Health, Nutrition, and Population Global Practice and Social, Urban, Rural, and Resilience Global Practice Qualitative Study of “Sunaula Hazar Din” Community Action for Nutrition Project Nepal Kaori Oshima, Monica Biradavolu, Chhitij Bashyal, and Manav Bhattarai JUNE 2017 Health, Nutrition, and Population Global Practice and Social, Urban, Rural, and Resilience Global Practice Qualitative Study of “Sunaula Hazar Din” Community Action for Nutrition Project Nepal Kaori Oshima, Monica Biradavolu, Chhitij Bashyal, and Manav Bhattarai June 2017 © 2017 International Bank for Reconstruction and Development / The World Bank 1818 H Street NW Washington DC 20433 Telephone: 202-473-1000 Internet: www.worldbank.org This work is a product of the staff of The World Bank with external contributions. The findings, interpretations, and conclusions expressed in this work do not necessarily reflect the views of The World Bank, its Board of Executive Directors, or the governments they represent. The World Bank does not guarantee the accuracy of the data included in this work. The boundaries, colors, denominations, and other information shown on any map in this work do not imply any judgment on the part of The World Bank concerning the legal status of any territory or the endorsement or acceptance of such boundaries. Rights and Permissions The material in this work is subject to copyright. Because The World Bank encourages dissemination of its knowledge, this work may be reproduced, in whole or in part, for noncommercial purposes as long as full attribution to this work is given. Any queries on rights and licenses, including subsidiary rights, should be addressed to World Bank Publications, The World Bank Group, 1818 H Street NW, Washington, DC 20433, USA; fax: 202-522-2625; e-mail: pubrights@worldbank.org. Contents Acknowledgments v Executive Summary vii Chapter One: Introduction 1 Background and Rationale 1 Study Objective 2 Chapter Two: Sunaula Hazar Din Project Design and Implementation Challenges 3 The Sunaula Hazar Din Project 3 Community-Driven Rapid Results Approach 3 Steps of the Rapid Results for Nutrition Initiative Cycle 4 Project Implementation Challenges 5 Chapter Three: Study Methodology 7 Site Selection 7 Study Population 9 Study Instruments 10 Data Collection Procedure and Analysis 11 Chapter Four: Key Findings 13 A. On the Sunaula Hazar Din Rapid Results Nutrition Initiative (RRNI) Cycle 13 a. RRNI Team Selection and Representativeness 13 b.  Selection of Focus Area 19 c.  Selection of Goal 20 B. Inclusion of Women 22 C. Inclusion of Minorities 24 D. Leadership 25 a. Coaches 25 b. Rapid Results for Nutrition Initiative Team Leaders 28 E. Local Government 29 F. Challenges and Opportunities in Implementation by Focus Area 32 Chapter Five: Emerging Lessons and Conclusion 37 References 43 Annex 1: Theory of Change of the SHD Project 45 Tables Table 1: S  tudy Districts, Wards, and Demographic Information 9 Table 2:  Focus Areas and Selected Goals in Our Sample 20 Contents iii Figures Figure 1:  Key Steps in the 100-Day Rapid Results Nutrition Initiative (RRNI) Cycle 4 Figure 2: The Sample Districts of the Qualitative Study 8 Boxes Box 1: Pathway 1—RRNI Team Selection Process Community-Led & RRNI Team Representative of Ward Population 15 Box 2A: Pathway 2—Elite Capture 16 Box 2B: Pathway 2—Elite Capture 17 Box 3: Pathway 3—Benign Capture 17 Box 4: Pathway 4—Population Density Issues 19 iv Qualitative Study of “Sunaula Hazar Din” Community Action for Nutrition Project Nepal Acknowledgments This qualitative study reviews the Sunaula Hazar Din (SHD), Community Action for Nutrition Project (2012–2017) of the Government of Nepal, which has been sup- ported by the World Bank. The study was jointly undertaken by the World Bank’s Health, Nutrition and Population Global Practice (GHNDR) and the Social, Urban, Rural, and Resilience Global Practice (GSURR). The study team was led by Kaori Oshima and Manav Bhattarai, and main authors included Monica Biradavolu and Chhitij Bashyal. E. Gail Richardson (Practice Manager, GHNDR South Asia); Rekha Menon (Practice Manager, GHNDR South Asia); Senait Nigiru Assefa (Practice Manager, GSURR Global Programs Unit); and Takuya Kamata (Country Manager, Nepal) provided overall guidance to the study team. Sean Bradley (Lead Social Development Specialist and head of the CDD Community of Practice) also provided valuable advice to the study team throughout the process. Jean Voss edited the report, and Jaya Karki, Ade- laide Barra, and Ajay Ram Dass provided timely and helpful administrative support. The study was financially supported by the South Asia Food and Nutrition Security Initiative (SAFANSI), and the team benefitted from continued support provided by Pauline Zwaans and Jamie Greenawalt. The study team met and held invaluable discussions with the members of the SHD Project Management Team of the Ministry of Federal Affairs and Local Develop- ment, Government of Nepal. The Oxford Policy Management Nepal team and their field researchers provided excellent data collection work, led by Suresh Tiwari, Hema Bhatt, and Dhruba Ghimire. The study team is grateful to the following list of colleagues who shared their insights as peer reviewers for the concept note and final draft, including Robert Wrobel (for both concept and final stages), Junko Onishi, Menno Mulder-Sibanda, Luiza Nora, and Ali Subandoro. Additional insights were provided by a number of World Bank practitioners working on social and rural development, or nutrition and health, in the process of study preparation and finalization: Mio Takada, Susumu Yoshida, Tekabe Ayalew Belay, Gogi Grewal, Michell Dong, Daniel Stein, Albertus Voetberg, Soham Sen, and Ritika Dsouza. Finally, the study team expresses deep gratitude to relevant agencies in Nepal, includ- ing the Ministry of Federal Affairs and Local Development and the Ministry of Health, as well as the District, Village, and Ward-level officials and stakeholders involved in the qualitative interviews and focus group discussions. Acknowledgments v Executive Summary Malnutrition remains a serious development challenge in Nepal. The country still has an alarmingly high rate of child malnutrition with 35.8% of children under five being stunted, 9.7% being wasted and 27.0% being underweight (Nepal Demographic and Health Survey. Key Indicators. 2016).1 Reduction in malnutrition is important for curbing disease and mortality and ensuring proper cognitive development to enhance economic productivity. The increasing body of evidence points to the importance of intervening within the “golden 1,000 days,” a period between conception and the age of two that is considered a “critical window of opportunity” to tackle the key risk factors of malnutrition that hamper a child’s long-term physical and cognitive devel- opment. A range of individual-level factors (such as chronic energy deficiency, poor hygienic practices, early marriage/pregnancy, poor infant and child feeding practices, etc.) and community-level factors (availability of nutritious food, poor access to safe drinking water, poor sanitation and hygiene, etc.) are associated with an increased risk for malnutrition. Launched in 2012, the Sunaula Hazar Din (SHD)—Community Action for Nutrition Project is the first project in Nepal that employs an innovative community-driven rapid results approach (RRA) to tackle both individual- and community-level factors and improve knowledge, attitudes and practices for improving nutritional outcomes. The project’s social mobilization and subsequent participatory subproject implementation cycle is called “Rapid Results for Nutrition Initiative (RRNI),” where each ward in a selected Village Development Committee (VDC) will commit to a series of nutrition relevant goals, each to be met within a 100-day period. A ‘coach’ (local facilitator) is assigned to each VDC to facilitate community-level teams created under the project (called “RRNI teams”) in achieving their respective goals. Given that the SHD is one of the first initiatives in Nepal and in the world to imple- ment the RRA to address malnutrition on a large scale, i.e., across 15 districts of the country, the implementation team and other stakeholders recognized the impor- tance of conducting a rigorous evaluation of the SHD. In this context, this in-depth 1 Nepal Demographic and Health Survey 2016, Key Indicators Report. USAID, New ERA and Ministry of Health. April 2017. Executive Summary vii qualitative study complements an ongoing quantitative, setting for the 100-day project; (3) implementation of randomized controlled trial impact evaluation, to better 100-day project including challenges faced and the solu- understand the process and local dynamics of the proj­ tions devised; (4) outcomes and sustainability of changes; ect implementation, especially at the community level, and (5) lessons learned. The study was conducted in six with particular focus on gender and social inclusion per- out of the fifteen SHD project intervention districts, and spectives (e.g., inclusion of low caste and other minority three different RRNI subprojects from each of the six groups). While promising, the project’s RRNI cycle and districts were selected jointly with the Government. A steps have been complex and challenging, offering valu- non-­ probabilistic, purposive sampling was done to enable able insights on how community-based nutrition-focused comparison of projects across various variables: (a) eco- programs can be implemented more effectively. logical topography (Hills and Terai), (b) three different project clusters (15 projects were divided into 3 clusters of In fact, the project faced multiple implementation con- 5 districts, and each was implemented with support from straints causing significant delays in the program’s prep- a different NGO), (c) five different most common focus aration and rollout. An earlier review found multiple areas, and (d) both moderate and high performance proj­ reasons, including: mismatch between the RRA model ects. All qualitative data were transcribed in English and and slow implementation mechanism of the government coded for rigorous analysis. structure; lack of common understanding and commu- nication among implementation actors in central, district While good practices and outcomes were observed in and local levels; low implementation capacity; lack of some wards, the data overall revealed multiple challenges incentives for busy local government officials to actively that the SHD participating wards faced. It provides a good participate in the project; and impact of the earthquake base to understand whether the 100-day rapid results in 2015 and the economic blockade at the India-Nepal approach was effective in the given context, and whether border. the community-driven model was adequately supported by the project to achieve its development objective. The objective of the study is, therefore, to promote a bet- »» Selection of community-based team members. When the ter understanding of implementation and outcomes of initial ward citizen’s forum failed to gather many the approach, considering such constraints and their pos- participants or provide sufficient orientation on the sible influence in community-driven projects, thereby pro- project in an inclusive and transparent manner, a viding design and implementation lessons and evidence to small group of individuals dominated the decision- practitioners and policy makers of the SHD and similar making step, and the subsequent procedures were development programs. Based on key findings, the study also controlled only by them. When such a capture also seeks to identify emerging lessons for other stakehold- happened at the beginning of the cycle, it was hard ers working on addressing malnutrition in Nepal to con- to reverse, and other community members did not sider, in order to stimulate future dialogue. have a sense of ownership or being included until the end. For the community’s decision-­ making pro- The qualitative study employed three methods for data cess to be effective, it is critical to have a clear set of collection: (a) key informant interviews (KIIs) with district rules and requirements, and to widely and clearly officials, VDC officials, Ward Citizens Forum chairper- communicate them up front. For example, the sons, coaches, and RRNI team members; (b) focus group minimum number of participants, or clear voting discussions (FGDs) with beneficiaries; and (c) observa- procedures that must be followed—and the deci- tions of project environment or outcomes. Both KIIs and sion would be considered invalid unless those con- FGDs were conducted using a common research guide ditions were met—would be useful. In some cases, framework, consisting of five sections corresponding to although not fully “ ­ community-driven,” resource specific stages of the RRNI project cycle: (1) RRNI team allocation and results of the RRNI project were selection and internal team dynamics (including social positively received by the community when a inclusion aspects); (2) selection of focus area and goal small group of members developed a clear work viii Qualitative Study of “Sunaula Hazar Din” Community Action for Nutrition Project Nepal plan and budget, and implemented the subproject used, developing a clear work plan and managing accordingly. the budget and the community’s expectations are »» Forming the RRNI team. Requiring nine members to both critical in successful project performance. form the project-specific RRNI team which would »» Inclusion of women was often done only in form, not be dissolved after each 100-day cycle was a seri- in function, since requirements to have women in ous challenge for the ward members, especially teams met with challenges such as cultural norms, in areas with a spread out population where ward time constraint with household chores, and threats members were already overstretched due to other felt and resistance by men. Even when women such programs and their daily work. In many sam- wished to actively participate, men dominated or ple wards, members were “selected” without even controlled the decision-making and implementa- being informed (hence without agreement), result- tion processes. It may be worth considering, when ing in confusion and inefficiency. This indicates forming or working with an existing women-only that in the designing phase, the SHD project did small group or separate meetings, to identify bot- not sufficiently assess the feasibility and effective- tlenecks and ways to resolve them, while strength- ness of the RRNI team model in the target areas. ening awareness raising and support from a wider A mapping exercise of existing local platforms and community (both men and women) including community groups, as well as consultations with through other existing platforms or groups. target populations, would be helpful to confirm the »» Inclusion of minorities was often hampered by lim- design efficiency, and when implementing a similar ited demand for participation among the minori- activity, to identify stronger synergies and partner- ties, often because of lack of understanding on ships to be built around the RRNI team. why their participation was important, and inad- »» Selection and implementation of community-driven subproj- equate guidelines and mechanisms from projects ects (focus area and goals) were affected by a range of to increase inclusion. For sustainability, it would be factors—such as budget, cultural norms, geograph- useful to more actively involve different levels of ical constraints, expected visibility of outputs in the local authority in related data gathering and 100 days, and targeted or broader benefit sharing. awareness raising. In collaboration, the project can A more customized facilitation for the geographi- provide incentives, such as rewarding a commu- cally and culturally diverse communities and provi- nity that was most inclusive in each VDC. If there sion of technical knowledge, by building capability are traditionally excluded groups, or if the group of facilitators or encouraging greater collaboration members are not used to speaking up at a general with local resource agencies and people, would meeting, separate meetings can be organized to help to strengthen this process. ensure their voices are heard. »» Targeting or broader benefit sharing. Targeting and »» Strong leadership and effective facilitation was critical for identifying beneficiaries was a contentious pro- proper project orientation and communication to cess, often fraught with confusion and skepticism the community, expectation management, conflict from stakeholders and nonselected groups, primar- resolution, transparent communication regarding ily due to inadequate provision and adherence to the definition of target beneficiaries, and organiz- inclusion and exclusion criteria for selection, men ing locally geared awareness raising activities such controlling the decision-making process, and low as road dramas. This requires greater investment transparency on how beneficiaries were selected. in building skills and capacity of facilitators and When disputes arose, teams resorted to universal team leaders. The 100-day timeframe of the RRNI targeting rather than selectively targeting the key model could not keep pace with and invest in the beneficiaries—the golden 1,000-day women. The need for customized and continued capacity build- data from this study cannot conclude whether the ing. Rewarding good performers would also be an targeted or universal approach is better. The data effective way to keep motivating these individu- is conclusive, however, that regardless of strategy als, and to positively stimulate others to improve. Executive Summary ix The coach was spending more time for the project the tools provided by the project), people were activities (he was a resident of the VDC), and was motivated and able to start new practices, and also able to effectively involve the local authority. the changes were considered as sustainable. As a Similarly, the implementation arrangement needs few good examples indicate, it is worth develop- to be built on diverse geographical factors includ- ing more entertaining ways to raise awareness by ing the number of households in a ward, because involving community members, combined with those factors are also notably affecting the project efforts to prepare a set of experts’ guidance and outcomes. good practices on key focus areas that are ready for »» The local government was given a critical role in sharing, especially with an emphasis on behavioral proposal approval and disbursement, the process changes. that was the primary bottleneck, causing delays in projects and subduing the initial momentum of The key findings of the in-depth qualitative analysis offer the teams. While communities seemed frustrated important implications for the community-based rapid with local agencies’ delays and low engagement results approach to address malnutrition. The study find- during implementation, the latter pointed to lim- ings suggest a need for practitioners and policy makers to ited adherence of teams to government financial more closely assess different players and modalities oper- policies, their own existing workload and other ating in Nepal, and to consolidate and/or partner with systemic challenges for their inabilities. Securing these programs by providing roles, incentives, and train- local government’s buy-in and active participation ing that are aligned with the country’s governance system. from the onset of the project by providing exten- Addressing malnutrition in a sustainable manner needs sive orientation and training, and assigning clear the entire community—men, women, minority groups, roles to play with associated incentives based on and local leaders and authorities—to raise awareness on the existing systems, is critical to build trust by the importance and concrete ways of improving knowl- communities. edge and behavior, and on meaningfully including women »» Focus area specific lessons. The data revealed both in decision-making and implementation processes. opportunities and challenges faced by RRNI teams that were specific to the focus area they selected. Although SHD faced multiple implementation chal- There are important lessons to learn from the lenges, a point to note about its design is that there was an examples for future implementation efforts, par- effort to have the demand-driven steps be embedded in ticularly to improve intended nutritional behaviors the local governance structures. Considering the country’s or outcomes by involving communities. A point shift toward federalism and more authority to be given that was confirmed across the sample wards was to the newly formed local bodies (at the level of a cluster that when the tangible outputs were combined of VDCs), lessons from SHD may provide some food for with related training and proper awareness raising thought to any new nutrition programs designed either by on their behaviors (e.g., demonstrations on how to government or development partners with a community- use a constructed latrine, or to wash hands using led approach. x Qualitative Study of “Sunaula Hazar Din” Community Action for Nutrition Project Nepal Chapter one Introduction Background and Rationale Nepal has an alarmingly high rate of child malnutrition with 35.8% of children under five being stunted, 9.7% being wasted and 27.0% being underweight (Nepal Demo- graphic and Health Survey, Key Indicators Report, 2016).2 Although there is a declining trend compared to the rates 5 years ago, these rates are still exceedingly high. For example, the percentage of stunting children under five is the third highest among 11 countries in the South-East Asia region, estimated in 2015,3 after Timor-Leste and India, and followed by Indonesia, Bangladesh, and Myanmar. The economic costs of malnutri- tion are very high—an estimated 2–3% of GDP (US$250 to 375 million) is lost every year in Nepal due to vitamin and mineral deficiencies alone. Improving nutrition con- tributes to productivity, economic development, and poverty reduction by reducing disease and mortality, improving physical work capacity, cognitive development, and school performance. However, the window of opportunity for improving nutrition is small—it is called the “First 1,000 days,” from the first day of pregnancy through the first two years of life. The damage to physical growth, brain development, and human capital formation that occurs during this period due to inadequate nutrition is extensive and largely irreversible. The main focus in Nepal on improving nutrition, therefore, is to accelerate the reduction of child malnutrition, and interventions must focus on the risk factors that influence nutritional outcomes during this critical period. To reduce child malnutrition, policies and programs have focused on addressing indi- vidual as well as community-level behaviors or characteristics that are known to be associated with higher risks for malnutrition. Such individual-level factors include, for example, early marriage and pregnancy, chronic energy deficiency and anemia, exces- sive physical workloads, indoor air pollution and smoking, poor hygienic practices, and related frequent episodes of illnesses such as diarrhea. Infant and young child 2 Nepal Demographic and Health Survey 2016, Key Indicators Report. USAID, New ERA and Ministry of Health. April 2017. 3 World Health Organization 2016. Strategic Action Plan to reduce the double burden of malnutrition in the South- East Asia region: 2016–2025. The data come from the countries: Bangladesh, Bhutan, Democratic People’s Republic of Korea, India, Indonesia, Maldives, Myanmar, Nepal, Sri Lanka, Thailand, and Timor-Leste. Introduction 1 care practices, such as low levels of early initiation of ment the quantitative evaluation, it was agreed that an breastfeeding, delays in health care seeking behavior, and in-depth qualitative study would be conducted to better low utilization levels of zinc with oral rehydration solu- understand the process and local dynamics of the imple- tion (ORS) to treat diarrhea can also lead to malnutrition. mentation of the innovative community-driven Rapid Community-wide features, such as the limited availability Results Approach. In zooming into activities and functions of an appropriate variety of food products, poor access to of community-based teams and stakeholders, the quali- safe drinking water, poor hygiene and sanitation practices, tative study pays special attention to gender and social and low attainment of education among girls, are often inclusion perspectives (e.g., low caste and other minority perpetuated by cultural beliefs and practices. groups), as well as to five most commonly selected focus areas on nutrition. Launched in 2012, the Sunaula Hazar Din (SHD)—­Community Action for Nutrition Project is the first project in Nepal and in the region that seeks to improve attitudes and practices Study Objective known to improve nutritional outcomes of women and The objective of the study is to promote a better under- young children through an innovative community-driven standing of implementation and outcomes of the rapid results approach. It seeks to specifically address community-driven rapid results approach under the ­ challenges at both individual and community levels, using SHD project. By examining the implementation steps a “rapid results approach” (RRA) where communities are and social and political dynamics at the community level motivated to achieve a self-selected goal in 100-day cycles, in detail, it seeks to provide design and implementation with support from facilitators, called “coaches.” Commu- lessons and evidence to practitioners and policy makers nities under the SHD choose the goals related to improved of the SHD and similar development programs. Given nutrition that best suit their particular setting, formulate that the SHD project closed in June 2017, based on key budgets and detailed work plans to achieve these goals, findings, the study also seeks to identify emerging lessons and are provided grants to implement these activities. It for other stakeholders working on addressing malnutri- has the potential to catalyze behavioral changes that can tion in Nepal to consider, in order to stimulate future last and, therefore, to strengthen nutrition security and dialogue. The study therefore does not intend to “evalu- actions in Nepal. ate” the project’s implementation or final outcomes—it While promising, the SHD project implementation pro- instead seeks to unpack perceptions and experiences of cess has been complex and challenging, offering valuable the SHD’s community-level stakeholders, and to help bet- insights on how community-based nutrition-focused pro- ter understand why and how any observed changes iden- grams can be implemented more effectively. Given that tified by the quantitative evaluation may have occurred. the SHD is one of the first initiatives in Nepal and in the By providing in-depth understanding of the SHD’s rapid world to implement the RRA to address malnutrition on a results approach, the study will also complement another large scale, i.e., across 15 districts of the country, from the ongoing study focusing on the project-level operational onset, the implementation team and other stakeholders and governance modalities used in various nutrition pro- have recognized the importance of conducting a rigor- grams by other development partners in Nepal.5 ous evaluation and relevant studies of SHD’s experiences and achievements. This important study is a quantitative impact evaluation employing randomized controlled trials to measure the impact in nutrition outcomes.4 To comple- 4 It is being conducted by the Development Impact Evaluation (DIME) team of 5 Forthcoming. Manav Bhattarai, Chhitij Bashyal and Abeyah A. Al-Omair. the World Bank, and the results are expected to be available by December 2017. “Operational Strategy to Address Malnutrition in Nepal.” World Bank. 2 Qualitative Study of “Sunaula Hazar Din” Community Action for Nutrition Project Nepal Chapter two Sunaula Hazar Din Project Design and Implementation Challenges The Sunaula Hazar Din Project The Community Action for Nutrition Initiative Project—locally known as “Sunaula Hazar Din (SHD)”—was a five-year program launched by the Government of Nepal in 2012. It was implemented by the Ministry of Federal Affairs and Local Develop- ment (MOFALD) across 15 districts (out of 75 districts) which were selected based on population, stunting levels, poverty levels and absence of overlapping interventions by other partners. Within the selected districts, SHD targeted 25% of the most disadvan- taged Village Development Committees (VDCs), in total approximately 1,100 VDCs. The project sought to improve practices that contribute to reduced undernutrition of women of reproductive age and children under the age of two. “Sunaula Hazar Din,” or in English “Golden first 1,000 days,” refers to the period between conception and 24 months of age, when children are most vulnerable to malnutrition. By improving knowledge, attitude and practices among 1,000-days women, adolescent women, and children, the project in the long run would address the key risk factors for child malnu- trition and create demand for nutrition related services and products. The theory of change of the SHD project is described in Annex 1. Community-Driven Rapid Results Approach To improve knowledge, attitudes and practices critical for nutritional outcomes, the SHD project employed an innovative community-driven rapid results approach (RRA).6 “Community-driven,” because in this approach, participating communities plan, select, and implement a nutrition-related goal, with support provided from the 6 See “Improving Nutrition in Communities through behavior change: the Sunaula Hazar Din Program in Nepal”: http://www-wds.worldbank.org/external/default/WDSContentServer/WDSP/IB/2015/01/07/000469252_ 20150107164957/Rendered/PDF/935310BRI0Box300CBBC010.14.140FINAL.pdf Sunaula Hazar Din Project Design and Implementation Challenges 3 facilitators hired by the project. The premise is that when motivated and adequately supported, communities are Steps of the Rapid able to identify their needs and resources (for this project, Results for Nutrition on nutrition), make decisions on specific goals to address Initiative Cycle those needs and formulate plans to help achieve the Each 100-day “RRNI” cycle follows six key steps as goals. Communities are supported through small grants described below and in Figure 1: (maximum US$ 3,000) to implement their self-selected 1.  Selection of ‘focus area’ and RRNI team activities. Furthermore, the SHD is unique because the members: The first step involves a coach con- community-driven process from planning to implementa- ducting a workshop with a Ward Citizen Forum tion would be done in 100 days. This is why it is called (WCF). The WCF is a ward-level body with 20–25 “Rapid Results Approach”—which is a results-focused members representing various socio-economic, learning process aimed at jump-starting major change ethnic, religious and age groups of the ward. The efforts and enhancing implementation capacity. It tack- WCF identifies needs of the ward in nutrition les large-scale medium- and long-term change efforts improvement and selects an appropriate “Focus through a series of small-scale, results producing and Area” to address that need. There are 15 Focus momentum building initiatives.7 The approach creates Areas pre-identified by the SHD project, and the motivation and confidence by defining goals and monitor- WCF members select one focus area for the par- ing results in short periods and mobilizing communities ticular 100-day cycle. The WCF, with guidance to act and coordinate. While this particular approach was from the coach, also selects nine people from the new to Nepal, there is a strong track record in Nepal of ward to form the “RRNI team.” community-driven development.8 2. Selection of RRNI team leader and 100- day project goal: The second step is a meet- The project’s social mobilization and subsequent partici- ing among the RRNI team members to develop a patory subproject implementation cycle is called “Rapid goal (i.e., activity) within the selected Focus Area, Results for Nutrition Initiative (RRNI),” where each ward in a selected VDC commits to a series of nutrition relevant goals, each to be met within a 100-day period. A ‘coach’ Figure 1. Key steps in the 100-day (local facilitator) is assigned to each VDC (therefore, wards Rapid Results Nutrition Initiative in one VDC will have the same person as their “coach”) to (RRNI) Cycle facilitate community-level teams created under the proj­ ect (called “RRNI teams”) in achieving their respective 1) Selection of focus area & RRNI goals. The role of the coaches is not to provide expert team members at opinions or participate in making decisions for the com- Ward Citizen Forum munity, but to help RRNI teams and other community members complete the entire RRNI cycle and achieve the intended goal. 2) Selection of RRNI 6) Review of results team leader & 100- and sustainability day subproject goal RRNI Project Cycle 7 Sylvester, Obong’o (2012) “Rapid Results Approach/Initiative. Institutional- ization of Results Based Management in Kenya Public Service” Mimeo. Com- 3) Preparation and monwealth Association for Public Administration and Management. 5) 100-day submission of 8 For instance, Manandhar et al. (2004) found, in “Effect of a participatory inter- implementation proposal vention with women’s groups on birth outcomes in Nepal: cluster-randomised controlled trial,” that birth outcomes in a poor rural population improved greatly through a low cost, potentially sustainable and scalable, participatory 4) Review and approval of intervention with women’s groups. http://www.who.int/rpc/meetings/Lancet proposal Manandhar.pdf 4 Qualitative Study of “Sunaula Hazar Din” Community Action for Nutrition Project Nepal to be achieved during the 100-day RRNI cycle. internal review conducted in March 2015 (World Bank For example, if the WCF had selected a focus 2015b),9 and underwent restructuring in June 2015, to area of “improved handwashing,” then the RRNI address recurring challenges. To better understand and team would decide what kind of concrete activi- dig deeper on experiences of the unique community- ties they would work on in order to achieve better driven rapid results approach at the community level, it handwashing practice in 100 days. They would is important to understand what kind of overall imple- also select a RRNI team leader to coordinate team mentation challenges existed at the project level that affect members and activities. one’s understanding on incidences and outcomes at the 3. Preparation and submission of proposal: community level: Once a particular goal/activity is chosen, the »» Mismatch between the rapid results model and the slow RRNI team develops a detailed 100-day work implementation mechanism. There were delays involv- plan to help achieve that goal, along with the bud- ing proposal approval and grant disbursement get required to execute the plan. through local government systems, which cre- 4. Review and approval of the proposal: The ated confusion and frustration among the project submitted work plan and budgets are reviewed implementers (i.e., project management units and and then approved by the local government—by coaches) and beneficiaries (i.e., RNNI teams and a Village Development Committee (VDC) for target community members). It also delayed in projects less than $1,000 and by a District Devel- many communities the completion of the RRNI opment Committee (DDC) for projects between cycle, which hindered the rapid results model to US$1,000 and US$3,000, and the necessary funds fully realize its essence of building a momentum are released to the communities. for action through “enthusiasm” by celebrating 5. 100-day implementation: Communities achievements in 100 days. Due to such inefficien- start to execute their plan and achieve their goal cies, only one cycle could be completed as opposed within 100 days. Implementation is divided into to envisioned three cycles per year. Proposals for the first 50 days and second 50 days, separated the first 100-days cycle began to get approved only by a mid-term review to assess the progress, cel- in the third quarter of the fiscal year, with imple- ebrate successes so far and make necessary course mentation happening mostly in the last quarter. corrections. »» Lack of common understanding and communication among 6. Review of results and sustainability: Finally, various implementing actors created further challenges. after this 100-day cycle, the community organizes Effective project implementation depended on a ‘sustainability review’ to assess the achievements, collaboration between government officials at dis- celebrate successes and plan for sustainability of trict and VDC levels and national service provider the practices. The coach (and sometimes an out- (NSP) staff, such as district supervisors and coaches side monitor) also assess whether or not the goal (facilitators). NSPs are local non-governmental has been achieved. If the community has been organizations (NGOs) contracted by the project unsuccessful, it can apply for another cycle to try to provide project implementation and facilitation to achieve the same goal or another one. If the support, including provision of coaches and district community has been successful, it can next apply supervisors. There were three NSPs covering the for two goals at once. entire SHD project districts. The NSP complained about and tried to work around the slow bureau- cratic process, whereas the local governments felt Project Implementation they were bypassed by the project, reducing the Challenges level of government ownership at local levels. When the project began across 15 districts, it faced a range of implementation challenges, as noted in an 9 “Sunaula Hazar Din—Mid-term Review,” 2015. Sunaula Hazar Din Project Design and Implementation Challenges 5 »» Low implementation capacity further hampered the imple- supplements in earthquake affected districts). Several mentation process. The poor qualification of coaches changes to institutional arrangements were proposed to and supervisors hired by the NSP; inadequate increase the overall effectiveness of project management, human resources in the project team, particularly and to increase the efficiency and quality of the RRNI for monitoring and evaluation; and financial man- cycle (World Bank 2015a): agement and district-level coordination created »» Reallocating responsibilities, with the VDC’s role difficulties in timely processing of proposals. focused on approving the proposals (previ- »» Limited incentives, in the midst of other responsibilities, dis- ously a VDC could approve proposals only up couraged key government stakeholders from actively supporting to US$1,000 and a DDC had to approve those the project. Project coordinators, district focal persons, between US$1,000 and 3,000), whereas the DDC’s steering committees, VDC secretaries and local role focused on supervision and monitoring of all development officers were already overstretched RRNI projects. The VDCs also had to make rec- with other responsibilities for various projects, ommendations for proposals between US$1,000 causing a challenge as they were the key actors in to US$3,000. The VDC was also responsible for reviewing and approving RRNI team’s work plans, ensuring a 60-day turnaround between proposal and for processing the fund disbursement. submission and disbursement of funds. »» The above-mentioned implementation challenges were exac- »» Expanding capacity of both government and NSP imple- erbated by two events: (a) a massive earthquake that menters, by adding coaches (2 per VDC) and super- hit Nepal in April 2015 which not only damaged visors (2 per district)—however, while the project some of the tangible and intangible outputs that successfully added district supervisors and pro- the SHD project had achieved to date but also dis- gram assistants in each district, the number of placed or disoriented the communities and ben- coaches was never increased. The contracts of eficiaries and (b) an almost five month unofficial two NSPs (out of three) were discontinued due to political blockade at the Nepal-India border with poor performance and for making late payments ongoing Terai unrest following promulgation of to coaches, often by more than 6 months. Coaches the constitution in September 2015 which greatly and NSP supervisors were provided with booster impacted project implementation in Terai districts. retraining sessions, and DDC and VDC officials were provided with reorientation sessions on the In response to these challenges, the project was restruc- project approach, targets and implementation tured in June 2015, in order to improve its efficiency modalities. (along with an added component on provision of nutrition 6 Qualitative Study of “Sunaula Hazar Din” Community Action for Nutrition Project Nepal Chapter three Study Methodology Site Selection The study was conducted in 6 out of the 15 SHD project intervention districts, namely Sunsari, Udayapur, Sindhuli, Dhanusa, Sarlahi and Makawanpur (see Figure 2). Across each of the six districts, three different RRNI subprojects were selected jointly with the Government (Ministry of Federal Affairs and Local Development) SHD proj- ect implementation team based on the following criteria: a. “Hill” or “Terai”—ecological belts/terrain: The sampled project dis- tricts were to be representative of two ecological belts in Nepal (i.e., Hill and Terai), where different topographical and other social characteristics exist, which could have strong implications on how communities work together on a community-driven rapid results approach, and how household and social dynamics operate in the process. In general, communities in Terai (flat plains of the south) are more densely populated and have easier access to roads and markets compared to the hills—which could affect, for example, frequency of meetings among key stakeholders or procurement of materials to achieve a specific goal. b. Project implementation clusters (representation of 3 NSPs): Under the project, 15 SHD intervention districts were divided into 3 clusters, con- sisting of 5 districts each, supported by a separate “national service provider (NSP),” namely a partner nongovernmental organization which followed a common implementation framework established by the project. To avoid bias caused by only analyzing communities facilitated by a specific NSP, the sam- pled districts had to be representative of 3 project implementation clusters (i.e., all three NSPs). c. Selected focus areas: To allow comparison and a more in-depth analysis, the sample projects were chosen from the five most popular focus areas selected by the beneficiary communities (out of the 15 focus areas eligible under the project) across 15 project intervention districts—though the implementation modality (RRNI cycle steps) were common across different focus areas. The five selected focus areas were: end open defecation, increase animal protein Study Methodology 7 Figure 2. The sample districts of the qualitative study The boundaries and names shown and the designations used on this map do not imply official endorsement and authors do not warrant the map or its features are either spatially or temporally accurate or fit for a particular use. consumption, appropriate handwashing, use of subprojects in each of the six districts were chosen, clean and safe water, and reduce indoor smoke. paying attention to the following performances: d. Impact of earthquake: To focus on the »» Exemplary achievements in certain focus areas intended community-driven RRA steps and dif- or activities; ferences in outcomes across the sample RRNI »» Examples of successful and unsuccessful RRNI teams and wards, the study avoided those districts team management and community leadership or VDCs that were heavily affected by the earth- as noted by the project implementation team; quake in April 2015. This was due to the concern »» Examples of successful and unsuccessful that respondents might only end up talking about behavioral change through 100-day initiatives the earthquake and its impact on the project, as noted by the project implementation team; rather than social dynamics and other political »» Examples of successful and unsuccessful economy factors. approaches in conflict resolution; government e. 100-day project performance and fea- agency coordination; financial management; tures: In order to compare, as much as pos- procurement and transparency in the commu- sible, outcomes of different project performance nity; etc. (if any); and and features across samples, three sample RRNI »» Continuation and sustainability of initiatives long after cycle completion (if any). 8 Qualitative Study of “Sunaula Hazar Din” Community Action for Nutrition Project Nepal Table 1. Study districts, wards, and demographic information Demographic Information Cluster Ecological No. of Population Population Total No. Belt District Ward Households of Female of Male Population 1 Terai Sunsari A 52 128 142 270 B 429 1,022 835 1,857 C 342 1,205 1,033 2,238 Hill Udayapur A 56 154 124 278 B 41 109 114 223 C 39 100 95 195 2 Terai Dhanusa A 96 272 255 527 B 186 474 544 1,018 C 135 335 404 739 Terai Sarlahi A 241 738 789 1,527 B 161 475 496 971 C 59 183 187 370 3 Hill Sindhuli A 145 400 389 789 B 86 247 223 470 C 43 104 93 197 Hill Makawanpur A 307 952 925 1,877 B 53 160 154 314 C 119 344 321 665 Source: Field Implementation Record and Nepal Census (2011). Using these criteria, 18 study sites (which are at the ward level) across 16 VDCs were identified (see Table 1). The Study Population number of study sites was equal across the two ecological The study captured perspectives and experiences of six belts of Hill and Terai, and each of the selected districts types of key stakeholders who were directly associated consisted of three sites. The five most popular focus areas with implementation of 100-day RRNI projects. The were selected out of the total of fifteen focus areas. The description of six study group populations and qualita- number of samples taken in each of the five focus areas tive methods used are presented in detail below, and total: corresponded to the actual number of projects in those 6 DDC officials; 16 VDC officials; 16 coaches; 18 Ward five focus areas across all project intervention districts, and Citizen Forum chairs; 72 RRNI team members; and the final list included the following focus areas (and their 36 beneficiaries. respective sample sizes): 1.  DDC officials with direct engagement in the 1.  End open defecation (6 sites) RRNI process, particularly in proposal review, 2. Increase animal protein consumption (4 sites) approval, disbursement approval, monitoring 3. Appropriate handwashing (4 sites) and evaluation, and coordination. Key informant 4. Use of clean and safe water (3 sites) interviews were conducted with 1 DDC official in 5. Reduce indoor smoke (1 site) each of the 6 sampled districts. Study Methodology 9 2. VDC official with direct engagement in RRNI ensure representation of people from various disadvan- process, particularly in proposal review, approval, taged groups, which included Dalits, Janajatis, and Mad- disbursement approval, monitoring and evalua- hesis. Socioeconomic status of these groups reported that tion, and coordination. Key informant interviews the Hill and Terai Dalits had the highest poverty levels were conducted with 1 VDC official in each of the (42 percent), followed by Terai Middle caste (29 percent), 16 sampled VDCs. and Hill and Terai Janajatis (26–28 percent),10 all of which 3. RRNI coach responsible for facilitating RRNI generated further differences in terms of how effectively teams during all steps of the project cycle. The and inclusively community teams worked toward achiev- coaches were recruited by the National Service ing their goals. Janajatis, for instance, have their own dis- Provider (NSP) from the same VDC where the tinct traditional, social and cultural identity, which makes project was to be operated, or from a neighbor- some of them prone to being endangered or marginalized ing VDC when such eligible candidates were in the society. Likewise, the caste system, which catego- not available in the VDC. Key informant inter- rizes people into hierarchical groups based on division of views were conducted with 1 coach in each of the labor, roles and functions, has discriminated economically, 16 sampled VDCs. politically and culturally against “lower” castes such as 4. Ward Citizen Forum (WCF) chairperson Dalits through “untouchable” and caste-based practices engaged in RRNI focus area selection and RRNI across both Hills and Terai. Madhesis, or people of Mad- team selection, proposal review, implementation hesi origin, have been discriminated primarily on regional support and project monitoring. Key informant identity and language. 11 interviews were conducted with 1 WCF chairper- son in each of the 18 sampled projects. 5. RRNI team members were usually 9 in each Study Instruments ward/RRNI team, responsible for implemen- The qualitative study employed three methods for data tation and leading communities to achieve the collection: (a) key informant interviews (KIIs); (b) focus 100-day goals. Key informant interviews were group discussions (FGDs); and (c) observations. As dis- conducted with 3–4 RRNI team members of each cussed above, KIIs were conducted with district officials, of the 18 sampled projects. VDC officials, WCF chairpersons, coaches and RRNI 6. Beneficiaries of respective 100-day projects team members, whereas FGDs were conducted with only in the communities (who were not RRNI team beneficiaries. Observations of project environment or members). Two separate focus groups for men and outcomes were recorded by field researchers at project women were conducted in each of the 18 sampled sites during their visit to the study sites to help identify wards. Focus groups with women consisted of pertinent topics to probe and discuss during FGDs and project main target beneficiaries, such as mothers, interviews. pregnant women and adolescents. Focus groups with men consisted of either male beneficiaries Both KIIs and FGD were conducted using a common (such as in the case of ‘end open defecation’ proj- research guide framework, consisting of five sections cor- ects that sought to improve community-dependent responding to specific stages of the RRNI project cycle factors not reliant on one gender) or men who had (see Figure 1), and lessons learned. Such internal consis­ facilitative roles in improving nutritional outcomes tency across all instruments allowed for exploring a range (such as in the case of ‘increasing consumption of protein among pregnant and lactating women’ projects that benefit from active facilitation by 10 NPC/CBS. 2013. Nepal Thematic Report on Food Security and Nutrition. male members of the household and community). Supported by: WFP, WB, Ausaid, and UNICEF. Kathmandu, Nepal. 11 Asian Development Bank. 2010 “Overview of Gender Equality and Social In addition to considerations for gender sensitivities in Inclusion in Nepal.” https://www.adb.org/sites/default/files/institutional- selecting focus group members, the study also sought to document/32237/cga-nep-2010.pdf 10 Qualitative Study of “Sunaula Hazar Din” Community Action for Nutrition Project Nepal of perspectives and experiences of various stakeholders 1.  Training was provided to field supervisors and around one particular issue or process covered in that spe- researchers by a World Bank team and the local cific section. The description of each section is provided research firm, including orientation on the SHD in detail below: project and its objectives, qualitative methods, 1.  Section 1: RRNI team selection and inter- and study instruments, together with extensive nal team dynamics: To understand the level hands-on practical sessions both in the classroom of knowledge and experience of stakeholders in and in the field before final deployment; the first critical step of the RRNI cycle, namely, 2. Study instrument standardization was done RRNI team formation and collaborative working through a field testing of a draft research guide process, the questions sought to explore the level (for focus group discussions, interviews, etc.) devel- of social inclusion in teams (i.e., representation oped by the World Bank research team. The abil- and nature of participation of women, Dalits, ity of field researchers was also assessed; religious minorities), the process of selecting team 3. Full-scale field data collection was con- members, roles and responsibilities of each team ducted after verifying that all field researchers member, and so on. had mastered the necessary skills and knowledge 2. Section 2: Selection of focus area and goal for conducting high quality data collection. With setting for 100-day project: The section coordination support from Ministry officials, six explored the process and factors that influenced teams, each comprised of 2 field researchers and focus area and goal selection. 1 field supervisor, were deployed simultaneously 3. Section 3: Implementation of the 100-day across 6 districts; project: The questions in this section focused on 4. Quality assurance and data management the challenges faced, and the solutions devised, was made on a continuous basis by the local with regards to team dynamics, financial manage- research firm, field supervisors, and the World Bank ment, community mobilization, and interaction team. Each field researcher produced detailed between and among various stakeholders such as observation notes explaining the challenges in government officials, coaches, beneficiaries, etc. data collection, general features of the community 4. Section 4: Outcomes and sustainability and projects and any other salient observations to of changes: This section asked about observed help understand the data. Field researchers were changes, unexpected changes, and sustainability responsible for preparing handwritten transcripts of such changes, especially related to changes in from an audio record in Nepali, which were later behavior or practices for improving nutritional translated into English and submitted for analysis outcomes. by the World Bank research team; and 5. Section 5: Lessons learned: The last set of 5. Data analysis was conducted using NVivo 11. questions collected recommendations for pro- A coding scheme was prepared with codes and grammatic improvement in enhancing the quality sub-codes based on the sections of the interview of implementation. guides and the key research questions. Codes were added if new themes emerged from the data. To manage the large volume of data, the analysis Data Collection occurred in a two-step process. First, the data from Procedure and Analysis each of the five selected focus areas was written up in mini analyses on the key research questions, The World Bank research team partnered with a local and on themes that emerged organically from the research firm to assist in qualitative data collection. After data. Then, the mini analyses were collated into obtaining an ethical clearance from the Nepal Health one report that constituted the findings presented Research Council, data collection and analysis took the in Chapter 4. following procedures: Study Methodology 11 Chapter four Key Findings A. On the Sunaula Hazar Din Rapid Results Nutrition Initiative (RRNI) cycle a. RRNI Team Selection and Representativeness The first step in the RRNI cycle was the selection of a nine-member RRNI team by the members of the Ward Citizen Forum, at a meeting organized and moderated by the SHD coach.12 The composition of the RRNI team was intended to reflect the eth- nic, gender, caste, class, and occupational diversity of the ward, and the RRNI team was tasked with implementing the SHD project. The data show that there are four different ways that the RRNI team selection process unfolded on the ground, with varying implications for the implementation rollout. The four pathways are derived after analyzing data on the following questions: What was the team selection process? Who participated in the team selection meeting? What cri- teria were used for member selection? If the respondent did not personally participate during selection, were they aware of the selection process and criteria? Was the team representative of the caste, class, religion, gender, and occupational composition of the village? What were the roles and responsibilities of the team members? The pathways are outlined below, with illustrative examples. To provide a fuller pic- ture, each pathway follows the data from team selection to implementation rollout. The first pathway represents the majority of the wards in the sample, where the team selection process proceeded as planned. The Ward Citizen Forum (WCF) was involved in the selection of the RRNI team, the RRNI team members were aware of their spe- cific roles, and an effort was made to include both genders and the various caste, eth- nicities, and occupational groups in the wards. Even though beneficiaries were unlikely 12 See pages 4–5 for detailed steps on the RRNI cycle. Key Findings 13 to know the details of the team selection process or crite- members of the nine-member team. However, unlike the ria, they were likely to be aware of the RRNI team’s work pathway of elite capture, the powerful did not usurp all plan and goals. the resources, but channeled some effectively to intended beneficiaries. Even though not community led, this sce- The illustrative example contrasts two wards, both of nario still resulted in community benefits. This pathway is which followed the prescribed guidelines for RRNI team referred to as “benign capture.” selection. Both wards are in the Hills, and with accessibil- ity issues that challenged implementation processes, both The fourth pathway is specific to the Hill wards in the chose the focus area of “Increasing Animal Protein,” both sample. Four wards in the sample fit into this category. chose the same goal of establishing community chicken Some of these wards, thinly populated to begin with, were coops and distributing eggs, and both faced budget delays. further diminished numerically due to migration for work However, the beneficiaries in one ward were satisfied with of male members of the households. It was difficult to the project’s achievements, and disappointed in the other. find enough people who were interested to participate as It is useful to explore the reasons to understand how, even RRNI team members. The RRNI team members said in though the team selection process was similar in both interviews that they were reluctant to become members cases, differences in implementation approaches led to as they did not have the time to volunteer. The specific differences in implementation outcomes. challenges this presented for implementation rollout were considered. This pathway is referred to as “population The remaining three pathways described below occurred density issues.” We included this in the report because of in a handful of cases each. They are included here to its implication for future implementation efforts in Nepal’s allow lessons to emerge on the full range of possible path- Hill districts. ways of RRNI team selection processes. The second pathway for RRNI team selection process Pathway 1: According to Plan—Involvement and subsequent implementation rollout was one of “elite of WCF and Attempt at Achieving capture”: a handful of local powerful people controlled Representation, with Varying Results the entire project, leaving little room for community- for Project Beneficiaries driven selection processes. Two out of eighteen wards in Most wards in the sample followed the protocol of involv- the study—both in the Terai region of Nepal—fit into ing the Ward Citizens Forum (WCF) and using the WCF this category. Some RRNI team members interviewed meeting as the platform through which to introduce were themselves unaware of the selection process and the SHD project, explain its objectives and modalities, criteria, and had none to very little involvement in the select RRNI team members and select a focus area and implementation. The beneficiaries interviewed in focus a goal. Respondents also stated that attempts were made groups also knew little about who was on the RRNI team, to include women and ensure representation from vari- or the selection criteria, or about other project decisions. ous castes, ethnicities, occupations, and religious groups. The project was not community driven, and project ben- Some respondents also stated, although others challenged efits, distributed in a scattershot manner, were considered it, that political parties were not represented because that of little value by the beneficiaries who felt that the funds is not a criterion that was used for selection. were seized by a few powerful people. This process did not ensure that the beneficiary commu- The third pathway presents evidence of some elements nity was equally aware about the selection processes; in of elite capture. Only one ward in the sample—in the fact, it was rare for either male or female participants in Terai region—fits into this category. There was a lack of beneficiary focus groups to know much about the details transparency in the selection of RRNI team members, of how the RRNI team was selected, and sometimes to and after selection, decision-making power was restricted even be aware of who exactly was on the RRNI team. in the hands of only a few, sidelining the remaining 14 Qualitative Study of “Sunaula Hazar Din” Community Action for Nutrition Project Nepal Having a community-led process of team selection did more inclusive approach that required ward residents to not always ensure favorable implementation outcomes. contribute to the project, it did not work. It is worth con- Box  1 contrasts two wards in two different Hill districts sidering whether the community would have been more that chose “increase consumption of animal protein” as accepting of the approach taken by Ward B if the RRNI their focus area, decided to raise chickens and distribute team and the SHD project had anticipated the challenge eggs to households with pregnant women and children of budget delays, and planned to manage community below the age of two, and faced implementation chal- expectations differently. lenges with budget delays. However, beneficiaries of only one ward were positive about the project. The reasons for the difference are considered. Pathway 2: Elite Capture—Beneficiaries Viewed the Project as Not Beneficial Budget delays are a known factor in Nepal. The contrast- In the cases that fit under Pathway 2, the entire project— ing cases in Box 1 highlight what it takes for projects to from the selection of the RRNI team, the focus area, the overcome this commonplace challenge. It required the goal, and the work plan, to budget and implementation— ability of private citizens to fund project activities until was controlled by a handful of locals, who became RRNI the money could be released. When another ward tried a team members. There was no functioning RRNI team in Box 1. Pathway 1—RRNI Team Selection Process Community-Led & RRNI Team Representative of Ward Population Wards A and B are Hill wards. Both wards followed the RRNI team selection process guideline, involving the WCF and forming a team representative of the constituents of the ward. Ward A, with 307 households and 1,877 residents was almost twice as large as Ward B, with 145 households and 789 residents. Ward A first targeted 75 households, later raising this number to 120. Ward B targeted 40 households. Both wards reported sig- nificant implementation challenges due to the difficult terrain, including administrative challenges of opening a bank account that took one full day to walk to, the problems of transporting materials for building the hen coops and bringing the chickens and feed, as well as distribution challenges which required beneficiaries to walk for hours to collect the eggs. “We had to go to Hetauda to open a bank account . . . there were other banks that are more accessible, but it was not allowed. To reach Hetauda, we first have to walk 6–7 hours to Lother and then another 2 hours to reach Hetauda . . . you have to return the next day . . .” [Coach, Ward A] “Even though we built the hen coops in the middle of the ward, the ward is big, and for some people who live uphill, it can take 1.5 hours, or even 2 hours if you walk slowly, to come to collect the eggs. Then they have to walk for even longer to go back uphill.” [Coach, Ward B] Both wards also reported significant implementation delays. However, the two wards diverged in overall project performance. Ward A—despite being the larger ward with a greater target population—received positive reviews from beneficiaries, whereas Ward B beneficiaries said that after some time, the number of people who collected eggs dwindled considerably. There is one key difference that is important for community-based projects to consider, related to how the two wards handled bud- get delays. Ward A relied on the RRNI team members taking personal loans and being repaid once project funds were deposited into their bank accounts. In Ward B, the beneficiaries were required to pay Rs 10 a week to receive the eggs, later raised to Rs 50, which was justified as necessary to help tide over the financial difficulties of maintaining the hen coop while there were delays in project disbursement, and a means of ensuring sustainability of the project. Beneficiaries refused to pay. “. . . we first decided to distribute the egg without any charge. If we had continued like that, it would have been better . . . the coach said that if we collected money, it would be much easier to sustain the program for a longer duration. . . . so we decided to collect ten rupees every week and again later it was increased to 50 rupees. The [beneficiaries] didn’t agree with the increment and so we decided to end the program . . .” [RRNI team member] A female focus group participant from Ward B said, “in the end, only 3 or 4 were left who took the eggs.” This is not to suggest that how they dealt with budget delays is the only difference between the two wards. However, the two wards did take different approaches to overcome this obstacle, and there are lessons to be learned from this for future project implementation. Key Findings 15 Box 2a. Pathway 2—Elite Capture Ward C, in the Terai region, has 52 households and 270 residents. The ward chose “Reduce Indoor Smoke” as their focus area, and the goal was to distribute smokeless stoves. However, respondents were unable to articulate a clear work plan on how beneficiaries were selected. RRNI team members were unaware about why they were selected: “I did not even know about my selection. The Team leader selected me and only she knows the reason.” “I am unknown about who selected me . . . If we had gathered I would have known, but they selected me secretly . . . I was not present during the selection period, I was busy with my own work.” Respondents representing various constituencies in the village spoke openly about elite capture: “Nobody knew about SHD. . . . Only after a long time we came to know about its purpose. People who were of higher class and of higher power were in the team.” [RRNI team member] “In the stove distribution project, even though I am the coordinator [of WCF], they did not tell me anything, nor did I participate in any of the pro- grams. I am unknown about who the coach selected.” [WCF member] “. . . everything was done by 2–4 individuals who formed a group and did the work.” [Male focus group respondent] The consequence of elite capture was that a handful of powerful individuals, and their relatives, received a disproportionate share of project benefits. These powerful few gave small monetary incentives to entice the non-powerful RRNI team members to work on their behalf, promises which were not kept and that resulted in further social disruption. The villagers viewed the distributed goods as having little value. “. . . the funds and goods that were received from SHD such as for filter, stove . . . was given to the relatives of the coach and some other powerful people . . . their relatives were always there. If others had to be given, then they tried to do it in the least expensive way possible . . . they took any remaining money.” [Male focus group respondent] “At first they said they would give 300 rupees as allowance for a day but they did not give even 200 rupees. We worked very hard for 10 days but they did not give us any money . . . We collected bricks and soil for the construction of the stove. The money had come in the committee but they did not give us. Instead they divided it among themselves. . . . They gave only 200 rupees after much quarrels and fights . . .” [RRNI team member] “They gave stoves to everyone, and they also had said they would give pipe which emits smoke outside. But they never gave us pipes. They just made stove of mud, how will the smoke go out? Even now, the smoke does not go out.” [Female focus group respondent] There was change in awareness on the adverse effects of smoke on health, although respondents did not articulate this in terms of effects specifically of the SHD project. Respondents did not think the project was sustainable because the goods distributed were outmoded. “Who will use this stove, when all the houses want to use [LPG] gas for cooking?” [Female focus group respondent] the manner envisaged by the project. The coach colluded example, where elements of elite capture were present, with, or was already part of, the local elite. The issue of but did not result in the beneficiaries giving negative representativeness of the RRNI team became moot, and reviews about the project. the ward residents did not perceive the benefits that they received as being of any value. Pathway 3: Benign Capture— Elite capture is well documented and well understood Some Elements of Elite Capture, in development projects. The examples above high- but Ultimately Villagers Benefitted light the need for continued vigilance to prevent misuse There were elements of non-transparency and lack of of resources and social disruption. On the other hand, representativeness of the RRNI team as described in interestingly, Pathway 3 below provides a countervailing Pathway 2. However, there were crucial differences that 16 Qualitative Study of “Sunaula Hazar Din” Community Action for Nutrition Project Nepal Box 2b. Pathway 2—Elite Capture Ward D, also in the Terai region, was the largest ward (by population) in the sample, with 342 households and 2,238 residents. The ward chose Open Defecation Free (ODF). This was a Muslim-majority ward. There was no community buy-in when SHD was launched, and the WCF was unaware of the team selection process. Two or three RRNI team members were in complete control of the project, while others were “in name only.” Unlike the case of Ward A above, the coach, being new, was not part of the powerful. The coach was hired in a rush, was not given training, and had to learn on the job. The work plan was to construct 20 toilets, but respondents were unsure on what basis the 20 were selected. There were arguments and fights about receiving project benefits, including physical fights. In the end, it is unclear whether the stated plan was conducted systematically. Beneficiaries felt that those who raised their voices in objection got more benefits, and it was not the neediest who ultimately benefitted the most. None of the focus group respondents thought that the desired results were achieved. “. . . 15 lakhs were allocated for our whole VDC . . . When we surveyed all 9 wards, in total no more than 7–8 lakhs worth of work has been done, that is what we felt. This is because some households only received 4 rings, and those who raised voices got 5 and those who forced more got 6, those who didn’t speak at all got 3 . . . there was no differentiation between rich and poor . . . the neediest didn’t get . . .” [Male focus group respondents] Box 3. Pathway 3—Benign Capture Ward E, in the Terai region, has 135 households and 739 residents. The ward chose ODF as their focus area and built toilets. Like Pathway 2, the selection process was done by the coach in a non-transparent manner, there was lack of representation on the team, and the implementation was in the hands of a few powerful people. “No meeting was called to select the team. . . . Suddenly, the coach arrived in the WCF meeting and told us that she had a program named SHD and she needed to select people from different wards. She selected all 9 members . . . There was no rule and regulation for selection. . . .” [RRNI team leader] “I was not present during the meeting; I was out of the village. When I returned, they had already written my name. . . . [RRNI team member] Furthermore, three respondents—all RRNI members—said they were members “in name only,” and others more powerful than them made key decisions. We didn’t attend every meeting as we were not called during important ones. I heard that meetings were held only with three or four important people.” “They only used to ask me to sign the documents. I rarely attended the meetings. I don’t know much.” “I was a member in name only. They never gave me any responsibility.” On whether all castes were represented, respondents thought that certain castes had been left out, and named the castes that they felt should have been included. “. . . some people who should have been on the team were left out. There were people from Bheriyar and Dhanuk caste in the team. We have other castes in our ward, however they were not present during the WCF meeting . . . if they had informed us earlier, we would have included Hajam and Sharma. We thought that it would be easy to work if people from all the caste were included, they would rightly speak for themselves bring their problems and perspectives. I think Sharma and Hajam were left.” [RRNI team leader] “. . . if Lohar caste was added, it would have been better.” [RRNI team member] “We didn’t include Sahus, I feel bad about it. If we had included them, it would have been better.” [WCF member] However, unlike in Pathway 2, the elite in this case developed a clear work plan, listed targeted households, and convinced ward residents about the rationale for the targets when complaints of being left out arose among non-beneficiaries. The RRNI completed the work outlined in their work plan. The coach (and the handful involved in the project) were not viewed as only working for theirs, and their relatives’, personal benefit. (continued) Key Findings 17 Box 3. Continued The project targeted 23 households that were in a weak financial position, and had pregnant women and/or children under the age of two years. They were provided 3 rings and a seat, with the expectation that the households would themselves construct a door, walls and a roof. Those not included as beneficiaries protested, but were told that they had the financial capability to build on their own. “At the individual level as an RRNI member, we were all facing the same problem, so we decided to go in a group and explained to ward residents that pregnant women, those families who had children less than two years old, and those who could not afford, are the focus of this program and they will be helped. We explained our priorities and suggested to them to build a toilet by themselves as they are financially capable. Villagers were convinced.” [RRNI team member] While some community members perceived that more toilets could have been constructed with the available funds, there was gen- eral agreement across all respondents—both in interviews and focus groups—that the project had brought changes in knowledge (about the benefits of not defecating in the open), as well as changes in practice (fewer people are openly defecating and people have started to use soap and water as well). There have also been unexpected changes: “I did not expect that the non-beneficiaries would also make toilets. I had thought that if the project would make toilets, they would destroy them, but instead they began to make their own. They had a conservative thought that toilets were made in the cities, and villages do not need them. But now I can see this has changed. People have learned that it is good to build a toilet and it is for their daughters’ and daughters-in-laws’ safety and the health of their children. I never expected this would happen.” [RRNI team member] ultimately led to different project outcomes. Beneficiaries long distances for daily chores, and hesitate to add to their expressed satisfaction with the project. burden by undertaking volunteer work on development projects. Furthermore, many residents migrate for work, It is instructive to consider the difference between elite leaving behind an even more diminished population, and capture and benign capture. Both pathways had elements the residents that remain are overworked. of the project being controlled by a handful of powerful individuals, with beneficiaries wondering about misuse of However, it would be incorrect to conclude that all Hills funds. There was distribution of goods and materials in are the same. Some wards in the Hills are accessible both (mud stoves in one case, and rings/seats for latrines by road, even though unpaved, where a vehicle can be in the other). The difference is that there was a clearly used, whereas other wards in the Hill district can only be articulated work plan that was followed in the path- reached on foot. Better connected wards did not face the way of benign capture, and the community was aware issue of community reluctance to participate in project of the rationale behind it. Even though the project was work. Thus, although there were 9 wards in the sample not community led, there was community buy-in, which that were in the Hills, only 4 wards presented this chal- resulted in outcomes that were viewed positively by proj­ lenge. The observation notes of the research team about ect recipients. accessibility of two wards, both in the Hills are contrasted. “Most of the wards in the VDC were accessible by roads, though Pathway 4: Population Density Issues— unpaved, and [the VDC] also has basic facilities of electricity, Community Reluctant to Participate in communication etc.” [Observation note of research team in Sindhuli district] Hills Due to Inaccessibility and Migration Four Hill wards in the sample, all thinly populated, “To reach the District Headquarters from this village takes two reported the challenge of finding enough members to form days. However, to return to the village takes 5 days. To go from RRNI teams. The hilly terrain, combined with a lack of Ward No. 4 to Ward No. 5 in this VDC takes 5 hours by walk roads, means that activities that take hours to complete in . . . there is no scope to use any vehicle. . . .” [Observation other regions in Nepal take days in the hills. People walk note of research team in Udayapur district] 18 Qualitative Study of “Sunaula Hazar Din” Community Action for Nutrition Project Nepal Box 4. Pathway 4—population density issues Ward F is the smallest ward (by population) in the sample—with 39 households and 159 residents. It is difficult to find enough people to participate in development programs. “It’s a bit difficult to work in the village. There are many committees. When the people are told that they should stay and work, they say they don’t have time.” [Coach] RRNI team members in this ward were uninformed about their selection into the teams. “I don’t know who was there during the team formation. I stayed at home. My children, i.e., sons and daughter went there in the gathering and wrote down my name . . . . I didn’t go for the training because once, when I had been there, and they just said ‘hurry up, you need to sign’ and so I did. . . .” [RRNI team member] “I came to know that I am an RRNI Team member today only.” [RRNI team member] Ward members who had reluctantly given their names as members found that they had to personally incur financial costs for SHD project work, which resulted in reinforcing their belief that such projects were not for the benefit of the villagers, but only meant to enrich the more powerful VDC and DDC officials. To complete a piece of work for the SHD project, three RRNI members went to the district headquarters, a return trip that took several days. They were given 3,000 rupees each from the project for their lodging, transport, food, etc. The work kept getting delayed because the DDC administrators were not available, and they were forced to stay for 15 days. They ran out of their allo- cated money and had to pay out-of-pocket. Such experiences added to the frustration and anger felt by the villagers. “Such projects are only a playground for DDC and VDC officials to make money in the name of development by fooling locals . . .” The following quote is illustrative on the reluctance of selection was generally demand driven, in part because the community members to show enthusiasm for projects: community thought that all 15 focus areas were relevant. Respondents explained that the choice of focus areas was “There were only 4 people present during the WCF gathering narrowed down by the visibility of project outputs within for the first time. We had informed villagers about the meeting the 100-day timeframe of the project, the available bud- but they didn’t attend.” [RRNI team member] get, and cultural norms. “In our remote village, there are many hills and people have to walk around a hill to reach each other’s house. So, the main problem was the absence of the people of the community.” Focus area selection was generally [RRNI team member] demand driven Across wards, respondents from all categories—RRNI team members, male and female beneficiaries, VDC/ Box 4 highlights that inaccessibility issues are a huge ward level functionaries—generally thought that the cho- challenge, and if not paid attention to, can result in both sen focus area was needed in the community. In one case, misuse of resources, and the building up of resentment when the DDC disagreed with the selected focus area, the and suspicion toward development projects in the local community advocated for its need: population. “We have 15 different working areas. Those 15 areas were explained to the group and suggestions came that priority should b.  Selection of Focus Area be given to free open defecation in our VDC, but this decision After the RRNI team was selected, the next step in the was not supported by the DDC. So, the people of the RRNI cycle was the selection of a focus area out of a menu of team went up to the DDC and insisted that this was the most 15 possible focus areas. The data show that the focus area needed focus area.” [Coach] Key Findings 19 Selection of focus area based on visibility Selection of focus area based on budget of project outputs in 100 days One ward rejected ODF because they thought the budget The ability to show visible project outcomes in 100 days was too little to construct toilets. Another ward elected to determined choice of focus area. As a DDC official stated, distribute smokeless stoves because the budget was per- “a 100-days approach is good. There is a need for small projects that ceived to be appropriate for this activity. give quick returns.” For example, one ward chose handwash- ing over increasing consumption of animal protein. The Selection of focus area based on cultural norms reason given was that the animal protein project required While more data is needed to validate this claim, one DDC a lot of effort to build hen coops and buy and rear chick- respondent claimed that Madhesi communities in the Terai ens. It was not possible to buy hens that were ready to lay region were not accustomed to rearing chickens, and only eggs from Day 1 of the project. Or, projects ended up Muslim communities in the Terai raise hens for eggs. buying hens that were old enough that they could poten- tially give eggs as soon as the project started, but there was always the worry that these older hens would die soon. c.  Selection of Goal After narrowing down 1 of a 15 possible focus areas, the “As I see it, consuming eggs is an invisible type of outcome. In next step for the RRNI team was to select a goal and hundred days, the outcomes must be shown. Hens can’t be found develop a work plan to achieve the goal. In the Hills, that have just started laying eggs; they are not sold often. You teams selected goals based on whether the goal could be need to invest a lot of effort to bring the hens to that stage and achieved given the difficulty of transportation of materi- we only have 100 days. Projects would end up with only those als in the hilly terrain. Another critical issue was whether hens which have nearly come to the age of losing the capacity the targeted groups should cover a more narrowly defined to lay eggs. We bring them and keep them, after laying eggs nutrition-related population (e.g., pregnant women and for two–three months, they would stop laying; they would die.” children below 5 years of age), or include marginalized [Coach] populations as a whole regardless of age or gender. The selection was made based also on the 100-day timeframe The RRNI team member in the same ward reasoned that that was specific to the rapid results approach. The voices the community could more easily feed eggs themselves to of people therefore show the importance of making sure pregnant women and children, and it was better to use the the project design fits well in the context and strengthens, funds for the handwashing project, with visible outcomes not hinders, demand-driven investments. and ability to spread the benefits universally (more on uni- versal vs. targeted benefits in the next section). Selection of goal based on terrain “We thought that this hand washing project would be the best. Table 2 shows the differences in selected goals within the We can see the outcomes soon. We could choose the egg dis- same focus area. tribution program, but we thought that we could feed the eggs to the pregnant women and children on our own. We thought Table 2. Focus areas and selected that this project would be useful for everybody.” [RRNI team goals in our sample member] Number Focus Area Selected Goal of Wards Respondents who selected “Increasing Animal Protein” as Animal protein Raise chickens and distribute 3 their focus area felt that the project should have lasted for eggs 2–3 years, not 100 days. Distribute eggs and milk 1 “These types of programs should run for 2–4 years instead of bought from the market 100 days.” [Focus group participant] Handwashing Distribute drums to store 2 water Distribute buckets and soap 2 20 Qualitative Study of “Sunaula Hazar Din” Community Action for Nutrition Project Nepal Four wards chose “increasing consumption of animal pro- “The people say we didn’t get anything; they will break the tein” as their focus area. Three—all in the Hills—chose heads [of those who say they should not get benefits] with sticks to rear the chickens themselves and distribute the eggs and ask, “how come we are not getting it?” Thinking of this, we that were laid. This activity was labor-, skill-, and cost- decided that all should be provided the benefits.” intensive. The fourth ward—in the Terai belt—opted to distribute eggs and milk bought from the market. One The decision on broader targets sometimes went well, and possible explanation for this is that the Hill wards did not other times did not. Consider the case of two wards in have the option to choose the easier route of buying and the same VDC, under the same coach, that decided to selling eggs, because these wards are inaccessible, making target all economically disadvantaged. The VDC decided it a challenge to reach the nearest market. It is a hardship that every ward would implement Open Defecation Free to procure large quantities of milk and eggs for distribu- (ODF) and construct toilets for those who could not afford tion. Similarly, the two wards that selected the distribu- it, and only provide materials (but not assist in the con- tion of buckets and soaps initially considered the option struction) to those who were financially secure. The SHD of procuring 200-litre drums used for storing water. The project funds would be used so that the VDC would be large drums are more useful than the smaller buckets, and able to declare itself “Open Defecation Free.” However, the ability to store large quantities of water increases the one ward was twice the size of the other, in terms of total likelihood that the project would lead to sustainable hand- population. Also, the larger ward had a larger number of washing practices. However, the wards finally decided marginalized community members. After drawing up a against it because the cost of transporting the large drums list of the economically disadvantaged households, the uphill was cost prohibitive. smaller ward concluded that it needed to build 22 toilets, and the larger one 27. However, after community mem- Nutrition-focused targeting vs. targeting bers who were not included as beneficiaries resisted this plan, the totals were altered: the smaller ward decided to more broadly build 27 instead of 22, but the jump in the larger ward was In selecting goals, several RRNI teams discussed whether severe—from 27 to 144. In the end, the budget proved to target households that had “thousand days mothers” sufficient for the smaller ward, the beneficiaries were (pregnant women and children under the age of 2 years), satisfied, and the RRNI team and community members or spread the benefits more broadly to the economically felt energized by their accomplishment. Comparatively, disadvantaged, regardless of their “thousand days” status. the larger ward did not have a sufficient budget to meet A VDC-level respondent expressed the dilemma: ev­erybody’s expectations, and there was confusion about “SHD was for pregnant women and children below 5 years . . . who should be included. The coach explained it thus: there was an agreement between the DDC and VDC for this. “[The larger ward] . . . had more population. There were But how much support would this provide to the very poor peo- also more Dalits and more marginalized. . . . Other RRNI ple? That is why we decided there should be ODF for the whole teams had less population so the funds were sufficient, but the VDC, toilets should be given to households that don’t have them, available funds did nothing in [the larger ward]. It was dif- and there should be improvement in health for all.” ficult to maintain the program. It was very difficult to provide everything to everyone. It was also difficult for them [RRNI Or, as stated by an RRNI member: team members] to make a decision on who should get toi- “. . . targeting makes it difficult to run programs, and brings lets.” [Coach] disputes between people.” The case above reflects the need to make budgets avail- Another concurred that the broader targeting creates less able according to population size of the different wards, social conflict: and the failure of the larger ward to manage beneficiary selection. Key Findings 21 Another ward worked its rationale backwards after decid- “Regarding the animals, there was no option because of the ing that they wanted to choose a focus area that would 100-days project except to bring the egg laying chickens.” allow them to provide benefits for all residents. They [DDC respondent] decided that in order to target everybody within the avail- able budget, they must choose handwashing. “We decided to distribute drums and cement to all the houses. Example of demand-driven goal selection After a ward selected handwashing as their focus area, We didn’t leave any one behind. There are some rich people, the RRNI team selected the goal of distributing 200-litre some poor, some upper caste and some lower caste, but we did drums in which to store water. In addition, they wanted not differentiate. [RRNI team member] to distribute buckets and soaps, but they changed the goal after receiving community feedback: Unsurprisingly, this approach caused less social disruption: “We had planned to provide materials like buckets, towels and “The program didn’t separate—this is higher group; this is soaps but the community people said that they wanted cement lower! It has given us equally. We are satisfied. Everyone got it instead to build a platform, and they will buy buckets and soaps equally.” [Male focus group participant] by themselves.” [RRNI team member] However, some RRNI teams tried another approach, deciding to use both criteria and target the economically B. Inclusion of Women disadvantaged “golden thousand days mothers.” Some Every ward in the sample included women on RRNI decided to target in the first phase, and make it broader teams. Yet, gender inclusion was not a simple task. in the second phase of the project. When the commu- Unequal gender norms that prevent women from occupy- nity resisted the beneficiaries’ list and demanded inclu- ing public roles resulted in low interest among women to sion, successful RRNI teams anticipated the problem and participate. SHD projects felt pressured to include women resolved it through careful engagement and reasoning (for to fulfill the requirements outlined in project documents. more details, see subsection D on leadership of coaches.) Women were coerced to join, but even if they showed an inclination to participate, their voices were not heard, and The data from this study cannot conclude whether one they were not given decision-making power. This was even approach is better than the other. The data is conclusive, the case where the RRNI team leader was a woman. Proj- however, that regardless of strategy used, managing the ects paid attention to form and appearance, but missed budget and the community’s expectations are both critical striving for gender equality in substance and content. in successful project performance. When asked whether women were included in RRNI Selection of goal based teams, respondents across all wards and all respondent on “100‑days” timeframe types answered overwhelmingly in the affirmative. Indeed, Presented above is an example above of how a ward all 18 wards in the sample included women in their RRNI rejected the focus area of “increasing animal protein con- teams, and 12 out of 18 RRNI teams in our sample had sumption” because of the need to show visible outcomes a female RRNI team leader. However, the participation in 100 days. In another ward that selected this focus area, of women in project implementation followed the rules the rapid results approach influenced their selection of on paper, rather than the spirit, of the project’s objectives. a goal. This ward considered rearing buffaloes or goats Several issues arose regarding the inclusion and participa- for their milk production, but determined that raising a tion of women. calf (which was cheaper than buying a full-grown animal) would take too long to show results within the project’s 100-days cycle. 22 Qualitative Study of “Sunaula Hazar Din” Community Action for Nutrition Project Nepal Difficult to include women “In our Muslim community, females are not allowed to leave Respondents who participated in team selection pro- their house. Family members don’t show interest in letting them cesses (e.g., coach, Ward Citizen Forum representative, take adult literacy classes. They say, “Why should women etc.) expressed difficulty in persuading women to be on study? Why should they go outside?” They have concerns that teams. Women declined because unequal gender norms the women will become clever. Most of the women in this VDC presented roadblocks: women’s time was consumed with who are Muslim do not have citizenship. People think that if household work, families prevented women from partici- citizenship is obtained, women will ask for their rights and run pating in the public domain, and women’s involvement away. However, to be in the RRNI team, citizenship number is outside the home was viewed as a threat by male house- required. Without citizenship, it is hard to have bank account.” hold members. [Coach, Terai district, ODF project] “Women’s involvement was compulsory. . . . In our societies, girls are not allowed to work outside especially among the Man- Women forced to be on RRNI teams dals and Bheriyars . . . we had explained and suggested the Women were coerced to be on RRNI teams, presumably names of Mandals. . . .” [Ward citizen forum represent- to follow project guidelines on RRNI team composition. ative, Terai district, ODF project] However, after joining, women either willingly chose not to participate because they were uninterested, or their “In comparison to females, more males were involved in the participation was not actively sought. As one respondent teams . . . there is a huge discrimination between males and put it, women were members “in name only.” females still in our society. The men can openly attend meetings and do whatever they wish, but females are compelled to look “I was not interested [in RRNI membership]. I declined after the cattle and children. They are not given the opportunity many times but they forced me. After they approached me many to show their talent.” [Coach, Hill district, animal pro- times, I agreed . . . I am in the team because the program was for tein project] the welfare of the villagers. . . . I have household chores but they said that it would be enough if I could manage only one hour. I “Making women leaders is quite challenging as illiterate men did not have to do any work. We didn’t attend every meeting as start to feel threatened. They start to quarrel and it ends up we were not called. . . .” [Female RRNI team member, breaking up families. Literacy projects should be initiated at the Hill district, animal protein project] same time as development work. It helps to at least have literate people in the group. It is more difficult to work with illiterate “I don’t know what my role was, I had to write my name that’s people and on top of that illiterate women.” [Coach, Hill all, I went to write my name that also for one time only. After district, handwashing project] that I didn’t go. Others used to go . . .” [Female RRNI team member, Hill district, handwashing project] While finding women to include on teams was challeng- ing everywhere, there was an added concern reported in Women do not overcome gender roles one ward in our sample that had a largely Muslim popu- when included lation. Not only were families reluctant to let women step Even when they attended meetings, women followed tra- outside the home, an additional issue was that of citizen- ditional gender roles. ship. Muslim women do not necessarily have a “citizen- ship card,” which was noted by the coach as a prerequisite “There were many males and rich people so we didn’t talk in for becoming an RRNI team member since this card is front of them.” [Female RRNI team member, Terai dis- essential to holding a bank account. trict, ODF project] “Most of them are Muslims in the ward. They were not given the permission to go outside and work. . . .” [VDC respon­ dent, Terai district, ODF project] Key Findings 23 Even when women were enthusiastic However, another respondent mentioned that having about participation, men controlled female members can be advantageous—since women’s lines are shorter in banks, the monetary transactions at the implementation processes banks can be completed faster. Some women who were included on RRNI teams were enthusiastic about participating, and knowledgeable about nutrition issues. Yet, their voices were not heard How gendered beneficiary selection affects and they were not involved in decision-making processes. perception of projects “I have a wish that by working with educated people I can learn The wards in the sample that chose “Increasing Animal something which would be good for me. I can keep records, for Protein” identified beneficiaries selectively (rather than example, of the materials that were used in construction. Fur- universally) to only include those households with preg- ther I raised my voice and said to the team, “Only you will do nant women and children under the age of two. One the whole work, let us do some of the work so that we can also effect of this was that these projects came to be viewed learn and do something for our locality . . .” [Female RRNI through a gender lens as villagers perceived the project team member, Terai district, ODF project] to be a “women’s project,” not one that affected all ward residents equally. “They [male coach and male RRNI team leader] behaved as if they didn’t want to tell us though we wanted to know more. We don’t know how much fund was received and Inclusion of women leaders does not result how much was spent. They themselves did everything. They in women’s greater voice purchased solar but we don’t even know how much they spent In another ward where the RRNI team was headed by a on it. To tell you you the truth, they have not provided us with woman, the team leader did not ensure that other female the information they had to . . . I withdrew 25 thousand and members were given equal responsibilities. Having a another female RRNI member withdrew 22 thousand, and female leader did not increase the chance that the other we just gave the money to Team Leader sir and Secretary sir.” women on the team got an opportunity for their voices to [Female RRNI team member, Hill district, animal be heard. protein project] “Personally, I felt that women were not given the chance to get involved in any programs like the men were.” [Ward Citizen C. Inclusion of Minorities Forum representative, Terai district, ODF project] Projects tried to compose RRNI teams that reflected the caste distribution in their villages and an effort was made to include lower castes. However, inclusion of Dalit and How gender affects implementation religious minorities was complicated. Minority groups processes were reluctant to participate. At the same time, majority Respondents mentioned that having female RRNI team groups did not attempt (or were not trained) to under- members sometimes sped up the implementation process, stand the deeper reasons why minorities might not want whereas in other cases, women’s circumstances reduced to engage with project personnel or powerful groups in vil- efficiency. A coach mentioned in an interview that wom- lages, and took as unproblematic the reasons that minor- en’s lack of access to their own means of transportation ity groups give for lack of participation. increased the work completion time. “The wards which had more female members took a little more Like the issue of the inclusion of women, when asked in time for the work completion. Females took more time since they interviews whether RRNI team composition was repre- had no vehicles to reach markets.” [Coach, Terai district, sentative of all castes and religions in the village, most clean water project] respondents across wards said that RRNI teams were diverse and there was equal representation of all. The one exception was the ward described in Box 3 where 24 Qualitative Study of “Sunaula Hazar Din” Community Action for Nutrition Project Nepal respondents gave clear indication that more caste groups “There are those who follow Christianity but it has not affected should have been included on the RRNI team. much. They have not said that their religion was not repre- sented.” [Coach, Hill district, handwashing project] However, when probed further, respondents made state- ments that contradicted their earlier assertions of inclu- sion. These statements were related to minority groups, Minority groups may distance themselves and the minority status was derived from a combination In selecting RRNI team members, several wards tried to of being a minority numerically, or because of their lower actively include minority groups, especially Dalits. They caste status, or belonging to a religion different from the explained to Dalits that projects required them to be majority Hindu religion. Instances of such exclusion did included. However, once minority groups refused to par- not predominate in the data; in fact, several Dalit group ticipate, the majority considered this to be unproblematic. members who participated in RRNI teams were active Respondents rationalized the refusal of minority groups participants. However, it is instructive to learn from the in terms of the minority groups’ lack of interest in the occasions where minority exclusion emerged in the data, project. No respondent thought of the issue as resulting and to unpack how respondents rationalized exclusion. from historical exclusion or that it required extra effort on the part of the majority group to help overcome the structural barriers to participation. Difficult to work with minority groups “We have B.K. [lower caste group] also here but they are In one ward in the Terai district, the coach explained that not interested to participate. They refused . . . even when we most RRNI team members were from the Mandal caste told them we have to include them in the team, they refused.” because 90% of the ward residents belonged to this caste. [RRNI team member, Hill district, handwashing The coach (somewhat improbably) asserted that “Even project] though there are Muslims from Ward Number 1, they live in Ward Number 2. There are no Muslims in Ward Number 1.” Conse- “There are 10–15 Dalit households. When we told them that quently, Muslims were excluded from the RRNI team. A we have to include them, they did not come to the meeting. It was more likely explanation for the exclusion was provided in during the month of Asad and Shrawan. They had a lot of an interview in the same ward with an RRNI team mem- work to do, they have their own problems . . . maybe those are the ber, who said that collaborative work is difficult when dif- reasons why they didn’t want to be included.” [RRNI team ferent groups are brought together. member, Hill district, handwashing project] “Only those people who follow Hindu religion were included. They said it would be difficult to work with people from other religions.” [RRNI team member, Terai district, ODF D. Leadership project] a.  Coaches Leadership emerged as critical to projects that performed better. SHD coaches with strong leadership skills, a clear Minority groups may be invisible idea of project objectives, and local knowledge on manag- Minority groups might not consider their exclusion to be ing government systems guided communities to overcome exceptional, nor are they actively and purposively included obstacles, and work together toward desired outcomes. A by the majority. For example, no other respondent men- case study of one ward in Terai district that benefitted tioned Christians in one ward, but the coach said that from the leadership of the coach, and was able to achieve while there were no Christians on the RRNI team, this results despite alleged corruption, resistance from the was not a problem because the minority group had not DDC and VDC officials, and budget delays is useful to expressed their lack of representation. The coach did not consider. think it was necessary to actively pursue their inclusion, and rationalized that since the group did not complain, there was no problem. Key Findings 25 It was necessary that coaches be trained well. A DDC by all members of family including men and women. The coach respondent highlighted the importance of training, and from his own initiation and collaboration with other NGO was how the training for coaches could be improved: successful in helping every household build toilets. Community members were well sensitized on the use of toilets and they made “The first thing is that the coach should have is local knowledge contributions and maintained the toilets. The community were about how to work in a VDC. . . . The coach should have happy with the project.” [Observation note] the knowledge about our financial sources and systems, and the working process . . . There were foreign trainers to train the coaches but our coaches here did not have ability to understand Work accomplished despite alleged what they said. Coaches went to the village and did their work corruption and resistance at DDC but I am still not sure the level at which they functioned. This and VDC levels and budget delays program could have been more effective but because of these types The coach’s work is especially noteworthy because results of factors it was not as effective as it could have been.” [DDC were achieved despite allegations of corrupt practice at respondent] the DDC and VDC, and objections raised by the DDC on the selected focus area. Case study of a ward that benefitted “Both DDC and VDC used to say that they will do our work from the leadership of the coach only if we give them some money.” [RRNI team member] The coach of one ward in our sample was universally praised and credited with what the ward achieved in their The coach played a critical leadership role in all the com- ODF project.13 A range of respondents—from RRNI ponents that enhanced project performance: developing a team members to ward representatives to beneficiaries— clear work plan, ensuring community buy-in, placing an had positive reviews of the project, and were satisfied with emphasis on awareness raising in addition to the construc- what had been accomplished. A WCF respondent said: tion of toilets, drawing up a clearly defined beneficiary list “We did what our coach guided us to do. Without our coach, that was communicated in a transparent process to ward nothing could have been accomplished. We can’t really take the residents, mobilizing the citizenry who participated in the credit as the coach told us what to do and what not to do.” project, and involving VDC officials. Each is discussed [WCF respondent] below. Similarly, focus group participants also praised the coach: Work plan “He was very good . . . he never scolded anyone . . . but always The coach led the RRNI team in developing a strong checked on things in a timely fashion . . . Everyone in the ward is work plan, which involved constructing toilets in 30 house- happy. They praise the coach for the work he did . . .” [Female holds. The remaining households were provided with FG participants] some materials and/or were asked to pay for the received materials (e.g., one FG respondent said that she had to pay The research team expressed the same opinions in their NPR 2,500 for the seat). The overall goal was to ensure observation note: that every household had a toilet so that the ward could be declared an “Open Defecation Free” zone. “The coach appeared to be the role model for the district. His efforts were praised by even DDC officials and the community people. We could see neat and clean toilets which are being used SHD project sought community buy-in When the project was introduced into the community, the coach ensured that the prescribed steps were followed, 13 This is not to suggest that this was the only ward in our sample where respon- i.e., going through the WCF, involving the DDC and dents were satisfied with the work of the coach. We have selected to highlight VDC officials, and forming an RRNI team that was rep- this particular ward because the coach was universally praised across all data resentative. In both in-depth interviews and focus groups, points, and focusing on one case allows us to provide a more in-depth analysis. 26 Qualitative Study of “Sunaula Hazar Din” Community Action for Nutrition Project Nepal respondents opined that the coach ensured that the RRNI thousand people, then we cannot distribute to all. If there is team members included different castes, occupations, Rs 100 and 100 people, then we will distribute one rupee to education levels and genders, and the overall perception each person, and it will be finished after buying just one choco- was that the team represented the needs of the commu- late. So let us give more to those who are needy.” [Coach] nity. The one exception was that the women on the RRNI team did not feel included in the decision making, and At the end, there was overall satisfaction that the project expressed dissatisfaction at not being given any responsi- targeted the neediest: bilities, even though they had the interest and were part of the implementing team. “Only the needy ones were given [materials and labor for toi- let construction] . . . the poor were not discriminated against . . . we felt this was good . . .” [Female FG participant] Awareness raising emphasized When asked about the criteria for the selection of RRNI team members, respondents said that they selected those Citizen effort who were active and would not hesitate to go door to door Ward members were mobilized to build toilets, and took to spread word about the project, and raise awareness on pride in their hard work. toilet use. The coach suggested the use of street plays to “We did a lot of hard work. We made the cement paste with change people’s perceptions on open defecation and its our own bare hands and built the toilet. Everyone got their hands relationship to ill health. injured. But we were successful.” [RRNI team member] “. . . the road drama regarding this project was very much effec- tive as it immediately changed the thoughts of the people and There was recognition that the SHD kick started a larger everybody wholeheartedly took part in the project.” [RRNI movement among citizens to build toilets in the ward, team member] under the encouragement of the coach. “Government started the process and citizens made it possible. Beneficiaries clearly defined If the government hadn’t put effort, and if the coach was not Beneficiary selection emerged as a problematic issue in there to encourage, it was not possible for citizens to make it pos- the data from several wards, where community members sible.” [FG participant] expressed dissatisfaction for not being selected to receive project benefits, or not being aware of why they were not Two other factors, besides the coach’s leadership efforts, selected. In the case of this ward, there was initial con- were significant in successful implementation. First, the fusion on how to distribute resources, as revealed in an coach was a resident of the ward, and second, the VDC interview with an RRNI team member: played an active role in promoting the project’s completion. “There was confusion on how to provide funds as there were the people with high, moderate and low economic status.” Coach is a resident of this ward The coach is a 32-year old male, who resides in this ward. However, the issue was satisfactorily resolved through dia- This is important because the coach can dedicate time to logue and transparency. the project instead of spending time traveling for work. “The Yadav community is more here. Then there are Telis, fol- lowed by Mahattos and Kalwaar. The Yadavs said they should Incentives to complete toilet construction also get [project benefits]. The other community people said work by engaging VDC officials that they are only 1–2 households of their community, so they The VDC encouraged the different wards and RRNI should also get. We said, let’s all sit together and decide to give teams to complete the work by announcing that the fastest to those who are more poor. If there is Rs 100 and there are team/ward would receive a reward of Rs 5,000. Key Findings 27 “We had a scheme for the team that would finish the work the “. . . in our Terai, the ones who have toilets are considered as fastest, which was won by Ward No. 4’s team. In my opinion people with prestige . . . people envy those villages with toilets too they were the most active group. They finished the work . . . so this has caused improvement in marriage . . . the girls of ahead of time and were awarded Rs 5,000. This also moti- Saptari and Siraha [outer Terai districts] are getting mar- vated other teams to work faster.” [VDC respondent] ried in our village too . . . this is a good achievement . . .” [Male FG participants] A strong coach, together with an active VDC, a dedicated RRNI team, and a mobilized citizenry, ensured that there Changes will continue were both visible and unexpected changes in the village. Respondents were confident that these changes were sus- tainable since the construction of toilets went together Changes seen in cleaner environment with behavior change. Once habits change, or new habits are formed among the young, they are likely to be sus- and health outcomes tained. This example highlights how a coach’s leadership Respondents spoke about visible changes in their village— can guide RRNI teams to work together, help communi- less open defecation leading to better health and a cleaner ties understand and accept the project’s objectives, and environment. One respondent attributed his understand- bring change. ing of the link between proper toilet use and health to the SHD project: “Before this program was put forward to us, we didn’t know b.  Rapid Results for Nutrition that the construction of proper toilets and use of soap and water Initiative Team Leaders Projects benefitted when the selected RRNI Team Leader for cleaning hands could prevent the occurrence of diseases.” displayed leadership skills, in some cases working along- [WCF] side a skilled coach, and at other times, compensating for the absence of a strong coach. Unexpected change: the fact that toilets could be constructed at all Careful choice and work plans made by the The unexpected change mentioned most often was the wonderment that any change occurred at all. The coach RRNI leadership to optimize the outcome Projects benefitted when RRNI team leaders considered himself seemed amazed that such a project could accom- the various options thoughtfully, and guided their teams to plish results in the Terai belt, presumably because accom- desired outcomes. Let us consider the case of a ward that plishing community-driven projects, which requires selected handwashing as their focus area. Projects that significant participation and commitment among local selected handwashing opted either to distribute drums, groups, is particularly hard in the plains region in Nepal. or distribute buckets and soaps. An RRNI team leader “We had never thought that this VDC—which is in Madesh— explained why a decision was made to distribute 200-litre would ever be free from open defecation but it stood to be the drums (200-litre drums have twenty times the capacity of second best VDC in the district. This was the most unexpected the average household bucket). He argued that drums had change.” [Coach] more capacity and could be covered, which increased the likelihood that they would be used for the intended pur- pose and would create sustainable handwashing practices. Better marriage prospects because having toilets is a prestige issue “Initially we decided to distribute buckets for washing hands but Another unexpected change mentioned by male FGD we realized that there would not be any covering for the bucket. participants was an improvement in marriage prospects And it would be very dirty to use that water. This distribution of the residents of this VDC, because villages with toilet of bucket was cancelled. There are many family members in a facilities have greater prestige. house, on average 6 people living together. Then the water in the bucket would not be sufficient for them. 12 to 15 buckets will 28 Qualitative Study of “Sunaula Hazar Din” Community Action for Nutrition Project Nepal also not be sufficient. Carrying buckets repeatedly for each use on the singular shoulders of the team leader (and the team may be difficult, people may discontinue using it, and they may leader was a woman). again follow their previous bad habits and not maintain per- “During the meeting, everyone agreed to take turns to look after sonal hygiene. We thought that drum distribution would be the the chickens. But it didn’t go as per the discussion. I used to look best way for the hand washing project since it can hold enough after the chickens, feed them, clean their utensils, give medicine, water. We can fit the tap in the end and can be used for many manage their shed and even bring feed from the market. I had to days. We get covers for the drum also which will keep the water do it all alone. Nobody from team came for help. They used to clean and safe and can be used numerous times. If the water is participate only during the time of eggs distribution. [RRNI made safer for use by using covers, it can be used in the future team leader, Hill region, animal protein] for personal use, even for drinking if scarcity occurs. If the area near the drum is kept clean, then it will make the environment clean and there will be no spread of disease.” The coach also commented on the work burden of the team leader: Furthermore, before the distribution of drums, villagers “Once the feed was bought, it was delivered to a central point in were given training on proper handwashing by RRNI the vehicle. From there, she [RRNI team leader in whose team members, and when drums were distributed, people land the hen coop was built] had to carry it herself [on were told about their proper use. her back]. An insect called khajuro bit her so much on the way that she had to take bed rest for days. She worked very hard “If we had told them SHD is a program which has come with for the program. I felt guilty sometimes seeing what all she has an aim of distributing drums, they would not have used it prop- done for the village and the problems she faced.” [Coach, Hill erly. They may have used for other purposes and our program region, animal protein] would not have been successful. We personally went to every house and told them that drums are for washing hands not for other purposes. We made them understand that we had to fill While free-ride could occur in any team, this particular water in the drum, keep soap beside it and wash hands with case with the female leader also indicated a risk of the clean water in a proper manner. Now if you go and observe, pattern of the “second shift” 14 that gender scholars write outside every house you will see a blue drum.” about. When the project was perceived as a women’s proj­ ect, and when women volunteer to work on development project particularly on leadership roles, it means that Residents were also given half a sack of cement to con- women would enter the “workforce” (which is not per- struct a platform near the drum to place washed utensils ceived as their “traditional” role) and, in effect, work two on, since residents did not practice hygienic ways of stor- shifts: one at work and the second at home. ing utensils. Risk of “free-riding” by other members while E. LOCAL GOVERNMENT the RRNI leader takes on a greater share Local government officials played a crucial role in SHD of the responsibility projects. From the perspective of the community, the chief In one Hill ward they selected Animal Protein as their complaint was budget delays, which sometimes resulted in focus area, and decided to grow chickens themselves to RRNI team members having to provide personal funds distribute eggs to beneficiaries. The team decided that for project expenses. Respondents also mentioned cases the chicken coop would be placed in the team leader’s of corrupt officials. When local officials were positively house and other RRNI members would assist and take involved in project implementation, the community turns in caring for the hens. However, the members did not contribute their share of work, and the bulk of the 14 Hochschild, Arlie, and Anne Machung. The second shift: Working families and the labor- and time-intensive job of rearing the chickens fell revolution at home. Penguin, 2012. Key Findings 29 expressed satisfaction and the projects moved forward DDC but I don’t think anyone would take action. I have also well. From the perspective of the local officials, they were shown my dissatisfaction to them and threatened to leave the job not provided with sufficient orientation on SHD projects, but it’s my loss at the end of the day. I am having a tough time and community members were not trained well in how to as a coach.” [Coach, Terai region, ODF] work within existing government systems. One respondent laid the blame on the unsuccessful com- Budget delays are the chief complaint pletion of the SHD project solely on the shoulders of the Budget delays resulted in the need for RRNI team mem- local government officials, claiming that the wards had bers to provide personal funds for project expenses, in the done everything per the plan but the district officials in project ending sooner than originally planned, in raising remote regions of the Hills did not work for the welfare suspicions among community members that the project of the citizens. was going to cheat people, in spending the majority of “The leaders of the district level have no intentions to work implementation time period on awareness raising, and for the benefits of the people living in remote areas like this. in dampening the motivation of coaches to work on the The fault was of the DDC and VDC for the program being project’s behalf. unsuccessful. The WCF formed RRNI teams. The teams have “The main challenge for the program was delay in release of collected data on the needs of the community and prepared and budget. The RRNI members were forced to buy feeds for the hens presented the proposals in the DDC for addressing those needs on their own expense.” [WCF respondent, Hill region, but the files have been pending since one and half years. I have animal protein] felt that the programs have been delayed because of non-response of the VDC and DDC to the proposals presented for addressing “They passed only 30% of the budget in the first phase. I the identified problems in the community . . . There is a chal- provided the project a loan of Rs 40,000.” [RRNI team lenge in interacting with the LDO, focal person and even the leader, Hill region, animal protein] senior officials of the VDC. We present the proposal and present it to coach who again presents it to the VDC. There is a certain “Due to insufficient budget, they were unable to feed the hens time interval in which the VDC sends it to the district. As the and they finished the project in 60–65 days only.” [VDC offi- VDC increases this interval and doesn’t submit our proposal on cial, Hill region, animal protein] time, the response gets delayed. It’s difficult for the coach too . . . “During the first 50 days, only 20% budget was provided I don’t find the RRNI team responsible for all the delays . . . We but expenditure was 80%. It was difficult to convince fami- built cages; we raised hens; yet, this could not be successful . . . lies that toilet construction would happen . . . Even though the Our RRNI team informed me time and again that they were DDC and VDC supported us in finding a trainer, estimating facing difficulty in raising the hens and asked why the budget the budget, and buying materials, no one could support with the had been delayed; why they could not operate bank account from timely release of the budget. I can understand the system issues the banks that were accessible to them. They had mentioned that with budget release but there are delays at the district level itself it would have been easier for them to work if these had been for no obvious reason. This creates difficulty in working with done.” [WCF, Hill region, animal protein] the community. They start to think that these are the kinds of programs that are notorious in cheating people and building false expectations.” [Coach, Terai region, ODF] Local government officials demand bribes to move projects forward “If salary is not provided in time, the enthusiasm is also lost. Respondents shared that local officials can be corrupt and We can work under pressure but would not come from our heart. refuse to work unless they are paid bribes. Even if I receive salary every quarter that would be good for me. I am tired of going to DDC for NPR. 10,000. NPR. 10,000 “Sometimes we need to feed them with money. You know, bribe! is not sufficient even for the bus fare. I need to go, work, eat, Yes, we need to bribe them.” [RRNI team member, Hill sleep and return. I need to go to the VDC 5 times in a month . . region, handwashing] . I did not receive salary for last 10 month. I have informed the 30 Qualitative Study of “Sunaula Hazar Din” Community Action for Nutrition Project Nepal “. . . we had to go and take the signatures of the government “There should be more monitoring by local government . . .” officials and we had to visit again and again . . . Even if we [FG participants, Hill region, handwashing] could meet, they had to be bribed . . . it was hard . . . [RRNI “We didn’t see the involvement of DDC officials at all. As I team member, Hill region, animal protein] said earlier DDC should come for supervision. All the materials “Both DDC and VDC used to say that they will do our work come through DDC and if the DDC does not work properly, we only if we give them some money.” [RRNI team member, don’t get anything. The responsibility of training has been given Terai region, ODF] to District Public Health Officer (DPHO) but people from district public health department never come. Since the Public “We didn’t agree to provide bribe to the officials. But, if we did Health officials showed no interest, we had to ask the health not give them, they had said that they would not do our work. So, post workers to provide the training. We got the notice about the as we had to keep our work in mind ahead, so we somehow had training to be given by the district public health officials and the to bribe them . . . I do not remember the amount given . . . That training was for women. We gathered them for training next day is known by our brother . . . he had said that it is not a big sum. no one came from the district. They did not send anyone. We had As the work was completed, so I did not want to ask on the issue to plead people from the health post to come. There should be . . .” [RRNI team member, Hill region, handwashing] more active involvement of the DDC, after seeing the activeness of district and district development, district public health should DDC and VDC officials on their part complained that be able to maintain itself.” [WCF, Terai region, indoor the RRNI teams were not well trained on how to work smoke] within the existing government financial systems. For example, if a certain amount was deducted at source as One respondent explained that the VDC Secretary lived part of a government rule, RRNI teams and ward resi- in the district center, which was at a distance from the dents thought that the amount was being deducted due ward, and this created coordination problems and a to corruption. resultant lack of involvement and delays. “They don’t have knowledge about how to arrange the bills, they “The Food and Nutrition Committee must decide on the budget don’t know about PAN and VAT. They don’t know what bill proposal of the RRNI team. Their weakness is being unable to is used and in which bill how much is deducted. They think the arrange meetings on time. The secretary of the VDC itself is the secretary might take benefits [bribes]. They don’t know about coordinator of the nutrition committee. The VDC secretary lives the system. It’s very difficult to work.” [VDC Secretary, Hill in the [District HQ] and the other members of the committee region, animal protein] live here in the village. That is why the process is taking so long . . .” [WCF, Hill region, animal protein] Not enough involvement by local government The most consistent recommendation offered by respon- When local government officials were actively involved dents of all categories was that local government officials and conducted monitoring visits, ward residents involved at the VDC and DDC levels should be more involved in in the project were very appreciative. Note that three projects, especially to monitor the work of projects. RRNI team members from the same VDC appreciated “VDC representative didn’t come to monitor the work. If they the help received from the VDC official. had come to monitor, it would have been easier for me.” [RRNI “The secretary from our VDC helped us very much. He helped team leader, Hill region, animal protein] us to contact the place where we could get the drum. He man- aged the people to bring the drums to the village and supported “. . . there was not much contribution from the VDC level. To us in every problem. He advised us about the management and my knowledge, the VDC has never sent anyone for monitoring implementation of the projects. Due to his help, we were able to . . . DDC level respondents have visited, but not the VDC.” perform well for the first time otherwise it would have been tough [WCF, Hill region, animal protein] for us.” [RRNI team leader, Hill region, handwashing] Key Findings 31 “The Secretary of the VDC went himself to every house of “The rates that the district had for transportation of goods was ward number 1. That was so very useful to us. The political not sufficient for our Hills. It costs more to bring materials to our leaders also went along with them to the whole VDC. After see- remote ward, but the DDC would not allow it.” [RRNI team ing the work, they corrected the wrong ones and encouraged the member, Hill region, handwashing] people. They taught us to do the work in this way and that way. They improved our working process.” [RRNI team leader, Hill region, handwashing] Not sufficient orientation provided to local government on SHD projects “The people from VDC used to monitor. And we were Local government officials were not clear on project happy about that.” [RRNI team member, Hill region, objectives, and sometimes paid more attention to reach- handwashing] ing the end goal without paying attention to the process. “In the beginning of the project DDC was not clear and the Interested and involved VDC officials had a good under- VDC was also not clear about how the project should be done. standing, and realistic expectations of what the SHD proj­ We were not informed about how we should run the project so ect could accomplish and how it could push the VDC’s we were only focused to finish the project as soon as possible development agenda forward. For one VDC respondent, and there was no prior planning. If anything needed to be done the SHD was akin to a “booster shot” for the commu- urgently there used to be difficulties.” [VDC respondent, nity. The project raised awareness on health and gave an Hill region, animal protein] impetus to the VDC to implement actionable tasks with the available funds. However, they felt the project could not go on forever, and in this sense, it was not sustainable. However, by being given a kick start, the project could F.  Challenges and have long-lasting impact. Opportunities in “The program put forward by Sunaula Hazaar Din is a great Implementation program which has gained a lot of success and has provided by Focus Area opportunity for the village people to learn something new. Some The data revealed both opportunities and challenges people had criticized the concept as they think that for the village faced by RRNI teams that were specific to the focus area to be economically well and healthy, there must always be con- they selected. There are important lessons to learn from struction work going on. Some call it throwing sand into a lake. the examples below for future implementation efforts, But no, this isn’t something that the program will keep on doing particularly to improve intended nutritional behaviors or for years but something to start a better future. The initiatives outcomes by involving communities. put forward by Sunaula Hazaar Din may not be a sustainable one, it isn’t supposed to be. It was just a step to make people aware about their health and take control over it. The program Consumption of animal protein The challenges faced in interventions to increase con- has gained a huge success . . . [VDC respondent, Terai sumption of animal protein ranged from effective ben- region, ODF project] eficiary targeting to technical capability constraints in carrying out activities as described in detail below: Local government policies are inconvenient Some local government policies were obstacles to the Difficult to ensure that only target implementation. populations are beneficiaries “There were challenges due to the policy of the DDC and VDC. Even though the eggs were distributed to women who fell They kept on calling us to [District HQ] to open the bank in the category of “1,000-day mothers,” sometimes the account and we kept on following them though it took us three receiver of the eggs shared with other household mem- days!” [WCF respondent, Hill region, animal protein] bers who did not fall into the target population. For exam- ple, in one ward the targeted women were given a weekly 32 Qualitative Study of “Sunaula Hazar Din” Community Action for Nutrition Project Nepal quota of eggs so that they could consume one a day, but as that. It was felt that training on techniques of poultry farming the quote below shows, the entire household ate the eggs, was required.” [RRNI team member] and the weekly quota was all consumed in one day. “It was difficult while constructing the hencoops. There was a “In some of the cases, the whole family prepared the eggs and plan of using tin. The training was given before the building consumed them together finishing all on the same day.” [RRNI of the hen coops. We were not told that tin is not good. The tin team member] produces too much heat to the hens and so it was decided that we should buy wood.” [RRNI team member] Or, ward residents would fudge their pregnancy status to receive the benefits. The above mentioned challenges point to the need (a) to “Some people who were not pregnant also claimed that they consider household and community-level dynamics in ben- were.” [RRNI team member] efit sharing and (b) to provide capacity-building programs to carry out interventions. Aside from the challenges, the study found some success in sustaining behavior change Residents who were just above the cut-off mark to be a and continued adoption and expansion of poultry farm- target beneficiary resisted their exclusion. ing in communities as described below: “The women who had children who had crossed two years . . . they also asked for eggs which created problems.” [RRNI team member] Behavior change in communities While respondents stated that behavior change of egg consumption was more likely among financially better- When the excluded protested, one RRNI team felt that off households, nevertheless, the program successfully they had no option but to give eggs to them. induced change among some individuals. “We estimated 75 beneficiaries but during distribution, there “Around 18–20% of the households out of 65 [beneficiary were more than 100. After that it became nearly double. Those households] have started raising hens, consuming eggs and who were left out from the beneficiary demanded for the eggs feeding the children. They have developed a clear understanding from the beginning as their name was not in the list. Team mem- that they can also gain benefits from the hens they rear them- bers told that the eggs were already distributed and they wouldn’t selves. They have been encouraged because they have witnessed get it . . . Every time new people were added. There was no other the changes themselves . . .” [RRNI team member, Hill option than including their names as they were from our own district, animal protein] community.” [RRNI team members] “I have seen a daughter-in-law near my house . . . she used to feed eggs to her child during the implementation of the SHD No training given on how to raise chickens program. Later, even after this program got completed, I have Wards that decided to build hen coops and raise chick- seen her buying eggs from the mobile market.” [RRNI team ens themselves said that they would have benefitted from member, Hill district, animal protein] prior training on the procedures. Lack of training resulted in the hens dying or not giving sufficient eggs. Increase in poultry farming in the community “Not all, but one or two of the hens died as well. We had not “I have already given an example of such change. I have seen received training regarding poultry farming so we had no idea it in one of the houses downhill. He has started rearing 16 or about the way and amount of feed the hens had to be given and 17 chickens. The pregnant woman in that house will deliver the how they had to be raised. We asked the coach. He also didn’t baby after three or four months. He started this being inspired by have any idea. He had not raised any chickens before, nor had this program. This way, the changes have been there. I had not he received training. We asked the people who were experienced expected these changes would be there. They will give continu- in poultry farming, we also learned that the hens should also be ity to the changes because other people will get inspired simi- given vitamins from time to time. We had no information about larly. The people will start poultry farming for their benefits. Key Findings 33 Other people will buy from them knowing that its consumption “We forced everyone to build toilets. For those who didn’t fol- is necessary.” [WCF representative, Hill district, animal low our suggestion, we took help from the police administration. protein] The illiterate people used to fight. They were not ready to build toilets. They were not ready to understand us specially some old aged women. Therefore, we had to call for help from the police Ending Open Defecation administration. In our ward, Ward 4, people were stable so it The study found that a campaign approach of declaring a was not difficult. Other wards had a bigger problem.” [RRNI village an open defecation free zone seemed to support the team member] behavior change, which reinforced or sustained the com- munity’s pride in staying clean. Some communities, inter- “They [ward residents] were threatened. If they did not estingly, took drastic measures, such as policing, enforcing build the toilets, their children would not be taught to read and sanctions, denying government services and certificates, write, they would not be able to complete their VDC works and and even fining individuals and households that did not they would not be able to go foreign countries to work. After comply. threatening, they built the toilets with their own money. We helped only the poor ones.” [RRNI team member] Behavior change in communities Two female focus group respondents corroborated these “Yes, there have been lots of changes after the building of toilets findings: . . . the main change is in the habits . . . we feel ashamed to defecate in the open . . .” [Female focus group participant, Person 4: “Now, we have toilet license, so our works through Terai district, ODF] VDC is easily done.” Person 2: “They wouldn’t even make Birth Certificates for our Pride in community due to improved children if we hadn’t built toilets.” cleanliness “When I went to another village that was a little bit dirty, I In another ward, village residents were told that if they described my village and invited them to visit our village and built toilets, they would receive other benefits, such as get- learn from it.” [Male focus group participant, Ward 4] ting a citizenship card or a passport. “They told us that they would help us in making citizenship and passport. [FG participant] Sanctions against villagers to ensure toilet construction and toilet use Wards that chose ODF wanted the SHD project to make Fines of NPR 5,000 were imposed for not using toilets. the entire VDC “open defecation free.” Project funds were Villagers were told that children of ‘offenders’ would be utilized to build toilets for some households, but the funds barred from getting admission into school. Those who were not sufficient to cover everybody. Those not covered had completed the toilet work in their homes were given by project funds were required to construct on their own, a ‘red card’. Without the red card, villagers were unable or were given some materials but had to provide the rest to complete any official work, e.g., birth registration, mar- on their own. When ward residents did not agree to build riage registration, farm registration, etc. Loudspeakers toilets, more extreme measures were taken. were used to broadcast messages to encourage toilet use, but also to let people know what would happen if they In one ward, the police were called in, and residents were failed to comply. The police were recruited into the effort threatened with other sanctions, such as being barred as well: from completing any work that needed VDC approval, or “We told the police that if someone goes outside of the house [to not being allowed to send their family members to work defecate], then catch him and take him. There should be a fine overseas. of Rs 5,000 . . . take him to the prison and beat him so that 34 Qualitative Study of “Sunaula Hazar Din” Community Action for Nutrition Project Nepal no one goes outside due to fear and threat of arrest. Now what Showing the promise of development more can we do than this? We cannot take the stick into our own programs hands...” [RRNI team member] In the wards that successfully completed the handwashing project, the projects boosted the promise of development Handwashing and Using programs in general. Clean Water “. . . no one thought that they would distribute drums . . . We The study found that interventions in handwashing and did not have hope that these programs would do anything. But using clean water helped promote behavior change in when we saw them come, entering each village, entering each communities successfully, which resulted in improved ward, entering each house, we felt, ‘yes, this organization will perception about development programs and an over- do something . . . They did it. That was unexpected.” [Female all increase in the capability of community members focus group participants, Hill region, handwashing] engaged in implementation. Increased community management skills; Behavior change in communities increased community-driven demands “Earlier, before this hand washing project, if we washed our hands with soap, our elders would say, “what a show off ”! for health, not just for physical Now everybody understands the importance.” [Female FG infrastructure projects participant, Hill district, handwashing project] “. . . other than changes seen immediately after the project of improved handwashing, using toilets, what I have observed is that the capacity to manage the program has been increased in Behavior change in communities communities. Secondly, there is effectiveness in the demands . . . for using clean water not only drainage, water and roads are the demands . . . the “Everyone was happy to get filters. There is an increased aware- things that are required in our daily social life and in health, ness in drinking clean water. Even if filter breaks, they will have also come under demands or things that should also be go and buy new ones rather than drink dirty water.” [RRNI demanded  .  .   ” [DDC representative, Hill region, team member, Terai district, clean water] handwashing] Key Findings 35 Chapter five Emerging Lessons and Conclusion While good practices and outcomes were observed in some wards, our data overall revealed multiple challenges that the SHD participating wards faced. It provides a good base to understand whether the 100-day rapid results approach was effective in the given context, and whether the community-driven model was adequately sup- ported by the project to achieve its development objective. This section summarizes emerging lessons learned from the data analysis, and offers concluding remarks. Summary of emerging lessons from the key findings Selection of community-based team members. When the initial ward citizen’s forum failed to gather a large number of participants, or provide sufficient orientation on the project in an inclusive and transparent manner, a small group of individuals dominated the decision-making step, and the subsequent procedures were also con- trolled only by them. When such a capture happened at the beginning of the cycle, it was hard to be reversed, and other community members did not have a sense of own- ership or being included until the end. For the community’s decision-making process to be effective, it is critical to have a clear set of rules and requirements, and to widely and clearly communicate them up front. For example, the minimum number of par- ticipants, or clear voting procedures that must be followed—and the decision would be considered invalid unless those conditions were met—would be useful. In some cases, although not fully “community driven,” resource allocation and results of the RRNI project were positively received by the community when a small group of members developed a clear work plan and budget, and implemented the subproject accordingly. Forming the RRNI team. Requiring nine members to form the project-specific RRNI team which would be dissolved after each 100-day cycle was a serious chal- lenge for the ward members, especially in areas with a spread-out population where ward members were already overstretched due to other such programs and their daily work. In many sample wards, members were “selected” without even being informed (hence without agreement), resulting in confusion and inefficiency. It indicates that in the designing phase, the SHD project did not sufficiently assess the feasibility and effectiveness of the RRNI team model in the target areas. A mapping exercise of Emerging Lessons and Conclusion 37 existing local platforms and community groups, as well as providing such training would logistically need more time consultations with target populations, would be helpful to and investment. confirm the design efficiency, and even when implement- ing the same activity, to identify stronger synergies and Targeting or broader benefit sharing. One of the partnerships to be built around the RRNI team. most contentious and conflict-prone steps in SHD was how beneficiaries were selected. Decisions on who should Selection and implementation of a community- benefit was left to the community without clear guidance driven subproject (focus area and goals).15 The from the project, which caused confusion and doubts in selection of focus areas and goals were naturally affected different stakeholders—rather than allowing positive by a range of factors, such as budget, cultural norms, “flexibility”—and made the process receptive to capture geographical constraints, expected visibility of outputs and to the influence by local governments who some- in 100 days, and targeted or universal benefit sharing. In times had other incentives. While the SHD’s main target this decision-making process, however, the local stake- beneficiaries were the golden 1,000-day women, many holders seemed to be consistent in reporting that the of the sample RRNI teams selected activities to benefit SHD did not provide much customized facilitation for the entire community in order to avoid disputes. It was the geographically and culturally diverse communities it even more challenging to ensure that the target women covered. While some communities demonstrated capac- benefit from the project, when decisions and implementa- ity to improve their plans and activities to better fit their tion were controlled by male members. Frustration was need in the course of the RRNI cycle, lack of enough exacerbated when transparency or information sharing provision of customized guidance seemed to create chal- was missing regarding how the team reached the final lenges to certain communities in optimizing their demand decision on beneficiaries. The data from this study cannot (need) and supply (intervention). In SHD, community- conclude whether the targeted or universal approach was level needs and assets diagnostics were to be done by the better. The data is conclusive, however, that regardless of RRNI team through facilitative support, but coaches strategy used, developing a clear work plan and manag- were not sufficiently trained either. The limitation with ing the budget and the community’s expectations are both technical knowledge was also observed during the imple- critical in successful project performance. mentation. For example, an RRNI team struggled with how to raise hens and some hens died, as no one around Inclusion of women. While every ward in our sam- them including the coach had relevant experience. For ple included women on RRNI teams as required by the ­community-driven investments to be efficient and effective project, their participation was only in form, rather than in achieving intended goals, and even more so to achieve in substance. Reasons ranged from cultural norms, time local ownership and sustainability of knowledge, attitude, constraint with the household chores, to threats felt and and practices in the long run, continued and customized resistance by men. When forced to be on the team, women training and capacity building as well as support from were either uninterested, or their participation was not experts are essential. Therefore, facilitators need to be actively sought. Even when women showed an inclination trained and encouraged to connect the community with to actively participate, men dominated or controlled the relevant local resource people. SHD’s 100-day timeframe decision-making and implementation processes. When did not have the scope to provide extensive training or women did take a leading role, working for the project capacity building—especially in the remote areas where also meant that they were adding extra work and respon- sibility on their shoulders because the members (espe- cially male) still assumed that women would keep their 15 According to the midline impact evaluation by DIME, 92% of the households full responsibility at home. If the project wants to ensure interviewed believed that the focus area implemented in their community was that the voice of the target 1,000-day women would be the best choice for their community, 76% of households reported they have directly benefitted from the project, and 90% reported that their community meaningfully reflected in the processes and outcomes of has benefitted from the project. However, it is worth looking in depth whether the project, it may be worth considering to form or work the communities’ decision on the focus area and the goal was an optimal choice with an existing women-only small group or have separate to address malnutrition of the target populations. 38 Qualitative Study of “Sunaula Hazar Din” Community Action for Nutrition Project Nepal meetings to identify the bottlenecks and ways to resolve in this area. The coach was spending more time for the them, while strengthening awareness raising and support project activities (he was resident of the ward), and was from a wider community (both men and women) includ- also able to effectively involve the local authority. As the ing through other existing platforms or groups. voices of several coaches indicated, the implementation arrangement needs to be built on diverse geographical Inclusion of minorities. Often, though not always, factors, including the number of households in a ward, minority groups were reluctant to participate, while because those factors are also notably affecting the project majority groups did not attempt or were not trained to outcomes. The 100-day timeframe of the RRNI model understand the deeper reasons why minorities might not could not keep pace with and invest in the need for cus- want to engage with project personnel or RRNI pro- tomized and continued capacity building. Rewarding cesses. There is, however, an opportunity for improv- good performers would also be an effective way to keep ing people’s awareness on why inclusion of minorities motivating these individuals, and to positively stimulate is important. There are cases, as observed in a sample others to improve. ward, where community members regretted that they could not include people from a certain caste in the initial Conflict resolution. The sample wards show that discussion. More opportunities for inclusion of diverse the community groups do have capabilities to address group members may have taken place if the project had local-level issues, though it takes time and good leader- initially set up and implemented clearer guidelines and ship. When a community was able to identify and address mechanisms. For sustainability, it would be useful to more challenges during the implementation phase, subproject actively involve different levels of the local authority in outcomes turned out to be positive, and sometimes even related data gathering and awareness raising. In collabo- created positive externalities. For example, in a ward ration, the project can provide incentives such as reward- where better-off community members had initially been ing a community that was most inclusive in each VDC. unhappy that only the poor members were given support If there are traditionally excluded groups, or, if the group to construct a latrine, after having been explained by the members are not used to speaking up at a general meet- RRNI team about the rationale, these better-off members ing, separate meetings can be organized to ensure their started to construct latrines with their own money. Over- voices are heard. In either case, initial inclusion of these all, it appears that conflict resolution was left to each com- minority groups takes time, and the project needs to be munity’s capacity and experience, without a clear policy ready for that by design. at the project level. Traditional mechanisms and platforms to resolve local conflicts—such as the WCF and in some Leadership and effective facilitation. A case study cases engagement of community leaders and political of the coach who demonstrated universally praised lead- party representatives, along with VDC and DDC officials, ership in a usually socially fragmented Terai ward was should be incorporated in the project design as much as probably an ideal scenario the SHD project envisioned. possible. Such existing complaint handling mechanisms It proved that good leadership was able to accomplish the can be examined up front, and relevant information and following activities: clear orientation and communication awareness such as on how to report complaints need to be of the project to the entire community; expectation man- communicated across the entire community. agement and complaints handling in a timely manner; transparent communication regarding the definition of Local government. The primary bottlenecks faced by target beneficiaries; and locally geared awareness raising the SHD communities in terms of the implementation activities such as road dramas. As a consequence, commu- flow were at proposal approval and budget disbursement nity members collaborated well to accomplish the proj­ phases by the DDC officials. Systemic challenges like these ect ahead of time with a strong sense of ownership, and increased frustration and subdued the initial momentum was awarded by the VDC. It confirmed the importance of community groups to achieve their goals. In addition to of building skills and capacity of facilitators and commu- the local government’s workload and willingness to be sup- nity members, and the SHD could have invested more portive, the data indicated that the SHD was not designed Emerging Lessons and Conclusion 39 to build a systematic, meaningful partnership with the approval processes, the Rapid Results 100-day model in local government (both VDC and DDC levels) through SHD has been overall implemented in isolation without its implementation. Local governments complained that investing in strengthening capacity of and linkages with they felt bypassed and the communities were not trained other existing platforms, including the local government on basic procedures (e.g., financial management), causing and technical experts. Relying almost exclusively on troubles with their government requirement. On the other available capacity and willingness of the community or hand, communities were frustrated or disappointed that coaches to take the activities forward made it challeng- local governments were not engaged, rarely showed up, ing to optimize the community-driven investment. It is or rather hindered implementation. Local development also worth noting, however, that there were still promis- projects that require close collaboration between local ing cases and practices observed at the community level government and communities have both a risk of severely that were able to actually improve knowledge and prac- damaging, or an opportunity of significantly improving, tice of the target population. Good practices were born the trust and relationship with each other. Securing local in an environment where there were strong buy-in, trans- government’s buy-in and active participation from the parency, and well-developed plans, combined with local onset of the project by providing extensive orientation leadership. These findings suggest a need for practitioners and training, and assigning clear roles to play with associ- and policy makers to more closely assess different play- ated incentives based on the existing systems, is critical to ers and modalities operating in Nepal, and to consolidate build trust with communities. and/or partner with these programs by providing roles, incentives, and training that are aligned with the country’s Focus area specific lessons. The data revealed both governance system. Addressing malnutrition in a sustain- opportunities and challenges faced by RRNI teams that able manner needs the entire community—men, women, were specific to the focus area they selected. There are minority groups, and local leaders and authorities—to important lessons to learn from the examples for future raise awareness on the importance and concrete ways of implementation efforts, particularly to improve intended improving knowledge and behavior, and on meaningfully nutritional behaviors or outcomes by involving com- including women in decision-making and implementation munities. A point that was confirmed across the sample processes. wards was that when the tangible outputs were com- bined with related training and proper awareness raising As the SHD closed in June 2017, it is critical for the on their behaviors (e.g., demonstrations on how to use a SHD team and other stakeholders in the government to constructed latrine, or to wash hands using the tools pro- discuss the findings and lessons with practitioners and vided by the project), people were motivated and able to policy makers of relevant programs, so that follow-up start new practices, and the changes were considered as activities or improved design for those programs can be sustainable. As a few good examples indicate, it is worth considered. There are a dozen government and donor- developing more entertaining ways to raise awareness by funded programs to address malnutrition in Nepal,16 involving community members, combined with efforts to as well as programs on related key sectors such as rural prepare a set of experts’ guidance and good practices on water and sanitation (including ODF), and food secu- key focus areas that are ready for sharing, especially with rity and entrepreneurship support for women.17 Many an emphasis on behavioral changes. of these programs have been using community-based Concluding remarks 16 See a complementary study “Operational Strategy for addressing malnutri- The key findings of the in-depth qualitative analysis tion in Nepal” that reviews the political economy, institutional structures, and implementation arrangements currently in place in Nepal to address malnu- offered important implications for the community-based trition, including the programs supported by the WB, USAID, UNICEF, etc. rapid results approach to address malnutrition. Despite (forthcoming), World Bank. its design that had linkages with the local governance 17 For example, there are pilot activities such as “Women’s Enterprising Initia- structure, such as the fund flow and proposal review and tives to Ensure Community Food and Nutrition Security in Upland Nuwakot” that can provide insights. 40 Qualitative Study of “Sunaula Hazar Din” Community Action for Nutrition Project Nepal approaches,18 and therefore, there are common risks fiduciary mechanisms and flexible timeframes, a point to and opportunities to learn from each other, including on note about its design is that there was an effort to have gender, social inclusion and mobilization, capacity build- the demand-driven steps be embedded in the local gover­ ing, and nutrition-specific aspects. It is also important nance structures, such as the use of WCF, and local gov- to review existing community-based platforms beyond ernments’ involvement in proposal approval and funds nutrition-focused programs, and to identify synergies and ­ disbursement. Considering the country’s shift toward entry points for nutrition activities that can be integrated federalism and more authority to be given to the newly and leveraged by local-level institutions that are consisting formed local bodies (at the level of a cluster of VDCs), les- of, or at least close to, the target population. sons from SHD may provide some food for thought to any new nutrition programs designed either by government Although SHD faced multiple implementation chal- or development partners with a community-led approach. lenges, due mainly to lack of adequate capacity building, 18 To name a few, a program to provide special grants to VDCs to establish a female community health volunteer fund; or Community-Led Total Sanitation program on ODF and other sanitation activities. Emerging Lessons and Conclusion 41 References Asian Development Bank (2010). “Overview of Gender Equality and Social Inclusion in Nepal.” NPC/CBS. 2013. Nepal Thematic Report on Food Security and Nutrition. Supported by: WFP, WB, Ausaid, and UNICEF. Kathmandu, Nepal. Sylvester, Obong’o (2012) “Rapid Results Approach/Iniative. Institutionalization of Results Based Management in Kenya Public Service” Mimeo. Commwealth Association for Public Administration and Management. http://www.capam .org/_documents/rapidresultsapproach.kenya.pdf World Bank (2015a) “Restructuring Paper on a Proposed Project Restructuring of the Community Action for Nutrition Project (Sunaula Hazar Din).” June 17, 2015. World Bank (2015b) “Sunaula Hazar Din—Mid-term Review” March 31, 2015. World Bank Group (forthcoming). “Nepal Sunaula Hazar Din Community Action for Nutrition Project: Impact Evaluation Midline Report.” World Health Organization 2016. Strategic Action Plan to reduce the double burden of malnutrition in the South-East Asia Region: 2016–2025. References 43 Annex 1 Theory of Change of the SHD Project The theory of change of the SHD project is illustrated in the diagram below. Improved long-term nutritional status (stunting) Improved Knowledge, Attitudes and Practices (KAP) for improving nutritional outcomes among reproductive age and 1,000 days women Demand-side factors Supply-side factors Individuals (women, men, children) Conditional Grants Household (family, household dynamics, assets) Implementers Community/society (cohesion, culture, inequality) Implementing institutions Political-economy (political, economic, social) Public/private health system Other public programs & services (Ex: needs, assets, level of awareness, women’s Markets status, inequality of power and influence) Infrastructure (Ex: financial resources, quality of health facilities and services, access to markets, capability and effectiveness of implementing agencies) Proper diet of 1,000 days Clean, sanitary and hygienic Use of preventive & Age at pregnancy, number women and children environment curative health services and spacing Dietary intake of protein Reduce indoor smoke & Vaccination of children Delay marriage and Proper breast-feeding air pollution Use of contraceptives pregnancy of Use safe & clean water adolescence girls Maintain adequate weight Intake of iron-folic acid & End open defecation deworming medicine Increase schooling of Reduce workload of adolescence girls pregnant women Hand-washing practices Timely seeking of Sanitation of schools health services 100-day projects POOR NUTRITIONAL STATUS IN MOST DISADVANTAGED VDCS OF 15 DISTRICTS Key causal risk factors Insufficient intake and absorption of nutrients and excessive energy expenditure Community-wide factors (poor sanitation and hygiene practices) Low demand for preventive and curative health services (vaccination, IFA, deworming) Early age of pregnancy among reproductive age women Theory of Change of the SHD Project 45 H e a l t h , N u t r i t i o n , a n d P o pu l a t i o n G l o b a l P r a c t i c e a n d S o c i a l , U r b a n , Ru r a l , a n d Res i l i en c e G l o b a l P r a c t i c e 1818 H Street, NW Washington, D.C. 20433 USA Telephone: 202-473-1000