion of total members) was 70 percent, then the averaged participation score for the microfinance society turns out to be 0.766. Adaptability To understand the adaptability of the institution, it is necessary to identify, through a focus group discussion with different organizational stakeholders, the major challenges and opportunities that arose in the organizational context/environment during a specified past period. We use the past five years as a 52 convenient period because organizational stakeholders are likely to retain a reasonably clear memory of this time period and because it is likely to represent a long-enough time period for at least some challenges or opportunities for the organization to have emerged. To assess organizational adaptability, we identify six domains of organizational functioning: · strategic planning; · finance and accounting; · operations and maintenance; · human resources; · marketing; and · information management. Environmental opportunities and threats could affect a single functional domain or more than one at once. The crucial point here is to identify the different organizational functions that a given external challenge affected, and whether the organization was effectively able to address these threats or opportunities. The institution's adaptability score is the proportion of challenges/opportunities it successfully addressed in different domains. Example: Suppose the microfinance society we are assessing encountered two challenges. The first was that the support it was receiving from one of its external donors was halved, as a consequence of which it had to lay off personnel. The second challenge concerned several borrowers who defaulted on the loans they had taken. As a result of reduced external support, there was additional pressure on the financial services that the microfinance society was providing in the village. Finally, it had to close down a training center it had started. This set of events means that it encountered challenges to three functional domains: personnel, two challenges related to finance, and operations. Suppose further that it was able to address one of the financial challenges by finding new sources of support from different aid organizations, and also restart its training center, but that it had not yet been able to address the problem of default, nor had it been able to hire people to fill empty positions. To calculate the adaptability score of the society, first count the domains in which it encountered challenges; in our example this number is 3. Successfully addressing all the challenges in a given domain increases the adaptability score by .33. Given that it successfully addressed the challenges it encountered in operations (restarted training center), and one of the two financial challenges (identified a new source of support, but did not address the problem of significant levels of default), and that it also did not successfully address the personnel difficulties, its adaptability score is .33 + .33/2 = .50. Final Calculation of the IAPA Index To calculate the final score for the IAPA governance index, one needs to add the scores received in all the four dimensions (note that they each vary between 0 and 1), and divide by 4 since there are four dimensions. In our case, the IAPA Index score turns out to be 0.70 + 0.625 + 0.766 + 0.50 / 4 = 0.65. Table A1.1 summarizes the four component measures used in calculating the IAPA Governance Index Scores. 53 Table A1.1 Calculating the IAPA Governance Index Score Dimension Aspects Potential information Measure source Inclusion 1. General coverage 1. Focus group 1. Member households as proportion of 2. Coverage of marginalized discussion all eligible households groups 2. Organizational 2. Marginal group member households records as proportion of all eligible marginal 3. household surveys group households Accountability 1. Information 1. Organizational 1. Upward and downward accountability 2. Power informants 2. Focus group discussion Participation 1. Frequency of meetings 1. Organizational 1. Actual meetings as a proportion of 2. Attendance bylaws; constitution desired meetings 2. Focus group 2. Proportion of members attending discussion meetings Adaptability 1. Successful response to 1. Focus group 1. Proportion of challenges in different environmental challenges and discussion domains addressed successfully opportunities The score for each component of the index falls between 0 and 1. The index score is calculated by adding the scores for each component and dividing the sum by 4. The IAPA Governance Index score is a continuous measure between 0 and 1. It can be especially helpful when used to address questions about whether high governance capacity is correlated with high livelihoods-related outcomes, whether particular organizational design features or characteristics promote governance capacity, or simply to compare the governance capacity of different rural institutions. However, the index score can also be interpreted on its own through the following suggested ranks: Superior or outstanding 0.75 to 1.00 High Less than 0.75 to 0.50 Low Less than 0.50 to 0.25 Minimal Less than 0.25 54 Appendix 2. Institutional Network Analysis and Mapping 1. What is institutional mapping? Institutional mapping for rural governance and livelihoods is the visual presentation of different institutions and their relationships in a given social context. Institutional maps present a large amount of information about characteristics of different institutions and how institutions are linked. 2. Why do institutional mapping? The effectiveness of rural institutions depends on a variety of factors, among them their relationship with other institutions within and outside the local context. Institutional maps provide an immediate sense of institutions and their links in a context. Thus institutional maps are powerful visual tools. Knowledge about capacities of institutions and the strength of their links provides an indication of key institutions and leverage points through which new policies can be introduced to maximal effect. Information on institutions and their links can be used to identify how and to what extent social and policy outcomes can be attributed to institutional arrangements versus other features of the social/political/economic context. 3. How to map institutions? The first step in institutional mapping is to identify the suite of institutions in a given rural context--formal and informal, governance and social service delivery, sectoral and general purpose, traditional and introduced. The framework below can serve as a checklist to identify institutions in a location. Table A2.1 Checklist of Institutions According to Formal/Informal and Public/Market/Collective Dimensions Public Market Collective Formal 1. Local 1. Service organizations 1. Membership organizations administrative 2. Private voluntary 2. Cooperatives (producer, savings, organizations organizations consumer, labor, marketing) 2. Local government 3. Private businesses 3. NGOs organizations 4. Banks 4. Women's, youth, religion, and 3. Education and 5. Contract farming ethnic group­based organizations training institutions arrangements Informal 1. Discussion 1. Exchange groups/norms 1. Community-based organizations meetings 2. Marketing arrangements 2. Self-help groups 2. Political party 3. Technology-sharing 3. Savings and credit networks arrangements (tractors, associations/revolving funds ploughs, cell phones) 4. User groups 4. Sharecropping 5. Gender and work norms arrangements 6. Caste institutions 5. Information-sharing norms The cells in table A2.1 can be filled out with the names and characteristics of many different institutions. A few examples are provided above for illustration but note that depending on context the specific institutions listed can be more or less informal. Institutional characteristics should be identified in terms of such intuitively important criteria as size, personnel, and budget; the services they perform for their constituents; and their relevant functional domains. 4. How to Combine Institutional Mapping with Governance Outcomes Once the relevant institutions have been identified, mapping them requires basic information about their relationships. Such information can be collected in focus group discussions with the most important 55 institutional stakeholders. Information about links between institutions should cover at least four aspects: whether a link exists, and if so, whether it concerns information flow, influence flow, or resource flow. For the sake of analytical convenience, this measure is binary: if any of the above aspects of a link is answered positively, it is coded as 1, and if it is absent, it is coded as 0. · Example of institutional mapping exercise: For a given village/settlement, one may generate a table of basic information related to its institutions and their links. Table A2.2 charts hypothetical data for five institutions. With this information, maps for institutional relationships in a given context can be generated easily using a number of available mapping software applications. Such software applications also allow users to select different measures of institutional linkages/configuration. Table A2.2 Hypothetical Data for Institutions in a Given Context Marketing coop Labor-sharing Irrigation Extension Rotating credit group institution agency group Marketing 1, 0, 1, 1 0, 0, 0, 0 1, 1, 0, 1 1, 0, 0, 1 coop Labor-sharing 1, 0, 0, 0 1, 0, 1, 1 0, 0, 0, 0 0, 0, 0, 0 group Irrigation 1, 0, 1, 1 1, 1, 1, 0 0, 0, 0, 0 1, 0, 1, 1 institution Extension 1, 1, 1, 1 1, 1, 0, 0 1, 0, 0, 0 0, 0, 0, 0 agency Rotating 0, 0, 1, 1 1, 0, 1, 1 1, 1, 1, 1 0, 0, 0, 0 credit group The numbers in each cell indicate whether a link between the corresponding row and column institutions exists and whether flows of information, influence, and resources exist. The information can be condensed by summing up the positive links so as to yield an overall indicator of the strength of each link (see figure A2.1). The map indicates, for example, that there are no links between extension-agency and credit groups, and that the strongest links are between credit and irrigation, and between marketing and extension groups. Table A2.2 does not contain any information about the institutions themselves. But this information can be added to the network so as to supplement the information on strength of links, and the direction of flow of influences. For example, we can use the IAPA Governance Index for each institution so that the size of each institutional node is proportional to the index score for that institution (see the technical note on the IAPA index for calculating the IAPA score for each institution. (See table A2.3.) Table A2.3 Hypothetical Governance and Livelihoods Scores for Institutions Governance Livelihoods Marketing Coop 0.75 0.7 Labor sharing group 0.4 0.9 Irrigation instn. 0.75 0.68 Extension agency 0.4 0.35 Rotating credit group 0.63 0.89 56 In figure A2.1, the governance capacity of the labor and extension groups (as denoted by the IAPA index) is so small as to be almost invisible. Marketing and irrigation groups have the greatest governance index scores. In addition, the square shape of three of the nodes denotes that they are informal institutions while the circular shape denotes formal institutions. Figure A2.1 5. How to Combine Institutional Mapping with Livelihoods Capitals Outcomes Institutional maps can also be supplemented with information about how different institutions contribute to livelihoods capitals. Although livelihoods is an inherently complex concept, for the purposes of institutional mapping we focus on contributions of specific institutions to different forms of livelihoods capitals. In a livelihoods framework, the influences of different institutional structures on livelihoods outcomes in turn lead to changes in stocks of five forms of livelihoods capitals: human, social, physical, natural, and financial. In assessing the role of each institution to livelihoods capitals, we adopt an approach in which we seek to identify whether the institution contributes to a particular form of capital, and score the institution between 0 and 1 based on the number of different types of capitals to which the institution contributes positively (If the institutions contributes to one type of livelihoods capital, it receives a score of 0.2; for two types of livelihoods capitals, it receives a score of 0.4, and so forth). Scored in this fashion, the IAPA score and the score for contribution to livelihoods capital both vary between 0 and 1. The information about the contribution to livelihoods capitals can either be listed next to the institution or incorporated within the node itself. In addition to representing the information on each institutional node visually, it is also possible through institutional mapping and network analysis software to undertake statistical analysis of the relationships between livelihoods scores and IAPA governance scores. In combination, the information about institutions and their links can present a highly comprehensive and concise picture of the institutional context within which policy interventions must unfold at a local level. 6. Caution in Institutional Mapping As the number of local institutions grows for a given study, the institutional map can quickly become unwieldy and uninformative. Therefore, it is essential to keep in mind the main issues to be conveyed by the institutional map. These pertain to the chief institutional relationships that are to be highlighted, the degree of articulation among local institutions, and who benefits from institutional relationships. 57 Institutional maps should always be accompanied by a careful text description that points out the highlights of institutional configuration. 58 Appendix 3. Membership-Based Rural Producer Organizations in Ethiopia, Vietnam, Yemen, and Orissa (India) A3.1 ETHIOPIA'S SMALLHOLDER AND AGRICULTURAL MARKETING COOPERATIVES The basic goal of the studied rural producer organizations in Ethiopia was similar: to promote pro-poor development by better linking farmers and rural households to markets. For example, some of the key problems for agricultural producers in Ethiopia continue to be high transactions costs associated with trading agricultural commodities. These are reflected in poor information systems and infrastructure and weak private-sector capacity, preventing exchange of price information and surpluses from markets (Alemu, Gabre-Medhin, and Dejene 2006; Alemu, de Groote, and Bacha 2006; Gabre-Medhin 2001; Gabre- Medhin et al. 2003). These general features aside, the marketing chains involved in the movement of crops from farmers' fields to markets are quite complex, with multiple links between producers, assemblers, traders, and consumers (for cereals) and export markets (for coffee). Although coffee production in Ethiopia is highly privatized in comparison with the production and marketing process for grains and cereals, the links in both sets of marketing chains have constraints related to "inequitable bargaining power, information asymmetries relating to commodity prices and quality, issues of quantities and scale economies, as well as standards and measures" (Tanguy and Spielman 2008, 10). In this context of relatively strong constraints on the performance of cooperatives, it is instructive to examine governance and livelihoods outcomes associated with the examined cooperatives in relation to inclusion of local residents, levels of participation and accountability, and contribution to livelihoods of members. Although marketing cooperatives are widely prevalent in Ethiopia, their links to smallholder farmers remain limited at best. In a survey of 205 cooperatives in 2005, Tanguy and Spielman found that only 9 percent of smallholders were members. Nonmembership is explained by a variety of factors, among them uncertainties about benefits from membership, concerns about cooperative management, and rejection of membership bids. In addition, participation in cooperatives reveals a pattern seen elsewhere: the poorest and the richest households find participation less beneficial, with the greater benefits going to the middle-income farmers (Tanguy and Spielman 2008; Weinberger and Jutting 2001). While these factors explain the low level of inclusion from the point of view of farmers, it is also likely the case that cooperatives themselves do not encourage high levels of inclusion so as to keep costs of coordination among heterogeneous members low. Thus even if cooperatives declare in principle that their membership is open to all, specific membership criteria related to residential location and costs of membership (by buying shares whose prices may be too high) can make it difficult for poorer households to become members. Similarly, although the members of the cooperatives are well aware of the existence and functions of the cooperative, just over a third of them are active in the cooperative's meetings and in its decision making. To add to the relatively pessimistic picture of the involvement of members in cooperative functioning, cooperative members tend not to find decision making transparent, or decision makers accountable (Tanguy and Spielman 2008, 24­31). These overall concerns about the nature of cooperative governance and performance related to inclusion, participation, and accountability are reflected to some degree in the contributions of cooperatives to livelihoods-related benefits. Two types of livelihoods benefits are important for rural households-- managing environmental risks and increases to overall livelihoods. In both areas the functioning and performance of the surveyed cooperatives can be improved substantially. 59 The reasons for low contributions to the management of environmental risks seem relatively clear: cooperatives are not designed to help members ameliorate fluctuations in livelihoods owing to floods, droughts, or insect infestations. Insurance of different kinds is more often a service other rural institutions--often private firms with assistance from governments--provide. With regard to increases in contributions to livelihoods, two points emerge clearly from the study of cooperatives in Ethiopia: First, cooperatives provide benefits geared primarily to the needs of members, and even low entry costs can significantly reduce the proportion of households who are cooperative members. Second, cooperatives are expected to provide not just marketing services but also a large number of other services, and the multiplicity of assigned responsibilities is often a strain on existing institutional capacity within cooperatives. Thus, even when cooperative involvement in marketing increases the prices members receive for their products, this happens for a relatively small number of households. Even with a low threshold of whether a cooperative had marketed any of its members' produce in the past year, only 60 percent of the cooperatives were able to respond in the affirmative. The analysis of the data on cooperatives in Ethiopia thus clearly uncovers a number of interesting results: Ethiopian cooperatives are not highly inclusive. They are also not particularly participatory or accountable institutions. They tend to provide a variety of services to their members, but provision of marketing services is not very widespread, in part because they are also expected to provide a variety of other services. Although member participation is positively associated with better service provision, high inclusion and member heterogeneity are negatively associated with effective service provision for improved livelihoods. A3.2 VIETNAMESE RURAL COOPERATIVES Cooperatives are widespread in Vietnam and are one of the more prominent examples of local rural institutions. The field research in Vietnam focused on four examples of cooperatives that varied in their size (just over 40 members to more than 1,100 members), degree of formal organization, and socialecological context (from the northern highlands to the southern delta region). The selected cooperatives sought to improve the kinds of services available to farmers for marketing a variety of different crops such as rice, cassava, and maize, and livestock and fish. The rural agricultural context in Vietnam is characterized by a variety of production, distribution, and marketing-related constraints. Low productivity of traditional crops such as rice, maize, and cassava; high mortality in aquaculture; underemployment; lack of needed knowledge; low levels of and access to capital; and increasing environmental pollution mean that cooperatives and RPOs can play a significant role in enhancing agricultural productivity and production levels. At the same time, the fact that rural producers organized to establish cooperatives in all the selected cases typically in the aftermath of some agricultural crisis or only after external support became available points to substantial obstacles to local collective action. The different cooperatives selected for the study are connected to other local institutions to varying degrees. Thus the Chieng Kheo Farmer Union in Son La has a complex set of relationships to more than 11 other local and extralocal institutions. In contrast, the Nam Thinh Clam Association has substantial relationships with only five other institutions. But in all cases the influence of the Communist Party and the government is visible and strong (Dinh 2008, 16). Similarly, there are substantial variations among the different cooperatives with respect to the complexity and formality of internal organization, resource availability, and strength of leadership, and in their changes since formation. These differences can be juxtaposed against differences in levels of inclusion of local residents and the participation of members in decision making, and the accountability of decision makers to members. 60 Broadly speaking, the larger cooperatives also seem to have more developed and formalized internal structures. However, governance outcomes seem to be associated with a number of different factors that echo patterns observed in Vietnam to a significant degree. Inclusion is higher the lower the costs of entry and membership in the cooperative. Higher levels of interference in the functioning of the cooperatives tend to lower the extent to which the cooperative's decision makers are accountable to their members. Further, cooperatives that are more directly focused on providing a narrow set of services to their members--the two associations focusing on improvement of clam cultivation and marketing--tend also to provide a superior level of services to members. In the Vietnamese case, there is a strong positive association between better governance and livelihood outcomes. None of the cooperatives are fully inclusive, but higher levels of member participation help mobilize higher levels of capital, better resolution of internal conflicts, greater accountability, and better livelihoods outcomes for members. For one of the cooperatives that does not conform to this pattern--a formally organized public cooperative--the explanation is likely related to high levels of external interference in the cooperative's internal affairs and decision making. The evidence collected on cooperatives in Vietnam allows a number of generalizations, some similar to those for Ethiopian cooperatives and others in interesting contrast. The studied cooperatives are not highly inclusive institutions. They also suffer significant intervention and influence in their internal decision making by Communist Party officials. Cooperatives whose functioning is strongly shaped by government tend to be more inclusive, but often are less effective. This is especially true for the agricultural cooperatives that were studied. Better governance in terms of participation and accountability is also associated with stronger contributions to improvements in livelihoods outcomes for members. A3.3 AGRICULTURAL COOPERATIVES IN YEMEN The study of rural producer organizations in Yemen is important for reasons similar to those on which the study of cooperatives in Vietnam and Ethiopia is based, but also in conjunction with the Bank- financed Rainfed Agriculture and Livestock Project and the work of the Social Fund for Development. Yemeni agriculture is an integral and important part of the national economy as agriculture and fisheries form a quarter of the total GDP. But agriculture also suffers problems of low integration of markets for factor input, commodities, and consumption goods; low levels of information and technological support to farmers; weak marketing networks; low productivity; and high transactions costs, as is the case for Ethiopia and parts of Vietnam (Colburn 2002; Sharief 2008a, 1­2). The benefits to improved performance of agricultural cooperatives can be especially high in light of the high costs of food imports and the growing Yemeni population (Varisco 1991). Agricultural cooperatives are also potentially important because of the high fragmentation of land holdings, low access of farmers to agricultural inputs such as irrigation, lack of access to credit and extension services, and periodic floods, rainfall failure, and animal disease outbreaks (Abdolmaola and Ali 1987). At the same time, the cooperative sector in Yemen needs significant support to perform to its potential. At present it receives relatively little support from donors. Although most cooperatives depend on government support through the Agricultural and Fisheries Promotion Fund, little of this support is directed toward improvements in the capacity of cooperatives to provide services to their members. The cooperative movement started only in 1992 and is a part of the growth of Yemen's civil society (Carapico 1998). Though it has now spread to all governorates, the total number of agricultural cooperative societies is still only about 850 with a membership of 130,000 households. The growth of cooperatives has been relatively rapid in the past 15 years, with the total increasing from less than 200 in 61 1990 to more than 850 by 2006. The increase in capital investment in the cooperatives has been even greater, from 141M Yrls to more than 5000M Yrls during the same period. Agricultural cooperatives function in an institutionally rich landscape with a significant number of different kinds of formal and informal and public, private, and civil society institutions. The four locations selected for the study were in Bani Matar, Al Mahweet, Abs, and Tuban districts, representing a variety of social and ecological contexts for the study and a range of values as regards the strength of the cooperatives. But the districts display similar constraints to agricultural production as discussed earlier in relation to Yemen and farmers need greater assistance with access to credit, marketing, extension, and irrigation than they are able to achieve on their own. The selected cooperatives vary significantly in their age (established between 1973 and 1997), size (from 100 to 5,000 members), degree of formality in internal structures and decision making, resource endowments, institutional capacity, and relationships with other institutions, especially the DLCs and the Agricultural Marketing Centers. These variations are related to the functioning and performance of the cooperatives. Overall, the cooperatives have not been very successful in addressing agricultural constraints faced by farmers. The services they have provided are not viewed by farmers as being very important. Nor do they perform very well in terms of their degree of inclusion, accountability to members, and member participation in decision making. The cooperatives are mostly inattentive to questions of gender equity. One reason for the low performance of the selected cooperatives lies in the fact they were all organized in response to external incentives rather than perceptions about local needs. At the same time, external resources to support the cooperatives have not remained available, leading to a loss of local interest in the cooperatives. Because the cooperatives are unable to provide similar services as are available to farmers from private operators, they do not see any reason to be involved closely with the cooperatives. Most view the cooperatives as inefficient (76 percent), corrupt (88 percent), and impermanent (66 percent) (Sharief 2008a). One cooperative that stands out from the rest is different in terms of its ability to contribute to livelihoods and to do relatively well in terms of governance. In this cooperative, a significant number of members produce mangoes and are relatively well-off. It is very likely that their wealth positively influences both their ability to hold the cooperative accountable and the positive performance of the cooperative. A3.4 ORISSA'S SELF-HELP GROUPS In Orissa the study focused on rural SHGs as the most prominent example of membership-based RPOs. SHGs are obviously different in their objectives and organization from cooperatives: They directly target a narrower set of constraints to rural or urban growth--access to credit--and they do so by focusing on relatively small-scale activities. The common features of agricultural cooperatives and rural microfinance SHGs lie in their effort to improve livelihoods benefits to rural populations, and do so by focusing on specific population groups. The comparison between cooperatives and SHGs provides some interesting and useful lessons, particularly as a result of the contrast between cooperatives and SHGs. The SHGs in Orissa aim primarily to address the credit needs of their members, one of the main bottlenecks for rural economic and productive activities. The rate of growth of SHGs in Orissa has been rapid and impressive, with nearly 0.23 million groups currently estimated to be in existence (Sasmal 2008b). 62 Efforts to build SHGs were promoted by a combination of state NGOs, the National Bank for Agriculture and Rural Development, some international donor support, and initiatives undertaken by the state government. With increasing numbers, efforts to create organizational linkages among existing SHGs have also been undertaken. Orissa's Self Help Cooperative Act has helped provide both a legal framework and legal status to the SHGs and their federations and cooperatives. Federations of SHGs have emerged now in 16 districts in Orissa, and the foundations for similar federations have been laid down in the remaining 14 districts as well. The study focused on a cluster of 12 SHGs in four gram panchayats in two districts in Orissa: Bhadrak and Kalahandi. Bhadrak has a higher per capita income compared with Kalahandi, and better human development indicators as well. The SHGs in both districts were promoted by NGOs; the Orissa state government's initiative to assist SHGs, Mission Shakti; and other local government agencies such as watershed committees. Indeed, a large number of other local organizations also exist in the studied locations in addition the SHGs, including other SHGs. All the studied SHGs were formed in the last decade, and as is true for most SHGs in Orissa, the membership of these SHGs is typically below 20 persons to avoid registration requirements under the Indian Company Registration Act 11(A), 1956. However, the small membership also keeps transactions costs down. Membership in the SHGs is open to any local resident, although there is a focus on women members. Most SHGs have either retained the same number of members as were present when the group was formed, or have lost some members since formation. Given the size of the groups, the organizational structure for most of them is quite flat, and decision making and functioning is typically highly participatory and equitable. A set of bylaws form the basis for decisions, and they tend to be similar across the SHGs. These include a small amount of saving each month (Rs 10 to 40 each month; approximately $0.25 to $1.00 at current exchange rates), use of a bank to manage the collective savings, relatively small amounts of loans to members (maximum loan amount is approximately $80.00), and interest rates ranging from 2 percent to 5 percent per month. The closest institutional linkages for the SHGs exist with the local bank and the gram panchayats, although many of them have also retained close connections with the agency that had initially facilitated their formation. However, most SHGs do not have horizontal linkages with each other either within the same village or across villages. Nor do they have many institutional links with other government agencies. On the whole, the SHGs in the study were inclusive to the extent their small size permits. Leadership of the SHGs tends to go to more well educated and articulate members, but there is usually a high rate of participation and strong accountability of leaders to members--loans cannot be denied to members without valid reasons, accounts are available to all for examination, and major financial transactions are discussed and decided in group meetings. The SHGs have also made small but significant contributions to member livelihoods, through individually and jointly oriented income-generating activities. Smaller groups that have not been able to mobilize significant savings, or those suffering from tensions among members, tend not to advance loans but focus on managing member savings. But all SHGs have helped meet emergency and other consumption needs of members, supported purchase of livestock and other livelihood opportunities, and helped reduce seasonal vulnerability of members. They have also been instrumental in helping generate higher levels of social capital and promoting a sense of greater empowerment among most members. Nonetheless, there are important differences in the levels of performance of the SHGs in Bhadrak versus those in Kalahandi. Members of SHGs in Kalahandi are less familiar with, less satisfied with, and less informed about their SHGs in comparison with those in Bhadrak. This difference can be explained partly by the greater poverty of households in Kalahandi and the lower institutional capacities of the SHGs. 63 What also emerges clearly from the evidence is the limited ability of the SHGs and their members to absorb risks, promote large-scale income- generating activities, and link with other organizations and groups. 64 Appendix 4: Natural Resource Management Institutions in Ethiopia, Yemen, Orissa, and Vietnam A4.1 SMALLHOLDER COOPERATIVES IN ETHIOPIA AND THEIR ROLE IN IRRIGATION AND WATER MANAGEMENT Despite extensive water resources in Ethiopia, water remains a scarce resource for most of the country's small farmers, who rely on single cropping systems depending on rainfall (Kloos 1991). The dependence of the Ethiopian farmers on temporally and spatially uncertain rainfall is the main reason for the notorious droughts and famines in Ethiopia. Traditional irrigation systems are the dominant mechanism for irrigation in Ethiopia, serving more than 250,000 hectares of land. Collectively, both medium- and large-scale modern irrigation arrangements benefit less than half that area (Van den Berg and Ruben 2006; Spielman and Kelemework 2008). And in all, irrigated land is no more than 5 percent to 6 percent of total cultivated land in Ethiopia. As a result, the potential impact of better irrigation can be revolutionary in terms of agricultural productivity in Ethiopia. Indeed, a number of efforts of the Ethiopian government recognize the basic importance of irrigation for improved agricultural production and greater food security. Its investments in irrigation are aimed to insulate smallholders from unpredictable and variable rainfall and have a positive impact on industrial development and economic growth as well. The government's Agriculture Development Led Industrialization (ADLI) policy is built on this assumption. The research study sites in Ethiopia were selected from a range of agroclimatic zones and with different socioeconomic characteristics. Cooperatives in these sites are an increasingly important agent of agricultural transformation as the government looks to them for input supplies and output marketing. However, only 3 percent of farmers are members of an irrigation cooperative, water users' association, or any other water management organization (Spielman and Kelemework 2008, 7). Nor is the membership in collective institutions widespread among those who use irrigation--less than a tenth of farmers with irrigated land are members of irrigation institutions. Thus there is significant potential for a greater role for irrigation institutions in improving access to water for farmers, and for applying the lessons from the study to scale up irrigation-related interventions in water management and distribution. Most irrigation cooperatives in Ethiopia are organized to serve households that fall within the command area of a specific irrigation scheme and do so in accordance with guidelines developed by the federal, regional, and/or Woreda authorities. Thus, the fortunes of the cooperative and its membership are often tied to decisions over which they have relatively little influence. Analogously, the organizational design and structure of different irrigation cooperatives are similar across the organizations. The views of members about the effectiveness, responsiveness, transparency, and accountability of the irrigation cooperatives are fairly divided. In a survey of their perceptions roughly half of the members agreed that their cooperatives were working well on a variety of indicators, whereas the other half disagreed! Typically, the irrigation cooperatives are open to any local residents who have irrigable land, and in that sense are not exclusive organizations. There is some bias toward the richer residents of the locality, which in part may be explained by the fact that the poorer households are less likely to own irrigable land. The 65 cooperatives are open to close involvement of members in meetings and discussions around decisions, but tend not to have high levels of deep participation in terms of consultative decision making and regular rotation of leadership positions. Only one of the five studied cooperatives could be characterized as having high levels of participation (Spielman and Kelemework 2008, 16). The tendency toward low to medium levels of inclusion and participation among the cooperatives is not matched by the level of the cooperatives' responsiveness and accountability to members and their interests. Apart from one cooperative that is essentially nonfunctional, all the remaining cooperatives are highly responsive to their members' interests and demands, and also have medium to high levels of accountability. The irrigation cooperatives' impact on livelihoods is difficult to evaluate because of the multiple causal links between what they do and its effect on cooperative members' livelihoods. It is even more difficult to ascertain whether the cooperatives have helped empower their members. It is likely that at least some benefits from the cooperatives are becoming available at least to the middle-income members, but limited to those who own land in an irrigation scheme's command area. Overall, the irrigation cooperatives in the studied locations in Ethiopia seem to be more or less local initiatives whose membership is restricted to households located in a specific geographical area. Although the level of inclusion in the cooperatives is naturally restricted, the cooperatives have high levels of participation and are responsive and accountable institutions. They are well connected to higher level decision makers, and serve their members' interests well. Increasingly, there are pressures on these cooperatives to add to the menu of services they provide; what is unclear, however, is whether the addition of new services will reduce the effectiveness of existing service provision even if the scope of the cooperative is narrowly defined at present. The contributions of irrigation cooperatives to improvements in livelihoods from the cultivation of cash crops also suggests that efforts to increase agricultural commercialization are likely to enjoy greater success when they are accompanied by efforts to improve access to irrigation and water management (Spielman and Kelemework 2008). Given the importance of irrigation to agricultural development in Ethiopia and the strong emphasis on the role of cooperatives, irrigation cooperatives can be an important actor among local natural resource management institutions. They are always likely to be relatively limited in those they serve, but expansion of areas where irrigation is available will also increase the inclusiveness of irrigation cooperatives. The success of cooperative members in gaining a measure of responsiveness and the willingness of government officials to remain engaged with the cooperatives likely bodes well for the future of such cooperatives in Ethiopia. The research also highlights the importance of sequencing of external versus internal support in the development of new local institutions. Government and NGO actors have played a clear role in promoting and organizing cooperatives. Their continued engagement is necessary to facilitate the effective functioning and sustainable organizational development of these new institutions. Single shot, standard packages of inputs and guidelines are likely to be neither well suited to local conditions nor effective for long-run development of new institutions. Over time, as local organizations gain maturity and strength, the disengagement of external actors and greater mobilization of local support is equally important to prevent the emergence and consolidation of dependency relations and movement toward sustainable development. The exact timing of withdrawal of external support is likely to be highly dependent on local conditions. 66 A4.2 PANI PANCHAYATS IN ORISSA Irrigation schemes in India are publicly created and managed for the most part. However, state irrigation departments in most parts of India are characterized by problems in maintaining the irrigation infrastructure and systems, limited capacity to collect water charges, staff shortages, and declining public allocation for repair and maintenance of infrastructure. At the same time, irrigation continues to be critical to self-reliance in food production. Since the 1970s, participatory irrigation management has received significant attention from government decision makers as a strategy to make irrigation management more efficient and easily accessible. The first national water policy adopted by the Indian government in 1987 emphasized the role of farmers in management of irrigation systems, especially in distribution and collection of water charges. The government of Orissa's state water policy of 1994 is in accordance with the national water policy. By 2000 the state water policy had begun to promote pani panchayats, or water users' associations, in all irrigation schemes. Today, the legal standing for these associations is provided by the Pani Panchayat Act (2002) and Pani Panchayat Rules (2003) such that they can be registered as legal bodies for the operation and maintenance of irrigation systems and receive capacity building and financial assistance from the state government. More than 15,000 such associations have been formed with the capacity to manage water delivery for approximately 1.4 million hectares of irrigated land (Sasmal 2008c). The command area of each pani panchayat comprises the area irrigated by several distribution outlets (chaks). The landholders for each distribution outlet select a representative to the water users' association's executive committee. All the farmers who have land in the command area of the association form the general body of the association. In addition to flow irrigation water users' associations, Orissa also has lift irrigation associations. The study focused on six water users' associations, four from Bhadrak and two from Kalahandi district. The formation of these associations was encouraged by the government in an effort to persuade users to take greater control over the maintenance and repair of the system and distribution of its water. Because the water users' associations are formed according to a common set of rules encoded in the Pani Panchayat Rules, the basic management and organizational structure of all associations is similar. However, there are significant variations across the studied associations in their size, members' awareness of the rules and organizational roles, internalization of organizational objectives, and commitment to service delivery. At the same time, the members of the different associations do not view the benefits of membership to be very high in relation to the tasks they are supposed to perform for maintaining the system. Nor are the different associations able to get sufficient funds they need to repair, maintain, and run the system. For most of the associations, the main source of funds is the state government; smaller amounts are taken in as membership fees. In contrast, associations formed around lift irrigation tend to rely more on water charges and membership fees than grants from the government. The water users' associations function in a way that helps the better-off farmers and those in leadership positions. Limited funds are used as they become available to repair and maintain channels that are along the land held by leaders and their relatives. Meetings of the executive committee and the general body are held infrequently at best. Participation in such meetings is low, and members have little control over the decisions made by the leadership. 67 Despite indifferent performance on governance indicators, the water users' associations have been successful in improving livelihoods benefits for a proportion of their members. On the average, members attribute a 10 percent increase in output to the association's activities, but these benefits are available only to some of the farmers. Typically, these are farmers whose land lies along the few distribution channels that have been repaired. Few respondents believe that the water users' associations have contributed much to the livelihoods of members. In the case of water users' associations dependant on flow irrigation in Orissa, elite capture of local institutions is evident. Most of the organizations are led by individuals who are also local luminaries involved in political parties. The politicization of the institutions has concomitant effects on the way it functions and delivers benefits to constituents. In contrast, associations built around distribution of water from lift irrigation schemes tend to be more responsive and accountable to their members, in part because they depend more on fees they charge members for distributing the water. In addition, the boundaries of the areas they manage are relatively small, which also contributes to higher levels of member participation and organizational accountability. A4.3 TRIBAL STRUCTURES IN WATER MANAGEMENT IN YEMEN Water resource management institutions in Yemen derive their importance from the fact that Yemen is a water-stressed country. Its annual consumption of water exceeds its renewable water resources, with more than 90 percent of water devoted to agricultural use. Groundwater resources are threatened with overexploitation, reaching up to 135 percent of recharge in places, leading to an urgent need to improve irrigation efficiency, develop a sustainable water strategy, and recharge groundwater. Threats related to desertification, soil degradation, and climate change are additional problems requiring better water management. In this context, the role of tribal leaders and elders in water management is very important. Application of traditional laws in relation to agriculture, grazing, common lands utilization, and irrigation is widespread. The role of tribal leaders is also obvious in the protection and use of vegetation. And tribal structures are critical in the mobilization of communities for flood control, protecting riverbanks, and constructing and repairing terraces. Within the selected sites, spate and masque irrigation are the two main forms of irrigation. Under spate irrigation, elementary stone or earthen bunds together with diversion canals are used to divert flood waters from riverbeds or valleys (wadis) into fields. Runoff can also be harvested from small surfaces and diverted by gravity directly onto fields using ditches (masques). In general, surface water irrigation based on harvesting of excess rainfall is the primary irrigation method in Yemen. The basic principle for allocating water is that landowners located upstream in a wadi irrigate their lands before those located downstream. Tribal structures and decision makers in the four selected locations are well connected to most of the other local institutions, but less so with higher-level or central institutional arrangements or decision makers. In contrast, many of the newer institutions that have been created tend to have fewer links with local institutions. These include, for example, the water users' associations and water users' groups for two different Bank-assisted projects to improve spate irrigation and for soil and groundwater conservation. Local residents often view tribal structures as relatively inclusive and participatory. They help represent local concerns to government offices and connect community members with new projects. They are also 68 important in mobilizing community contributions and settling local disputes. However, the relevance of tribal structures in helping allocate water or maintain irrigation structures is lower in areas close to formal administrative centers. Nor are the tribal structures very effective in addressing large-scale damages or undertaking preventive measures to reduce problems resulting from flooding or droughts. The tension between the capacities and shortcomings of traditional tribal institutions and new irrigation institutions created by external interventions are obvious in Yemen. Traditional institutions are well connected to local populations, work to represent local interests and concerns, and mediate between external interventions and local communities. But they have limited ability to change existing rules for allocating water or mobilizing large-scale support. Because existing rules systematically favor landowners in upper catchments of valleys and rivers, new community institutions such as the water users' associations and groups seek local involvement on a more equitable basis. But the new community- based institutions are not well connected to communities or households, nor do they enjoy the same legitimacy as do tribal leaders and institutional structures. One of the key puzzles in understanding local institutional governance revolves around this tension between traditional and newly introduced institutions. A4.4 FOREST MANAGEMENT INSTITUTIONS IN VIETNAM Forests are a prominent example of a multiproduct resource that generates both private and public benefits. Forests provide rural populations with fodder, firewood, and timber, and urban populations with wood for furniture. They also sequester carbon and contain high levels of biodiversity, especially in the tropics. They are especially important for the livelihoods of the poor and serve as safety nets for the poorest during periods of stressed livelihoods (Sunderlin et al. 2005). The sustainable and equitable use of forests and their benefits is crucial to many poor households in Vietnam as well. The study in Vietnam focused on forests in four provinces, with special attention to the relationship between institutional design, governance, and livelihoods outcomes. The four selected sites were drawn from the main forest types of Vietnam located in the northern uplands, the north central coast, and the central highlands. Within each location, there are a number of different types of forestry institutions. Interestingly, all the forestry institutions selected for the research were created with the help of donor organizations. The finances for the institutions are derived from external donors or from the government, but there is little generation of resources from within the communes or villages to run the organization. Although the studied forestry institutions at the local level are highly inclusive, there are important variations in the nature of participation and the extent to which decision makers in the forestry institutions are accountable to local residents. Women are excluded from membership. High levels of poverty and ethnic differences have been critical determinants of low performance levels, and in many cases benefits from forests have been allocated quite unequally among villagers. One reason for the low performance of the local forestry institutions in the communes is their low level of interaction and integration with higher-level institutions at the district level. Local forest groups may be very effective in connecting to people in the villages, but in the absence of strong connections with higher-level institutions they have little influence on longer- term sustainable forest use. In the absence of close connections both with higher- and lower-level actors, it is unclear if donor-facilitated local forestry institutions will survive after the project life is completed. Indeed, in one of the studied provinces the local forestry institutions have faded away. 69 Appendix 5 Local Governance Institutions in Afghanistan, Ethiopia, Vietnam, and Yemen A5.1 THE COMMUNITY DEVELOPMENT COUNCILS OF AFGHANISTAN The community development councils (CDCs) in Afghanistan are elected local governments created through the National Solidarity Program (NSP) in 2003. In the postconflict environment of Afghanistan, effective local governance of development projects is a prized objective. The CDCs are viewed by many as a means of reaching this objective. According to a recent news report written by Afghanistan's Minister of Rural Rehabilitation and Development, In [Afghanistan's] precarious environment, the NSP has been revolutionary. Since its inception in 2003, the NSP has reached over 15.4 million Afghans. These communities have democratically elected community development councils in 352 of Afghanistan's 364 districts. It's helped finance more than 35,000 projects. Communities are experiencing development for the first time. Projects are dictated from below, avoiding any disconnect between donors and the communities they serve. Some have opted to take advantage of nearby rivers to build small hydropower facilities. Others have built new roads that connect to markets, and district centers. More than 500,000 households have benefited from small-scale irrigation projects (Zia 2008). This strongly positive assessment of the most widespread and comprehensive local governance and development initiative in Afghanistan underlines the importance that the national government places on the success of this initiative. The CDCs are introduced in a community governance context that has been largely informal and widely variable across different provinces. They are intended to enable wider involvement of community members--including women--in development and other decisions that affect the community. However, greater inclusion of women is a challenge in all the studied cases, and where ethnic minorities are present the challenge of inclusion also includes ethnicity. Given the history of women's exclusion from decision making and public involvement in governance institutions, even imperfect application of the formal electoral and voting provisions of the NSP represents an advance. The CDCs, in addition to their role in governance, are also viewed as a means to select and implement community-level development projects. There is some tension between the governance and development functions attributed to the CDCs because of the different emphasis placed on each by different partners in the program. One of the ways in which this tension is evident concerns the allocation of funds to the CDCs for development projects. Each CDC has a funding ceiling of $60K for the block grant it can receive, with no more than $200 per household. This provision creates strong incentives for dividing communities larger than 300 households so as to maximize the available block grant per capita. There are a number of variations in the specific electoral procedures used for selecting the leadership for a CDC, as also for selecting projects. These have typically emerged in response to local and provincial level variations. The availability of resources for implementing development projects may be one of the strongest motivations for the formation of CDCs and likely accounts for the additional roles that CDCs have assumed in some cases as around conflict resolution. One of the hopes of the NSP seems to be that as the CDCs gain legitimacy and local consent for the development projects they facilitate, they will over time also gain legitimacy as local governance institutions (Nixon 2008a; Nixon 2008b, 20). 70 Although some studies of CDCs have remarked on the unity and solidarity that the CDCs have produced (Barakat 2006), inclusion through CDCs has not typically extended across gender divisions. One way in which women's inclusion has been affected in some parts of the country is through the creation of women's CDCs. The gendered nature of inclusion is also similar for representation and responsiveness. Although the CDCs are responsive to the interests of their members, that is less true for women's interests. Accountability of the CDC decision makers to community members is certainly not institutionalized at present, especially with low attention to reelection of CDC decision makers. An examination of the relationship of the CDCs with other institutions in the study sites suggests that stronger connections to such institutions, either through formal channels or through shared personnel, enhances the prospects for greater legitimacy of the CDCs. As new institutions, the CDCs are more inclusive, participatory, and accountable in comparison with the more traditional institutions with which they coexist or which they have in some cases displaced. But it is also the case that CDCs have in many instances been effective by allying with existing institutions. Thus collaboration with preexisting organizational arrangements has often contributed to the success of the CDCs in terms of improved livelihoods and better governance outcomes. Continued effectiveness of the CDCs and consequently their local legitimacy and acceptance is likely to depend in substantial measure on their being able to provide members continued access to resources. Continued access to resources with high levels of performance on governance indicators such as participation and accountability can lead to sustainable gains in governance. There is a real risk, otherwise, that the existing gains of the CDCs will be lost when external support ceases to be available after a single cycle of disbursements (Perrin , Van Der Riet, and Sjorslev 2006). A5.2 GRAM PANCHAYATS IN ORISSA Local governance in India has been traced back to the precolonial and even ancient period, but the modern experiment with local governance in the context of development is more directly traceable to panchayats under the community development program in the 1950s. The 73rd amendment to the Indian constitution recognized various shortcomings in the earlier experiments as it provided for gram sabhas (village assemblies) and gram panchayats (village councils), direct elections to the councils, and legislative measures for devolution of power and resources to the gram panchayats. In response to the 73rd amendment, state governments in India have created Panchayati Raj Institutions (PRI) at the local level but with substantial differences in the powers, resources, and capacities allowed to these institutions. In Orissa, a number of the key objectives of the constitutional amendment have been met. These include reservation for women and the SC/ST (scheduled caste and tribe) groups, elections, and devolution of powers. The gram panchayats are the central rural institution in the studied sites in Orissa with a large number of and strong connections to the other institutions. They are supposed to undertake a range of development functions. But in practice they have found it difficult to do so owing to limited financial devolution or autonomy. There are other reasons that make the gram panchayats less effective. Few rural residents, including the office holders in the gram panchayats, have adequate knowledge of the different activities and tasks that the panchayats are supposed to perform, and the services they are expected to deliver. Nor do the panchayats have much control over the local officials of the line agencies of the state government. In addition, communication between the panchayat office holders and villagers is poor, government officials exercise strong control over the projects that the panchayat selects for execution, and panchayat officials have little discretion in implementing development projects. 71 The effects of these organizational constraints are visible in the performance of the gram panchayats. Most of the studied panchayats ranked low on the four major governance criteria on which their performance was assessed: inclusion, accountability, participation, and adaptive capacities. Although women and marginal groups are formally included in the working of the panchayats, their share in the benefits allocated through the panchayats is low and their actual ability to influence decision making also remains low. Decentralized governance institutions have been ineffective in other ways. One of the results of devolution of project and implementation activities to the gram panchayat level has been competition between the officials of the line agencies and elected members of the PRI over decision making. Attempts to minimize the role of elected representatives at various levels and control over information by line agencies mean that the role of panchayat institutions in decision making has remained limited. Changing this situation will require greater authority to panchayats and panchayat institutions over line agency officials, a prospect that would be resisted by most civil servants. Gram panchayats were also viewed by respondents as operating mainly along party lines, responding to party interests rather than to the interests of their constituents, and more accountable to higher level decision makers than to villagers. At best, the loyalties of the gram panchayat elected officials is to party loyalists within the village. One reason for the politicization of local development and governance is the lack of access to discretionary funds. But another is the extent to which political parties affect election outcomes even in local elections. Further, many of the elected officials see panchayats as a potential stepping stone to higher-level political office. The contributions of the gram panchayats to local livelihoods are primarily through their role in the implementation of centrally initiated projects to provide employment and create infrastructure. These projects are usually selected without much systematic planning, and often without much input from villagers. They have certainly resulted in improved infrastructure in the studied villages. However, these improvements have occurred with little attention to sustainability of benefits, and have not been very equitably distributed. In summary, the gram panchayats seem to be institutions with much potential to shape local governance and improve development outcomes. However a combination of circumstances and constraints has worked to reduce their effectiveness. These include continuing efforts by line agencies to control information and limit involvement of elected officials, low access to discretionary funds, inability to select projects in response to local needs, and high levels of political influence over local decision making. The high levels of institutionally mandated inclusion have had some positive impacts in terms of accountability and livelihoods outcomes, particularly in situations where women decision makers are not simply proxies for their male relatives. Further administrative and fiscal devolution is essential if the gram panchayats are to fulfill their potential and meet the expectations placed on them. A5.3 DEMOCRACY COMMITTEES IN VIETNAM Decentralization of governance institutions in Vietnam is a relatively recent phenomenon, especially in comparison with economic decentralization that can be traced back to the mid-1980s. In a context where the government is in principle committed to more decentralized decision making at the local level but accepts the existence of fragmented and inconsistent management of the process, the role of new institutions to shape local governance has become increasingly important. Grassroots democracy steering committees, created under the provisions of the Grassroots Democracy Decree in 1998, are the means through which the government has attempted to invigorate democratic governance processes at the commune level. Although all communes have created grassroots steering 72 committees in the wake of the decree, their performance is highly variable (Tran 2008, 2). This variation is visible in their links with existing institutions and the degree to which they are actually able to elicit local participation, whether the primary motivation for their activities are from above or a reflection of local interests. In part because political decentralization processes in Vietnam are recent, the important local governance institutions are affiliated for the most part to government agencies. For the same reason, the formal structure of local governance and responsibilities assigned to the steering committees are common across the four selected locations where research was carried out. The steering committee is charged with coordinating the activities of the local units of the Communist Party; different mass organizations such as the Women's and Youth Unions and the Farmers', Senior Citizens', and Veterans Associations; the People's Council; and the People's Committee. Some of the more important decisions of the steering committee concern infrastructure development at the local level. The committee is supposed to coordinate the major local infrastructure decisions about what to construct, local versus government contribution share in construction costs, and how to undertake infrastructure development. However, the extent of the steering committee's influence on these decisions has been relatively low for the most part as institutions predating it have sought to maintain and exercise their own influence. The contributions of the steering committees in the study sites to local livelihoods have thus been relatively low. Their performance on different governance measures has also not been noteworthy. Overall levels of inclusion are high, as is participation, especially in terms of attendance at meetings. But accountability to constituents is either nonexistent or low, and the organizational structure and upward accountability mechanisms make decision making rigid and inflexible. Some of the larger issues raised by the study of the steering committees are similar to those related to local governance institutions in other countries. These include low capacity to manage the assigned tasks, limited financial devolution, strong influence from higher-level officials on decision making within the organization, and absence of monitoring capacity. In many ways, the structural organization of the committees works to prevent them from being effective. To the extent any committee manages to be effective, it is because of strong leadership that is able to overcome existing institutional barriers. A5.4 DISTRICT LOCAL COUNCILS IN YEMEN Effective local governance and local government structures in Yemen are viewed as being an antidote to growing poverty and inequality, a means to bridge disjunctures between traditional and modern institutional structures, and necessary to promote greater inclusion of those marginalized from the development process. Attempts to create formal local governance structures in Yemen can be traced back to the local councils for cooperative development (LCCD) of the early 1990s that sought to improve the functioning of cooperatives. Following the demise of these institutions when external support became unavailable, the government attempted to replace them with local development associations (LDAs). The Local Authority Law of 2000 laid the foundations for a new set of local governance structures in the form of the district local councils (DLCs). These institutions are viewed as responsible for managing budgets for local development and construction projects, monitoring and supervising the use of resources and provision of services, implementing various national policies at the local level and providing feedback to improve them, and coordinating the implementation of development programs together with other government agencies. 73 In the study sites, a number of changes have resulted in local development­related decision making as a result of the establishment of the DLCs. More decisions are now made at the district rather than at the central level. The annual budget of local governments has increased significantly, ranging from 15 percent to nearly 90 percent. The DLCs are extensively involved in infrastructure development, making decisions about the kind of projects and services that should be provided to local communities, and project implementation. However, these improvements in the functioning of the DLCs have generated only limited benefits to poorer households. There were few variations across the studied DLCs in terms of their structure or organization. Additionally, DLCs in the study sites are well connected to line agencies but their links with local community associations are however relatively weak. As a result, their performance on governance indicators of inclusion, participation, and accountability can be improved significantly. The DLCs have especially been remiss in the inclusion of disadvantaged groups including women, have secured only low levels of participation from their constituents, and are accountable primarily to higher-level authorities. Interviewed households expressed a number of reservations about the DLCs. Their complaints centered around the inability of DLCs to provide social services and needed livelihoods opportunities; lack of transparency, information, or accountability; and absence of links to decision makers within the DLCs. 74 Appendix 6. Data for Assessing the Overall Relationship Among Selected Factors (Based on Institutional Case Studies for Study Countries) Country Institution Policy Long-Term Name Location Name Domain Context Poverty Context Support Afghanistan Bamiyan AG Inhospitable Negative Absent Afghanistan Balkh AG Inhospitable Negative Absent Afghanistan Heart AG Inhospitable Negative Absent Afghanistan Nangerhar AG Inhospitable Negative Absent Ethiopia Shebedino AG Medium Positive Absent Ethiopia Goro Gutu AG Medium Negative Absent Ethiopia Oromia AG Medium Positive Absent Ethiopia Tigray AG Medium Negative Absent Orissa Ambroli GOV Medium Positive Absent Orissa Sabrang GOV Medium Positive Medium Orissa Lanjigarh GOV Medium Negative Absent Orissa Gunduri GOV Medium Negative Medium Orissa Maa Mangla NRM Medium Positive Absent Orissa Rameswarjew NRM Medium Positive Absent Orissa Maa Kandeswari NRM Medium Positive Absent Orissa Baba Sapneswar NRM Inhospitable Positive Absent Orissa Maa Nilakanteswari NRM Medium Negative Absent Orissa Ma Brahmani NRM Medium Negative Absent Orissa Mother Teresa AG Enabling Positive Absent Orissa Gopal Jew AG Enabling Positive Absent Orissa Maa Mangla AG Enabling Positive Absent Orissa Maa Durgati AG Enabling Positive Absent Orissa Maa Santoshi AG Enabling Positive Absent Orissa Maa Saraswati AG Enabling Positive Absent Orissa NilKanteswari AG Enabling Negative Absent Orissa Routpada AG Enabling Negative Absent Orissa NilKanteswari AG Enabling Negative Absent Orissa Chandi AG Enabling Negative Absent Orissa Bharavi AG Enabling Negative Absent Orissa Mouli AG Enabling Negative Absent Vietnam Chieng Kheo AG Enabling Positive Absent Vietnam Nam Thinh AG Medium Negative Absent Vietnam Chu a Thai AG Medium Negative Absent Vietnam Phuong Dong AG Enabling Positive Absent Vietnam Buot, Son La NRM Medium Positive Medium 75 Country Institution Policy Long-Term Name Location Name Domain Context Poverty Context Support Vietnam Nam Thinh forest NRM Medium Negative Absent Vietnam Chu A Thai Forest NRM Medium Negative Absent Vietnam Dong Hai Forest NRM Medium Positive Absent Vietnam Chieng Kheo GOV Enabling Positive Absent Vietnam Nam Thinh GOV Enabling Negative Absent Vietnam Truong Long GOV Enabling Negative Absent Vietnam Ia Ake GOV Enabling Positive Absent Yemen Tuban GOV Medium Negative Absent Yemen Al Mawheet GOV Medium Negative Absent Yemen Abs GOV Medium Negative Absent Yemen Bani Matar GOV Medium Negative Absent Yemen Al Shatri AG Enabling Negative Absent Yemen Aljar AG Enabling Negative Absent Yemen Bani Shehab AG Enabling Negative Absent Yemen Al Argoob AG Enabling Negative Absent Yemen Tuban NRM Inhospitable Negative Absent Yemen Al Mawheet NRM Inhospitable Negative Absent Yemen Abs NRM Inhospitable Negative Absent Yemen Bani Matar NRM Inhospitable Negative Absent 76 Country Name Location Name Inclusion Participation Accountability Adaptability Afghanistan Bamiyan Low Low Low Low Afghanistan Balkh Low Low Low Low Afghanistan Heart Low Low Low Low Afghanistan Nangerhar Low Low Low Low Ethiopia Shebedino High Low Low Low Ethiopia Goro Gutu Medium Low High Low Ethiopia Oromia Medium Low High Low Ethiopia Tigray Medium Low Medium Low Orissa Ambroli High High Medium Medium Orissa Sabrang High Medium Medium Medium Orissa Lanjigarh Low Low Low Low Orissa Gunduri Low Medium Medium Low Orissa Maa Mangla Low Medium Low Low Orissa Rameswarjew Low Low Low Low Orissa Maa Kandeswari Low Medium High Low Orissa Baba Sapneswar Medium High High Low Orissa Maa Nilakanteswari Low Low Low Low Orissa Ma Brahmani Low Low Low Low Orissa Mother Teresa Low Medium High Medium Orissa Gopal Jew Low Medium High Medium Orissa Maa Mangla Low Medium High Medium Orissa Maa Durgati Low High Medium Low Orissa Maa Santoshi Medium Low High Medium Orissa Maa Saraswati Low High Medium Low Orissa NilKanteswari Low Low Medium Medium Orissa Routpada Low Low Low Low Orissa NilKanteswari Low Low Low Low Orissa Chandi Low Medium High Medium Orissa Bharavi Low Medium Medium Low Orissa Mouli Low Low Medium Low Vietnam Chieng Kheo High Medium Low Low Vietnam Nam Thinh Low High High Medium Vietnam Chu a Thai Medium Medium High Medium Vietnam Phuong Dong Medium Medium High High Vietnam Buot, Son La Medium High High High Vietnam Nam thing forest High Low Low Low Vietnam Chu A Thai Forest High Medium Low Medium Vietnam Dong Hai Forest Low High High Low Vietnam Chieng Kheo Medium High Low Low Vietnam Nam Thinh Medium Low Low Low 77 Country Name Location Name Inclusion Participation Accountability Adaptability Vietnam Truong Long High Low Low Low Vietnam Ia Ake Low High Medium Low Yemen Tuban Medium High High Medium Yemen Al Mawheet Low High High Low Yemen Abs Low High Medium Low Yemen Bani Matar Low High Medium Low Yemen Al Shatri Medium Medium High Low Yemen Aljar Low Low High Medium Yemen Bani Shehab Low Low Low Low Yemen Al Argoob Low Low Low Low Yemen Tuban High Medium Low Medium Yemen Al Mawheet High Medium Medium High Yemen Abs High Medium Low High Yemen Bani Matar High Medium Low High 78 Local Country Institutional Livelihoods Equity Name Location Name Linkages Improvement Outcomes Afghanistan Bamiyan Low Medium Low Afghanistan Balkh Low Medium Low Afghanistan Heart Low Medium Low Afghanistan Nangerhar Low Medium Low Ethiopia Shebedino Medium High Low Ethiopia Goro Gutu Medium High Low Ethiopia Oromia Low High Low Ethiopia Tigray Low Medium Low Orissa Ambroli High Medium Medium Orissa Sabrang High Medium Medium Orissa Lanjigarh Medium Low Low Orissa Gunduri High Low Medium Orissa Maa Mangla Low Low Low Orissa Rameswarjew Low Low Low Orissa Maa Kandeswari Medium Low Medium Orissa Baba Sapneswar Medium Medium Medium Orissa Maa Nilakanteswari Low Low Low Orissa Ma Brahmani Low Low Low Orissa Mother Teresa High Medium High Orissa Gopal Jew Medium Medium High Orissa Maa Mangla High Medium High Orissa Maa Durgati Medium Low High Orissa Maa Santoshi Medium High Medium Orissa Maa Saraswati Medium Low Medium Orissa NilKanteswari Low Medium High Orissa Routpada Medium Low Low Orissa NilKanteswari Low Low Low Orissa Chandi Medium Medium High Orissa Bharavi Low Low Medium Orissa Mouli Low Low Medium Vietnam Chieng Kheo Medium Low Low Vietnam Nam Thinh Low Medium High Vietnam Chu a Thai Medium High Medium Vietnam Phuong Dong Medium Medium Medium Vietnam Buot, Son La Medium High High Vietnam Nam Thinh forest Low Low Low Vietnam Chu A Thai Forest Medium High High Vietnam Dong Hai Forest Medium Low Low Vietnam Chieng Kheo Medium High Low 79 Local Country Institutional Livelihoods Equity Name Location Name Linkages Improvement Outcomes Vietnam Nam Thinh Low Medium Medium Vietnam Truong Long Medium Medium Medium Vietnam Ia Ake Medium Medium Low Yemen Tuban Medium Medium Medium Yemen Al Mawheet Low Low Medium Yemen Abs Low Low Medium Yemen Bani Matar Low Low Medium Yemen Al Shatri Medium Medium High Yemen Aljar Medium High High Yemen Bani Shehab Medium Medium High Yemen Al Argoob Medium Low Medium Yemen Tuban High Low Medium Yemen Al Mawheet Medium Low Low Yemen Abs Medium Low Low Yemen Bani Matar Low Low Low 80